Hypoglycaemia in Clinical Diabetes

Hypoglycaemia in Clinical Diabetes

Hypoglycaemia in Clinical Diabetes

  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>Second Edition

Other titles <strong>in</strong> the Wiley <strong>Diabetes</strong> <strong>in</strong> Practice SeriesExercise and Sport <strong>in</strong> <strong>Diabetes</strong> Second EditionEdited by D<strong>in</strong>esh Nagi047002206XComplementary Therapies and the Management of <strong>Diabetes</strong> and Vascular DiseaseEdited by Patricia Dunn<strong>in</strong>g047001458X<strong>Diabetes</strong> <strong>in</strong> Cl<strong>in</strong>ical Practice: Questions and Answers from Case StudiesN. Katsilambros, E. Diakoumopoulou, I. Ioannidis, S. Liatis, K. Makrilakis, N. Tentolouris and P. Tsapogas978 0470 035221Obesity and <strong>Diabetes</strong>Edited by Anthony Barnett and Sudhesh Kumar0470848987Prevention of Type 2 <strong>Diabetes</strong>Edited by Manfred Ganz0470857331<strong>Diabetes</strong> – Chronic Complications Second EditionEdited by Kenneth Shaw and Michael Cumm<strong>in</strong>gs0470865972The Metabolic SyndromeEdited by Christopher Byrne and Sarah Wild0470025115Psychology <strong>in</strong> <strong>Diabetes</strong> Care Second EditionEdited by Frank J. Snoek and T. Chas Sk<strong>in</strong>ner0470023848The Foot <strong>in</strong> <strong>Diabetes</strong> Fourth EditionEdited by Andrew J.M. Boulton, Peter R. Cavanagh and Gerry Rayman0470015047Gastro<strong>in</strong>test<strong>in</strong>al Function <strong>in</strong> <strong>Diabetes</strong> MellitusEdited by Michael Horowitz and Melv<strong>in</strong> Samson047189916XDiabetic NephropathyEdited by Christoph Hasslacher0471489921Nutritional Management of <strong>Diabetes</strong> MellitusEdited by Gary Frost, Anne Dornhorst and Robert Moses0471497517<strong>Diabetes</strong> <strong>in</strong> Pregnancy: An International Approach to Diagnosis and ManagementEdited by Anne Dornhorst and David R. Hadden047196204X

<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>Second EditionEdited byBrian M. FrierThe Royal Infirmary of Ed<strong>in</strong>burgh, Scotland, UKMiles FisherGlasgow Royal Infirmary, Scotland, UK

Copyright © 2007John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester,West Sussex PO19 8SQ, EnglandTelephone (+44) 1243 779777Email (for orders and customer service enquiries): cs-books@wiley.co.ukVisit our Home Page on www.wiley.comAll Rights Reserved. No part of this publication may be reproduced, stored <strong>in</strong> a retrieval system or transmitted <strong>in</strong>any form or by any means, electronic, mechanical, photocopy<strong>in</strong>g, record<strong>in</strong>g, scann<strong>in</strong>g or otherwise, except underthe terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the CopyrightLicens<strong>in</strong>g Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission <strong>in</strong> writ<strong>in</strong>g ofthe Publisher. Requests to the Publisher should be addressed to the Permissions Department, John Wiley & SonsLtd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, or emailed topermreq@wiley.co.uk, or faxed to (+44) 1243 770620.Designations used by companies to dist<strong>in</strong>guish their products are often claimed as trademarks. All brand namesand product names used <strong>in</strong> this book are trade names, service marks, trademarks or registered trademarks of theirrespective owners. The Publisher is not associated with any product or vendor mentioned <strong>in</strong> this book.This publication is designed to provide accurate and authoritative <strong>in</strong>formation <strong>in</strong> regard to the subject mattercovered. It is sold on the understand<strong>in</strong>g that the Publisher is not engaged <strong>in</strong> render<strong>in</strong>g professional services. Ifprofessional advice or other expert assistance is required, the services of a competent professional should besought.Other Wiley Editorial OfficesJohn Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USAJossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USAWiley-VCH Verlag GmbH, Boschstr. 12, D-69469 We<strong>in</strong>heim, GermanyJohn Wiley & Sons Australia Ltd, 42 McDougall Street, Milton, Queensland 4064, AustraliaJohn Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, J<strong>in</strong> X<strong>in</strong>g Distripark, S<strong>in</strong>gapore 129809John Wiley & Sons Canada Ltd, 6045 Freemont Blvd, Mississauga, Ontario, L5R 4J3Wiley also publishes its books <strong>in</strong> a variety of electronic formats. Some content that appears <strong>in</strong> pr<strong>in</strong>t may not beavailable <strong>in</strong> electronic books.Anniversary Logo Design: Richard J. PacificoLibrary of Congress Catalog<strong>in</strong>g <strong>in</strong> Publication Data<strong>Hypoglycaemia</strong> <strong>in</strong> cl<strong>in</strong>ical diabetes / edited by Brian M. Frier and Miles Fisher. — 2nd ed.p. ; cm.Includes bibliographical references and <strong>in</strong>dex.ISBN 978-0-470-01844-6 (cloth : alk. paper)1. Hypoglycemia. 2. <strong>Diabetes</strong>—Treatment—Complications. 3. Hypoglycemic agents—Sideeffects. I. Frier, Brian M. II. Fisher, Miles. III. Title: Hypoglycemia <strong>in</strong> cl<strong>in</strong>ical diabetes.[DNLM: 1. Hypoglycemia—complications. 2. Hypoglycemia—physiopathology. 3. <strong>Diabetes</strong>Complications. 4. Insul<strong>in</strong>—adverse effects. WK 880 H9963 2007]RC662.2.H965 2007616.4 ′ 66—dc222007012095British Library Catalogu<strong>in</strong>g <strong>in</strong> Publication DataA catalogue record for this book is available from the British LibraryISBN 978-0-470-01844-6Typeset <strong>in</strong> 10/12pt Times by Integra Software Services Pvt. Ltd, Pondicherry, IndiaPr<strong>in</strong>ted and bound <strong>in</strong> Great Brita<strong>in</strong> by Antony Rowe Ltd, Chippenham, WiltshireThis book is pr<strong>in</strong>ted on acid-free paper responsibly manufactured from susta<strong>in</strong>able forestry <strong>in</strong> which at least twotrees are planted for each one used for paper production.

ToEmily, Ben and Marc

ContentsPrefaceContributorsixxi1 Normal Glucose Metabolism and Responses to <strong>Hypoglycaemia</strong> 1Ian A. Macdonald and Paromita K<strong>in</strong>g2 Symptoms of <strong>Hypoglycaemia</strong> and Effects on Mental Performanceand Emotions 25Ian J. Deary3 Frequency, Causes and Risk Factors for <strong>Hypoglycaemia</strong> <strong>in</strong> Type 1<strong>Diabetes</strong> 49Mark W.J. Strachan4 Nocturnal <strong>Hypoglycaemia</strong> 83Simon R. Heller5 Moderators, Monitor<strong>in</strong>g and Management of <strong>Hypoglycaemia</strong> 101Tristan Richardson and David Kerr6 Counterregulatory Deficiencies <strong>in</strong> <strong>Diabetes</strong> 121David Kerr and Tristan Richardson7 Impaired Awareness of <strong>Hypoglycaemia</strong> 141Brian M. Frier8 Risks of Strict Glycaemic Control 171Stephanie A. Amiel9 <strong>Hypoglycaemia</strong> <strong>in</strong> Children with <strong>Diabetes</strong> 191Krystyna A. Matyka10 <strong>Hypoglycaemia</strong> <strong>in</strong> Pregnancy 217Ann E. Gold and Donald W.M. Pearson11 <strong>Hypoglycaemia</strong> <strong>in</strong> Type 2 <strong>Diabetes</strong> and <strong>in</strong> Elderly People 239Nicola N. Zammitt and Brian M. Frier12 Mortality, Cardiovascular Morbidity and Possible Effects of<strong>Hypoglycaemia</strong> on Diabetic Complications 265Miles Fisher and Simon R. Heller

viiiCONTENTS13 Long-term Effects of <strong>Hypoglycaemia</strong> on Cognitive Function and theBra<strong>in</strong> <strong>in</strong> <strong>Diabetes</strong> 285Petros Perros and Ian J. Deary14 Liv<strong>in</strong>g with <strong>Hypoglycaemia</strong> 309Brian M. FrierIndex 333

PrefaceIn the second edition of this book, we have cont<strong>in</strong>ued to emphasise the cl<strong>in</strong>ical significanceof hypoglycaemia to the person who has diabetes, particularly when receiv<strong>in</strong>g treatmentwith <strong>in</strong>sul<strong>in</strong>. S<strong>in</strong>ce the first edition of the book was published <strong>in</strong> 1999, new therapies haveemerged, <strong>in</strong>clud<strong>in</strong>g new <strong>in</strong>sul<strong>in</strong> analogues and <strong>in</strong>haled <strong>in</strong>sul<strong>in</strong>, and monitor<strong>in</strong>g systems arenow available that can provide cont<strong>in</strong>uous record<strong>in</strong>g of blood glucose. However, far fromm<strong>in</strong>imis<strong>in</strong>g the risk of hypoglycaemia <strong>in</strong> cl<strong>in</strong>ical practice, the newer treatments have beenshown to be as liable to cause hypoglycaemia as before, while cont<strong>in</strong>uous blood glucosemonitor<strong>in</strong>g has revealed that this side-effect of <strong>in</strong>sul<strong>in</strong> therapy is even more common thanwas believed previously. The frequency of severe hypoglycaemia <strong>in</strong> vulnerable groups suchas children and elderly people receiv<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> therapy is unacceptably high, and presentspotentially serious risks to health as well as dim<strong>in</strong>ish<strong>in</strong>g their quality of life. Much scientificresearch <strong>in</strong> recent years has focused on the effects of hypoglycaemia on the bra<strong>in</strong>, provid<strong>in</strong>ga greater understand<strong>in</strong>g of the protean effects of this metabolic abnormality. New data andconcepts have been <strong>in</strong>corporated <strong>in</strong> this edition, particularly where these are of importanceto cl<strong>in</strong>ical practice.In updat<strong>in</strong>g and revis<strong>in</strong>g this book about hypoglycaemia, particular emphasis has beengiven to the risk factors for hypoglycaemia and how these may be reduced or avoided. Newchapters have been <strong>in</strong>cluded to discuss recognised moderators of hypoglycaemia and the roleof new glucose monitor<strong>in</strong>g systems, to address the <strong>in</strong>creas<strong>in</strong>g problem of hypoglycaemia <strong>in</strong>people with type 2 diabetes and the elderly person, and to acknowledge the major importanceof nocturnal hypoglycaemia, which is frequently not identified <strong>in</strong> cl<strong>in</strong>ical practice but canhave serious consequences, not only <strong>in</strong> its immediate morbidity, but also <strong>in</strong> promot<strong>in</strong>g thedevelopment of the acquired syndromes of hypoglycaemia.We are grateful for the expert assistance and support of the colleagues who havecontributed chapters, some of whom are new as authors for this edition. All have skillfullyhighlighted the relevance of the enhancement of scientific knowledge <strong>in</strong> this field to theeveryday management of diabetes, which we hope will assist all members of the diabetesteam <strong>in</strong> their efforts to prevent and manage the extremely common but unwanted scourgethat is hypoglycaemia.Brian M. FrierMiles Fisher

ContributorsProfessor Stephanie A. Amiel, R.D. Lawrence Professor of <strong>Diabetes</strong>, Departmentof Medic<strong>in</strong>e, K<strong>in</strong>g’s College Hospital, Bessemer Road, London, SE5 9PJ (e-mail:stephanie.amiel@kcl.ac.uk)Professor Ian J. Deary, Department of Psychology, University of Ed<strong>in</strong>burgh, 7 GeorgeSquare, Ed<strong>in</strong>burgh, EH8 9JZ (e-mail: ian.deary@ed.ac.uk)Dr Miles Fisher, Consultant Physician, Glasgow Royal Infirmary, Glasgow, G4 0SF (e-mail:miles.fisher@northglasgow.scot.nhs.uk)Professor Brian M. Frier, Consultant Physician, Department of <strong>Diabetes</strong>, Royal Infirmary,51 Little France Crescent, Ed<strong>in</strong>burgh, EH16 4SA (e-mail: brian.frier@luht.scot.nhs.uk)Dr Ann E. Gold, Consultant Physician, Wards 27 and 28, Aberdeen Royal Infirmary,Foresterhill, Aberdeen, AB25 2ZN (e-mail: ann.gold@nhs.net)Professor Simon R. Heller, Professor of Cl<strong>in</strong>ical <strong>Diabetes</strong>, Cl<strong>in</strong>ical Sciences Centre, Departmentof <strong>Diabetes</strong> & Endocr<strong>in</strong>ology, Northern General Hospital, Herries Road, Sheffield, S57AU (e-mail: s.heller@sheffield.ac.uk)Dr David Kerr, Consultant Physician, The Royal Bournemouth Hospital, Castle Lane East,Bournemouth, BH7 7DW (e-mail: David.Kerr@rbch.nhs.uk)Dr Paromita K<strong>in</strong>g, Consultant Physician, Jenny O’Neill <strong>Diabetes</strong> Centre, Derbyshire RoyalInfirmary, London Road, Derby, DE1 2QY (e-mail: Paru.K<strong>in</strong>g@derbyhospitals.nhs.uk)Professor Ian A. Macdonald, Professor of Metabolic Physiology, Department of Physiologyand Pharmacology, Medical School, Queen’s Medical Centre, Nott<strong>in</strong>gham, NG7 2UH(e-mail: Ian.Macdonald@nott<strong>in</strong>gham.ac.uk)Dr Krystyna A. Matyka, Senior Lecturer <strong>in</strong> Paediatrics, University of Warwick MedicalSchool, Division of Cl<strong>in</strong>ical Sciences, CSB Research W<strong>in</strong>g, UHCW Trust, Clifford BridgeRoad, Coventry, CV2 2DX (e-mail: K.A.Matyka@warwick.ac.uk)Dr Donald W.M. Pearson, Consultant Physician, Aberdeen Royal Infirmary, Foresterhill,Aberdeen, AB25 2ZN (e-mail: Dwm.Pearson@arh.grampian.scot.nhs.uk)Dr Petros Perros, Consultant Endocr<strong>in</strong>ologist, Ward 15, Freeman Hospital, Freeman Road,Newcastle-Upon-Tyne, NE7 7DN (e-mail: Petros.Perros@ncl.ac.uk)Dr Tristan Richardson, Consultant Physician, Bournemouth <strong>Diabetes</strong> and Endocr<strong>in</strong>eCentre, The Royal Bournemouth Hospital, Castle Lane East, Bournemouth, BH7 7DW(e-mail: Tristan.Richardson@rbch.nhs.uk)

xiiCONTRIBUTORSDr Mark W.J. Strachan, Consultant Physician, Metabolic Unit, Western General Hospital,Crewe Road, Ed<strong>in</strong>burgh, EH4 2XU (e-mail: mark.strachan@luht.scot.nhs.uk)Dr Nicola N. Zammitt, Specialist Registrar, Department of <strong>Diabetes</strong>, Royal Infirmaryof Ed<strong>in</strong>burgh, 51 Little France Crescent, Ed<strong>in</strong>burgh, EH16 4SA (e-mail: nicolazammitt@hotmail.com)

1 Normal Glucose Metabolismand Responses to<strong>Hypoglycaemia</strong>Ian A. Macdonald and Paromita K<strong>in</strong>gINTRODUCTIONControl of blood glucose is a fundamental feature of homeostasis, i.e., the process by whichthe <strong>in</strong>ternal environment of the body is ma<strong>in</strong>ta<strong>in</strong>ed stable allow<strong>in</strong>g optimal function. Bloodglucose concentrations are regulated with<strong>in</strong> a narrow range (which <strong>in</strong> humans is known asnormoglycaemia or euglycaemia) despite wide variability <strong>in</strong> carbohydrate <strong>in</strong>take and physicalactivity. Teleologically, the upper limit is defended because high glucose concentrationscause microvascular complications, and the lower limit, because the bra<strong>in</strong> cannot functionwithout an adequate supply of glucose. In this chapter the mechanisms that protect aga<strong>in</strong>sthypoglycaemia <strong>in</strong> healthy <strong>in</strong>dividuals and the physiological consequences of low glucoseconcentrations are discussed.NORMAL GLUCOSE HOMEOSTASISHumans evolved as hunter-gatherers and, unlike people today, did not consume regularmeals. Mechanisms therefore evolved for the body to store food when it was <strong>in</strong> abundance,and to use these stores to provide an adequate supply of energy, <strong>in</strong> particular <strong>in</strong> the form ofglucose when food was scarce. Cahill (1971) orig<strong>in</strong>ally described the ‘rules of the metabolicgame’ which humans had to follow to ensure their survival. These rules were modified byTattersall (personal communication) and are as follows:1. Ma<strong>in</strong>ta<strong>in</strong> glucose with<strong>in</strong> very narrow limits.2. Ma<strong>in</strong>ta<strong>in</strong> an emergency energy source (glycogen) which can be tapped quickly for flee<strong>in</strong>gor fight<strong>in</strong>g.3. Waste not want not, i.e., store (fat and prote<strong>in</strong>) <strong>in</strong> times of plenty.4. Use every trick <strong>in</strong> the book to ma<strong>in</strong>ta<strong>in</strong> prote<strong>in</strong> reserves.<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

2 NORMAL GLUCOSE METABOLISM AND RESPONSESInsul<strong>in</strong> and glucagon are the two hormones controll<strong>in</strong>g glucose homeostasis, and thereforethe mechanisms enabl<strong>in</strong>g the ‘rules’ to be followed. The most important processes governedby these hormones are:• Glycogen synthesis and breakdown (glycogenolysis): Glycogen, a carbohydrate, is anenergy source stored <strong>in</strong> the liver and skeletal muscle. Liver glycogen is broken downto provide glucose for all tissues, whereas the breakdown of muscle glycogen results <strong>in</strong>lactate formation.• Gluconeogenesis: This is the production of glucose <strong>in</strong> the liver from precursors: glycerol,lactate and am<strong>in</strong>o acids (<strong>in</strong> particular alan<strong>in</strong>e). The process can also occur <strong>in</strong> the kidneys,but this site is not important under physiological conditions.• Glucose uptake and metabolism (glycolysis) by skeletal muscle and adipose tissue.The actions of <strong>in</strong>sul<strong>in</strong> and glucagon are summarised <strong>in</strong> Boxes 1.1 and 1.2. Insul<strong>in</strong> is ananabolic hormone, reduc<strong>in</strong>g glucose output by the liver (hepatic glucose output), <strong>in</strong>creas<strong>in</strong>gthe uptake of glucose by muscle and adipose tissue (<strong>in</strong>creas<strong>in</strong>g peripheral uptake) and<strong>in</strong>creas<strong>in</strong>g prote<strong>in</strong> and fat formation. Glucagon opposes the actions of <strong>in</strong>sul<strong>in</strong> <strong>in</strong> the liver.Thus <strong>in</strong>sul<strong>in</strong> tends to reduce, and glucagon to <strong>in</strong>crease, blood glucose concentrations.Box 1.1Actions of <strong>in</strong>sul<strong>in</strong>Liver↑ Glycogen synthesis (↑ glycogen synthetase activity)↑ Glycolysis↑ Lipid formation↑ Prote<strong>in</strong> formation↓ Glycogenolysis (↓ phosphorylase activity)↓ Gluconeogenesis↓ Ketone formationMuscle↑ Uptake of glucoseam<strong>in</strong>o acidsketonepotassium↑ Glycolysis↑ Synthesis of glycogenprote<strong>in</strong>↓ Prote<strong>in</strong> catabolism↓ Release of am<strong>in</strong>o acidsAdipose tissue↑ Uptake glucosepotassiumStorage of triglyceride

NORMAL GLUCOSE HOMEOSTASIS 3Box 1.2Actions of glucagonLiver↑ Glycogenolysis↑ Gluconeogenesis↑ Extraction of alan<strong>in</strong>e↑ KetogenesisNo significant peripheral actionThe metabolic effects of <strong>in</strong>sul<strong>in</strong> and glucagon and their relationship to glucose homeostasisare best considered <strong>in</strong> relationship to fast<strong>in</strong>g and the postprandial state (Siegal and Kreisberg,1975). In both these situations it is the relative and not absolute concentrations of thesehormones that are important.Fast<strong>in</strong>g (Figure 1.1a)Dur<strong>in</strong>g fast<strong>in</strong>g, <strong>in</strong>sul<strong>in</strong> concentrations are reduced and glucagon <strong>in</strong>creased, which ma<strong>in</strong>ta<strong>in</strong>sblood glucose concentrations <strong>in</strong> accordance with rule 1 above. The net effect is toreduce peripheral glucose utilisation, to <strong>in</strong>crease hepatic glucose production and to providenon-glucose fuels for tissues not entirely dependent on glucose. After a short (for exampleovernight) fast, glucose production needs to be 5–6 g/h to ma<strong>in</strong>ta<strong>in</strong> blood glucose concentrations,with the bra<strong>in</strong> us<strong>in</strong>g 80% of this. Glycogenolysis provides 60–80% and gluconeogenesis20–40% of the required glucose. In prolonged fasts, glycogen becomes depletedand glucose production is primarily from gluconeogenesis, with an <strong>in</strong>creas<strong>in</strong>g proportionfrom the kidney compared to the liver. In extreme situations renal gluconeogenesis cancontribute as much as 45% of glucose production. Thus glycogen is the short term or ‘emergency’fuel source (rule 2), with gluconeogenesis predom<strong>in</strong>at<strong>in</strong>g dur<strong>in</strong>g more prolongedfasts. The follow<strong>in</strong>g metabolic alterations enable this <strong>in</strong>crease <strong>in</strong> glucose productionto occur:• Muscle: Glucose uptake and oxidative metabolism are reduced and fatty acid oxidation<strong>in</strong>creased. Am<strong>in</strong>o acids are released.• Adipose tissue: There are reductions <strong>in</strong> glucose uptake and triglyceride storage. The<strong>in</strong>crease <strong>in</strong> the activity of the enzyme hormone-sensitive lipase results <strong>in</strong> hydrolysisof triglyceride to glycerol (a gluconeogenic precursor) and fatty acids, which can bemetabolised.• Liver: Increased cAMP concentrations result <strong>in</strong> <strong>in</strong>creased glycogenolysis and gluconeogenesisthus <strong>in</strong>creas<strong>in</strong>g hepatic glucose output. The uptake of gluconeogenic precursors(i.e. am<strong>in</strong>o acids, glycerol, lactate and pyruvate) is also <strong>in</strong>creased. Ketone bodies areproduced <strong>in</strong> the liver from fatty acids. This process is normally <strong>in</strong>hibited by <strong>in</strong>sul<strong>in</strong> andstimulated by glucagon, thus the hormonal changes dur<strong>in</strong>g fast<strong>in</strong>g lead to an <strong>in</strong>crease <strong>in</strong>ketone production. Fatty acids are also a metabolic fuel used by the liver and provide asource of energy for the reactions <strong>in</strong>volved <strong>in</strong> gluconeogenesis.

4 NORMAL GLUCOSE METABOLISM AND RESPONSESglucagon, <strong>in</strong>sul<strong>in</strong> Insul<strong>in</strong>, glucagon(a)FASTING(b)POST PRANDIALlactateGlycogenGlucoseGlycogenAm<strong>in</strong>o acidsProte<strong>in</strong>MuscleAm<strong>in</strong>o acidsProte<strong>in</strong>TGGlucoseFAGlycerolphosphateFAGlycerolphosphateFree FAGlycerolFree FATriglycerideTriglycerideAdipose TissueAm<strong>in</strong>o acidslactateGlycogenKetonesKetonesGlycogenGlucoseFAGlycerolFFA, TG,lipoprote<strong>in</strong>GlucoseFAGlycerolHepatic glucoseoutputLiverFFA, TG,Hepatic glucoseoutputFigure 1.1 Metabolic pathways for glucose homeostasis <strong>in</strong> muscle, adipose tissue and liver dur<strong>in</strong>gfast<strong>in</strong>g (left) and postprandially (right). FA = fatty acids; TG = triglyceride (associated CO 2 productionexcluded for clarity)The reduced <strong>in</strong>sul<strong>in</strong> : glucagon ratio favours a catabolic state, but the effect on fatmetabolism is greater than prote<strong>in</strong>, and thus muscle is relatively preserved (rule 4). Theseadaptations meant that not only did hunter-gatherers have sufficient muscle power topursue their next meal, but also that bra<strong>in</strong> function was optimally ma<strong>in</strong>ta<strong>in</strong>ed to help themdo this.

Fed state (Figure 1.1b)EFFECTS OF GLUCOSE DEPRIVATION 5In the fed state, <strong>in</strong> accordance with the rules of the metabolic game, excess food is stored asglycogen, prote<strong>in</strong> and fat (rule 3). The rise <strong>in</strong> glucose concentrations results <strong>in</strong> an <strong>in</strong>crease<strong>in</strong> <strong>in</strong>sul<strong>in</strong> and reduction <strong>in</strong> glucagon secretion. This balance favours glucose utilisation,reduction of glucose production and <strong>in</strong>creases glycogen, triglyceride and prote<strong>in</strong> formation.The follow<strong>in</strong>g changes enable these processes to occur:• Muscle: Insul<strong>in</strong> <strong>in</strong>creases glucose transport, oxidative metabolism and glycogen synthesis.Am<strong>in</strong>o acid release is <strong>in</strong>hibited and prote<strong>in</strong> synthesis is <strong>in</strong>creased.• Adipose tissue: In the fat cells, glucose transport is <strong>in</strong>creased, while lipolysis is <strong>in</strong>hibited.At the same time the enzyme lipoprote<strong>in</strong> lipase, located <strong>in</strong> the capillaries, is activatedand causes triglyceride to be broken down to fatty acids and glycerol. The fatty acidsare taken up <strong>in</strong>to the fat cells and re-esterified to triglyceride (us<strong>in</strong>g glycerol phosphatederived from glucose) before be<strong>in</strong>g stored.• Liver: Glucose uptake is <strong>in</strong>creased <strong>in</strong> proportion to plasma glucose, a process which doesnot need <strong>in</strong>sul<strong>in</strong>. However, <strong>in</strong>sul<strong>in</strong> does decrease cAMP concentrations, which results <strong>in</strong> an<strong>in</strong>crease <strong>in</strong> glycogen synthesis and the <strong>in</strong>hibition of glycogenolysis and gluconeogenesis.These effects ‘reta<strong>in</strong>’ glucose <strong>in</strong> the liver and reduce hepatic glucose output.This complex <strong>in</strong>terplay between <strong>in</strong>sul<strong>in</strong> and glucagon ma<strong>in</strong>ta<strong>in</strong>s euglycaemia and enablesthe rules of the metabolic game to be followed, ensur<strong>in</strong>g not only the survival of thehunter-gatherer, but also of modern humans.EFFECTS OF GLUCOSE DEPRIVATION ON CENTRAL NERVOUSSYSTEM METABOLISMThe bra<strong>in</strong> constitutes only 2% of body weight, but consumes 20% of the body’s oxygenand receives 15% of its cardiac output (Sokaloff, 1989). It is almost totally dependenton carbohydrate as a fuel and s<strong>in</strong>ce it cannot store or synthesise glucose, depends ona cont<strong>in</strong>uous supply from circulat<strong>in</strong>g blood. The bra<strong>in</strong> conta<strong>in</strong>s the enzymes needed tometabolise fuels other than glucose such as lactate, ketones and am<strong>in</strong>o acids, but underphysiological conditions their use is limited by <strong>in</strong>sufficient quantities <strong>in</strong> the blood or slowrates of transport across the blood-bra<strong>in</strong> barrier. When arterial blood glucose falls below3 mmol/l, cerebral metabolism and function decl<strong>in</strong>e.Metabolism of glucose by the bra<strong>in</strong> releases energy, and also generates neurotransmitterssuch as gamma am<strong>in</strong>o butyric acid (GABA) and acetylchol<strong>in</strong>e, together with phospholipidsneeded for cell membrane synthesis. When blood glucose concentration falls, changes <strong>in</strong>the synthesis of these products may occur with<strong>in</strong> m<strong>in</strong>utes because of reduced glucosemetabolism, which can alter cerebral function. This is likely to be a factor <strong>in</strong> produc<strong>in</strong>gthe subtle changes <strong>in</strong> cerebral function detectable at blood glucose concentrations as highas 3 mmol/l, which is not sufficiently low to cause a major depletion <strong>in</strong> ATP or creat<strong>in</strong>ephosphate, the bra<strong>in</strong>’s two ma<strong>in</strong> sources of energy (McCall, 1993).

6 NORMAL GLUCOSE METABOLISM AND RESPONSESIsotope techniques and Positron Emission Tomography (PET) allow the study ofmetabolism <strong>in</strong> different parts of the bra<strong>in</strong> and show regional variations <strong>in</strong> metabolismdur<strong>in</strong>g hypoglycaemia. The neocortex, hippocampus, hypothalamus and cerebellum are mostsensitive to hypoglycaemia, whereas metabolism is relatively preserved <strong>in</strong> the thalamusand bra<strong>in</strong>stem. Changes <strong>in</strong> cerebral function are <strong>in</strong>itially reversible, but dur<strong>in</strong>g prolongedsevere hypoglycaemia, general energy failure (due to the depletion of ATP and creat<strong>in</strong>ephosphate) can cause permanent cerebral damage. Pathologically this is caused by selectiveneuronal necrosis most likely due to ‘excitotox<strong>in</strong>’ damage. Local energy failure <strong>in</strong>duces the<strong>in</strong>trasynaptic release of glutamate or aspartate, and failure of reuptake of the neurotransmitters<strong>in</strong>creases their concentrations. This leads to the activation of N-methyl-D-aspartate(NMDA) receptors caus<strong>in</strong>g cerebral damage. One study <strong>in</strong> rats has shown that an experimentalcompound called AP7, which blocks the NMDA receptor, can prevent 90% of thecerebral damage associated with severe hypoglycaemia (Wieloch, 1985). In humans withfatal hypoglycaemia, protracted neuroglycopenia causes lam<strong>in</strong>ar necrosis <strong>in</strong> the cerebralcortex and diffuse demyel<strong>in</strong>ation. Regional differences <strong>in</strong> neuronal necrosis are seen, withthe basal ganglia and hippocampus be<strong>in</strong>g sensitive, but the hypothalamus and cerebellumbe<strong>in</strong>g relatively spared (Auer and Siesjö, 1988; Sieber and Traysman, 1992).The bra<strong>in</strong> is very sensitive to acute hypoglycaemia, but can adapt to chronic fuel deprivation.For example, dur<strong>in</strong>g starvation, it can metabolise ketones for up to 60% of itsenergy requirements (Owen et al., 1967). Glucose transport can also be <strong>in</strong>creased <strong>in</strong> theface of hypoglycaemia. Normally, glucose is transported <strong>in</strong>to tissues us<strong>in</strong>g prote<strong>in</strong>s calledglucose transporters (GLUT) (Bell et al., 1990). This transport occurs down a concentrationgradient faster than it would by simple diffusion and does not require energy (facilitateddiffusion). There are several of these transporters, with GLUT 1 be<strong>in</strong>g responsible fortransport<strong>in</strong>g glucose across the blood-bra<strong>in</strong> barrier and GLUT 3 for transport<strong>in</strong>g glucose<strong>in</strong>to neurones (Figure 1.2). Chronic hypoglycaemia <strong>in</strong> animals (McCall et al., 1986) and<strong>in</strong> humans (Boyle et al., 1995) <strong>in</strong>creases cerebral glucose uptake, which is thought to bepromoted by an <strong>in</strong>crease <strong>in</strong> the production and action of GLUT 1 prote<strong>in</strong>. It has not beenFigure 1.2Transport of glucose <strong>in</strong>to the bra<strong>in</strong> across the blood–bra<strong>in</strong> barrier

COUNTERREGULATION DURING HYPOGLYCAEMIA 7established whether this adaptation is of major benefit <strong>in</strong> protect<strong>in</strong>g bra<strong>in</strong> function dur<strong>in</strong>ghypoglycaemia.COUNTERREGULATION DURING HYPOGLYCAEMIAThe potentially serious effects of hypoglycaemia on cerebral function mean that not onlyare stable blood glucose concentrations ma<strong>in</strong>ta<strong>in</strong>ed under physiological conditions, but alsoif hypoglycaemia occurs, mechanisms have developed to combat it. In cl<strong>in</strong>ical practice, thepr<strong>in</strong>cipal causes of hypoglycaemia are iatrogenic (as side-effects of <strong>in</strong>sul<strong>in</strong> and sulphonylureasused to treat diabetes) and excessive alcohol consumption. Insul<strong>in</strong> secret<strong>in</strong>g tumours(such as <strong>in</strong>sul<strong>in</strong>oma) are rare. The mechanisms that correct hypoglycaemia are called counterregulation,because the hormones <strong>in</strong>volved oppose the action of <strong>in</strong>sul<strong>in</strong> and therefore arethe counterregulatory hormones. The processes of counterregulation were identified <strong>in</strong> themid 1970s and early 1980s, us<strong>in</strong>g either a bolus <strong>in</strong>jection or cont<strong>in</strong>uous <strong>in</strong>fusion of <strong>in</strong>sul<strong>in</strong>to <strong>in</strong>duce hypoglycaemia (Cryer, 1981; Gerich, 1988). The response to the bolus <strong>in</strong>jectionof 0.1 U/kg <strong>in</strong>sul<strong>in</strong> <strong>in</strong> a normal subject is shown <strong>in</strong> Figure 1.3. Blood glucose concentrationsdecl<strong>in</strong>e with<strong>in</strong> m<strong>in</strong>utes of the adm<strong>in</strong>istration of <strong>in</strong>sul<strong>in</strong> and reach a nadir after 20–30 m<strong>in</strong>utes,then gradually rise to near normal by two hours after the <strong>in</strong>sul<strong>in</strong> was adm<strong>in</strong>istered. The factFigure 1.3 (a) Glucose and (b) <strong>in</strong>sul<strong>in</strong> concentrations after <strong>in</strong>travenous <strong>in</strong>jection of <strong>in</strong>sul<strong>in</strong> 0.1 U/kgat time 0. Reproduced from Garber et al. (1976) by permission of the Journal of Cl<strong>in</strong>ical Investigation

8 NORMAL GLUCOSE METABOLISM AND RESPONSESthat blood glucose starts to rise when plasma <strong>in</strong>sul<strong>in</strong> concentrations are still ten times thebasel<strong>in</strong>e values means that it is not simply the reduction <strong>in</strong> <strong>in</strong>sul<strong>in</strong> that reverses hypoglycaemia,but active counterregulation must also occur. Many hormones are released whenblood glucose is lowered (see below), but glucagon, the catecholam<strong>in</strong>es, growth hormoneand cortisol are regarded as be<strong>in</strong>g the most important.Several studies have determ<strong>in</strong>ed the relative importance of these hormones by produc<strong>in</strong>gisolated deficiencies of each hormone (by block<strong>in</strong>g its release or action) and assess<strong>in</strong>gthe subsequent response to adm<strong>in</strong>istration of <strong>in</strong>sul<strong>in</strong>. These studies are exemplified <strong>in</strong>Figure 1.4 which assesses the relative importance of glucagon, adrenal<strong>in</strong>e (ep<strong>in</strong>ephr<strong>in</strong>e) andgrowth hormone <strong>in</strong> the counterregulation of short term hypoglycaemia. Somatostat<strong>in</strong> <strong>in</strong>fusionblocks glucagon and growth hormone secretion and significantly impairs glucose recovery(Figure 1.4a). If growth hormone is replaced <strong>in</strong> the same model to produce isolated glucagondeficiency (Figure 1.4b), and glucagon replaced to produce isolated growth hormone deficiency(Figure 1.4c), it is clear that it is glucagon and not growth hormone that is responsiblefor acute counterregulation. Comb<strong>in</strong>ed alpha and beta adrenoceptor blockade us<strong>in</strong>g phentolam<strong>in</strong>eand propranolol <strong>in</strong>fusions or adrenalectomy (Figure 1.4d), can be used to evaluatethe role of the catecholam<strong>in</strong>es. These and other studies demonstrate that glucagon is themost important counterregulatory hormone whereas catecholam<strong>in</strong>es provide a backup ifglucagon is deficient (for example <strong>in</strong> type 1 diabetes, see Chapters 6 and 7). Cortisol andFigure 1.4 Glucose recovery from acute hypoglycaemia. Glucose concentration follow<strong>in</strong>g an <strong>in</strong>travenous<strong>in</strong>jection of <strong>in</strong>sul<strong>in</strong> of 0.05 U/kg at time 0; after (a) sal<strong>in</strong>e <strong>in</strong>fusion (cont<strong>in</strong>uous l<strong>in</strong>e) andsomatostat<strong>in</strong>, (b) somatostat<strong>in</strong> and growth hormone (GH), (c) somatostat<strong>in</strong> and glucagon, (d) comb<strong>in</strong>edalpha and beta blockade with phentolam<strong>in</strong>e and propranolol <strong>in</strong>fusions or adrenalectomy, (e) somatostat<strong>in</strong>with alpha and beta blockade, and (f) somatostat<strong>in</strong> <strong>in</strong> adrenalectomised patients. Sal<strong>in</strong>e <strong>in</strong>fusion= cont<strong>in</strong>uous l<strong>in</strong>es; experimental study = broken l<strong>in</strong>es. Reproduced from Cryer (1981) courtesyof the American <strong>Diabetes</strong> Association (ep<strong>in</strong>ephr<strong>in</strong>e = adrenal<strong>in</strong>e)

COUNTERREGULATION DURING HYPOGLYCAEMIA 9growth hormone are important only <strong>in</strong> prolonged hypoglycaemia. Therefore if glucagonand catecholam<strong>in</strong>es are both deficient, as <strong>in</strong> longstand<strong>in</strong>g type 1 diabetes, counterregulationis seriously compromised, and the <strong>in</strong>dividual is defenceless aga<strong>in</strong>st acute hypoglycaemia(Cryer, 1981).Glucagon and catecholam<strong>in</strong>es <strong>in</strong>crease glycogenolysis and stimulate gluconeogenesis.Catecholam<strong>in</strong>es also reduce glucose utilisation peripherally and <strong>in</strong>hibit <strong>in</strong>sul<strong>in</strong> secretion.Cortisol and growth hormone <strong>in</strong>crease gluconeogenesis and reduce glucose utilisation. Therole of the other hormones (see below) <strong>in</strong> counterregulation is unclear, but they are unlikelyto make a significant contribution. F<strong>in</strong>ally, there is evidence that dur<strong>in</strong>g profound hypoglycaemia(blood glucose below 1.7 mmol/l), hepatic glucose output is stimulated directly,although the mechanism is unknown. This is termed hepatic autoregulation.The depth, as well as the duration, of hypoglycaemia is important <strong>in</strong> determ<strong>in</strong><strong>in</strong>g themagnitude of the counterregulatory hormone response. Studies us<strong>in</strong>g ‘hyper<strong>in</strong>sul<strong>in</strong>aemicclamps’ show a hierarchical response of hormone production. In this technique, <strong>in</strong>sul<strong>in</strong> is<strong>in</strong>fused at a constant rate and a glucose <strong>in</strong>fusion rate varied to ma<strong>in</strong>ta<strong>in</strong> blood glucoseconcentrations with<strong>in</strong> ±02 mmol/l of target concentrations. This permits the controlledevaluation of the counterregulatory hormone response at vary<strong>in</strong>g degrees of hypoglycaemia.It also demonstrates that glucagon, catecholam<strong>in</strong>es and growth hormone start to be secretedat a blood glucose concentration of 3.5–3.7 mmol/l, with cortisol produced at a lower glucoseof 3.0 mmol/l (Mitrakou et al., 1991). The counterregulatory response is <strong>in</strong>itiated beforeimpairment <strong>in</strong> cerebral function commences, usually at a blood glucose concentration ofapproximately 3.0 mmol/l (Heller and Macdonald, 1996).The magnitude of the hormonal response also depends on the length of the hypoglycaemicepisode. The counterregulatory hormonal response commences up to 20 m<strong>in</strong>utesafter hypoglycaemia is achieved and cont<strong>in</strong>ues to rise for 60 m<strong>in</strong>utes (Kerr et al., 1989).In contrast, this response is attenuated as a result of a previous episode of hypoglycaemia(with<strong>in</strong> a few days) (reviewed by Heller and Macdonald, 1996) and even by prolongedexercise the day before hypoglycaemia is <strong>in</strong>duced. Galassetti et al. (2001) showed that <strong>in</strong>non-diabetic subjects three hours of moderate <strong>in</strong>tensity exercise the previous day markedlydecreased the counterregulatory response to hypoglycaemia <strong>in</strong>duced by the <strong>in</strong>fusion of<strong>in</strong>sul<strong>in</strong>, and that the reduced counterregulatory response was more marked <strong>in</strong> men than<strong>in</strong> women.Although the primary role of the counterregulatory hormones is on glucose metabolism,any effects on fatty acid utilisation can have an <strong>in</strong>direct effect on blood glucose. Thus,the <strong>in</strong>crease <strong>in</strong> plasma ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) (and activation of the sympathetic nervoussystem) that is seen <strong>in</strong> hypoglycaemia can stimulate lipolysis of triglyceride <strong>in</strong> adipose tissueand muscle and release fatty acids which can be used as an alternative fuel to glucose, mak<strong>in</strong>gmore glucose available for the CNS. Enoksson et al. (2003) demonstrated that patients withtype 1 diabetes, who had lower plasma ep<strong>in</strong>ephr<strong>in</strong>e responses to hypoglycaemia than nondiabeticcontrols, also had reduced rates of lipolysis <strong>in</strong> adipose tissue and skeletal muscle,mak<strong>in</strong>g them more dependent on glucose as a fuel and therefore at risk of develop<strong>in</strong>g a moresevere hypoglycaemia.The complex counterregulatory and homeostatic mechanisms described above are thoughtto be mostly under the control of the central nervous system. Evidence for this comesfrom studies <strong>in</strong> dogs, where glucose was <strong>in</strong>fused <strong>in</strong>to the carotid and vertebral arteries toma<strong>in</strong>ta<strong>in</strong> euglycaemia <strong>in</strong> the bra<strong>in</strong>. Despite peripheral hypoglycaemia, glucagon did not<strong>in</strong>crease and responses of the other counterregulatory hormones were blunted. This, and

10 NORMAL GLUCOSE METABOLISM AND RESPONSESother studies <strong>in</strong> rats, led to the hypothesis that the ventromedial nucleus of the hypothalamus(VMH), which does not have a blood–bra<strong>in</strong> barrier, acts as a glucose-sensor and co-ord<strong>in</strong>atescounterregulation (Borg et al., 1997). However, evidence exists that other parts of the bra<strong>in</strong>may also be <strong>in</strong>volved <strong>in</strong> mediat<strong>in</strong>g counterregulation.It is now clear that glucose-sens<strong>in</strong>g neurones can <strong>in</strong>volve either glucok<strong>in</strong>ase or ATPsensitiveK + channels (Lev<strong>in</strong> et al., 2004). In rats, the VMH has ATP-sensitive K + channelswhich seem to be <strong>in</strong>volved <strong>in</strong> the counterregulatory responses to hypoglycaemia, as <strong>in</strong>jectionof the sulphonylurea, glibenclamide, directly <strong>in</strong>to the VMH suppressed hormonal responsesto systemic hypoglycaemia (Evans et al., 2004).The existence of hepatic autoregulation suggests that some peripheral control should exist.Studies produc<strong>in</strong>g central euglycaemia and hepatic portal venous hypoglycaemia <strong>in</strong> dogshave provided evidence for hepatic glucose sensors and suggest that these sensors, as wellas those <strong>in</strong> the bra<strong>in</strong>, are important <strong>in</strong> the regulation of glucose (Hamilton-Wessler et al.,1994). However, this topic is somewhat controversial and more recent studies on dogs havefailed to demonstrate an effect of hepatic sensory nerves on the responses to hypoglycaemia(Jackson et al., 2000). Moreover, studies <strong>in</strong> humans by Heptulla et al. (2001) showedthat provid<strong>in</strong>g glucose orally rather than <strong>in</strong>travenously dur<strong>in</strong>g a hypoglycaemic hyper<strong>in</strong>sul<strong>in</strong>aemicclamp actually enhanced the counterregulatory hormone responses rather thanreduced them.HORMONAL CHANGES DURING HYPOGLYCAEMIA<strong>Hypoglycaemia</strong> <strong>in</strong>duces the secretion of various hormones, some of which are responsible forcounterregulation, many of the physiological changes that occur as a consequence of lower<strong>in</strong>gblood glucose and contribute to symptom generation (see Chapter 2), The stimulation of theautonomic nervous system is central to many of these changes.Activation of the Autonomic Nervous SystemThe autonomic nervous system comprises sympathetic and parasympathetic components(Figure 1.5). Fibres from the sympathetic division leave the sp<strong>in</strong>al cord with the ventral rootsfrom the first thoracic to the third or fourth lumbar nerves to synapse <strong>in</strong> the sympatheticcha<strong>in</strong> or visceral ganglia, and the long postganglionic fibres are <strong>in</strong>corporated <strong>in</strong> somaticnerves. The parasympathetic pathways orig<strong>in</strong>ate <strong>in</strong> the nuclei of cranial nerves III, VII, IXand X, and travel with the vagus nerve. A second component, the sacral outflow, suppliesthe pelvic viscera via the pelvic branches of the second to fourth sp<strong>in</strong>al nerves. The ganglia<strong>in</strong> both cases are located near the organs supplied, and the postganglionic neurones aretherefore short.Selective activation of both components of the autonomic system occurs dur<strong>in</strong>g hypoglycaemia.The sympathetic nervous system <strong>in</strong> particular is responsible for many of thephysiological changes dur<strong>in</strong>g hypoglycaemia and the evidence for its activation can beobta<strong>in</strong>ed <strong>in</strong>directly by observ<strong>in</strong>g functional changes such as cardiovascular responses (consideredbelow), measur<strong>in</strong>g plasma catecholam<strong>in</strong>es which gives a general <strong>in</strong>dex of sympatheticactivation, or by directly record<strong>in</strong>g sympathetic activity.

HORMONAL CHANGES DURING HYPOGLYCAEMIA 11Figure 1.5 Anatomy of the autonomic nervous system. Pre = preganglionic neurones; post = postganglionicneurones; RC = ramus communicansDirect record<strong>in</strong>gs are possible from sympathetic nerves supply<strong>in</strong>g skeletal muscle and sk<strong>in</strong>.Sympathetic neural activity <strong>in</strong> skeletal muscle <strong>in</strong>volves vasoconstrictor fibres which <strong>in</strong>nervateblood vessels and are <strong>in</strong>volved <strong>in</strong> controll<strong>in</strong>g blood pressure. Dur<strong>in</strong>g hypoglycaemia (<strong>in</strong>ducedby <strong>in</strong>sul<strong>in</strong>), the frequency and amplitude of muscle sympathetic activity are <strong>in</strong>creased asblood glucose falls, with an <strong>in</strong>crease <strong>in</strong> activity eight m<strong>in</strong>utes after <strong>in</strong>sul<strong>in</strong> is <strong>in</strong>jected<strong>in</strong>travenously, peak<strong>in</strong>g at 25–30 m<strong>in</strong>utes co<strong>in</strong>cident with the glucose nadir, and persist<strong>in</strong>g

12 NORMAL GLUCOSE METABOLISM AND RESPONSESFigure 1.6 (a) Muscle sympathetic activity dur<strong>in</strong>g euglycaemia and hypoglycaemia. Reproducedfrom Fagius et al. (1986) courtesy of the American <strong>Diabetes</strong> Association. (b) Sk<strong>in</strong> sympatheticactivity dur<strong>in</strong>g euglycaemia and hypoglycaemia. Reproduced from Berne and Fagius (1986), with k<strong>in</strong>dpermission from Spr<strong>in</strong>ger Science and Bus<strong>in</strong>ess Mediafor 90 m<strong>in</strong>utes after euglycaemia is restored (Figure 1.6a) (Fagius et al., 1986). Dur<strong>in</strong>ghypoglycaemia, a sudden <strong>in</strong>crease <strong>in</strong> sk<strong>in</strong> sympathetic activity is seen, which co<strong>in</strong>cideswith the onset of sweat<strong>in</strong>g. This sweat<strong>in</strong>g leads to vasodilatation of sk<strong>in</strong> blood vessels,which is also contributed to by a reduction <strong>in</strong> sympathetic stimulation of the vasoconstrictorcomponents of sk<strong>in</strong> arterio-venous anastomoses (Figure 1.6b) (Berne and Fagius, 1986).These effects (at least <strong>in</strong>itially) <strong>in</strong>crease total sk<strong>in</strong> blood flow and promote heat loss fromthe body.Activation of both muscle and sk<strong>in</strong> sympathetic nerve activity are thought to be centrallymediated. Tissue neuroglycopenia can be produced by 2-deoxy-D-glucose, a glucoseanalogue, without <strong>in</strong>creas<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>. Infusion of this analogue causes stimulation of muscleand sk<strong>in</strong> sympathetic activity demonstrat<strong>in</strong>g that it is the hypoglycaemia per se, and not the<strong>in</strong>sul<strong>in</strong> used to <strong>in</strong>duce it, which is responsible for the sympathetic activation (Fagius andBerne, 1989).The activation of the parasympathetic nervous system (vagus nerve) dur<strong>in</strong>g hypoglycaemiacannot be measured directly. The most useful <strong>in</strong>dex of parasympathetic function is themeasurement of plasma pancreatic polypeptide, the peptide hormone secreted by the PP cellsof the pancreas, which is released <strong>in</strong> response to vagal stimulation.

HORMONAL CHANGES DURING HYPOGLYCAEMIA 13Neuroendocr<strong>in</strong>e Activation (Box 1.3)Insul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia was used to study pituitary function as early as the 1940s.The development of assays for adrenocorticotrophic hormone (ACTH) and growth hormone(GH) allowed the direct measurement of pituitary function dur<strong>in</strong>g hypoglycaemia <strong>in</strong> the1960s, and many of the processes govern<strong>in</strong>g these changes were unravelled before elucidationof the counterregulatory system. The studies are comparable to those evaluat<strong>in</strong>gcounterregulation, <strong>in</strong> that potential regulatory factors are blocked to measure the hormonalresponse to hypoglycaemia with and without the regulat<strong>in</strong>g factor.Box 1.3Neuroendocr<strong>in</strong>e activationHypothalamus ↑ Corticotrophic releas<strong>in</strong>g hormone↑ Growth hormone releas<strong>in</strong>g hormoneAnterior Pituitary ↑ Adrenocorticotrophic hormone↑ Beta endorph<strong>in</strong>↑ Growth hormone↑ Prolact<strong>in</strong>↔ Thyrotroph<strong>in</strong>↔ Gonadotroph<strong>in</strong>sPosterior pituitary ↑ Vasopress<strong>in</strong>↑ Oxytoc<strong>in</strong>Pancreas ↑ Glucagon↑ Pancreatic polypeptide↑ Insul<strong>in</strong>Adrenal ↑ Cortisol↑ Ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e)↑ AldosteroneOthers ↑ Parathyroid hormone↑ Gastr<strong>in</strong>↑ Somatostat<strong>in</strong> (28)Hypothalamus and anterior pituitaryACTH, GH and prolact<strong>in</strong> concentrations <strong>in</strong>crease dur<strong>in</strong>g hypoglycaemia, but there is nochange <strong>in</strong> thyrotroph<strong>in</strong> or gonadotroph<strong>in</strong> secretion. The secretion of these pituitary hormonesis controlled by releas<strong>in</strong>g factors which are produced <strong>in</strong> the median em<strong>in</strong>ence of the hypothalamus,secreted <strong>in</strong>to the hypophyseal portal vessels and then pass to the pituitary gland(Figure 1.7). The mechanisms regulat<strong>in</strong>g the releas<strong>in</strong>g factors are <strong>in</strong>completely understood,but may <strong>in</strong>volve the ventromedial nucleus, one site where bra<strong>in</strong> glucose sensors are situated(Fish et al., 1986).

14 NORMAL GLUCOSE METABOLISM AND RESPONSESFigure 1.7Anatomy of the hypothalamus and pituitary gland• ACTH: Secretion is governed by release of corticotroph<strong>in</strong> releas<strong>in</strong>g hormone (CRH) fromthe hypothalamus; alpha adrenoceptors stimulate CRH release, and beta adrenoceptorshave an <strong>in</strong>hibitory action. A variety of neurotransmitters control the release of CRH <strong>in</strong>tothe portal vessels, <strong>in</strong>clud<strong>in</strong>g seroton<strong>in</strong> and acetylchol<strong>in</strong>e which are stimulatory and GABAwhich is <strong>in</strong>hibitory. The <strong>in</strong>crease <strong>in</strong> ACTH causes cortisol to be secreted from the corticesof the adrenal glands.• Beta endorph<strong>in</strong>s are derived from the same precursors as ACTH and are co-secreted withit. The role of endorph<strong>in</strong>s <strong>in</strong> counterregulation is uncerta<strong>in</strong>, but they may <strong>in</strong>fluence thesecretion of the other pituitary hormones dur<strong>in</strong>g hypoglycaemia.• GH: Growth hormone secretion is governed by two hypothalamic hormones: growthhormone releas<strong>in</strong>g hormone (GHRH) which stimulates GH secretion, and somatostat<strong>in</strong>which is <strong>in</strong>hibitory. GHRH secretion is stimulated by dopam<strong>in</strong>e, GABA, opiates andthrough alpha adrenoceptors, whereas it is <strong>in</strong>hibited by seroton<strong>in</strong> and beta adrenoceptors.A study <strong>in</strong> rats showed that bioassayable GH and GHRH are depleted <strong>in</strong> the pituitary andhypothalamus respectively after <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia (Katz et al., 1967).• Prolact<strong>in</strong>: The mechanisms underly<strong>in</strong>g its secretion are not established. Prolact<strong>in</strong> secretionis normally under the <strong>in</strong>hibitory control of dopam<strong>in</strong>e, but evidence also exists forreleas<strong>in</strong>g factors be<strong>in</strong>g produced dur<strong>in</strong>g hypoglycaemia. Prolact<strong>in</strong> does not contribute tocounterregulation.Posterior pituitaryVasopress<strong>in</strong> and oxytoc<strong>in</strong> both <strong>in</strong>crease dur<strong>in</strong>g hypoglycaemia (Fisher et al., 1987). Theirsecretion is under hormonal and neurotransmitter control <strong>in</strong> a similar way to the hypothalamic

PHYSIOLOGICAL RESPONSES 15hormones. Vasopress<strong>in</strong> has glycolytic actions and oxytoc<strong>in</strong> <strong>in</strong>creases hepatic glucose output<strong>in</strong> dogs, but their contribution to glucose counterregulation is uncerta<strong>in</strong>.Pancreas• Glucagon: The mechanisms of glucagon secretion dur<strong>in</strong>g hypoglycaemia are still not fullyunderstood. Although activation of the autonomic nervous system stimulates its release,this pathway has been shown to be less important <strong>in</strong> humans. A reduction <strong>in</strong> glucoseconcentrations may have a direct effect on the glucagon-secret<strong>in</strong>g pancreatic alpha cells,or the reduced beta cell activity (reduced <strong>in</strong>sul<strong>in</strong> secretion), which also occurs with lowblood glucose, may release the tonic <strong>in</strong>hibition of glucagon secretion. However, suchmechanisms would be disturbed <strong>in</strong> type 1 diabetes, where hypoglycaemia is normallyassociated with high plasma <strong>in</strong>sul<strong>in</strong> levels and there is no direct effect of beta cell-derived<strong>in</strong>sul<strong>in</strong> on the alpha cells.• Somatostat<strong>in</strong>: This is thought of as a pancreatic hormone produced from D cells of theislets of Langerhans, but it is also secreted <strong>in</strong> other parts of the gastro<strong>in</strong>test<strong>in</strong>al tract. Thereare a number of structurally different polypeptides derived from prosomatostat<strong>in</strong>: thesomatostat<strong>in</strong>-14 peptide is secreted from D cells, and somatostat<strong>in</strong>-28 from the gastro<strong>in</strong>test<strong>in</strong>altract. The plasma concentration of somatostat<strong>in</strong>-28 <strong>in</strong>creases dur<strong>in</strong>g hypoglycaemia(Francis and Ens<strong>in</strong>ck, 1987). The normal action of somatostat<strong>in</strong> is to <strong>in</strong>hibit the secretionboth of <strong>in</strong>sul<strong>in</strong> and glucagon, but somatostat<strong>in</strong>-28 <strong>in</strong>hibits <strong>in</strong>sul<strong>in</strong> ten times more effectivelythan glucagon, and thus may have a role <strong>in</strong> counterregulation by suppress<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>release.• Pancreatic polypeptide: This peptide has no known role <strong>in</strong> counterregulation, but its releasedur<strong>in</strong>g hypoglycaemia is stimulated by chol<strong>in</strong>ergic fibres through muscar<strong>in</strong>ic receptorsand is a useful marker of parasympathetic activity.Adrenal and Ren<strong>in</strong>–Angiotens<strong>in</strong> systemThe processes govern<strong>in</strong>g the <strong>in</strong>crease <strong>in</strong> cortisol dur<strong>in</strong>g hypoglycaemia are discussed above.The rise <strong>in</strong> catecholam<strong>in</strong>es, <strong>in</strong> particular ep<strong>in</strong>ephr<strong>in</strong>e from the adrenal medulla, which occurswhen blood glucose is lowered, is controlled by sympathetic fibres <strong>in</strong> the splanchnic nerve.The <strong>in</strong>crease <strong>in</strong> ren<strong>in</strong>, and therefore angiotens<strong>in</strong> and aldosterone, dur<strong>in</strong>g hypoglycaemia isstimulated primarily by the <strong>in</strong>tra-renal effects of <strong>in</strong>creased catecholam<strong>in</strong>es, mediated throughbeta adrenoceptors, although the <strong>in</strong>crease <strong>in</strong> ACTH and hypokalaemia due to hypoglycaemiacontributes (Trovati et al., 1988; Jungman et al., 1989). These changes do not have asignificant role <strong>in</strong> counterregulation, although angiotens<strong>in</strong> II has glycolytic actions <strong>in</strong> vitro.PHYSIOLOGICAL RESPONSESHaemodynamic Changes (Box 1.4)The haemodynamic changes dur<strong>in</strong>g hypoglycaemia (Hilsted, 1993) are mostly caused bythe activation of the sympathetic nervous system and an <strong>in</strong>crease <strong>in</strong> circulat<strong>in</strong>g ep<strong>in</strong>ephr<strong>in</strong>e.

16 NORMAL GLUCOSE METABOLISM AND RESPONSESBox 1.4Haemodynamic changes↑↑↑↓↑Heart rateSystolic blood pressureCardiac outputPeripheral resistanceMyocardial contractilityAn <strong>in</strong>crease <strong>in</strong> heart rate (tachycardia), myocardial contractility and cardiac output occurs,which is mediated through beta 1 adrenoceptors, but <strong>in</strong>creas<strong>in</strong>g vagal tone counteracts thiseffect so the <strong>in</strong>crease is transient. Peripheral resistance, estimated from mean arterial pressuredivided by cardiac output, is reduced. A comb<strong>in</strong>ation of the <strong>in</strong>crease <strong>in</strong> cardiac output andreduction <strong>in</strong> peripheral resistance results <strong>in</strong> an <strong>in</strong>crease <strong>in</strong> systolic and a decrease <strong>in</strong> diastolicpressure, i.e. a widen<strong>in</strong>g of pulse pressure without a change <strong>in</strong> mean arterial pressure.Changes <strong>in</strong> Regional Blood Flow (Box 1.5 and Figure 1.8)• Cerebral blood flow: Early work produced conflict<strong>in</strong>g results, but these studies were<strong>in</strong> subjects receiv<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> shock therapy, and the vary<strong>in</strong>g effects of convulsions andaltered level of consciousness may have <strong>in</strong>fluenced the outcome. Subsequent studies haveconsistently shown an <strong>in</strong>crease <strong>in</strong> cerebral blood flow dur<strong>in</strong>g hypoglycaemia despite theuse of different methods of measurement (isotopic, s<strong>in</strong>gle photon emission computedtomography (SPECT) and Doppler ultrasound). In most of the studies blood glucoseconcentration was less than 2 mmol/l before a change was observed. In animals, hypoglycaemiais associated with loss of cerebral autoregulation (the ability of the bra<strong>in</strong> to ma<strong>in</strong>ta<strong>in</strong>cerebral blood flow despite variability <strong>in</strong> cardiac output) through beta adrenoceptor stimulation,but the exact mechanisms are unknown (Bryan, 1990; Sieber and Traysman,1992).• Gastro<strong>in</strong>test<strong>in</strong>al system: Total splanchnic blood flow (supply<strong>in</strong>g the <strong>in</strong>test<strong>in</strong>es, liver,spleen and stomach) is <strong>in</strong>creased and splanchnic vascular resistance reduced as assessedby the bromosulphthale<strong>in</strong> extraction technique (Bearn et al., 1952). Superior mesentericartery blood flow measured us<strong>in</strong>g Doppler ultrasound <strong>in</strong>creases dur<strong>in</strong>g hypoglycaemiadue to beta adrenoceptor stimulation (Braatvedt et al., 1993). Radioisotope scann<strong>in</strong>g hasBox 1.5Changes <strong>in</strong> regional blood flow↑↑↓↑↓Cerebral flowTotal splanchnic flowSplenic flowSk<strong>in</strong> flow variable (early ↑, late ↓)Muscle flowRenal flow

PHYSIOLOGICAL RESPONSES 17Figure 1.8Changes <strong>in</strong> regional blood flow dur<strong>in</strong>g hypoglycaemiademonstrated a reduction <strong>in</strong> splenic activity dur<strong>in</strong>g hypoglycaemia (Fisher et al., 1990),which is thought to be a consequence of alpha adrenoceptor-mediated reduction <strong>in</strong> bloodflow. These changes would all be expected to <strong>in</strong>crease hepatic blood flow.• Sk<strong>in</strong>: The control of blood flow to the sk<strong>in</strong> is complex and different mechanisms predom<strong>in</strong>ate<strong>in</strong> different areas. Studies of the effect of hypoglycaemia on sk<strong>in</strong> blood flow are<strong>in</strong>consistent partly because different methods have been used for blood flow measurementand <strong>in</strong>duction of hypoglycaemia, as well as differences <strong>in</strong> the part of the bodystudied. Def<strong>in</strong>itive conclusions are therefore not possible. Studies us<strong>in</strong>g the dorsum ofthe foot and the face (cheek and forehead) have consistently shown an <strong>in</strong>itial vasodilatationand <strong>in</strong>crease <strong>in</strong> blood flow followed by later vasoconstriction at a blood glucose of2.5 mmol/l (Maggs et al., 1994). These f<strong>in</strong>d<strong>in</strong>gs are consistent with the cl<strong>in</strong>ical pictureof <strong>in</strong>itial flush<strong>in</strong>g and later pallor, with an early rise <strong>in</strong> sk<strong>in</strong> blood flow followed by alater fall.• Muscle blood flow: A variety of techniques have been used to study muscle blood flow(<strong>in</strong>clud<strong>in</strong>g venous occlusion plethysmography, isotopic clearance techniques and the useof thermal conductivity meters). All studies have consistently shown an <strong>in</strong>crease <strong>in</strong> muscle

18 NORMAL GLUCOSE METABOLISM AND RESPONSESblood flow dur<strong>in</strong>g hypoglycaemia irrespective of sk<strong>in</strong> blood flow. This change is mediatedby beta 2 adrenoceptors (Abramson et al., 1966; Allwood et al., 1959).• Kidney: Inul<strong>in</strong> and sodium hippurate clearance can be used to estimate glomerular filtrationrate and renal blood flow respectively. Both decrease dur<strong>in</strong>g hypoglycaemia (Patrick et al.,1989) and catecholam<strong>in</strong>es and ren<strong>in</strong> are implicated <strong>in</strong> <strong>in</strong>itiat<strong>in</strong>g the changes.The changes <strong>in</strong> blood flow <strong>in</strong> various organs, like the haemodynamic changes, are mostlymediated by the activation of the sympathetic nervous system or circulat<strong>in</strong>g ep<strong>in</strong>ephr<strong>in</strong>e. Themajority either protect aga<strong>in</strong>st hypoglycaemia or <strong>in</strong>crease substrate delivery to vital organs.The <strong>in</strong>crease <strong>in</strong> cerebral blood flow <strong>in</strong>creases substrate delivery to the bra<strong>in</strong>. Increas<strong>in</strong>gmuscle flow enhances the release and washout of gluconeogenic precursors. The <strong>in</strong>crease <strong>in</strong>splanchnic blood flow and reduction <strong>in</strong> splenic blood flow serve to <strong>in</strong>crease hepatic bloodflow to maximise hepatic glucose production. Meanwhile, blood is diverted away fromorgans such as the kidney and spleen, which are not required <strong>in</strong> the acute response to themetabolic stress.Functional Changes (Box 1.6)• Sweat<strong>in</strong>g: Sweat<strong>in</strong>g is mediated by sympathetic chol<strong>in</strong>ergic nerves, although other neurotransmitterssuch as vasoactive <strong>in</strong>test<strong>in</strong>al peptide and bradyk<strong>in</strong><strong>in</strong> may also be <strong>in</strong>volved.The activation of the sympathetic <strong>in</strong>nervation of the sk<strong>in</strong> as described above results <strong>in</strong>the sudden onset of sweat<strong>in</strong>g. Sweat<strong>in</strong>g is one of the first physiological responses tooccur dur<strong>in</strong>g hypoglycaemia and can be demonstrated with<strong>in</strong> ten m<strong>in</strong>utes of achiev<strong>in</strong>ga blood glucose of 2.5 mmol/l (Maggs et al., 1994). It co<strong>in</strong>cides with the onset ofother measures of autonomic activation such as an <strong>in</strong>crease <strong>in</strong> heart rate and tremor(Figure 1.9).• Tremor: Trembl<strong>in</strong>g and shak<strong>in</strong>g are characteristic features of hypoglycaemia and resultfrom an <strong>in</strong>crease <strong>in</strong> physiological tremor. The rise <strong>in</strong> cardiac output and vasodilatationoccurr<strong>in</strong>g dur<strong>in</strong>g hypoglycaemia <strong>in</strong>crease the level of physiological tremor and thisis exacerbated by beta adrenoceptor stimulation associated with <strong>in</strong>creased ep<strong>in</strong>ephr<strong>in</strong>econcentrations (Kerr et al., 1990). S<strong>in</strong>ce adrenalectomy does not entirely abolish tremor,other components such as the activation of muscle sympathetic activity must be<strong>in</strong>volved.Box 1.6Functional changes↑↑↓↓↑↑Sweat<strong>in</strong>g (sudden onset)TremorCore temperatureIntraocular pressureJejunal activityGastric empty<strong>in</strong>g

PHYSIOLOGICAL RESPONSES 19We do not have rights to reproduce thisfigure electronicallyFigure 1.9 Sudden onset of sweat<strong>in</strong>g, tremor and <strong>in</strong>crease <strong>in</strong> heart rate dur<strong>in</strong>g the <strong>in</strong>duction of hypoglycaemia.Reproduced from <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>: Cl<strong>in</strong>ical and Physiological Aspects (edsB.M. Frier and M. Fisher), © 1993 Edward Arnold, by permission of Edward Arnold (Publishers) Ltd• Temperature: Despite a beta adrenoceptor-mediated <strong>in</strong>crease <strong>in</strong> metabolic rate, coretemperature falls dur<strong>in</strong>g hypoglycaemia. The mechanisms by which this occurs dependon whether the environment is warm or cold. In a warm environment, heat is lost becauseof sweat<strong>in</strong>g and <strong>in</strong>creased heat conduction from vasodilatation. <strong>Hypoglycaemia</strong> reducescore temperature by 03 C and sk<strong>in</strong> temperature up to 2 C (depend<strong>in</strong>g on the part of thebody measured) after 60 m<strong>in</strong>utes (Maggs et al., 1994). Shiver<strong>in</strong>g is reduced <strong>in</strong> the cold,and together with vasodilatation and sweat<strong>in</strong>g this causes a substantial reduction <strong>in</strong> core

20 NORMAL GLUCOSE METABOLISM AND RESPONSEStemperature (Gale et al., 1983). In rats, mortality was <strong>in</strong>creased <strong>in</strong> animals whose coretemperature was prevented from fall<strong>in</strong>g dur<strong>in</strong>g hypoglycaemia (Buchanan et al., 1991).In humans there is anecdotal evidence from subjects undergo<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> shock therapythat those who had a rise <strong>in</strong> body temperature showed delayed neurological recovery(Ramos et al., 1968). These f<strong>in</strong>d<strong>in</strong>gs support the hypothesis that the fall <strong>in</strong> core temperaturereduces metabolic rate, allow<strong>in</strong>g hypoglycaemia to be better tolerated, and thus thechanges <strong>in</strong> body temperature are of survival value. The beneficial effects are likely tobe limited, particularly <strong>in</strong> a cold environment, where the impairment of cerebral functionmeans subjects may not realise they are cold, caus<strong>in</strong>g them to be at risk of severehypothermia.• Other functional changes <strong>in</strong>clude a reduction <strong>in</strong> <strong>in</strong>traocular pressure, greater jejunal butnot gastric motility and <strong>in</strong>consistent abnormalities of liver function tests. An <strong>in</strong>crease<strong>in</strong> gastric empty<strong>in</strong>g occurs dur<strong>in</strong>g hypoglycaemia (Schvarcz et al., 1995), which maybe protective <strong>in</strong> that carbohydrate delivery to the <strong>in</strong>test<strong>in</strong>e is <strong>in</strong>creased, enabl<strong>in</strong>g fasterglucose absorption and reversal of hypoglycaemia.CONCLUSIONS• Homeostatic mechanisms exist to ma<strong>in</strong>ta<strong>in</strong> glucose concentration with<strong>in</strong> narrow limitsdespite a wide variety of circumstances.• The dependence of the central nervous system on glucose has led to a complex series ofbiochemical, functional and haemodynamic changes aimed at restor<strong>in</strong>g glucose concentrations,produc<strong>in</strong>g symptoms and protect<strong>in</strong>g the body <strong>in</strong> general, and central nervous system<strong>in</strong> particular, aga<strong>in</strong>st the effects of a low blood glucose (Figure 1.10).• Many symptoms of hypoglycaemia result from the activation of the autonomic nervoussystem and help to warn the <strong>in</strong>dividual that blood glucose is low. This encourages the<strong>in</strong>gestion of carbohydrate, so help<strong>in</strong>g to restore glucose concentrations <strong>in</strong> addition tocounterregulation.• Faster gastric empty<strong>in</strong>g and the changes <strong>in</strong> regional blood flow which also occur as aresult of the activation of the autonomic nervous system <strong>in</strong>crease substrate delivery.• The greater cerebral blood flow <strong>in</strong>creases glucose delivery to the bra<strong>in</strong> (although lossof autoregulation is undesirable), and the <strong>in</strong>creased splanchnic flow results <strong>in</strong> a greaterdelivery of gluconeogenic precursors to the liver.• Activation of the autonomic nervous system also <strong>in</strong>creases sweat<strong>in</strong>g, and together with the<strong>in</strong>hibition of shiver<strong>in</strong>g, this predisposes to hypothermia, which may be neuroprotective.ACKNOWLEDGEMENTSWe would like to thank Professor Robert Tattersall for read<strong>in</strong>g the chapter and for his helpfulsuggestions.

REFERENCES 21NORMAL GLUCOSE HOMEOSTASISINSULIN : GLUCAGON RATIOExcess Insul<strong>in</strong>HYPOGLYCAEMIADETECTION BY BRAIN GLUCOSE SENSORSPANCREASPITUITARYADRENALACTIVATION OFAUTONOMICNERVOUS SYSTEMGLUCAGON(GH and CORTISOL)CATECHOLAMINESCOUNTERREGULATIONGASTRIC EMPTYINGANDCHANGES IN BLOOD FLOWSWEATING(<strong>in</strong>creased substrateproduction)SYMPTOMS(Autonomic)CEREBRALBLOOD FLOWSPLANCHNICBLOOD FLOW(eat<strong>in</strong>g)HYPOTHERMIARESTORE BLOODGLUCOSEINCREASE SUBSTRATEDELIVERYPROTECTION OFVITAL ORGANSFigure 1.10Glucose homeostasis and the correction of hypoglycaemiaREFERENCESAbramson EA, Arky RA, Woeber KA (1966). Effects of propranolol on the hormonal and metabolicresponses to <strong>in</strong>sul<strong>in</strong> <strong>in</strong>duced hypoglycaemia. Lancet ii: 1386–9.Allwood MJ, Hensel H, Papenberg J (1959). Muscle and sk<strong>in</strong> blood flow <strong>in</strong> the human forearm dur<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> hypoglycaemia. Journal of Physiology 147: 269–73.Auer RN, Siesjö BK (1988). Biological differences between ischaemia, hypoglycaemia and epilepsy.Annals of Neurology 24: 699–707.

22 NORMAL GLUCOSE METABOLISM AND RESPONSESBearn AG, Bull<strong>in</strong>g BH, Sherlock S (1952). The response of the liver to <strong>in</strong>sul<strong>in</strong> <strong>in</strong> normal subjects and<strong>in</strong> diabetes mellitus: hepatic ve<strong>in</strong> catheterisation studies. Cl<strong>in</strong>ical Science 11: 151–64.Bell GI, Karyano T, Buse JB, Burant CF, Takeda T, L<strong>in</strong> D et al. (1990). Molecular biology ofmammalian glucose transporters. <strong>Diabetes</strong> Care 13: 198–208.Berne C, Fagius J (1986). Sk<strong>in</strong> sympathetic activity dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> <strong>in</strong>duced hypoglycaemia.Diabetologia 29: 855–60.Borg MA, Sherw<strong>in</strong> RS, Borg WP, Tamborlane WV, Shulman GI (1997). Local ventromedial hypothalamusglucose perfusion blocks counterregulation dur<strong>in</strong>g systemic hypoglycemia <strong>in</strong> awake rats.Journal of Cl<strong>in</strong>ical Investigation 99: 361–5.Boyle PJ, Kempers SE, O’Conner AM, Nagy RJ (1995). Bra<strong>in</strong> glucose uptake and unawarenessof hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 333: 1726–31.Braatvedt GD, Flynn MD, Stanners A, Halliwell M, Corrall RJM (1993). Splanchnic blood flow <strong>in</strong>man: evidence for mediation via a -adrenergic mechanism. Cl<strong>in</strong>ical Science 84: 201–7.Bryan RM (1990). Cerebral blood flow and energy metabolism dur<strong>in</strong>g stress. American Journal ofPhysiology 259: H269–80.Buchanan TA, Cane P, Eng CC, Sipos GF, Lee C (1991). Hypothermia is critical for survival dur<strong>in</strong>gprolonged <strong>in</strong>sul<strong>in</strong> <strong>in</strong>duced hypoglycemia <strong>in</strong> rats. Metabolism 40: 330–4.Cahill GF (1971). Physiology of <strong>in</strong>sul<strong>in</strong> <strong>in</strong> man. <strong>Diabetes</strong> 20: 785–99.Cryer PE (1981). Glucose counterregulation <strong>in</strong> man. <strong>Diabetes</strong> 30: 261–4.Enoksson S, Caprio SK, Rife F, Shulman GI, Tamborlane WV, Sherw<strong>in</strong> RS (2003). Defective activationof skeletal muscle and adipose tissue lipolysis <strong>in</strong> type 1 diabetes mellitus dur<strong>in</strong>g hypoglycemia.Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 88: 1503–11.Evans ML, McCrimmon RJ, Flanagan DE, Keshavarz T, Fan X, McNay EC et al. (2004). HypothalamicATP-sensitive K + channels play a key role <strong>in</strong> sens<strong>in</strong>g hypoglycemia and trigger<strong>in</strong>g counterregulatoryep<strong>in</strong>ephr<strong>in</strong>e and glucagon responses. <strong>Diabetes</strong> 53: 2542–51.Fagius J, Niklasson F, Berne C (1986). Sympathetic outflow <strong>in</strong> human muscle nerves <strong>in</strong>creases dur<strong>in</strong>ghypoglycemia. <strong>Diabetes</strong> 35: 1124–9.Fagius J, Berne C (1989). Changes <strong>in</strong> sympathetic nerve activity <strong>in</strong>duced by 2-deoxy-D-glucose<strong>in</strong>fusion <strong>in</strong> humans. American Journal of Endocr<strong>in</strong>ology 256: E714–21.Fish HR, Chernow B, O’Brian JT (1986). Endocr<strong>in</strong>e and neurophysiologic responses of the pituitaryof <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia: a review. Metabolism 35: 763–80.Fisher M, Baylis PH, Frier BM (1987). Plasma oxytoc<strong>in</strong>, arg<strong>in</strong><strong>in</strong>e vasopress<strong>in</strong> and atrial natriureticpeptide response to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong> man. Cl<strong>in</strong>ical Endocr<strong>in</strong>ology 26:179–85.Fisher M, Gillen G, Hepburn DA, Dargie HJ, Barnett E, Frier BM (1990). Splenic responses to acute<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong> humans. Cl<strong>in</strong>ical Science 78: 469–74.Francis BH, Ens<strong>in</strong>ck JW (1987). Differential alterations of the circulat<strong>in</strong>g prosomatostat<strong>in</strong>-derivedpeptides dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> <strong>in</strong>duced hypoglycemia <strong>in</strong> man. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology andMetabolism 65: 880–4.Gale EAM, Bennet J, Macdonald IA, Holst JJ, Mathews JA (1983). The physiological effects of<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong> man: responses at differ<strong>in</strong>g levels of blood glucose. Cl<strong>in</strong>icalScience 65: 262–71.Galassetti P, Neill AR, Tate D, Ertl AC, Wasserman DH, Davis SN (2001). Sexual dimorphism <strong>in</strong> counterregulatoryresponses to hypoglycemia after antecedent exercise. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ologyand Metabolism 86: 3516–24.Garber AJ, Cryer PE, Santiago JV, Hammond MW, Pagliara AS, Kipnis DM (1976). The role ofadrenergic mechanisms <strong>in</strong> the substrate and hormonal response to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong>man. Journal of Cl<strong>in</strong>ical Investigation 58: 7–15.Gerich JE (1988). Glucose counterregulation and its impact <strong>in</strong> diabetes mellitus. <strong>Diabetes</strong> 37:1608–17.

REFERENCES 23Hamilton-Wessler M, Bergman RN, Halter JB, Watanabe RM, Donovan CM (1994). The role ofliver glucosensors <strong>in</strong> the <strong>in</strong>tegrated sympathetic response <strong>in</strong>duced by deep hypoglycemia <strong>in</strong> dogs.<strong>Diabetes</strong> 43: 1052–60.Heller SR, Macdonald IA (1996). The measurement of cognitive function dur<strong>in</strong>g acute hypoglycaemia:experimental limitations and their effect on the study of hypoglycaemia unawareness. DiabeticMedic<strong>in</strong>e 13: 607–15.Heptulla RA, Tamborlane WV, Ma TY-Z, Rife F, Sherw<strong>in</strong> RS (2001). Oral glucose augments thecounterregulatory hormone response dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> humans. Journal ofCl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 86: 645–8.Hilsted J (1993). Cardiovascular changes dur<strong>in</strong>g hypoglycaemia. Cl<strong>in</strong>ical Physiology 13: 1–10.Jackson PA, Card<strong>in</strong> S, Coffey CS, Neal DW, Allen EJ, Penazola AR et al. (2000). Effect of hepaticdenervation on the counterregulatory response to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> the dog. AmericanJournal of Endocr<strong>in</strong>ology 279: E1249–57.Jungman E, Konzog C, Holl E, Fassib<strong>in</strong>der W, Schoffl<strong>in</strong>g K (1989). Effect of a human atrial natureticpeptide on blood glucose concentrations and hormone stimulation dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia<strong>in</strong> healthy man. European Journal of Cl<strong>in</strong>ical Pharmacology 36: 593–7.Katz SH, Dhariwal APS, McCann SM (1967). Effects of hypoglycaemia on the content of pituitarygrowth hormone (GH) and hypothalamic growth hormone releas<strong>in</strong>g factor (GNRH) <strong>in</strong> the rat.Endocr<strong>in</strong>ology 81: 333–9.Kerr D, Macdonald IA, Tattersall RB (1989). Influence of duration of hypoglycemia on the hormonalcounterregulatory response <strong>in</strong> normal subjects. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism68: 118–22.Kerr D, Macdonald IA, Heller SR, Tattersall RB (1990). A randomised double-bl<strong>in</strong>d placebo controlledtrial of the effects of Metoprolol CR, Atenolol and Propranolol LA on the physiological responsesto hypoglycaemia <strong>in</strong> the non-diabetic. British Journal of Cl<strong>in</strong>ical Pharmacology 29: 685–94.Lev<strong>in</strong> BE, Routh VH, Kang L, Sanders NM, Dunn-Meynell AA (2004). Neuronal glucosens<strong>in</strong>g: Whatdo we know after 50 years? <strong>Diabetes</strong> 53: 2521–8.Maggs DG, Scott AR, Macdonald IA (1994). Thermoregulatory responses to hyper<strong>in</strong>sul<strong>in</strong>emic hypoglycemiaand euglycemia <strong>in</strong> humans. American Journal of Physiology 267: R1266–72.McCall AL, Fixman LB, Flem<strong>in</strong>g N, Tornheim K, Chick W, Ruderman ND (1986). Chronic hypoglycemia<strong>in</strong>creases bra<strong>in</strong> glucose transport. American Journal of Endocr<strong>in</strong>ology 251: E442–5.McCall AL (1993). Effects of glucose deprivation on glucose metabolism <strong>in</strong> the central nervous system.In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>: Cl<strong>in</strong>ical and Physiological Aspects. Frier BM and Fisher M,eds. Edward Arnold, London: 56–71.Mitrakou A, Ryan C, Veneman T, Mokan M, Jenssen T, Kiss I et al. (1991). Hierarchy of glycemicthresholds for counterregulatory hormone secretion, symptoms and cerebral dysfunction. AmericanJournal of Endocr<strong>in</strong>ology 266: E67–74.Owen OE, Morgan AP, Kemp HG, Sullivan JM, Herrara MG, Cahill GF Jr (1967). Bra<strong>in</strong> metabolismdur<strong>in</strong>g fast<strong>in</strong>g. Journal of Cl<strong>in</strong>ical Investigation 46: 1589–95.Patrick AW, Hepburn DA, Craig KJ, Thompson I, Swa<strong>in</strong>son CD, Frier BM (1989). The effects of acute<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia on renal function <strong>in</strong> normal human subjects. Diabetic Medic<strong>in</strong>e 6:703–8.Ramos E, Zorilla E, Hadley WB (1968). Fever as a manifestation of hypoglycemia. Journal of theAmerican Medical Association 205: 590–2.Schvarcz E, Palmér M, Åman J, Berne C (1995). Hypoglycemia <strong>in</strong>creases the gastric empty<strong>in</strong>g rate<strong>in</strong> healthy subjects. <strong>Diabetes</strong> Care 18: 674–6.Sieber FE, Traysman RJ (1992). Special Issues: Glucose and the bra<strong>in</strong>. Critical Care Medic<strong>in</strong>e 20:104–14.Siegal AM, Kreisberg RA (1975). Metabolic homeostasis: <strong>in</strong>sul<strong>in</strong>-glucagon <strong>in</strong>teractions. In: <strong>Diabetes</strong>Mellitus (4th edition). Sussman KE and Metz RJS, eds. American <strong>Diabetes</strong> Association, New York:29–35.

24 NORMAL GLUCOSE METABOLISM AND RESPONSESSokaloff (1989). Circulation and energy metabolism of the bra<strong>in</strong>. In: Basic Neurochemistry. Siegel G,Agranoff B, Albers RW and Mol<strong>in</strong>off P, eds. Raven Press, New York: 565–90.Trovati M, Massucco P, Mularoni E, Cavalot F, Anfossi G, Matiello L, Emanuelli G (1988). Insul<strong>in</strong><strong>in</strong>ducedhypoglycaemia <strong>in</strong>creases plasma concentrations of angiotens<strong>in</strong> II and does not modify atrialnaturetic polypeptide secretion <strong>in</strong> man. Diabetologia 31: 816–20Wieloch T (1985). Hypoglycemia <strong>in</strong>duced neuronal damage prevented by an N-Methyl D-Aspartateantagonist. Science 230: 681–3.

2 Symptoms of <strong>Hypoglycaemia</strong>and Effects on MentalPerformance and EmotionsIan J. DearyINTRODUCTIONThis chapter describes the symptoms that are perceived dur<strong>in</strong>g acute hypoglycaemia, andthe changes <strong>in</strong> mental functions and emotions that occur dur<strong>in</strong>g this metabolic state.The most obvious benefit to a person of know<strong>in</strong>g about the symptoms of hypoglycaemiais the ability to recognise the onset of a hypoglycaemic episode as early as possible. This isof key importance <strong>in</strong> <strong>in</strong>form<strong>in</strong>g and educat<strong>in</strong>g people with diabetes. Moreover, if a personwith diabetes understands which mental functions are affected by hypoglycaemia he or shecan judge which activities may be most threatened <strong>in</strong> this state.SYMPTOMS OF HYPOGLYCAEMIAIdentify<strong>in</strong>g the SymptomsThe physiological responses to hypoglycaemia are described <strong>in</strong> Chapter 1. The responseto hypoglycaemia results <strong>in</strong> physical symptoms, which raises several questions (McAulayet al., 2001b). Can we compile a comprehensive list of symptoms of hypoglycaemia? Whichare the more common symptoms of hypoglycaemia? Are there early warn<strong>in</strong>g symptoms ofhypoglycaemia? Do people differ <strong>in</strong> how quickly and accurately they detect or recognisehypoglycaemia? Do people differ <strong>in</strong> the set of symptoms of hypoglycaemia they experience?How can <strong>in</strong>dividuals dist<strong>in</strong>guish the symptoms of hypoglycaemia from other bodily changes?The total symptom complexThe most basic question is: what symptoms do people report when they develophypoglycaemia?In humans the symptoms associated with hypoglycaemia were first recorded when <strong>in</strong>sul<strong>in</strong>became available for the treatment of diabetes (Fletcher and Campbell, 1922). A list ofcharacteristic symptoms was described (Table 2.1). It was noted: that some symptoms<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

Table 2.1 Common symptoms associated with hypoglycaemiaSymptoms associated withhypoglycaemia as derived frompopulation studies (afterHepburn, 1993)Percentage (m<strong>in</strong>imum–maximum) ofpeople report<strong>in</strong>g the given symptomas associated with hypoglycaemia(after Hepburn, 1993)Symptoms of hypoglycaemiaas noted by Fletcher andCampbell (1922)Symptoms (and percentage [to nearest5%]) of people endors<strong>in</strong>g symptoms asassociated with hypoglycaemia; after Coxet al., 1993a, Figure 2)Sweat<strong>in</strong>g 47–84 Sweat<strong>in</strong>g Sweat<strong>in</strong>g (80)Trembl<strong>in</strong>g 32–78 Tremulousness Trembl<strong>in</strong>g (65)Weakness 28–71 Weakness Fatigue/weak (70)Visual disturbance 24–60 Diplopia Blurred vision (20)Hunger 39–49 Excessive hunger Hunger (60)Pound<strong>in</strong>g heart 8–62 Change <strong>in</strong> pulse rate Pound<strong>in</strong>g heart (55)Difficulty with speak<strong>in</strong>g 7–41 Dysarthria, sensory and Slurred speech (40)motor aphasiaT<strong>in</strong>gl<strong>in</strong>g around the mouth 10–39 – Numb lips (50)Dizz<strong>in</strong>ess 11–41 Vertigo, fa<strong>in</strong>tness,Light-headed/dizzy (60)syncopeHeadache 24–36 Headache (30)Anxiety 10–44 Nervousness, anxiety,excitement,emotional upsetNervous/tense (65)Nausea 5–20 – –Difficulty concentrat<strong>in</strong>g 31–75 – Difficulty concentrat<strong>in</strong>g (80)Tiredness 38–46 –Drows<strong>in</strong>ess 16–33 – Drowsy–sleepy (40)Confusion 13–53 Confusion,disorientation,‘goneness’PallorIncoord<strong>in</strong>ation Uncoord<strong>in</strong>ated (75)Feel<strong>in</strong>g of heat orcoldEmotional <strong>in</strong>stabilityCold sweats (40)Slowed th<strong>in</strong>k<strong>in</strong>g (70)

SYMPTOMS OF HYPOGLYCAEMIA 27appeared before others dur<strong>in</strong>g hypoglycaemia; that the blood glucose level at which subjectsbecame aware of hypoglycaemia was characteristic for the <strong>in</strong>dividual; that there were large<strong>in</strong>dividual differences <strong>in</strong> the levels of blood glucose at which awareness of hypoglycaemiacommenced; and that the preced<strong>in</strong>g blood glucose concentration could affect the onset ofsymptoms.Lists of common symptoms of hypoglycaemia have been compiled from more recentresearch. Hepburn (1993) summarised eight population studies of the symptoms ofhypoglycaemia experienced by adults and children with <strong>in</strong>sul<strong>in</strong>-treated diabetes, and Coxet al. (1993a) also produced a list of symptoms (Table 2.1). It is evident that the three listsof symptoms do not differ greatly, and that Fletcher and Campbell’s early report (1922)had captured many of the symptoms found <strong>in</strong> subsequent, more structured <strong>in</strong>vestigations.However, their report omitted to mention some symptoms such as tiredness, drows<strong>in</strong>ess anddifficulty concentrat<strong>in</strong>g, though it did <strong>in</strong>clude others – such as pallor (a sign rather thana symptom), <strong>in</strong>coord<strong>in</strong>ation and feel<strong>in</strong>gs of temperature change – that are emphasised byother researchers as regularly perceived symptoms. Table 2.1 establishes a useful group ofsymptoms that are commonly reported <strong>in</strong> hypoglycaemia.The way we ask people to describe their symptoms of hypoglycaemia can alter what theytell us. The rank order of symptoms alters considerably if patients are asked to <strong>in</strong>dicate therelevance of each symptom rather than merely to identify that the symptom is associatedwith hypoglycaemia (Cox et al., 1993a). With regard to the criterion of relevance, the mostuseful symptoms <strong>in</strong> detect<strong>in</strong>g hypoglycaemia are as follows:• sweat<strong>in</strong>g;• trembl<strong>in</strong>g;• difficulty concentrat<strong>in</strong>g;• nervousness, tenseness;• light-headedness, dizz<strong>in</strong>ess.The <strong>in</strong>itial symptomsAnother important question is: which hypoglycaemic symptoms appear early dur<strong>in</strong>g anepisode? The symptoms of hypoglycaemia that appear first and offer early warn<strong>in</strong>g of theonset of hypoglycaemia (Hepburn, 1993) are as follows:• trembl<strong>in</strong>g;• sweat<strong>in</strong>g;• tiredness;• difficulty concentrat<strong>in</strong>g;• hunger.This knowledge is obviously useful for the prompt detection and treatment of hypoglycaemia.

28 SYMPTOMS OF HYPOGLYCAEMIAThe Validity of Symptom BeliefsThe <strong>in</strong>dividuality of hypoglycaemic symptom clustersA great deal of the <strong>in</strong>terest <strong>in</strong> symptoms of hypoglycaemia has been stimulated by concernsabout patient education. It is helpful to let patients know the range of symptoms found <strong>in</strong>hypoglycaemia and to <strong>in</strong>form them of the early warn<strong>in</strong>g symptoms reported by other peoplewith diabetes, much as we all tend to know the range of symptoms that are experiencedwith the common cold. Many surveys and laboratory studies have shown that people differconsiderably <strong>in</strong> the symptoms of hypoglycaemia they experience (Cox et al., 1993a). Inaddition to learn<strong>in</strong>g the generally reported symptoms, <strong>in</strong>dividuals with diabetes should beencouraged to learn about their own typical symptoms of hypoglycaemia.Correctly <strong>in</strong>terpret<strong>in</strong>g symptoms as represent<strong>in</strong>g hypoglycaemiaSymptoms of hypoglycaemia do not appear on top of the bodily equivalent of a blank sheetof paper. Sometimes we experience symptoms when there is noth<strong>in</strong>g wrong with our bodilyfunctions; on the other hand, sometimes we fail to notice any symptoms when the body ismalfunction<strong>in</strong>g. The alert person with diabetes who is on the lookout for hypoglycaemiamust make two sorts of decisions. First, symptoms of hypoglycaemia must be detected andcorrectly identified. It would be dangerous for a patient to ignore symptoms of hypoglycaemiabecause he or she thought they were related to someth<strong>in</strong>g else. Second, symptoms that havenoth<strong>in</strong>g to do with hypoglycaemia must be excluded. Unwanted hyperglycaemia could occurif patients treated themselves for hypoglycaemia when the symptoms had another cause.These two ma<strong>in</strong> types of error are a failure to treat hypoglycaemia when blood glucose islow, and <strong>in</strong>appropriate treatment when blood glucose is acceptable or high (Cox et al., 1985;1993a) (Figure 2.1).Blood glucose concentration – symptom report correlationsDo patients’ reports of symptoms of hypoglycaemia bear any relation to their concurrentblood glucose concentrations? After all, the pr<strong>in</strong>cipal aim of educat<strong>in</strong>g people to beFigure 2.1Consequences of correct and <strong>in</strong>correct perception of hypoglycaemic symptoms

SYMPTOMS OF HYPOGLYCAEMIA 29aware of symptoms of hypoglycaemia is that they become alert to low and potentiallydangerous levels of blood glucose. To answer the above question some researchers haveemployed a field study approach where people are <strong>in</strong>vited to list any symptoms they areexperienc<strong>in</strong>g, and then measure and record their blood glucose concentration several timesa day for weeks. As a result of these studies it is known that each person has somesymptoms that are most reliably associated with their actual blood glucose concentrations(Pennebaker et al., 1981). Some of the symptoms that people report dur<strong>in</strong>g hypoglycaemiaare more closely related to their actual blood glucose concentrations than are others, and ifwe can identify each <strong>in</strong>dividual’s most <strong>in</strong>formative symptoms, we can <strong>in</strong>struct people to paymore attention to them.The follow<strong>in</strong>g symptoms are most consistently associated with actual blood glucoseconcentrations (Pennebaker et al., 1981):• hunger (<strong>in</strong> 53% of people);• trembl<strong>in</strong>g (<strong>in</strong> 33%);• weakness (<strong>in</strong> 27%);• light-headedness (<strong>in</strong> 20%);• pound<strong>in</strong>g heart and fast heart rate (both 17%).The same symptoms are not <strong>in</strong>formative for everyone. There were 27% of people for whomweakness was significantly associated with hypoglycaemia, but there were 7% <strong>in</strong> whom itwas a good symptom of hyperglycaemia! Most people reported more than three symptomsthat were strongly associated with the measured blood glucose concentration.It is evident that an <strong>in</strong>dividual’s symptoms are idiosyncratic. If we can help a patientto identify the symptoms of hypoglycaemia peculiar to him or her, which relate toactual blood glucose concentrations, then, by attend<strong>in</strong>g to these symptoms, the personshould be especially accurate <strong>in</strong> recognis<strong>in</strong>g hypoglycaemia. People who have one ormore reliable symptom(s) of hypoglycaemia correctly recognise half of their episodes ofhypoglycaemia (def<strong>in</strong>ed as a blood glucose less than 3.95 mmol/l [Cox et al., 1993a]).Those who have four or more reliable symptoms recognise a blood glucose below3.95 mmol/l three-quarters of the time. The field study method has suggested that attentionto the follow<strong>in</strong>g symptoms was particularly useful <strong>in</strong> detect<strong>in</strong>g actual low blood glucoseconcentrations:• nervousness/tenseness;• slowed th<strong>in</strong>k<strong>in</strong>g;• trembl<strong>in</strong>g;• light-headedness/dizz<strong>in</strong>ess;• difficulty concentrat<strong>in</strong>g;• pound<strong>in</strong>g heart;• lack of co-ord<strong>in</strong>ation.

30 SYMPTOMS OF HYPOGLYCAEMIAClassify<strong>in</strong>g Symptoms of <strong>Hypoglycaemia</strong>Until now the symptoms of hypoglycaemia have been treated as a homogeneous whole. Canthese symptoms be divided <strong>in</strong>to different groups?<strong>Hypoglycaemia</strong> has effects on more than one part of the body, and the symptoms ofhypoglycaemia reflect this. First, the direct effects of a low blood glucose concentrationon the bra<strong>in</strong> – especially the cerebral cortex – cause neuroglycopenic symptoms. Second,autonomic symptoms result from activation of parts of the autonomic nervous system.F<strong>in</strong>ally, there may be some non-specific symptoms that are not directly generated by eitherof these two mechanisms. It is only recently that scientific <strong>in</strong>vestigations have taken placeto confirm the idea that these separable groups of hypoglycaemic symptoms exist.As suggested above, there are at least two dist<strong>in</strong>ct groups of symptoms dur<strong>in</strong>g the body’sreaction to hypoglycaemia (Hepburn et al., 1991):• Autonomic, with symptoms such as trembl<strong>in</strong>g, anxiety, sweat<strong>in</strong>g and warmness.• Neuroglycopenic, with symptoms such as drows<strong>in</strong>ess, confusion, tiredness, <strong>in</strong>ability toconcentrate and difficulty speak<strong>in</strong>g.This <strong>in</strong>formation can assist with patient education by supply<strong>in</strong>g evidence for separablegroups of symptoms, and by <strong>in</strong>dicat<strong>in</strong>g which symptoms belong to each group. Someneuroglycopenic symptoms, such as the <strong>in</strong>ability to concentrate, weakness and drows<strong>in</strong>ess,are among the earliest detectable symptoms, but patients tend to rely more on autonomicsymptoms when detect<strong>in</strong>g the onset of hypoglycaemia. Pay<strong>in</strong>g more attention to thepotentially useful, early neuroglycopenic symptoms could help with the early detection ofhypoglycaemia.Similar groups of symptoms of hypoglycaemia have been discovered by ask<strong>in</strong>g people torecall the symptoms they typically noticed dur<strong>in</strong>g hypoglycaemia. However, <strong>in</strong> addition tothe two groups described above, a general feel<strong>in</strong>g of malaise is added (Deary et al., 1993):• Autonomic: e.g. sweat<strong>in</strong>g, palpitations, shak<strong>in</strong>g and hunger.• Neuroglycopenic: e.g. confusion, drows<strong>in</strong>ess, odd behaviour, speech difficulty and<strong>in</strong>coord<strong>in</strong>ation.• General malaise: e.g. headache and nausea.These 11 symptoms are so reliably reported by people and so clearly separable <strong>in</strong>to thesethree groups, that they are used as the ‘Ed<strong>in</strong>burgh <strong>Hypoglycaemia</strong> Scale’ (Deary et al.,1993). Table 2.2 shows how different researchers have found similar groups of autonomicand hypoglycaemic symptoms.In addition to the above studies that used patients’ self-reported symptoms, physiologicalstudies have also confirmed that the symptoms of hypoglycaemia can be divided <strong>in</strong>to autonomicand neuroglycopenic groups. Symptoms such as sweat<strong>in</strong>g, hunger, pound<strong>in</strong>g heart,t<strong>in</strong>gl<strong>in</strong>g, nervousness and feel<strong>in</strong>g shaky/tremulous (autonomic symptoms) can be reduced oreven prevented by drugs that block neurotransmission with<strong>in</strong> the autonomic nervous system(Towler et al., 1993), confirm<strong>in</strong>g that these symptoms are caused by the autonomic responseto hypoglycaemia. Symptoms such as warmth, weakness, difficulty th<strong>in</strong>k<strong>in</strong>g/confusion,

SYMPTOMS OF HYPOGLYCAEMIA 31Table 2.2Different authors’ lists of autonomic and neuroglycopenic symptoms of hypoglycaemiaAutonomicNeuroglycopenicDeary et al.(1993)Towler et al.(1993)We<strong>in</strong>ger et al.(1995)Deary et al.(1993)Towler et al. (1993)Sweat<strong>in</strong>g Sweaty Sweat<strong>in</strong>g Confusion Difficulty th<strong>in</strong>k<strong>in</strong>g/confusedPalpitation Heart pound<strong>in</strong>g Pound<strong>in</strong>g heart, Drows<strong>in</strong>ess Tired/drowsyfast pulseShak<strong>in</strong>g Shaky/tremulous Trembl<strong>in</strong>g OddbehaviourHunger Hungry SpeechDifficulty speak<strong>in</strong>gdifficultyT<strong>in</strong>gl<strong>in</strong>gIncoord<strong>in</strong>ationNervous/anxious Tense WeakBreath<strong>in</strong>g hardWarmfeel<strong>in</strong>g tired/drowsy, feel<strong>in</strong>g fa<strong>in</strong>t, difficulty speak<strong>in</strong>g, dizz<strong>in</strong>ess and blurred vision (neuroglycopenicsymptoms) are not prevented by drugs that block the autonomic nervous system.Therefore, neuroglycopenic symptoms are not mediated via the autonomic nervous systemand are thought to be caused by the direct effect of glucose deprivation on the bra<strong>in</strong>. Thistype of research has also observed that people tend to rely on autonomic symptoms to detecthypoglycaemia, even when neuroglycopenic symptoms are just as prom<strong>in</strong>ent (Towler et al.,1993). Once more, this suggests that more emphasis should be placed on education of thepotential importance of neuroglycopenic symptoms for the early warn<strong>in</strong>g of hypoglycaemia.Symptoms might gather <strong>in</strong>to slightly different group<strong>in</strong>gs depend<strong>in</strong>g on the situation. Thesymptom group<strong>in</strong>gs <strong>in</strong> the Ed<strong>in</strong>burgh <strong>Hypoglycaemia</strong> Scale were developed from diabeticpatients’ retrospective reports. However, when people are asked to rate the same group ofsymptoms dur<strong>in</strong>g acute, experimentally-<strong>in</strong>duced moderate hypoglycaemia, a slightly differentpattern emerges (McCrimmon et al., 2003). In Table 2.3 there is an autonomic group<strong>in</strong>g, andthe s<strong>in</strong>gle neuroglycopenia group has divided <strong>in</strong>to two symptom groups: one with mostlycognitive symptoms and the other with more general symptoms. This division probably arosebecause the subjects <strong>in</strong> the studies used to form Table 2.3 were engaged <strong>in</strong> cognitive tasksTable 2.3 Symptom group<strong>in</strong>gs of the Ed<strong>in</strong>burgh <strong>Hypoglycaemia</strong> Scaledur<strong>in</strong>g experimentally-<strong>in</strong>duced hypoglycaemiaNeuroglycopenic symptomsCognitive dysfunction Neuroglycopenia Autonomic symptomsInability to concentrate Drows<strong>in</strong>ess Sweat<strong>in</strong>gBlurred vision Tiredness Trembl<strong>in</strong>gAnxiety Hunger WarmnessConfusionWeaknessDifficulty speak<strong>in</strong>gDouble vision

32 SYMPTOMS OF HYPOGLYCAEMIAdur<strong>in</strong>g the period of hypoglycaemia. Therefore, they would be especially aware of cognitiveshortcom<strong>in</strong>gs, mak<strong>in</strong>g this group of symptoms more prom<strong>in</strong>ent and coherent.Symptoms <strong>in</strong> Children and Older PeopleChildren often have difficulty <strong>in</strong> recognis<strong>in</strong>g symptoms of hypoglycaemia, and they showmarked variability <strong>in</strong> symptoms between episodes of hypoglycaemia (Macfarlane and Smith,1988). Trembl<strong>in</strong>g and sweat<strong>in</strong>g are often the first symptoms recognised by children. From<strong>in</strong>terviews with the parents of children (aged up to 16 years) with type 1 diabetes, andwith some of their children, more is known about the frequency of symptoms of hypoglycaemia<strong>in</strong> children (McCrimmon et al., 1995; Ross et al., 1998) (Table 2.4). The mostfrequently reported sign that parents observed was pallor (noted by 88%). The parentsTable 2.4 Symptoms of hypoglycaemia <strong>in</strong> children (derived from Ross et al., 1998)Frequency of rat<strong>in</strong>g (%)Symptom Parents’ reports Children’s reportsCorrelation betweenparents’ and children’s<strong>in</strong>tensity rat<strong>in</strong>gs aTearful 73 47 040 dHeadache 73 65 033 dIrritable 85 65 016Uncoord<strong>in</strong>ated 62 56 018Naughty 47 31 023 bWeak 79 83 021 bAggressive 75 62 026 cTrembl<strong>in</strong>g 79 88 025 bSleep<strong>in</strong>ess 63 69 027 cNightmares 33 19 033 dSweat<strong>in</strong>g 76 73 028 cSlurred speech 53 45 028 cBlurred vision 52 55 030 cTummy pa<strong>in</strong> 67 41 036 dFeel<strong>in</strong>g sick 63 53 032 cHungry 74 84 019Yawn<strong>in</strong>g 48 45 020 bOdd behaviour 65 50 022 bWarmness 57 68 013Restless 61 57 021 bDaydream<strong>in</strong>g 70 48 014Argumentative 64 50 021 bPound<strong>in</strong>g heart 21 44 002Confused 75 70 041 dT<strong>in</strong>gl<strong>in</strong>g lips 20 24 −001Dizz<strong>in</strong>ess 66 87 028 cTired 83 76 026 cFeel<strong>in</strong>g awful 92 79 020a Correlations: p of z (corrected for ties) 1.0 represents perfect agreement; 0 represents no agreement.b p

SYMPTOMS OF HYPOGLYCAEMIA 33frequently reported symptoms of behavioural disturbance such as irritability, argumentativenessand aggression. This latter group of symptoms is not prom<strong>in</strong>ent <strong>in</strong> adults, althoughthe Ed<strong>in</strong>burgh <strong>Hypoglycaemia</strong> Scale <strong>in</strong>cludes ‘odd behaviour’ as an adult neuroglycopenicsymptom. Others had previously noted the prom<strong>in</strong>ence of symptoms such as irritability,aggression and disobedience <strong>in</strong> the parents’ reports of their children’s symptoms of hypoglycaemia(Macfarlane and Smith, 1988; Macfarlane et al., 1989). Parents tend to under-reportthe subjective symptoms of hypoglycaemia, such as weakness and dizz<strong>in</strong>ess, but generallythere is good agreement between parents and their children about the most prom<strong>in</strong>entsymptoms of childhood hypoglycaemia (McCrimmon et al., 1995; Ross et al., 1998).Separate groups of autonomic and neuroglycopenic symptoms were not found <strong>in</strong> childrenwith type 1 diabetes (McCrimmon et al., 1995; Ross et al., 1998). These symptoms arereported together by children and are not dist<strong>in</strong>guished as separate groups, whereas the groupof symptoms related to behavioural disturbance is clearly reported as a dist<strong>in</strong>ct group. In aref<strong>in</strong>ement of the earlier study by McCrimmon et al. (1995), Ross et al. (1998) found thatparents could dist<strong>in</strong>guish between autonomic and neuroglycopenic symptoms.People with <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes report symptoms dur<strong>in</strong>g hypoglycaemia thatseparate <strong>in</strong>to autonomic and neuroglycopenic groups (Henderson et al., 2003). Elderlypatients with type 2 diabetes treated with <strong>in</strong>sul<strong>in</strong> commonly report neurological symptomsof hypoglycaemia which may be mis<strong>in</strong>terpreted as features of cerebrovascular disease, suchas transient ischaemic attacks (Jaap et al., 1998). The age-specific differences <strong>in</strong> the groupsof hypoglycaemic symptoms, classified us<strong>in</strong>g statistical techniques (Pr<strong>in</strong>cipal ComponentAnalysis), are shown <strong>in</strong> Table 2.5. Health professionals and carers who are <strong>in</strong>volved <strong>in</strong>the treatment and education of diabetic patients should be aware of which symptoms arecommon at either end of the age spectrum.From Symptom Perception to ActionPeople with diabetes are better at estimat<strong>in</strong>g their blood glucose <strong>in</strong> natural, everyday situations,as opposed to cl<strong>in</strong>ical laboratory sett<strong>in</strong>gs (Cox et al., 1985). In some ways this issurpris<strong>in</strong>g as natural hypoglycaemia often occurs at a time when it is unexpected. In this situation,attention toward symptoms will not be as actively directed toward detection as <strong>in</strong> thelaboratory sett<strong>in</strong>g where it is usually anticipated. Furthermore, hypoglycaemia <strong>in</strong> everydaylife occurs on the background of other bodily feel<strong>in</strong>gs and must be separated from othercauses of the same symptoms. For example, exercise and various acute illnesses can provokesweat<strong>in</strong>g <strong>in</strong> people with diabetes, <strong>in</strong>dependently of their association with hypoglycaemia.In a real-life situation a person must detect symptoms of hypoglycaemia and then <strong>in</strong>terpretthem. Failure to detect symptoms can lead to a failure to treat hypoglycaemia, but detectionTable 2.5 Classification of symptoms of hypoglycaemia us<strong>in</strong>g Pr<strong>in</strong>cipal Components Analysis <strong>in</strong>patients with <strong>in</strong>sul<strong>in</strong>-treated diabetes depend<strong>in</strong>g on age groupChildren (pre pubertal) Adults ElderlyAutonomic/neuroglycopenic Autonomic AutonomicNeuroglycopenicNeuroglycopenicBehavioural Non-specific malaise Neurological

34 SYMPTOMS OF HYPOGLYCAEMIAwithout the correct <strong>in</strong>terpretation of the cause of the symptoms is equally dangerous. Furthermore,it is obvious that someone who does <strong>in</strong>terpret symptoms correctly as be<strong>in</strong>g causedby hypoglycaemia, but who does not take action to treat the low blood glucose, will be atthe same risk. These and other steps toward the avoidance of severe hypoglycaemia demonstratethe key role of education about symptoms <strong>in</strong> people with diabetes (Gonder-Fredericket al., 1997). The psycho-educational programmes of Blood Glucose Awareness Tra<strong>in</strong><strong>in</strong>g(BGAT; Schach<strong>in</strong>ger et al., 2005) and Hypoglycemia Anticipation, Awareness and TreatmentTra<strong>in</strong><strong>in</strong>g (HAATT; Cox et al., 2004) have led to better recognition of hypoglycaemicstates and reduced frequency of hypoglycaemia.Symptom generationFigure 2.2 outl<strong>in</strong>es the stages that <strong>in</strong>tervene between low blood glucose occurr<strong>in</strong>g <strong>in</strong> an<strong>in</strong>dividual and the implementation of effective treatment (Gonder-Frederick et al., 1997). Itis <strong>in</strong>terest<strong>in</strong>g to note the importance of behavioural factors <strong>in</strong> generat<strong>in</strong>g states of low bloodglucose. Episodes of low blood glucose are most likely to come about because of changes<strong>in</strong> rout<strong>in</strong>e aspects of diabetes management, such as tak<strong>in</strong>g extra <strong>in</strong>sul<strong>in</strong>, eat<strong>in</strong>g less food ortak<strong>in</strong>g more exercise (Clarke et al., 1997). These factors predict more than 85% of episodesof hypoglycaemia <strong>in</strong> people with diabetes.In the presence of an <strong>in</strong>tact physiological response to low blood glucose, autonomic andneuroglycopenic symptoms, and symptoms of general malaise, are generated (Figure 2.2).The degree of hypoglycaemia, the person’s quality of glycaemic control and any recentepisodes of hypoglycaemia may all affect the magnitude of the body’s physiological response.Recent, preced<strong>in</strong>g hypoglycaemia can reduce the symptomatic and counterregulatoryhormonal responses to subsequent hypoglycaemia, result<strong>in</strong>g <strong>in</strong> a dim<strong>in</strong>ished awareness ofsymptoms. This effect of ‘antecedent’ hypoglycaemia is described <strong>in</strong> Chapter 7. Gender doesnot appear to <strong>in</strong>fluence the symptomatic response to hypoglycaemia (Geddes et al., 2006).At the second stage <strong>in</strong> Figure 2.2 comes the actual generation of physical symptoms.Among the variables that can <strong>in</strong>fluence this stage is the prior <strong>in</strong>gestion of caffe<strong>in</strong>e, whichhas been shown to enhance the <strong>in</strong>tensity of the autonomic and neuroglycopenic symptomsof experimentally <strong>in</strong>duced hypoglycaemia (Debrah et al., 1996) (see Chapter 5). Caffe<strong>in</strong>emay act by <strong>in</strong>creas<strong>in</strong>g the <strong>in</strong>tensity of symptoms of hypoglycaemia to perceptible levels,much as a magnify<strong>in</strong>g glass enables one to read otherwise too-small pr<strong>in</strong>t.Symptom detectionThe occurrence of physiological changes <strong>in</strong> the body does not guarantee that a personwill detect symptoms (Gonder-Frederick et al., 1997). If attention is directed to physicalchanges, people are more likely to detect symptoms than if their attention is held elsewhere.Everyone has had the experience of feel<strong>in</strong>g less discomfort, and be<strong>in</strong>g less likely to detecta physical symptom, when be<strong>in</strong>g distracted by someth<strong>in</strong>g divert<strong>in</strong>g. The personal relevanceof the symptom may affect detection; for example, a person with heart disease may be verylikely to detect palpitations (Cox et al., 1993a). The activity of the person at the time ofthe physiological change is obviously important. Hypoglycaemic symptoms will be moreobvious to the person engaged <strong>in</strong> active mental effort (McCrimmon et al., 2003), such assitt<strong>in</strong>g an exam<strong>in</strong>ation, than to the person relax<strong>in</strong>g and watch<strong>in</strong>g television. A doctor engaged<strong>in</strong> microsurgery may be very sensitive to the onset of tremor. In one laboratory study the

SYMPTOMS OF HYPOGLYCAEMIA 35Figure 2.2 A model for the occurrence and avoidance of severe hypoglycaemia (after Gonder-Frederick et al., 1997). Note on the right-hand side of each stage the factors that affect its occurrencepeople who had higher anxiety levels were better at detect<strong>in</strong>g symptoms of hypoglycaemia(Ryan et al., 2002).The autonomic symptoms of hypoglycaemia are often emphasised <strong>in</strong> the detection of hypoglycaemia.However, a strong case can be made for an equal emphasis on neuroglycopenicsymptoms (Gonder-Frederick et al., 1997; McCrimmon et al., 2003) because:• performance on mental tasks deteriorates dur<strong>in</strong>g hypoglycaemia, and subjective awarenessof this decrement beg<strong>in</strong>s at very mild levels of hypoglycaemia• the difference <strong>in</strong> glycaemic thresholds for the onset of autonomic and neuroglycopenicsymptoms is so small that it is unlikely to be detected when blood glucose decl<strong>in</strong>es rapidly

36 SYMPTOMS OF HYPOGLYCAEMIA• neuroglycopenic symptoms are as strongly related to actual blood glucose concentrationsas are autonomic symptoms (Cox et al., 1993a)• people with <strong>in</strong>sul<strong>in</strong>-treated diabetes cite autonomic and neuroglycopenic symptoms withequal frequency as the primary warn<strong>in</strong>g symptoms of hypoglycaemia (Hepburn et al., 1991).Low blood glucose detection – symptom <strong>in</strong>terpretationThe correct detection of a symptom of hypoglycaemia does not always lead to correct<strong>in</strong>terpretation (Figure 2.2). After correctly detect<strong>in</strong>g relevant symptoms, people fail to detectabout 26% of episodes of low blood glucose (Gonder-Frederick et al., 1997). There areseveral factors that could break a perfect relationship between detection and recognition,and some are discussed below. However, it should be appreciated that symptom detection(<strong>in</strong>ternal cues) is not mandatory for the detection of low blood glucose. Self-test<strong>in</strong>g of bloodglucose or the <strong>in</strong>formation of family members (external cues) can lead to the successfulrecognition of hypoglycaemia without the patient hav<strong>in</strong>g detected the episode by symptomaticperception (see Box 2.1).Correct knowledge about symptoms of hypoglycaemia is necessary for the detection oflow blood glucose. The lack of such knowledge among elderly people with diabetes <strong>in</strong>particular gives cause for concern (Mutch and D<strong>in</strong>gwall-Fordyce, 1985). Of 161 diabeticpeople between ages 60 and 87, all of whom were <strong>in</strong>ject<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> or tak<strong>in</strong>g a sulphonylurea,only 22% had ever been told the symptoms of hypoglycaemia and 9% knew no symptomsat all! The percentages of the <strong>in</strong>sul<strong>in</strong>-treated diabetic patients who knew that the follow<strong>in</strong>gsymptoms were associated with hypoglycaemia were as follows:• sweat<strong>in</strong>g (82%);• palpitations (62%);• confusion (53%);• hunger (51%);• <strong>in</strong>ability to concentrate (50%);• speech problems (41%);• sleep<strong>in</strong>ess (33%).Box 2.1Identification of hypoglycaemia• Internal cues – autonomic, neuroglycopenic and non-specific symptoms• External cues – relationship of <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jection to meals, exercise and experienceof self-management• Blood glucose monitor<strong>in</strong>g• Information from observers (e.g. relatives, friends, colleagues)

ACUTE HYPOGLYCAEMIA AND COGNITIVE FUNCTIONING 37In the midst of this ignorance, much hypoglycaemia may not be treated because of a lackof knowledge of the symptoms of hypoglycaemia which would aid their recognition.Most, if not all, of the symptoms of hypoglycaemia can be expla<strong>in</strong>ed by other physicalconditions. Therefore, correct symptom detection may be usurped by <strong>in</strong>correct attribution ofthe cause. For example, hav<strong>in</strong>g completed some strenuous activity, an athlete may attributethe symptoms of sweat<strong>in</strong>g and palpitations to physical exertion. An obvious problem <strong>in</strong>detect<strong>in</strong>g a low blood glucose is the fact that the organ responsible for the detectionand <strong>in</strong>terpretation of symptoms – the bra<strong>in</strong>, especially the cerebral cortex – is impaired.Thus, impaired concentration and lowered consciousness levels can beget even more severehypoglycaemia.Symptom scor<strong>in</strong>g systemsThe controversy about the effect of human <strong>in</strong>sul<strong>in</strong> on symptom awareness (Chapter 7)stimulated the development of scor<strong>in</strong>g systems for hypoglycaemia to allow comparativestudies between <strong>in</strong>sul<strong>in</strong> species. This produced scor<strong>in</strong>g systems such as the Ed<strong>in</strong>burgh<strong>Hypoglycaemia</strong> Scale (Deary et al., 1993), and any such system must be validated forresearch application. It is important to note that the nature and <strong>in</strong>tensity of <strong>in</strong>dividualsymptoms are as important as, if not more important than, the number of symptoms generatedby hypoglycaemia. The concepts <strong>in</strong>volved are discussed <strong>in</strong> detail by Hepburn (1993).More <strong>in</strong>formation on the symptoms of hypoglycaemia is provided by McAulay et al.(2001b).ACUTE HYPOGLYCAEMIA AND COGNITIVE FUNCTIONINGSymptoms are subjective reports of bodily sensations. With respect to hypoglycaemia someof these reports – especially neuroglycopenic symptoms – perta<strong>in</strong> to altered cognitive(mental ability) function<strong>in</strong>g. Do reports of ‘confusion’ and ‘difficulty th<strong>in</strong>k<strong>in</strong>g’ (Table 2.2)concur with objective mental test performance <strong>in</strong> hypoglycaemia? Before experimentalhypoglycaemia became an accepted <strong>in</strong>vestigative tool <strong>in</strong> diabetes, expert cl<strong>in</strong>ical observersnoted impairments of cognitive functions despite clear consciousness dur<strong>in</strong>g hypoglycaemia(Fletcher and Campbell, 1922; Wilder, 1943). Cognitive functions <strong>in</strong>clude the follow<strong>in</strong>g sortsof mental activity: orientation and attention, perception, memory (verbal and non-verbal),language, construction, reason<strong>in</strong>g, executive function and motor performance. Early studies(Russell and Rix-Trot, 1975) established that the follow<strong>in</strong>g abilities become disrupted belowblood glucose levels of about 3.0 mmol/l:• f<strong>in</strong>e motor co-ord<strong>in</strong>ation;• mental speed;• concentration;• some memory functions.

38 SYMPTOMS OF HYPOGLYCAEMIAThe hyper<strong>in</strong>sul<strong>in</strong>aemic glucose clamp technique allows more controlled experiments ofacute hypo- and hyperglycaemia. However, although this technique is used <strong>in</strong> most studiesof cognitive function <strong>in</strong> hypoglycaemia, it does not mimic the physiological or temporalcharacteristics of ‘natural’ or <strong>in</strong>tercurrent episodes of hypoglycaemia experienced by peoplewith type 1 diabetes. From laboratory experiments us<strong>in</strong>g the glucose clamp technique it wasfound that blood glucose concentrations between 3.1 and 3.4 mmol/l caused the follow<strong>in</strong>geffects (Holmes, 1987; Deary, 1993):• Slowed reaction times (this experiment <strong>in</strong>volves mak<strong>in</strong>g a fast response when a lightappears on a computer screen. <strong>Hypoglycaemia</strong> had more effect on reaction times whenthe reaction <strong>in</strong>volved mak<strong>in</strong>g a decision).• Slowed mental arithmetic.• Impaired verbal fluency (<strong>in</strong> this test one has to th<strong>in</strong>k of words beg<strong>in</strong>n<strong>in</strong>g with a givenletter, probably <strong>in</strong>volv<strong>in</strong>g the frontal lobes of the bra<strong>in</strong>).• Impaired performance <strong>in</strong> parts of the Stroop test (<strong>in</strong> this test one has to read aloud a seriesof <strong>in</strong>k colours when words are pr<strong>in</strong>ted <strong>in</strong> a different colour from that of the name, e.g. theword RED pr<strong>in</strong>ted <strong>in</strong> green <strong>in</strong>k).Some mental functions were spared dur<strong>in</strong>g hypoglycaemia, for example:• simple motor (like the speed of tapp<strong>in</strong>g) and sensory skills;• the speed of read<strong>in</strong>g words aloud.By 1993 over 16 studies had <strong>in</strong>vestigated cognitive functions dur<strong>in</strong>g acute and mild–moderate hypoglycaemia (Deary, 1993). The levels of blood glucose ranged from 2.0 to3.7 mmol/l. The way that hypoglycaemia was <strong>in</strong>duced varied among studies, as did themethods of blood sampl<strong>in</strong>g (e.g. arterialised or venous blood). Moreover, the ability levelsof the people <strong>in</strong> different studies varied, and there was much heterogeneity <strong>in</strong> the testbatteries used to assess mental performance. An authoritative statement as to the mentalfunctions disrupted dur<strong>in</strong>g hypoglycaemia is still not possible. However, <strong>in</strong> at least one ormore of the studies a number of tests were significantly impaired dur<strong>in</strong>g hypoglycaemia(Box 2.2).Few areas of mental function are preserved at normal levels dur<strong>in</strong>g acute hypoglycaemia.There is a general dampen<strong>in</strong>g of many abilities that <strong>in</strong>volve conscious mental effort. In theface of so many deleterious effects, what mental functions rema<strong>in</strong> <strong>in</strong>tact dur<strong>in</strong>g acute hypoglycaemia?At blood glucose concentrations similar to those <strong>in</strong>dicated above, the follow<strong>in</strong>gmental tests are not significantly impaired:• f<strong>in</strong>ger tapp<strong>in</strong>g;• forward digit span (repeat<strong>in</strong>g back a list of numbers <strong>in</strong> the same order);• simple reaction time;• elementary sensory process<strong>in</strong>g.

ACUTE HYPOGLYCAEMIA AND COGNITIVE FUNCTIONING 39Box 2.2Cognitive function tests impaired dur<strong>in</strong>g acute hypoglycaemia• Trail mak<strong>in</strong>g (<strong>in</strong>volv<strong>in</strong>g visual scann<strong>in</strong>g and mental flexibility)• Digit symbol (speed of replac<strong>in</strong>g a list of numbers with abstract codes)• Reaction time (especially <strong>in</strong>volv<strong>in</strong>g a decision)• Mental arithmetic• Verbal fluency• Stroop test• Grooved pegboard (a test of f<strong>in</strong>e manual dexterity)• Pursuit rotor (a test of eye–hand co-ord<strong>in</strong>ation)• Letter cancellation (strik<strong>in</strong>g out occurrences of a given letter <strong>in</strong> a page of letters)• Delayed verbal memory• Backward digit span (repeat<strong>in</strong>g back a list of numbers backwards)• Story recallThus tests which <strong>in</strong>volve speeded responses and which are more cognitively complex andattention-demand<strong>in</strong>g tend to show impairment dur<strong>in</strong>g hypoglycaemia (Deary, 1993). Hellerand Macdonald (1996) have concluded that:• even quite severe degrees of hypoglycaemia do not impair simple motor functions;• choice reaction time (where a mental decision of some k<strong>in</strong>d is needed before react<strong>in</strong>g toa stimulus) is affected at higher blood glucose concentrations more than simple reactiontime;• speed of respond<strong>in</strong>g is sometimes slowed <strong>in</strong> a task <strong>in</strong> which accuracy is preserved;• many aspects of mental performance become impaired when blood glucose falls belowabout 3.0 mmol/l;• there are important <strong>in</strong>dividual differences; some people’s mental performance is alreadyimpaired above a blood glucose of 3.0 mmol/l, whereas others cont<strong>in</strong>ue to function wellat lower levels;• the speed of response of the bra<strong>in</strong> <strong>in</strong> mak<strong>in</strong>g decisions slows down dur<strong>in</strong>g hypoglycaemia(Tallroth et al., 1990; Jones et al., 1990);• it can take as long as 40 to 90 m<strong>in</strong>utes after blood glucose returns to normal for the bra<strong>in</strong>to recover fully (Blackman et al., 1992; L<strong>in</strong>dgren et al., 1996).

40 SYMPTOMS OF HYPOGLYCAEMIAFigure 2.3 Cognitive effects of hypoglycaemia. After Deary (1998). This Article was published <strong>in</strong><strong>Diabetes</strong> Annual 11, SM Marshall, PD Home and RA Rizza (eds), 97–118, Copyright Elsevier 1998Determ<strong>in</strong><strong>in</strong>g whether mental performance is impaired at all dur<strong>in</strong>g mild to moderatehypoglycaemia, while a person is still fully conscious, is only the beg<strong>in</strong>n<strong>in</strong>g of this l<strong>in</strong>eof <strong>in</strong>vestigation. The next question to ask is whether some particular functions are moresusceptible and some less so? Figure 2.3 encapsulates this problem and illustrates three otherimportant questions about the cognitive effects of acute hypoglycaemia:1. What factors affect the degree of cognitive impairment dur<strong>in</strong>g hypoglycaemia, other thanthe level of blood glucose?2. Do impairments <strong>in</strong> laboratory cognitive tasks have a bear<strong>in</strong>g on mental performance <strong>in</strong>real life?3. Which basic bra<strong>in</strong> functions are disturbed dur<strong>in</strong>g acute hypoglycaemia?Influences on the Degree of and Threshold for Cognitive DysfunctionDur<strong>in</strong>g Acute <strong>Hypoglycaemia</strong>Although, on average, impairment of mental performance is worse dur<strong>in</strong>g hypoglycaemia,some people do not change or may even improve (Pramm<strong>in</strong>g et al., 1986; Hoffman et al.,

ACUTE HYPOGLYCAEMIA AND COGNITIVE FUNCTIONING 411989). It is not yet certa<strong>in</strong> whether such <strong>in</strong>dividual differences <strong>in</strong> responses are stable(Gonder-Frederick et al., 1994; Driesen et al., 1995). The follow<strong>in</strong>g factors might <strong>in</strong>creasea person’s degree of cognitive impairment dur<strong>in</strong>g acute hypoglycaemia:• male sex (Draelos et al., 1995; but this is disputed for people with type 2 diabetes byBremer et al., 2006);• impaired hypoglycaemia awareness (Gold et al., 1995b);• type 1 diabetes (Wirsen et al., 1992);• high IQ (Gold et al., 1995a).Does glycaemic control affect the cognitive impact of hypoglycaemia? People with type1 diabetes on <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy atta<strong>in</strong> glycaemic control that is nearer to normalthan most people treated with conventional <strong>in</strong>sul<strong>in</strong> treatment. As a result, the frequency ofsevere hypoglycaemia is <strong>in</strong>creased and is associated with a greater risk of impaired hypoglycaemiaawareness. Neuroendocr<strong>in</strong>e responses to hypoglycaemia are reduced <strong>in</strong> magnitudeand beg<strong>in</strong> at lower absolute blood glucose concentrations than <strong>in</strong> people with less strictglycaemic control. However, the hypoglycaemic threshold for cognitive dysfunction maynot change <strong>in</strong> a similar fashion. Diabetic patients on <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy reported autonomicand neuroglycopenic symptoms at blood glucose concentrations of about 2.4 and2.3 mmol/l respectively, whereas <strong>in</strong> those with less strict glycaemic control and <strong>in</strong> nondiabetic<strong>in</strong>dividuals, these symptoms commenced at between 2.8 and 3.0 mmol/l. However,<strong>in</strong> all three groups, the accuracy and speed <strong>in</strong> a reaction time test deteriorated significantly atblood glucose concentrations between 2.8 and 3.0 mmol/l (Amiel et al., 1991; Maran et al.,1995). Therefore, people with <strong>in</strong>sul<strong>in</strong>-treated diabetes who have strict glycaemic controlhave the misfortune that the deterioration <strong>in</strong> their mental performance beg<strong>in</strong>s before theonset of warn<strong>in</strong>g symptoms of hypoglycaemia. By contrast, neurophysiological responses(P300 event-related potentials), which have been l<strong>in</strong>ked to various measures of cognitivefunction, occur at lower blood glucose concentrations, suggest<strong>in</strong>g that cerebral adaptationhas occurred (Ziegler et al., 1992). The effects of quality of glycaemic control, antecedenthypoglycaemia and impaired hypoglycaemia awareness on the mental performance responsesto hypoglycaemia, and the relation of these responses to the perceptions of the symptoms ofhypoglycaemia, are important topics still under study (see Chapters 7 and 8). The <strong>in</strong>terrelationof these factors makes the field complex, and progress is further hampered by the lackof consensus agreement on a validated battery of cognitive tests for use <strong>in</strong> hypoglycaemia.Are the Cognitive Changes Dur<strong>in</strong>g Acute <strong>Hypoglycaemia</strong> Importantand Valid?Do the impairments of mental test performance actually have implications for real-lifefunctions? In addition, are the mental changes dur<strong>in</strong>g hypoglycaemia a result of impairments<strong>in</strong> basic bra<strong>in</strong> functions?One common, important and potentially dangerous area of real-life function<strong>in</strong>g is driv<strong>in</strong>g(see Chapter 14), which <strong>in</strong>volves many cognitive abilities <strong>in</strong>clud<strong>in</strong>g psychomotor controland divided attention. Cox and colleagues (1993b; 2000) employed a sophisticated driv<strong>in</strong>gsimulator and had people ‘drive’ on this dur<strong>in</strong>g controlled hypoglycaemia us<strong>in</strong>g a glucose

42 SYMPTOMS OF HYPOGLYCAEMIAclamp technique. With very mild hypoglycaemia (blood glucose below 3.8 mmol/l) thediabetic drivers committed significant driv<strong>in</strong>g errors, and dur<strong>in</strong>g hypoglycaemia the patientsoften drove very slowly, possibly us<strong>in</strong>g a compensatory mechanism to avoid errors. Despitethis, more global errors of driv<strong>in</strong>g were committed and about half of the participants,despite demonstrat<strong>in</strong>g a seriously impaired ability to drive, said they felt competent to driveirrespective of their low blood glucose! It cannot be stated with certa<strong>in</strong>ty that the f<strong>in</strong>d<strong>in</strong>gsobta<strong>in</strong>ed <strong>in</strong> a driv<strong>in</strong>g simulator will apply to real-life driv<strong>in</strong>g. However, studies that exam<strong>in</strong>ethe practical cognitive effects of hypoglycaemia are <strong>in</strong>valuable and more are required.Just as more studies that exam<strong>in</strong>e the practical cognitive aspects of hypoglycaemia wouldbe useful, so would more studies of the bra<strong>in</strong>’s process<strong>in</strong>g efficiency. Cognitive tests typically<strong>in</strong>volve a melange of <strong>in</strong>separable mental processes, and yet very specific aspects of the humanbra<strong>in</strong>’s activities can be measured <strong>in</strong> the cl<strong>in</strong>ical laboratory (Massaro, 1993). Studies of thecognitive effects of hypoglycaemia have thus begun to address the impairments to variouscognitive doma<strong>in</strong>s <strong>in</strong> more detail. Basic, specific aspects of visual and auditory process<strong>in</strong>ghave been exam<strong>in</strong>ed dur<strong>in</strong>g acute hypoglycaemia <strong>in</strong> non-diabetic humans. Standard tests ofvisual acuity – those that are measured by an optometrist – are not affected by hypoglycaemia,but other aspects of vision are affected (McCrimmon et al., 1996). These <strong>in</strong>clude:• contrast sensitivity (the ability to discrim<strong>in</strong>ate fa<strong>in</strong>t patterns);• <strong>in</strong>spection time (the ability to see what is <strong>in</strong> a pattern when it is shown for a very briefperiod of time);• visual change detection (the ability to spot a small, quick change <strong>in</strong> a pattern);• visual movement detection (the ability to spot brief movement <strong>in</strong> a pattern).This means that the ability to see the environment changes <strong>in</strong> important ways dur<strong>in</strong>g hypoglycaemia.Visual acuity is preserved, as tested by the ability to read black letters on a whitebackground. However, most visual activity is not like that; many of the th<strong>in</strong>gs we see happenquickly and <strong>in</strong> relatively poor light. When the level of contrast falls, or discrim<strong>in</strong>ationsmust be made under pressure of time, visual process<strong>in</strong>g is impaired dur<strong>in</strong>g hypoglycaemia.However, at about the same degree of hypoglycaemia, the ability to dist<strong>in</strong>guish one colourfrom another does not appear to be impaired (Hardy et al., 1995). Speed of auditoryprocess<strong>in</strong>g also appears to be impaired by hypoglycaemia, and the ability to discrim<strong>in</strong>atethe loudness of two tones is disrupted (McCrimmon et al., 1997; Strachan et al., 2003).This suggests that the ability to understand language may be compromised dur<strong>in</strong>g hypoglycaemia.However, despite there be<strong>in</strong>g disruption to central nervous system process<strong>in</strong>g dur<strong>in</strong>ghypoglycaemia, no disturbance has been detected <strong>in</strong> peripheral nerve conduction (Strachanet al., 2001).If basic <strong>in</strong>formation process<strong>in</strong>g provides a fundamental limitation to how well the bra<strong>in</strong>is operat<strong>in</strong>g, then at a higher level of function, attention is important <strong>in</strong> carry<strong>in</strong>g out anumber of cognitive functions. A detailed study of a number of different aspects of attentiondur<strong>in</strong>g hypoglycaemia found that the abilities to attend selectively and to switch attentionas necessary both deteriorated (McAulay et al., 2001a; 2005).In turn, attention is necessary <strong>in</strong> order to learn and form new memories. There hasnow been detailed study of the different aspects of memory dur<strong>in</strong>g acute, <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>ducedhypoglycaemia (Sommerfield et al., 2003a; 2003b; Deary et al., 2003). Most memory

ACUTE HYPOGLYCAEMIA AND EMOTIONS 43systems are disrupted dur<strong>in</strong>g hypoglycaemia. However, some are especially badly affected:long-term memory, which is the ability to reta<strong>in</strong> new <strong>in</strong>formation after many m<strong>in</strong>utes andmuch distraction; work<strong>in</strong>g memory, which is the ability to reta<strong>in</strong> and manipulate <strong>in</strong>formationat the same time; and prospective memory, which is the ability to remember to do th<strong>in</strong>gs (as<strong>in</strong> a shopp<strong>in</strong>g list) (Warren et al., 2007). Indeed, the ability to perform one tricky work<strong>in</strong>gmemory task was obliterated dur<strong>in</strong>g hypoglycaemia (Deary et al., 2003). That is, no matterhow good the person was at perform<strong>in</strong>g the task dur<strong>in</strong>g euglycaemia, the same task couldnot be done dur<strong>in</strong>g moderate hypoglycaemia.Further <strong>in</strong>formation on hypoglycaemia and cognitive function is available <strong>in</strong> a review byWarren and Frier (2005).ACUTE HYPOGLYCAEMIA AND EMOTIONSMood change is part of the experience of hypoglycaemia. Moods are emotion-like experiencesthat are quite general rather than applied to specific situations. Psychologists recognisethree basic moods:• energetic arousal (a tendency to feel lively and active rather than tired and sluggish);• tense arousal (a tendency to feel anxious and nervous versus relaxed and calm);• hedonic tone (a tendency to feel happy versus sad).When people are asked to rate their mood states dur<strong>in</strong>g hypoglycaemia <strong>in</strong>duced <strong>in</strong> thelaboratory, changes occur <strong>in</strong> all of these basic mood states. People feel less energetic, moretense and less happy (Gold et al., 1995c; McCrimmon et al., 1999a; Hermanns et al., 2003).Dur<strong>in</strong>g hypoglycaemia the emotional arousal <strong>in</strong> response to stimuli becomes more <strong>in</strong>tense(Hermanns et al., 2003). In addition, some people become more irritable and have angryfeel<strong>in</strong>gs dur<strong>in</strong>g hypoglycaemia (Merbis et al., 1996; McCrimmon et al., 1999b). The feel<strong>in</strong>gof low energy takes over half an hour to be restored to normal levels, whereas the feel<strong>in</strong>gs oftenseness and unhapp<strong>in</strong>ess disappear when blood glucose returns to normal. The prolongedfeel<strong>in</strong>g of low energy after hypoglycaemia may affect work performance, so that when hypoglycaemiahas been treated, an immediate return to the normal state should not be expected.In addition to some people experienc<strong>in</strong>g a low, tense, washed-out, angry mood state,hypoglycaemia alters the way some people look at their life problems. When junior doctorswere asked to assess their career prospects dur<strong>in</strong>g controlled hypoglycaemia, they weremore pessimistic (McCrimmon et al., 1995) and if a general state of pessimism is commondur<strong>in</strong>g hypoglycaemia, it would be a poor state from which to make personal decisions. It ispossible that the change <strong>in</strong> mood states dur<strong>in</strong>g hypoglycaemia is one of the causes of adultsadmitt<strong>in</strong>g to more ‘odd behaviour’ (Deary et al., 1993). Altered mood may also account<strong>in</strong> part for symptoms of behavioural disturbance that are so prom<strong>in</strong>ent <strong>in</strong> the responses tohypoglycaemia of children with diabetes (McCrimmon et al., 1995).In addition to emotional responses as a result of hypoglycaemia, some people haveemotional responses <strong>in</strong> anticipation of hypoglycaemia. In Ed<strong>in</strong>burgh one young man with<strong>in</strong>sul<strong>in</strong>-treated diabetes developed a phobic anxiety state; his phobia related to becom<strong>in</strong>gcomatose as a result of hypoglycaemia (Gold et al., 1997a). Such a case is exceptional,but many people with diabetes are frightened of hypoglycaemia (see Chapter 14). The

44 SYMPTOMS OF HYPOGLYCAEMIA<strong>Hypoglycaemia</strong> Fear Survey (HFS), which comes <strong>in</strong> two parts, measures this tendency(Cox et al., 1987). The first part asks people several questions concern<strong>in</strong>g how much theyworry about hypoglycaemia (e.g. ‘Do you worry if you have no one around you dur<strong>in</strong>g a[hypoglycaemic] reaction?’). The second part asks several questions about what people doto avoid hypoglycaemia (e.g. ‘Do you eat large snacks at bedtime?’). People with greaterfear of hypoglycaemia (Polonsky et al., 1992; Hepburn et al., 1994)• have more anxious personalities <strong>in</strong> general;• are more likely to confuse symptoms of anxiety for those of hypoglycaemia;• report hav<strong>in</strong>g had more episodes of hypoglycaemia.It is not yet known whether people who experience more hypoglycaemia become worriersabout it, or whether people who are worriers <strong>in</strong> general just worry more about hypoglycaemiaas well. Perhaps both are true. However, it does seem likely that the experience of more severehypoglycaemia <strong>in</strong> the past and the development of impaired awareness of hypoglycaemialead to <strong>in</strong>creased worry about subsequent hypoglycaemia (Gold et al., 1997b).CONCLUSIONS• Because people with diabetes are closely <strong>in</strong>volved <strong>in</strong> their own treatment it is importantthat they and their educators know about the ma<strong>in</strong> side-effects and sequelae of the disorderand its treatments.• Accurate knowledge of the symptoms of hypoglycaemia may be used to avoid the dangersof hypoglycaemia.• The progressively more serious impairment <strong>in</strong> cognitive function that occurs as bloodglucose decl<strong>in</strong>es provides knowledge about the bra<strong>in</strong>’s compromised state dur<strong>in</strong>g hypoglycaemia:basic functions such as visual process<strong>in</strong>g deteriorate and driv<strong>in</strong>g becomesdangerously error-prone. Performance on a host of mental tests becomes worse dur<strong>in</strong>ghypoglycaemia.• Some of the neuroglycopenic symptoms of hypoglycaemia are thought to be subjectiveimpressions of impaired cognitive function: these impressions are fully supported by theresults of objective cognitive test<strong>in</strong>g.• Moderate hypoglycaemia may <strong>in</strong>duce a state of anxious tension, unhapp<strong>in</strong>ess and lowenergy, and even irritability and anger. Thus hypoglycaemia importantly touches theemotions as well as <strong>in</strong>duc<strong>in</strong>g bodily symptoms and affect<strong>in</strong>g mental performance.REFERENCESAmiel SA, Pott<strong>in</strong>ger RC, Archibald HR, Chusney G, Cunnah DTF, Prior PF, Gale EAM (1991). Effectof antecedent glucose control on cerebral function dur<strong>in</strong>g hypoglycemia. <strong>Diabetes</strong> Care 14: 109–18.Blackman JD, Towle VL, Sturis J, Lewis GF, Spire J-P, Polonsky KS (1992). Hypoglycemic thresholdsfor cognitive dysfunction <strong>in</strong> IDDM. <strong>Diabetes</strong> 41: 392–9.

REFERENCES 45Bremer JP, Baron M, Peters H, Oltmanns KM, Kern W, Fehm HL et al. (2006). Hormonal, subjectiveand neurocognitive responses to brief hypoglycemia <strong>in</strong> postmenopausal women and age-matchedmen with type 2 diabetes mellitus. Metabolism 55: 331–8.Clarke WL, Cox DJ, Gonder-Frederick LA, Julian D, Schlundt D, Polonsky W (1997). The relationshipbetween nonrout<strong>in</strong>e use of <strong>in</strong>sul<strong>in</strong>, food, and exercise and the occurrence of hypoglycemia <strong>in</strong>adults with IDDM and vary<strong>in</strong>g degrees of hypoglycemia awareness and metabolic control. <strong>Diabetes</strong>Education 23: 55–8.Cox DJ, Clarke WL, Gonder-Frederick L, Pohl S, Hoover C, Snyder A et al. (1985). Accuracy ofperceiv<strong>in</strong>g blood glucose <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 8: 529–36.Cox DJ, Irv<strong>in</strong>e A, Gonder-Frederick L, Nowacek G, Butterfield J (1987). Fear of hypoglycemia:quantification, validation and utilization. <strong>Diabetes</strong> Care 10: 617–21.Cox DJ, Gonder-Frederick L, Antoun B, Cryer PE, Clarke WL (1993a). Perceived symptoms <strong>in</strong> therecognition of hypoglycemia. <strong>Diabetes</strong> Care 16: 519–27.Cox D, Gonder-Frederick L, Clarke W (1993b). Driv<strong>in</strong>g decrements <strong>in</strong> type 1 diabetes dur<strong>in</strong>g moderatehypoglycemia. <strong>Diabetes</strong> 42: 239–43.Cox DJ, Gonder-Frederick LA, Kovatchev BP, Julian DM, Clarke WL (2000). Progressive hypoglycemia’simpact on driv<strong>in</strong>g simulation performance. Occurrence, awareness and correction.<strong>Diabetes</strong> Care 23: 163–70.Cox DL, Kovatchev B, Koev D, Koeva L, Dachev S, Tcharaktchiev D et al. (2004). Hypoglycemiaanticipation, awareness and treatment tra<strong>in</strong><strong>in</strong>g (HAAT) reduces occurrence of severe hypoglycemiaamong adults with type 1 diabetes. International Journal of Behavioral Medic<strong>in</strong>e 11: 212–18.Deary IJ (1993). Effects of hypoglycaemia on cognitive function. In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>:Cl<strong>in</strong>ical and Physiological Aspects. Frier BM and Fisher M, eds. Edward Arnold, London: 80–92.Deary IJ, Hepburn DA, MacLeod KM, Frier M (1993). Partition<strong>in</strong>g the symptoms of hypoglycaemiaus<strong>in</strong>g multi-sample confirmatory factor analysis. Diabetologia 36: 771–7.Deary IJ (1998). The effects of diabetes on cognitive function. In: <strong>Diabetes</strong> Annual 11. Marshall SM,Home PD and Rizza RA eds. Elsevier, London: 97–118.Deary IJ, Sommerfield AJ, McAulay V, Frier BM (2003). Moderate hypoglycaemia obliterates work<strong>in</strong>gmemory <strong>in</strong> humans with and without <strong>in</strong>sul<strong>in</strong> treated diabetes. Journal of Neurology, Neurosurgeryand Psychiatry 74: 277–82.Debrah T, Sherw<strong>in</strong> RS, Murphy J, Kerr D (1996). Effect of caffe<strong>in</strong>e on recognition of and physiologicalresponses to hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 347: 19–24.Draelos MT, Jacobson AM, We<strong>in</strong>ger K, Widom B, Ryan CM, F<strong>in</strong>kelste<strong>in</strong> DM, Simonson DC (1995).Cognitive function <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus dur<strong>in</strong>g hyperglycemia andhypoglycemia. American Journal of Medic<strong>in</strong>e 98: 135–44.Driesen NR, Cox DJ, Gonder-Frederick L, Clarke W (1995). Reaction time impairment <strong>in</strong> <strong>in</strong>sul<strong>in</strong>dependentdiabetes: task complexity, blood glucose levels, and <strong>in</strong>dividual differences. Neuropsychology9: 246–54.Fletcher AA, Campbell WR (1922). The blood sugar follow<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> adm<strong>in</strong>istration and the symptomcomplex-hypoglycemia. Journal of Metabolic Research 2: 637–49.Geddes J, Warren, RE, Sommerfield AJ, McAulay V, Strachan MWJ, Allen KV et al. (2006). Absenceof sexual dimorphism <strong>in</strong> the symptomatic responses to hypoglycemia <strong>in</strong> adults with and withouttype 1 diabetes. <strong>Diabetes</strong> Care 29: 1667–9.Gold AE, Deary IJ, MacLeod KM, Frier BM (1995a). The effect of IQ level on the degree of cognitivedeterioration experienced dur<strong>in</strong>g acute hypoglycemia <strong>in</strong> normal humans. Intelligence 20: 267–90.Gold AE, MacLeod KM, Deary IJ, Frier BM (1995b). Hypoglycemia-<strong>in</strong>duced cognitive dysfunction<strong>in</strong> diabetes mellitus: effect of hypoglycemia unawareness. Physiology and Behavior 58: 501–11.Gold AE, MacLeod KM, Frier BM, Deary IJ (1995c). Changes <strong>in</strong> mood dur<strong>in</strong>g acute hypoglycemia<strong>in</strong> healthy subjects. Journal of Personality and Social Psychology 68: 498–504.Gold AE, Deary IJ, Frier BM (1997a). <strong>Hypoglycaemia</strong> and non-cognitive aspects of psychologicalfunction <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent (type 1) diabetes mellitus (IDDM). Diabetic Medic<strong>in</strong>e 14: 111–18.

46 SYMPTOMS OF HYPOGLYCAEMIAGold AE, Frier BM, MacLeod KM, Deary IJ (1997b). A structural equation model for predictors ofsevere hypoglycaemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Diabetic Medic<strong>in</strong>e 14:309–15.Gonder-Frederick L, Cox D, Driesen NR, Ryan CM, Clarke W (1994). Individual differences <strong>in</strong>neurobehavioral disruption dur<strong>in</strong>g mild and moderate hypoglycemia <strong>in</strong> adults with IDDM. <strong>Diabetes</strong>43: 1407–12.Gonder-Frederick L, Cox D, Kovatchev B, Schlundt D, Clarke W (1997). A biopsychobehavioralmodel of risk of severe hypoglycemia. <strong>Diabetes</strong> Care 20: 161–9.Hardy KJ, Scase MO, Foster DH, Scarpello JH (1995). Effect of short term changes <strong>in</strong> blood glucoseon visual pathway function <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dependent diabetes. British Journal of Ophthalmology 79:38–41.Heller SR, Macdonald IA (1996). The measurement of cognitive function dur<strong>in</strong>g acute hypoglycaemia:experimental limitations and their effects on the study of hypoglycaemia unawareness. DiabeticMedic<strong>in</strong>e 13: 607–15.Henderson JN, Allen KV, Deary IJ, Frier BM (2003). <strong>Hypoglycaemia</strong> <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes:frequency, symptoms and impaired awareness. Diabetic Medic<strong>in</strong>e 20: 1016–21.Hepburn DA, Deary IJ, Frier BM, Patrick AW, Qu<strong>in</strong>n JD, Fisher M (1991). Symptoms of acute <strong>in</strong>sul<strong>in</strong><strong>in</strong>ducedhypoglycemia <strong>in</strong> humans with and without IDDM. Factor-analysis approach. <strong>Diabetes</strong> Care14: 949–57.Hepburn DA (1993). Symptoms of hypoglycaemia. In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>: Cl<strong>in</strong>ical andPhysiological Aspects. Frier BM and Fisher M eds. Edward Arnold, London: 93–103.Hepburn DA, Deary IJ, MacLeod KM, Frier BM (1994). Structural equation model<strong>in</strong>g of symptoms,awareness and fear of hypoglycemia, and personality <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-treated diabetes.<strong>Diabetes</strong> Care 17: 1273–80.Hermanns N, Kubiak T, Kulzer B, Haak T (2003). Emotional changes dur<strong>in</strong>g experimentally-<strong>in</strong>ducedhypoglycaemia <strong>in</strong> type 1 diabetes. Biological Psychology 63: 15–44.Hoffman RG, Speelman DJ, H<strong>in</strong>nen DA, Conley KL, Guthrie RA, Knapp RK (1989). Changes <strong>in</strong>cortical function<strong>in</strong>g with acute hypoglycemia and hyperglycemia <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care12: 193–7.Holmes CS (1987). Metabolic control and auditory <strong>in</strong>formation process<strong>in</strong>g at altered glucose levels <strong>in</strong><strong>in</strong>sul<strong>in</strong> dependent diabetes. Bra<strong>in</strong> and Cognition 6: 161–74.Jaap AJ, Jones GC, McCrimmon RJ, Deary IJ, Frier BM (1998). Perceived symptoms of hypoglycaemia<strong>in</strong> elderly type 2 diabetic patients treated with <strong>in</strong>sul<strong>in</strong>. Diabetic Medic<strong>in</strong>e 15: 398–401.Jones TW, McCarthy G, Tamborlane WV, Caprio S, Roessler E, Kraemer D et al. (1990). Mildhypoglycemia and impairment of bra<strong>in</strong>stem and cortical evoked potentials <strong>in</strong> healthy subjects.<strong>Diabetes</strong> 39: 1550–5.L<strong>in</strong>dgren M, Eckert B, Stenberg G, Agardh C-D (1996). Restitution of neurophysiological functions,performance, and subjective symptoms after moderate <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong>non-diabetic men. Diabetic Medic<strong>in</strong>e 13: 218–25.Maran A, Lomas J, Macdonald IA, Amiel SA (1995). Lack of preservation of higher bra<strong>in</strong>function dur<strong>in</strong>g hypoglycaemia <strong>in</strong> patients with <strong>in</strong>tensively-treated IDDM. Diabetologia 38:1412–18.Massaro DW (1993). Information process<strong>in</strong>g models: microscopes of the m<strong>in</strong>d. Annual Review ofPsychology 44: 383–425.Macfarlane PI, Smith CS (1988). Perceptions of hypoglycaemia <strong>in</strong> childhood diabetes mellitus: aquestionnaire study. Practical <strong>Diabetes</strong> 5: 56–8.Macfarlane PI, Walters M, Stutchfield P, Smith CS (1989). A prospective study of symptomatichypoglycaemia <strong>in</strong> childhood diabetes. Diabetic Medic<strong>in</strong>e 6: 627–30.McAulay V, Deary IJ, Ferguson SC, Frier BM (2001a). Acute hypoglycemia <strong>in</strong> humanscauses attentional dysfunction while nonverbal <strong>in</strong>telligence is preserved. <strong>Diabetes</strong> Care 24:1745–50.

REFERENCES 47McAulay V, Deary IJ, Frier BM (2001b). Symptoms of hypoglycaemia <strong>in</strong> people with diabetes.Diabetic Medic<strong>in</strong>e 18: 690–705.McAulay V, Deary IJ, Sommerfield AJ, Frier BM (2005). Attentional function<strong>in</strong>g is impaired dur<strong>in</strong>gacute hypoglycaemia <strong>in</strong> people with type 1 diabetes. Diabetic Medic<strong>in</strong>e 23: 26–31.McCrimmon RJ, Gold AE, Deary IJ, Kelnar CJH, Frier BM (1995). Symptoms of hypoglycemia <strong>in</strong>children with IDDM. <strong>Diabetes</strong> Care 18: 858–61.McCrimmon RJ, Deary IJ, Huntly BJP, MacLeod KJ, Frier BM (1996). Visual <strong>in</strong>formation process<strong>in</strong>gdur<strong>in</strong>g controlled hypoglycaemia <strong>in</strong> humans. Bra<strong>in</strong> 119: 1277–87.McCrimmon RJ, Deary IJ, Frier BM (1997). Auditory <strong>in</strong>formation process<strong>in</strong>g dur<strong>in</strong>g acute <strong>in</strong>sul<strong>in</strong><strong>in</strong>ducedhypoglycaemia <strong>in</strong> non-diabetic human subjects. Neuropsychologia 35: 1547–53.McCrimmon RJ, Deary IJ, Frier BM (1999a). Appraisal of mood and personality dur<strong>in</strong>g hypoglycaemia.Physiology and Behavior 67: 27–33.McCrimmon RJ, Ew<strong>in</strong>g FME, Frier BM, Deary IJ (1999b). Anger-state dur<strong>in</strong>g acute <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>ducedhypoglycaemia. Physiology and Behavior 67: 35–9.McCrimmon RJ, Deary IJ, Gold AE, Hepburn DA, MacLeod KM, Ew<strong>in</strong>g FME, Frier BM (2003).Symptoms reported dur<strong>in</strong>g experimental hypoglycaemia: effect of method of <strong>in</strong>duction of hypoglycaemiaand of diabetes per se. Diabetic Medic<strong>in</strong>e 20: 507–9.Merbis MAE, Snoek FJ, Kanc K, He<strong>in</strong>e RJ (1996). <strong>Hypoglycaemia</strong> <strong>in</strong>duces emotional disruption.Patient Education and Counsel<strong>in</strong>g 29: 117–22.Mutch WJ, D<strong>in</strong>gwall-Fordyce I (1985). Is it a hypo? Knowledge of symptoms of hypoglycaemia <strong>in</strong>elderly diabetic patients. Diabetic Medic<strong>in</strong>e 2: 54–6.Pennebaker JW, Cox DJ, Gonder-Frederick L, Wunsch MG, Evans WS, Pohl S (1981). Physicalsymptoms related to blood glucose <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetics. Psychosomatic Medic<strong>in</strong>e 43:489–500.Polonsky WH, Davis CL, Jacobson AM, Anderson BJ (1992). Correlates of hypoglycemic fear <strong>in</strong> type1 and type 2 diabetes mellitus. Health Psychology 11: 199–202.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Theilgaard A, P<strong>in</strong>ner EM, B<strong>in</strong>der C (1986). Cognitive function dur<strong>in</strong>ghypoglycaemia <strong>in</strong> type 1 diabetes mellitus. British Medical Journal 292: 647–50.Ross LA, McCrimmon RJ, Frier BM, Kelnar CJH, Deary IJ (1998). Hypoglycaemic symptoms reportedby children with type 1 diabetes mellitus and by their parents. Diabetic Medic<strong>in</strong>e 15: 836–43.Russell PN, Rix-Trot HM (1975). An exploratory study of some behavioural consequences of <strong>in</strong>sul<strong>in</strong><strong>in</strong>ducedhypoglycaemia. New Zealand Medical Journal 81: 337–40.Ryan CM, Dulay D, Suprasongs<strong>in</strong> C, Becker DJ (2002). Detection of symptoms by adolescents andyoung adults with type 1 diabetes dur<strong>in</strong>g experimental <strong>in</strong>duction of mild hypoglycemia: role ofhormonal and psychological variables. <strong>Diabetes</strong> Care 25: 852–8.Schach<strong>in</strong>ger H, Hegar K, Hermanns N, Straumann M, Keller U, Fehm-Wolfsdorf G et al. (2005).Randomized controlled cl<strong>in</strong>ical trial of Blood Glucose Awareness Tra<strong>in</strong><strong>in</strong>g (BGAT III) <strong>in</strong> Switzerlandand Germany. Journal of Behavioral Medic<strong>in</strong>e 28: 587–94.Sommerfield AJ, Deary IJ, McAulay V, Frier BM (2003a). Moderate hypoglycemia impairs multiplememory functions <strong>in</strong> healthy adults. Neuropsychology 17: 125–32.Sommerfield AJ, Deary IJ, McAulay V, Frier BM (2003b). Short-term, delayed, and work<strong>in</strong>g memoryare impaired dur<strong>in</strong>g hypoglycemia <strong>in</strong> <strong>in</strong>dividuals with type 1 diabetes. <strong>Diabetes</strong> Care 26: 390–6.Strachan MWJ, Deary IJ, Ew<strong>in</strong>g FME, Ferguson SC, Young MJ, Frier BM (2001). Acute hypoglycemiaimpairs the function<strong>in</strong>g of the central but not peripheral nervous system. Physiology & Behavior72: 83–92.Strachan MWJ, Ew<strong>in</strong>g FME, Frier BM, McCrimmon RJ, Deary IJ (2003). Effects of acute hypoglycaemiaon auditory <strong>in</strong>formation process<strong>in</strong>g <strong>in</strong> adults with type 1 diabetes. Diabetologia 46:97–105.Tallroth G, L<strong>in</strong>dgren M, Stenberg G, Rosen I, Agardh C-D (1990). Neurophysiological changes dur<strong>in</strong>g<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia and <strong>in</strong> the recovery period follow<strong>in</strong>g glucose <strong>in</strong>fusion <strong>in</strong> type 1(<strong>in</strong>sul<strong>in</strong>-dependent) diabetes mellitus and <strong>in</strong> normal man. Diabetologia 33: 319–23.

48 SYMPTOMS OF HYPOGLYCAEMIATowler DA, Havl<strong>in</strong> CE, Craft S, Cryer P (1993). Mechanism of awareness of hypoglycemia: perceptionof neurogenic (predom<strong>in</strong>antly chol<strong>in</strong>ergic) rather than neuroglycopenic symptoms. <strong>Diabetes</strong> 42:1791–8.Warren RE, Frier BM (2005). <strong>Hypoglycaemia</strong> and cognitive function. <strong>Diabetes</strong>, Obesity andMetabolism 7: 493–503.Warren RE, Zammitt NN, Deary IJ, Frier BM (2007). The effects of acute hypoglycaemia on memoryacquisition and recall and prospective memory <strong>in</strong> type 1 diabetes. Diabetologia 50: 178–85.We<strong>in</strong>ger K, Jacobson AM, Draelos MT, F<strong>in</strong>kelste<strong>in</strong> DM, Simonson DC (1995). Blood glucose estimationand symptoms dur<strong>in</strong>g hyperglycemia and hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. American Journal of Medic<strong>in</strong>e 98: 22–31.Wilder J (1943). Psychological problems <strong>in</strong> hypoglycemia. American Journal of Digestive Diseases10: 428–35.Wirsen A, Tallroth G, L<strong>in</strong>dgren M, Agardh C-D (1992). Neuropsychological performance differsbetween type 1 diabetic and normal men dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia. Diabetic Medic<strong>in</strong>e9: 156–65.Ziegler D, Hub<strong>in</strong>ger A, Muhlen H, Gries FA (1992). Effects of previous glycaemic control on theonset and magnitude of cognitive dysfunction dur<strong>in</strong>g hypoglycaemia <strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong>-dependent)diabetic patients. Diabetologia 35: 828–34.

3 Frequency, Causes and RiskFactors for <strong>Hypoglycaemia</strong> <strong>in</strong>Type 1 <strong>Diabetes</strong>Mark W.J. StrachanINTRODUCTION<strong>Hypoglycaemia</strong> was first described <strong>in</strong> humans <strong>in</strong> the early years of the 20th century, but didnot become firmly established as a pathophysiological entity until the discovery of <strong>in</strong>sul<strong>in</strong><strong>in</strong> 1922. Despite the substantial advances <strong>in</strong> <strong>in</strong>sul<strong>in</strong> therapy and blood glucose monitor<strong>in</strong>gthat have occurred <strong>in</strong> the subsequent 80 years, hypoglycaemia rema<strong>in</strong>s the most commoncomplication of type 1 diabetes (The <strong>Diabetes</strong> Control and Complications Trial ResearchGroup, 1993) and generates as much anxiety <strong>in</strong> patients as the threat of advanced diabeticcomplications, such as bl<strong>in</strong>dness or renal failure (Pramm<strong>in</strong>g et al., 1991). Few people withtype 1 diabetes escape <strong>in</strong>termittent exposure and, as a result, hypoglycaemia is the pr<strong>in</strong>cipallimit<strong>in</strong>g factor <strong>in</strong> achiev<strong>in</strong>g good glycaemic control (Cryer, 1994; Cryer et al., 2003). Themagnitude of the psychological and physical consequences of hypoglycaemia cannot beoverestimated and is considered <strong>in</strong> detail <strong>in</strong> other chapters of this book. In this chapterthe frequency of hypoglycaemia is described <strong>in</strong> people with type 1 diabetes, along with itsunderly<strong>in</strong>g causes and risk factors.DEFINITIONS OF HYPOGLYCAEMIAAny attempt to consider critically the frequency of hypoglycaemia <strong>in</strong> cl<strong>in</strong>ical practice,requires def<strong>in</strong>itive criteria for what constitutes an episode of hypoglycaemia. This poses animmediate difficulty because researchers of the epidemiology of hypoglycaemia have notemployed common def<strong>in</strong>itions with shared specifications.Biochemical Def<strong>in</strong>itions of <strong>Hypoglycaemia</strong>At first glance, it would seem sensible to employ a biochemical def<strong>in</strong>ition of hypoglycaemia,specify<strong>in</strong>g a given blood glucose concentration, below which hypoglycaemia wouldbe deemed to occur. However, it is not possible to provide such a precise biochemical criterionfor the diagnosis of hypoglycaemia (Service, 1995). As blood glucose concentrations<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

50 FREQUENCY, CAUSES AND RISK FACTORSdecl<strong>in</strong>e, a hierarchy of events occur at <strong>in</strong>dividual glycaemic thresholds, commenc<strong>in</strong>g withcounterregulation (arterialised blood glucose ∼38 mmol/l), impairment of different cognitivefunctions (∼32–26 mmol/l) and the onset of symptoms and neurophysiological changes(∼32–24 mmol/l). In cl<strong>in</strong>ical practice, however, it is usual for venous or capillary bloodglucose levels to be measured, and these are lower than contemporaneous arterialised bloodglucose concentrations (which are usually measured <strong>in</strong> research studies of hypoglycaemia)(Heller and Macdonald, 1996).In the non-diabetic <strong>in</strong>dividual, venous blood glucose concentrations below 3.0 mmol/lmay occur follow<strong>in</strong>g an overnight fast or dur<strong>in</strong>g the course of a prolonged oral glucosetolerance test (Service, 1995). Moreover, as is discussed later, the blood glucose thresholdsfor the onset of symptoms and counterregulation <strong>in</strong> patients with type 1 diabetesmay vary depend<strong>in</strong>g on the preced<strong>in</strong>g or prevail<strong>in</strong>g glycaemic control. Thus, patients withpoor glycaemic control may experience symptoms of hypoglycaemia at venous plasmaglucose concentrations substantially higher than 3.0 mmol/l (Boyle et al., 1988). Patientswith preced<strong>in</strong>g strict glycaemic control may not experience the onset of symptoms of hypoglycaemiauntil venous plasma glucose concentrations have decl<strong>in</strong>ed to below 2.0 mmol/l(Boyle et al., 1995). On a pragmatic basis, <strong>in</strong> rout<strong>in</strong>e cl<strong>in</strong>ical practice, <strong>Diabetes</strong> UK hasrecommended that <strong>in</strong>dividuals with diabetes should try to ensure that their blood glucoseconcentrations do not fall below 4.0 mmol/l (O’Neill, 1997), but this does not def<strong>in</strong>e hypoglycaemia.The American <strong>Diabetes</strong> Association has proposed a blood glucose concentrationof 3.9 mmol/l as represent<strong>in</strong>g hypoglycaemia (ADA Workgroup on Hypoglycemia, 2005),but this has been challenged as be<strong>in</strong>g too high.Cl<strong>in</strong>ical Def<strong>in</strong>itions of <strong>Hypoglycaemia</strong>The <strong>in</strong>ability to agree on a biochemical def<strong>in</strong>ition for hypoglycaemia requires <strong>in</strong>stead theapplication of cl<strong>in</strong>ical criteria (Box 3.1). The difficulty here is that because the symptomsof hypoglycaemia are not specific, and vary between <strong>in</strong>dividuals (see Chapter 2), the use ofsymptomatology alone may be unreliable and may result <strong>in</strong> the <strong>in</strong>clusion of episodes thatare not true hypoglycaemia. In one prospective study, where capillary blood glucose wasmeasured whenever a patient had symptoms suggestive of hypoglycaemia, only 29% of suchepisodes were accompanied by evidence of biochemical hypoglycaemia (i.e., blood glucose

FREQUENCY OF HYPOGLYCAEMIA 51Box 3.1Cl<strong>in</strong>ical def<strong>in</strong>itions of hypoglycaemia• Asymptomatic hypoglycaemia – low blood glucose identified on rout<strong>in</strong>e blood test,with no associated symptoms.• Mild symptomatic hypoglycaemia ∗ – symptoms suggestive of hypoglycaemia;episode successfully treated by the patient alone.• Severe hypoglycaemia ∗ – assistance from a third party is required to effect treatment.• Profound hypoglycaemia ∗ – associated with permanent neurological deficits ordeath.∗ Contemporaneous demonstration of a low blood glucose concentration is not mandatory but, particularly<strong>in</strong> the case of mild episodes, if it is available it does make the def<strong>in</strong>ition more robust.before an episode of hypoglycaemia can be considered to be severe, nor does itspecify any particular method of treat<strong>in</strong>g the low blood glucose. There is a substantialdifference <strong>in</strong> the degree of neuroglycopenia dur<strong>in</strong>g an episode of hypoglycaemiathat has resolved follow<strong>in</strong>g the oral adm<strong>in</strong>istration of carbohydrate, and one that hasresulted <strong>in</strong> coma and required the adm<strong>in</strong>istration of <strong>in</strong>tramuscular glucagon or <strong>in</strong>travenousglucose.Mild hypoglycaemia is considered to occur when an episode is self-treated by the affected<strong>in</strong>dividual. Although the record<strong>in</strong>g of episodes of severe hypoglycaemia is generally regardedas be<strong>in</strong>g a fairly robust measure, mild hypoglycaemia is a much ‘softer’ end-po<strong>in</strong>t. Symptomperception by the patient may be difficult and the glycaemic thresholds at which <strong>in</strong>dividualsdevelop symptoms of hypoglycaemia are very variable. In many cl<strong>in</strong>ical research studies, itis not uncommon to <strong>in</strong>clude a composite def<strong>in</strong>ition of mild hypoglycaemia, which <strong>in</strong>cludessymptomatic episodes that respond to self-treatment (with or without a confirmatory bloodtest) and asymptomatic episodes below an arbitrary biochemical blood glucose value (e.g.3.5 mmol/l) that have been identified through rout<strong>in</strong>e blood glucose monitor<strong>in</strong>g (Andersonet al., 1997). The <strong>in</strong>troduction by some <strong>in</strong>vestigators of a further category of ‘moderate’hypoglycaemia is not helpful and confuses the dist<strong>in</strong>ction between mild and severeepisodes.FREQUENCY OF HYPOGLYCAEMIAThe true frequency of hypoglycaemia <strong>in</strong> people with type 1 diabetes is difficult to estimateaccurately and not simply because of differences <strong>in</strong> the def<strong>in</strong>itions of hypoglycaemia. Mostepisodes occur at home, at work or dur<strong>in</strong>g leisure activities, without <strong>in</strong>volv<strong>in</strong>g medical,nurs<strong>in</strong>g or paramedical staff, and the subsequent recall of such episodes by patients isgenerally poor, particularly with regard to mild hypoglycaemia. Prospective studies thereforehave a strong advantage over retrospective studies <strong>in</strong> their respective abilities to documentthe frequency of hypoglycaemia with accuracy. Another crucially important feature is thenature of the population of people with type 1 diabetes under study – factors such as the

52 FREQUENCY, CAUSES AND RISK FACTORSquality of glycaemic control (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1997) and the presence of impaired awareness of hypoglycaemia (Gold et al., 1997) havea major <strong>in</strong>fluence on the frequency of hypoglycaemia. Participants <strong>in</strong> cl<strong>in</strong>ical <strong>in</strong>terventionalstudies are often not representative of the wider body of people with type 1 diabetes. Forexample, <strong>in</strong> the DCCT (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1993), the subjects were young, motivated and received substantial professional support(particularly those <strong>in</strong> the <strong>in</strong>tensive group). They were pre-selected depend<strong>in</strong>g on their historyof hypoglycaemia. Thus, <strong>in</strong>dividuals were excluded if, <strong>in</strong> the previous two years, theyhad experienced more than one episode of severe hypoglycaemia caus<strong>in</strong>g neurologicalimpairment, without experienc<strong>in</strong>g warn<strong>in</strong>g symptoms, or more than two episodes of seizureor coma, regardless of attributed cause. Such exclusions will have <strong>in</strong>evitably <strong>in</strong>fluenced thebasel<strong>in</strong>e frequency of severe hypoglycaemia <strong>in</strong> the study groups.Small studies estimat<strong>in</strong>g the frequency of hypoglycaemia are likely to be biased becausechance variations <strong>in</strong> the prevalence of risk factors for hypoglycaemia can magnify (ordim<strong>in</strong>ish) the frequency of hypoglycaemic events to a much greater extent than would beobserved <strong>in</strong> larger <strong>in</strong>vestigations. There is also likely to be a major period effect <strong>in</strong> studiesof the prevalence of hypoglycaemia. In the post-DCCT era, it might be anticipated thatthe <strong>in</strong>cidence of hypoglycaemia would rise as patients and diabetes healthcare professionalssought to tighten glycaemic control (Johnson et al., 2002). This trend may be counterbalancedby the <strong>in</strong>creased use of home blood glucose monitor<strong>in</strong>g, improved educational programmesand the greater use of <strong>in</strong>sul<strong>in</strong> analogues and cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion therapy(Chase et al., 2001).Thus, estimates of the frequency of hypoglycaemia must be considered <strong>in</strong> relation to thedef<strong>in</strong>itions employed and the characteristics of the patients studied. Comparisons betweenpresent-day and historical studies must be made <strong>in</strong> the knowledge that treatment targets,methods of monitor<strong>in</strong>g blood glucose and <strong>in</strong>sul<strong>in</strong> therapy have changed greatly <strong>in</strong> recentyears.Frequency of Mild, Symptomatic <strong>Hypoglycaemia</strong>The features of several studies that have exam<strong>in</strong>ed the frequency of mild hypoglycaemia,either prospectively or retrospectively, <strong>in</strong> adults with type 1 diabetes are shown <strong>in</strong> Table 3.1.Rates of mild hypoglycaemia vary substantially, rang<strong>in</strong>g from 8 to 160 episodes per patientper year. However, it is extremely difficult to make direct comparisons between <strong>in</strong>dividualstudies because of differences <strong>in</strong> methodology and patient characteristics.Retrospective studiesTwo retrospective studies by a Danish group (Pedersen-Bjergaard et al., 2001; Pedersen-Bjergaard et al., 2004) were very similar <strong>in</strong> their patient characteristics, most of whom wereadm<strong>in</strong>ister<strong>in</strong>g four or more <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> per day and had moderate glycaemic control.On be<strong>in</strong>g asked to recall the number of episodes of mild, symptomatic hypoglycaemiaexperienced <strong>in</strong> the preced<strong>in</strong>g week, the subjects reported a frequency of two episodes perpatient per week. The strength of these studies lies <strong>in</strong> the large numbers of patients exam<strong>in</strong>edand the short period of recall. This should have m<strong>in</strong>imised <strong>in</strong>accuracy, but may have beenan unrepresentative time frame.

Table 3.1 Frequency of mild hypoglycaemia <strong>in</strong> adults with type 1 diabetesAuthorsNumber ofpatients Follow-up a Treatment b cHbA 1c (%)Def<strong>in</strong>ition ofhypoglycaemiaFrequency d(number/pt/yr)Pramm<strong>in</strong>g et al., 1991 411 1 week (P) 78% twice daily 87 Symptomatic 94Janssen et al., 2000 31 6 weeks (P) All multiple 72 BG < 35 mmol/l 160Pedersen-Bjergaard 201 1 week (R) 86% multiple 86 Symptomatic 104et al., 2001Pedersen-Bjergaard 170 12 months (P) 87% ≥ 4 <strong>in</strong>j 84 Symptomatic 88et al., 2003aPedersen-Bjergaard 1076 1 week (R) 72% multiple 86 Symptomatic 104et al., 2004Donnelly et al., 2005 94 4 weeks (P) 49% S&I f 83 Symptomatic 42Leckie et al., 2005 243 (27 T2) e 12 months (P) 80% multiple 91 Symptomatic 8UK <strong>Hypoglycaemia</strong>Study Group, 200746 (< 5 yrs)54 (> 15 yrs)9–12 months All ≥ 2 <strong>in</strong>jections 73 Symptomatic and/orBG < 30 mmol/l3578 29This table does not <strong>in</strong>clude early studies which tended to focus on rates of mild hypoglycaemia <strong>in</strong> particular subsets of patients, e.g. with impaired awareness of hypoglycaemia or <strong>in</strong>dividuals us<strong>in</strong>gporc<strong>in</strong>e versus human <strong>in</strong>sul<strong>in</strong>.a Duration of follow-up: P = prospective estimation of hypoglycaemia frequency; R = retrospective.b Multiple refers to preprandial <strong>in</strong>jections (<strong>in</strong>j) of soluble (or analogue) <strong>in</strong>sul<strong>in</strong> and bedtime isophane <strong>in</strong>sul<strong>in</strong>.c Mean HbA1c for the subjects under study.d Frequency of hypoglycaemia <strong>in</strong> each study has been adjusted to represent an annual rate per person.e 27 Patients <strong>in</strong> this study had type 2 diabetesf Patients were on soluble and isophane <strong>in</strong>sul<strong>in</strong>s, but frequency of <strong>in</strong>jections not specified.BG = capillary blood glucose concentration.

54 FREQUENCY, CAUSES AND RISK FACTORSProspective studiesProspective studies offer the potential to provide more conv<strong>in</strong>c<strong>in</strong>g data on frequency of mildhypoglycaemia, but substantial differences <strong>in</strong> prevalence were aga<strong>in</strong> reported. In an earlierstudy of 441 patients with type 1 diabetes, managed pr<strong>in</strong>cipally with a twice daily <strong>in</strong>sul<strong>in</strong>regimen conta<strong>in</strong><strong>in</strong>g soluble and isophane <strong>in</strong>sul<strong>in</strong>s, the weekly average was 1.8 episodes ofmild symptomatic hypoglycaemia (Pramm<strong>in</strong>g et al., 1991). These patients had moderateglycaemic control and the period of assessment was one week. This study may be of lessrelevance today <strong>in</strong> view of the current use of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy and <strong>in</strong>sul<strong>in</strong> analogues,yet it is <strong>in</strong>terest<strong>in</strong>g that the rate of mild hypoglycaemia is unchanged today.A Danish prospective study (Pedersen-Bjergaard et al., 2003a) <strong>in</strong>cluded patients withsimilar characteristics to those <strong>in</strong> two retrospective studies by the same group (Pedersen-Bjergaard et al., 2001; Pedersen-Bjergaard et al., 2004). <strong>Hypoglycaemia</strong> was recordedmonthly with episodes of mild symptomatic hypoglycaemia be<strong>in</strong>g reported for the preced<strong>in</strong>gweek. Mild hypoglycaemia occurred on average 1.7 times per patient per week. Subjectswere also asked to perform a monthly five-po<strong>in</strong>t blood glucose profile and to record <strong>in</strong> additionany blood glucose value below 3.0 mmol/l. Measurements demonstrat<strong>in</strong>g biochemicalhypoglycaemia represented 3.7% of all blood glucose read<strong>in</strong>gs.In a community-based study <strong>in</strong> Tayside, Scotland, 94 adults with type 1 diabetes wereselected at random from a regional diabetes database and were asked to record episodes ofhypoglycaemia prospectively over one month (Donnelly et al., 2005). Their median age was40 years, median duration of diabetes was 18 years and median HbA 1c was 8.3%. Biphasic<strong>in</strong>sul<strong>in</strong> was used by 35% of participants and 49% used a comb<strong>in</strong>ation of <strong>in</strong>termediate andshort-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>s (although the frequency of <strong>in</strong>jections was not reported). A total of325 episodes of mild hypoglycaemia occurred, represent<strong>in</strong>g a rate of 41.5 episodes perperson per year, i.e., approximately half the rate reported by Pramm<strong>in</strong>g et al., (1991) andPedersen-Bjergaard et al., (2003a). The study has weaknesses; it was relatively small anddata on frequency of blood glucose monitor<strong>in</strong>g were limited. The precise criteria for def<strong>in</strong><strong>in</strong>gmild hypoglycaemia were not clearly described and, <strong>in</strong> particular, the role of contemporaneousmonitor<strong>in</strong>g data was not specified. Nevertheless, the subjects were probably veryrepresentative of the population of people with type 1 diabetes <strong>in</strong> that region.Similar data have been reported from a multicentre study from the United K<strong>in</strong>gdom (UK<strong>Hypoglycaemia</strong> Study Group, 2007). The primary aim of this study was to compare thefrequencies of hypoglycaemia <strong>in</strong> <strong>in</strong>dividuals with different types and durations of diabetes,receiv<strong>in</strong>g different treatment modalities. As part of the study, 50 adults with type 1 diabetesof duration less than five years and 57 adults with type 1 diabetes of greater than 15 yearsduration were recruited. All subjects used two or more <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> per day andtheir glycaemic control was good (mean HbA 1c < 80%). The participants were followed forbetween 9–12 months (mean 10 months) and were asked to report all episodes of symptomatic,self-treated hypoglycaemia and episodes where blood glucose was less 3.0 mmol/l,regardless of symptomatology. Subjects were given forms to record such episodes and wereencouraged to record contemporaneous blood glucose levels. To maximise compliance,subjects were asked to send <strong>in</strong> completed forms every month to the local research centre,<strong>in</strong>clud<strong>in</strong>g when no episodes of hypoglycaemia had occurred. If no forms were received,telephone contact was made with the subjects. Us<strong>in</strong>g this robust methodology, mean ratesof hypoglycaemia of 35 and 29 episodes per person per year were reported for the shortand long duration groups respectively. The distribution of episodes was much skewed, with

FREQUENCY OF HYPOGLYCAEMIA 55some <strong>in</strong>dividuals report<strong>in</strong>g no events and others <strong>in</strong> excess of 200 per year; overall, approximately85% of <strong>in</strong>dividuals experienced at least one episode of mild hypoglycaemia. Theauthors did not report the relative proportion of symptomatic and asymptomatic episodes, orthe proportion of symptomatic episodes with a corroborative blood test, but these data areextremely <strong>in</strong>formative and <strong>in</strong>dicate that duration of diabetes has little impact on the frequencyof mild hypoglycaemia, an observation that has been made before (Pedersen-Bjergaardet al., 2004).In a 12 month prospective study of 243 <strong>in</strong>sul<strong>in</strong>-treated adults, Leckie and colleaguesreported that mild, symptomatic hypoglycaemia occurred with a frequency of only eightepisodes per patient per year (Leckie et al., 2005). This is one of the lowest rates to bereported <strong>in</strong> a large group of people, most of whom had type 1 diabetes. However, severalreasons may account for this result. The subjects had suboptimal glycaemic control, with amean HbA 1c of 9.1%. A small number of people with <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes were<strong>in</strong>cluded, who might be expected to have had a lower overall frequency of hypoglycaemia.The prevalence of impaired awareness of hypoglycaemia, a recognised risk factor for hypoglycaemia(see Chapter 7), was exceptionally low <strong>in</strong> this cohort at 3%. The proportion ofsubjects us<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues was not reported, but it would certa<strong>in</strong>ly have been higherthan <strong>in</strong> studies from the early 1990s. The ma<strong>in</strong> strength of the study, namely a long periodof follow up, may have been an <strong>in</strong>advertent weakness by caus<strong>in</strong>g ‘patient fatigue’, i.e.,participants may have been less assiduous <strong>in</strong> record<strong>in</strong>g episodes of mild hypoglycaemia asthe study progressed. F<strong>in</strong>ally, this was primarily a study of the impact of hypoglycaemiaoccurr<strong>in</strong>g <strong>in</strong> the work place. Thus, all of the participants were <strong>in</strong> full-time employment and sorepresented an atypical group of <strong>in</strong>dividuals who may have adopted strategies to reduce thefrequency of hypoglycaemia because of the potentially adverse effects that this could haveon their jobs.Frequency of Asymptomatic, Biochemical <strong>Hypoglycaemia</strong>Aside from the problem of specify<strong>in</strong>g a biochemical threshold for hypoglycaemia, therecan be little doubt that asymptomatic hypoglycaemia is even more common than symptomatic,mild hypoglycaemia. However, a major determ<strong>in</strong>ant of the frequency of documentedasymptomatic hypoglycaemia is, necessarily, the frequency with which blood glucose ismeasured.Traditionally, two different approaches have been used to <strong>in</strong>vestigate this phenomenon:(a) tak<strong>in</strong>g multiple sequential blood samples <strong>in</strong> a controlled, experimental sett<strong>in</strong>g, over a prespecifiedtime period; and (b) <strong>in</strong>vit<strong>in</strong>g patients <strong>in</strong> the community to perform capillary bloodtests at multiple, pre-set time po<strong>in</strong>ts. Provid<strong>in</strong>g the sampl<strong>in</strong>g frame is sufficiently frequent,the advantage of the former strategy is that it will capture all episodes of biochemicalhypoglycaemia, however brief, dur<strong>in</strong>g the time period under study. The disadvantage isthat it is labour <strong>in</strong>tensive for <strong>in</strong>vestigators and so only a small number of subjects can bestudied over a relatively short time period, such as 12 or 24 hours. By contrast, the use ofhome blood glucose monitor<strong>in</strong>g, particularly with meters that store the results electronically,allows larger numbers of subjects to be studied over very prolonged time periods. Typically,subjects are asked to perform periodic seven-po<strong>in</strong>t profiles, i.e., blood glucose estimationsbefore each meal, two hours after food and dur<strong>in</strong>g the night. The obvious disadvantages ofthis mode of <strong>in</strong>vestigation are that subjects may forget to perform the relevant monitor<strong>in</strong>g,

56 FREQUENCY, CAUSES AND RISK FACTORSor may do so at the wrong times, and that episodes of asymptomatic hypoglycaemia mayoccur outside the sampl<strong>in</strong>g time-frames and, thus, be missed.Thorste<strong>in</strong>sson et al., (1986) exam<strong>in</strong>ed seven-po<strong>in</strong>t capillary blood glucose profiles <strong>in</strong>99 adults with type 1 diabetes and demonstrated that the frequency of biochemical hypoglycaemiawas <strong>in</strong>versely related, <strong>in</strong> a curvil<strong>in</strong>ear manner, to the median blood glucoseconcentration. Thus, for example, <strong>in</strong> patients who had a median blood glucose concentrationof 5.0 mmol/l, 10% of blood glucose levels were less than 3.0 mmol/l. By contrast, only 2.5%of blood glucose levels were below 3.0 mmol/l, <strong>in</strong> patients whose median blood glucoseconcentration was 10 mmol/l (Figure 3.1; Thorste<strong>in</strong>sson et al., 1986).In a separate study, nocturnal blood glucose profiles were exam<strong>in</strong>ed <strong>in</strong> 31 patients withtype 1 diabetes us<strong>in</strong>g multiple <strong>in</strong>jection therapy with soluble and isophane (NPH) <strong>in</strong>sul<strong>in</strong>(mean HbA 1c 8.6%). Venous blood samples were taken every 30 m<strong>in</strong>utes from 11 p.m. until7.30 a.m.. Nocturnal hypoglycaemia (blood glucose less than 3.0 mmol/l) occurred on 29%of occasions and 67% of these episodes were asymptomatic (Vervoort et al., 1996). Six<strong>in</strong>dividuals were studied on two separate nights and the blood glucose profiles, perhapspredictably, showed considerable <strong>in</strong>tra-<strong>in</strong>dividual variation.Figure 3.1 Correlation between the median blood glucose concentration and the frequencies of bloodglucose concentrations below 4.0, 3.0, 2.5 and 2.0 mmol/l <strong>in</strong> adults with type 1 diabetes. Solid l<strong>in</strong>esrepresent data from 70 adults on twice daily <strong>in</strong>sul<strong>in</strong> therapy and dotted l<strong>in</strong>es represent data from20 adults treated with cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong>. Approximately 10% of read<strong>in</strong>gs were below3.0 mmol/l, when median blood glucose concentration was 5.0 mmol/l. Reproduced with permissionfrom Thorste<strong>in</strong>sson et al. (1986) © John Wiley & Sons, Ltd

FREQUENCY OF HYPOGLYCAEMIA 57Janssen et al., (2000) have studied, prospectively, the frequency of biochemical hypoglycaemiaover a six week period <strong>in</strong> 31 people with type 1 diabetes with a mean HbA 1c of7.2% (all had a HbA 1c ≤ 83%). Subjects were all us<strong>in</strong>g a multiple <strong>in</strong>jection regimen withsoluble and isophane <strong>in</strong>sul<strong>in</strong>s and performed one seven-po<strong>in</strong>t blood glucose profile and sixfour-po<strong>in</strong>t profiles each week. No overnight read<strong>in</strong>gs were performed. Patients completeda mean of 82% of the required monitor<strong>in</strong>g schedule and overall experienced a mean of18.7 episodes of hypoglycaemia (i.e., approximately 160 episodes per year). The range waswide, however, at between zero and 41 episodes per <strong>in</strong>dividual, over the six weeks ofthe study.The more widespread application of cont<strong>in</strong>uous glucose monitor<strong>in</strong>g systems may help toelucidate the frequency of mild hypoglycaemia with greater accuracy. In one study, 65%of people with type 1 diabetes monitored over three days had an episode of asymptomatichypoglycaemia (<strong>in</strong>terstitial glucose < 33 mmol/l) (Bode et al., 2005). However, much workstill needs to be done to clarify the relationship between <strong>in</strong>terstitial glucose and blood glucoseconcentrations, before this can be regarded as a robust tool for detect<strong>in</strong>g hypoglycaemia (seeChapter 5).Frequency of Severe <strong>Hypoglycaemia</strong>As with mild hypoglycaemia, direct comparisons between <strong>in</strong>dividual studies on the frequencyof severe hypoglycaemia are not straightforward. However, it is a more robust end-po<strong>in</strong>tthan mild hypoglycaemia and, because episodes typically have a more profound effect on<strong>in</strong>dividuals, retrospective recall is much more reliable. At the conclusion of their prospectivestudy, Pedersen-Bjergaard et al. (2003b), showed that 90% of subjects recalled correctlytheir experience of severe hypoglycaemia over the preced<strong>in</strong>g year (Figure 3.2). Subjectswho had experienced a high <strong>in</strong>cidence of severe hypoglycaemia (prospectively recorded),retrospectively underestimated the overall rate by around 15%.Table 3.2 lists some of the large surveys (each <strong>in</strong> excess of 100 participants) that haveexam<strong>in</strong>ed the frequency of severe hypoglycaemia. The table focuses primarily on studiesexam<strong>in</strong><strong>in</strong>g unselected groups of <strong>in</strong>dividuals with type 1 diabetes, and so excludes <strong>in</strong>terventiontrials (The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1993; Reichard andPihl, 1994; MacLeod et al., 1995) and studies that exam<strong>in</strong>ed particular sub-groups of patients,such as people with impaired awareness of hypoglycaemia (Gold et al., 1994; MacLeodet al., 1994; Clarke et al., 1995), differ<strong>in</strong>g durations of diabetes (UK <strong>Hypoglycaemia</strong> StudyGroup, 2007) or subjects who have received <strong>in</strong>tensive therapy or education (Muhlhauseret al., 1985; Pampanelli et al., 1996; Bott et al., 1997). A study that exam<strong>in</strong>ed a mixedgroup of children, adolescents and adults (Allen et al., 2001) has also been excluded. Thefrequency of severe hypoglycaemia (def<strong>in</strong>ed as episodes requir<strong>in</strong>g third party assistance) <strong>in</strong>the studies listed <strong>in</strong> Table 3.2 is remarkably consistent at 1.0 to 1.6 episodes per patientper year.However, it is important to note that the frequency of severe hypoglycaemia <strong>in</strong> unselectedpopulations does not follow a Gaussian distribution (Figure 3.3). The distribution isheavily skewed such that the majority of <strong>in</strong>dividuals do not experience any severe hypoglycaemia<strong>in</strong> a given year, while a small number of <strong>in</strong>dividuals have recurrent episodes. Inthe studies highlighted <strong>in</strong> Table 3.2, between 30–40% of <strong>in</strong>dividuals experienced at leastone episode of severe hypoglycaemia over the period <strong>in</strong> question. The proportion affected

58 FREQUENCY, CAUSES AND RISK FACTORS15Recalled severe hypoglycaemia(episodes per patient-year)10500 5 10 15Prospectively recorded severe hypoglycaemia(episodes per patient-year)Figure 3.2 Correlation between prospectively recorded and retrospectively recalled rate of severehypoglycaemia over the same one year period <strong>in</strong> 230 people with type 1 diabetes. Marker sizes areweighted by the number of cases. R 2 = 066; p < 0001. Reproduced with permission from Pedersen-Bjergaard et al. (2003b) © John Wiley & Sons, Ltd<strong>in</strong> the study by Pramm<strong>in</strong>g et al., (1991) was much lower, but the period of follow up wasonly one week. This skewed distribution serves to emphasise further the importance ofpatient selection <strong>in</strong> ascerta<strong>in</strong><strong>in</strong>g the frequency of severe hypoglycaemia with accuracy, as theexclusion of a relatively small number of people at high risk would substantially reduce theoverall risk.In three of the studies of unselected adults with type 1 diabetes, severe hypoglycaemia wasfurther subdivided to exam<strong>in</strong>e episodes associated with more significant neuroglycopenia,i.e., those result<strong>in</strong>g <strong>in</strong> coma and/or seizures (Table 3.2) (ter Braak et al., 2000; Pedersen-Bjergaard et al., 2003a; Pedersen-Bjergaard et al., 2004). Furthermore, <strong>in</strong> the study byMuhlhauser et al., (1998) only episodes treated with <strong>in</strong>tra-muscular glucagon or <strong>in</strong>travenousglucose were addressed. Predictably, such events were rarer and represented about onequarter of all episodes of severe hypoglycaemia.Emergency and hospital services will occasionally be <strong>in</strong>volved <strong>in</strong> the management ofsevere hypoglycaemia and there are data on the use of such agencies. This clearly hasto be <strong>in</strong>terpreted with caution, as the majority of all hypoglycaemia is managed <strong>in</strong> thecommunity, without <strong>in</strong>volvement of (para)cl<strong>in</strong>ical staff. Individuals admitted to hospitalwith hypoglycaemia are probably atypical, and have an <strong>in</strong>creased prevalence of alcoholdependence and mental illness (Hart and Frier, 1998). An early study from Australia reportedthat, over one year, 3.5% of people attend<strong>in</strong>g an urban diabetes cl<strong>in</strong>ic had an episode ofhypoglycaemia severe enough to warrant referral to hospital (Moses et al., 1985). Morerecent data from Tayside, Scotland, demonstrated that 7.1% of people with type 1 diabetes

Table 3.2 Frequency of severe hypoglycaemia <strong>in</strong> adults with type 1 diabetesAuthors Number of patients Follow-up a Treatment b HbA 1cc(%)Def<strong>in</strong>ition ofhypoglycaemiaFrequency d(episodes/pt/yr)ProportionAffected (%)Pramm<strong>in</strong>g et al., 1991 411 1 week (P) 78% twice daily 87 Third party 14 3MacLeod et al., 1993 600 (56 T2) e 12 months (R) 76% twice daily 107 A1 Third party 16 29Muhlhauser et al.,1998684 12 months (R) 70% > 2 <strong>in</strong>j 80 Glucagon/glucose 021 13ter Braak et al., 2000 195 12 months (R) 82% <strong>in</strong>tensive 78 Third partyComa/seizurePedersen-Bjergaardet al., 2001Pedersen-Bjergaardet al., 2003aPedersen-Bjergaardet al., 200415 4104207 24 months (R) 86% ≥ 4 <strong>in</strong>j 86 Third party 11 NR170 12 months (P) 87% ≥ 4 <strong>in</strong>j 84 Third partyComa/seizure1076 12 months (R) 72% ≥ 4 <strong>in</strong>j 86 Third partyComa/seizure11 390313 37035Leckie et al., 2005 243 (27 T2) e 12 months (P) 80% multiple 91 Third party 098 34UK <strong>Hypoglycaemia</strong>Study Group46 (15 years)9–12 months All ≥ 2 <strong>in</strong>jections 7.37.8Third party 1.13.22246This table only considers studies exam<strong>in</strong><strong>in</strong>g <strong>in</strong> excess of 100 subjects and does not <strong>in</strong>clude early studies which tended to focus on rates of severe hypoglycaemia <strong>in</strong> particular subsets of patients, e.g.with impaired awareness of hypoglycaemia or <strong>in</strong>dividuals us<strong>in</strong>g porc<strong>in</strong>e versus human <strong>in</strong>sul<strong>in</strong>.a Duration of follow-up: P = prospective estimation of hypoglycaemia frequency; R = retrospective.b Multiple refers to preprandial <strong>in</strong>jections (<strong>in</strong>j) of soluble (or analogue) <strong>in</strong>sul<strong>in</strong> and bedtime isophane <strong>in</strong>sul<strong>in</strong>.c Mean HbA1c for the subjects under study; A1 = figure is HbA 1 .d Frequency of hypoglycaemia <strong>in</strong> each study has been adjusted to represent an annual rate per person.NR = data not reported.e 56 <strong>in</strong>clude patients with type 2 diabetes.

60 FREQUENCY, CAUSES AND RISK FACTORS90807060% of patients504030201000 2 4 6 8 10Severe hypoglycaemia, episodes per yearFigure 3.3 Distribution of self-reported number of episodes of severe hypoglycaemia dur<strong>in</strong>g thepreced<strong>in</strong>g year <strong>in</strong> 1049 unselected patients with type 1 diabetes (light bars) and 209 patients selectedby criteria to mimic the characteristics of the DCCT cohort (dark bars). Reproduced with permissionfrom Pedersen-Bjergaard et al. (2004) © John Wiley & Sons, Ltdhad an episode of hypoglycaemia that required contact with emergency medical teams overone year, with an <strong>in</strong>cidence rate of 0.12 per patient per year (Leese et al., 2003). Thus, onlyabout one <strong>in</strong> 10 of all episodes of severe hypoglycaemia result <strong>in</strong> contact with emergencyservices.Frequency of <strong>Hypoglycaemia</strong>: Summary and ConclusionsThe literature on frequency of hypoglycaemia is heterogeneous and <strong>in</strong>consistent and, thusthere are considerable methodological limitations <strong>in</strong> our ability to ascerta<strong>in</strong> accurately thefrequency of hypoglycaemia. The def<strong>in</strong>itive study, a prolonged, prospective evaluation of alarge number of unselected people with type 1 diabetes, outside ‘trial’ conditions, has yet to beperformed. Exist<strong>in</strong>g data on severe hypoglycaemia are probably fairly accurate, but estimatesof mild hypoglycaemia should be regarded with caution. However, the ‘average’ patientwith type 1 diabetes will probably experience about 1–2 episodes of mild hypoglycaemiaper week and be exposed to several thousand episodes over a lifetime with diabetes. Overall,severe hypoglycaemia may be expected to occur once or twice each year, but the distributionis heavily skewed, such that most <strong>in</strong>dividuals will be unaffected while a small number willhave multiple episodes. In the early 1980s, Robert Tattersall’s group <strong>in</strong> Nott<strong>in</strong>gham reportedthat admission to hospital as a consequence of hypoglycaemia represented the ‘tip of aniceberg’ of all episodes of hypoglycaemia (Potter et al., 1982), and this observation doesnot appear to have changed.

CAUSES OF HYPOGLYCAEMIACAUSES OF HYPOGLYCAEMIA 61Although advances <strong>in</strong> <strong>in</strong>sul<strong>in</strong> therapy have been made over the last 80 years, the adm<strong>in</strong>istrationof exogenous <strong>in</strong>sul<strong>in</strong> rema<strong>in</strong>s a very crude means of manag<strong>in</strong>g type 1 diabetes. Thetime-action profiles of the modern <strong>in</strong>sul<strong>in</strong> analogues do not mimic the physiological changes<strong>in</strong> plasma <strong>in</strong>sul<strong>in</strong> concentrations that occur <strong>in</strong> non-diabetic <strong>in</strong>dividuals (Figure 3.4) and,crucially, concentrations of exogenous <strong>in</strong>sul<strong>in</strong> cannot respond to changes <strong>in</strong> blood glucoseconcentration. Therefore, at its most fundamental level, hypoglycaemia <strong>in</strong> people with type1 diabetes is the result of an imbalance between <strong>in</strong>sul<strong>in</strong>-mediated glucose efflux from theblood stream and the amount of glucose enter<strong>in</strong>g the circulation from <strong>in</strong>gested carbohydrateand from the liver. Cryer et al. (2003), have grouped the causes of hypoglycaemia <strong>in</strong>type 1 diabetes <strong>in</strong>to six categories (Box 3.2) depend<strong>in</strong>g on their relative effects on <strong>in</strong>sul<strong>in</strong>Figure 3.4 Mean 24 hour plasma glucose and <strong>in</strong>sul<strong>in</strong> profiles <strong>in</strong> 12 healthy non-diabetic <strong>in</strong>dividuals.Shaded areas represent 95% confidence <strong>in</strong>tervals. Glucose levels rema<strong>in</strong> with<strong>in</strong> tight limits, while thereis considerable variation <strong>in</strong> <strong>in</strong>sul<strong>in</strong> concentrations, particularly around meal times. Repr<strong>in</strong>ted from TheLancet, 358, Owens et al., Insul<strong>in</strong>s today and beyond 739–746 (2001), with permission from ElsevierBox 3.2Causes of hypoglycaemia <strong>in</strong> type 1 diabetes1. Inappropriate <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jection – e.g. excessive dose, <strong>in</strong>appropriate time, <strong>in</strong>appropriate<strong>in</strong>sul<strong>in</strong> formulation.2. Inadequate exogenous carbohydrate – e.g. missed meal or snack, overnight fast.3. Increased carbohydrate utilisation – e.g. exercise.4. Decreased endogenous glucose production – e.g. excessive alcohol consumption.5. Increased <strong>in</strong>sul<strong>in</strong> sensitivity – e.g. night time, exercise, weight loss.6. Decreased <strong>in</strong>sul<strong>in</strong> clearance – e.g. renal failure.Modified from Cryer et al., 2003.

62 FREQUENCY, CAUSES AND RISK FACTORSconcentrations or sensitivity and on glucose entry <strong>in</strong>to the circulation. Although this classificationis not perfect, it is a useful start<strong>in</strong>g po<strong>in</strong>t for consider<strong>in</strong>g the causes of an episodeof hypoglycaemia.Patient ErrorIt is common after an episode of hypoglycaemia for the person with diabetes or, <strong>in</strong>deeda member of the diabetes team, to try to ascerta<strong>in</strong> why the episode occurred. Althoughthere is always a risk of ‘spurious attribution’, <strong>in</strong> many <strong>in</strong>stances an obvious cause canbe identified and this is often the result of an error of judgement. Carbohydrate <strong>in</strong>takemay have been <strong>in</strong>adequate because a meal was missed or delayed, or simply conta<strong>in</strong>ed an<strong>in</strong>sufficient content of carbohydrate. Alternatively, the patient may have <strong>in</strong>jected too much<strong>in</strong>sul<strong>in</strong> relative to the amount of carbohydrate <strong>in</strong> the meal. It is also not uncommon for<strong>in</strong>sul<strong>in</strong> to be adm<strong>in</strong>istered at an <strong>in</strong>appropriate time or for the ‘wrong’ <strong>in</strong>sul<strong>in</strong> to be <strong>in</strong>jectedaccidentally, e.g. a rapid-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogue is adm<strong>in</strong>istered at a time when basal <strong>in</strong>sul<strong>in</strong>should have been given. An often unrecognised problem is the <strong>in</strong>adequate re-suspensionof isophane (or lente) <strong>in</strong>sul<strong>in</strong>s or of fixed mixtures of <strong>in</strong>sul<strong>in</strong>. If these <strong>in</strong>sul<strong>in</strong>s are notfully resuspended prior to subcutaneous <strong>in</strong>jection, the <strong>in</strong>sul<strong>in</strong> may be absorbed at variablerates result<strong>in</strong>g <strong>in</strong> unpredictable <strong>in</strong>sul<strong>in</strong> levels and, thus, an <strong>in</strong>creased risk of hypoglycaemia(Owens et al., 2001). Deliberate overdose of <strong>in</strong>sul<strong>in</strong> is rare, but may result <strong>in</strong> protractedhypoglycaemia.AlcoholThe relationship of alcohol to hypoglycaemia is considered <strong>in</strong> detail <strong>in</strong> Chapter 5. Surveysbased on patient <strong>in</strong>terviews have implicated alcohol <strong>in</strong> up to one fifth of episodes ofsevere hypoglycaemia requir<strong>in</strong>g hospital admission (Potter et al., 1982; Moses et al., 1985;Feher et al., 1989; Hart and Frier, 1998). In a recent study of 141 people treated forsevere hypoglycaemia <strong>in</strong> three emergency centres <strong>in</strong> Copenhagen, alcohol was detected<strong>in</strong> the blood of 17% of the subjects (Pedersen-Bjergaard et al., 2005). The median bloodalcohol concentration was 11 mmol/l. Alcohol <strong>in</strong>hibits gluconeogenesis and so may directlycontribute to the development of hypoglycaemia. In addition, there are some data to suggestthat alcohol attenuates the counterregulatory response to hypoglycaemia (Avogaro et al.,1993; Turner et al., 2001). However, the ma<strong>in</strong> impact of alcohol probably centres onits ability to impair awareness of hypoglycaemia and so h<strong>in</strong>der the ability of <strong>in</strong>dividualsto take appropriate corrective therapy (Kerr et al., 1990). Thus, an <strong>in</strong>dividual underthe <strong>in</strong>fluence of alcohol may not recognise that he or she is hypoglycaemic, and evenif the symptoms are recognised, the person may not have the capacity to self-treat. Anepisode of mild hypoglycaemia may therefore be converted rapidly <strong>in</strong>to a severe episode.Moreover, friends, colleagues or bystanders may presume that the neuroglycopenic symptomsand signs exhibited by the <strong>in</strong>dividual are a consequence of alcohol <strong>in</strong>toxicationand so aga<strong>in</strong> appropriate treatment and medical help may not be provided. It is forthese reasons that alcohol is implicated <strong>in</strong> many <strong>in</strong>stances of profound and protracted<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia associated with permanent neurological damage (Arky et al.,1968).

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 63ExerciseExercise is recommended for people with type 1 diabetes because of its positive physiologicaland psychological effects. However, exercise can <strong>in</strong>crease the risk of hypoglycaemia bothdur<strong>in</strong>g the physical activity itself and <strong>in</strong> the recovery period (see Chapter 14). The reasonsfor the high <strong>in</strong>cidence of exercise-<strong>in</strong>duced hypoglycaemia <strong>in</strong> adults with type 1 diabeteshave not been fully elucidated. Moderate exercise <strong>in</strong> non-diabetic <strong>in</strong>dividuals causes a fall<strong>in</strong> plasma <strong>in</strong>sul<strong>in</strong> to 40–50% of pre-exercise levels (Galassetti et al., 2003). This normalphysiological decl<strong>in</strong>e cannot occur when <strong>in</strong>sul<strong>in</strong> is be<strong>in</strong>g adm<strong>in</strong>istered exogenously, unlessthe dose is reduced before exercise commences (Sonnenberg et al., 1990). An acute <strong>in</strong>crease<strong>in</strong> <strong>in</strong>sul<strong>in</strong> sensitivity follow<strong>in</strong>g exercise also <strong>in</strong>creases the risk of hypoglycaemia (Sonnenberget al., 1990).The metabolic and counterregulatory hormonal responses to acute hypoglycaemia andexercise are qualitatively very similar. Antecedent exercise blunts the counterregulatoryresponse to subsequent acute hypoglycaemia (Galassetti et al., 2001a; Galassetti et al.,2001b). The <strong>in</strong>verse situation also applies, i.e., antecedent hypoglycaemia dim<strong>in</strong>ishes thecounterregulatory response to subsequent exercise (Galassetti et al., 2003). This means thatat the very time that metabolic substrate requirements are <strong>in</strong>creas<strong>in</strong>g, there is the potentialfor an acute failure of endogenous glucose production. Thus, impaired counterregulatoryresponses may be an important mechanism <strong>in</strong> promot<strong>in</strong>g exercise-related hypoglycaemia <strong>in</strong>type 1 diabetes.RISK FACTORS FOR SEVERE HYPOGLYCAEMIAIn a relatively high proportion of cases – <strong>in</strong> some series as high as 40% (Potter et al., 1982;Feher et al., 1989) – it is not possible to identify the precipitat<strong>in</strong>g cause of an episode ofhypoglycaemia. Indeed, this figure may be even higher, because the phenomenon of ‘spuriousattribution’ means that <strong>in</strong> some <strong>in</strong>stances the perceived precipitant of a given episode mayhave been an <strong>in</strong>nocent bystander. It has, therefore, been <strong>in</strong>creas<strong>in</strong>gly recognised that diabeteshealthcare professionals must look beyond conventional precipitat<strong>in</strong>g factors and considerother phenomena which may be associated with an <strong>in</strong>creased risk of hypoglycaemia, namely,the risk factors for hypoglycaemia (Table 3.3). Many of these are discussed <strong>in</strong> more detail<strong>in</strong> other chapters.Intensive Insul<strong>in</strong> TherapyRandomised trials, most notably the DCCT, have provided substantial data on the epidemiologyof hypoglycaemia <strong>in</strong> adults with type 1 diabetes and, <strong>in</strong> particular, on the impact of<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy.The <strong>Diabetes</strong> Control and Complications Trial (DCCT)The DDCT was a landmark study and provided diabetes specialists with the long-awaitedproof that strict glycaemic control limited the <strong>in</strong>cidence and severity of microvascularcomplications <strong>in</strong> people with type 1 diabetes. A total of 1441 patients with type 1 diabetes

Table 3.3 Risk factors for severe hypoglycaemia <strong>in</strong> adults with type 1 diabetesStudyIntensivePrevioustherapy Low HbA 1c episode DurationImpairedawarenessC-peptidenegative Sex AgeInsul<strong>in</strong>dose ACE activityMuhlhauser et al., 1985 − − −Muhlhauser et al., 1987 − −MacLeod et al., 1993 − + − − −Gold et al., 1994 +EURODIAB, 1994 + +MacLeod et al., 1994 +SDIS, 1994 +Clarke et al., 1995 +Pampanelli, et al., 1996 −Bott et al., 1997 − + + + +DCCT, 1997 + + + + + M + +Gold et al., 1997 − + + + +Muhlhauser et al., 1998 − − + − + + − − −Pedersen-Bjergaard et al., 2001 + ∗ + + + +ter Braak et al., 2000 − − + − − +Leese et al., 2003 − + − +Pedersen-Bjergaard et al., 2003 − + ∗ + − − − +Pedersen-Bjergaard et al., 2004 − − + − − − −Leckie et al., 2005 − + − F − −UK <strong>Hypoglycaemia</strong> + +Study Group, 2007This table exam<strong>in</strong>es the risk factors for severe hypoglycaemia that have been most commonly exam<strong>in</strong>ed <strong>in</strong> adults with type 1 diabetes. Studies exam<strong>in</strong><strong>in</strong>g predom<strong>in</strong>antly mild hypoglycaemia werenot <strong>in</strong>cluded. +=positive association between risk factor and severe hypoglycaemia; −=no association between risk factor and severe hypoglycaemia. ∗ = only a risk factor if awareness ofhypoglycaemia not <strong>in</strong>cluded. M = severe hypoglycaemia more common <strong>in</strong> men; F = severe hypoglycaemia more common <strong>in</strong> women.

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 65were randomly assigned either to <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy (based on multiple <strong>in</strong>jection <strong>in</strong>sul<strong>in</strong>regimens or cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion therapy) or to conventional <strong>in</strong>sul<strong>in</strong>therapy (one or two <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jections daily) (The <strong>Diabetes</strong> Control and Complications TrialResearch Group, 1993). In the conventional group, patients did not generally perform homeblood glucose monitor<strong>in</strong>g, cl<strong>in</strong>ical reviews were undertaken every three months and patientswere not <strong>in</strong>formed about their HbA 1c result, unless it was <strong>in</strong> excess of 13%. By contrast, <strong>in</strong>the <strong>in</strong>tensive therapy group, subjects performed frequent home blood glucose monitor<strong>in</strong>g,had monthly visits with the study team and also had frequent telephone contact to achieve asstrict glycaemic control as possible. Over a mean follow-up of 6.5 years, the average HbA 1cconcentration <strong>in</strong> the <strong>in</strong>tensive group was approximately 7.0% and <strong>in</strong> the conventional groupapproximately 8.8% (The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1993).The subjects recruited to this study were not typical of the wider population of people withtype 1 diabetes. The participants had greater motivation and <strong>in</strong>dividuals were not permittedto take part if, <strong>in</strong> the preced<strong>in</strong>g two years, they had experienced more than one episodeof severe neurological impairment without warn<strong>in</strong>g symptoms of hypoglycaemia, or morethan two episodes of seizure or coma, regardless of attributed cause. Moreover, dur<strong>in</strong>g thestudy itself, the occurrence of an episode of severe hypoglycaemia <strong>in</strong> an <strong>in</strong>dividual patientprompted a review of conventional risk factors and, <strong>in</strong> <strong>in</strong>stances where probable causes wereidentified, corrective actions such as re-educat<strong>in</strong>g the patient were undertaken (The <strong>Diabetes</strong>Control and Complications Trial Research Group, 1997).Despite all these factors, 3788 episodes of severe hypoglycaemia occurred <strong>in</strong> the 1441patients over the course of the study and, of these, 1027 were associated with coma and/orseizure (The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1997). The rateof severe hypoglycaemia <strong>in</strong> the <strong>in</strong>tensively-treated patients was 0.61 per patient per year,while that <strong>in</strong> the conventionally-treated group was 0.19 per patient per year, i.e., a three-folddifference. Over the mean of 6.5 years of follow-up, 65% of the <strong>in</strong>tensive group experiencedat least one episode of severe hypoglycaemia, compared with 35% of the <strong>in</strong>dividuals <strong>in</strong> theconventional group.The Stockholm <strong>Diabetes</strong> Intervention Study (SDIS)The SDIS was a much smaller study than the DCCT, but had similar aims and objectives.A group of 102 patients with type 1 diabetes were recruited and 89 rema<strong>in</strong>ed after 7.5 yearsof follow-up (Reichard and Pihl, 1994). The mean HbA 1c was 7.1% <strong>in</strong> the <strong>in</strong>tensive groupand 8.5% <strong>in</strong> the conventional group. Severe hypoglycaemia, def<strong>in</strong>ed as episodes requir<strong>in</strong>gthird party assistance, occurred <strong>in</strong> 80% of subjects <strong>in</strong> the <strong>in</strong>tensive group and 58% of those <strong>in</strong>the conventional group over the follow-up period. Overall, the rate of severe hypoglycaemiawas 1.1 per patient per year <strong>in</strong> the <strong>in</strong>tensive group and 0.4 per patient per year <strong>in</strong> the<strong>in</strong>tensive group. Other risk factors for severe hypoglycaemia were not reported.The Bucharest-Dusseldorf StudyThe DCCT was a multicentre study; 27 out of the 29 centres reported that <strong>in</strong>tensive therapywas associated with an <strong>in</strong>creased risk of severe hypoglycaemia, but <strong>in</strong> two centres no<strong>in</strong>creased risk was observed (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1997). This led some <strong>in</strong>vestigators to claim that with appropriate education and tra<strong>in</strong><strong>in</strong>g

66 FREQUENCY, CAUSES AND RISK FACTORS<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy need not necessarily be associated with an <strong>in</strong>creased risk of hypoglycaemia(Plank et al., 2004). In the Bucharest-Dusseldorf Study, 300 <strong>in</strong>dividuals withtype 1 diabetes were randomised to: (a) conventional therapy for one year and then to oneyear of <strong>in</strong>tensive therapy; (b) two years of <strong>in</strong>tensive therapy; or (c) a four-day <strong>in</strong>-patientgroup teach<strong>in</strong>g programme with conventional <strong>in</strong>sul<strong>in</strong> therapy for one year (Muhlhauseret al., 1987). Glycated haemoglob<strong>in</strong> (HbA 1 ), rema<strong>in</strong>ed unchanged at around 12–13% dur<strong>in</strong>gconventional therapy, but fell to ∼95% dur<strong>in</strong>g <strong>in</strong>tensive therapy. In the first year of thestudy, severe hypoglycaemia occurred <strong>in</strong> 6% of the <strong>in</strong>tensively-treated patients and 12% ofthe conventionally-treated patients, and <strong>in</strong> year two the proportion of patients experienc<strong>in</strong>gsevere hypoglycaemia <strong>in</strong> both <strong>in</strong>tensive groups fell to 3–4%.Observational dataThe Dusseldorf team reported observational data on the impact of <strong>in</strong>tensive therapy on thefrequency of severe hypoglycaemia (Bott et al., 1997). A total of 636 people with type1 diabetes who had participated <strong>in</strong> a structured five-day <strong>in</strong>-patient treatment and teach<strong>in</strong>gprogramme for <strong>in</strong>tensification of <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> one of ten different hospitals <strong>in</strong> Germany,were re-exam<strong>in</strong>ed at <strong>in</strong>tervals over six years. The mean HbA 1c fell from 8.3% to 7.6%.Severe hypoglycaemia, def<strong>in</strong>ed as episodes treated with <strong>in</strong>tramuscular glucagon or <strong>in</strong>travenousglucose, decreased from 0.28 episodes per patient per year <strong>in</strong> the year preced<strong>in</strong>g theprogramme to 0.17 episodes per patient per year afterwards (Bott et al., 1997). The variation<strong>in</strong> <strong>in</strong>cidence of severe hypoglycaemia between different centres ranged from 0.05–0.27episodes per patient per year.Pampanelli and colleagues (1996) from Perugia reported retrospective data on 112 <strong>in</strong>dividualswho had been commenced on <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy (preprandial soluble <strong>in</strong>sul<strong>in</strong>and bedtime isophane <strong>in</strong>sul<strong>in</strong>) at diagnosis of diabetes and who were attend<strong>in</strong>g cl<strong>in</strong>ic atleast four times per year. Mean HbA 1c was 7.2% and mean duration of diabetes was 7.8years. Frequency of mild hypoglycaemia was estimated by a review of the patients’ bloodglucose monitor<strong>in</strong>g diaries <strong>in</strong> the n<strong>in</strong>e months prior to study, and the overall rate was 35.6episodes per patient per year. Severe hypoglycaemia, def<strong>in</strong>ed as episodes requir<strong>in</strong>g thirdparty assistance, was assessed retrospectively over the duration of diabetes and was recalledby only six patients (represent<strong>in</strong>g an overall rate of 0.001 episodes per patient per year).SummaryThus, <strong>in</strong> the DCCT and the SDIS, <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy was associated with a nearlythree-fold <strong>in</strong>crease <strong>in</strong> the risk of severe hypoglycaemia. By contrast, <strong>in</strong> other studies, the<strong>in</strong>cidence of severe hypoglycaemia fell when <strong>in</strong>tensive therapy was coupled with a detailededucation programme. The teams from Dusseldorf and Perugia would argue that the highquality of their education programmes resulted <strong>in</strong> the low frequencies of observed severehypoglycaemia. Although this may be the case, the importance of patient selection <strong>in</strong> allof these studies cannot be over-emphasised. In their study of 1076 adults with type 1diabetes from the United K<strong>in</strong>gdom and Denmark, Pedersen-Bjergaard et al. (2004) exam<strong>in</strong>eda subset of 230 <strong>in</strong>dividuals whose cl<strong>in</strong>ical characteristics were similar to patients enrolled<strong>in</strong> the DCCT. This sub-group accounted for only 5.4% of all reported episodes of severehypoglycaemia, with an overall rate of 0.35 episodes per patient per year, i.e., approximately

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 67one quarter that of the entire group. Risk factors for severe hypoglycaemia <strong>in</strong> this group(impaired awareness and ret<strong>in</strong>opathy) were different from that of the study population as awhole (see Table 3.3). Thus, the DCCT patients were not representative of people with type1 diabetes, whereas the education programme undertaken <strong>in</strong> Dusseldorf is not someth<strong>in</strong>gthat is widely replicated <strong>in</strong> ma<strong>in</strong>stream diabetes practice. Risk factors for hypoglycaemiamay differ accord<strong>in</strong>g to the specific characteristics of <strong>in</strong>dividuals be<strong>in</strong>g studied.Thus, <strong>in</strong> conclusion, although most specialists would accept that severe hypoglycaemia ismore common <strong>in</strong> patients receiv<strong>in</strong>g <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy, it is not <strong>in</strong>evitable and betterpatient education may actually reduce the <strong>in</strong>cidence.Strict Glycaemic ControlStrict glycaemic control, as evidenced by glycated haemoglob<strong>in</strong> concentrations that approachthe upper end of the non-diabetic range, is closely l<strong>in</strong>ked to <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy. S<strong>in</strong>cethe DCCT results were published, glycated haemoglob<strong>in</strong> concentrations at or around thislevel have become the usual target for many people with type 1 diabetes. In the DCCT, therewas a quadratic relationship between HbA 1c and risk of severe hypoglycaemia (Figure 3.5),with the risk of hypoglycaemia <strong>in</strong>creas<strong>in</strong>g as HbA 1c decreased (The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1997). However, the atta<strong>in</strong>ed HbA 1c did not accountfor all the difference <strong>in</strong> risk of severe hypoglycaemia between the two arms of the study,as subjects <strong>in</strong> the <strong>in</strong>tensive group still had an excess risk of severe hypoglycaemia after10080Rate per 100 Patient Years604020*** **0* * ** * * **5 6 7 8 9 10 11 12 13 14HbA 1c (%) Dur<strong>in</strong>g StudyFigure 3.5 Risk of severe hypoglycaemia as a function of monthly updated HbA 1c <strong>in</strong> the <strong>Diabetes</strong>Control and Complications Trial. The circles represent data from the <strong>in</strong>tensive group and asterisks datafrom the conventional group. The bold solid and bold dashed l<strong>in</strong>es represent the regression plots foreach group, and the non-bold dashed l<strong>in</strong>es on either side show the upper and lower 95% confidencebands; DCCT (1997). Copyright © 1997 American <strong>Diabetes</strong> Association. Repr<strong>in</strong>ted with permissionfrom The American <strong>Diabetes</strong> Association

68 FREQUENCY, CAUSES AND RISK FACTORSstatistical adjustment for HbA 1c concentration. Indeed, <strong>in</strong> the <strong>in</strong>tensive group, only about 5%of the variation <strong>in</strong> frequency of severe hypoglycaemia could be expla<strong>in</strong>ed by the glycatedhaemoglob<strong>in</strong> concentration (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1997).Other studies have reported variable relationships between glycaemic control andfrequency of severe hypoglycaemia. In the study by Bott et al., (1997) a lower mean HbA 1cwas associated with severe hypoglycaemia, but there was no l<strong>in</strong>ear or quadratic relationshipbetween HbA 1c and severe hypoglycaemia. In the EURODIAB IDDM ComplicationsStudy, severe hypoglycaemia (def<strong>in</strong>ed as episodes requir<strong>in</strong>g third party assistance) occurred<strong>in</strong> 32% of <strong>in</strong>dividuals over 12 months. A clear relationship to glycated haemoglob<strong>in</strong> wasevident, <strong>in</strong> that 40% of <strong>in</strong>dividuals with an HbA 1c < 54% were affected compared with24% of <strong>in</strong>dividuals with a HbA 1c > 78% (The EURODIAB IDDM Complications StudyGroup, 1994). In the study of workplace hypoglycaemia, recurrent severe hypoglycaemiawas associated with strict glycaemic control, but no l<strong>in</strong>k was found <strong>in</strong> <strong>in</strong>dividuals who hadexperienced only one episode (Leckie et al., 2005). In several other studies, no relationshipwas observed between severe hypoglycaemia and glycated haemoglob<strong>in</strong> (Muhlhauseret al., 1985; Muhlhauser et al., 1987; MacLeod et al., 1993; Gold et al., 1997; Muhlhauseret al., 1998; ter Braak et al., 2000; Pedersen-Bjergaard et al., 2001; Leese et al., 2003;Pedersen-Bjergaard et al., 2003a; Pedersen-Bjergaard et al., 2004) after adjustment for otherrisk factors.Glycated haemoglob<strong>in</strong> does not, of course, provide the entire picture about an <strong>in</strong>dividual’sglycaemic control and although HbA 1c may not predict risk of hypoglycaemia, low meanhome blood glucose concentrations and excessive variability <strong>in</strong> blood glucose can identify<strong>in</strong>dividuals more prone to hypoglycaemia (Cox et al., 1994; Janssen et al., 2000).Thus, a straightforward relationship does not exist between severe hypoglycaemia andglycaemic control. For an episode of mild hypoglycaemia to progress to one that causessignificant neuroglycopenia and impairs consciousness, other factors must operate, whichnegate the normal symptomatic and hormonal responses to hypoglycaemia.Acquired <strong>Hypoglycaemia</strong> SyndromesWith<strong>in</strong> each treatment group of the DCCT, the number of previous episodes of severehypoglycaemia was the strongest predictor of risk of future episodes (The <strong>Diabetes</strong> Controland Complications Trial Research Group, 1997). Moreover, approximately 30% of patients<strong>in</strong> each group experienced a second episode of severe hypoglycaemia with<strong>in</strong> four monthsfollow<strong>in</strong>g an <strong>in</strong>itial episode. The importance of a previous history of severe hypoglycaemia<strong>in</strong> predict<strong>in</strong>g future risk has been replicated <strong>in</strong> several other studies (MacLeod et al., 1993;Bott et al., 1997; Gold et al., 1997; Muhlhauser et al., 1998). Moreover, as demonstrated <strong>in</strong>Table 3.3, many studies have also l<strong>in</strong>ked <strong>in</strong>creased duration of diabetes with an <strong>in</strong>creasedrisk of severe hypoglycaemia.Cryer has suggested that the <strong>in</strong>tegrity of the glucose counterregulatory system may be apivotal factor <strong>in</strong> determ<strong>in</strong><strong>in</strong>g whether the relative or absolute hyper<strong>in</strong>sul<strong>in</strong>ism that frequentlyoccurs <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetes ultimately results <strong>in</strong> the development of hypoglycaemia(Cryer, 1994; Cryer et al., 2003). Three acquired hypoglycaemia syndromes are associatedwith an <strong>in</strong>creased risk of severe hypoglycaemia <strong>in</strong> people with type 1 diabetes. These areconsidered <strong>in</strong> greater detail <strong>in</strong> Chapters 6 and 7.

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 69Counterregulatory hormonal deficiencies<strong>Hypoglycaemia</strong>-<strong>in</strong>duced secretion of glucagon decl<strong>in</strong>es <strong>in</strong> most patients with<strong>in</strong> five years ofdevelop<strong>in</strong>g type 1 diabetes (Gerich et al., 1973; Bolli et al., 1983). A defective ep<strong>in</strong>ephr<strong>in</strong>eresponse to hypoglycaemia may develop some years later (Bolli et al., 1983; Hirsch andShamoon, 1987; Dagogo-Jack et al., 1993). As with the glucagon response, the impairedep<strong>in</strong>ephr<strong>in</strong>e response is hypoglycaemia-specific but, <strong>in</strong> contrast to glucagon, it exhibits athreshold effect – i.e., an ep<strong>in</strong>ephr<strong>in</strong>e response can still be elicited by hypoglycaemia, but onlyat a lower blood glucose concentration (Dagogo-Jack et al., 1993). If hypoglycaemia develops<strong>in</strong> patients who have this comb<strong>in</strong>ed counterregulatory hormonal deficiency, glucose recoverymay be severely compromised (see Chapter 6). Subject<strong>in</strong>g such patients to <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong>therapy <strong>in</strong>creased the risk of severe hypoglycaemia by 25 times, compared with subjectswho had an <strong>in</strong>tact ep<strong>in</strong>ephr<strong>in</strong>e response (White et al., 1983).Impaired awareness of hypoglycaemiaIn many patients with <strong>in</strong>sul<strong>in</strong>-treated diabetes, the hypoglycaemia symptom profile alterswith time, result<strong>in</strong>g <strong>in</strong> impaired perception of the onset of hypoglycaemia (see Chapter 7).Commonly, autonomic warn<strong>in</strong>g symptoms are dim<strong>in</strong>ished and neuroglycopenic symptomspredom<strong>in</strong>ate. Around 25% of people with type 1 diabetes develop impaired awareness ofhypoglycaemia and the prevalence of this problem <strong>in</strong>creases with the duration of <strong>in</strong>sul<strong>in</strong>treatment (Hepburn et al., 1990; Gerich et al., 1991; Pramm<strong>in</strong>g et al., 1991). Prospectivestudies have demonstrated that the frequency of severe hypoglycaemia is <strong>in</strong>creased up tosix-fold <strong>in</strong> patients with impaired awareness compared to those with normal hypoglycaemiaawareness (Gold et al., 1994; Clarke et al., 1995).<strong>Hypoglycaemia</strong>-associated central autonomic failureThe above acquired hypoglycaemia syndromes tend to segregate together cl<strong>in</strong>ically. Patientswith glycated haemoglob<strong>in</strong> concentrations close to the non-diabetic range are at greater riskof develop<strong>in</strong>g impaired awareness (Boyle et al., 1995; K<strong>in</strong>sley et al., 1995; Pampanelli et al.,1996), while the glycaemic thresholds for the onset of symptoms and responses are altered<strong>in</strong> patients with impaired awareness (Grimaldi et al., 1990; Hepburn et al., 1991; Mokanet al., 1994; Bacatselos et al., 1995). Cryer has suggested that these acquired abnormalitiesrepresent a form of central ‘<strong>Hypoglycaemia</strong>-Associated Autonomic Failure’ (HAAF) <strong>in</strong> type1 diabetes, speculat<strong>in</strong>g that recurrent severe hypoglycaemia is the primary cause (Cryer,1992; Cryer et al., 2003). If hypoglycaemia is the precipitant, then it is possible to see how avicious cycle may become established with the development of the acquired hypoglycaemiasyndromes promot<strong>in</strong>g further episodes of severe hypoglycaemia.SummaryThere is a well-known adage that ‘hypoglycaemia begets hypoglycaemia’. People who haveexperienced one episode of severe hypoglycaemia are much more likely to develop furtherepisodes, and the greatest risk occurs <strong>in</strong> the weeks and months after the <strong>in</strong>dex event. Severehypoglycaemia becomes a more common problem <strong>in</strong> people with long-stand<strong>in</strong>g type 1

70 FREQUENCY, CAUSES AND RISK FACTORSdiabetes (UK <strong>Hypoglycaemia</strong> Group, 2007). This is the legacy of the burden ofhypoglycaemia that such <strong>in</strong>dividuals have experienced over many years with diabetes.The impaired symptomatic and counterregulatory responses to hypoglycaemia dramatically<strong>in</strong>crease the likelihood that an episode of mild hypoglycaemia will progress to a more severeevent.Genetic Predisposition to <strong>Hypoglycaemia</strong>Most of the precipitants and risk factors for hypoglycaemia have been known about formany years. In 2001, Pedersen-Bjergaard et al. (2001) reported a novel risk factor for severehypoglycaemia <strong>in</strong> adults with type 1 diabetes and raised the notion that some <strong>in</strong>dividualsmay have an <strong>in</strong>herent genetic susceptibility to hypoglycaemia. The Danish <strong>in</strong>vestigatorsnoted the similarity between endurance exercise and hypoglycaemia <strong>in</strong> that both are statesof limited metabolic fuel availability. Previous studies had l<strong>in</strong>ked exercise performance to aparticular polymorphism of the angiotens<strong>in</strong>-convert<strong>in</strong>g enzyme (ACE) gene. Specifically, the<strong>in</strong>sertion (I) allele, which resulted <strong>in</strong> low serum ACE activity, was associated with superiorperformance capacity compared with the deletion (D) allele. In an <strong>in</strong>itial retrospective surveyof 207 adults with type 1 diabetes, patients with the DD genotype had a 3.2-fold <strong>in</strong>creasedrisk of severe hypoglycaemia <strong>in</strong> the preced<strong>in</strong>g two years, compared with <strong>in</strong>dividuals withthe II genotype. There was also a significant relationship between serum ACE activity,with a 1.4 <strong>in</strong>crement <strong>in</strong> risk of severe hypoglycaemia for every 10 U/l rise <strong>in</strong> serum ACEconcentration (Figure 3.6). The serum ACE activity was directly l<strong>in</strong>ked to ACE genotype,and it rema<strong>in</strong>ed a significant risk factor even after adjustment for conventional risk factors.Moreover, serum ACE activity was a stronger risk factor for severe hypoglycaemia <strong>in</strong>C-peptide negative <strong>in</strong>dividuals with impaired awareness, than <strong>in</strong> other groups (relative risk1.7 per 10 U/l; Figure 3.7). No significant relationship was observed between serum ACEactivity or genotype and frequency of mild hypoglycaemia.Figure 3.6 Risk of severe hypoglycaemia accord<strong>in</strong>g to serum ACE activity <strong>in</strong> 207 patients withtype 1 diabetes, untreated with ACE <strong>in</strong>hibitors or angiotens<strong>in</strong>-2 receptor antagonists. Broken l<strong>in</strong>esrepresent 95% confidence <strong>in</strong>tervals. Repr<strong>in</strong>ted from The Lancet, Pedersen-Bjergaard et al. (2001) withpermission from Elsevier

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 71Figure 3.7 Association between severe hypoglycaemia and serum ACE activity accord<strong>in</strong>g toC-peptide status and self-estimated awareness of hypoglycaemia. Repr<strong>in</strong>ted from The Lancet, Pedersen-Bjergaard et al. (2001) with permission from ElsevierThese f<strong>in</strong>d<strong>in</strong>gs were replicated <strong>in</strong> a prospective study for one year <strong>in</strong> 107 adults (Pedersen-Bjergaard et al., 2003a). Serum ACE activity <strong>in</strong> the fourth quartile was associated with a2.7-fold <strong>in</strong>creased risk of severe hypoglycaemia compared to activity <strong>in</strong> the lowest quartile.Compared to subjects with the II genotype, <strong>in</strong>dividuals with the DD genotype had a 1.8-fold<strong>in</strong>creased risk of severe hypoglycaemia, although this did not reach statistical significance.Higher serum ACE concentrations were also associated with an <strong>in</strong>creased risk of severehypoglycaemia <strong>in</strong> Swedish children and adolescents with type 1 diabetes (Nordfeldt andSamuelsson, 2003).The Danish group have put forward a number of possible mechanisms to expla<strong>in</strong> theirobservations (Pedersen-Bjergaard et al., 2001; Pedersen-Bjergaard et al., 2003a). They speculatethat low levels of serum ACE may be associated with less cognitive deterioration dur<strong>in</strong>gacute hypoglycaemia, thereby <strong>in</strong>creas<strong>in</strong>g the likelihood that remedial action to correct hypoglycaemiais taken. Alternatively, low serum ACE activity might enhance counterregulationor reduce production of toxic substances, e.g. reactive oxygen species, dur<strong>in</strong>g hypoglycaemia.All these putative mechanisms rema<strong>in</strong> highly speculative, but the authors also raiseone other <strong>in</strong>trigu<strong>in</strong>g possibility: namely that ACE <strong>in</strong>hibition might reduce the frequency ofhypoglycaemia. This may seem counter<strong>in</strong>tuitive, because previous population-based studiessuggested an association between severe hypoglycaemia and use of ACE <strong>in</strong>hibitors (Her<strong>in</strong>gset al., 1996; Morris et al., 1997). However, these studies are flawed and a re-exam<strong>in</strong>ation ofthe role of ACE <strong>in</strong>hibitors <strong>in</strong> ameliorat<strong>in</strong>g the impact of hypoglycaemia seems warranted.Recent data from one centre <strong>in</strong> Scotland have not demonstrated such a strong l<strong>in</strong>k betweenserum ACE and severe hypoglycaemia <strong>in</strong> adults with type 1 diabetes (Zammitt et al., 2007),nor has a study of children and adolescents with type 1 diabetes <strong>in</strong> Australia (Bulsaraet al., 2007). This should, therefore, serve as a rem<strong>in</strong>der that genetic susceptibility to anybiological variable may not be the same <strong>in</strong> different populations and re<strong>in</strong>forces the need forthe Danish observations to be exam<strong>in</strong>ed <strong>in</strong> other countries and ethnic groups. However, thestudies by Pedersen-Bjergaard and colleagues raise the <strong>in</strong>trigu<strong>in</strong>g prospect that there maybe yet other genetic factors that <strong>in</strong>fluence susceptibility to hypoglycaemia. Identification of

72 FREQUENCY, CAUSES AND RISK FACTORSthese may help stratify risk <strong>in</strong> <strong>in</strong>dividuals with type 1 diabetes and may ultimately lead tothe development of novel therapeutic <strong>in</strong>terventions to prevent or ameliorate the impact ofhypoglycaemia.Absent Endogenous Insul<strong>in</strong> SecretionSeveral studies have demonstrated the importance of endogenous <strong>in</strong>sul<strong>in</strong> secretion <strong>in</strong> def<strong>in</strong><strong>in</strong>grisk of hypoglycaemia. Individuals who are C-peptide negative, i.e., who have no endogenous<strong>in</strong>sul<strong>in</strong> production, have an approximately two- to fourfold <strong>in</strong>creased risk of severe hypoglycaemiacompared to people with detectable C-peptide (Bott et al., 1997; The <strong>Diabetes</strong> Controland Complications Trial Research Group, 1997; Muhlhauser et al., 1998; Pedersen-Bjergaardet al., 2001). These data mirror the cl<strong>in</strong>ical experience that severe hypoglycaemia is rare <strong>in</strong>the 12 months after the diagnosis of type 1 diabetes (Davis et al., 1997), when significantconcentrations of endogenous <strong>in</strong>sul<strong>in</strong> can be measured. The presumption is that the abilityof endogenous <strong>in</strong>sul<strong>in</strong> levels to fall <strong>in</strong> the face of a decl<strong>in</strong><strong>in</strong>g blood glucose concentrationprovides an additional layer of protection <strong>in</strong> the defences aga<strong>in</strong>st hypoglycaemia and thusreduces overall risk.Dose and Type of Insul<strong>in</strong>Few diabetes specialists who were practis<strong>in</strong>g <strong>in</strong> the mid-1980s will forget the controversythat surrounded the <strong>in</strong>troduction of human <strong>in</strong>sul<strong>in</strong>. A substantial and vocal m<strong>in</strong>ority ofpatients with type 1 diabetes claimed that the change from animal-derived to human <strong>in</strong>sul<strong>in</strong>was associated with a loss of warn<strong>in</strong>g symptoms of hypoglycaemia and thus an <strong>in</strong>creasedrisk. These claims were subject to considerable scientific scrut<strong>in</strong>y and ultimately a systematicreview of the evidence found no evidence to support the premise that treatment with human,as opposed to animal, <strong>in</strong>sul<strong>in</strong> was associated with an <strong>in</strong>creased risk of hypoglycaemia (Aireyet al., 2000).S<strong>in</strong>ce that time, several <strong>in</strong>sul<strong>in</strong> analogues have been developed and their <strong>in</strong>troduction<strong>in</strong>to cl<strong>in</strong>ical practice has been accompanied by the publication of studies that purport todemonstrate that use of the <strong>in</strong>sul<strong>in</strong> analogues is associated with a lower frequency ofhypoglycaemia, <strong>in</strong> the face of stable or improved glycaemic control (Anderson et al., 1997;Garg et al., 2004; Hermansen et al., 2004). However, recent systematic reviews suggestthat neither short-act<strong>in</strong>g (Siebenhofer et al., 2004) nor long-act<strong>in</strong>g (Warren et al., 2004)<strong>in</strong>sul<strong>in</strong> analogues are associated with cl<strong>in</strong>ically significant lower rates of hypoglycaemia. Theusual caveats of such cl<strong>in</strong>ical trials apply <strong>in</strong> terms of patient characteristics and record<strong>in</strong>gof hypoglycaemia. In one observational study from Colorado, the frequency of severehypoglycaemia rose <strong>in</strong> cl<strong>in</strong>ic patients <strong>in</strong> the immediate aftermath of the DCCT as effortsto <strong>in</strong>tensify <strong>in</strong>sul<strong>in</strong> therapy were <strong>in</strong>stituted, but from 1996 onwards, the rates of severehypoglycaemia decl<strong>in</strong>ed and the authors l<strong>in</strong>ked this to the <strong>in</strong>troduction of <strong>in</strong>sul<strong>in</strong> lispro(Chase et al., 2001). This study was subject to the effects of numerous confounders, butfurther data from rout<strong>in</strong>e cl<strong>in</strong>ical practice are required to help clarify the impact of <strong>in</strong>sul<strong>in</strong>analogues on risk of hypoglycaemia. The <strong>in</strong>troduction of <strong>in</strong>haled <strong>in</strong>sul<strong>in</strong> (Exubera) has notbeen associated with a lower risk of hypoglycaemia <strong>in</strong> either type 1 or type 2 diabetes; whencompared with <strong>in</strong>sul<strong>in</strong> adm<strong>in</strong>istered by the subcutaneous route, rates of severe hypoglycaemiawere either equivalent (Hollander et al., 2004) or higher (Skyler et al., 2005).

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA 73Higher doses of <strong>in</strong>sul<strong>in</strong> have also been associated with an <strong>in</strong>creased risk of hypoglycaemia<strong>in</strong> some studies (The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1997;ter Braak et al., 2000), but not <strong>in</strong> others (Table 3.3). Several possible explanations mayaccount for this relationship – high <strong>in</strong>sul<strong>in</strong> doses may be a sign of less endogenous <strong>in</strong>sul<strong>in</strong>production, or of efforts to achieve strict glycaemic control. It may also reflect sub-optimalcompliance with <strong>in</strong>sul<strong>in</strong> therapy and so <strong>in</strong>dicate a pattern of behaviour that predisposes tomarked fluctuations <strong>in</strong> glucose control.SleepAs has already been discussed <strong>in</strong> the section on asymptomatic hypoglycaemia, nocturnalhypoglycaemia is very common (Vervoort et al., 1996). In the DCCT, 43% of episodesof severe hypoglycaemia occurred between midnight and 8.00 a.m., and 55% of episodesoccurred when <strong>in</strong>dividuals were asleep (The DCCT Research Group, 1991). The potentialfor nocturnal hypoglycaemia engenders significant anxiety among patients, particularly <strong>in</strong><strong>in</strong>dividuals who live alone. Patients worry that they will not awaken when hypoglycaemiaoccurs and that they may be left <strong>in</strong>capacitated or die as a consequence. Many <strong>in</strong>dividuals,as a result, ma<strong>in</strong>ta<strong>in</strong> higher blood glucose concentrations at bedtime to reduce the risk ofnocturnal hypoglycaemia. <strong>Hypoglycaemia</strong> appears to be more common at night becausecounterregulatory hormone responses are blunted dur<strong>in</strong>g sleep <strong>in</strong> people with type 1 diabetes(Jones et al., 1998; Banarer and Cryer, 2003). Moreover, hypoglycaemia awareness is alsoreduced dur<strong>in</strong>g sleep (Banarer and Cryer, 2003), and so ultimately sleep impairs both thephysiological and behavioural responses to hypoglycaemia. Unsurpris<strong>in</strong>gly, considerableeffort has been directed at develop<strong>in</strong>g strategies to reduce the risk of nocturnal hypoglycaemiaand these are discussed <strong>in</strong> Chapter 4.Microvascular ComplicationsIn a retrospective study of 44 patients with type 1 diabetes with impaired renal function(serum creat<strong>in</strong><strong>in</strong>e ≥ 133 umol/l and prote<strong>in</strong>uria), the <strong>in</strong>cidence of severe hypoglycaemiawas five times higher than <strong>in</strong> matched subjects with normal renal function (Muhlhauseret al., 1991). In other studies, severe hypoglycaemia was l<strong>in</strong>ked with nephropathy (terBraak et al., 2000), peripheral neuropathy and ret<strong>in</strong>opathy (Pedersen-Bjergaard et al., 2004).Although it is generally recognised that <strong>in</strong>sul<strong>in</strong> requirement decl<strong>in</strong>es <strong>in</strong> advanced renaldisease, with reduced clearance of <strong>in</strong>sul<strong>in</strong>, this association between hypoglycaemia andnephropathy (and other microvascular complications) could be confounded by many otherfactors, e.g. concomitant drug therapy (ter Braak et al., 2000) and the co-existence ofacquired hypoglycaemia syndromes which, like microvascular complications, are l<strong>in</strong>ked with<strong>in</strong>creas<strong>in</strong>g duration of diabetes.The role of peripheral autonomic neuropathy <strong>in</strong> <strong>in</strong>creas<strong>in</strong>g the risk of severe hypoglycaemiahas been considered <strong>in</strong> several studies and deserves mention. In the EURODIABIDDM Complications Study, the presence of abnormal cardiovascular reflexes was associatedwith a 1.7-fold <strong>in</strong>creased risk of severe hypoglycaemia (Stephenson et al., 1996).Gold et al. (1997) also demonstrated that autonomic neuropathy was associated with asmall <strong>in</strong>creased risk of severe hypoglycaemia <strong>in</strong> 60 patients with type 1 diabetes, but nosuch relationship was demonstrated <strong>in</strong> the DCCT (The DCCT Research Group, 1991). The

74 FREQUENCY, CAUSES AND RISK FACTORSmechanism underly<strong>in</strong>g this association rema<strong>in</strong>s to be fully elucidated. Peripheral autonomicneuropathy often coexists with impaired hypoglycaemia awareness <strong>in</strong> patients with type 1diabetes, presumably because both conditions are associated with diabetes of long duration(Hepburn et al., 1990), but impaired awareness can readily occur <strong>in</strong> the absence of peripheralautonomic neuropathy (Hepburn et al., 1990; Ryder et al., 1990; Bacatselos et al., 1995).It is well established that severe autonomic neuropathy is associated with ‘gastroparesisdiabeticorum’, which can cause marked sw<strong>in</strong>gs <strong>in</strong> blood glucose concentration. Althoughsuch severe gastroparesis is now rare, delayed gastric empty<strong>in</strong>g is relatively common (Konget al., 1996) and may expla<strong>in</strong> at least part of the association between hypoglycaemia andautonomic neuropathy.Social and Psychological FactorsPsychological factors clearly play a crucial role <strong>in</strong> determ<strong>in</strong><strong>in</strong>g an <strong>in</strong>dividual’s likelihood ofdevelop<strong>in</strong>g severe hypoglycaemia. Low mood (Gonder-Frederick and Cox, 1997), emotionalcop<strong>in</strong>g (Bott et al., 1997) and socio-economic status (Muhlhauser et al., 1998; Leese et al.,2003) have been l<strong>in</strong>ked to risk of severe hypoglycaemia and so too have other morestraightforward behavioural factors such an <strong>in</strong>dividual’s propensity to carry a supply ofcarbohydrate for emergency use (Bott et al., 1997) and their determ<strong>in</strong>ation to achievenormoglycaemia (Muhlhauser et al., 1998).S<strong>in</strong>ce the 1980s, Cox and colleagues at the University of Virg<strong>in</strong>ia, USA, have carried outsem<strong>in</strong>al <strong>in</strong>vestigations to explore the psychological impact of hypoglycaemia on people withtype 1 diabetes. They developed a Fear of <strong>Hypoglycaemia</strong> scale, which sought to quantifythe anxieties that people with type 1 diabetes have with respect to hypoglycaemia and theextent to which <strong>in</strong>dividuals take steps to avoid experienc<strong>in</strong>g such episodes (Cox et al., 1987).Unsurpris<strong>in</strong>gly, there is a close association between fear of hypoglycaemia and perceivedrisk of future severe hypoglycaemia (Gonder-Frederick et al., 1997). In many <strong>in</strong>stances thisis appropriate, <strong>in</strong> that ‘fear’ rat<strong>in</strong>gs are often high <strong>in</strong> people who have impaired awarenessof hypoglycaemia and/or have experienced multiple episodes <strong>in</strong> the past (Gold et al., 1996).However, <strong>in</strong> other people, fear of hypoglycaemia may be high while absolute risk is low –such <strong>in</strong>dividuals often display high levels of trait anxiety or have had a traumatic previousexperience of hypoglycaemia. It is often extremely difficult to persuade such <strong>in</strong>dividuals toma<strong>in</strong>ta<strong>in</strong> strict glycaemic control. Conversely, there are people who have a very low fearof hypoglycaemia, despite a propensity to recurrent episodes. Such <strong>in</strong>dividuals may fail totake appropriate precautionary measures, thereby putt<strong>in</strong>g themselves and others at risk ifhypoglycaemia occurs, for example, when that <strong>in</strong>dividual is driv<strong>in</strong>g a car.Endocr<strong>in</strong>opathiesEndocr<strong>in</strong>e disorders, such as Addison’s disease and hypopituitarism, which are associatedwith a deficiency of counterregulatory hormones, can be associated with an <strong>in</strong>creased riskof hypoglycaemia <strong>in</strong> adults with type 1 diabetes. These are uncommon <strong>in</strong> everyday diabetespractice, but cl<strong>in</strong>icians should ma<strong>in</strong>ta<strong>in</strong> a high <strong>in</strong>dex of suspicion particularly <strong>in</strong> the patientwho simultaneously demonstrates a marked and otherwise unexpla<strong>in</strong>ed decl<strong>in</strong>e <strong>in</strong> <strong>in</strong>sul<strong>in</strong>requirements.

CONCLUSIONS 75Other Risk FactorsThe list of other possible risk factors for hypoglycaemia is a long one. Smok<strong>in</strong>g is a relativelynovel marker of <strong>in</strong>creased risk (Pedersen-Bjergaard et al., 2004), but may be confoundedby other diabetes-related and lifestyle factors, e.g. regular use of alcohol (ter Braak et al.,2000). In the DCCT, men and adolescents were at <strong>in</strong>creased risk (The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1997), but these associations have not been replicatedwith any consistency <strong>in</strong> other studies (Table 3.3). Risks associated with pregnancy areaddressed <strong>in</strong> Chapter 10.Causes and Risk Factors for <strong>Hypoglycaemia</strong>: Summary andConclusionsMany of the risk factors described above are <strong>in</strong>ter-related and any given <strong>in</strong>dividual mayhave more than one, which clearly will <strong>in</strong>crease their overall risk. However, at its mostfundamental level, hypoglycaemia <strong>in</strong> adults with type 1 diabetes is a consequence of the<strong>in</strong>ability of exogenous <strong>in</strong>sul<strong>in</strong> levels to fall <strong>in</strong> response to a decl<strong>in</strong><strong>in</strong>g blood glucose concentration.In people who have had type 1 diabetes for several years, the situation is exacerbatedbecause of a failure of the normal physiological counterregulatory defence mechanisms,which <strong>in</strong> non-diabetic <strong>in</strong>dividuals serve to <strong>in</strong>crease exogenous glucose production andgenerate warn<strong>in</strong>g symptoms. This counterregulatory failure worsens with <strong>in</strong>creas<strong>in</strong>g duration ofdiabetes, particularly if glycaemic control is strict and the <strong>in</strong>dividual has experienced recurrentepisodes of severe hypoglycaemia. Other factors may come <strong>in</strong> to play, for example particularbehavioural patterns and the time-action profiles of the exogenous <strong>in</strong>sul<strong>in</strong>. Recent data alsoraise the possibility that there may be <strong>in</strong>herent genetic susceptibility to hypoglycaemia, butthis needs to be affirmed <strong>in</strong> wider populations and the underly<strong>in</strong>g mechanisms more clearlydissected. Novel strategies to reduce overall risk of hypoglycaemia are urgently required.CONCLUSIONS• There are no def<strong>in</strong>itive criteria for what constitutes hypoglycaemia, but most specialistsdist<strong>in</strong>guish mild from severe episodes depend<strong>in</strong>g on whether or not the <strong>in</strong>dividual is ableto self-treat.• The ‘average’ adult with type 1 diabetes will experience many thousands of episodes of mildhypoglycaemia over a lifetime, with a typical frequency of one to two episodes per week.• Severe hypoglycaemia is less common, and on average occurs once or twice every year,with an annual prevalence of around 30%. However, the distribution is heavily skewed,such that many <strong>in</strong>dividuals are unaffected over a calendar year, while a small numberexperience recurrent episodes.• Severe hypoglycaemia requir<strong>in</strong>g treatment with <strong>in</strong>tramuscular glucagon and/or <strong>in</strong>travenousglucose is even less common, and the majority of all episodes of hypoglycaemia aremanaged <strong>in</strong> the community by the patient and/or relatives and friends, without recourseto emergency services.

76 FREQUENCY, CAUSES AND RISK FACTORS• <strong>Hypoglycaemia</strong> occurs when there is an imbalance between <strong>in</strong>sul<strong>in</strong>-mediated glucosedisposal and glucose <strong>in</strong>flux <strong>in</strong>to the circulation from the liver and exogenous carbohydrate.Typical precipitants <strong>in</strong>clude patient error <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dosage, alcohol andexercise.• Plasma concentrations of exogenous <strong>in</strong>sul<strong>in</strong> cannot decl<strong>in</strong>e <strong>in</strong> response to fall<strong>in</strong>g bloodglucose and the time action profiles of current <strong>in</strong>sul<strong>in</strong>s do not accurately mimic the normalphysiological variation <strong>in</strong> <strong>in</strong>sul<strong>in</strong>. In a high proportion of cases, no underly<strong>in</strong>g precipitantof a given episode of hypoglycaemia can be identified.• The DCCT and other <strong>in</strong>tervention studies have provided substantial <strong>in</strong>formation of theepidemiology of severe hypoglycaemia, but <strong>in</strong>dividuals participat<strong>in</strong>g <strong>in</strong> such studies maynot be representative of the wider population who have type 1 diabetes.• A major determ<strong>in</strong>ant of <strong>in</strong>creased risk of severe hypoglycaemia is <strong>Hypoglycaemia</strong>-Associated Autonomic Failure, which is a consequence of exposure to recurrent episodesof hypoglycaemia. This is a feature of <strong>in</strong>dividuals with long-stand<strong>in</strong>g diabetes, <strong>in</strong>tensive<strong>in</strong>sul<strong>in</strong> therapy and strict glycaemic control.• Recent data suggest that there may be specific genetic factors that predispose to an<strong>in</strong>creased risk of hypoglycaemia.REFERENCESAirey CM, Williams DRR, Mart<strong>in</strong> PG, Bennett CMT, Spoor PA (2000). <strong>Hypoglycaemia</strong> <strong>in</strong>duced byexogenous <strong>in</strong>sul<strong>in</strong> – ‘human’ and animal <strong>in</strong>sul<strong>in</strong> compared. Diabetic Medic<strong>in</strong>e 17: 416–32.Allen C, LeClaire T, Palta M, Daniels K, Meredith M, D’Alessio DJ, for the Wiscons<strong>in</strong> <strong>Diabetes</strong>Registry Project (2001). Risk factors for frequent and severe hypoglycemia <strong>in</strong> type 1 diabetes.<strong>Diabetes</strong> Care 24: 1878–81.American <strong>Diabetes</strong> Association Workgroup on Hypoglycemia (2005). Def<strong>in</strong><strong>in</strong>g and Report<strong>in</strong>g Hypoglycemia<strong>in</strong> <strong>Diabetes</strong>. <strong>Diabetes</strong> Care 28: 1245–9.Anderson JH, Brunelle RL, Koivisto VA, Pfutzner A, Trautmann ME, Vignati L, DiMarchi R, TheMulticenter Insul<strong>in</strong> Lispro Study Group (1997). Reduction of postprandial hyperglycemia andfrequency of hypoglycemia <strong>in</strong> IDDM patients on <strong>in</strong>sul<strong>in</strong>-analog treatment. <strong>Diabetes</strong> 46: 265–70.Arky RA, Veverbrants E, Abramson EA (1968). Irreversible hypoglycemia. A complication of alcoholand <strong>in</strong>sul<strong>in</strong>. Journal of the American Medical Association 206: 575–8.Avogaro A, Beltramello P, Gnudi L, Maran A, Valerio A, Miola M et al. (1993). Alcohol <strong>in</strong>takeimpairs glucose counterregulation dur<strong>in</strong>g acute <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> IDDM patients.Evidence for a critical role of free fatty acids. <strong>Diabetes</strong> 42: 1626–34.Bacatselos SO, Karamitsos DT, Kourtoglou GI, Zambulis CX, Yovos JG, Vyzantiadis AT (1995).<strong>Hypoglycaemia</strong> unawareness <strong>in</strong> type 1 diabetic patients under conventional <strong>in</strong>sul<strong>in</strong> treatment.<strong>Diabetes</strong> Nutrition and Metabolism 8: 267–75.Banarer S, Cryer PE (2003). Sleep-related hypoglycemia-associated autonomic failure <strong>in</strong> type 1diabetes. <strong>Diabetes</strong> 52: 1195–1203.Bode BW, Schwartz S, Stubbs HA, Block JE (2005). Glycemic characteristics <strong>in</strong> cont<strong>in</strong>uously monitoredpatients with type 1 and type 2 diabetes. Normative values. <strong>Diabetes</strong> Care 28: 2361–6.Bolli G, De Feo P, Compagnucci P, Cartech<strong>in</strong>i MG, Angeletti G, Santeusanio F et al. (1983).Abnormal glucose counterregulation <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Interaction of anti<strong>in</strong>sul<strong>in</strong>antibodies and impaired glucagon and ep<strong>in</strong>ephr<strong>in</strong>e secretion. <strong>Diabetes</strong> 32: 134–41.

REFERENCES 77Bott S, Bott U, Berger M, Muhlhauser I (1997). Intensified <strong>in</strong>sul<strong>in</strong> therapy and the risk of severehypoglycaemia. Diabetologia 40: 926–32.Boyle PJ, Schwartz NS, Shah SD, Clutter WE, Cryer PE (1988). Plasma glucose concentrations at theonset of hypoglycemic symptoms <strong>in</strong> patients with poorly controlled diabetes and <strong>in</strong> non-diabetics.New England Journal of Medic<strong>in</strong>e 318: 1487–92.Boyle PJ, Kempers SF, O’Connor AM, Nagy RJ (1995). Bra<strong>in</strong> glucose uptake and unawarenessof hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 333: 1726–31.Bulsara MK, Holman CDJ, van Bockxmeer FM, Davis EA, Gallego PH, Beilby JP, Palmer LJet al. (2007). The relationship between ACE genotype and risk of severe hypoglycemia <strong>in</strong> a largepopulation-based cohort of children and adolescents with type 1 diabetes. Diabetologia 50: 965–71.Chase HP, Lockspeiser T, Peery B, Shepherd M, MacKenzie T, Anderson J, Garg SK (2001). Theimpact of the <strong>Diabetes</strong> Control and Complications Trial and Humalog <strong>in</strong>sul<strong>in</strong> on glycohemoglob<strong>in</strong>levels and severe hypoglycemia <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care 24: 430–4.Clarke WL, Cox DJ, Gonder-Frederick LA, Julian D, Schlundt D, Polonsky W (1995). Reducedawareness of hypoglycemia <strong>in</strong> adults with IDDM. A prospective study of hypoglycemic frequencyand associated symptoms. <strong>Diabetes</strong> Care 18: 517–22.Cox DJ, Irv<strong>in</strong>e A, Gonder-Frederick L, Nowacek G, Butterfield J (1987). Fear of hypoglycemia:quantification, validation and utilization. <strong>Diabetes</strong> Care 10: 617–21.Cox DJ, Kovatchev BP, Julian DM, Gonder-Frederick LA, Polonsky WH, Schlundt DG, Clarke WL(1994). Frequency of severe hypoglycemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus can be predictedfrom self-monitor<strong>in</strong>g blood glucose data. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 79:1659–62.Cryer PE (1992). Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure<strong>in</strong> IDDM. A vicious cycle. <strong>Diabetes</strong> 41: 255–260.Cryer PE (1994). Hypoglycemia: the limit<strong>in</strong>g factor <strong>in</strong> the management of IDDM. <strong>Diabetes</strong> 43:1378–89.Cryer PE, Davis SN, Shamoon H (2003). Hypoglycemia <strong>in</strong> diabetes. <strong>Diabetes</strong> Care 26: 1902–12.Dagogo-Jack SE, Craft S, Cryer PE (1993). Hypoglycemia-associated autonomic failure <strong>in</strong> <strong>in</strong>sul<strong>in</strong>dependentdiabetes mellitus. Recent antecedent hypoglycemia reduces autonomic responses to,symptoms of, and defense aga<strong>in</strong>st subsequent hypoglycemia. Journal of Cl<strong>in</strong>ical Investigation 91:819–28.Davis EA, Keat<strong>in</strong>g B, Byrne GC, Russell M, Jones TW (1997). Hypoglycemia: Incidence and cl<strong>in</strong>icalpredictors <strong>in</strong> a large population-based sample of children and adolescents with IDDM. <strong>Diabetes</strong>Care 20: 22–5.Donnelly LA, Morris AD, Frier BM, Ellis JD, Donnan PT, Durrant R et al., for the DARTS/MEMOCollaboration (2005). Frequency and predictors of hypoglycaemia <strong>in</strong> type 1 and <strong>in</strong>sul<strong>in</strong>-treatedtype 2 diabetes: a population-based study. Diabetic Medic<strong>in</strong>e 22: 749–55.Feher MD, Grout P, Kennedy A, Elkeles RS, Touquet R (1989). <strong>Hypoglycaemia</strong> <strong>in</strong> an <strong>in</strong>ner-cityaccident and emergency department: a 12-month survey. Archives of Emergency Medic<strong>in</strong>e 6: 183–8.Galassetti P, Mann S, Tate D, Neill RA, Costa F, Wasserman DH, Davis SN (2001a). Effects ofantecedent prolonged exercise on subsequent counterregulatory responses to hypoglycemia. AmericanJournal of Endocr<strong>in</strong>ology 280: E908–17.Galassetti P, Neill AR, Tate D, Ertl AC, Wasserman DH, Davis SN (2001b). Sexual dimorphism<strong>in</strong> counterregulatory responses to hypoglycemia after antecedent exercise. Journal of Cl<strong>in</strong>icalEndocr<strong>in</strong>ology and Metabolism 86: 3516–24.Galassetti P, Tate D, Neill RA, Morrey S, Wasserman DH, Davis SN (2003). Effect of antecedenthypoglycemia on counterregulatory responses to subsequent euglycemic exercise <strong>in</strong> type 1 diabetes.<strong>Diabetes</strong> 52: 1761–9.Garg SK, Gottlieb PA, Hisatomi ME, D’Souza A, Walker AJ, Izuora KE, Chase HP (2004). Improvedglycemic control without an <strong>in</strong>crease <strong>in</strong> severe hypoglycemic episodes <strong>in</strong> <strong>in</strong>tensively treated patients

78 FREQUENCY, CAUSES AND RISK FACTORSwith type 1 diabetes receiv<strong>in</strong>g morn<strong>in</strong>g, even<strong>in</strong>g, or split dose <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e. <strong>Diabetes</strong> Researchand Cl<strong>in</strong>ical Practice 66: 49–56.Gerich J, Langlois M, Noacco C, Karam J, Forsham P (1973). Lack of glucagon response to hypoglycemia<strong>in</strong> diabetes: evidence for an <strong>in</strong>tr<strong>in</strong>sic pancreatic alpha-cell defect. Science 182: 171–3.Gerich JE, Mokan M, Veneman T, Korytkowski M, Mitrakou A (1991). Hypoglycemia unawareness.Endocr<strong>in</strong>e Reviews 12: 356–71.Gold AE, MacLeod KM, Frier BM (1994). Frequency of severe hypoglycemia <strong>in</strong> patients with type 1diabetes with impaired awareness of hypoglycemia. <strong>Diabetes</strong> Care 17: 697–703.Gold AE, Frier BM, MacLeod KM, Deary IJ (1997). A structural equation model for predictors ofsevere hypoglycaemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Diabetic Medic<strong>in</strong>e 14:309–15.Gonder-Frederick L, Clarke WL, Cox DJ (1997). The emotional, social, and behavioural implicationsof <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia. Sem<strong>in</strong>ars <strong>in</strong> Cl<strong>in</strong>ical Neuropsychiatry 2: 57–65.Gonder-Frederick LA, Cox D (1997). The relationship between psychological depression and frequencyof severe hypoglycemia (SH) <strong>in</strong> IDDM. <strong>Diabetes</strong> 46: 268A(Abstract)Grimaldi A, Bosquet F, Davidoff P, Digy JP, Sachon C, Landault C et al. (1990). Unawareness ofhypoglycemia by <strong>in</strong>sul<strong>in</strong>-dependent diabetics. Hormone and Metabolic Research 22: 90–5.Hart SP, Frier BM (1998). Causes, management and morbidity of acute hypoglycaemia <strong>in</strong> adultsrequir<strong>in</strong>g hospital admission. Quarterly Journal of Medic<strong>in</strong>e 91: 505–10.Heller SR, Macdonald IA (1996). The measurement of cognitive function dur<strong>in</strong>g acute hypoglycaemia:experimental limitations and their effect on the study of hypoglycaemia unawareness. DiabeticMedic<strong>in</strong>e 13: 607–15.Hepburn DA, Patrick AW, Ead<strong>in</strong>gton DW, Ew<strong>in</strong>g DJ, Frier BM (1990). Unawareness of hypoglycaemia<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients: prevalence and relationship to autonomic neuropathy. DiabeticMedic<strong>in</strong>e 7: 711–17.Hepburn DA, Patrick AW, Brash HM, Thomson I, Frier BM (1991). <strong>Hypoglycaemia</strong> unawareness<strong>in</strong> type 1 diabetes: a lower plasma glucose is required to stimulate sympatho-adrenal activation.Diabetic Medic<strong>in</strong>e 8: 934–45.Her<strong>in</strong>gs RMC, de Boer A, Stricker BHC, Leufkens HGM, Porsius A (1996). Hypoglycemia associatedwith the use of <strong>in</strong>hibitors of angiotens<strong>in</strong>-convert<strong>in</strong>g enzyme. Lancet 345: 1195–8.Hermansen K, Fonta<strong>in</strong>e P, Kukolja KK, Peterkova V, Leth G, Gall M-A (2004). Insul<strong>in</strong> analogues(<strong>in</strong>sul<strong>in</strong> detemir and <strong>in</strong>sul<strong>in</strong> aspart) versus traditional human <strong>in</strong>sul<strong>in</strong>s (NPH <strong>in</strong>sul<strong>in</strong> and regularhuman <strong>in</strong>sul<strong>in</strong>) <strong>in</strong> basal-bolus therapy for patients with type 1 diabetes. Diabetologia 47: 622–9.Hirsch BR, Shamoon H (1987). Defective ep<strong>in</strong>ephr<strong>in</strong>e and growth hormone responses <strong>in</strong> type 1diabetes are stimulus specific. <strong>Diabetes</strong> 36: 20–6.Hollander PA, Blonde L, Rowe R, Mehta AE, Milburn JL, Hershon KS et al., the Exubera Phase IIIStudy Group (2004). Efficacy and safety of <strong>in</strong>haled <strong>in</strong>sul<strong>in</strong> (Exubera) compared with subcutaneous<strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> patients with type 2 diabetes. <strong>Diabetes</strong> Care 27: 2356–62.Janssen MMJ, Snoek FJ, de Jongh RT, Casteleijn S, Deville W, He<strong>in</strong>e RJ (2000). Biological andbehavioural determ<strong>in</strong>ants of the frequency of mild, biochemical hypoglycaemia <strong>in</strong> patients with Type 1diabetes on multiple <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jection therapy. <strong>Diabetes</strong> Metabolism Research and Reviews 16: 157–63.Johnson ES, Koepsell TD, Reiber G, Stergachis A, Platt R (2002). Increas<strong>in</strong>g <strong>in</strong>cidence of serioushypoglycemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong> users. Journal of Cl<strong>in</strong>ical Epidemiology 55: 253–9.Jones TW, Porter P, Sherw<strong>in</strong> RS, Davis EA, O’Leary P, Frazer F et al. (1998). Decreased ep<strong>in</strong>ephr<strong>in</strong>eresponses to hypoglycemia dur<strong>in</strong>g sleep. New England Journal of Medic<strong>in</strong>e 338: 1657–62.Kerr D, Macdonald IA, Heller SR, Tattersall RB (1990). Alcohol causes hypoglycaemic unawareness<strong>in</strong> healthy volunteers and patients with type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabetes. Diabetologia 33: 216–21.K<strong>in</strong>sley BT, Widom B, Simonson DC (1995). Differential regulation of counterregulatory hormonesecretion and symptoms dur<strong>in</strong>g hypoglycemia <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 18: 17–26.Kong M-F, Macdonald IA, Tattersall RB (1996). Gastric empty<strong>in</strong>g <strong>in</strong> diabetes. Diabetic Medic<strong>in</strong>e 13:112–9.

REFERENCES 79Leckie AM, Graham MK, Grant JB, Ritchie PJ, Frier BM (2005). Frequency, severity, and morbidityof hypoglycemia occurr<strong>in</strong>g <strong>in</strong> the workplace <strong>in</strong> people with <strong>in</strong>sul<strong>in</strong>-treated diabetes. <strong>Diabetes</strong> Care28: 1333–8.Leese GP, Wang J, Broomhall J, Kelly P, Marsden A, Morrison W et al., for the DARTS/MEMOCollaboration (2003). Frequency of severe hypoglycemia requir<strong>in</strong>g emergency department treatment<strong>in</strong> type 1 and type 2 diabetes. A population-based study of health service resource use. <strong>Diabetes</strong>Care 26: 1176–80.MacLeod KM, Hepburn DA, Frier BM (1993). Frequency and morbidity of severe hypoglycaemia <strong>in</strong><strong>in</strong>sul<strong>in</strong>-treated diabetic patients. Diabetic Medic<strong>in</strong>e 10: 238–45.MacLeod KM, Deary IJ, Graham KS, Hepburn DA, Frier BM (1994). <strong>Hypoglycaemia</strong> unawareness <strong>in</strong>adult patients with type 1 diabetes: relationship to severe hypoglycaemia and cognitive impairment.<strong>Diabetes</strong>, Nutrition and Metabolism 7: 205–11.MacLeod KM, Gold AE, Frier BM (1995). Frequency, severity and symptomatology of hypoglycaemia:a comparative trial of human and porc<strong>in</strong>e <strong>in</strong>sul<strong>in</strong>s <strong>in</strong> type 1 diabetic patients. Diabetic Medic<strong>in</strong>e 12:134–41.Mokan M, Mitrakou A, Veneman T, Ryan C, Korytkowski M, Cryer PE, Gerich J (1994). Hypoglycemiaunawareness <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 17: 1397–403.Morris AD, Boyle DIR, McMahon AD, Pearce H, Evans JMM, Newton RW et al., for theDARTS/MEMO Collaboration (1997). ACE <strong>in</strong>hibitor use is associated with hospitalization forsevere hypoglycemia <strong>in</strong> patients with diabetes. <strong>Diabetes</strong> Care 20: 1363–7.Moses RG, Hubert PA, Lewis-Driver DJ (1985). Severe hypoglycaemia. A one-year prospective survey<strong>in</strong> Wollongong. Medical Journal of Australia 142: 294–6.Muhlhauser I, Berger M, Sonnenberg G, Koch J, Jorgens V, Schernthaner G, Scholz V (1985).Incidence and management of severe hypoglycemia <strong>in</strong> 434 adults with <strong>in</strong>sul<strong>in</strong>-dependent diabetesmellitus. <strong>Diabetes</strong> Care 8: 268–73.Muhlhauser I, Bruckner I, Berger M, Cheta D, Jorgens V, Ionescu-Tirgoviste C et al. (1987). Evaluationof an <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> treatment and teach<strong>in</strong>g programme as rout<strong>in</strong>e management of Type 1(<strong>in</strong>sul<strong>in</strong>-dependent) diabetes. The Bucharest-Düsseldorf Study. Diabetologia 30: 681–90.Muhlhauser I, Toth G, Sawicki PT, Berger M (1991). Severe hypoglycemia <strong>in</strong> type 1 diabetic patientswith impaired kidney function. <strong>Diabetes</strong> Care 14: 344–6.Muhlhauser I, Overmann H, Bender R, Bott U, Berger M (1998). Risk factors for severe hypoglycaemia<strong>in</strong> adult patients with type 1 diabetes. Diabetologia 41: 1274–82.Nordfeldt S, Samuelsson U (2003). Serum ACE predicts severe hypoglycemia <strong>in</strong> children and adolescentswith type 1 diabetes. <strong>Diabetes</strong> Care 26: 274–8.O’Neill, S (1997). Make four the floor. Balance Jan/Feb: 20–3.Owens DR, Z<strong>in</strong>man B, Bolli GB (2001). Insul<strong>in</strong>s today and beyond. Lancet 358: 739–46.Pampanelli S, Fanelli C, Lalli C, Ciofetta M, Del S<strong>in</strong>daco P, Lepore M et al. (1996). Long-term<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> IDDM: effects on HbA 1c , risk for severe and mild hypoglycaemia, statusof counterregulation and awareness of hypoglycaemia. Diabetologia 39: 677–86.Pedersen-Bjergaard U, Agerholm-Larsen B, Pramm<strong>in</strong>g S, Hougaard P, Thorste<strong>in</strong>sson B (2001).Activity of angiotens<strong>in</strong>-convert<strong>in</strong>g enzyme and risk of severe hypoglycaemia <strong>in</strong> type 1 diabetesmellitus. Lancet 357: 1248–53.Pedersen-Bjergaard U, Agerholm-Larsen B, Pramm<strong>in</strong>g S, Hougaard P, Thorste<strong>in</strong>sson B (2003a).Prediction of severe hypoglycaemia by angiotens<strong>in</strong>-convert<strong>in</strong>g enzyme activity and genotype <strong>in</strong> type1 diabetes. Diabetologia 46: 89–96.Pedersen-Bjergaard U, Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B (2003b). Recall of severe hypoglycaemia andself-estimated state of awareness <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Metabolism Research and Reviews19: 232–40.Pedersen-Bjergaard U, Pramm<strong>in</strong>g S, Heller SR, Wallace TM, Jorgensen HV, Matthews DR et al.(2004). Severe hypoglycaemia <strong>in</strong> 1076 adult patients with type 1 diabetes: <strong>in</strong>fluence of risk markersand selection. <strong>Diabetes</strong> Metabolism Research and Reviews 20: 479–86.

80 FREQUENCY, CAUSES AND RISK FACTORSPedersen-Bjergaard U, Reubsaet JLE, Nielsen SL, Pedersen-Bjergaard S, Perrild H, Pramm<strong>in</strong>g S,Thorste<strong>in</strong>sson B (2005). Psychoactive drugs, alcohol, and severe hypoglycemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treateddiabetes: Analysis of 141 cases. American Journal of Medic<strong>in</strong>e 118: 307–10.Plank J, Kohler G, Rakovac I, Semlitsch BM, Horvath K, Bock G, et al. (2004). Long-term evaluationof a structured out-patient education programme for <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> patients with type1 diabetes: a 12 year follow-up. Diabetologia 47: 1370–5.Potter J, Clarke P, Gale EAM, Dave SH, Tattersall RB (1982). Insul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia<strong>in</strong> an accident and emergency department: the tip of an iceberg? British Medical Journal 285:1180–2.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Bendtson I, B<strong>in</strong>der C (1990). The relationship between symptomaticand biochemical hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetic patients. Journal of Internal Medic<strong>in</strong>e228: 641–6.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Bendtson I, B<strong>in</strong>der C (1991). Symptomatic hypoglycaemia <strong>in</strong> 411 type1 diabetic patients. Diabetic Medic<strong>in</strong>e 8: 217–22.Reichard P, Pihl M (1994). Mortality and treatment side-effects dur<strong>in</strong>g long-term <strong>in</strong>tensified conventional<strong>in</strong>sul<strong>in</strong> treatment <strong>in</strong> the Stockholm <strong>Diabetes</strong> Intervention Study. <strong>Diabetes</strong> 43: 313–7.Ryder REJ, Owens DR, Hayes TM, Ghatei MA, Bloom SR (1990). Unawareness of hypoglycaemiaand <strong>in</strong>adequate hypoglycaemic counterregulation: no causal relationship with diabetic autonomicneuropathy. British Medical Journal 301: 783–7.Service FJ (1995). Hypoglycemic disorders. New England Journal of Medic<strong>in</strong>e 332: 1144–52.Siebenhofer A, Plank J, Berghold A, Horvath K, Sawicki PT, Beck P, Pieber TR (2004). Meta-analysisof short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues <strong>in</strong> adult patients with type 1 diabetes: cont<strong>in</strong>uous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusionversus <strong>in</strong>jection therapy. Diabetologia 47: 1895–1905.Skyler JS, We<strong>in</strong>stock RS, Rask<strong>in</strong> P, Yale J-F, Barrett E, Gerich JE, Gerste<strong>in</strong> HC, the Inhaled Insul<strong>in</strong>Phase III type 1 <strong>Diabetes</strong> Study Group (2005). <strong>Diabetes</strong> Care 28: 1630–35.Sonnenberg GE, Kemmer FW, Berger M (1990). Exercise <strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabeticpatients treated with cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion. Prevention of exercise <strong>in</strong>ducedhypoglycaemia. Diabetologia 33: 696–703.Stephenson JM, Kempler P, Cavallo Per<strong>in</strong> P, Fuller JH, the EURODIAB IDDM Complications StudyGroup (1996). Is autonomic neuropathy a risk factor for severe hypoglycaemia? The EURODIABIDDM Complications Study. Diabetologia 39: 1372–6.ter Braak EWMT, Appelman AMMF, van de Laak MF, Stolk RP, Van Haeften TW, Erkelens DW(2000). Cl<strong>in</strong>ical characteristics of type 1 diabetic patients with and without severe hypoglycemia.<strong>Diabetes</strong> Care 23: 1467–71.The DCCT Research Group (1991). Epidemiology of severe hypoglycemia <strong>in</strong> the <strong>Diabetes</strong> Controland Complications Trial. American Journal of Medic<strong>in</strong>e 90: 450–9.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1993). The effect of <strong>in</strong>tensive treatmentof diabetes on the development and progression of long-term complications <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. New England Journal of Medic<strong>in</strong>e 329: 977–86.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1997). Hypoglycemia <strong>in</strong> the <strong>Diabetes</strong>Control and Complications Trial. <strong>Diabetes</strong> 46: 271–86.The EURODIAB IDDM Complications Study Group (1994). Microvascular and acute complications<strong>in</strong> IDDM patients: the EURODIAB IDDM Complications Study. Diabetologia 37:278–85.Thorste<strong>in</strong>sson B, Pramm<strong>in</strong>g S, Lauritzen T, B<strong>in</strong>der C (1986). Frequency of daytime biochemical hypoglycaemia<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients: relation to daily median blood glucose concentrations.Diabetic Medic<strong>in</strong>e 3: 147–51.Turner BC, Jenk<strong>in</strong>s E, Kerr D, Sherw<strong>in</strong> RS, Cavan DA (2001). The effect of even<strong>in</strong>g alcohol consumptionon next-morn<strong>in</strong>g glucose control <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care 24: 1888–93.UK <strong>Hypoglycaemia</strong> Study Group (2007). Risk of hypoglycaemia <strong>in</strong> types 1 and 2 diabetes: effects oftreatment modalities and their duration. Diabetologia 50: 1140–47.

REFERENCES 81Vervoort G, Goldschmidt HMG, van Doorn LG (1996). Nocturnal blood glucose profiles <strong>in</strong> patientswith type 1 diabetes mellitus on multiple (> 4) daily <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jection regimens. Diabetic Medic<strong>in</strong>e13: 794–9.Warren E, Weatherley-Jones E, Chilcott J, Beverley C (2004). Systematic review and economicevaluation of a long-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogue, <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e. Health Technology Assessment 8 (45):iii, 1–57.White NH, Skor DA, Cryer PE, Levandoski LA, Bier DM, Santiago JV (1983). Identification of type 1diabetic patients at <strong>in</strong>creased risk for hypoglycemia dur<strong>in</strong>g <strong>in</strong>tensive therapy. New England Journalof Medic<strong>in</strong>e 308: 485–91.Zammitt NN, Geddes J, Warren RE, Marioni R, Ashby JP, Frier BM (2007). Serum Angiotens<strong>in</strong>-Convert<strong>in</strong>g Enzyme and frequency of severe hypoglycaemia <strong>in</strong> type 1 diabetes: does a relationshipexist? Diabetic Medic<strong>in</strong>e, <strong>in</strong> press.

4 Nocturnal <strong>Hypoglycaemia</strong>Simon R. HellerINTRODUCTIONPeople with type 1 diabetes, who owe their lives to <strong>in</strong>sul<strong>in</strong>, worry as much abouthypoglycaemia as they do about the prospect of develop<strong>in</strong>g serious diabetic complications,and episodes occurr<strong>in</strong>g dur<strong>in</strong>g sleep are feared more than any other. Several studies haveconfirmed how commonly hypoglycaemia occurs dur<strong>in</strong>g the night, and the possibility ofthe development of severe nocturnal episodes can drive parents to sleep <strong>in</strong> their child’sbedroom for many years. However, nocturnal hypoglycaemia may have consequences beyonda domestic emergency. Asymptomatic episodes may contribute to impaired hypoglycaemiaawareness and deficient counterregulation, while nocturnal hypoglycaemia may also be associatedwith cognitive impairment and is implicated <strong>in</strong> the ‘Dead <strong>in</strong> Bed’ syndrome (seeChapter 12) with the <strong>in</strong>creased risk of sudden death dur<strong>in</strong>g sleep <strong>in</strong> young people with type1 diabetes. In this chapter, the reasons why nocturnal hypoglycaemia occurs so frequently<strong>in</strong> type 1 diabetes are discussed. Also, some cl<strong>in</strong>ical approaches are described that mighthelp to reduce its frequency and severity and some controversial topics are considered suchas the Somogyi effect.EPIDEMIOLOGY – HOW COMMON IS NOCTURNALHYPOGLYCAEMIA?Nocturnal hypoglycaemia has always been of major concern to <strong>in</strong>dividuals with diabetes,with many patients experienc<strong>in</strong>g severe episodes, but it was not until the 1970s that thisproblem was studied systematically. Patients with type 1 diabetes on rout<strong>in</strong>e <strong>in</strong>sul<strong>in</strong> therapywere admitted for overnight monitor<strong>in</strong>g us<strong>in</strong>g <strong>in</strong>termittent venous blood sampl<strong>in</strong>g (Gale andTattersall, 1979). Nocturnal hypoglycaemia was reported <strong>in</strong> 22 of 39 adults, who had whatwould now be considered poorly controlled diabetes and were be<strong>in</strong>g treated with one ortwo <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> each day. S<strong>in</strong>ce then a number of <strong>in</strong>vestigators have exam<strong>in</strong>ed thefrequency of nocturnal hypoglycaemia, either us<strong>in</strong>g patients’ self-reports or by measur<strong>in</strong>gblood glucose at <strong>in</strong>tervals overnight. Although some studies have reported relatively lowrates, others have demonstrated how common nocturnal hypoglycaemia can be. Reportedrates of hypoglycaemia vary between 7 and 60% both <strong>in</strong> children and adults (Table 4.1).<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

Table 4.1 Frequency of nocturnal hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetesStudy NumberAdults (A) orchildren (C) TreatmentSampl<strong>in</strong>g (bloodglucose (BG) orcont<strong>in</strong>uous(CGM))Percentage withglucose below:(mmol/l)3.0 2.5 2.0Proportionasymptomatic Mean HbA 1cGale andTattersall, 1979Dornan et al.,1981Pramm<strong>in</strong>g et al.,1985Wh<strong>in</strong>cup andMilner, 1987Bendtson et al.,1988Shalwitz et al.,1990Vervoort et al.,199639 A BD <strong>in</strong>sul<strong>in</strong> BG 56 6482 A BD <strong>in</strong>sul<strong>in</strong> BG 27 27 9.5 (A1)58 A Conventional BG 29 22 9 9 9.1 (A1)71 C BD <strong>in</strong>sul<strong>in</strong> OD<strong>in</strong>sul<strong>in</strong>BG 34 111023 A MDI BG 30 9 9 57 8.3 (A1)25 CSII 44 36 12 66 7.0 (A1c)135 C Conventional BG 14 7 2 100 10.3 (A1)31 A Basal bolus BG 29 67 8.6 (A1c)Porter et al., 1997 61 C BD BG 17 13 8Beregszaszi et al., 150 C BD/MIT BG 47 491997Matyka et al., 29 C BD BG 45 100 8.8 (A1c)1999bBoland et al., 2001 56 C MIT/CSII CGM 70 100Kaufman et al.,200247 C MIT/CSII CGM 35 27 100 8.6BD = twice daily; CSII = cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion; OD = once daily; MDI = multiple doses of <strong>in</strong>sul<strong>in</strong>; CGM = cont<strong>in</strong>uous glucose monitor<strong>in</strong>g; MIT = multiple <strong>in</strong>jection therapy;BG = blood glucose

EPIDEMIOLOGY OF NOCTURNAL HYPOGLYCAEMIA? 8576Glucose (mmol/l)54321021:0023:00 01:00 03:00Time (hours)05:00 07:00Figure 4.1 Overnight glucose profiles of one child with type 1 diabetes who was hypoglycaemic onboth study nights. The shaded area represents the range of glucose values from the overnight profilesof the children without diabetesAlthough this is clearly dependent upon the blood glucose level that is designated asrepresent<strong>in</strong>g hypoglycaemia, it is <strong>in</strong>terest<strong>in</strong>g that the studies that reported the lowest ratesof hypoglycaemia had sampled blood glucose less frequently, and it is conceivable thathypoglycaemia at other times of the night was simply not identified. Furthermore, most earlystudies <strong>in</strong>volved patients with poor glycaemic control. It seems likely that with <strong>in</strong>tensive<strong>in</strong>sul<strong>in</strong> therapy now be<strong>in</strong>g offered to most patients, rates of nocturnal hypoglycaemia areeven higher. This certa<strong>in</strong>ly appears to be true <strong>in</strong> children. In one alarm<strong>in</strong>g study of childrentreated with multiple <strong>in</strong>jection therapy, the rates of nocturnal hypoglycaemia exceeded 50%(Beregszaszi et al., 1997). It is noteworthy that children have been directly observed tosleep through most nocturnal episodes despite hav<strong>in</strong>g a blood glucose value below 2 mmol/l(Matyka et al., 1999b) (Figure 4.1).The <strong>in</strong>troduction of cont<strong>in</strong>uous blood glucose monitor<strong>in</strong>g with glucose be<strong>in</strong>g measuredevery few m<strong>in</strong>utes has allowed more frequent record<strong>in</strong>g dur<strong>in</strong>g sleep. Initial reports us<strong>in</strong>gthese methods have suggested that hypoglycaemic events may be even more commonthan was proposed previously (Table 4.1). However, although cont<strong>in</strong>uous blood glucoserecord<strong>in</strong>g appears to have revealed a high rate of nocturnal episodes <strong>in</strong> some studies,it may over-estimate the overall frequency (McGowan et al., 2002). The technologymeasures glucose <strong>in</strong> the extra cellular space and not <strong>in</strong> the blood vessels, and the relationshipbetween the glucose concentrations <strong>in</strong> these sites is not clear (Kulcu et al., 2003).It is also possible that the technology itself produces an <strong>in</strong>built bias (Wentholt et al.,2005). Nevertheless, not only is nocturnal hypoglycaemia common when measured byconventional blood glucose sampl<strong>in</strong>g but it is also often of long duration with someepisodes last<strong>in</strong>g for more than three hours (Gale and Tattersall, 1979; Matyka et al.,1999b).

86 NOCTURNAL HYPOGLYCAEMIACAUSES OF NOCTURNAL HYPOGLYCAEMIAThe Limitations of Therapeutic Insul<strong>in</strong> DeliveryThe ability of people with <strong>in</strong>sul<strong>in</strong>-treated diabetes to ma<strong>in</strong>ta<strong>in</strong> strict glucose targets andprevent long-term tissue damage is compromised by the deficiencies of currently availablemethods of <strong>in</strong>sul<strong>in</strong> delivery. In non-diabetic <strong>in</strong>dividuals, endogenous secretion of <strong>in</strong>sul<strong>in</strong>precisely meets demand. When food is eaten, <strong>in</strong>sul<strong>in</strong> secretion <strong>in</strong>creases rapidly to matchnutrient <strong>in</strong>take, particularly when the <strong>in</strong>gested food conta<strong>in</strong>s carbohydrate. In between meals,basal <strong>in</strong>sul<strong>in</strong> secretion falls to a low but consistent level to ma<strong>in</strong>ta<strong>in</strong> basal metabolism,without the risk of hypoglycaemia, even dur<strong>in</strong>g prolonged fast<strong>in</strong>g that lasts for hours ordays. In contrast, the limitations of deliver<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> by subcutaneous <strong>in</strong>jection to patientswho can no longer produce endogenous <strong>in</strong>sul<strong>in</strong>, leads not only to <strong>in</strong>adequate plasma <strong>in</strong>sul<strong>in</strong>concentrations dur<strong>in</strong>g eat<strong>in</strong>g and the immediate postprandial period, but also to <strong>in</strong>appropriatelyraised plasma <strong>in</strong>sul<strong>in</strong> levels <strong>in</strong> the post-absorptive phase (Rizza et al., 1980). Thisresults <strong>in</strong> high postprandial blood glucose concentrations <strong>in</strong> the hour or so after eat<strong>in</strong>g anda tendency to cause hypoglycaemia <strong>in</strong> the period before the next meal. Individuals areparticularly likely to be affected dur<strong>in</strong>g the night, when the <strong>in</strong>terval between <strong>in</strong>gestion offood may be several hours (Box 4.1). As discussed below, developments <strong>in</strong> <strong>in</strong>sul<strong>in</strong> deliveryby us<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues or cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion offer some benefitover conventional <strong>in</strong>sul<strong>in</strong>, although these approaches mitigate rather than cure the problemof nocturnal hypoglycaemia.Impaired Counterregulatory Responses to <strong>Hypoglycaemia</strong> dur<strong>in</strong>gNocturnal <strong>Hypoglycaemia</strong>Various mechanisms contribute to the additional risk of develop<strong>in</strong>g hypoglycaemia dur<strong>in</strong>g thenight. Although an earlier study suggested that hormonal responses to hypoglycaemia mightbe <strong>in</strong>creased dur<strong>in</strong>g nocturnal episodes (Bendtson et al., 1993), recent work has <strong>in</strong>dicatedthat counterregulatory defences are generally impaired. Matyka et al. (1999a) studied 29 prepubertalchildren overnight <strong>in</strong> their own homes on two separate occasions. They confirmedthat not only was nocturnal hypoglycaemia common, but also that prolonged episodes ofhypoglycaemia were not accompanied by <strong>in</strong>creases <strong>in</strong> ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) and otherBox 4.1Factors contribut<strong>in</strong>g to the development of nocturnal hypoglycaemia1. Long period between meals (especially <strong>in</strong> children).2. Inconsistency of conventional subcutaneous basal <strong>in</strong>sul<strong>in</strong> delivery – nocturnalhyper<strong>in</strong>sul<strong>in</strong>aemia.3. Unawareness of early symptoms of hypoglycaemia when asleep.4. Dim<strong>in</strong>ished counterregulatory hormone release and symptomatic response <strong>in</strong> sup<strong>in</strong>eposture and effect of sleep per se.

CAUSES OF NOCTURNAL HYPOGLYCAEMIA 87protective endocr<strong>in</strong>e responses. S<strong>in</strong>ce protective counterregulatory responses mitigate theseverity of hypoglycaemia by <strong>in</strong>creas<strong>in</strong>g hepatic glucose output and reduc<strong>in</strong>g peripheralglucose uptake, as well as enhanc<strong>in</strong>g some of the warn<strong>in</strong>g autonomic symptoms, defectiveresponses dur<strong>in</strong>g the night may have a major effect <strong>in</strong> <strong>in</strong>creas<strong>in</strong>g the risk of nocturnalhypoglycaemia.One important question that emerges from this work is why nocturnal hypoglycaemiashould provoke different physiological responses to those that occur dur<strong>in</strong>g the day. It nowappears that impaired responses are the result of additional factors at night, <strong>in</strong> particular asup<strong>in</strong>e posture and also sleep per se.Sup<strong>in</strong>e PostureTwo studies have <strong>in</strong>vestigated the effect of posture on counterregulatory hormonal defencesto hypoglycaemia. Hirsch et al. (1991) compared the physiological responses to hypoglycaemia<strong>in</strong>duced us<strong>in</strong>g a hyper<strong>in</strong>sul<strong>in</strong>aemic clamp <strong>in</strong> young people with type 1 diabetes, whowere either <strong>in</strong> an upright or a horizontal position. Increments <strong>in</strong> hypoglycaemic symptomscores were more than 50% lower when patients rema<strong>in</strong>ed sup<strong>in</strong>e compared to the usual<strong>in</strong>crease that was observed when they were stand<strong>in</strong>g erect. These f<strong>in</strong>d<strong>in</strong>gs have beenconfirmed by other researchers, who also demonstrated that plasma ep<strong>in</strong>ephr<strong>in</strong>e levels werethree times lower dur<strong>in</strong>g hypoglycaemia <strong>in</strong> the sup<strong>in</strong>e position compared to concentrationswhen subjects were stand<strong>in</strong>g (Rob<strong>in</strong>son et al., 1994). The precise physiological explanationis not entirely clear but may be related to the recruitment of adrenoreceptors which areprimed by an upright posture. Whatever the cause, these observations suggest that patientsare more vulnerable to progress<strong>in</strong>g to a severe episode of hypoglycaemia when ly<strong>in</strong>g horizontal<strong>in</strong> bed, because of a reduction <strong>in</strong> symptom <strong>in</strong>tensity and <strong>in</strong> magnitude of hormonalcounterregulation.SleepMost studies have reported suppression of physiological protection by counterregulatorymechanisms dur<strong>in</strong>g hypoglycaemia. Jones et al. (1998) lowered blood glucose to 2.8 mmol/lboth dur<strong>in</strong>g the daytime and at night, and demonstrated dim<strong>in</strong>ished ep<strong>in</strong>ephr<strong>in</strong>e responses<strong>in</strong> diabetic and non-diabetic adolescents while they were asleep when compared to thebrisk responses while they were awake, whether dur<strong>in</strong>g daytime hours or dur<strong>in</strong>g the night(Figure 4.2). Banerer and Cryer (2003) confirmed these observations <strong>in</strong> patients with type 1diabetes but, <strong>in</strong>terest<strong>in</strong>gly, <strong>in</strong> non-diabetic subjects no difference <strong>in</strong> ep<strong>in</strong>ephr<strong>in</strong>e responsewas observed between these states. Dim<strong>in</strong>ished counterregulatory responses, <strong>in</strong>clud<strong>in</strong>gep<strong>in</strong>ephr<strong>in</strong>e, have also been demonstrated dur<strong>in</strong>g spontaneous nocturnal hypoglycaemicepisodes <strong>in</strong> children with diabetes (Matyka et al., 1999a). Therefore, sleep appears to beassociated with dim<strong>in</strong>ished catecholam<strong>in</strong>e and symptomatic responses to hypoglycaemiawith a reduction <strong>in</strong> waken<strong>in</strong>g dur<strong>in</strong>g a hypoglycaemic episode. The dim<strong>in</strong>ished responsemay occur because the glycaemic thresholds for activation of these responses have beenreset to a lower glucose level (Gais et al., 2003), so that more profound hypoglycaemia isnecessary to provoke a similar response to that observed <strong>in</strong> subjects when they are awake.As the <strong>in</strong>vestigators <strong>in</strong>volved <strong>in</strong> these studies have po<strong>in</strong>ted out, this <strong>in</strong>creases the risk of a

88 NOCTURNAL HYPOGLYCAEMIAPlasma Ep<strong>in</strong>ephr<strong>in</strong>e (pg/ml)350300250200150100500Patients with <strong>Diabetes</strong>Normal Subjects450Daytime, awakeDaytime, awake400Nighttime, asleepNighttime, asleepNighttime, awake350300250200150100500–60 –40 –20 0 20 40 60 –60 –40 –20 0 20 40 60Time (m<strong>in</strong>utes)Figure 4.2 Mean (± SE) plasma ep<strong>in</strong>ephr<strong>in</strong>e concentrations <strong>in</strong> eight patients with type 1 diabetesand six normal subjects dur<strong>in</strong>g periods of hypoglycaemia when they were awake dur<strong>in</strong>g the day, awakeat night and asleep at night. (To convert plasma ep<strong>in</strong>ephr<strong>in</strong>e values to picomoles per litre, multiply by5.458. The zero on the x-axis <strong>in</strong>dicates the beg<strong>in</strong>n<strong>in</strong>g of the hypoglycaemic period). Reproduced withpermission from Jones et al. (1998). Copyright © 1998 Massachusetts Medical Societysevere hypoglycaemic episode occurr<strong>in</strong>g at night. However, the mechanisms that disturb thephysiological responses to hypoglycaemia dur<strong>in</strong>g sleep rema<strong>in</strong> unknown.Sleep is not a unitary process (Oswald, 1987). Sleep is dom<strong>in</strong>ated by Slow Wave Sleep(SWS) dur<strong>in</strong>g the first third of the night, and by the cyclical appearance of Rapid EyeMovement (REM) sleep dur<strong>in</strong>g the latter two thirds. Autonomic activity dur<strong>in</strong>g SWS isrelatively steady but <strong>in</strong> REM sleep (desynchronisation of the EEG, with absence of activity <strong>in</strong>the anti-gravity and periodic eye movements) modulations <strong>in</strong> respiratory and cardiovascularevents occur with other changes <strong>in</strong> the autonomic nervous system. These differences <strong>in</strong>autonomic activity between SWS and REM sleep suggest that the effect of hypoglycaemiaon autonomic responses may vary depend<strong>in</strong>g on which stage of sleep is be<strong>in</strong>g experienced.However, to date no studies have been published that have explored the effect of differentphases of sleep on counterregulatory responses to hypoglycaemia.CONSEQUENCES OF NOCTURNAL HYPOGLYCAEMIAImpaired Awareness of <strong>Hypoglycaemia</strong>The demonstration <strong>in</strong> the early 1990s that repeated exposure to hypoglycaemia leads toimpaired physiological defences to subsequent episodes and a reduction <strong>in</strong> the <strong>in</strong>tensity ofsymptoms (Heller and Cryer, 1991; Dagogo-Jack et al., 1993) identified a mechanism thatexpla<strong>in</strong>s why some <strong>in</strong>dividuals lose their hypoglycaemia warn<strong>in</strong>g symptoms. Further studiesdemonstrated that such episodes did not need to be symptomatic to produce alterations <strong>in</strong>physiological responses. Veneman et al. (1993) <strong>in</strong>duced hypoglycaemia <strong>in</strong> 10 non-diabeticsubjects overnight and tested their physiological responses to hypoglycaemia the follow<strong>in</strong>gmorn<strong>in</strong>g. They reported lower symptomatic and hormonal responses when compared to anon-hypoglycaemia control night. These relatively mild levels of hypoglycaemia are rarely

CONSEQUENCES OF NOCTURNAL HYPOGLYCAEMIA 89reported by patients because they rema<strong>in</strong> asleep, but may contribute to the development ofimpaired awareness of hypoglycaemia.The ‘Dead <strong>in</strong> Bed’ SyndromeThe possible contribution of hypoglycaemia to cardiac arrhythmias and sudden death isdiscussed <strong>in</strong> detail <strong>in</strong> Chapter 12. However, s<strong>in</strong>ce it has been proposed that nocturnalhypoglycaemia may be a specific precipitant, it merits a brief mention here. A detailed surveyof unexpected deaths <strong>in</strong> young people with type 1 diabetes <strong>in</strong> the early 1990s highlighted arare but dist<strong>in</strong>ct mode of death <strong>in</strong> which young people were found ly<strong>in</strong>g <strong>in</strong> an undisturbed bed.Autopsy failed to reveal a structural cause but circumstantial evidence implicated nocturnalhypoglycaemia as a precipitant. Not all affected <strong>in</strong>dividuals had strict glycaemic control, butmany were known to have been susceptible to develop<strong>in</strong>g nocturnal hypoglycaemia. Thistype of sudden and unexpected death has s<strong>in</strong>ce been confirmed <strong>in</strong> epidemiological surveys<strong>in</strong> other countries, possibly account<strong>in</strong>g for between 5–10% of all deaths under the age of40 <strong>in</strong> people with diabetes. <strong>Hypoglycaemia</strong> causes an <strong>in</strong>crease <strong>in</strong> the QT <strong>in</strong>terval and onepossible explanation is that an episode of nocturnal hypoglycaemia triggers a ventriculararrhythmia <strong>in</strong> a susceptible <strong>in</strong>dividual. However, further work is required to establish thisplausible hypothesis as the cause of these deaths.Neurological Consequences of Nocturnal <strong>Hypoglycaemia</strong> on CerebralFunctionThe possibility that recurrent exposure to hypoglycaemia, particularly occurr<strong>in</strong>g dur<strong>in</strong>g sleep,might <strong>in</strong>sidiously damage cerebral function and cause permanent cognitive impairment wasraised by the f<strong>in</strong>d<strong>in</strong>g that children who had developed type 1 diabetes before the age offive years, exhibited cognitive impairment when compared with non-diabetic controls (Ryanet al., 1985). This observation was replicated <strong>in</strong> different studies us<strong>in</strong>g a range of cognitivetests (Bjorgaas et al., 1997; Rovet and Ehrlich, 1999).However, the relationship to hypoglycaemia, particularly dur<strong>in</strong>g the night, was unclear(Golden et al., 1989); hypoglycaemia-<strong>in</strong>duced convulsions have also been implicated, and itis possible that other factors associated with early-onset diabetes may contribute to cognitiveimpairment.Several studies have explored the extent to which nocturnal episodes may affect performanceon the follow<strong>in</strong>g day. Bendtson et al. (1992) found no difference <strong>in</strong> cognitive performanceamong adults with type 1 diabetes when tested on the morn<strong>in</strong>g after an episode ofnocturnal hypoglycaemia <strong>in</strong> comparison with a night with no hypoglycaemia. Similar f<strong>in</strong>d<strong>in</strong>gshave been reported after the <strong>in</strong>duction of experimental hypoglycaemia dur<strong>in</strong>g the night (K<strong>in</strong>get al., 1998) although the subjects were more fatigued on the follow<strong>in</strong>g morn<strong>in</strong>g. It seemsthat even children are relatively unaffected by a s<strong>in</strong>gle episode of nocturnal hypoglycaemia.Matyka et al. (1999b) tested pre-pubertal children after episodes of prolonged, spontaneouslyoccurr<strong>in</strong>g, hypoglycaemia that occurred dur<strong>in</strong>g sleep, many of which lasted several hours,and found no deleterious effect on cognitive function on the follow<strong>in</strong>g morn<strong>in</strong>g, althoughmood was adversely affected.In summary, although it seems plausible that recurrent nocturnal hypoglycaemia mightcontribute to cognitive decl<strong>in</strong>e, on the available evidence the verdict rema<strong>in</strong>s unproven.

90 NOCTURNAL HYPOGLYCAEMIACAN NOCTURNAL HYPOGLYCAEMIA BE PREDICTED?The ability to predict whether a nocturnal hypoglycaemic episode is likely, by measur<strong>in</strong>gglucose concentrations at bedtime, has generally been tested <strong>in</strong> studies us<strong>in</strong>g <strong>in</strong>termittentblood glucose sampl<strong>in</strong>g to detect hypoglycaemia. It is likely that such a design will failto identify all episodes. The results of two early studies that observed patients who werereceiv<strong>in</strong>g one or two daily <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong>, suggested that a bedtime glucose concentrationof 6.0 mmol/l or less <strong>in</strong>dicated an 80% chance of a period of biochemical hypoglycaemiadur<strong>in</strong>g the night (Pramm<strong>in</strong>g et al., 1985; Wh<strong>in</strong>cup and Milner, 1987). Furthermore, theseearly studies were generally undertaken among patients who were tak<strong>in</strong>g two <strong>in</strong>jectionsof conventional <strong>in</strong>sul<strong>in</strong>s each day. Studies of children have suggested that blood glucosemeasurements at bedtime are less predictive of hypoglycaemia <strong>in</strong> the first half of thenight, although a value of < 75 mmol/l does <strong>in</strong>dicate an <strong>in</strong>creased risk (Matyka et al.,1999b). That study and others have also shown that a low fast<strong>in</strong>g blood glucose is astrong <strong>in</strong>dicator that hypoglycaemia has occurred <strong>in</strong> the latter half of the night (Matyka,2002).S<strong>in</strong>ce severe nocturnal hypoglycaemia is relatively rare, it is much more difficult toestablish to what extent bedtime blood glucose values are predictive. A previous historyof severe hypoglycaemia, impaired awareness of hypoglycaemia or strict control with lowHbA 1c values, will confer a greater chance of a severe episode but are a poor guide to therisk of develop<strong>in</strong>g a severe nocturnal episode on any particular night.THE SOMOGYI PHENOMENON: THE CONCEPT OF REBOUNDHYPERGLYCAEMIAIn the late 1930s, a Hungarian biochemist, Michael Somogyi, work<strong>in</strong>g <strong>in</strong> St Louis, USA,suggested that nocturnal hypoglycaemia might provoke rebound hyperglycaemia on thefollow<strong>in</strong>g morn<strong>in</strong>g, and he supported his hypothesis with a demonstration that reduc<strong>in</strong>geven<strong>in</strong>g doses of <strong>in</strong>sul<strong>in</strong> led to a reduction <strong>in</strong> fast<strong>in</strong>g ur<strong>in</strong>ary glycosuria (Somogyi, 1959). Heproposed that nocturnal hypoglycaemia provokes a counterregulatory response with rises <strong>in</strong>plasma ep<strong>in</strong>ephr<strong>in</strong>e, cortisol and growth hormone result<strong>in</strong>g <strong>in</strong> the release of glucose from theliver and <strong>in</strong>hibition of the effects of <strong>in</strong>sul<strong>in</strong> over the next few hours. The logical conclusionfrom his hypothesis was that this ‘rebound’ elevated fast<strong>in</strong>g blood glucose <strong>in</strong> the morn<strong>in</strong>gshould be treated, not by an <strong>in</strong>crease <strong>in</strong> the even<strong>in</strong>g dose of <strong>in</strong>sul<strong>in</strong>, but paradoxically by areduction. The idea of ‘rebound hyperglycaemia’ follow<strong>in</strong>g nocturnal hypoglycaemia, (alsoknown as the Somogyi phenomenon) as an explanation for a high fast<strong>in</strong>g blood glucose<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated patients, has proved to be very attractive to many diabetes healthcareprofessionals who firmly believe <strong>in</strong> its existence. The consequences are important as patientsare often advised to reduce their even<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> dose, particularly if they compla<strong>in</strong> ofnocturnal hypoglycaemia. However, its cl<strong>in</strong>ical relevance was challenged over 20 years agoand repeated studies have established that fast<strong>in</strong>g hyperglycaemia is largely a result of fall<strong>in</strong>gplasma <strong>in</strong>sul<strong>in</strong> concentrations dur<strong>in</strong>g the night, as the subcutaneous depot of <strong>in</strong>sul<strong>in</strong> that was<strong>in</strong>jected the day before is dissipated.Gale et al. (1980) demonstrated that periods of hypoglycaemia dur<strong>in</strong>g the night wereoften prolonged and were not accompanied by a large rise <strong>in</strong> counterregulatory hormones.

THE SOMOGYI PHENOMENON 91Probability of hypoglycaemic episodes <strong>in</strong> the night1. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21Morn<strong>in</strong>g blood glucose (mmol/I)Figure 4.3 Risk of nocturnal hypoglycaemia accord<strong>in</strong>g to fast<strong>in</strong>g morn<strong>in</strong>g blood glucose (95% Cl) <strong>in</strong>594 nights. Black bars = hypoglycaemic nights; shaded bars = possibly hypoglycaemic nights. Reproducedfrom Hoi-Hansen et al. (2005). With k<strong>in</strong>d permission from Spr<strong>in</strong>ger Science and Bus<strong>in</strong>ess MediaAlthough fast<strong>in</strong>g glucose concentrations were frequently high <strong>in</strong> the patients they studied,this was related directly to a wan<strong>in</strong>g of circulat<strong>in</strong>g plasma <strong>in</strong>sul<strong>in</strong> concentrations. Some<strong>in</strong>vestigators have demonstrated that when hypoglycaemia is experimentally-<strong>in</strong>duced dur<strong>in</strong>gthe night, this can raise blood glucose on the follow<strong>in</strong>g morn<strong>in</strong>g, even if circulat<strong>in</strong>g plasma<strong>in</strong>sul<strong>in</strong> concentrations are ma<strong>in</strong>ta<strong>in</strong>ed (Perriello et al., 1988). However, the additional <strong>in</strong>crease<strong>in</strong> the fast<strong>in</strong>g blood glucose concentration is modest (around 2.0 mmol/l) and its cl<strong>in</strong>icalrelevance is questionable. Other researchers have found no effect on daytime concentrationsof blood glucose after lower<strong>in</strong>g blood glucose to hypoglycaemic levels dur<strong>in</strong>g the night(Hirsch et al., 1990). Careful analysis of data collected both by self-monitor<strong>in</strong>g of bloodglucose (Havl<strong>in</strong> and Cryer, 1987) and by cont<strong>in</strong>uous glucose monitor<strong>in</strong>g (Hoi-Hansen et al.,2005) (Figure 4.3) dur<strong>in</strong>g everyday activities, has also shown that nocturnal hypoglycaemiadoes not provoke rebound fast<strong>in</strong>g hyperglycaemia.Many <strong>in</strong>sul<strong>in</strong>-treated diabetic patients experience high fast<strong>in</strong>g blood glucose levels butthis common cl<strong>in</strong>ical problem is essentially a consequence of <strong>in</strong>adequate basal <strong>in</strong>sul<strong>in</strong>replacement overnight. The important mechanisms that contribute to fast<strong>in</strong>g hyperglycaemiaappear to be a comb<strong>in</strong>ation of wan<strong>in</strong>g plasma <strong>in</strong>sul<strong>in</strong> levels and glucose release fromthe liver secondary to nocturnal spikes of growth hormone secretion (Campbell et al.,1985), a physiological process termed the ‘dawn phenomenon’. High blood glucose levelsfollow<strong>in</strong>g symptomatic nocturnal hypoglycaemia may also result from excessive <strong>in</strong>takeof oral carbohydrate, <strong>in</strong>gested as treatment of the hypoglycaemia, rather than a powerfulcounterregulatory response, which is usually suppressed at night. The cl<strong>in</strong>ical message istherefore clear: fast<strong>in</strong>g hyperglycaemia <strong>in</strong>dicates a need for adjust<strong>in</strong>g basal <strong>in</strong>sul<strong>in</strong> <strong>in</strong> termsof type or tim<strong>in</strong>g rather than reduc<strong>in</strong>g the dose. Some useful cl<strong>in</strong>ical steps to be undertaken<strong>in</strong> patients present<strong>in</strong>g with this problem are listed <strong>in</strong> Box 4.2.

92 NOCTURNAL HYPOGLYCAEMIABox 4.2 Cl<strong>in</strong>ical approach to high fast<strong>in</strong>g blood glucose complicated by nocturnalhypoglycaemia1. Measure blood glucose at 2–3 a.m. over a few days.2. If nocturnal hypoglycaemia is present, ensure that basal <strong>in</strong>sul<strong>in</strong> is taken at bedtime(i.e., split pre-mixed even<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>).3. Progressively <strong>in</strong>crease bedtime long-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> <strong>in</strong> doses of 2–4 units whilecheck<strong>in</strong>g with 3 a.m. blood glucose measurements that this is not precipitat<strong>in</strong>gnocturnal hypoglycaemia4. Use a long-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogue, either glarg<strong>in</strong>e or detemir.5. Teach patients to take an appropriate (but not excessive) quantity of a high energyglucose dr<strong>in</strong>k, orange juice or glucose as sweets or tablets to treat nocturnal hypoglycaemicepisodes.6. When available, consider obta<strong>in</strong><strong>in</strong>g a cont<strong>in</strong>uous glucose monitor<strong>in</strong>g profile.CLINICAL SOLUTIONS (BOX 4.3)Dietary MeasuresA time honoured approach to reduc<strong>in</strong>g the frequency of nocturnal hypoglycaemia has beento counteract the effect of <strong>in</strong>sul<strong>in</strong> by ensur<strong>in</strong>g that patients eat a bedtime snack. This seemedparticularly important for children who go to bed early and sleep for many hours betweentheir even<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> dose and breakfast on the follow<strong>in</strong>g day. It clearly makes sense for all<strong>in</strong>dividuals tak<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> to measure their blood glucose before bed and to take additionalfood if their blood glucose is low. However, the extent to which protective eat<strong>in</strong>g can preventnocturnal hypoglycaemia <strong>in</strong> patients tak<strong>in</strong>g <strong>in</strong>termittent <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> is limited.Box 4.3Potential remedies for problematical nocturnal hypoglycaemia1. Long and rapid-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues.2. -agonists (terbutal<strong>in</strong>e or salbutamol).3. Appropriate snacks:– uncooked cornstarch– high prote<strong>in</strong> foods.4. CSII.

CLINICAL SOLUTIONS 93The effectiveness of different approaches has generally been assessed by measur<strong>in</strong>g theeffectiveness of different snacks <strong>in</strong> prevent<strong>in</strong>g biochemical or symptomatic episodes ofhypoglycaemia. Before the advent of cont<strong>in</strong>uous blood glucose monitor<strong>in</strong>g, such studiesdemanded regular blood sampl<strong>in</strong>g for measurement of glucose and s<strong>in</strong>ce this generallyrequired admission to hospital for study, these <strong>in</strong>vestigations were not exactly exam<strong>in</strong><strong>in</strong>ga typical cl<strong>in</strong>ical situation. Furthermore, the number of subjects studied has generally beenrelatively small, and although sufficient to measure changes <strong>in</strong> blood glucose concentrationor frequency of biochemical hypoglycaemia, the studies have been <strong>in</strong>adequately poweredto determ<strong>in</strong>e effects on the frequency of severe episodes. Some studies have exam<strong>in</strong>edthe potential of carbohydrate foods that are absorbed slowly and thus able to counter thehypoglycaemic effect of <strong>in</strong>sul<strong>in</strong> over longer periods. Uncooked cornstarch, which has a verylow glycaemic <strong>in</strong>dex, has been studied <strong>in</strong>tensively, particularly because it is used successfullyto prevent severe hypoglycaemia <strong>in</strong> glycogen storage diseases (Goldberg and Slonim, 1993).In one study, when young people were given uncooked cornstarch <strong>in</strong>corporated <strong>in</strong>to a normalbedtime snack, the <strong>in</strong>cidence of symptomatic and biochemical nocturnal hypoglycaemiawas three-fold lower (Kaufman and Devgan, 1996). A bl<strong>in</strong>ded randomised trial reported asimilarly lower frequency of nocturnal episodes (Kaufman et al., 1997). Other work hasexam<strong>in</strong>ed the effect of snacks rich <strong>in</strong> prote<strong>in</strong> or fat.Different types of snack were compared aga<strong>in</strong>st placebo <strong>in</strong> a trial us<strong>in</strong>g a crossoverdesign <strong>in</strong> 15 adults with type 1 diabetes (Kalergis et al., 2003). When patients retired tobed with blood glucose greater than 10.0 mmol/l, nocturnal episodes were not observed.At a pre-bedtime glucose below 7.0 mmol/l, a high prote<strong>in</strong> snack prevented any episodeof nocturnal hypoglycaemia by contrast with either a standard snack or one conta<strong>in</strong><strong>in</strong>gcornstarch (Figure 4.4). Thus the cl<strong>in</strong>ical data measur<strong>in</strong>g the effectiveness of cornstarch are100%80%% Nights60%40%20%0% Pl S CS Prot Pl S CS Prot Pl S CS Prot10 mmol/LBedtime blood glucose categoryFigure 4.4 Effect of snacks and bedtime blood glucose concentration on frequency of nocturnalhypoglycaemia. Open box = neither nocturnal hypoglycaemia nor morn<strong>in</strong>g hypoglycaemia; solidblack box = hypoglycaemia; striped box = morn<strong>in</strong>g hypoglycaemia only. Snack type: Pl = placebo;S = standard diet; CS = cornstarch; prot = prote<strong>in</strong>

94 NOCTURNAL HYPOGLYCAEMIAconflict<strong>in</strong>g. S<strong>in</strong>ce uncooked cornstarch is not easy to prepare <strong>in</strong> a digestible form, it is notsurpris<strong>in</strong>g that it is not used widely to prevent nocturnal hypoglycaemia.An alternative approach to dietary supplements is to reduce the rate of absorption ofcarbohydrates us<strong>in</strong>g a disaccharidase <strong>in</strong>hibitor such as acarbose. Three studies of theseagents have exam<strong>in</strong>ed their effects <strong>in</strong> patients with type 1 diabetes. McCulloch et al. (1983)studied the effect of acarbose on the risk of overnight hypoglycaemia, and found that therisk of symptomatic nocturnal hypoglycaemia was lowered by 39%. Taira et al. (2000)reported similar benefits us<strong>in</strong>g voglibose. However, a recent study reported no benefit ofacarbose over placebo when both pharmaceutical and snack<strong>in</strong>g <strong>in</strong>terventions were comparedwith respect to their effect on prevent<strong>in</strong>g hypoglycaemia (Raju et al., 2006). In the light ofthese limited and conflict<strong>in</strong>g data it seems unlikely that acarbose will never be widely used,particularly as sucrose-conta<strong>in</strong><strong>in</strong>g products cannot be used as a treatment for hypoglycaemiawhen acarbose is be<strong>in</strong>g taken.Pharmaceutical InterventionsThere are <strong>in</strong>dications that -agonists may have some use <strong>in</strong> reduc<strong>in</strong>g the risk of nocturnalhypoglycaemia. For some years <strong>in</strong>haled terbutal<strong>in</strong>e has been proposed as a method ofelevat<strong>in</strong>g blood glucose. In the early 1990s, Wiethop and Cryer (1993) demonstrated thatits oral or subcutaneous delivery follow<strong>in</strong>g <strong>in</strong>duced hypoglycaemia, led to a rise <strong>in</strong> bloodglucose compared to placebo, an effect that lasted for some hours. More recently, Wright andWales (2003) reported that children with type 1 diabetes who were receiv<strong>in</strong>g treatment forasthma had fewer episodes of nocturnal hypoglycaemia when compared to a non-asthmaticgroup of children <strong>in</strong> a survey last<strong>in</strong>g three months, and they implicated a beneficial effectof -agonist therapy. Raju et al. (2006) also compared the effect of <strong>in</strong>haled terbutal<strong>in</strong>e withother therapeutic remedies such as cornstarch, standard snacks and acarbose on nocturnalblood glucose levels <strong>in</strong> patients with type 1 diabetes. They found that terbutal<strong>in</strong>e preventednocturnal hypoglycaemia <strong>in</strong> all 15 subjects. However, although this treatment offered thegreatest protection aga<strong>in</strong>st hypoglycaemia at night, it also led to the highest fast<strong>in</strong>g bloodglucose among the different remedies, <strong>in</strong>dicat<strong>in</strong>g that additional work needs to be done toestablish this treatment as a realistic therapeutic option.Tim<strong>in</strong>g and Type of Insul<strong>in</strong>, Includ<strong>in</strong>g Insul<strong>in</strong> AnaloguesThe <strong>in</strong>troduction of <strong>in</strong>sul<strong>in</strong> analogues with pharmacok<strong>in</strong>etic properties that bear more resemblanceto physiological <strong>in</strong>sul<strong>in</strong> profiles <strong>in</strong> non-diabetic <strong>in</strong>dividuals highlighted the potentialof such preparations to lower the risk of hypoglycaemia. Over a full 24 hours, the overallfrequency of symptomatic hypoglycaemia <strong>in</strong> trials of rapid-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues hasbeen modestly lower than with conventional <strong>in</strong>sul<strong>in</strong>s, but a consistent f<strong>in</strong>d<strong>in</strong>g has been alower rate of nocturnal hypoglycaemia. It appears that the tendency of conventional soluble<strong>in</strong>sul<strong>in</strong> to self-associate <strong>in</strong>to hexamers at therapeutic concentrations leads to <strong>in</strong>creas<strong>in</strong>gplasma <strong>in</strong>sul<strong>in</strong> levels with repeated <strong>in</strong>jection. S<strong>in</strong>ce rapid-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues separate<strong>in</strong>to s<strong>in</strong>gle molecules much more readily, accumulation of <strong>in</strong>sul<strong>in</strong> is less likely and therisk of nocturnal hypoglycaemia is subsequently lower. The frequency of nocturnal hypoglycaemiaobserved <strong>in</strong> cl<strong>in</strong>ical trials has been variable. Rates of nocturnal hypoglycaemiahave been over 50% lower <strong>in</strong> some trials <strong>in</strong>volv<strong>in</strong>g patients with type 1 diabetes with strict

CONCLUSIONS 95glycaemic control (Heller et al., 1999; Heller et al., 2004), and although this has generallybeen demonstrated for symptomatic episodes, these data might reflect a lower frequency ofsevere hypoglycaemia (Holleman et al., 1997).The long-act<strong>in</strong>g analogues – <strong>in</strong>sul<strong>in</strong>s glarg<strong>in</strong>e and detemir – which provide basal <strong>in</strong>sul<strong>in</strong>replacement, also appear to lower the risk of nocturnal hypoglycaemia. They have a longerduration of action when compared to isophane (NPH) <strong>in</strong>sul<strong>in</strong>, together with less of a peak<strong>in</strong> their time-action profile and a more consistent duration of action (Barnett, 2003). Bothof these properties probably contribute to the lower rates of nocturnal hypoglycaemia thathave been observed dur<strong>in</strong>g cl<strong>in</strong>ical trials. Relative risk reductions of around 30% have beenreported for long-act<strong>in</strong>g preparations <strong>in</strong> trials <strong>in</strong>volv<strong>in</strong>g patients with type 1 (De Leeuw et al.,2005; Pieber et al., 2000) and with type 2 diabetes (Hermansen et al., 2006; Yki-Jarv<strong>in</strong>enet al., 2000).The comb<strong>in</strong>ation of both rapid and long-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues might be expected to havea particularly powerful effect <strong>in</strong> lower<strong>in</strong>g the risk of nocturnal hypoglycaemia. In the fewstudies compar<strong>in</strong>g comb<strong>in</strong>ations of <strong>in</strong>sul<strong>in</strong> analogues to conventional <strong>in</strong>sul<strong>in</strong>s this appears tobe the case. Hermansen et al. (2004) reported a 55% lower rate of symptomatic nocturnalhypoglycaemia when us<strong>in</strong>g an <strong>in</strong>sul<strong>in</strong> detemir/aspart comb<strong>in</strong>ation as basal-bolus therapy <strong>in</strong>patients with type 1 diabetes, which was accompanied by a modest but significant fall <strong>in</strong> HbA 1cof 0.2%. Ashwell et al. (2006) observed a fall <strong>in</strong> HbA 1c of 0.5% us<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e andlispro <strong>in</strong> a basal-bolus regimen <strong>in</strong> patients with type 1 diabetes, while nocturnal hypoglycaemiawas 44% lower <strong>in</strong> frequency. However, nocturnal hypoglycaemia was not eradicated,lead<strong>in</strong>g to the conclusion that although <strong>in</strong>sul<strong>in</strong> analogues may help to reduce the side-effectsof <strong>in</strong>sul<strong>in</strong> therapy, they do not approach the requirements of therapeutic <strong>in</strong>sul<strong>in</strong> delivery.Cont<strong>in</strong>uous Subcutaneous Insul<strong>in</strong> Infusion (CSII)S<strong>in</strong>ce nocturnal hypoglycaemia is largely the result of <strong>in</strong>adequate basal <strong>in</strong>sul<strong>in</strong> replacement,one would expect that the most effective method of basal <strong>in</strong>sul<strong>in</strong> delivery currently available,CSII, could limit the frequency of nocturnal hypoglycaemia (Pickup and Keen, 2002).However, early studies reported surpris<strong>in</strong>gly little effect on hypoglycaemia, perhaps becauseof a failure to tra<strong>in</strong> patients <strong>in</strong> the essential related skills of carbohydrate and <strong>in</strong>sul<strong>in</strong>dose adjustment. More recent work has <strong>in</strong>dicated that CSII can reduce overall rates ofhypoglycaemia (Bode et al., 1996; Boland et al., 1999; Kanc et al., 1998), but specificreport<strong>in</strong>g on rates of nocturnal episodes is unusual. Furthermore, few trials have used arandomised design, suggest<strong>in</strong>g that a lower risk might relate to other characteristics ofthose who use CSII rather than to the technology itself. Thus the amount to which modernpump therapy lowers the risk of nocturnal hypoglycaemia has still be to be established.Nevertheless, <strong>in</strong> those who have experienced recurrent nocturnal episodes and who have notimproved after a trial of <strong>in</strong>sul<strong>in</strong> analogues, it seems worthwhile undertak<strong>in</strong>g a trial of CSII.CONCLUSIONS• Nocturnal hypoglycaemia rema<strong>in</strong>s an unresolved cl<strong>in</strong>ical side-effect of <strong>in</strong>sul<strong>in</strong> therapyprevent<strong>in</strong>g the atta<strong>in</strong>ment of strict glycaemic control for many people. It contributes tomorbidity, and perhaps mortality, <strong>in</strong> patients with type 1 diabetes.

96 NOCTURNAL HYPOGLYCAEMIA• Nocturnal hypoglycaemia is caused chiefly by the limitations of current methods of<strong>in</strong>sul<strong>in</strong> delivery. The <strong>in</strong>ability of subcutaneous <strong>in</strong>sul<strong>in</strong> therapy, particularly conventionalbasal <strong>in</strong>sul<strong>in</strong>s, to ma<strong>in</strong>ta<strong>in</strong> low-level stable <strong>in</strong>sul<strong>in</strong> concentrations overnight, leads both tonocturnal hypoglycaemia and fast<strong>in</strong>g hyperglycaemia.• In addition to the limitations of <strong>in</strong>sul<strong>in</strong> delivery, hypoglycaemia is also more commonas a result of dim<strong>in</strong>ished counterregulatory responses overnight, associated <strong>in</strong> part withsleep, which has a specific <strong>in</strong>hibitory effect on physiological defences to hypoglycaemia,and a sup<strong>in</strong>e posture which suppresses autonomic responses.• Nocturnal hypoglycaemia is a particular problem <strong>in</strong> children, partly as a consequence ofthe long period of fast<strong>in</strong>g between their even<strong>in</strong>g meal and their breakfast.• The risk of nocturnal hypoglycaemia is greatest <strong>in</strong> those who have a bedtime glucosebelow 7.0 mmol/l. Bedtime snacks can reduce the risk dur<strong>in</strong>g the early part of the night.Specific foods (uncooked cornstarch or prote<strong>in</strong> rich snacks) reduce the risk of nocturnalhypoglycaemia <strong>in</strong> some studies.• Rebound hyperglycaemia (the ‘Somogyi phenomenon’) is rarely caused by counterregulatoryhormone release provoked by an overnight hypoglycaemic episode, s<strong>in</strong>ce hormonalsecretion is generally suppressed. It is ma<strong>in</strong>ly a consequence of wan<strong>in</strong>g of circulat<strong>in</strong>gplasma <strong>in</strong>sul<strong>in</strong> levels and should be treated by an adjustment <strong>in</strong> the tim<strong>in</strong>g and type of<strong>in</strong>sul<strong>in</strong> used rather than a reduction <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dose.• The problem of nocturnal hypoglycaemia may respond to the use of <strong>in</strong>sul<strong>in</strong> analogues(both rapid and long-act<strong>in</strong>g) or to CSII with an <strong>in</strong>sul<strong>in</strong> pump. However, its eradicationwill depend upon new methods of <strong>in</strong>sul<strong>in</strong> delivery.REFERENCESAshwell SG, Amiel SA, Bilous RW, Dashora U, Heller SR, Hepburn DA et al. (2006). Improvedglycaemic control with <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e plus <strong>in</strong>sul<strong>in</strong> lispro: a multicentre, randomized, cross-overtrial <strong>in</strong> people with type 1 diabetes. Diabetic Medic<strong>in</strong>e 23: 285–92.Banarer S, Cryer PE (2003). Sleep-related hypoglycemia-associated autonomic failure <strong>in</strong> type 1diabetes: reduced awaken<strong>in</strong>g from sleep dur<strong>in</strong>g hypoglycemia. <strong>Diabetes</strong> 52: 1195–203.Barnett AH (2003). A review of basal <strong>in</strong>sul<strong>in</strong>s. Diabetic Medic<strong>in</strong>e 20: 873–85.Bendtson I, Kverneland A, Pramm<strong>in</strong>g S, B<strong>in</strong>der C (1988). Incidence of nocturnal hypoglycaemia <strong>in</strong><strong>in</strong>sul<strong>in</strong>-dependent diabetes. Acta Endocr<strong>in</strong>ologica 223: 543–8.Bendtson I, Gade J, Theilgard A, B<strong>in</strong>der C (1992). Cognitive function <strong>in</strong> type 1 IDD patients afternocturnal hypoglycaemia. Diabetologia 35: 898–903.Bendtson I, Rosenfalck AM, B<strong>in</strong>der C (1993). Nocturnal versus diurnal hormonal counterregulation tohypoglycemia <strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabetic patients. Acta Endocr<strong>in</strong>ologica 128: 109–15.Beregszaszi M, Tubiana-Rufi N, Benali K, Noel M, Bloch J, Czernichow P (1997). Nocturnal hypoglycemia<strong>in</strong> children and adolescents with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: prevalence and riskfactors. Journal of Pediatrics 131: 27–33.Bjorgaas M, Gimse R, Vik T, Sand T (1997). Cognitive function <strong>in</strong> type 1 diabetic children with andwithout episodes of severe hypoglycaemia. Acta Paediatrica 86: 148–53.Bode BW, Steed RD, Davidson PC (1996). Reduction <strong>in</strong> severe hypoglycemia with long-term cont<strong>in</strong>uoussubcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion <strong>in</strong> type I diabetes. <strong>Diabetes</strong> Care 19: 324–7.

REFERENCES 97Boland EA, Grey M, Oesterle A, Fredrickson L, Tamborlane WV (1999). Cont<strong>in</strong>uous subcutaneous<strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion. A new way to lower risk of severe hypoglycemia, improve metabolic control, andenhance cop<strong>in</strong>g <strong>in</strong> adolescents with type 1 diabetes. <strong>Diabetes</strong> Care 22: 1779–84.Boland E, Monsod T, Delucia M, Brandt CA, Fernando S, Tamborlane WV (2001). Limitations ofconventional methods of self-monitor<strong>in</strong>g of blood glucose: lessons learned from 3 days of cont<strong>in</strong>uousglucose sens<strong>in</strong>g <strong>in</strong> pediatric patients with type 1 diabetes. <strong>Diabetes</strong> Care 24: 1858–62.Campbell PJ, Bolli GB, Cryer PE, Gerich JE (1985). Pathogenesis of the dawn phenomenon <strong>in</strong> patientswith <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Accelerated glucose production and impaired glucoseutilization due to nocturnal surges <strong>in</strong> growth hormone secretion. New England Journal of Medic<strong>in</strong>e312: 1473–9.Dagogo-Jack SE, Craft S, Cryer PE (1993). Hypoglycemia-associated autonomic failure <strong>in</strong> <strong>in</strong>sul<strong>in</strong>dependentdiabetes mellitus. Journal of Cl<strong>in</strong>ical Investigation 91: 819–28.De Leeuw I, Vague P, Selam JL, Skeie S, Lang H, Draeger E, Elte JW (2005). Insul<strong>in</strong> detemir used<strong>in</strong> basal-bolus therapy <strong>in</strong> people with type 1 diabetes is associated with a lower risk of nocturnalhypoglycaemia and less weight ga<strong>in</strong> over 12 months <strong>in</strong> comparison to NPH <strong>in</strong>sul<strong>in</strong>. <strong>Diabetes</strong> Obesityand Metabolism 7: 73–82.Dornan TL, Orde Peckar C, Mayon-White VA, Knight AH, Moore RA, Hockaday TDR et al. (1981).Unsuspected hypoglycaemia, haemoglob<strong>in</strong> A1 and diabetic control. Quarterly Journal of Medic<strong>in</strong>e197: 31–8.Gais S, Born J, Peters A, Schultes B, He<strong>in</strong>dl B, Fehm HL, Werner K (2003). Hypoglycemia counterregulationdur<strong>in</strong>g sleep. Sleep 26: 55–9.Gale EA, Tattersall RB (1979). Unrecognised nocturnal hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetics.Lancet 1: 1049–52.Gale EA, Kurtz AB, Tattersall RB (1980). In search of the Somogyi effect. Lancet 2: 279–82.Goldberg T, Slonim AE (1993). Nutrition therapy for hepatic glycogen storage diseases. Journal ofthe American Dietetic Association 93: 1423–30.Golden MP, Ingersoll GM, Brack CJ, Russell BA, Wright JC, Huberty TJ (1989). Longitud<strong>in</strong>alrelationship of asymptomatic hypoglycemia to cognitive function <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 12:89–93.Havl<strong>in</strong> CE, Cryer PE (1987). Nocturnal hypoglycemia does not commonly result <strong>in</strong> major morn<strong>in</strong>ghyperglycemia <strong>in</strong> patients with diabetes mellitus. <strong>Diabetes</strong> Care 10: 141–7.Heller SR, Cryer PE (1991). Reduced neuroendocr<strong>in</strong>e and symptomatic responses to subsequenthypoglycemia after 1 episode of hypoglycemia <strong>in</strong> non-diabetic humans. <strong>Diabetes</strong> 40: 223–6.Heller SR, Amiel SA, Mansell P (1999). Effect of the fast-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analog lispro on the risk ofnocturnal hypoglycemia dur<strong>in</strong>g <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy. <strong>Diabetes</strong> Care 22: 1607–11.Heller SR, Colagiuri S, Vaaler S, Wolffenbuttel BH, Koelendorf K, Friberg HH et al. (2004). <strong>Hypoglycaemia</strong>with <strong>in</strong>sul<strong>in</strong> aspart: a double-bl<strong>in</strong>d, randomised, crossover trial <strong>in</strong> subjects with type 1diabetes. Diabetic Medic<strong>in</strong>e 21: 769–75.Hermansen K, Fonta<strong>in</strong>e P, Kukolja KK, Peterkova V, Leth G, Gall MA (2004). Insul<strong>in</strong> analogues(<strong>in</strong>sul<strong>in</strong> detemir and <strong>in</strong>sul<strong>in</strong> aspart) versus traditional human <strong>in</strong>sul<strong>in</strong>s (NPH <strong>in</strong>sul<strong>in</strong> and regularhuman <strong>in</strong>sul<strong>in</strong>) <strong>in</strong> basal-bolus therapy for patients with type 1 diabetes. Diabetologia 47: 622–9.Hermansen K, Davies M, Derez<strong>in</strong>ski T, Mart<strong>in</strong>ez Ravn G, Clauson P, Home P (2006). A 26-week,randomized, parallel, treat-to-target trial compar<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> detemir with NPH <strong>in</strong>sul<strong>in</strong> as add-ontherapy to oral glucose-lower<strong>in</strong>g drugs <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-naive people with type 2 diabetes. <strong>Diabetes</strong> Care29: 1269–74.Hirsch IB, Smith LJ, Havl<strong>in</strong> CE, Shah SD, Clutter WE, Cryer PE (1990). Failure of nocturnalhypoglycemia to cause daytime hyperglycemia <strong>in</strong> patients with IDDM. <strong>Diabetes</strong> Care 13: 133–42.Hirsch IB, Heller SR, Cryer PE (1991). Increased symptoms of hypoglycaemia <strong>in</strong> the stand<strong>in</strong>g position<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Cl<strong>in</strong>ical Science 80: 583–6.Hoi-Hansen T, Pedersen-Bjergaard U, Thorste<strong>in</strong>sson B (2005). The Somogyi phenomenon revisitedus<strong>in</strong>g cont<strong>in</strong>uous glucose monitor<strong>in</strong>g <strong>in</strong> daily life. Diabetologia 48: 2437–8.

98 NOCTURNAL HYPOGLYCAEMIAHolleman F, Schmitt H, Rottiers R, Rees A, Symanowski S, Anderson JH, The Benelux-UK Insul<strong>in</strong>Lispro Study Group (1997). Reduced frequency of severe hypoglycemia and coma <strong>in</strong> well-controlledIDDM patients treated with <strong>in</strong>sul<strong>in</strong> lispro. <strong>Diabetes</strong> Care 20: 1827–32.Jones TW, Porter P, Sherw<strong>in</strong> RS, Davis EA, O’Leary P, Frazer F et al. (1998). Decreased ep<strong>in</strong>ephr<strong>in</strong>eresponses to hypoglycemia dur<strong>in</strong>g sleep. New England Journal of Medic<strong>in</strong>e 338: 1657–62.Kalergis M, Schiffr<strong>in</strong> A, Gougeon R, Jones PJ, Yale JF (2003). Impact of bedtime snack compositionon prevention of nocturnal hypoglycemia <strong>in</strong> adults with type 1 diabetes undergo<strong>in</strong>g <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong>management us<strong>in</strong>g lispro <strong>in</strong>sul<strong>in</strong> before meals: a randomized, placebo-controlled, crossover trial.<strong>Diabetes</strong> Care 26: 9–15.Kanc K, Janssen MM, Keulen ET, Jacobs MA, Popp-Snijders C, Snoek FJ, He<strong>in</strong>e RJ (1998). Substitutionof night-time cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion therapy for bedtime NPH <strong>in</strong>sul<strong>in</strong> <strong>in</strong> amultiple <strong>in</strong>jection regimen improves counterregulatory hormonal responses and warn<strong>in</strong>g symptomsof hypoglycaemia <strong>in</strong> IDDM. Diabetologia 41: 322–9.Kaufman FR, Devgan S (1996). Use of uncooked cornstarch to avert nocturnal hypoglycemia <strong>in</strong>children and adolescents with type I diabetes. Journal of <strong>Diabetes</strong> and Its Complications 10: 84–7.Kaufman FR, Halvorson M, Kaufman ND (1997). Evaluation of a snack bar conta<strong>in</strong><strong>in</strong>g uncookedcornstarch <strong>in</strong> subjects with diabetes. <strong>Diabetes</strong> Research and Cl<strong>in</strong>ical Practice 35: 27–33.Kaufman FR, Aust<strong>in</strong> J, Ne<strong>in</strong>ste<strong>in</strong> A, Jeng L, Halvorson M, Devoe DJ, Pitukcheewanont P (2002).Nocturnal hypoglycemia detected with the Cont<strong>in</strong>uous Glucose Monitor<strong>in</strong>g System <strong>in</strong> pediatricpatients with type 1 diabetes. Journal of Pediatrics 141: 625–30.K<strong>in</strong>g P, Kong MF, Park<strong>in</strong> H, Macdonald IA, Tattersall RB (1998). Well-be<strong>in</strong>g, cerebral function, andphysical fatigue after nocturnal hypoglycemia <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 21: 341–5.Kulcu E, Tamada JA, Reach G, Potts RO, Lesho MJ (2003). Physiological differences between<strong>in</strong>terstitial glucose and blood glucose measured <strong>in</strong> human subjects. <strong>Diabetes</strong> Care 26: 2405–9.Matyka KA, Crowne EC, Havel PJ, Macdonald IA, Matthews D, Dunger DB (1999a). Counterregulationdur<strong>in</strong>g spontaneous nocturnal hypoglycemia <strong>in</strong> prepubertal children with type 1 diabetes.<strong>Diabetes</strong> Care 22: 1144–50.Matyka KA, Wigg L, Pramm<strong>in</strong>g S, Stores G, Dunger DB (1999b). Cognitive function and mood afterprofound nocturnal hypoglycaemia <strong>in</strong> prepubertal children with conventional <strong>in</strong>sul<strong>in</strong> treatment fordiabetes. Archives of Disease <strong>in</strong> Childhood, 81: 138–42.Matyka KA (2002). Sweet dreams? Nocturnal hypoglycemia <strong>in</strong> children with type 1 diabetes. Pediatric<strong>Diabetes</strong> 3: 74–81.McCulloch DK, Kurtz AB, Tattersall RB (1983). A new approach to the treatment of nocturnalhypoglycemia us<strong>in</strong>g alpha-glucosidase <strong>in</strong>hibition. <strong>Diabetes</strong> Care 6: 483–7.McGowan K, Thomas W, Moran A (2002). Spurious report<strong>in</strong>g of nocturnal hypoglycemia by CGMS<strong>in</strong> patients with tightly controlled type 1 diabetes. <strong>Diabetes</strong> Care 25: 1499–503.Oswald I (1987). The normal record of sleep. In: A Textbook of Cl<strong>in</strong>ical Neurophysiology. Halliday AM,Butler SR and Paul R, eds. John Wiley & Sons Inc., New York: 173–85.Perriello G, De Feo P, Torlone E, Calc<strong>in</strong>aro F, Ventura MM, Basta G et al. (1988). The effectof asymptomatic nocturnal hypoglycemia on glycemic control <strong>in</strong> diabetes mellitus. New EnglandJournal of Medic<strong>in</strong>e 319: 1233–9.Pickup J, Keen H (2002). Cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion at 25 years: evidence base for theexpand<strong>in</strong>g use of <strong>in</strong>sul<strong>in</strong> pump therapy <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care 25: 593–8.Pieber TR, Eugene-Jolch<strong>in</strong>e I, Derobert E (2000). Efficacy and safety of HOE 901 versus NPH <strong>in</strong>sul<strong>in</strong><strong>in</strong> patients with type 1 diabetes. The European Study Group of HOE 901 <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong>Care 23: 157–62.Porter PA, Keat<strong>in</strong>g B, Byrne G, Jones TW (1997). Incidence and predictive criteria of nocturnalhypoglycemia <strong>in</strong> young children with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Journal of Pediatrics130: 366–72.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Bendtson I, Ronn B, B<strong>in</strong>der C (1985). Nocturnal hypoglycaemia <strong>in</strong>patients receiv<strong>in</strong>g conventional treatment with <strong>in</strong>sul<strong>in</strong>. British Medical Journal 291: 376–9.

REFERENCES 99Raju B, Arbelaez AM, Breckenridge SM, Cryer PE (2006). Nocturnal hypoglycemia <strong>in</strong> type 1 diabetes:an assessment of preventive bedtime treatments. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism91: 2087–92.Rizza RA, Gerich JE, Haymond MW, Westland RE, Hall LD, Clemens AH, Service FJ (1980).Control of blood sugar <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes: comparison of an artificial endocr<strong>in</strong>e pancreas,cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion and <strong>in</strong>tensified conventional <strong>in</strong>sul<strong>in</strong> therapy. New EnglandJournal of Medic<strong>in</strong>e 303: 1313–18.Rob<strong>in</strong>son AM, Park<strong>in</strong> HM, Macdonald IA, Tattersall RB (1994). Physiological response to posturalchange dur<strong>in</strong>g mild hypoglycaemia <strong>in</strong> patients with IDDM. Diabetologia 37: 1241–50.Rovet JF, Ehrlich RM (1999). The effect of hypoglycemic seizures on cognitive function <strong>in</strong> childrenwith diabetes: a 7-year prospective study. Journal of Pediatrics 134: 503–6.Ryan C, Vega A, Drash A (1985). Cognitive deficits <strong>in</strong> adolescents who developed diabetes early <strong>in</strong>life. Pediatrics 75: 921–7.Shalwitz RA, Farkas-Hirsch R, White NH, Santiago JV (1990). Prevalence and consequences ofnocturnal hypoglycemia among conventionally treated children with diabetes mellitus. Journal ofPediatrics 116: 685–9.Somogyi M (1959). Exacerbation of diabetes by excess <strong>in</strong>sul<strong>in</strong> action. American Journal of Medic<strong>in</strong>e26: 169–91.Taira M, Takasu N, Komiya I, Taira T, Tanaka H (2000). Voglibose adm<strong>in</strong>istration before the even<strong>in</strong>gmeal improves nocturnal hypoglycemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetic patients with <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong>therapy. Metabolism 49: 440–3.Veneman T, Mitrakou A, Mokan M, Cryer P, Gerich J (1993.) Induction of hypoglycemia unawarenessby asymptomatic nocturnal hypoglycemia. <strong>Diabetes</strong> 42: 1233–7.Vervoort G, Goldschmidt HM, van Doorn LG (1996). Nocturnal blood glucose profiles <strong>in</strong> patients withtype 1 diabetes mellitus on multiple (> or = 4) daily <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jection regimens. Diabetic Medic<strong>in</strong>e13: 794–9.Wentholt IM, Vollebregt MA, Hart AA, Hoekstra JB, DeVries JH (2005). Comparison of a needletypeand a microdialysis cont<strong>in</strong>uous glucose monitor <strong>in</strong> type 1 diabetic patients. <strong>Diabetes</strong> Care 28:2871–6.Wh<strong>in</strong>cup G, Milner RD (1987). Prediction and management of nocturnal hypoglycaemia <strong>in</strong> diabetes.Archives of Disease <strong>in</strong> Childhood 62: 333–7.Wiethop BV, Cryer PE (1993). Alan<strong>in</strong>e and terbutal<strong>in</strong>e <strong>in</strong> treatment of hypoglycemia <strong>in</strong> IDDM.<strong>Diabetes</strong> Care 16: 1131–6.Wright NP, Wales JK (2003). The <strong>in</strong>cidence of hypoglycaemia <strong>in</strong> children with type 1 diabetes andtreated asthma. Archives of Disease <strong>in</strong> Childhood 88: 155–6.Yki-Jarv<strong>in</strong>en H, Dressler A, Ziemen M (2000). Less nocturnal hypoglycemia and better post-d<strong>in</strong>nerglucose control with bedtime <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e compared with bedtime NPH <strong>in</strong>sul<strong>in</strong> dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>comb<strong>in</strong>ation therapy <strong>in</strong> type 2 diabetes. HOE 901/3002 Study Group. <strong>Diabetes</strong> Care 23: 1130–6.

5 Moderators, Monitor<strong>in</strong>g andManagement of <strong>Hypoglycaemia</strong>Tristan Richardson and David KerrINTRODUCTIONDespite advances <strong>in</strong> <strong>in</strong>sul<strong>in</strong> pharmacology and delivery and <strong>in</strong> patient education, the lifetimefrequency of symptomatic hypoglycaemia rema<strong>in</strong>s substantial, with the average patient likelyto experience thousands of episodes over the course of their life with <strong>in</strong>sul<strong>in</strong>-treated diabetes.Furthermore, there are a number of everyday factors that cont<strong>in</strong>ue to <strong>in</strong>fluence (moderate) therisk, presentation and rate of recovery from low blood glucose concentrations (Table 5.1, andsee Chapter 3). An understand<strong>in</strong>g of the predispos<strong>in</strong>g factors that <strong>in</strong>fluence hypoglycaemia isimportant to allow appropriate advice and education to be given to patients, and to appreciatethe relative importance of different moderators and reduce their risks (real and perceived),particularly of caus<strong>in</strong>g recurrent severe hypoglycaemia.RISK FACTORS FOR THE DEVELOPMENTOF HYPOGLYCAEMIAIn physiological terms, current <strong>in</strong>sul<strong>in</strong> regimens are less than ideal. Imperfect <strong>in</strong>sul<strong>in</strong> is oftengiven at the ‘wrong’ time, <strong>in</strong> the ‘wrong’ place and at the ‘wrong’ dose. Unsurpris<strong>in</strong>gly, amismatch between <strong>in</strong>sul<strong>in</strong> and carbohydrate absorption frequently occurs, lead<strong>in</strong>g to over<strong>in</strong>sul<strong>in</strong>isationand the potential risk of hypoglycaemia. Several other factors can <strong>in</strong>fluence<strong>in</strong>sul<strong>in</strong> absorption after subcutaneous <strong>in</strong>jection <strong>in</strong>clud<strong>in</strong>g:• depth of the <strong>in</strong>jection;• angle of the needle for giv<strong>in</strong>g the <strong>in</strong>jection;• site of the <strong>in</strong>jection;• presence of lipohypertrophy at <strong>in</strong>jection site;• time of day that <strong>in</strong>jection is given;• phase of the menstrual cycle;• relationship with exercise;<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

102 MODERATORS, MONITORING AND MANAGEMENTTable 5.1Causes of hypoglycaemia <strong>in</strong> patients with diabetesChange <strong>in</strong> <strong>in</strong>sul<strong>in</strong>sensitivityChange <strong>in</strong> <strong>in</strong>sul<strong>in</strong>pharmaco-dynamicsAltered <strong>in</strong>sul<strong>in</strong>:carbohydrate ratioOther related conditions‘Honeymoon period’<strong>in</strong> newly diagnosedtype 1 diabetesPost-partumChange <strong>in</strong> <strong>in</strong>jection site Unplanned exercise Endocr<strong>in</strong>e dysfunction(Addison’s disease,hypopituitarism)Change <strong>in</strong> <strong>in</strong>sul<strong>in</strong> Change <strong>in</strong> social PsychologicalformulationcircumstancesMenstruation Change <strong>in</strong> temperature Breastfeed<strong>in</strong>gAlcoholGastroparesisRenal failureExerciseMalabsorption e.g.Coeliac Disease• ambient temperature;• psychological factors such as mood;• use of other medic<strong>in</strong>es affect<strong>in</strong>g sk<strong>in</strong> blood flow;• for pre-mixed <strong>in</strong>sul<strong>in</strong> preparations – whether the <strong>in</strong>sul<strong>in</strong> has been shaken adequately beforethe <strong>in</strong>jection.Although there is now greater awareness of the importance of educat<strong>in</strong>g patients aboutthe nuances of dietary carbohydrate count<strong>in</strong>g, it is important to remember that the absorptionof food is also variable with<strong>in</strong> <strong>in</strong>dividuals, be<strong>in</strong>g affected by the constituents and size of ameal, and the speed with which it is eaten. In addition, gastric empty<strong>in</strong>g is variable with<strong>in</strong>an <strong>in</strong>dividual with diabetes and is <strong>in</strong>fluenced by the status of the autonomic nervous system(Feldman and Schiller, 1983; V<strong>in</strong>ik et al., 2003).In addition, improv<strong>in</strong>g glycaemic control and lower<strong>in</strong>g HbA 1c per se are also associatedwith a trebl<strong>in</strong>g of risk for hypoglycaemia (The <strong>Diabetes</strong> Control and Complications TrialResearch Group, 1995). However, this alone does not wholly predict the occurrence ofhypoglycaemia. In the <strong>in</strong>tensively treated group <strong>in</strong> the DCCT, HbA 1c only accounted for60% of the risk of severe hypoglycaemia (The <strong>Diabetes</strong> Control and Complications TrialResearch Group, 1997). Additional risk factors have also been identified, many of which arediscussed <strong>in</strong> detail <strong>in</strong> Chapter 3:• previous episodes of severe hypoglycaemia (Bott et al., 1997; The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1997);• long duration of type 1 diabetes (Cox et al., 1994; The <strong>Diabetes</strong> Control and ComplicationsTrial Research Group, 1997);• <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy (The DCCT Research Group, 1991);• strict glycaemic control (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1997);

LIFESTYLE MODERATORS 103• absolute <strong>in</strong>sul<strong>in</strong> deficiency (Bott et al., 1997; The <strong>Diabetes</strong> Control and ComplicationsTrial Research Group, 1997);• sleep (Pramm<strong>in</strong>g et al., 1990; The <strong>Diabetes</strong> Control and Complications Trial ResearchGroup, 1997);• impaired hypoglycaemia awareness (Gold et al., 1994);• alcohol (Richardson et al., 2005b);• exercise (MacDonald, 1987);• pregnancy (Rosen et al., 1995);• impaired renal function (Muhlhauser et al., 1991).Despite these risk factors be<strong>in</strong>g recognised, it is often difficult to untangle the majorprecipitat<strong>in</strong>g factors for a given episode of severe hypoglycaemia. Patient recall is oftenvague, preconceptions may have been applied and amnesia for an event is common. Anecdotally,around a third to half of episodes rema<strong>in</strong> unexpla<strong>in</strong>ed <strong>in</strong> rout<strong>in</strong>e cl<strong>in</strong>ical practice,although patient ‘error’ is still perceived (by healthcare professionals) as the most likely‘cause’ of a hypoglycaemic episode (The DCCT Research Group, 1991).LIFESTYLE MODERATORSNumerous lifestyle <strong>in</strong>fluences predispose towards hypoglycaemia, and some of these, suchas pregnancy (Chapter 10) and exercise (Chapter 14), are discussed elsewhere <strong>in</strong> this book.Alcohol and <strong>Hypoglycaemia</strong>Alcohol is an important risk factor for hypoglycaemia for <strong>in</strong>dividuals treated with <strong>in</strong>sul<strong>in</strong>,with estimates suggest<strong>in</strong>g that up to 20% of severe hypoglycaemic events are attributableto its use (Potter et al., 1982; Nilsson et al., 1988). However, there is noth<strong>in</strong>g to suggestthat (<strong>in</strong> general terms) people with type 1 diabetes adopt a different approach to their use ofalcohol than the rest of the population.Alcohol has been associated with hypoglycaemia <strong>in</strong> several ways:• Ingestion of even small amounts may impair the ability of the <strong>in</strong>dividual to detect theonset of hypoglycaemia at a stage when they are still able to take appropriate action, i.e.,eat some carbohydrate.• <strong>Hypoglycaemia</strong> per se may be mistaken for <strong>in</strong>toxication by observers, with legal andhealth consequences.• Alcohol has been shown <strong>in</strong> some studies to impact directly on gluconeogenesis and/or thecounterregulatory responses to hypoglycaemia (Turner et al., 2001; Kerr et al., 2007).• Recent data <strong>in</strong>dicate that small amounts of alcohol can augment the cognitive deficitassociated with hypoglycaemia <strong>in</strong> <strong>in</strong>dividuals with type 1 diabetes (Cheyne et al., 2004).

104 MODERATORS, MONITORING AND MANAGEMENTIn the past, biochemical hypoglycaemia (<strong>in</strong> non-diabetic <strong>in</strong>dividuals) associated withalcohol <strong>in</strong>toxication was attributed to the toxic effects of impurities associated with theproduction of illicit dr<strong>in</strong>ks (‘hooch’ or ‘moonsh<strong>in</strong>e’). These <strong>in</strong>cluded methanol, gasol<strong>in</strong>e andethyl acetate (Brown and Harvey, 1941). More recently, it was thought that the biochemicaleffects of alcohol were associated with hypoglycaemia <strong>in</strong> three ways:• <strong>in</strong>hibition of gluconeogenesis;• potentiation of the effects of exercise and glucose-lower<strong>in</strong>g agents;• caus<strong>in</strong>g reactive hypoglycaemia <strong>in</strong> susceptible <strong>in</strong>dividuals.More than 90% of an ethanol load is metabolised by the liver, be<strong>in</strong>g catalysed by alcoholdehydrogenase <strong>in</strong>to acetate. The rate-limit<strong>in</strong>g step is ultimately dependent upon the availabilityof nicot<strong>in</strong>amide di-nucleotide (NAD+) <strong>in</strong> a glycogen-replete state. In conditionswhere glycogen stores are depleted and blood glucose is be<strong>in</strong>g ma<strong>in</strong>ta<strong>in</strong>ed by hepatic glucoseproduction (e.g. <strong>in</strong> malnourished <strong>in</strong>dividuals and after prolonged exercise), alcohol <strong>in</strong>gestionmay lead directly to a fall <strong>in</strong> blood glucose. In contrast, studies have failed to showany short-term effect of alcohol consumed with a meal (Kerr et al., 1990; Avogaro et al.,1993), or when given <strong>in</strong>travenously after an overnight fast (Kolaczynski et al., 1988). This isprobably because the suppression of gluconeogenesis by ethanol has little effect on hepaticglucose output <strong>in</strong> well-fed subjects (G<strong>in</strong> et al., 1992) and may <strong>in</strong> fact be counterbalancedby a reduction <strong>in</strong> peripheral glucose uptake (Koivisto et al., 1993). In well-nourished, nondiabeticsubjects, very little evidence is available to suggest that alcohol has any significanteffect on glucose homeostasis (Trojan et al., 1999).However, alcohol has been shown to suppress lipolysis acutely (Avogaro et al., 1993).Follow<strong>in</strong>g <strong>in</strong>gestion of alcohol, a reduction <strong>in</strong> plasma levels of free fatty acids is associatedwith a reduction <strong>in</strong> gluconeogenesis and an <strong>in</strong>creased risk of hypoglycaemia <strong>in</strong> type1 diabetes (Avogaro et al., 1993). In other words, any potential effect of alcohol uponprevail<strong>in</strong>g glucose is likely to be maximal at a time when glucose homeostasis is dependentupon free fatty acid production, e.g. overnight, when lipolysis <strong>in</strong>creases to promote gluconeogenesis(Hagstrom-Toft et al., 1997). The alcohol-<strong>in</strong>duced suppression of lipolysis maythen predispose to hypoglycaemia the next morn<strong>in</strong>g (Figures 5.1 and 5.2). The predispositionto delayed hypoglycaemia <strong>in</strong> type 1 diabetes may be augmented by relative hyper<strong>in</strong>sul<strong>in</strong>aemia,which <strong>in</strong> turn further suppresses lipolysis. However, it is likely that other factorsare <strong>in</strong>volved, as <strong>in</strong> well-fed subjects the suppression of gluconeogenesis by alcohol per semay have little effect on hepatic glucose output (G<strong>in</strong> et al., 1992), and is considered to becounterbalanced by a reduction <strong>in</strong> peripheral glucose uptake (Koivisto et al., 1993).In a cl<strong>in</strong>ical context the ma<strong>in</strong> concern for <strong>in</strong>sul<strong>in</strong>-treated <strong>in</strong>dividuals is that moderatealcohol <strong>in</strong>take (6–9 units) can acutely dim<strong>in</strong>ish hypoglycaemia awareness (Moriarty et al.,1993) and impair counterregulatory responses to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia (Yki-Jarv<strong>in</strong>en and Nikkila, 1985; Pukaka<strong>in</strong>en et al., 1991). Glucagon release has been shown tobe suppressed by alcohol <strong>in</strong> some studies (Rasmussen et al., 2001), but not <strong>in</strong> others (Kerret al., 1990). However, the prolonged hypoglycaemic effect of alcohol follow<strong>in</strong>g its <strong>in</strong>gestionis more likely to implicate either cortisol or growth hormone responses to hypoglycaemia(Figure 5.3). In animal studies, alcohol has been shown to stimulate corticotroph<strong>in</strong>-releas<strong>in</strong>ghormone and thus <strong>in</strong>crease plasma cortisol (Rivier et al., 1984). A rise <strong>in</strong> circulat<strong>in</strong>g glucocorticoidcan <strong>in</strong>hibit growth hormone release (Wehrenberg et al., 1990), which <strong>in</strong> turn may

LIFESTYLE MODERATORS 10520 22 24 2 4 6 8 10 12 14TimeAdrenal<strong>in</strong>e FFA Growth Hormone GlucoseFigure 5.1 Pictoral show<strong>in</strong>g steady nocturnal and next-day glucose concentrations, a normal rise <strong>in</strong>growth hormone and free fatty acids overnight, and a normal counterregulatory response should therebe a predisposition to hypoglycaemiaAlcoholPotential forhypoglycaemia20 22 24 2 4 6 8 10 12 14TimeAdrenal<strong>in</strong>e FFA GH GlucoseFigure 5.2 Pictoral <strong>in</strong>dicat<strong>in</strong>g a slow decl<strong>in</strong>e <strong>in</strong> plasma glucose follow<strong>in</strong>g even<strong>in</strong>g alcohol <strong>in</strong>gestion,suppression of free fatty acids and overnight growth hormone release, and the <strong>in</strong>ability to counterregulatean <strong>in</strong>creased predisposition to hypoglycaemia and impaired awareness of hypoglycaemiabe further reduced through the direct <strong>in</strong>hibition of growth hormone release through the acute<strong>in</strong>gestion of alcohol (Conway and Mauceri, 1991).In type 1 diabetes, the <strong>in</strong>fluence of alcohol <strong>in</strong>gestion on the immediate counterregulatoryresponses has been <strong>in</strong>vestigated by a hyper<strong>in</strong>sul<strong>in</strong>aemic clamp study (Kerr et al.,2007). Ingestion of modest amounts of alcohol (to plasma levels of less than 50 mg/dl)

106 MODERATORS, MONITORING AND MANAGEMENT5004003002001000Free <strong>in</strong>sul<strong>in</strong> (pmol/l)W<strong>in</strong>eWater20 22 24 02 04 06 08 10 12Time (24 h)8007006005004003002001000W<strong>in</strong>eWaterCortisol (nmol/l)20 22 24 02 04 06 08 10 12Time (24 h) hormone (µg/l)Time (24 h)W<strong>in</strong>eWater20 22 24 02 04 06 08 10 129080706050403020100Glucagon (ng/l)Time (24 h)W<strong>in</strong>eWater20 22 24 02 04 06 08 10 12Figure 5.3 Counterregulatory hormone responses follow<strong>in</strong>g earlier <strong>in</strong>gestion of alcohol <strong>in</strong>dicat<strong>in</strong>gsuppression of overnight growth hormone secretion. Reproduced from Turner et al., 2001, withpermission from The American <strong>Diabetes</strong> Associationappears to attenuate the usual growth hormone response to mild hypoglycaemia (Figure 5.4).Other <strong>in</strong>vestigators have reported that acute and susta<strong>in</strong>ed alcohol <strong>in</strong>gestion <strong>in</strong> non-diabeticsubjects can suppress growth hormone release <strong>in</strong> response to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia(Kolaczynski et al., 1988; Kerr et al., 1990), and also <strong>in</strong> <strong>in</strong>dividuals suffer<strong>in</strong>g from reactivehypoglycaemia (Avogaro et al., 1993). In association with blunt<strong>in</strong>g of the hormonal counterregulatoryresponses to hypoglycaemia follow<strong>in</strong>g alcohol <strong>in</strong>gestion, <strong>in</strong>sul<strong>in</strong> sensitivity alsoappears to be <strong>in</strong>creased (T<strong>in</strong>g and Lautt, 2006), thus perhaps directly suppress<strong>in</strong>g hepaticglucose production.Recent studies have also reported blunt<strong>in</strong>g of the ep<strong>in</strong>ephr<strong>in</strong>e response to hypoglycaemiaafter alcohol (Figure 5.5). The reduction is catecholam<strong>in</strong>e response was mirrored by areduction <strong>in</strong> hypoglycaemia awareness, which has been described previously with alcohol(Kerr et al., 1990).This blunt<strong>in</strong>g of the adreno-medullary response could result from a number of differentmechanisms:• Antecedent nocturnal hypoglycaemia could predispose to delayed hypoglycaemia(Veneman et al., 1993) (i.e., hypoglycaemia leads to more hypoglycaemia).• Sleep patterns are altered by alcohol, which may <strong>in</strong>crease the time spent <strong>in</strong> ‘deep’ non-REMsleep, thereby <strong>in</strong>creas<strong>in</strong>g the predisposition towards hypoglycaemia through a reduction<strong>in</strong> the ability to counterregulate (Jones et al., 1998) (see Chapter 4).

LIFESTYLE MODERATORS 107Growthhormone(µg / l)35Growth hormone (mean and SE)Glc = 4.5 Glc = 4.5 / 2.8 Glc = 4.530252015alcoholE + PE + AH + PH + A10500 30 60 90 120 150 180 210 240Time (m<strong>in</strong>s)Figure 5.4 Growth hormone concentrations dur<strong>in</strong>g the four study conditions (E = euglycaemia,P = placebo, H = hypoglycaemia, A = alcohol and Glc = glucose)1200Plasma adrenal<strong>in</strong>e dur<strong>in</strong>g hyper<strong>in</strong>sul<strong>in</strong>aemic hypoglycaemic clamp1000Adrenal<strong>in</strong>e (ng/ml)8006004002000Basel<strong>in</strong>e euglycemia4.5mmol / lInitial hypoglycaemia2.5mmol / lEnd hypoglycaemia2.5mmol / lEuglycaemia4.5mmol / lAlcoholPlaceboFigure 5.5 Ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) response dur<strong>in</strong>g a hypoglycaemic clamp follow<strong>in</strong>g alcohol<strong>in</strong>gested 12 hours previously (unpublished data of authors)

108 MODERATORS, MONITORING AND MANAGEMENT• Other mechanisms, which could account for a failure to counterregulate through an <strong>in</strong>adequatecatecholam<strong>in</strong>e response, could be expla<strong>in</strong>ed by a reduction <strong>in</strong> GH-associated prim<strong>in</strong>gof the catecholam<strong>in</strong>e response.The blunted catecholam<strong>in</strong>e response to hypoglycaemia may expla<strong>in</strong> why symptom scoresare lower; this may account for the previously described, alcohol-<strong>in</strong>duced impairment ofhypoglycaemia awareness that is associated with delayed hypoglycaemia follow<strong>in</strong>g consumptionof alcohol at an earlier time (Kerr et al., 1990).The predisposition of alcohol to cause prolonged, recurrent hypoglycaemia with blunt<strong>in</strong>gof the counterregulatory response, may lead to delayed hypoglycaemia and impaired awarenessof hypoglycaemia. Knowledge of the <strong>in</strong>creased risks and dangers <strong>in</strong>volved with an<strong>in</strong>creased risk of severe hypoglycaemia is important so that patients can make appropriateadjustments to their lifestyles.Although patients often ask for guidance about alcohol and diabetes, the advice offeredcan be variable and conflict<strong>in</strong>g. For people treated with <strong>in</strong>sul<strong>in</strong>, it is often recommendedthat dietary carbohydrate should not be omitted and alcohol should be taken with, or shortlybefore, food. Patients should be advised that the risk of hypoglycaemia may extend for‘several hours’ after dr<strong>in</strong>k<strong>in</strong>g. It is an important cl<strong>in</strong>ical observation that, at blood alcohollevels that rema<strong>in</strong> with<strong>in</strong> the statutory limits for driv<strong>in</strong>g <strong>in</strong> the UK, autonomic and neuroglycopenicwarn<strong>in</strong>g symptoms of early hypoglycaemia can be impaired and the cognitivedeficits that are usually associated with mild hypoglycaemia are augmented (Cheyne et al.,2004). As is recommended for all non-diabetic drivers, the advice should be to consume noalcohol if an <strong>in</strong>dividual is plann<strong>in</strong>g to drive.A further consideration is the ‘morn<strong>in</strong>g after the night before’ phenomenon <strong>in</strong> terms ofhypoglycaemia risk. In a laboratory-based study, Turner et al. (2001) reported that <strong>in</strong>gestionof alcohol with an even<strong>in</strong>g meal <strong>in</strong>creased the risk of hypoglycaemia the next morn<strong>in</strong>g(Figure 5.6). More recently, a study of people with type 1 diabetes dur<strong>in</strong>g their normal dailylives and us<strong>in</strong>g a cont<strong>in</strong>uous glucose monitor<strong>in</strong>g system (CGMS), confirmed a predisposition2016W<strong>in</strong>eWaterGlucose(mmol / l)1284018 20 22 24 02 04 06 08 10 12Time (24 h)Figure 5.6 Change <strong>in</strong> overnight plasma glucose follow<strong>in</strong>g alcohol or placebo (Turner et al., 2001).The period of dr<strong>in</strong>k<strong>in</strong>g is <strong>in</strong>dicated by the shaded bar and the times of symptomatic hypoglycaemia are<strong>in</strong>dicated by the vertical arrows. Repr<strong>in</strong>ted with permission from The American <strong>Diabetes</strong> Association

LIFESTYLE MODERATORS 109mean change <strong>in</strong> <strong>in</strong>terstitial glucose (mmol / l) 20:00 22:00 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00–1.0–2.0–3.0–4.0–5.0period of <strong>in</strong>creasedrisk ofhypoglycaemiaFigure 5.7 Mean difference <strong>in</strong> <strong>in</strong>terstitial glucose between alcohol and placebo beverages follow<strong>in</strong>gstandardised meal at 19:00–20:00to delayed hypoglycaemia (Richardson et al., 2005b). In this study alcohol was shown toreduce average <strong>in</strong>terstitial tissue glucose (Figure 5.7) and <strong>in</strong>crease the risk of hypoglycaemiaover the follow<strong>in</strong>g 24 hours, with patients report<strong>in</strong>g more than twice as many hypoglycaemicepisodes per day.In the study of Richardson et al. (2005b), the average <strong>in</strong>terstitial glucose level was 1–2 mmol/l lower with alcohol, but the rate of hypoglycaemia the next day depended on theprevail<strong>in</strong>g levels of glucose overnight. Patients who had a lower basel<strong>in</strong>e glucose at thebeg<strong>in</strong>n<strong>in</strong>g of the night or <strong>in</strong> the morn<strong>in</strong>g, were at a greater risk of develop<strong>in</strong>g alcohol-<strong>in</strong>ducedhypoglycaemia than those with preced<strong>in</strong>g hyperglycaemia. This <strong>in</strong>creased risk persisted fornearly a full day after alcohol <strong>in</strong>gestion (Richardson et al., 2005b).Caffe<strong>in</strong>eThe consumption of caffe<strong>in</strong>e has occurred for over 8000 years. Its impact on health –both good and bad – has been widely reported and disputed. Coffee (and caffe<strong>in</strong>e) is themost widely used stimulant <strong>in</strong> the world and is consumed by more than 50% of Britonsregularly. This amounts to about 75 million cups of coffee consumed every day. Caffe<strong>in</strong>eis also consumed <strong>in</strong> a variety of different formats such as tea and soft dr<strong>in</strong>ks and <strong>in</strong> ‘overthe counter’ remedies for coughs and colds. Dur<strong>in</strong>g the period from 1960 to 1982, thelevel of consumption of caffe<strong>in</strong>e-conta<strong>in</strong><strong>in</strong>g products rose by 231% (Gilbert, 1984), and<strong>in</strong> the UK, the average caffe<strong>in</strong>e consumption by adults was estimated to be 400 mg daily(Gilbert, 1984). Children, who do not dr<strong>in</strong>k coffee, may consume equivalent amounts <strong>in</strong>soft dr<strong>in</strong>ks.Caffe<strong>in</strong>e exerts a variety of pharmacological actions at diverse sites, both centrallyand peripherally, pr<strong>in</strong>cipally through adenos<strong>in</strong>e receptor antagonism. Although it has beensuggested that the amount of caffe<strong>in</strong>e consumed is related to the risk of develop<strong>in</strong>g orprotect<strong>in</strong>g aga<strong>in</strong>st type 2 diabetes (Pereira et al., 2006), the discussion here is focused onthe <strong>in</strong>fluence that caffe<strong>in</strong>e has on the physiological responses to a fall <strong>in</strong> blood glucose.

110 MODERATORS, MONITORING AND MANAGEMENTM<strong>in</strong>utes of <strong>in</strong>terstitial hypoglycaemia20016012080400DayplaceboDaycaffe<strong>in</strong>eNightplaceboNightcaffe<strong>in</strong>eFigure 5.8 Diurnal variation <strong>in</strong> time spent <strong>in</strong> hypoglycaemic range (<strong>in</strong>terstitial glucose < 35 mmol/l)compar<strong>in</strong>g caffe<strong>in</strong>e and placebo. Error bars <strong>in</strong>dicate the confidence <strong>in</strong>tervals of the meansPrevious studies have shown that <strong>in</strong>gestion of moderate amounts of caffe<strong>in</strong>e may be usefulby augment<strong>in</strong>g the symptomatic and hormonal responses to mild hypoglycaemia. Caffe<strong>in</strong>e(3–4 cups of drip-brewed coffee each day) enhances the symptomatic and sympathoadrenalresponses to hypoglycaemia <strong>in</strong> healthy non-diabetic volunteers and <strong>in</strong> patients with type 1diabetes (Kerr et al., 1993; Debrah et al., 1996). This may enhance the ability of <strong>in</strong>dividualsto perceive the onset of symptoms and take appropriate action by <strong>in</strong>gest<strong>in</strong>g carbohydratebefore neuroglycopenia develops. The beneficial effect of caffe<strong>in</strong>e on hypoglycaemia risk is<strong>in</strong>dependent of a change <strong>in</strong> glycaemic control (Watson et al., 2000). Putative mechanismshave <strong>in</strong>cluded an <strong>in</strong>crease <strong>in</strong> counterregulatory hormones, <strong>in</strong>clud<strong>in</strong>g ep<strong>in</strong>ephr<strong>in</strong>e, growthhormone and cortisol, but not norep<strong>in</strong>ephr<strong>in</strong>e (Debrah et al., 1996).More recently the <strong>in</strong>fluence of caffe<strong>in</strong>e on frequency of hypoglycaemia <strong>in</strong> patients withlong-stand<strong>in</strong>g type 1 diabetes has been studied us<strong>in</strong>g CGMS. By us<strong>in</strong>g cont<strong>in</strong>uous monitor<strong>in</strong>g,caffe<strong>in</strong>e appears to reduce the duration of nocturnal hypoglycaemia <strong>in</strong> subjects with type 1diabetes by almost 50% (Richardson et al., 2005a) (Figure 5.8). It has been suggested thata beneficial effect of the caffe<strong>in</strong>e-associated reduction <strong>in</strong> nocturnal hypoglycaemia may beto reduce the risk of develop<strong>in</strong>g impaired awareness of hypoglycaemia on the next day. Thecaffe<strong>in</strong>e-associated reduction <strong>in</strong> ‘antecedent’ nocturnal hypoglycaemia seen <strong>in</strong> these studiesmay expla<strong>in</strong> the augmentation <strong>in</strong> the symptomatic and hormonal responses to mild daytimehypoglycaemia described previously (Debrah et al., 1996).The relationship between autonomic dysfunction and the development of impaired awarenessof hypoglycaemia was unclear for many years (see Chapter 7). Most studies havesuggested that peripheral autonomic neuropathy is not associated with an <strong>in</strong>creased riskof severe hypoglycaemia (Pol<strong>in</strong>sky et al., 1980; Bjork et al., 1990; Ryder et al., 1990;The DCCT Research Group, 1991) although central autonomic dysfunction may be important(Evans et al., 2003). Although caffe<strong>in</strong>e improves parasympathetic autonomic function(Richardson et al., 2004), no correlation was found with the observed reduction <strong>in</strong> nocturnalhypoglycaemia associated with caffe<strong>in</strong>e. As mentioned earlier, it is possible that caffe<strong>in</strong>euncouples cerebral blood flow and glucose utilisation via antagonism of adenos<strong>in</strong>e receptors(Laurienti et al., 2003), attenuat<strong>in</strong>g the glucose supply to the bra<strong>in</strong> (reduced cerebral blood

MONITORING 111flow) while simultaneously <strong>in</strong>creas<strong>in</strong>g glucose demand, thus result<strong>in</strong>g <strong>in</strong> relative neuroglycopeniaand earlier release of counterregulatory hormones.Caffe<strong>in</strong>e may also act through an alteration <strong>in</strong> sleep pattern as adenos<strong>in</strong>e has been implicated<strong>in</strong> the physiological regulation of sleep. Caffe<strong>in</strong>e reduces non-rapid eye movement(REM) sleep (Landolt et al., 1995), a stage of sleep that is associated with attenuatedcounterregulatory responses to hypoglycaemia (Jones et al., 1998). Therefore, it couldbe hypothesised that caffe<strong>in</strong>e reduces the time spent <strong>in</strong> non-REM sleep, lessen<strong>in</strong>g theperiod dur<strong>in</strong>g which counterregulatory responses are suppressed. This may protect aga<strong>in</strong>stprolonged hypoglycaemia and may expla<strong>in</strong> the f<strong>in</strong>d<strong>in</strong>gs of fewer and shorter moderateepisodes of nocturnal hypoglycaemia <strong>in</strong> people us<strong>in</strong>g caffe<strong>in</strong>e. The suggested beneficialeffects of caffe<strong>in</strong>e on nocturnal hypoglycaemia would support the notion that caffe<strong>in</strong>euse should be encouraged <strong>in</strong> a population that is prone to the risks of severe nocturnalhypoglycaemia.MONITORINGIt is a s<strong>in</strong>e qua non that the best defence aga<strong>in</strong>st hypoglycaemia is the ability to recogniseit at an early stage and take appropriate action. Nevertheless, some people lose their abilityto detect hypoglycaemia:• Individuals may fail to develop appropriate warn<strong>in</strong>g symptoms.• Individuals may fail to recognise the warn<strong>in</strong>g symptoms as be<strong>in</strong>g related to hypoglycaemia.• Individuals may recognise the warn<strong>in</strong>g symptoms but may be unable to take appropriateaction because of neuroglycopenia.Memory impairment is commonly associated with hypoglycaemia (see Chapter 2) andthus patient recall may underestimate the true frequency of the problem. Furthermore thedef<strong>in</strong>ition of hypoglycaemia needs to be agreed by the patient, their relatives and their healthprofessionals especially for more modest events, which are often ‘accepted’ by patients asan <strong>in</strong>evitable part of <strong>in</strong>sul<strong>in</strong> treatment.As a consequence, it is important that patients have additional methods of detect<strong>in</strong>g lowblood glucose levels. Invariably this <strong>in</strong>volves f<strong>in</strong>ger stick measurements of blood glucoselevels us<strong>in</strong>g glucose meters. However, this method has problems per se related to:• poor technical performance, e.g. <strong>in</strong>adequate samples;• contam<strong>in</strong>ation (actually rare <strong>in</strong> rout<strong>in</strong>e practice);• technical limitations of the devices (Melki et al., 2006).Recently novel methods for measur<strong>in</strong>g glucose levels have been <strong>in</strong>troduced although atpresent none fulfils the ‘holy grail’ of non-<strong>in</strong>vasive glucose monitor<strong>in</strong>g. It is noteworthythat the newer methods of measur<strong>in</strong>g <strong>in</strong>terstitial glucose have highlighted the fact thathypoglycaemia rema<strong>in</strong>s a common problem <strong>in</strong> type 1 diabetes management and that manyepisodes rema<strong>in</strong> unrecognised (Cheyne and Kerr, 2002). These methods can, however, be auseful aid, allow<strong>in</strong>g patients (and their healthcare professionals) to determ<strong>in</strong>e the modulat<strong>in</strong>g

112 MODERATORS, MONITORING AND MANAGEMENT25.0Sensor Profile – before pump therapy25.0Sensor Profile – after pump therapySensor value (mmol/l) 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00Timesensor value20. 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00TimeFigure 5.9 CGMS profile before and after conversion to CSII <strong>in</strong> a patient with > 15 years of poorlycontrolled type 1 diabetesGlucose Concentration (mmol/L) 4:00 8:00 12:00 16:00 20:00 24:00TimeFigure 5.10 CGMS profile <strong>in</strong>dicat<strong>in</strong>g the extent of postprandial hyperglycaemia. Capillary bloodglucose measurements are shown <strong>in</strong> square boxes<strong>in</strong>fluence of a number of factors on glycaemic control and <strong>in</strong>sul<strong>in</strong> use (Figures 5.9 and 5.10).Unfortunately they have a number of limitations, as detailed <strong>in</strong> the next section.Cont<strong>in</strong>uous Glucose Monitor<strong>in</strong>g Systems (CGMS)There are limitations to the use of traditional blood glucose measurements for detect<strong>in</strong>glow blood glucose levels. Consequently, a great deal of time, effort and money has beenspent on develop<strong>in</strong>g new methods for the detection of low blood glucose levels. The mostpopular system at present is the use of a cont<strong>in</strong>uous glucose monitor<strong>in</strong>g system (CGMS)(Hoi-Hansen et al., 2005). However, technical considerations <strong>in</strong>fluence the use of <strong>in</strong>terstitialglucose monitor<strong>in</strong>g for the detection of hypoglycaemia, specifically:• There is a physiological lag between changes <strong>in</strong> blood and <strong>in</strong>terstitial glucose levels whichsuggest that CGMS may overestimate the duration of a hypoglycaemic event.• It is unclear as to the significance of changes at the level of <strong>in</strong>terstitial fluid compared tofluctuations <strong>in</strong> blood glucose level.

MONITORING 113• It is unclear whether changes <strong>in</strong> <strong>in</strong>terstitial glucose levels, measured <strong>in</strong> the anteriorabdom<strong>in</strong>al wall, mirror those occurr<strong>in</strong>g at the level of glucose sens<strong>in</strong>g neurones <strong>in</strong> thehypothalamus and elsewhere.• The record<strong>in</strong>gs from the CGMS are markedly <strong>in</strong>fluenced by the accuracy and frequencyof the results of f<strong>in</strong>gerstick blood test<strong>in</strong>g for calibration purposes.The first commercially available device, the CGMS (Medtronic M<strong>in</strong>imed, M<strong>in</strong>neapolis,USA), was designed to monitor glucose levels cont<strong>in</strong>uously <strong>in</strong> tissue fluid. The systemcomprised a disposable subcutaneous glucose sensor connected by a cable to a pager-sizedglucose monitor. The sensor measured <strong>in</strong>terstitial glucose levels every 10 seconds andaveraged over 5 m<strong>in</strong>utes, i.e., 288 measurements each day, and provided measurementsof glucose <strong>in</strong> the range of 2.2–22 mmol/l. The first version provided retrospective data.The latest generation sensor transfers data to an on-screen display <strong>in</strong> real-time for thepatient to adjust his or her own medication (Bode et al., 2004). One problem has been <strong>in</strong>def<strong>in</strong><strong>in</strong>g hypoglycaemia us<strong>in</strong>g these devices, although some groups have def<strong>in</strong>ed <strong>in</strong>terstitialhypoglycaemia accord<strong>in</strong>g to the level at which there is activation of the counterregulatoryhormone cascade and onset of neuroglycopenic symptoms respectively (Fanelli et al., 1994)(Table 5.2).There are also other limitations (<strong>in</strong>clud<strong>in</strong>g expense) <strong>in</strong> the use of cont<strong>in</strong>uous <strong>in</strong>terstitialmonitor<strong>in</strong>g devices. Cont<strong>in</strong>uous glucose monitor<strong>in</strong>g is a technique <strong>in</strong> its <strong>in</strong>fancy and as suchthere is debate whether it has the ability to detect ‘true’ hypoglycaemia (McGowan et al.,2002). This uncerta<strong>in</strong>ty relates to a number of factors <strong>in</strong>clud<strong>in</strong>g:Table 5.2 Guidel<strong>in</strong>es for <strong>in</strong>terpret<strong>in</strong>g <strong>in</strong>terstitial hypoglycaemia us<strong>in</strong>g CGMS as developedby the UK <strong>Hypoglycaemia</strong> Study Group (2007), and published <strong>in</strong> Department forTransport Road Safety Research Report No. 61, “Stratify<strong>in</strong>g hypoglycemic event risk <strong>in</strong><strong>in</strong>sul<strong>in</strong>-treated diabetes” (2006), pp. 68–9Def<strong>in</strong><strong>in</strong>g an episode of hypoglycaemiaThere must be four consecutive read<strong>in</strong>gs of 3.5 mmol/l or lower for an episode to beclassified as hypoglycaemia.The first of the read<strong>in</strong>gs (of 3.5 mmol/l or less) signifies the start of hypoglycaemia.The first read<strong>in</strong>g of 3.5 mmol/l or above signifies the end of hypoglycaemia.<strong>Hypoglycaemia</strong> ends fully when there are four or more consecutive read<strong>in</strong>gs above3.5 mmol/l.If dur<strong>in</strong>g hypoglycaemia the sensor value is 3.5 mmol/l or higher for 1–3 read<strong>in</strong>gs andthen goes back down below 3.5 mmol/l, then the whole episode is counted as onehypoglycaemic event and the short period above 3.5 mmol/l is <strong>in</strong>cluded <strong>in</strong> theduration of hypoglycaemia.To be labelled as moderate hypoglycaemia, the sensor has to read 3.0 mmol/l orbelow for at least four consecutive read<strong>in</strong>gs.If dur<strong>in</strong>g mild hypoglycaemia the read<strong>in</strong>g falls to 3.0 mmol/l or below for four ormore consecutive read<strong>in</strong>gs, the whole hypoglycaemic episode will be considered asmoderate hypoglycaemia.Prolonged hypoglycaemia is def<strong>in</strong>ed as last<strong>in</strong>g > 2 hours.

114 MODERATORS, MONITORING AND MANAGEMENT6.00Sensor value compared to blood glucoseGlucose (mmol/l) averageaverage BG0.000:00 1:00 2:00 3:00 4:00Time (hours)Figure 5.11 Comparison between <strong>in</strong>terstitial (CGMS) and plasma glucose under hyper<strong>in</strong>sul<strong>in</strong>aemichypoglycaemic clamp conditions• the potential error <strong>in</strong> record<strong>in</strong>g glucose values at the limit of the detection range;• the potential for artefact (a flat l<strong>in</strong>e might <strong>in</strong>dicate a low glucose value or a technicalproblem with the record<strong>in</strong>g);• differences between blood and extracellular <strong>in</strong>terstitial glucose.Cont<strong>in</strong>uous glucose monitor<strong>in</strong>g methodology does not directly sample blood glucosebut records glucose values <strong>in</strong> the extracellular <strong>in</strong>terstitial space. Glucose diffuses acrossthe capillary wall <strong>in</strong>to the <strong>in</strong>terstitial space before be<strong>in</strong>g transported <strong>in</strong>to cells where it ismetabolised or stored. The relationship between blood and <strong>in</strong>terstitial glucose is not wellunderstood and there are physiological and pharmacological reasons why the two may differ.Laboratory studies have demonstrated a variable relationship between blood and <strong>in</strong>terstitialglucose concentrations accord<strong>in</strong>g to whether blood glucose is fall<strong>in</strong>g or ris<strong>in</strong>g, and circulat<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> concentrations may also be important (Capl<strong>in</strong> et al., 2003).F<strong>in</strong>ally, it is important to note that the use of CGMS is best used to corroborate themore traditional method of record<strong>in</strong>g hypoglycaemia events, which is rely<strong>in</strong>g on patients’self-reports. Even tak<strong>in</strong>g <strong>in</strong>to account the uncerta<strong>in</strong>ty of which CGMS value representstrue hypoglycaemia, rates of low glucose are comparable whether measured by prospectivecollection of self-reported episodes or by cont<strong>in</strong>uous glucose monitor<strong>in</strong>g.Although there is <strong>in</strong>creas<strong>in</strong>g experience with the CGMS system for detection of hypoglycaemia,the relationship between <strong>in</strong>terstitial glucose levels measured from the anterior abdom<strong>in</strong>alwall and cerebral <strong>in</strong>terstitial levels is unknown. Furthermore, <strong>in</strong> non-diabetic <strong>in</strong>dividuals,the CGMS may overestimate the duration of hypoglycaemia as there appears to be a timelag between sensor-measured <strong>in</strong>terstitial tissue glucose and peripheral blood glucose levelsdur<strong>in</strong>g recovery from hypoglycaemia (Cheyne et al., 2002) (Figure 5.11). Overestimation ofnocturnal hypoglycaemia <strong>in</strong> patients with strictly controlled type 1 diabetes has also beenreported (McGowan et al., 2002). However, <strong>in</strong> more customary populations where capillaryglucose sensor calibration was more widely dispersed, accurate prediction of hypoglycaemiawith the CGMS has been demonstrated (McGowan et al., 2002; Capl<strong>in</strong> et al., 2003).

MANAGEMENT OF HYPOGLYCAEMIA 115MANAGEMENT OF HYPOGLYCAEMIAIt goes without say<strong>in</strong>g that prevention is better than cure and patients with diabetes needto be thoroughly and cont<strong>in</strong>uously educated about the potential risks of hypoglycaemia.<strong>Diabetes</strong> UK recommends a policy of 4 is the floor as a capillary blood glucose level for<strong>in</strong>tervention with regard to treat<strong>in</strong>g hypoglycaemia. The treatment of hypoglycaemia is bestregarded as a spectrum of <strong>in</strong>creas<strong>in</strong>g <strong>in</strong>tervention with at one end the simple <strong>in</strong>gestion of oralcarbohydrate and at the other, acute medical therapy <strong>in</strong> an <strong>in</strong>tensive care unit (Table 5.3).The simplest treatment, when the patient recognises the early warn<strong>in</strong>g symptoms (seeChapter 2), is to eat carbohydrate, which must be palatable, concentrated and portable.Glucose tablets (Dextrosol) are usually recommended <strong>in</strong> the UK, barley sugar <strong>in</strong> the USAand, <strong>in</strong> France, lumps of sugar (sucrose). Beverages such as soft dr<strong>in</strong>ks or orange juicewith a high glucose content are also suitable. The important factor is that short-act<strong>in</strong>gcarbohydrate should be followed by some form of longer-act<strong>in</strong>g carbohydrate such as bread orbiscuits.The second level of treatment is when the patient is clearly hypoglycaemic but cannotor will not take oral fast-act<strong>in</strong>g carbohydrate. People on <strong>in</strong>sul<strong>in</strong> will often not admit tobe<strong>in</strong>g hypoglycaemic and may react adversely to attempts to give them carbohydrate. Liquidglucose solutions are often unsatisfactory because the patient can spit them out. It is betterto use a commercially-available glucose gel such as GlucoGel (Diabetic Bio-diagnostics),which can be squeezed like toothpaste <strong>in</strong>to the mouth and is absorbed through the buccalmucosa. Although some doubts have been expressed about the effectiveness of oral glucosegels, relatives often seem to prefer to try this before resort<strong>in</strong>g to <strong>in</strong>ject<strong>in</strong>g glucagon. It shouldnot be used <strong>in</strong> semi-comatose patients or those at risk of aspirat<strong>in</strong>g. Jam or honey may bejust as effective.Glucagon promotes hepatic glycogenolysis and the glycaemic response to a dose of1 mg is essentially the same whether it is <strong>in</strong>jected subcutaneously, <strong>in</strong>tramuscularly or <strong>in</strong>travenously(Muhlhauser et al., 1985a). The advantage of glucagon is that it can be givenTable 5.3 The therapeutic spectrum of hypoglycaemia; complexity of treatment depends primarilyon duration (adapted from MacCuish, 1993)Duration of hypoglycaemiaInitial Management(m<strong>in</strong>utes)By patient By family By paramedicsIn hospital A&EdepartmentOngo<strong>in</strong>gmanagement (hours)In <strong>in</strong>tensive careOral carbohydrate(> 20 g)Oral carbohydrate(liquid/solid)Glucagon 1mgim or iv25 g glucose iv Mannitol (20%,20 ml)Glucagon 1 mg iv DexamethasoneGlucose gel 25 g glucose iv Dextrose/<strong>in</strong>sul<strong>in</strong><strong>in</strong>fusionGlucagon 1 mg imOxygenAnticonvulsantsSedation

116 MODERATORS, MONITORING AND MANAGEMENTby relatives or friends after m<strong>in</strong>imal tra<strong>in</strong><strong>in</strong>g. Paramedics can also use it at the patient’shome or <strong>in</strong> an ambulance. The disadvantages are that it takes longer (approximately 10m<strong>in</strong>utes) than <strong>in</strong>travenous glucose to restore consciousness and does not work <strong>in</strong> patientswho have deficient or absent hepatic glycogen stores (alcoholics or people with cachexia).Unfortunately, even where glucagon is available, relatives or friends may not use it. Inone study (Muhlhauser et al., 1985b), 53 of 123 episodes of severe hypoglycaemia weretreated by relatives or friends with glucagon, 30 by assist<strong>in</strong>g physicians and 44 requiredhospital admission. When glucagon was available but not used, it was because those whoknew how to use it were not present (20 cases) or were too anxious to do so (24 cases).In children with diabetes, Daneman et al. (1989) found that glucagon was used <strong>in</strong> only athird of households <strong>in</strong> which it was available – presumably because relatives were eithertoo frightened or poorly educated. Another problem is the limited shelf life; Ward et al.(1990) found that nearly three-quarters of patients knew about glucagon but only 20% hada supply that was <strong>in</strong> date. Its effect is less certa<strong>in</strong> if coma has been prolonged. In a studyof 100 patients brought to a hospital outpatient department, glucagon was immediatelyeffective <strong>in</strong> only 41% of patients whose mean estimated duration of coma was 50 m<strong>in</strong>utes(MacCuish et al., 1970).With<strong>in</strong> five m<strong>in</strong>utes the traditional dose of 50 ml of a 50% glucose (dextrose) solutionraises blood glucose from below 1 to > 12 mmol/l (Collier et al., 1987). This dose isnow considered to be too large by us<strong>in</strong>g too concentrated a solution, and 20% dextrosewill suffice. The ma<strong>in</strong> problem is the difficulty of giv<strong>in</strong>g an <strong>in</strong>travenous <strong>in</strong>jection toan uncooperative patient who may be refus<strong>in</strong>g or resist<strong>in</strong>g treatment or who has to bephysically restra<strong>in</strong>ed to receive an <strong>in</strong>travenous <strong>in</strong>jection. Extravasation of the concentratedglucose solution outside the ve<strong>in</strong> causes pa<strong>in</strong>ful phlebitis and, because of its hypertonicity,even an <strong>in</strong>travascular <strong>in</strong>jection can cause phlebitis or thrombosis. It is not thereforeused much outside hospitals, and few GPs carry glass vials of dextrose for thispurpose.When a patient known to have type 1 diabetes is admitted to hospital <strong>in</strong> hypoglycaemiccoma, but fails to recover after be<strong>in</strong>g given <strong>in</strong>travenous glucose, other causes ofcoma such as excessive <strong>in</strong>gestion of alcohol, self-poison<strong>in</strong>g with opiates or other drugs,acute vascular events such as subarachnoid haemorrhage or stroke, head <strong>in</strong>jury or other<strong>in</strong>tracranial catastrophes must be excluded. Cerebral oedema is a recognised sequel ofsevere hypoglycaemia, and urgent neuroimag<strong>in</strong>g is required to establish its presence; treatment(Table 5.3) is usually <strong>in</strong> an <strong>in</strong>tensive care unit as this complication has a highmortality.The most difficult decision is to know for how long to cont<strong>in</strong>ue treatment. Patients,who have made a full recovery after be<strong>in</strong>g unconscious for several days, may (anecdotally)appear subsequently to have significant cognitive impairment or permanent bra<strong>in</strong> damage.It is beyond the scope of this chapter to discuss the <strong>in</strong>vestigation and management ofhypoglycaemia <strong>in</strong> non-diabetic <strong>in</strong>dividuals.In conclusion, hypoglycaemia cont<strong>in</strong>ues to be a common problem <strong>in</strong> the management of<strong>in</strong>dividuals with type 1 diabetes. The use of newer technologies of cont<strong>in</strong>uous glucose monitor<strong>in</strong>ghas highlighted that it is almost impossible to elim<strong>in</strong>ate hypoglycaemia completelywith present <strong>in</strong>sul<strong>in</strong> therapy, although understand<strong>in</strong>g moderat<strong>in</strong>g factors such as alcohol and<strong>in</strong>clud<strong>in</strong>g them as a component of education programmes for people with <strong>in</strong>sul<strong>in</strong>-treateddiabetes may help to alleviate some of the anxiety associated with the risk of liv<strong>in</strong>g constantlywith the threat of hypoglycaemia.

REFERENCES 117CONCLUSIONS• <strong>Hypoglycaemia</strong> is associated with multiple risk factors.• Alcohol is a frequent cause of severe hypoglycaemia.• Alcohol is associated with delayed hypoglycaemia through impaired hypoglycaemiaawareness and an abnormal counterregulatory response.• Understand<strong>in</strong>g of the nature of alcohol-associated hypoglycaemia is the first step toreduc<strong>in</strong>g the risks associated with alcohol.• Caffe<strong>in</strong>e may augment the symptomatic and hormonal responses to hypoglycaemia andreduce nocturnal hypoglycaemia <strong>in</strong> <strong>in</strong>dividuals with type 1 diabetes.• Cont<strong>in</strong>uous Glucose Monitor<strong>in</strong>g is a useful adjunct to the management of the patient withtype 1 diabetes.• Treatment of hypoglycaemia can be regarded as a spectrum of <strong>in</strong>creas<strong>in</strong>g therapeuticcomplexity depend<strong>in</strong>g on the severity of the hypoglycaemia and the cl<strong>in</strong>ical status of thepatient.REFERENCESAvogaro A, Beltramello P, Gnudi L, Maran A, Valerio A, Miola M et al. (1993). Alcohol <strong>in</strong>takeimpairs glucose counterregulation dur<strong>in</strong>g acute <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> IDDM patients.Evidence for a critical role of free fatty acids. <strong>Diabetes</strong> 42: 1626–34.Bjork E, Palmer M, Schvarcz E, Berne C (1990). Incidence of severe hypoglycaemia <strong>in</strong> an unselectedpopulation of patients with <strong>in</strong>sul<strong>in</strong>-treated diabetes mellitus, with special reference to autonomicneuropathy. <strong>Diabetes</strong> Nutrition and Metabolism 4: 303–9.Bode B, Gross K, Rikalo N, Schwartz S, Wahl T, Page C et al. (2004). Alarms based on real-timesensor glucose values alert patients to hypo- and hyperglycemia: the guardian cont<strong>in</strong>uous monitor<strong>in</strong>gsystem. <strong>Diabetes</strong> Technology and Therapeutics 6: 105–13.Bott S, Bott U, Berger M, Muhlhauser I (1997). Intensified <strong>in</strong>sul<strong>in</strong> therapy and the risk of severehypoglycaemia. Diabetologia 40: 926–32.Brown TM, Harvey AM (1941). Spontaneous hypoglycemia <strong>in</strong> ‘smoke’ dr<strong>in</strong>kers. Journal of theAmerican Medical Association 17: 12–5.Capl<strong>in</strong> NJ, O’Leary P, Bulsara M, Davis EA, Jones TW (2003). Subcutaneous glucose sensor valuesclosely parallel blood glucose dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia. Diabetic Medic<strong>in</strong>e 20: 238–41.Cheyne EH, Cavan DA, Kerr D (2002). Performance of a cont<strong>in</strong>uous glucose monitor<strong>in</strong>g system dur<strong>in</strong>gcontrolled hypoglycemia <strong>in</strong> healthy volunteers. <strong>Diabetes</strong> Technology and Therapeutics 4: 607–13.Cheyne E, Kerr D (2002). Mak<strong>in</strong>g ‘sense’ of diabetes: us<strong>in</strong>g a cont<strong>in</strong>uous glucose sensor <strong>in</strong> cl<strong>in</strong>icalpractice. <strong>Diabetes</strong> Metabolism Research and Reviews 18 Suppl 1: S43–8.Cheyne EH, Sherw<strong>in</strong> RS, Lunt MJ, Cavan DA, Thomas PW, Kerr D (2004). Influence of alcoholon cognitive performance dur<strong>in</strong>g mild hypoglycaemia; implications for type 1 diabetes. DiabeticMedic<strong>in</strong>e 21: 230–7.Collier A, Steedman DJ, Patrick AW, Nimmo GR, Matthews DM, MacIntyre CA et al. (1987).Comparison of <strong>in</strong>travenous glucagon and dextrose <strong>in</strong> treatment of severe hypoglycemia <strong>in</strong> anAccident and Emergency Department. <strong>Diabetes</strong> Care 10: 712–5.Conway S, Mauceri H (1991). The <strong>in</strong>fluence of acute ethanol exposure on growth hormone release <strong>in</strong>female rats. Alcohol 8: 159–64.

118 MODERATORS, MONITORING AND MANAGEMENTCox DJ, Gonder-Frederick L, Julian DM, Clarke W (1994). Long-term follow-up of blood glucoseawareness tra<strong>in</strong><strong>in</strong>g. <strong>Diabetes</strong> Care 17: 1–5.Daneman D, Frank M, Perlman K, Tamm J, Ehrlich R (1989). Severe hypoglycemia <strong>in</strong> children with<strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: frequency and predispos<strong>in</strong>g factors. Journal of Pediatrics 115:681–5.Debrah K, Sherw<strong>in</strong> RS, Murphy J, Kerr D (1996). Effect of caffe<strong>in</strong>e on recognition of and physiologicalresponses to hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 347: 19–24.Evans SB, Wilk<strong>in</strong>son CW, Gronbeck P, Bennett JL, Taborsky GJ Jr, Figlewicz DP (2003). Inactivationof the PVN dur<strong>in</strong>g hypoglycemia partially simulates hypoglycemia-associated autonomic failure.American Journal of Physiology. 284: R57–65.Fanelli C, Pampanelli S, Epifano L, Rambotti AM, Ciofetta M, Moda F et al. (1994). Relative rolesof <strong>in</strong>sul<strong>in</strong> and hypoglycaemia on <strong>in</strong>duction of neuroendocr<strong>in</strong>e responses to, symptoms of, anddeterioration of cognitive function <strong>in</strong> hypoglycaemia <strong>in</strong> male and female humans. Diabetologia 37:797–807.Feldman M, Schiller LR (1983). Disorders of gastro<strong>in</strong>test<strong>in</strong>al motility associated with diabetes mellitus.Annals of Internal Medic<strong>in</strong>e 98: 378–84.Gilbert, RM (1984). Caffe<strong>in</strong>e consumption. In: The Methylxanth<strong>in</strong>es Beverages and Foods: Chemistry,Consumption and Health Effects. Spiller GA, ed. Alan R Liss, New York: 235–301.G<strong>in</strong> H, Morlat P, Ragnaud JM, Aubert<strong>in</strong> J (1992). Short-term effect of red w<strong>in</strong>e (consumed dur<strong>in</strong>gmeals) on <strong>in</strong>sul<strong>in</strong> requirement and glucose tolerance <strong>in</strong> diabetic patients. <strong>Diabetes</strong> Care 15: 546–8.Gold AE, MacLeod KM, Frier BM (1994). Frequency of severe hypoglycemia <strong>in</strong> patients with type Idiabetes with impaired awareness of hypoglycemia. <strong>Diabetes</strong> Care 17: 697–703.Hagstrom-Toft E, Bol<strong>in</strong>der J, Ungerstedt U, Arner P (1997). A circadian rhythm <strong>in</strong> lipid mobilizationwhich is altered <strong>in</strong> IDDM. Diabetologia 40: 1070–8.Hoi-Hansen T, Pedersen-Bjergaard U, Thorste<strong>in</strong>sson B (2005). Reproducibility and reliability ofhypoglycaemic episodes recorded with cont<strong>in</strong>uous glucose monitor<strong>in</strong>g system (CGMS) <strong>in</strong> daily life.Diabetic Medic<strong>in</strong>e 7: 858–62.Jones TW, Porter P, Sherw<strong>in</strong> RS, Davis EA, O’Leary P, Frazer F et al. (1998). Decreased ep<strong>in</strong>ephr<strong>in</strong>eresponses to hypoglycemia dur<strong>in</strong>g sleep. New England Journal of Medic<strong>in</strong>e 338: 1657–62.Kerr D, Macdonald IA, Heller SR, Tattersall RB (1990). Alcohol causes hypoglycaemic unawareness <strong>in</strong>healthy volunteers and patients with Type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabetes. Diabetologia 33: 216–21.Kerr D, Sherw<strong>in</strong> RS, Pavalkis F, Fayad PB, Sikorski L, Rife F et al. (1993). Effect of caffe<strong>in</strong>e onthe recognition of and responses to hypoglycemia <strong>in</strong> humans. Annals of Internal Medic<strong>in</strong>e 119:799–804.Kerr D, Cheyne EH, Thomas P, Sherw<strong>in</strong> RS (2007). Influence of acute alcohol <strong>in</strong>gestion on thehormonal responses to modest hypoglycaemia <strong>in</strong> patients with Type 1 diabetes. Diabetic Medic<strong>in</strong>e24: 312–16.Kolaczynski JW, Ylikahri R, Harkonen M, Koivisto VA (1988). The acute effect of ethanol oncounterregulatory response and recovery from <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia. Journal of Cl<strong>in</strong>icalEndocr<strong>in</strong>ology and Metabolism 67: 384–8.Koivisto VA, Tulokas S, Toivonen M, Haapa E, Pelkonen R (1993). Alcohol with a meal has noadverse effects on postprandial glucose homeostasis <strong>in</strong> diabetic patients. <strong>Diabetes</strong> Care 16: 1612–4.Landolt HP, Werth E, Borbely AA, Dijk DJ (1995). Caffe<strong>in</strong>e <strong>in</strong>take (200 mg) <strong>in</strong> the morn<strong>in</strong>g affectshuman sleep and EEG power spectra at night. Bra<strong>in</strong> Research 675: 67–74.Laurienti PJ, Field AS, Burdette JH, Maldjian JA, Yen YF, Moody DM (2003). Relationship betweencaffe<strong>in</strong>e-<strong>in</strong>duced changes <strong>in</strong> rest<strong>in</strong>g cerebral perfusion and blood oxygenation level-dependent signal.American Journal of Neuroradiology 24: 1607–11.MacCuish AC, Munro JF, Duncan LJP (1970). Treatment of hypoglycaemic coma with glucagon,<strong>in</strong>travenous dextrose, and mannitol <strong>in</strong>fusion <strong>in</strong> a hundred diabetics. Lancet ii: 946–9.MacCuish AC (1993). Treatment of hypoglycaemia. In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>: Cl<strong>in</strong>ical andPhysiological Aspects. Frier BM and Fisher M, eds. Edward Arnold, London: 212–21.

REFERENCES 119MacDonald MJ (1987). Postexercise late-onset hypoglycemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetic patients.<strong>Diabetes</strong> Care 10: 584–8.McGowan K, Thomas W, Moran A (2002). Spurious report<strong>in</strong>g of nocturnal hypoglycemia by CGMS<strong>in</strong> patients with tightly controlled type 1 diabetes. <strong>Diabetes</strong> Care 25: 1499–503.Melki V, Ayon F, Fernandez M, Hanaire-Brout<strong>in</strong> H (2006). Value and limitations of the Cont<strong>in</strong>uousGlucose Monitor<strong>in</strong>g System <strong>in</strong> the management of type 1 diabetes. <strong>Diabetes</strong> and Metabolism 32:123–9.Moriarty KT, Maggs DG, Macdonald IA, Tattersall RB (1993). Does ethanol cause hypoglycaemia <strong>in</strong>overnight fasted patients with type 1 diabetes? Diabetic Medic<strong>in</strong>e 10: 61–5.Muhlhauser I, Koch J, Berger M (1985a). Pharmacok<strong>in</strong>etics and bioavailability of <strong>in</strong>jected glucagon:differences between <strong>in</strong>tramuscular, subcutaneous, and <strong>in</strong>travenous adm<strong>in</strong>istration. <strong>Diabetes</strong> Care 8:39–42.Muhlhauser I, Berger M, Sonnenberg G, Koch J, Jorgens V, Schernthaner G et al. (1985b). Incidenceand management of severe hypoglycemia <strong>in</strong> 434 adults with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus.<strong>Diabetes</strong> Care 8: 268–73.Muhlhauser I, Toth G, Sawicki PT, Berger M (1991). Severe hypoglycemia <strong>in</strong> type 1 diabetic patientswith impaired kidney function. <strong>Diabetes</strong> Care 14: 344–46.Nilsson A, Tideholm B, Kalen J, Katzman P (1988). Incidence of severe hypoglycaemia and its causes<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetics. Acta Medica Scand<strong>in</strong>avica 224: 257–2.Pereira MA, Parker ED, Folsom AR (2006). Coffee consumption and risk of type 2 diabetes mellitus:an 11-year prospective study of 28 812 postmenopausal women. Archives of Internal Medic<strong>in</strong>e 26:1311–6.Pol<strong>in</strong>sky RJ, Kop<strong>in</strong> IJ, Ebert MH, Weise V (1980). The adrenal medullary response to hypoglycemia<strong>in</strong> patients with orthostatic hypotension. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 51:1401–6.Potter J, Clarke P, Gale EA, Dave SH, Tattersall RB (1982). Insul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong>an Accident and Emergency Department: the tip of an iceberg? British Medical Journal 285:1180–2.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Bendtson I, B<strong>in</strong>der C (1990). The relationship between symptomaticand biochemical hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetic patients. Journal of Internal Medic<strong>in</strong>e228: 641–6.Pukaka<strong>in</strong>en I, Koivisto VA, Yki-Jarv<strong>in</strong>en H (1991). No reduction <strong>in</strong> total hepatic glucose output by<strong>in</strong>hibition of gluconeogenesis with ethanol <strong>in</strong> NIDDM patients. <strong>Diabetes</strong> 40: 1319–27.Rasmussen BM, Orskov L, Schmitz O, Hermansen K (2001). Alcohol and glucose counterregulationdur<strong>in</strong>g acute <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> type 2 diabetic subjects. Metabolism 50:451–7.Richardson T, Rozkovec A, Thomas P, Ryder J, Meckes C, Kerr D (2004). Influence of caffe<strong>in</strong>e onheart rate variability <strong>in</strong> patients with long-stand<strong>in</strong>g type 1 diabetes. <strong>Diabetes</strong> Care 27: 1127–31.Richardson T, Thomas P, Ryder J, Kerr D (2005a). Influence of caffe<strong>in</strong>e on frequency of hypoglycemiadetected by cont<strong>in</strong>uous <strong>in</strong>terstitial glucose monitor<strong>in</strong>g system <strong>in</strong> patients with long-stand<strong>in</strong>g type 1diabetes. <strong>Diabetes</strong> Care 28: 1316–20.Richardson T, Weiss M, Thomas P, Kerr D (2005b). Day after the night before: <strong>in</strong>fluence ofeven<strong>in</strong>g alcohol on risk of hypoglycemia <strong>in</strong> patients with type 1 diabetes. <strong>Diabetes</strong> Care 28:1801–2.Rivier C, Bruhn T, Vale W (1984). Effect of ethanol on the hypothalamic-pituitary-adrenal axis <strong>in</strong>the rat: role of corticotrop<strong>in</strong>-releas<strong>in</strong>g factor (CRF). Journal of Pharmacology and ExperimentalTherapy 229: 127–31.Rosen B, Miodovnik M, Holcberg G, Khoury J, Siddiqi T (1995). <strong>Hypoglycaemia</strong>: the price of<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy for pregnant women with <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus. Obstetricsand Gynaecology 85: 417–22.

120 MODERATORS, MONITORING AND MANAGEMENTRyder RE, Owens DR, Hayes TM, Ghatei MA, Bloom SR (1990). Unawareness of hypoglycaemia and<strong>in</strong>adequate hypoglycaemic counterregulation: no causal relation with diabetic autonomic neuropathy.British Medical Journal 301: 783–7.The DCCT Research Group (1991). Epidemiology of severe hypoglycemia <strong>in</strong> the <strong>Diabetes</strong> Controland Complications Trial. American Journal of Medic<strong>in</strong>e 90: 450–9.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1995). Adverse events and theirassociation with treatment regimens <strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. <strong>Diabetes</strong> Care18: 1415–27.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1997). Hypoglycemia <strong>in</strong> the <strong>Diabetes</strong>Control and Complications Trial. The <strong>Diabetes</strong> Control and Complications Trial Research Group.<strong>Diabetes</strong> 46: 271–86.T<strong>in</strong>g JW, Lautt WW (2006). The effect of acute, chronic, and prenatal ethanol exposure on <strong>in</strong>sul<strong>in</strong>sensitivity. Pharmacology and Therapeutics 111: 346–73.Trojan N, Pavan P, Iori E, Vettore M, Marescotti MC, Macdonald IA et al. (1999). Effect of differenttimes of adm<strong>in</strong>istration of a s<strong>in</strong>gle ethanol dose on <strong>in</strong>sul<strong>in</strong> action, <strong>in</strong>sul<strong>in</strong> secretion and redox state.Diabetic Medic<strong>in</strong>e 16: 400–7.Turner BC, Jenk<strong>in</strong>s E, Kerr D, Sherw<strong>in</strong> RS, Cavan DA (2001). The effect of even<strong>in</strong>g alcohol consumptionon next-morn<strong>in</strong>g glucose control <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care 24: 1888–93.UK <strong>Hypoglycaemia</strong> Group (2007). Risk of hypoglycaemia <strong>in</strong> types 1 and 2 diabetes: effects oftreatment modalities and their duration. Diabetologia 50: 1140–7.Veneman T, Mitrakou A, Mokan M, Cryer P, Gerich J (1993). Induction of hypoglycemia unawarenessby asymptomatic nocturnal hypoglycemia. <strong>Diabetes</strong> 42: 1233–7.V<strong>in</strong>ik AI, Maser RE, Mitchell BD, Freeman R (2003). Diabetic autonomic neuropathy. <strong>Diabetes</strong> Care26: 1553–79.Ward CM, Stewart AW, Cutfield RG (1990). <strong>Hypoglycaemia</strong> <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dependent diabetic patientsattend<strong>in</strong>g an outpatients’ cl<strong>in</strong>ic. New Zealand Medical Journal 25: 339–41.Watson JM, Jenk<strong>in</strong>s EJ, Hamilton P, Lunt MJ, Kerr D (2000). Influence of caffe<strong>in</strong>e on the frequencyand perception of hypoglycemia <strong>in</strong> free-liv<strong>in</strong>g patients with type 1 diabetes. <strong>Diabetes</strong> Care 23:455–9.Wehrenberg WB, Janowski BA, Pier<strong>in</strong>g AW, Culler F, Jones KL (1990). Glucocorticoids: potent<strong>in</strong>hibitors and stimulators of growth hormone secretion. Endocr<strong>in</strong>ology 126: 3200–3.Yki-Jarv<strong>in</strong>en H, Nikkila EA (1985). Ethanol decreases glucose utilisation <strong>in</strong> healthy man. Journal ofCl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 61: 941–5.

6 Counterregulatory Deficiencies<strong>in</strong> <strong>Diabetes</strong>David Kerr and Tristan RichardsonINTRODUCTIONModern <strong>in</strong>tensive education programmes for people with <strong>in</strong>sul<strong>in</strong>-treated diabetes have failedto elim<strong>in</strong>ate hypoglycaemia, and this important side-effect of <strong>in</strong>sul<strong>in</strong> treatment may add tothe psychological, as well as the physical, burden associated with this condition. Attempts toimprove blood glucose control can <strong>in</strong>crease the risk of severe hypoglycaemia. For example,with<strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial (DCCT), improvements <strong>in</strong> HbA 1c levelswith <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy were associated with a three-fold <strong>in</strong>crease <strong>in</strong> the risk of severehypoglycaemia compared to <strong>in</strong>dividuals treated conventionally (The <strong>Diabetes</strong> Control andComplications Trial Research Group,1993). Despite the development of strategies to improveglycaemic control, while simultaneously try<strong>in</strong>g to reduce the risk of recurrent severe hypoglycaemia,less than half of people with type 1 diabetes consistently achieve HbA 1c levels < 75%(Jacquem<strong>in</strong>et et al., 2005), and for some <strong>in</strong>dividuals, fear of hypoglycaemia is the ma<strong>in</strong> barrierto achiev<strong>in</strong>g optimal glycaemic control (Cox et al., 1987) (see Chapter 14).The use of cont<strong>in</strong>uous glucose monitor<strong>in</strong>g has <strong>in</strong>dicated that the frequency and duration ofhypoglycaemic events <strong>in</strong> patients with type 1 diabetes have probably been under-recognised(Figure 6.1) (Cheyne and Kerr, 2002), especially <strong>in</strong> children (We<strong>in</strong>trob et al., 2004). In theUK <strong>in</strong> recent years, a number of structured education programmes have been developed, mostof which are based on a German model (Muhlhauser et al., 1987), and have been focusedon empower<strong>in</strong>g patients to alter <strong>in</strong>sul<strong>in</strong> doses more accurately accord<strong>in</strong>g to the carbohydratecontent of meals, the level of planned exercise, work demands and so on (DAFNE StudyGroup, 2002). Disappo<strong>in</strong>t<strong>in</strong>gly, despite <strong>in</strong>creas<strong>in</strong>g numbers of patients hav<strong>in</strong>g access to theseeducational programmes, along with the <strong>in</strong>creased availability of analogue <strong>in</strong>sul<strong>in</strong>s, and withthe evidence that these measures are reduc<strong>in</strong>g the number of patients who have microvascularcomplications (Figure 6.2), the rates of severe hypoglycaemia have not altered significantly(Figure 6.3) (Bulsara et al., 2004).Why is hypoglycaemia still such an important problem for people with type 1 diabetesdespite these significant improvements <strong>in</strong> technology and the delivery of care? The answerfrequently relates to the problem of defective glucose counterregulation and the often associateddifficulty of recognis<strong>in</strong>g a fall <strong>in</strong> blood glucose (impaired awareness of hypoglycaemia –see Chapter 7) at a time when appropriate self-corrective action can be taken.<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

122 COUNTERREGULATORY DEFICIENCIES IN DIABETES20.0Glucose (mmol/L)–5.012:00AM4:00AM8:00AM12:00PM4:00PM8:00PM12:00AMFigure 6.1 Interstitial glucose values over three days measured by cont<strong>in</strong>uous glucose monitor<strong>in</strong>gand show<strong>in</strong>g persistent nocturnal hypoglycaemia50%40%Ret<strong>in</strong>opathyKidney disease30%20%10%0%25 yearsFigure 6.2 Reduction <strong>in</strong> prevalence of microvascular complications over the past 25 years fromspecialist diabetes centres. Adapted from Ross<strong>in</strong>g (2005), with k<strong>in</strong>d permission from Spr<strong>in</strong>ger Scienceand Bus<strong>in</strong>ess MediaNORMAL GLUCOSE COUNTERREGULATIONUnder normal circumstances, the bra<strong>in</strong> uses glucose as its predom<strong>in</strong>ant fuel and is almostcompletely dependent upon a cont<strong>in</strong>uous supply of glucose from the peripheral circulationto ma<strong>in</strong>ta<strong>in</strong> normal function. Consequently, to protect the delivery of glucose to the bra<strong>in</strong>,a hierarchy of (counterregulatory) responses (Figure 6.4) are activated as peripheral bloodglucose falls below normal (Mitrakou et al., 1991). The ma<strong>in</strong> components of this systemof normal (i.e. non-diabetic) glucose counterregulation, which prevents or quickly correctshypoglycaemia, are as follows:• A reduction <strong>in</strong> pancreatic -cell <strong>in</strong>sul<strong>in</strong> secretion.• An <strong>in</strong>crease <strong>in</strong> pancreatic -cell glucagon secretion. If hypoglycaemia is prolonged anumber of other hormones are also released. These <strong>in</strong>clude ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e),growth hormone and cortisol.

NORMAL GLUCOSE COUNTERREGULATION 123A22Rate / 100 Patient years201816141210864Incidence rate21992 1994 1996Calender Year1998 20002002Figure 6.3 Rates of severe hypoglycaemia <strong>in</strong> 1335 children between 1992 and 2002. <strong>Hypoglycaemia</strong>has rema<strong>in</strong>ed problematical despite a fall <strong>in</strong> HbA 1c of 0.2% per year and <strong>in</strong>creased use of multiple daily<strong>in</strong>jections, <strong>in</strong>sul<strong>in</strong> analogues and pump therapy. Adapted from Bulsara et al. (2004) and reproducedcourtesy of The American <strong>Diabetes</strong> Association• Activation of the autonomic nervous system with the development of characteristicwarn<strong>in</strong>g symptoms.• Dur<strong>in</strong>g more profound hypoglycaemia (blood glucose < 20 mmol/l), glucose delivery tothe bra<strong>in</strong> is enhanced as a consequence of <strong>in</strong>creas<strong>in</strong>g cerebral blood flow (Thomas et al.,1997).• Hepatic autoregulation, whereby the liver is able to respond to hypoglycaemia by directly<strong>in</strong>creas<strong>in</strong>g glucose production <strong>in</strong> the absence of detectable hormonal stimulation. Hepaticautoregulation only appears to have an <strong>in</strong>fluential role <strong>in</strong> glucose counterregulation dur<strong>in</strong>gprolonged and severe hypoglycaemia (Tappy et al., 1999).The pr<strong>in</strong>cipal counterregulatory hormones, glucagon and ep<strong>in</strong>ephr<strong>in</strong>e, directly <strong>in</strong>crease theproduction of glucose by the liver as a consequence of the breakdown of hepatic glycogenstores (glycogenolysis) and the manufacture of glucose by gluconeogenesis. Ep<strong>in</strong>ephr<strong>in</strong>ealso promotes muscle glycogenolysis, proteolysis and lipolysis to provide substrates (lactate,alan<strong>in</strong>e and glycerol) for further gluconeogenesis (Figure 6.5).As blood glucose falls below normal, these hormonal responses are not secreted on an ‘allor noth<strong>in</strong>g’ basis. Individual hormones have specific blood glucose thresholds at which levelsbeg<strong>in</strong> to rise above their basel<strong>in</strong>e levels (Figure 6.6). For example, the glycaemic thresholdsfor the release of glucagon and ep<strong>in</strong>ephr<strong>in</strong>e are well above the thresholds for the generationof warn<strong>in</strong>g symptoms and impairment of higher bra<strong>in</strong> (cognitive) function (Mitrakou et al.,1991). These thresholds are not fixed but can be altered upwards or downwards accord<strong>in</strong>g to

124 COUNTERREGULATORY DEFICIENCIES IN DIABETESHypothalamusPituitary glandGrowth hormoneAutonomic Nervous SystemACTHPancreasAdrenalglandCortisolGlucagonNorep<strong>in</strong>ephr<strong>in</strong>eEp<strong>in</strong>ephr<strong>in</strong>eFigure 6.4The bra<strong>in</strong> is the key regulatory organ <strong>in</strong>volved <strong>in</strong> glucose counterregulationthe prevail<strong>in</strong>g quality of glycaemic control. They do not appear, however, to be <strong>in</strong>fluencedby the rate of fall of blood glucose with<strong>in</strong> the hyper- or euglycaemic range.The key organ <strong>in</strong> coord<strong>in</strong>at<strong>in</strong>g the hormonal and other responses to hypoglycaemia isthe bra<strong>in</strong>. Although numerous neural areas have been proposed as the control centres forcounterregulation, it is likely that neurones located with<strong>in</strong> the ventro-medial nuclei of thehypothalamus are essential for <strong>in</strong>tegrat<strong>in</strong>g the hormonal responses to a fall <strong>in</strong> peripheralblood glucose (Borg et al., 1994), possibly via ATP-sensitive K+ channels (McCrimmonet al., 2005), although other glucoreceptors outside the bra<strong>in</strong> are <strong>in</strong>volved <strong>in</strong> <strong>in</strong>itiat<strong>in</strong>g thecounterregulatory responses, most notably with<strong>in</strong> the liver (Smith et al., 2002). Althoughthe bra<strong>in</strong> was once thought to be <strong>in</strong>sensitive to <strong>in</strong>sul<strong>in</strong>, there is cl<strong>in</strong>ical evidence to suggestthat <strong>in</strong>sul<strong>in</strong> can act on the central nervous system (CNS) to <strong>in</strong>fluence the physiologicalresponses to hypoglycaemia (Kerr et al., 1991). Recently, <strong>in</strong> studies us<strong>in</strong>g bra<strong>in</strong>/neuronal<strong>in</strong>sul<strong>in</strong> receptor knockout mice (i.e., mice with absent <strong>in</strong>sul<strong>in</strong> receptor prote<strong>in</strong>s <strong>in</strong> the bra<strong>in</strong>),the <strong>in</strong>duction of hypoglycaemia was associated with an attenuated ep<strong>in</strong>ephr<strong>in</strong>e and an almostcompletely absent norep<strong>in</strong>ephr<strong>in</strong>e response, although glucagon release was unaffected whencompared to control animals. Therefore it appears that <strong>in</strong>sul<strong>in</strong> has a role <strong>in</strong> protect<strong>in</strong>g theCNS aga<strong>in</strong>st hypoglycaemia (Fisher et al., 2005).

NORMAL GLUCOSE COUNTERREGULATION 125Figure 6.5Pr<strong>in</strong>cipal metabolic effects of counterregulation <strong>in</strong> response to hypoglycaemiaGlucose (mmol / l)4.0Release of hormonesWarn<strong>in</strong>g symptoms3.0Cognitive impairmentIncreased bra<strong>in</strong> blood flow2.0Figure 6.6 Blood glucose thresholds for release of counterregulatory hormones, onset of warn<strong>in</strong>gsymptoms of hypoglycaemia and cognitive impairment as blood glucose falls below normal

126 COUNTERREGULATORY DEFICIENCIES IN DIABETESIn type 1 diabetes there may be multiple defects <strong>in</strong> normal glucose counterregulationwhich significantly <strong>in</strong>crease an <strong>in</strong>dividual’s risk of hypoglycaemia. Defects <strong>in</strong> hormonalcounterregulation can result as a consequence of the follow<strong>in</strong>g:• secretion of <strong>in</strong>adequate amounts of counterregulatory hormones;• alteration of the blood glucose threshold at which the hormones are released; i.e., a moreprofound hypoglycaemic stimulus is required before hormonal secretion <strong>in</strong>creases abovebasel<strong>in</strong>e;• dim<strong>in</strong>ished tissue sensitivity to a given plasma concentration of hormone.In type 1 diabetes, the most common scenario for <strong>in</strong>creas<strong>in</strong>g the risk of recurrent hypoglycaemia<strong>in</strong>cludes a reduced (but not absent) ep<strong>in</strong>ephr<strong>in</strong>e response to fall<strong>in</strong>g blood glucoselevels, susta<strong>in</strong>ed peripheral hyper<strong>in</strong>sul<strong>in</strong>aemia and an absent glucagon response – this constellationof defects is associated with a 25-fold <strong>in</strong>creased risk of severe hypoglycaemia <strong>in</strong>people who are on <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy. If impaired hypoglycaemia awareness is present,the risk is <strong>in</strong>creased six fold (see Chapter 7).DEFECTIVE HORMONAL GLUCOSE COUNTERREGULATIONThe most important defects <strong>in</strong> glucose counterregulation <strong>in</strong> type 1 diabetes are:• failure of circulat<strong>in</strong>g plasma <strong>in</strong>sul<strong>in</strong> levels to decl<strong>in</strong>e (i.e., as a consequence of exogenous<strong>in</strong>sul<strong>in</strong> adm<strong>in</strong>istration);• failure of glucagon secretion from pancreatic -cells;• attenuated ep<strong>in</strong>ephr<strong>in</strong>e response to hypoglycaemia.Several factors are known to <strong>in</strong>crease the risk of counterregulatory failure, <strong>in</strong>clud<strong>in</strong>g:• long duration of diabetes;• extremes of age;• improv<strong>in</strong>g glycaemic control;• sleep;• exercise;• recurrent hypoglycaemia – related to the degree rather than duration of antecedent hypoglycaemia(Davis et al., 2000a).The defects <strong>in</strong> glucose counterregulation are not ‘all or noth<strong>in</strong>g’ and are <strong>in</strong>fluenced by anumber of factors. Some of the defects may be reversible.

DEFECTIVE HORMONAL GLUCOSE COUNTERREGULATION 127GlucagonIn non-diabetic <strong>in</strong>dividuals, glucagon is secreted by pancreatic -cells with<strong>in</strong> 30 m<strong>in</strong>utes ofthe blood glucose fall<strong>in</strong>g below normal. It is released directly as a consequence of local tissueglucopenia and <strong>in</strong>directly by sympathetic neural <strong>in</strong>puts to the pancreas. It is unclear whetherlocal (pancreatic) tissue glucopenia or autonomic activation is the key process <strong>in</strong>volved <strong>in</strong>stimulat<strong>in</strong>g the release of glucagon dur<strong>in</strong>g hypoglycaemia.In people with type 1 diabetes, the glucagon secretory response to hypoglycaemia is<strong>in</strong>itially dim<strong>in</strong>ished and is subsequently lost with<strong>in</strong> a few years of the onset of diabetes(Gerich et al., 1973), although it can be released <strong>in</strong> response to other stimuli, such as exerciseor an <strong>in</strong>travenous <strong>in</strong>fusion of arg<strong>in</strong><strong>in</strong>e (Gerich and Bolli, 1993). This <strong>in</strong>dicates that the failureof the glucagon response is most probably a signall<strong>in</strong>g rather than a structural defect. Loss ofthe glucagon response to hypoglycaemia often coexists with cl<strong>in</strong>ical evidence of autonomicneuropathy but the latter is not <strong>in</strong>variably present (Bolli et al. 1983). Although recent studieshave suggested that an early sympathetic neuropathy limited to the pancreatic islets may bea potential mechanism, patients with pancreatic transplants (i.e., denervated islets) can stillproduce glucagon <strong>in</strong> response to hypoglycaemia (Diem et al., 1990). The cause of the defect<strong>in</strong> glucose counterregulation is not known but may <strong>in</strong>clude:• reduction <strong>in</strong> pancreatic -cell mass;• autonomic neuropathy;• local effect of <strong>in</strong>sul<strong>in</strong> on normal -cell function;• generalised hyper<strong>in</strong>sul<strong>in</strong>aemia;• chronic hyperglycaemia;• <strong>in</strong>creased pancreatic production of somatostat<strong>in</strong>;• accumulation of amyl<strong>in</strong> with<strong>in</strong> the islets;• local effect of <strong>in</strong>sul<strong>in</strong>-like growth factor-1.An alternative hypothesis to expla<strong>in</strong> the glucagon-counterregulatory defect suggests that<strong>in</strong> healthy <strong>in</strong>dividuals a decrease <strong>in</strong> <strong>in</strong>tra-islet <strong>in</strong>sul<strong>in</strong> <strong>in</strong> association with a decrease <strong>in</strong>-cell glucose is the usual signal for release of glucagon as peripheral blood glucose levelsfall (Samols et al., 1972). Support for this comes from the observation that <strong>in</strong>fusion ofthe -cell secretagogue, tolbutamide, prevents the glucagon response to hypoglycaemia <strong>in</strong>non-diabetic <strong>in</strong>dividuals (Banarer and Cryer, 2003). Similarly, supraphysiological levels of<strong>in</strong>sul<strong>in</strong> have been shown to impair glucagon release <strong>in</strong> response to moderate hypoglycaemia(Kerr et al., 1991). Blunt<strong>in</strong>g of the normal glucagon response to hypoglycaemia can alsobe achieved <strong>in</strong> healthy volunteers by <strong>in</strong>fusion of <strong>in</strong>sul<strong>in</strong>-like growth factor-1 (IGF-1), theputative mediator of the somatotrophic action of growth hormone, at least <strong>in</strong> non-diabetichumans (Kerr et al., 1993) (Figure 6.7). Whether IGF-1 is <strong>in</strong>volved <strong>in</strong> the pathogenesis ofglucagon counterregulatory failure <strong>in</strong> patients with type 1 diabetes is unclear.Alternatively, it is possible that chronic hyperglycaemia may directly impair pancreatic-cell function through the mechanism of glucose toxicity similar to the effect that this has on-cell function. Nevertheless, improvements <strong>in</strong> glycaemic control, follow<strong>in</strong>g <strong>in</strong>troduction of

128 COUNTERREGULATORY DEFICIENCIES IN DIABETESFigure 6.7 Infusion of <strong>in</strong>sul<strong>in</strong>-like growth factor 1 abolishes the expected rise <strong>in</strong> glucagon whenblood glucose was lowered to, and ma<strong>in</strong>ta<strong>in</strong>ed at, 2.8 mmol/l <strong>in</strong> healthy volunteers. Reproduced fromKerr et al. (1993) by permission of The Journal of Cl<strong>in</strong>ical Investigation<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy, <strong>in</strong>variably fail to restore the glucagon responses to hypoglycaemia(Amiel, 1991). The early appearance of an impaired response of glucagon, together with atemporal dissociation from deficiencies <strong>in</strong> other counterregulatory hormones, argues aga<strong>in</strong>stthis abnormality be<strong>in</strong>g mediated from with<strong>in</strong> the central nervous system.Impairment of the glucagon response to hypoglycaemia results <strong>in</strong> greater suppressionof hepatic glucose production by <strong>in</strong>sul<strong>in</strong>. S<strong>in</strong>ce the ability of <strong>in</strong>sul<strong>in</strong> to <strong>in</strong>crease glucoseutilisation is unaltered, glucose uptake by muscle and other tissues will exceed hepaticglucose production for much longer, result<strong>in</strong>g <strong>in</strong> more profound and prolonged hypoglycaemia.Whether or not the lack of a glucagon response is by itself sufficient to <strong>in</strong>crease thefrequency of severe hypoglycaemia is unclear; most patients with recurrent severe hypoglycaemiahave impairment both of glucagon and ep<strong>in</strong>ephr<strong>in</strong>e release.Catecholam<strong>in</strong>esAn attenuated ep<strong>in</strong>ephr<strong>in</strong>e response (from the adrenal medulla) to fall<strong>in</strong>g blood glucoselevels together with an absent glucagon response markedly <strong>in</strong>creases the risk of severehypoglycaemia <strong>in</strong> <strong>in</strong>dividuals with type 1 diabetes. It is unknown how common ep<strong>in</strong>ephr<strong>in</strong>ecounterregulatory failure is <strong>in</strong> type 1 diabetes, but it is related to the duration of diabetesand the prevail<strong>in</strong>g quality of glycaemic control. Estimates suggest this defect may exist <strong>in</strong>up to 45% of people with type 1 diabetes of long duration (Gerich and Bolli, 1993).The ep<strong>in</strong>ephr<strong>in</strong>e response to hypoglycaemia can be subnormal <strong>in</strong> patients with type 1diabetes who have no cl<strong>in</strong>ical evidence of autonomic neuropathy and also <strong>in</strong> <strong>in</strong>dividualswith secondary (pancreatic) diabetes. Like glucagon, the ep<strong>in</strong>ephr<strong>in</strong>e secretory deficiency isstimulus-specific to hypoglycaemia, rema<strong>in</strong><strong>in</strong>g <strong>in</strong>tact <strong>in</strong> response to exercise. The isolated

DEFECTIVE HORMONAL GLUCOSE COUNTERREGULATION 129failure of plasma ep<strong>in</strong>ephr<strong>in</strong>e concentrations to rise <strong>in</strong> response to hypoglycaemia doesnot appreciably impair glucose counterregulation – unless it is comb<strong>in</strong>ed with glucagondeficiency (as occurs <strong>in</strong> patients with long-stand<strong>in</strong>g type 1 diabetes), when the risk of severeand prolonged hypoglycaemia is significantly <strong>in</strong>creased.As mentioned above, deficient sympathoadrenal responses to hypoglycaemia are majorcomponents of the cl<strong>in</strong>ical problem of defective glucose counterregulation and impairedawareness of hypoglycaemia <strong>in</strong> type 1 diabetes (Cryer, 2004; 2005). This may occur as aresult of the follow<strong>in</strong>g:• reduced release of catecholam<strong>in</strong>es and acetyl chol<strong>in</strong>e;• possibly a reduction <strong>in</strong> tissue sensitivity to the actions of these substances, although thisis controversial.Of cl<strong>in</strong>ical importance is the observation that a hypoglycaemic episode per se, adverselyalters the subsequent sympathoadrenal responses to hypoglycaemia (Heller and Cryer, 1991).Cryer has termed this <strong>Hypoglycaemia</strong> Associated Autonomic Failure (HAAF) (Figure 6.8).In other words, a recent episode of hypoglycaemia, known as ‘antecedent hypoglycaemia’,(<strong>in</strong>clud<strong>in</strong>g an asymptomatic event) dim<strong>in</strong>ishes the sympathoadrenal, symptomatic and cognitiveresponses to subsequent hypoglycaemia. As a corollary, avoidance of hypoglycaemia(for as little as two weeks) markedly improves the responses (Cranston et al., 1994). Sympathoadrenalresponses to hypoglycaemia are also affected by recent exercise, whether the fall<strong>in</strong> blood glucose level occurs when an <strong>in</strong>dividual is awake or asleep (Jones et al., 1998;Banarer and Cryer, 2003) (Figure 6.9) and by the time of day or night (Merl et al., 2004).Insul<strong>in</strong> Deficient <strong>Diabetes</strong>(Imperfect Insul<strong>in</strong> Replacement)(No Insul<strong>in</strong>, No Glucagon)SleepAntecedent <strong>Hypoglycaemia</strong>Reduced SympathoadrenalResponses to <strong>Hypoglycaemia</strong>AntecedentExerciseReduced SympatheticNeural ResponsesReduced Ep<strong>in</strong>ephr<strong>in</strong>eResponses<strong>Hypoglycaemia</strong>UnawarenessDefective GlucoseCounterregulationRecurrent <strong>Hypoglycaemia</strong>Figure 6.8 <strong>Hypoglycaemia</strong> Associated Autonomic Failure (HAAF) <strong>in</strong> type 1 diabetes. Adapted fromCryer (2005) and reproduced courtesy of The American <strong>Diabetes</strong> Association

130 COUNTERREGULATORY DEFICIENCIES IN DIABETES(a)Nom<strong>in</strong>al glucose (mg / dl)(b)Nom<strong>in</strong>al glucose (mg / dl)8575 65 55 458575 65 55 45Ep<strong>in</strong>ephr<strong>in</strong>e (pg/ml)60050040030020010000 60 120 180 240 300Time (m<strong>in</strong>)300025002000150010005000pmol / lEp<strong>in</strong>ephr<strong>in</strong>e (pg / ml)6005004003002001000Night awakeMorn<strong>in</strong>g awake0 60 120 180 240 300Time (m<strong>in</strong>)300025002000150010005000pmol / lAsleepFigure 6.9 Plasma ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) <strong>in</strong> non-diabetic subjects (a) and <strong>in</strong> patients with type 1diabetes (b) studied <strong>in</strong> the morn<strong>in</strong>g while awake and dur<strong>in</strong>g the night while awake and asleep. Adaptedfrom Banarer and Cryer (2003) and reproduced courtesy of The American <strong>Diabetes</strong> Association(□ morn<strong>in</strong>g awake; • night awake; asleep)Cortisol and Growth HormoneGrowth hormone (GH) and cortisol are thought to become important glucose-rais<strong>in</strong>ghormones only after hypoglycaemia has been prolonged for more than one hour. However,defects <strong>in</strong> cortisol and GH release can cause profound and prolonged hypoglycaemia becauseof a reduction <strong>in</strong> hepatic glucose production and, to a lesser extent, by exaggeration of<strong>in</strong>sul<strong>in</strong>-stimulated glucose uptake by muscle.Abnormalities <strong>in</strong> growth hormone and cortisol secretion <strong>in</strong> response to hypoglycaemiaare characteristic of long-stand<strong>in</strong>g type 1 diabetes, affect<strong>in</strong>g up to a quarter of patientswho have had diabetes for more than ten years. In rare cases, coexistent endocr<strong>in</strong>e failuresuch as Addison’s disease or hypopituitarism also predisposes patients to severe hypoglycaemia.Pituitary failure, although uncommonly associated with type 1 diabetes, occasionallydevelops <strong>in</strong> young women as a consequence of ante-partum pituitary <strong>in</strong>farction. As an<strong>in</strong>tact hypothalamic–pituitary–adrenal axis is important for adequate counterregulation, thisaxis should be formally assessed <strong>in</strong> any <strong>in</strong>dividual with brittle diabetes present<strong>in</strong>g withunexpla<strong>in</strong>ed, recurrent hypoglycaemia (Hardy et al., 1994; Flanagan and Kerr, 1996).More commonly, <strong>in</strong>gestion of even modest amounts of alcohol can significantly attenuatenormal growth hormone secretion and <strong>in</strong>crease the risk of hypoglycaemia especially thefollow<strong>in</strong>g morn<strong>in</strong>g (see Chapter 5).MECHANISMS OF COUNTERREGULATORY FAILUREAt the onset of type 1 diabetes, hormonal counterregulation is usually normal but with<strong>in</strong> fiveyears of diagnosis, glucagon responses to hypoglycaemia become markedly impaired or evenabsent, although a glucagon response can occur if the hypoglycaemic stimulus is sufficientlyprofound (Frier et al. 1988; Hvidberg et al. 1998). After ten years of diabetes, patients usuallyhave a sub-optimal ep<strong>in</strong>ephr<strong>in</strong>e response to compound the absent glucagon response to a fall

MECHANISMS OF COUNTERREGULATORY FAILURE 131<strong>in</strong> blood glucose (White et al., 1985) (Figure 6.10). Thus, patients with type 1 diabetes oflong duration are at risk of severe and prolonged neuroglycopenia dur<strong>in</strong>g hypoglycaemia asa direct consequence of <strong>in</strong>adequate glucose counterregulation. Although attenuated growthhormone and cortisol responses are less common, they are late manifestations <strong>in</strong> terms ofdiabetes duration.As mentioned previously, these defects <strong>in</strong> glucose counterregulation are not ‘all or noth<strong>in</strong>g’changes but can be <strong>in</strong>fluenced by the prevail<strong>in</strong>g standard of glycaemic control and by thefrequency of hypoglycaemic episodes. Various theories relate to the cl<strong>in</strong>ical observation thatblood glucose thresholds for the release of counterregulatory hormone levels can change afterperiods of recurrent hypoglycaemia (Cryer, 2005). These may relate to changes at the levelof the CNS, which co-ord<strong>in</strong>ates the usual responses to low blood glucose levels. At presentthere is little evidence to suggest that the alterations associated with recurrent hypoglycaemiaoccur at glucose sensors outside the CNS, for example, with<strong>in</strong> the portal ve<strong>in</strong>.(a)type 1 diabetes1–5 years(b)type 1 diabetes14–31 yearsInsul<strong>in</strong>Insul<strong>in</strong>Glucagon (pg / ml) Ep<strong>in</strong>ephr<strong>in</strong>e (pg / ml) Glucose (mmol / I)54323002001000200150100Glucagon (pg / ml) Ep<strong>in</strong>ephr<strong>in</strong>e (pg / ml) Glucose (mmol / I)543230020010002001501000 30 60 90 120 150 0 30 60 90 120 150Time (m<strong>in</strong>)Time (m<strong>in</strong>)Figure 6.10 Influence of duration of diabetes on glucagon and ep<strong>in</strong>ephr<strong>in</strong>e responses to hypoglycaemia<strong>in</strong> patients with type 1 diabetes (•) after (a) 1–5 years (glucagon response is blunted whereasep<strong>in</strong>ephr<strong>in</strong>e release is preserved); and (b) with long-stand<strong>in</strong>g diabetes, both responses become severelyimpaired. = non-diabetic controls. Reproduced from Textbook of <strong>Diabetes</strong>, 2nd edition (1997)Pickup J. and William G. (eds) by permission of Blackwell Science Ltd. Data sourced from Bolliet al. (1983). Copyright © 1983 American <strong>Diabetes</strong> Association. Repr<strong>in</strong>ted with permission from TheAmerican <strong>Diabetes</strong> Association

132 COUNTERREGULATORY DEFICIENCIES IN DIABETESSystemic MediatorThe systemic mediator theory suggests that a substance is released <strong>in</strong> response to hypoglycaemiawhich attenuates subsequent sympathoadrenal responses to further episodes of hypoglycaemia.The <strong>in</strong>itial candidate for this was cortisol, based on two observations: first, theattenuat<strong>in</strong>g effect of antecedent hypoglycaemia on later sympathoadrenal responses is absent<strong>in</strong> patients with primary adrenocortical failure; and second, <strong>in</strong> healthy volunteers, follow<strong>in</strong>g<strong>in</strong>fusions of cortisol (to supraphysiological levels) dur<strong>in</strong>g euglycaemia, adrenomedullaryep<strong>in</strong>ephr<strong>in</strong>e secretion and muscle sympathetic neural activity were reduced dur<strong>in</strong>g subsequenthypoglycaemia (Davis et al., 1996; 1997). However, this effect of cortisol is lost ifthe prevail<strong>in</strong>g cortisol levels are lowered towards those seen dur<strong>in</strong>g hypoglycaemia or ifrecurrent hypoglycaemia is <strong>in</strong>duced <strong>in</strong> animals that are genetically modified to have absentadrenocortical responses (Raju et al., 2003; McGu<strong>in</strong>ess et al., 2005).Bra<strong>in</strong> Fuel TransportThis mechanism is based on the hypothesis that follow<strong>in</strong>g antecedent hypoglycaemia, glucosetransport from blood <strong>in</strong>to bra<strong>in</strong> tissue is <strong>in</strong>creased – <strong>in</strong> animals by <strong>in</strong>creas<strong>in</strong>g GLUT-1 transportacross the bra<strong>in</strong> microvasculature (McCall et al., 1986; Kumagai et al., 1995). In patients withtype 1 diabetes whose treatment resulted <strong>in</strong> near normal glucose levels, impaired awareness ofhypoglycaemia can develop – such patients are at <strong>in</strong>creased risk of seizures and coma. Boyleet al. (1995) tested the hypothesis that dur<strong>in</strong>g hypoglycaemia, these patients would have normalglucose uptake <strong>in</strong> the bra<strong>in</strong> and consequently that sympathoadrenal activation would not occur,result<strong>in</strong>g <strong>in</strong> impaired awareness of hypoglycaemia. They found that there was no significantchange <strong>in</strong> the uptake of glucose <strong>in</strong> the bra<strong>in</strong> among the patients with type 1 diabetes who hadthe lowest HbA 1c levels. Conversely, glucose uptake <strong>in</strong> the bra<strong>in</strong> fell <strong>in</strong> patients with less wellcontrolledtype 1 diabetes. The responses of plasma ep<strong>in</strong>ephr<strong>in</strong>e and pancreatic polypeptide andthe frequency of symptoms of hypoglycaemia were also lowest <strong>in</strong> the group with the lowestHbA 1c values. They concluded that dur<strong>in</strong>g hypoglycaemia, patients with nearly normal HbA 1cvalues have normal glucose uptake <strong>in</strong> the bra<strong>in</strong>, preserv<strong>in</strong>g cerebral metabolism, reduc<strong>in</strong>g theresponses of counterregulatory hormones, and caus<strong>in</strong>g impaired awareness of hypoglycaemia(Boyle et al., 1995). However, these f<strong>in</strong>d<strong>in</strong>gs occurred after days of prolonged hypoglycaemiawhich is <strong>in</strong> contrast to the cl<strong>in</strong>ical observation that attenuated sympathoadrenal responses occurwith<strong>in</strong> hours of a hypoglycaemic event.More recently, studies us<strong>in</strong>g positron emission tomography have found no change <strong>in</strong> bloodto-bra<strong>in</strong>glucose transport 24 hours after an episode of hypoglycaemia and no differencesbetween <strong>in</strong>dividuals with and without hypoglycaemia awareness (Segel et al., 2001; B<strong>in</strong>ghamet al., 2005). It rema<strong>in</strong>s possible, however, that there are changes <strong>in</strong> the transport of alternativecerebral fuels follow<strong>in</strong>g antecedent hypoglycaemia.Bra<strong>in</strong> MetabolismIt has been hypothesised that bra<strong>in</strong> metabolism per se is altered follow<strong>in</strong>g an episode ofhypoglycaemia. Most research <strong>in</strong> this area has focused on the ventromedial nucleus of thehypothalamus. Glucose deprivation <strong>in</strong> the VMH (by adm<strong>in</strong>istration of 2-deoxyglucose) activatesthe sympathoadrenal system and <strong>in</strong>creases glucagon secretion whereas local perfusion

AGE, OBESITY AND GLUCOSE COUNTERREGULATION 133GlucoseDeliveryUtilisationAdenos<strong>in</strong>eCaffe<strong>in</strong>eFigure 6.11 Caffe<strong>in</strong>e may act by uncoupl<strong>in</strong>g bra<strong>in</strong> glucose demand (<strong>in</strong>creased) and substrate delivery(decreased) through its actions on adenos<strong>in</strong>e receptors. Reproduced from Bra<strong>in</strong> Research Reviews, 17,Nehlig et al., 139–169, Copyright (1992), with permission from Elsevierof the area with glucose suppresses these responses dur<strong>in</strong>g systemic hypoglycaemia (Borget al., 1995; Borg et al., 1997). The mechanisms <strong>in</strong>volved are unknown but may be a consequenceof <strong>in</strong>creased glucok<strong>in</strong>ase activity to enhance glucose metabolism <strong>in</strong> neurones <strong>in</strong> theregion (Gabriely and Shamoon, 2005). However, it is likely that bra<strong>in</strong> metabolism <strong>in</strong> areasand other signall<strong>in</strong>g mechanisms with<strong>in</strong> the CNS are <strong>in</strong>fluenced by recurrent antecedenthypoglycaemia (Cryer, 2005).The bra<strong>in</strong> glycogen supercompensation hypothesis suggests that after a s<strong>in</strong>gle episodeof hypoglycaemia, there is a rebound <strong>in</strong>crease <strong>in</strong> glycogen formation <strong>in</strong> bra<strong>in</strong> astrocytes toprovide additional substrates (e.g. lactate) for bra<strong>in</strong> metabolism (Choi et al., 2003). Alterationsof substrate delivery to the bra<strong>in</strong> do appear to <strong>in</strong>fluence the magnitude of the hormonalcounterregulatory response to hypoglycaemia <strong>in</strong> healthy volunteers and <strong>in</strong> patients with type1 diabetes. Infusions of acetazolamide, a potent cerebral vasodilator, markedly attenuatesthese responses (Thomas et al., 1997) whereas <strong>in</strong>gestion of modest amounts of caffe<strong>in</strong>e (toreduce substrate delivery) augments the responses (Debrah et al., 1996). The mechanismsof the latter is unknown but may <strong>in</strong>volve antagonism of central adenos<strong>in</strong>e receptors withuncoupl<strong>in</strong>g of bra<strong>in</strong> blood flow (i.e., substrate delivery) and bra<strong>in</strong> glucose metabolism (i.e.,bra<strong>in</strong> glucose demand) result<strong>in</strong>g <strong>in</strong> relative cerebral neuroglycopenia (Figure 6.11). This isdiscussed <strong>in</strong> more detail <strong>in</strong> Chapter 5.AGE, OBESITY AND GLUCOSE COUNTERREGULATIONIn children with type 1 diabetes, the glucagon response to hypoglycaemia is markedly attenuatedcompared to non-diabetic <strong>in</strong>dividuals but compensated for by vigorous secretion of othercounterregulatory hormones, particularly ep<strong>in</strong>ephr<strong>in</strong>e, with the peak ep<strong>in</strong>ephr<strong>in</strong>e responsesbe<strong>in</strong>g almost two-fold higher than <strong>in</strong> adults (Amiel et al., 1987). The total sympathoadrenalresponses to hypoglycaemia are also <strong>in</strong>fluenced by pubertal stage (Ross et al., 2005).

134 COUNTERREGULATORY DEFICIENCIES IN DIABETESFurthermore, it appears that the glycaemic thresholds for the secretion of ep<strong>in</strong>ephr<strong>in</strong>e andgrowth hormone are set at a higher blood glucose level <strong>in</strong> non-diabetic children comparedto adults (Jones et al., 1991). In children with type 1 diabetes, the secretion of ep<strong>in</strong>ephr<strong>in</strong>e<strong>in</strong> response to hypoglycaemia commences at an even higher level. In children who havemarkedly elevated HbA 1c values, there is a further shift of the blood glucose threshold to ahigher level for the release of counterregulatory hormones.Advanced age, <strong>in</strong> otherwise healthy people, does not appear to dim<strong>in</strong>ish or delay counterregulatoryresponses to hypoglycaemia (Brierley et al., 1995), although the magnitudeof responses of ep<strong>in</strong>ephr<strong>in</strong>e and glucagon is lower at milder hypoglycaemic levels (around3.4 mmol/l ) compared to younger non-diabetic subjects, but is much more comparablewith a more profound hypoglycaemic stimulus (2.8 mmol/l) (Ortiz-Alonso et al., 1994) (SeeChapter 11). The magnitude of counterregulatory responses to low blood glucose levelsfollow<strong>in</strong>g preced<strong>in</strong>g hypoglycaemia also appears to depend on the gender of experimentalsubjects, with men hav<strong>in</strong>g blunted responses compared to women (Davis et al., 2000b).Both the autonomic nervous system and the hypothalamic–pituitary–adrenal axis areactivated <strong>in</strong> excess <strong>in</strong> the morbidly obese. Before and after bariatric surgery (averageweight loss 40 kg over 12 months), severely obese non-diabetic subjects, underwenta hyper<strong>in</strong>sul<strong>in</strong>aemic hypoglycaemic clamp (blood glucose 3.4 mmol/l). Before weightreduction, patients demonstrated brisk peak responses <strong>in</strong> glucagon, ep<strong>in</strong>ephr<strong>in</strong>e, pancreaticpolypeptide, and norep<strong>in</strong>ephr<strong>in</strong>e. After surgery and dur<strong>in</strong>g hypoglycaemia, all theseresponses were attenuated and most markedly so for glucagon, which was totally abolished<strong>in</strong> association with a marked improvement <strong>in</strong> <strong>in</strong>sul<strong>in</strong> sensitivity. In contrast,the growth hormone response was <strong>in</strong>creased after weight reduction (Guldstrand et al.,2003).HUMAN INSULIN AND COUNTERREGULATIONAt present there is no consistent evidence that the species of <strong>in</strong>sul<strong>in</strong> is an important determ<strong>in</strong>antof the counterregulatory response to hypoglycaemia. Over 25 cl<strong>in</strong>ical laboratory studieshave exam<strong>in</strong>ed the effect of <strong>in</strong>sul<strong>in</strong> species on the counterregulatory response to hypoglycaemia<strong>in</strong>duced by an <strong>in</strong>travenous bolus, <strong>in</strong>travenous <strong>in</strong>fusion, or subcutaneous <strong>in</strong>jection of<strong>in</strong>sul<strong>in</strong> (Fisher and Frier, 1993; Jorgensson et al., 1994). Most of the studies showed nosignificant differences between the hormonal responses. Two studies showed a reduction <strong>in</strong>the ep<strong>in</strong>ephr<strong>in</strong>e response to hypoglycaemia, and both of these studies also reported dim<strong>in</strong>ishedautonomic symptoms to hypoglycaemia after human <strong>in</strong>sul<strong>in</strong> (Schluter et al., 1982;He<strong>in</strong>e et al., 1989).A meta-analysis comparison of the effects of human and animal <strong>in</strong>sul<strong>in</strong> as well as ofthe adverse reaction profiles did not show cl<strong>in</strong>ically relevant differences between speciesespecially <strong>in</strong> terms of risk and responses to hypoglycaemia (Richter and Neises, 2005).TREATMENT OF COUNTERREGULATORY FAILUREAt present no treatment is available that will reverse the glucagon deficit that developswith<strong>in</strong> a few years of the onset of type 1 diabetes (Figure 6.12). However, there are strategies

TREATMENT OF COUNTERREGULATORY FAILURE 135GlucagonGlucose Glucose GlucoseTimeTimeGlucagonEp<strong>in</strong>ephr<strong>in</strong>eEp<strong>in</strong>ephr<strong>in</strong>eGlucagon andEp<strong>in</strong>ephr<strong>in</strong>eGlucagon andEp<strong>in</strong>ephr<strong>in</strong>eTimeFigure 6.12 Schematic representation of the consequences of defective glucagon, ep<strong>in</strong>ephr<strong>in</strong>e or acomb<strong>in</strong>ed defect of glucagon and ep<strong>in</strong>ephr<strong>in</strong>e release dur<strong>in</strong>g recovery from hypoglycaemiato reduce the risk of hypoglycaemia from caus<strong>in</strong>g further hypoglycaemia and promot<strong>in</strong>g<strong>Hypoglycaemia</strong> Associated Autonomic Failure:• relaxation of glycaemic targets;• use of multiple daily <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong>, utilis<strong>in</strong>g rapid-act<strong>in</strong>g and basal analogues (Bolli,2006);• consideration of switch<strong>in</strong>g to cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion therapy (Chaseet al., 2006);

136 COUNTERREGULATORY DEFICIENCIES IN DIABETES• <strong>in</strong>tensive education <strong>in</strong>clud<strong>in</strong>g carbohydrate count<strong>in</strong>g and appropriate blood glucose monitor<strong>in</strong>g;• use of novel technologies to aid diagnosis, e.g. cont<strong>in</strong>uous glucose monitor<strong>in</strong>g (Cheyneand Kerr, 2002);• discussion of patient factors, e.g. lipohypertrophy and other <strong>in</strong>jection site problems,alcohol, caffe<strong>in</strong>e consumption.Common psychological problems known to affect diabetes management adversely, such asanxiety, depression and eat<strong>in</strong>g disorders, have been extensively reported (Jacquem<strong>in</strong>et et al.,2005). Of particular relevance is the recognition of the role played by high levels of anxiety.Evidence-based treatment <strong>in</strong>terventions are available for treat<strong>in</strong>g anxiety <strong>in</strong> the non-diabeticpopulation; however a systematic review and meta-analysis of randomised controlled trialsof psychological <strong>in</strong>terventions for adults with diabetes has yet to be conducted. In cl<strong>in</strong>icalpractice it cont<strong>in</strong>ues to be recognised that there is a group of patients whose lives arecompletely disrupted by recurrent episodes of hyper and hypoglycaemia – the so-called‘brittle diabetic’. The outlook for such patients is usually poor (Tattersall et al., 1991).CONCLUSIONS• In non-diabetic <strong>in</strong>dividuals, cl<strong>in</strong>ically significant hypoglycaemia is an extremely rare eventbecause of effective glucose counterregulation. This <strong>in</strong>cludes suppression of endogenouspancreatic <strong>in</strong>sul<strong>in</strong> secretion, and release of glucagon, catecholam<strong>in</strong>es, cortisol and growthhormone.• The bra<strong>in</strong> is the critical organ for co-ord<strong>in</strong>ation of the physiological responses to lowblood glucose levels.• People with diabetes almost <strong>in</strong>evitably lose their ability to release glucagon <strong>in</strong> responseto a fall <strong>in</strong> blood glucose, with<strong>in</strong> five years of diagnosis. After ten years, a significantproportion of patients also has deficient ep<strong>in</strong>ephr<strong>in</strong>e responses and is at <strong>in</strong>creased risk ofmore protracted hypoglycaemia and neuroglycopenia.• Modern treatment of type 1 diabetes, <strong>in</strong>clud<strong>in</strong>g <strong>in</strong>tensive education and treatment regimensutilis<strong>in</strong>g new technologies, has failed to eradicate completely the problem of recurrenthypoglycaemia.REFERENCESAmiel SA, Simonson DC, Sherw<strong>in</strong> RS, Lauriano AA, Tamborlane WV (1987). Exaggeratedep<strong>in</strong>ephr<strong>in</strong>e responses to hypoglycemia <strong>in</strong> normal and <strong>in</strong>sul<strong>in</strong>-dependent diabetic children. Journalof Pediatrics 110: 832–7.Amiel S (1991). Glucose counter-regulation <strong>in</strong> health and disease: current concepts <strong>in</strong> hypoglycaemiarecognition and response. Quarterly Journal of Medic<strong>in</strong>e 293: 707–27.Banarer S, Cryer PE (2003). Sleep-related hypoglycemia-associated autonomic failure <strong>in</strong> type 1diabetes: reduced awaken<strong>in</strong>g from sleep dur<strong>in</strong>g hypoglycemia. <strong>Diabetes</strong> 52: 1195–203.

REFERENCES 137B<strong>in</strong>gham EM, Dunn JT, Smith D, Sutcliffe-Goulden J, Reed LJ, Mansden PK, Amiel SA (2005).Differential changes <strong>in</strong> bra<strong>in</strong> glucose metabolism dur<strong>in</strong>g hypoglycaemia accompany loss of hypoglycaemiaawareness <strong>in</strong> men with type 1 diabetes mellitus. An [11C]-3-O-methyl-D-glucose PETstudy. Diabetologia 48: 2080–9.Bolli G, De Feo P, Compagnucci P, Cartech<strong>in</strong>i MG, Angeletti G, Santeusanio F et al. (1983).Abnormal glucose counterregulation <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Interaction of anti<strong>in</strong>sul<strong>in</strong>antibodies and impaired glucagon and ep<strong>in</strong>ephr<strong>in</strong>e secretion. <strong>Diabetes</strong> 32: 134–41.Bolli GM (2006). Insul<strong>in</strong> treatment <strong>in</strong> type 1 diabetes. Endocr<strong>in</strong>e Practice 12 Suppl 1: 105–9.Borg WP, Dur<strong>in</strong>g MJ, Sherw<strong>in</strong> RS, Borg MA, Br<strong>in</strong>es ML, Shulman GI (1994). Ventromedial hypothalamiclesions <strong>in</strong> rats suppress counter regulatory responses to hypoglycemia. Journal of Cl<strong>in</strong>icalInvestigation 93: 1677–82.Borg WP, Sherw<strong>in</strong> RS, Dur<strong>in</strong>g MJ, Borg MA, Shulman GI (1995). Local ventromedial hypothalamusglucopenia triggers counterregulatory hormone release. <strong>Diabetes</strong> 44: 180–4.Borg MA, Sherw<strong>in</strong> RS, Borg WP, Tamborlane WV, Shulman GI (1997). Local ventromedial hypothalamusglucose perfusion blocks counterregulation dur<strong>in</strong>g systemic hypoglycemia <strong>in</strong> awake rats.Journal of Cl<strong>in</strong>ical Investigation 99: 361–5.Boyle PJ, Kempers SF, O’Connor AM, Nagy RJ (1995). Bra<strong>in</strong> glucose uptake and unawarenessof hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 333: 1726–31.Brierley EJ, Broughton DL, James OFW, Alberti KGMM (1995). Reduced awareness of hypoglycaemia<strong>in</strong> the elderly despite an <strong>in</strong>tact counter-regulatory response. Quarterly Journal of Medic<strong>in</strong>e 88: 439–45.Bulsara M, Holman CD, Davies EA, Jones TW (2004). The impact of a decade of chang<strong>in</strong>g treatmenton rates of severe hypoglycemia <strong>in</strong> a population-based cohort of children with type 1 diabetes.<strong>Diabetes</strong> Care 27: 2293–8.Chase HP, Horner BP, McFann K, Yetzer H, Gaston J, Banion C et al. (2006). The use of <strong>in</strong>sul<strong>in</strong> pumpswith meal bolus alarms <strong>in</strong> children with type 1 diabetes to improve glycemic control. <strong>Diabetes</strong> Care29: 1012–5.Cheyne E, Kerr D (2002) Mak<strong>in</strong>g ‘sense’ of diabetes: us<strong>in</strong>g a cont<strong>in</strong>uous glucose sensor <strong>in</strong> cl<strong>in</strong>icalpractice. <strong>Diabetes</strong> Metabolism Research Reviews 18 Suppl 1: S43–8.Choi I Y, Seaquist ER, Gruetter R (2003). Effect of hypoglycemia on bra<strong>in</strong> glycogen metabolism <strong>in</strong>vivo. Journal of Neuroscience Research 72: 25–32.Cox DJ, Irv<strong>in</strong>g GA, Gonder-Frederick L, Nowacek G, Butterfield J (1987). Fear of hypoglycemia:quantification, validation and utilization. <strong>Diabetes</strong> Care 10: 617–21.Cranston I, Lomas J, Maran A, Macdonald IA, Amiel SA (1994). Restoration of hypoglycaemiaawareness <strong>in</strong> patients with long-duration <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 344: 283–7.Cryer PE (2004). Diverse causes of hypoglycemia-associated autonomic failure <strong>in</strong> diabetes. NewEngland Journal of Medic<strong>in</strong>e 350: 2272–9.Cryer PE (2005). Mechanisms of hypoglycemia-associated autonomic failure and its componentsyndromes <strong>in</strong> diabetes. <strong>Diabetes</strong> 54: 3592–601.DAFNE Study Group (2002). Tra<strong>in</strong><strong>in</strong>g <strong>in</strong> flexible, <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> management to enable dietaryfreedom <strong>in</strong> people with type 1 diabetes: dose adjustment for normal eat<strong>in</strong>g (DAFNE) randomisedcontrolled trial. British Medical Journal 325: 746.Davis SN, Shavers C, Costa F, Mosqueda-Garcia R (1996). Role of cortisol <strong>in</strong> the pathogenesis ofdeficient counterregulation after antecedent hypoglycemia <strong>in</strong> normal humans. Journal of Cl<strong>in</strong>icalInvestigation 98: 680–91.Davis SN, Shavers C, Davis B, Costa F (1997). Prevention of an <strong>in</strong>crease <strong>in</strong> plasma cortisol dur<strong>in</strong>ghypoglycemia preserves subsequent counterregulatory responses. Journal of Cl<strong>in</strong>ical Investigation100: 429–38.Davis SN, Mann S, Galassetti P, Neil R, Tate D, Ertl AC, Costa F (2000a). Effects of differ<strong>in</strong>gdurations of antecedent hypoglycemia on counterregulatory responses to subsequent hypoglycemia<strong>in</strong> normal humans. <strong>Diabetes</strong> 49: 1897–903.

138 COUNTERREGULATORY DEFICIENCIES IN DIABETESDavis SN, Fowler F, Costa S (2000b). Hypoglycemic counterregulatory responses differ between menand women with type 1 diabetes. <strong>Diabetes</strong> 49: 65–72.Debrah K, Sherw<strong>in</strong> RS, Murphy J, Kerr D (1996). Effect of caffe<strong>in</strong>e on recognition of and physiologicalresponses to hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 347: 19–24Diem P, Redmon JB, Abid M, Moran A, Sutherland DE, Halter JB, Robertson RP (1990). Glucagon,catecholam<strong>in</strong>e and pancreatic polypeptide secretion <strong>in</strong> type I diabetic recipients of pancreas allografts.Journal of Cl<strong>in</strong>ical Investigation 86: 2008–13.Fisher M, Frier BM (1993). <strong>Hypoglycaemia</strong> and human <strong>in</strong>sul<strong>in</strong>. In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>:Cl<strong>in</strong>ical and Physiological Aspects. Frier BM and Fisher M, eds. Edward Arnold, London: 314–27.Fisher S, Brunn<strong>in</strong>g J, Lannon S, Kahn CR (2005). Insul<strong>in</strong> signal<strong>in</strong>g <strong>in</strong> the central nervous system iscritical for the normal sympathoadrenal response to hypoglycemia. <strong>Diabetes</strong> 54: 1447–51.Flanagan D, Kerr D (1996). Recurrent hypoglycemia and long-stand<strong>in</strong>g brittle diabetes. Endocr<strong>in</strong>ologyand Metabolism 3: 67–69.Frier BM, Fisher M, Gray CE, Beastall GH (1988). Counterregulatory hormonal responses to hypoglycaemia<strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabetes: evidence for dim<strong>in</strong>ished hypothalamic-pituitaryhormonal secretion. Diabetologia 31: 421–9.Gabriely I, Shamoon H (2005). Fructose normalizes specific counterregulatory responses to hypoglycemia<strong>in</strong> patients with type 1 diabetes. <strong>Diabetes</strong> 54: 609–16.Gerich JE, Langlois M, Noacco C, Karam JH, Forsham PH (1973). Lack of glucagon response tohypoglycemia <strong>in</strong> diabetes: evidence for an <strong>in</strong>tr<strong>in</strong>sic pancreatic alpha cell defect. Science 182: 171–3.Gerich JE, Bolli GB (1993). Counterregulatory failure. In: <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong>: Cl<strong>in</strong>ical andPhysiological Aspects. Frier BM and Fisher M, eds. Edward Arnold, London: 253–67.Guldstrand M, Ahren B, Wredl<strong>in</strong>g R, Backman L, L<strong>in</strong>s P, Adamson U (2003). Alteration of thecounterregulatory responses to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia and of cognitive function after massiveweight reduction <strong>in</strong> severely obese subjects. Metabolism 52: 900–7.Hardy KJ, Burge MR, Boyle P, Scarpello JHB (1994). A treatable cause of recurrent severe hypoglycemia.<strong>Diabetes</strong> Care 17: 722–4.He<strong>in</strong>e RJ, Van der Heyden EAP, Van der Veen EA (1989). Responses to human and porc<strong>in</strong>e <strong>in</strong>sul<strong>in</strong><strong>in</strong> healthy subjects. Lancet 334: 946–9.Heller SR, Cryer PE (1991). Reduced neuroendocr<strong>in</strong>e and symptomatic responses to subsequenthypoglycemia after 1 episode of hypoglycemia <strong>in</strong> non-diabetic humans. <strong>Diabetes</strong> 40: 223–6.Hvidberg A, Juel Christensen N, Hilsted J (1998). Counterregulatory hormones <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treateddiabetic patients admitted to an Accident and Emergency Department with hypoglycaemia. DiabeticMedic<strong>in</strong>e 15: 199–204.Jacquem<strong>in</strong>et S, Masseboeuff N, Rolland M, Grimaldi A, Sachon C (2005). Limitations of the so-called‘<strong>in</strong>tensified’ <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> type 1 diabetes mellitus. <strong>Diabetes</strong> Metabolism 31(4 Pt 2): 4S45–50.Jones TW, Boulware SD, Kraemer DT, Caprio S, Sherw<strong>in</strong> RS, Tamborlane WV (1991). Independenteffects of youth and poor diabetes control on responses to hypoglycemia <strong>in</strong> children. <strong>Diabetes</strong> 40:358–63.Jones TW, Porter P, Sherw<strong>in</strong> RS, Davis EA, O’Leary P, Frazer F et al. (1998). Decreased ep<strong>in</strong>ephr<strong>in</strong>eresponses to hypoglycemia dur<strong>in</strong>g sleep. New England Journal of Medic<strong>in</strong>e 338: 1657–62.Jorgensen L, Dejgaard A, Pramm<strong>in</strong>g SK (1994). Human <strong>in</strong>sul<strong>in</strong> and hypoglycaemia: a literature survey.Diabetic Medic<strong>in</strong>e 11: 925–34.Kerr D, Reza M, Smith N, Leatherdale BA (1991). Importance of <strong>in</strong>sul<strong>in</strong> <strong>in</strong> subjective, cognitive andhormonal responses to hypoglycemia <strong>in</strong> patients with IDDM. <strong>Diabetes</strong> 40: 1057–62.Kerr D, Tamborlane WV, Rife F, Sherw<strong>in</strong> RS (1993). Effect of <strong>in</strong>sul<strong>in</strong>-like growth factor-1 on theresponses to and recognition of hypoglycemia <strong>in</strong> humans. Journal of Cl<strong>in</strong>ical Investigation 91:141–7.Kumagai AK, Kang YS, Boado RJ, Pardridge WM (1995). Upregulation of blood-bra<strong>in</strong> barrier GLUT-1 glucose transporter prote<strong>in</strong> and mRNA <strong>in</strong> experimental chronic hypoglycemia. <strong>Diabetes</strong> 44:1399–404.

REFERENCES 139McCall AL, Fixman LB, Flem<strong>in</strong>g N, Tornheim K, Chick W, Ruderman NB (1986). Chronic hypoglycemia<strong>in</strong>creases bra<strong>in</strong> glucose transport. American Journal of Endocr<strong>in</strong>ology 251: E442–7.McCrimmon RJ, Evans ML, Fan X, McNay EC, Chan O, D<strong>in</strong>g Y et al. (2005). Activation ofATP-sensitive K+ channels <strong>in</strong> the ventromedial hypothalamus amplifies counterregulatory hormoneresponses to hypoglycemia <strong>in</strong> normal and recurrently hypoglycemic rats. <strong>Diabetes</strong> 54: 3169–74.McGu<strong>in</strong>ess O, Ansari T, Potts J, Jacobson L (2005). Glucocorticoid <strong>in</strong>sufficiency <strong>in</strong> corticotroph<strong>in</strong>releas<strong>in</strong>g hormone knockout mice does not preserve autonomic responses dur<strong>in</strong>g repeated hypoglycemiaclamps. <strong>Diabetes</strong> 54 (Suppl): A26 (abstract).Merl V, Kern W, Peters A, Oltmanns KM, Gais S, Born J et al. (2004). Differences between nighttimeand daytime hypoglycemia counterregulation <strong>in</strong> healthy humans. Metabolism 53: 894–8.Mitrakou A, Ryan C, Veneman T, Mokan M, Jenssen T, Kiss I et al. (1991). Hierarchy of glycemicthresholds for counterregulatory hormonal secretion, symptoms, and cerebral dysfunction. AmericanJournal of Endocr<strong>in</strong>ology 260: E67–74.Muhlhauser I, Bruckner I, Berger M, Cheta D, Jorgens V, Lonescu-Tirgoviste C et al. (1987). Evaluationof an <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> treatment and teach<strong>in</strong>g programme as rout<strong>in</strong>e management of type 1(<strong>in</strong>sul<strong>in</strong>-dependent) diabetes. The Bucharest–Dusseldorf Study. Diabetologia 30: 681–90.Nehlig A, Daval JL, Debry G (1992). Caffe<strong>in</strong>e and the central nervous system: mechanisms of action,biochemical, metabolic and psychostimulant effects. Bra<strong>in</strong> Research Reviews 17: 139–69.Ortiz-Alonso FJ, Galecki A, Herman WH, Smith MJ, Jacquez JA, Halter JB (1994). Hypoglycemiacounterregulation <strong>in</strong> elderly humans: relationship to glucose levels. American Journal ofEndocr<strong>in</strong>ology 267: E497–506.Raju B, McGregor VP, Cryer PE (2003). Cortisol elevations comparable to those that occurdur<strong>in</strong>g hypoglycemia do not cause hypoglycemia-associated autonomic failure. <strong>Diabetes</strong> 52:2083–89.Richter B, Neises G (2005). ‘Human’ <strong>in</strong>sul<strong>in</strong> versus animal <strong>in</strong>sul<strong>in</strong> <strong>in</strong> people with diabetes mellitus.Cochrane Database Systematic Reviews Jan 25(1): CD003816.Ross LA, Warren RE, Kelnar CJ, Frier BM. (2005). Pubertal stage and hypoglycaemia counterregulation<strong>in</strong> type 1 diabetes. Archives of Disease <strong>in</strong> Childhood 90: 190–4.Ross<strong>in</strong>g P. (2005) The chang<strong>in</strong>g epidemiology of diabetic microangiopathy <strong>in</strong> type 1 diabetes.Diabetologia 2005; 48: 1439–44.Samols E, Tyler J, Marks V (1972): Glucagon-<strong>in</strong>sul<strong>in</strong> <strong>in</strong>terrelationships. In: Glucagon: MolecularPhysiology, Cl<strong>in</strong>ical and Therapeutic Implications. Lefebvre P, Unger RH, eds. Pergamon, ElmsfordNY: 151–174.Schluter KJ, Petersen KG, Sontheimer K, Enzmann F, Kerp L (1982). Different counterregulatoryresponses to human <strong>in</strong>sul<strong>in</strong> (recomb<strong>in</strong>ant DNA) and purified pork <strong>in</strong>sul<strong>in</strong>. <strong>Diabetes</strong> Care 5 (suppl 2):78–81.Segel SA, Fanelli CG, Dence CS, Markham J, Videen TO, Paramore DS et al. (2001). Blood-to-bra<strong>in</strong>glucose transport, cerebral glucose metabolism and cerebral blood flow are not <strong>in</strong>creased follow<strong>in</strong>ghypoglycemia. <strong>Diabetes</strong> 50: 1911–17.Smith D, Pernet D, Reid H, B<strong>in</strong>gham E, Rosenthal JM, Macdonald IA et al. (2002). The role ofhepatic portal glucose sens<strong>in</strong>g <strong>in</strong> modulat<strong>in</strong>g responses to hypoglycaemia <strong>in</strong> man. Diabetologia 45:1416–24.Tappy L, Chiorelo R, Berger M (1999). Autoregulation of glucose production <strong>in</strong> health and disease.Current Op<strong>in</strong>ions <strong>in</strong> Cl<strong>in</strong>ical Nutrition and Metabolic Care 2: 161–4.Tattersall RB, Gregory R, Selby C, Kerr D, Heller S (1991). Course of brittle diabetes: 12 year followup. British Medical Journal 302: 1240–3.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1993). The effect of <strong>in</strong>tensive treatmentof diabetes on the development and progression of long-term complications <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. New England Journal of Medic<strong>in</strong>e 329: 977–86.Thomas M, Sherw<strong>in</strong> R, Murphy J, Kerr D (1997). Importance of cerebral blood flow on recognitionof and physiological responses to hypoglycemia. <strong>Diabetes</strong> 46: 829–33.

140 COUNTERREGULATORY DEFICIENCIES IN DIABETESWe<strong>in</strong>trob N, Schechter A, Benzaquen H, Shalit<strong>in</strong> S, Lilos P, Galatzer A, Phillip M (2004). Glycemicpatterns detected by cont<strong>in</strong>uous subcutaneous glucose sens<strong>in</strong>g <strong>in</strong> children and adolescents withtype 1 diabetes mellitus treated by multiple daily <strong>in</strong>jections versus cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong><strong>in</strong>fusion. Archives of Pediatric and Adolescent Medic<strong>in</strong>e 158: 677–84.White NH, G<strong>in</strong>gerich RL, Levandoski LA, Cryer PE, Santiago JV (1985). Plasma pancreatic polypeptideresponse to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia as a marker for defective glucose counterregulation<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes. <strong>Diabetes</strong> 34: 870–5.

7 Impaired Awareness of<strong>Hypoglycaemia</strong>Brian M. FrierDangerous hypoglycaemia may occur without warn<strong>in</strong>g symptoms– E.P. Josl<strong>in</strong> et al. (1922)INTRODUCTIONThe generation of symptoms <strong>in</strong> response to hypoglycaemia provides a fundamental defencefor the bra<strong>in</strong>, by alert<strong>in</strong>g the affected <strong>in</strong>dividual to the imm<strong>in</strong>ent development of neuroglycopenia(Chapter 2). This should provoke an appropriate response – obta<strong>in</strong><strong>in</strong>g and <strong>in</strong>gest<strong>in</strong>gsome form of carbohydrate to reverse the low blood glucose. If these warn<strong>in</strong>g symptomsfail to occur, or they are delayed until the blood glucose has fallen to a level that causesdisabl<strong>in</strong>g neuroglycopenia, serious consequences may ensue. When the normal warn<strong>in</strong>gmechanisms are deficient or are ignored and no avoid<strong>in</strong>g action is taken, severe hypoglycaemiamay occur, with progression to confusion, altered consciousness and eventual coma.An <strong>in</strong>adequate symptomatic warn<strong>in</strong>g often occurs <strong>in</strong> people with <strong>in</strong>sul<strong>in</strong>-treated diabetes, <strong>in</strong>various circumstances and with differ<strong>in</strong>g causes, and is described as impaired awarenessof hypoglycaemia or hypoglycaemia unawareness. This is an acquired abnormality that iseffectively a complication of <strong>in</strong>sul<strong>in</strong> therapy, and should be ranked alongside the microvascularcomplications of diabetes such as ret<strong>in</strong>opathy, neuropathy or nephropathy, because itsmorbidity can be just as serious and disabl<strong>in</strong>g.NORMAL RESPONSES TO HYPOGLYCAEMIAAcute hypoglycaemia <strong>in</strong>duces a series of changes – hormonal, neurophysiological, symptomaticand cognitive – which occur at different and def<strong>in</strong>ed blood glucose concentrations(Figure 7.1). The thresholds at which these changes are triggered have been described <strong>in</strong>non-diabetic humans, most occurr<strong>in</strong>g with<strong>in</strong> a relatively narrow range of blood glucoseconcentrations. In diabetic <strong>in</strong>dividuals these glycaemic thresholds are not static and permanent,but are dynamic and display plasticity, alter<strong>in</strong>g <strong>in</strong> response to external <strong>in</strong>fluences suchas changes <strong>in</strong> glycaemic control and exposure to extremes of blood glucose. Thus the bloodglucose level at which symptoms are activated can be modified through the ability of thebra<strong>in</strong> to adapt to environmental change, that is, its exposure to prevail<strong>in</strong>g blood glucoseconcentrations.<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

142 IMPAIRED AWARENESS OF HYPOGLYCAEMIAArterialised venous blood glucose concentration (mmol/L) mmol/LInhibition of endogenous<strong>in</strong>sul<strong>in</strong> secretion3.8 mmol/LCounterregulatoryhormone release• Glucagon• Ep<strong>in</strong>ephr<strong>in</strong>e3.2–2.8 mmol/LOnset of symptoms• Autonomic• Neuroglycopenic3.0–2.4 mmol/LNeurophysiologicaldysfunction• Evoked responses3.0 mmol/LWidespreadEEG changes2.8 mmol/LCognitivedysfunction• Inability toperform complextasks< 1.5 mmol/LSevereneuroglycopenia• Reducedconscious level• Convulsions• Coma0Figure 7.1 Hierarchy of endocr<strong>in</strong>e, symptomatic and neurological responses to acute hypoglycaemia<strong>in</strong> non-diabetic subjects. Glycaemic thresholds are based on glucose concentrations <strong>in</strong> arterialisedvenous blood. Modified from Textbook of <strong>Diabetes</strong>, 2nd edition (1997) (eds J. Pickup and G. Williams),by permission of Blackwell Science LtdDepriv<strong>in</strong>g the bra<strong>in</strong> of glucose causes it to malfunction, and cognitive impairment quicklybecomes evident as an overt manifestation of neuroglycopenia. Some of these features arerelatively subtle, and may not be detected immediately by the patient. A fall <strong>in</strong> blood glucosetriggers activation of the peripheral autonomic nervous system via central hypothalamic autonomiccentres with<strong>in</strong> the bra<strong>in</strong>, and stimulates the sympathoadrenal system. This promotestypical physiological responses <strong>in</strong>clud<strong>in</strong>g sweat<strong>in</strong>g, an <strong>in</strong>crease <strong>in</strong> rate and contractility of theheart (sensed as a pound<strong>in</strong>g heart), and tremor, these be<strong>in</strong>g some of the classical features ofthe autonomic reaction (Figure 7.2). Ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) is secreted <strong>in</strong> large quantitiesfrom the adrenal medullae and contributes to some of the symptoms ma<strong>in</strong>ly by heighten<strong>in</strong>gthe magnitude of the response. The early literature on hypoglycaemia and diabetes providesaccurate descriptions of the autonomic features of acute hypoglycaemia, and patients andphysicians alike commonly discussed hypoglycaemic ‘reactions’, a term that regrettably isnow seldom used. It emphatically describes the sudden, and often florid, onset of the autonomicfeatures of hypoglycaemia, which drive the <strong>in</strong>dividual to seek assistance or obta<strong>in</strong> asupply of glucose to relieve these unpleasant symptoms.‘Awareness’ of <strong>Hypoglycaemia</strong>The generation of typical physiological responses to hypoglycaemia is perceived throughsensory feedback to the bra<strong>in</strong>, and after central process<strong>in</strong>g, an appropriate motor responseis made. Much has been made by some commentators of the predom<strong>in</strong>ant importanceof autonomic symptoms <strong>in</strong> the detection of the onset of hypoglycaemia. This premise is

NORMAL RESPONSES TO HYPOGLYCAEMIA 143We do not have rights to reproduce thisfigure electronicallyFigure 7.2 Generation of neuroglycopenic and autonomic symptoms <strong>in</strong> response to hypoglycaemia.Autonomic activation and the <strong>in</strong>volvement of the sympatho-adrenal system <strong>in</strong> the stimulation of representativeend-organs associated with common autonomic symptoms of hypoglycaemia. Reproducedfrom <strong>Hypoglycaemia</strong> and <strong>Diabetes</strong> (eds B.M. Frier and M. Fisher), © 1993 Edward Arnold, bypermission of Edward Arnold (Publishers) Ltdbased partly on the laboratory-based observation of non-diabetic subjects that autonomicsymptoms commence at a higher blood glucose concentration (around 0.5 mmol/l) thanneuroglycopenic symptoms (Mitrakou et al., 1991). In everyday experience reported bypeople with <strong>in</strong>sul<strong>in</strong>-treated diabetes, a rapid decl<strong>in</strong>e <strong>in</strong> blood glucose does not permit asubjective dist<strong>in</strong>ction to be made between these different thresholds for the developmentof autonomic and neuroglycopenic symptoms, and people treated with <strong>in</strong>sul<strong>in</strong> identify bothtypes with equal frequency as their <strong>in</strong>itial warn<strong>in</strong>g symptoms (Hepburn et al., 1992).It has been assumed that because neuroglycopenia may <strong>in</strong>terfere with cognitive function,this will affect the <strong>in</strong>dividual’s ability to perceive and <strong>in</strong>terpret neuroglycopenic cues such asthe <strong>in</strong>ability to concentrate, drows<strong>in</strong>ess or difficulty with mentation. This may be true whena fall<strong>in</strong>g blood glucose is not treated and is allowed to drop to a level associated with severe

144 IMPAIRED AWARENESS OF HYPOGLYCAEMIAneuroglycopenia, but most patients detect (and often rely upon) neuroglycopenic symptomsdur<strong>in</strong>g early hypoglycaemia, and rate these as important as autonomic symptoms <strong>in</strong> provid<strong>in</strong>ga warn<strong>in</strong>g. It is the <strong>in</strong>itial perception of any symptom of hypoglycaemia, irrespective ofwhether this is autonomic, neuroglycopenic or simply a vague sensation of apprehension orloss of well-be<strong>in</strong>g (a common early feature described by many) which constitutes ‘awareness’of hypoglycaemia. Only the <strong>in</strong>itial warn<strong>in</strong>g symptoms are important <strong>in</strong> this respect, and notthe total spectrum or absolute number of symptoms, some of which occur too late to haveany value <strong>in</strong> alert<strong>in</strong>g the patient to the impend<strong>in</strong>g risk of a fall<strong>in</strong>g blood glucose. A majordifference between the autonomic and the neuroglycopenic symptomatic response is that,once triggered, the autonomic response quickly reaches a maximum <strong>in</strong>tensity which thengradually decl<strong>in</strong>es with time, whereas the neuroglycopenic response becomes more profoundthe further the blood glucose falls. This qualitative difference <strong>in</strong> response becomes importantif early cues are ignored or are not detected, as progressive neuroglycopenia will eventually<strong>in</strong>terfere with the <strong>in</strong>dividual’s ability to identify and self-treat the low blood glucose.When a person is fully awake, alert and on guard aga<strong>in</strong>st possible hypoglycaemia, this symptomaticwarn<strong>in</strong>g system generally works very effectively (Chapter 2). However, there are manytimes <strong>in</strong> everyday life when the symptoms may be either dim<strong>in</strong>ished or disregarded. This isparticularly so dur<strong>in</strong>g sleep when symptoms are seldom detected, or they may be ignoredif a person is distracted by other activities, such as watch<strong>in</strong>g an <strong>in</strong>terest<strong>in</strong>g programme ontelevision, participat<strong>in</strong>g <strong>in</strong> sport or concentrat<strong>in</strong>g on a task. Circumstances can modify thevalue of specific warn<strong>in</strong>g symptoms, mak<strong>in</strong>g them difficult to <strong>in</strong>terpret as features of hypoglycaemia.Examples <strong>in</strong>clude sweat<strong>in</strong>g on a hot day, shiver<strong>in</strong>g when the weather is cold orfeel<strong>in</strong>g drowsy dur<strong>in</strong>g a bor<strong>in</strong>g meet<strong>in</strong>g! All of these may represent early hypoglycaemiabut are attributed to other causes by the affected person. A list of the factors that <strong>in</strong>fluencenormal awareness of hypoglycaemia is shown <strong>in</strong> Box 7.1. The <strong>in</strong>tensity of symptoms canvary and the value of <strong>in</strong>dividual symptoms as warn<strong>in</strong>g features may not be constant <strong>in</strong> anys<strong>in</strong>gle <strong>in</strong>dividual. This is often not appreciated <strong>in</strong> the assessment of research f<strong>in</strong>d<strong>in</strong>gs, and it isdifficult to extrapolate the careful measurement of symptomatic responses to hypoglycaemia<strong>in</strong> studies performed <strong>in</strong> a laboratory sett<strong>in</strong>g, to the hurly-burly of everyday life.Box 7.1Factors <strong>in</strong>fluenc<strong>in</strong>g normal awareness of hypoglycaemiaInternalPhysiologicalRecent glycaemic controlDegree of neuroglycopeniaSymptom <strong>in</strong>tensity/sensitivityPsychologicalFocused attentionCongruence; denialCompet<strong>in</strong>g explanationsEducationKnowledgeSymptom beliefExternalDrugsBeta-adrenoceptor blockers (non-selective)Hypnotics, tranquillisersAlcoholEnvironmentalPostureDistraction

IMPAIRED AWARENESS OF HYPOGLYCAEMIA 145Warn<strong>in</strong>g symptoms provide <strong>in</strong>ternal cues, but most people with <strong>in</strong>sul<strong>in</strong>-treated diabetesalso rely on external cues based on their experience of the tim<strong>in</strong>g of <strong>in</strong>sul<strong>in</strong> adm<strong>in</strong>istration<strong>in</strong> relation to food, the effect of delay<strong>in</strong>g meals or the amount of food <strong>in</strong>gested, the effect ofexercise on blood glucose and many other factors that can <strong>in</strong>fluence short-term glycaemiccontrol. These cues are supplemented by blood glucose monitor<strong>in</strong>g, which gives an exactand objective measure of prevail<strong>in</strong>g glycaemia. Further useful feedback may be obta<strong>in</strong>edfrom observers such as relatives or friends, many of whom become adept at notic<strong>in</strong>g earlyneuroglycopenia before the onset of the patient’s subjective warn<strong>in</strong>g symptoms. ‘Awareness’of hypoglycaemia is therefore distilled from a comb<strong>in</strong>ation of resources, and has to belearned by people with newly diagnosed diabetes commenc<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> therapy. They have noprevious experience of symptoms of hypoglycaemia and must receive appropriate educationon the potential range of symptoms. Through experience they will recognise the clusterof symptoms peculiar to themselves, because symptoms are idiosyncratic. Awareness ofhypoglycaemia therefore assists <strong>in</strong> protect<strong>in</strong>g the <strong>in</strong>dividual from the risk of an unexpectedfall <strong>in</strong> blood glucose. When awareness of hypoglycaemia becomes impaired or is absentwhile a person is awake, the <strong>in</strong>dividual becomes progressively vulnerable to the developmentof severe hypoglycaemia.IMPAIRED AWARENESS OF HYPOGLYCAEMIADef<strong>in</strong>itionNo satisfactory or comprehensive def<strong>in</strong>ition of impaired hypoglycaemia awareness has beensuggested to date. Many laboratory-based studies of experimental hypoglycaemia have usedarbitrary def<strong>in</strong>itions based on witnessed observations of subjects who fail to develop classicalfeatures of hypoglycaemia, or the failure of physiological or hormonal responses toexceed twice the standard deviation from mean basal levels. These are statistical devices,which take no account of subjective reality, require the application of sophisticated andunphysiological glucose clamp procedures, and have little direct application to cl<strong>in</strong>icalmanagement.Asymptomatic biochemical hypoglycaemia occurs more frequently dur<strong>in</strong>g rout<strong>in</strong>e bloodglucose monitor<strong>in</strong>g <strong>in</strong> diabetic patients who report impaired awareness of hypoglycaemia(Gold et al., 1994; Clarke et al., 1995) and such a record may alert the cl<strong>in</strong>ician to thepossibility that an <strong>in</strong>dividual is develop<strong>in</strong>g this problem. A much higher rate of undetectedhypoglycaemia <strong>in</strong> people with impaired awareness has been demonstrated dur<strong>in</strong>g wak<strong>in</strong>ghours us<strong>in</strong>g cont<strong>in</strong>uous blood glucose monitor<strong>in</strong>g (Kubiak et al., 2004). However, <strong>in</strong> cl<strong>in</strong>icalpractice a careful history is essential <strong>in</strong> determ<strong>in</strong><strong>in</strong>g whether reduced warn<strong>in</strong>g symptoms ofhypoglycaemia are a significant problem, and if this is occurr<strong>in</strong>g consistently. Patients whoassert that they have a problem with perceiv<strong>in</strong>g the onset of symptoms of hypoglycaemiaare generally correct <strong>in</strong> this belief (Clarke et al., 1995), so that the identification of impairedawareness of hypoglycaemia should be based pr<strong>in</strong>cipally on cl<strong>in</strong>ical history. Validatedscor<strong>in</strong>g systems to assess awareness of hypoglycaemia have been described by Gold et al.(1994) and Clarke et al. (1995), and supportive <strong>in</strong>formation can be derived from simultaneous<strong>in</strong>spection of the <strong>in</strong>dividual’s blood glucose results. Detailed question<strong>in</strong>g of a patient about hisor her ability to detect the onset of hypoglycaemic symptoms may need to be supplementedby question<strong>in</strong>g close relatives, who often report a much higher rate of severe hypoglycaemia

146 IMPAIRED AWARENESS OF HYPOGLYCAEMIA(Heller et al., 1995; Jorgensen et al., 2003). This will provide a witnessed descriptionof how hypoglycaemia develops <strong>in</strong> a patient, with <strong>in</strong>formation on its true frequency andseverity. Patients often underestimate the frequency of severe hypoglycaemia, partly becauseof post-hypoglycaemia amnesia.ClassificationIn one study, Hepburn et al. (1990) subdivided hypoglycaemia awareness <strong>in</strong>to three categories:normal, partial and absent awareness. These were def<strong>in</strong>ed as follows:• Normal awareness: the <strong>in</strong>dividual is always aware of the onset of hypoglycaemia.• Partial awareness: the symptom profile has changed with a reduction either <strong>in</strong> the <strong>in</strong>tensityor <strong>in</strong> the number of symptoms and, <strong>in</strong> addition, the <strong>in</strong>dividual may be aware of someepisodes of some episodes of hypoglycaemia but not of others.• Absent awareness: the <strong>in</strong>dividual is no longer aware of any episode of hypoglycaemia.Although the subdivision <strong>in</strong>to partial and absent awareness is artificial, it reflects the naturalhistory of this cl<strong>in</strong>ical problem, illustrat<strong>in</strong>g the gradual progression of this disability, andemphasis<strong>in</strong>g that <strong>in</strong> some patients the abnormality is severe (absent awareness) althoughtotal absence of cl<strong>in</strong>ical manifestations of hypoglycaemia (particularly the neuroglycopenicfeatures) is exceptionally rare (Gold et al., 1994, Clarke et al., 1995). The problem maynot be simply an absence of symptoms, but rather that the time dur<strong>in</strong>g which warn<strong>in</strong>gsymptoms can be detected is extremely short, allow<strong>in</strong>g the affected <strong>in</strong>dividual a very limitedopportunity to take avoid<strong>in</strong>g action. Some patients describe how the onset of hypoglycaemiaappears to have become much more rapid compared with their previous experience andprogresses quickly to severe neuroglycopenia. However, impaired awareness may not necessarilyevolve <strong>in</strong>to total unawareness of hypoglycaemia, and may vary over time, presumablybecause of major <strong>in</strong>fluences of environmental factors on the generation and perception ofsymptoms.The above classification of awareness of hypoglycaemia is far from comprehensive. Inaddition, the state of hypoglycaemia awareness can be ascerta<strong>in</strong>ed only when the <strong>in</strong>dividualis <strong>in</strong> a physical state <strong>in</strong> which recognition of the onset of hypoglycaemia is possible.Therefore, if the person is asleep, <strong>in</strong>toxicated, <strong>in</strong>ebriated, anaesthetised or sedated, so thattheir conscious level is reduced, they are not able to perceive (as subjective symptoms)the normal physiological manifestations of hypoglycaemia. An <strong>in</strong>dividual’s awareness ofhypoglycaemia can be evaluated only if hypoglycaemia occurs while the <strong>in</strong>dividual is awake.A further prerequisite is that the person must have had previous experience of hypoglycaemiaat some time dur<strong>in</strong>g treatment with <strong>in</strong>sul<strong>in</strong>. In assess<strong>in</strong>g the present state ofhypoglycaemia awareness, it is desirable that the patient should have experienced one ormore episodes of hypoglycaemia (confirmed biochemically) with<strong>in</strong> a recent time <strong>in</strong>tervalsuch as the preced<strong>in</strong>g year, so that a comparison of the symptoms can be made with earlierepisodes of hypoglycaemia. A diagnosis of impaired hypoglycaemia awareness cannot beenterta<strong>in</strong>ed or surmised if a patient has either never been exposed previously to acute hypoglycaemiaor has only started to experience hypoglycaemic events very recently. Becausehypoglycaemia awareness and its impairment is a cont<strong>in</strong>uum rang<strong>in</strong>g from normality to

PREVALENCE OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 147complete <strong>in</strong>ability to detect the onset of hypoglycaemia, a classification of this conditionwill need to consider alterations <strong>in</strong> symptom <strong>in</strong>tensity as well as detection of hypoglycaemiaby any means and the ability of the patient to self-treat low blood glucose.PREVALENCE OF IMPAIRED AWARENESS OFHYPOGLYCAEMIAImpaired awareness of hypoglycaemia is common <strong>in</strong> people treated with <strong>in</strong>sul<strong>in</strong>. Althoughthe chronic form of this acquired condition ma<strong>in</strong>ly affects those with type 1 diabetes, itappears that a similar problem does eventually emerge <strong>in</strong> patients with type 2 diabetes whohave been treated with <strong>in</strong>sul<strong>in</strong> for several years (Hepburn et al., 1993a). A prevalence of 8%was observed <strong>in</strong> a cohort of 215 patients <strong>in</strong> Ed<strong>in</strong>burgh (Henderson et al., 2003). Becausefew patients with type 2 diabetes who require <strong>in</strong>sul<strong>in</strong> therapy survive for a sufficientlylong period to permit this complication to develop, impaired awareness is pr<strong>in</strong>cipally aproblem associated with type 1 diabetes. It is not known whether impaired awareness ofhypoglycaemia occurs <strong>in</strong> diabetic patients treated with oral antidiabetic agents.Impaired awareness of hypoglycaemia has been shown to be associated with strictglycaemic control (see Chapter 8), but significant modification of the symptomatic responseto hypoglycaemia does not occur unless the glycated haemoglob<strong>in</strong> concentration is with<strong>in</strong>the non-diabetic range (Boyle et al., 1995; K<strong>in</strong>sley et al., 1995; Pampanelli et al., 1996).Only a small proportion of people with <strong>in</strong>sul<strong>in</strong>-treated diabetes can susta<strong>in</strong> this degree ofsuper-optimal glycaemic control <strong>in</strong>def<strong>in</strong>itely. In the <strong>Diabetes</strong> Control and ComplicationsTrial (DCCT), with its extensive resources devoted to ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy,more than 40% of the patients <strong>in</strong> the group with strict glycaemic control achieved a HbA 1c of6.05% or less (the upper limit of the non-diabetic range) at some time dur<strong>in</strong>g the study, butonly 5% were able to ma<strong>in</strong>ta<strong>in</strong> this level of glycaemic control cont<strong>in</strong>uously (The <strong>Diabetes</strong>Control and Complications Trial Research Group, 1993). The proportion of any <strong>in</strong>sul<strong>in</strong>treateddiabetic population that can achieve this therapeutic goal will depend on local policiesregard<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> therapy, the expertise of local diabetes specialist teams, available resourcesand the enthusiasm of <strong>in</strong>dividual patients. With the exception of a few highly motivatedpatients, most people treated with <strong>in</strong>sul<strong>in</strong> are unable to ma<strong>in</strong>ta<strong>in</strong> strict glycaemic controlfor protracted periods. In cl<strong>in</strong>ical practice this ‘acute’ form of hypoglycaemia unawarenessis probably relatively uncommon. Nonetheless, the <strong>in</strong>fluence of strict glycaemic control onsymptomatic and counterregulatory responses to hypoglycaemia has been studied extensively,and has provided <strong>in</strong>sights <strong>in</strong>to the potential pathogenetic mechanisms underly<strong>in</strong>gimpaired awareness of hypoglycaemia.Reduced warn<strong>in</strong>g symptoms of hypoglycaemia (of vary<strong>in</strong>g severity) occur <strong>in</strong> approximatelyone quarter of all <strong>in</strong>sul<strong>in</strong>-treated patients. Cross-sectional population surveys <strong>in</strong>different European and North American populations of <strong>in</strong>sul<strong>in</strong>-treated diabetic patients, us<strong>in</strong>gsimilar methods of assessment, have given remarkably consistent estimates (Table 7.1).Impaired awareness of hypoglycaemia becomes more common with <strong>in</strong>creas<strong>in</strong>g duration of<strong>in</strong>sul<strong>in</strong>-treated diabetes (Hepburn et al., 1990), and almost 50% of patients experience hypoglycaemiawithout warn<strong>in</strong>g symptoms after 25 years or more of treatment (Pramm<strong>in</strong>g et al.,1991) (Figure 7.3). It appears therefore to be an acquired abnormality associated with <strong>in</strong>sul<strong>in</strong>therapy.

148 IMPAIRED AWARENESS OF HYPOGLYCAEMIATable 7.1 Prevalence of <strong>Hypoglycaemia</strong> Unawareness <strong>in</strong> population studies of<strong>in</strong>sul<strong>in</strong>-treated diabetesCountryNumber ofpatientsImpaired awareness ofhypoglycaemia (%)ReferenceScotland 302 23 Hepburn et al. (1990)Germany 523 25 Muhlhauser et al. (1991)Denmark 411 27 Pramm<strong>in</strong>g et al. (1991)USA 628 20 Orchard et al. (1991)Figure 7.3 Comparisons between the duration of diabetes and the percentage of 411 type 1 diabeticpatients report<strong>in</strong>g (a) changes <strong>in</strong> symptoms of hypoglycaemia, (b) sweat<strong>in</strong>g and/or tremor as one of thetwo card<strong>in</strong>al autonomic symptoms of hypoglycaemia, and (c) severe hypoglycaemic episodes withoutwarn<strong>in</strong>g symptoms. Values are medians; shaded areas show 95% confidence limits. Reproduced fromPramm<strong>in</strong>g et al. (1991) by permission of John Wiley & Sons, Ltd

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 149Frequency of Associated Severe <strong>Hypoglycaemia</strong>It is apparent that impaired awareness of hypoglycaemia is a major risk factor for severehypoglycaemia. In the DCCT, 36% of all episodes of severe hypoglycaemia occurred withno warn<strong>in</strong>g symptoms <strong>in</strong> patients while they were awake (The DCCT Research Group,1991). In a population study <strong>in</strong> Ed<strong>in</strong>burgh, retrospective assessment of the frequency ofsevere hypoglycaemia revealed that 90% of patients with impaired awareness of symptomsexperienced severe hypoglycaemia <strong>in</strong> the preced<strong>in</strong>g year, compared to 18% <strong>in</strong> a comparablegroup who had reta<strong>in</strong>ed normal awareness (Hepburn et al., 1990). Prospective studies haveconfirmed the <strong>in</strong>crease <strong>in</strong> frequency of mild and severe hypoglycaemia associated withimpaired awareness of hypoglycaemia (Gold et al., 1994; Clarke et al., 1995), with a sixfoldhigher frequency of severe hypoglycaemia be<strong>in</strong>g documented <strong>in</strong> people with impairedawareness (Gold et al., 1994) (Figure 7.4).Annual Prevalence(% Subjects)Incidence(Events per subject per year)Figure 7.4 Proportion of patients affected and event rates for severe hypoglycaemia <strong>in</strong> patientswith type 1 diabetes with normal (□, n= 31) or impaired (, n= 29) awareness of hypoglycaemia.Reproduced from Cryer and Frier (2004) by permission of John Wiley & Sons, Ltd. Data derived fromGold et al. (1994)PATHOGENESIS OF IMPAIRED AWARENESS OFHYPOGLYCAEMIAThe mechanisms underly<strong>in</strong>g impaired awareness of hypoglycaemia are not known and maybe multifactorial. Possible mechanisms are listed <strong>in</strong> Box 7.2.

150 IMPAIRED AWARENESS OF HYPOGLYCAEMIABox 7.2Impaired awareness of hypoglycaemia: possible mechanismsCNS adaptationChronic exposure to low blood glucose• glucose clamp (2.9 mmol/l) for 56 hours <strong>in</strong> non-diabetic subjects• <strong>in</strong>sul<strong>in</strong>oma <strong>in</strong> non-diabetic patients• strict glycaemic control <strong>in</strong> diabetic patientsRecurrent transient exposure to low blood glucose• antecedent hypoglycaemiaCNS glucoregulatory failure• counterregulatory deficiency (hypothalamic defect?)• hypoglycaemia associated (central) autonomic failure (HAAF)Peripheral nervous system dysfunction• peripheral autonomic neuropathy• reduced peripheral adrenoceptor sensitivityAltered Glycaemic Threshold for Initiation of SymptomsSymptoms of hypoglycaemia commence when the blood glucose reaches a specific level, andalthough this threshold may differ between <strong>in</strong>dividuals, it is usually constant and reproducible<strong>in</strong> the non-diabetic state (Vea et al., 1992). This blood glucose threshold for symptoms canbe modified by protracted hypoglycaemia (Boyle et al., 1994) and is not fixed <strong>in</strong> people withdiabetes who are treated with <strong>in</strong>sul<strong>in</strong>, with its dynamic nature be<strong>in</strong>g demonstrated <strong>in</strong> varioussituations. In cl<strong>in</strong>ical practice, it has long been recognised that <strong>in</strong>sul<strong>in</strong>-treated diabetic patientswho have poor glycaemic control experience symptoms of hypoglycaemia when their bloodglucose decl<strong>in</strong>es with<strong>in</strong> a hyperglycaemic range (Maddock and Krall, 1953) and this hasbeen shown to be associated with the onset of hypoglycaemic symptoms at a significantlyhigher blood glucose (4.3 mmol/l) compared to non-diabetic subjects (2.9 mmol/l) (Boyleet al., 1988). Conversely, strict glycaemic control modifies the glycaemic threshold for theonset of symptoms, which do not commence until blood glucose has decl<strong>in</strong>ed to a lowerlevel than that required <strong>in</strong> less well controlled patients to <strong>in</strong>itiate a symptomatic response(see Chapter 8).The term<strong>in</strong>ology that is used <strong>in</strong> relation to a change <strong>in</strong> the glycaemic threshold is potentiallyconfus<strong>in</strong>g. When a lower blood glucose is required to <strong>in</strong>itiate a response, whethersymptomatic, physiological or counterregulatory, the glycaemic threshold is said to be raised

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 151or elevated; that is, a more profound hypoglycaemic stimulus is necessary to trigger therelevant response. Thus, strict glycaemic control raises the glycaemic threshold for theonset of symptoms, which do not occur until blood glucose has decl<strong>in</strong>ed to a much lowerconcentration than would be observed <strong>in</strong> non-diabetic subjects.For many years, cl<strong>in</strong>icians have recognised that the glycaemic threshold for the onset ofhypoglycaemic symptoms is higher <strong>in</strong> patients with a long duration of type 1 diabetes whorequire a much lower blood glucose to provoke a symptomatic response. Lawrence (1941)wrote that ‘as years of <strong>in</strong>sul<strong>in</strong> life go on, sometimes only after 5–10 years, I f<strong>in</strong>d it almostthe rule that the type of <strong>in</strong>sul<strong>in</strong> reactions change, the premonitory autonomic symptoms aremissed out and the patient proceeds directly to the more serious manifestations affect<strong>in</strong>g thecentral nervous system’. He astutely suggested that ‘the tissues may become attuned to a lowersugar concentration’. Recent studies <strong>in</strong> animals and humans have shown that the bra<strong>in</strong> doesadapt to chronic exposure to low blood glucose (see below) but this may not be beneficialto the <strong>in</strong>dividual with diabetes who is treated with <strong>in</strong>sul<strong>in</strong>, i.e., it is a maladaptive response.An early study by Sussman et al. (1963) – revisited and extended by Hepburn et al.(1991) – showed that diabetic patients who had self-reported unawareness of hypoglycaemiadid mount a sympatho-adrenal response to acute hypoglycaemia, but that this occurred ata lower blood glucose concentration than comparable diabetic subjects who had normalsymptomatic awareness (Figure 7.5). However, the autonomic response was preceded by theWe do not have rights to reproduce thisfigure electronicallyFigure 7.5 Venous blood glucose concentrations for the onset of the autonomic reaction <strong>in</strong> responseto <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycaemia <strong>in</strong> <strong>in</strong>dividual non-diabetic control subjects, and <strong>in</strong> type 1 diabeticpatients with normal and impaired awareness of hypoglycaemia and with autonomic neuropathy.Mean + SEM is shown for each group. Data derived from Hepburn et al. (1991) and reproduced from<strong>Hypoglycaemia</strong> and <strong>Diabetes</strong> (eds B.M. Frier and M. Fisher), © 1993 Edward Arnold, by permissionof Edward Arnold (Publishers) Ltd

152 IMPAIRED AWARENESS OF HYPOGLYCAEMIAdevelopment of overt neuroglycopenia, which <strong>in</strong>terfered with perception of the autonomicwarn<strong>in</strong>g symptoms when they did eventually occur. This sequence of responses disrupts theability of the <strong>in</strong>dividual subject to take appropriate action to self-treat low blood glucose.Similar f<strong>in</strong>d<strong>in</strong>gs have been reported by others (Grimaldi et al., 1990; Mokan et al., 1994;Bacatselos et al., 1995).In these studies, the counterregulatory hormonal responses to hypoglycaemia were delayed(Grimaldi et al., 1990; Hepburn et al. 1991) and their glycaemic thresholds had also shiftedto occur at lower blood glucose concentrations (Mokan et al., 1994). This is consistentwith the reported observation that impaired awareness of hypoglycaemia co-segregates withcounterregulatory hormonal deficiency <strong>in</strong> people with longstand<strong>in</strong>g type 1 diabetes (Ryderet al., 1990). In addition, Mokan et al. (1994) reported that cognitive dysfunction andneuroglycopenic symptoms <strong>in</strong> people with impaired awareness occurred at lower bloodglucose levels than <strong>in</strong> people with type 1 diabetes who had normal awareness. This suggeststhat people with impaired awareness can function effectively with very low blood glucoseconcentrations, at which symptoms and cognitive impairment would normally occur <strong>in</strong> nondiabeticand aware diabetic subjects. The potential risk of this situation is apparent: it is ak<strong>in</strong>to walk<strong>in</strong>g along the edge of a cliff on a dark night. With such a narrow glycaemic warn<strong>in</strong>gzone the propensity to rapidly develop severe neuroglycopenia is high and the marg<strong>in</strong> forerror is dangerously narrow.The results of these laboratory-based experimental studies of diabetic patients who haveestablished hypoglycaemia unawareness are consistent with cl<strong>in</strong>ical observations of peoplewith this acquired problem. At one moment they appear to be cerebrat<strong>in</strong>g normally (despitetheir blood glucose be<strong>in</strong>g low) then they rapidly become confused or drowsy, often with avacant or dazed appearance and an <strong>in</strong>ertia to seek some form of carbohydrate to reverse theneuroglycopenia. They may have to rely on relatives, friends or colleagues to identify thehypoglycaemia and provide treatment. This becomes a serious emergency if the patient isalone or if the <strong>in</strong>sidious, but often rapid, development of neuroglycopenia goes unobserved.This expla<strong>in</strong>s the <strong>in</strong>creased risk of progression to severe hypoglycaemia, and the higher ratesreported <strong>in</strong> people with hypoglycaemia unawareness.Studies exam<strong>in</strong><strong>in</strong>g the effects of strict glycaemic control on symptomatic and counterregulatoryresponses to hypoglycaemia have also demonstrated a similar shift <strong>in</strong> glycaemicthreshold for autonomic symptoms and an acute sympatho-adrenal response. However, theeffect on glycaemic thresholds for neuroglycopenic symptoms and cognitive dysfunctionrema<strong>in</strong>s controversial (see Chapter 8).Peripheral Autonomic NeuropathyFor many years, peripheral autonomic neuropathy was considered to be the pr<strong>in</strong>cipalcause of impaired awareness of hypoglycaemia (Hoeldtke et al., 1982). This was basedon the assumption that the dim<strong>in</strong>ished secretion of ep<strong>in</strong>ephr<strong>in</strong>e <strong>in</strong> response to hypoglycaemia(Hilsted et al., 1981; Bott<strong>in</strong>i et al., 1997) would either prevent the generationof autonomic symptoms (such as sweat<strong>in</strong>g or a pound<strong>in</strong>g heart) or reduce their <strong>in</strong>tensity,result<strong>in</strong>g <strong>in</strong> an <strong>in</strong>ability to perceive the onset of hypoglycaemia. Thus, autonomicneuropathy would <strong>in</strong>terfere with the normal physiological responses stimulated by autonomicactivation.

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 153There are various reasons why this hypothesis is unlikely:• Although ep<strong>in</strong>ephr<strong>in</strong>e can augment the <strong>in</strong>tensity of a few autonomic symptoms of hypoglycaemia,it has a very limited role <strong>in</strong> their generation, which is modulated by sympatheticneural activation; the reduced secretory response of ep<strong>in</strong>ephr<strong>in</strong>e <strong>in</strong> autonomic neuropathyis compensated by an <strong>in</strong>crease <strong>in</strong> sensitivity of peripheral beta-adrenoceptors (Hilstedet al., 1987).• Diabetic subjects with autonomic neuropathy have normal physiological responses andexperience typical autonomic symptoms dur<strong>in</strong>g hypoglycaemia (Hilsted et al., 1981;Hepburn et al., 1993b), and no relationship has been found between autonomic dysfunctionand hypoglycaemic symptoms (Berl<strong>in</strong> et al., 1987).• Impaired awareness of hypoglycaemia co-segregates with deficient counterregulatoryhormonal responses and not with autonomic neuropathy (Ryder et al., 1990).• The prevalence of autonomic neuropathy is similar <strong>in</strong> patients with type 1 diabetes oflong duration (more than 15 years), whether or not they have impaired awareness ofhypoglycaemia (Hepburn et al., 1990).• Although impaired awareness of hypoglycaemia is a major risk factor for severe hypoglycaemia,the latter is either no more common <strong>in</strong> type 1 diabetic patients with autonomicneuropathy (Bjork et al., 1990; The DCCT Research Group, 1991), or is only modestly<strong>in</strong>creased (Stephenson et al., 1996).• Autonomic neuropathy is not a determ<strong>in</strong>ant of whether glycaemic thresholds for autonomic(<strong>in</strong>clud<strong>in</strong>g symptomatic) responses to hypoglycaemia are affected by antecedenthypoglycaemia (Dagogo-Jack et al., 1993).Both impaired awareness of hypoglycaemia and peripheral autonomic neuropathy arecommon <strong>in</strong> people with type 1 diabetes of long duration, and frequently coexist. This doesnot prove a causal relationship, and it would appear that peripheral autonomic dysfunctiondoes not have a prom<strong>in</strong>ent role <strong>in</strong> the pathogenesis of this syndrome.However, reduced sensitivity of cardiac beta-adrenoceptors to catecholam<strong>in</strong>es has beenobserved <strong>in</strong> patients with type 1 diabetes who have impaired awareness of hypoglycaemia(Berl<strong>in</strong> et al., 1987). <strong>Hypoglycaemia</strong> per se reduces beta-adrenergic sensitivity <strong>in</strong> type 1diabetes (Fritsche et al., 1998), and this sensitivity is <strong>in</strong>creased after avoidance of hypoglycaemiafor four months <strong>in</strong> people who have impaired awareness (Fritsche et al., 2001).The improved beta-adrenergic sensitivity correlated with a rise <strong>in</strong> autonomic symptomscores. Maladaptation of tissue sensitivity to catecholam<strong>in</strong>es may therefore contribute tothe development of hypoglycaemia unawareness even though autonomic neuropathy is notpresent.<strong>Hypoglycaemia</strong> Associated Autonomic FailureThe co-segregation of impaired hypoglycaemia awareness with counterregulatory deficiencysuggests that they share a common underly<strong>in</strong>g pathogenetic mechanism. These acquiredabnormalities associated with hypoglycaemia <strong>in</strong> type 1 diabetes (Box 7.3) are characterisedby a high frequency of severe hypoglycaemia and a common pathophysiological

154 IMPAIRED AWARENESS OF HYPOGLYCAEMIABox 7.3Acquired syndromes associated with hypoglycaemia <strong>in</strong> type 1 diabetes• Counterregulatory deficiency• Impaired hypoglycaemia awareness• Altered glycaemic thresholds for counterregulatory and symptomatic responsesFigure 7.6 Schematic diagram of the concept of hypoglycaemia associated autonomic failure(HAAF), based on Cryer (1992)feature, namely the elevated glycaemic thresholds (or lower blood glucose concentrations)that are required to trigger symptomatic and hormonal secretory responses. Inother words, more profound hypoglycaemia is necessary to produce the usual symptomaticand counterregulatory responses to acute hypoglycaemia. Cryer (1992) has designatedthis group of abnormalities as a form of ‘hypoglycaemia associated autonomicfailure’ (HAAF), and has speculated that recurrent severe hypoglycaemia may be theprimary problem which establishes a vicious circle (Figure 7.6). It seems likely that thisdefect resides with<strong>in</strong> the central nervous system. The possible mechanisms underly<strong>in</strong>gHAAF and related hypoglycaemia syndromes <strong>in</strong> diabetes have been reviewed <strong>in</strong> detail(Cryer, 2005).

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 155Central Nervous System Adaptation to <strong>Hypoglycaemia</strong>Some people with <strong>in</strong>sul<strong>in</strong>-treated diabetes rema<strong>in</strong> lucid, with no evidence of impaired cognitivefunction, when their blood glucose is low (often well below 3.5 mmol/l). Biochemicalhypoglycaemia that is asymptomatic is commonly recorded by patients who have impairedawareness of hypoglycaemia, and they appear to have developed a neurological adaptationto chronic neuroglycopenia. The altered glycaemic threshold prevents the onset of warn<strong>in</strong>gsymptoms and cognitive dysfunction until the blood glucose falls to a dangerously low level,which is extremely undesirable when striv<strong>in</strong>g for safe cl<strong>in</strong>ical management of <strong>in</strong>sul<strong>in</strong>-treateddiabetes.Although the human bra<strong>in</strong> is dependent on a cont<strong>in</strong>uous supply of glucose for normalfunction, it can adapt to prolonged exposure to hypoglycaemia. This adaptation processtakes at least several hours and possibly a few days to occur. Short-term exposure toacute hypoglycaemia (blood glucose 2.5 mmol/l) for 60 m<strong>in</strong>utes <strong>in</strong> non-diabetic subjectsshowed no improvement <strong>in</strong> cognitive function and no reduction <strong>in</strong> symptom scoresdur<strong>in</strong>g this brief time <strong>in</strong>terval (Gold et al., 1995a). However, when non-diabetic subjectswere subjected to chronic hypoglycaemia (blood glucose 2.9 mmol/l) for 56 hours, us<strong>in</strong>ga glucose clamp, significant cerebral adaptation did occur (Boyle et al., 1994). Theresponses to acute hypoglycaemia (blood glucose 2.5 mmol/l) were compared before andafter the period of chronic hypoglycaemia. Bra<strong>in</strong> glucose uptake was <strong>in</strong>itially reducedwhen blood glucose was below 3.6 mmol/l, but after a period of chronic hypoglycaemiauptake was preserved and cerebral function was ma<strong>in</strong>ta<strong>in</strong>ed (Figure 7.7), demonstrat<strong>in</strong>gan effect of cerebral adaptation to chronic neuroglycopenia. The glycaemic thresholds forthe onset of symptoms, counterregulatory hormonal secretion and cognitive dysfunction,were all modified and occurred at much lower blood glucose concentrations. A similarphenomenon has been observed <strong>in</strong> non-diabetic patients who had an <strong>in</strong>sul<strong>in</strong>oma caus<strong>in</strong>gchronic hypoglycaemia; symptomatic responses to acute hypoglycaemia were blunted andcounterregulatory hormonal responses were impaired but cognitive function was unaffected(Mitrakou et al., 1993). Surgical removal of the <strong>in</strong>sul<strong>in</strong>-secret<strong>in</strong>g tumour reversedthese abnormalities <strong>in</strong>dicat<strong>in</strong>g that they had resulted from cerebral adaptation to chronichypoglycaemia.Strict glycaemic control <strong>in</strong> people with <strong>in</strong>sul<strong>in</strong>-treated diabetes also alters the glycaemicthresholds for the development of counterregulatory hormones and symptoms (Chapter 8),so that a lower blood glucose concentration is required to trigger these responses. Theobservation that this requires a reduction <strong>in</strong> HbA 1c to with<strong>in</strong> the non-diabetic range(K<strong>in</strong>sley et al., 1995) suggests that the median daily blood glucose <strong>in</strong> these <strong>in</strong>dividualsis relatively low, and the frequency of biochemical (and symptomatic) hypoglycaemiawill be greater than <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients who are not as wellcontrolled (Thorste<strong>in</strong>sson et al., 1986). Boyle et al. (1995) have shown that those patientswho had near normal HbA 1c values, ma<strong>in</strong>ta<strong>in</strong>ed normal uptake of glucose by the bra<strong>in</strong>dur<strong>in</strong>g hypoglycaemia, so preserv<strong>in</strong>g cerebral metabolism, reduc<strong>in</strong>g the counterregulatoryresponses to hypoglycaemia and dim<strong>in</strong>ish<strong>in</strong>g symptomatic awareness. Although thiscapacity to ma<strong>in</strong>ta<strong>in</strong>, and even <strong>in</strong>crease, cerebral blood glucose uptake dur<strong>in</strong>g hypoglycaemiais a protective response for the bra<strong>in</strong> <strong>in</strong> these patients with strict glycaemiccontrol, it is considered to be maladaptive, because it suppresses the normal symptomaticwarn<strong>in</strong>g of responses and so risks the development of much more profoundneuroglycopenia.

156 IMPAIRED AWARENESS OF HYPOGLYCAEMIAFigure 7.7 Rates of bra<strong>in</strong> glucose uptake, ep<strong>in</strong>ephr<strong>in</strong>e concentration and total symptoms of hypoglycaemia<strong>in</strong> non-diabetic subjects before and after prolonged hypoglycaemia. Initial day of <strong>in</strong>vestigation(hatched); after 56 hours of hypoglycaemia (solid). ∗ Significant difference from basel<strong>in</strong>e for each ofthe two days. Reproduced from Boyle (1997), Diabetologia, 40, S69–S74. With k<strong>in</strong>d permission ofSpr<strong>in</strong>ger Science and Bus<strong>in</strong>ess MediaDur<strong>in</strong>g hypoglycaemia, glucose transport <strong>in</strong>to the bra<strong>in</strong> becomes rate-limit<strong>in</strong>g, and bra<strong>in</strong>energy metabolism deteriorates. The adaptive response results from an <strong>in</strong>creased utilisationof glucose by the bra<strong>in</strong>. In rodents, the transport of glucose across the blood–bra<strong>in</strong> barrier is<strong>in</strong>creased after several days of chronic hypoglycaemia (McCall et al., 1986). Further studies<strong>in</strong> rats of glucose transport activity across the blood–bra<strong>in</strong> barrier have shown that whenblood glucose was kept below 2.0 mmol/l for several days, changes <strong>in</strong> expression of theglucose transporter, GLUT-1, <strong>in</strong> bra<strong>in</strong> microvasculature occurred <strong>in</strong> response to the chronichypoglycaemia (Kumagai et al., 1995). This <strong>in</strong>crease <strong>in</strong> GLUT-1 activity was responsiblefor the compensatory <strong>in</strong>crease <strong>in</strong> glucose transport across the blood–bra<strong>in</strong> barrier.

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA 157Antecedent (Episodic) <strong>Hypoglycaemia</strong>It has been recognised for many years that severe hypoglycaemia is associated with furtherepisodes of severe hypoglycaemia, and one episode may <strong>in</strong>fluence the cl<strong>in</strong>ical manifestationsof another occurr<strong>in</strong>g soon afterwards (Sever<strong>in</strong>ghaus, 1926). In recent years, several studieshave shown that the symptomatic and counterregulatory responses to an episode of acutehypoglycaemia are dim<strong>in</strong>ished if a preced<strong>in</strong>g (or antecedent) episode of hypoglycaemiahas occurred with<strong>in</strong> the previous 24 hours. Several studies have been performed <strong>in</strong> nondiabeticsubjects (Table 7.2) and <strong>in</strong> people with <strong>in</strong>sul<strong>in</strong>-treated diabetes (Table 7.3). Althoughthese studies differ considerably <strong>in</strong> design and methods of <strong>in</strong>duc<strong>in</strong>g hypoglycaemia, <strong>in</strong>general it appears that antecedent hypoglycaemia of between one and two hours durationhas a significant <strong>in</strong>fluence on the magnitude of the symptomatic and counterregulatoryresponses to subsequent hypoglycaemia occurr<strong>in</strong>g with<strong>in</strong> the follow<strong>in</strong>g 24 to 48 hours(Figure 7.8).The glycaemic thresholds for symptomatic and counterregulatory hormonal responses arealtered by antecedent hypoglycaemia, and the degree to which subsequent responses areblunted are determ<strong>in</strong>ed by the duration and depth of antecedent hypoglycaemia (Davis et al.,1997). Some of the physiological responses (e.g. sweat<strong>in</strong>g) may be blunted for longer thanother responses follow<strong>in</strong>g antecedent hypoglycaemia (George et al., 1995). Recurrent, shortlived(15 m<strong>in</strong>utes) episodes of hypoglycaemia on four consecutive days, had no effect oncounterregulatory and symptomatic responses <strong>in</strong> non-diabetic subjects (Peters et al., 1995),and so transient reductions <strong>in</strong> blood glucose may not produce this effect. Davis et al. (2000)have observed that a short duration of antecedent hypoglycaemia (20 m<strong>in</strong>utes to lowerand raise blood glucose from 3.9 to 2.9 mmol/l with the blood glucose be<strong>in</strong>g ma<strong>in</strong>ta<strong>in</strong>edfor five m<strong>in</strong>utes at 2.9 mmol/l) did not affect symptomatic awareness of hypoglycaemia,but did blunt the counterregulatory hormonal responses. Antecedent hypoglycaemia alsoreduces the counterregulatory responses to exercise on the follow<strong>in</strong>g day, both <strong>in</strong> nondiabetic(Davis et al., 2000b) and type 1 diabetic subjects (Galassetti et al., 2003) and<strong>in</strong>fluences the metabolic responses, particularly dim<strong>in</strong>ish<strong>in</strong>g endogenous glucose production<strong>in</strong> response to exercise. This may promote exercise-<strong>in</strong>duced hypoglycaemia <strong>in</strong> type 1diabetes.Some studies have exam<strong>in</strong>ed the effect of antecedent hypoglycaemia on cognitive function,but <strong>in</strong> many the methods of assessment were <strong>in</strong>adequate and <strong>in</strong>sufficient to provide def<strong>in</strong>itiveevidence of a change <strong>in</strong> cognitive response. Although some ma<strong>in</strong>ta<strong>in</strong> that the glycaemicthreshold for cognitive dysfunction is not altered by hypoglycaemia (see Chapter 8), an<strong>in</strong>creas<strong>in</strong>g number of studies have suggested that this does shift to a lower blood glucoseconcentration <strong>in</strong> the same manner as the thresholds for autonomic and counterregulatoryresponses (Veneman et al., 1993; Ovalle et al., 1998; Fanelli et al., 1998). Consistentwith these observations, a study from Germany <strong>in</strong> non-diabetic men has shown thatafter a s<strong>in</strong>gle episode of antecedent hypoglycaemia, subsequent hypoglycaemia had lesseffect on auditory-evoked bra<strong>in</strong> potentials and short-term memory (Fruehwald-Schulteset al., 2000), demonstrat<strong>in</strong>g cerebral adaptation that preserves cognitive function. Nocturnal(episodic) hypoglycaemia, which is frequently not identified by patients, has been proposedas a mechanism for the <strong>in</strong>duction of hypoglycaemia unawareness <strong>in</strong> people who giveno history of recurrent hypoglycaemia (Veneman et al., 1993). The possible mechanismsof cerebral adaptation caus<strong>in</strong>g impaired awareness of hypoglycaemia are summarised <strong>in</strong>Box 7.4.

Table 7.2 Studies of antecedent hypoglycaemia (AH) <strong>in</strong> non-diabetic humansReferences No. of subjectsMethod of <strong>in</strong>duction,nadir BG (mmol/l) andduration of AHInterval before testhypoglycaemiaTest hypo: methodand BG nadir(mmol/l)Effect of AH on subsequent responsesto hypoglycaemiaHeller and Cryer(1991)Widom and Simonson(1992)9 Clamp (3.0) 2 h 18 h Clamp (2.8) Reduced symptoms and CR responses10 Clamp (2.2–2.8) 1 h 4 consecutive dailyhyposStepped clamp(2.2)Veneman et al. (1993) 10 iv. <strong>in</strong>fusion (2.2–2.5) 2 h 7 h Stepped clamp(2.3)Mellman et al. (1994) 9 Clamp (3.2) 2 h 1.5 h Stepped clamp(2.8)Elevated BG thresholds for symptomsand CR responsesNocturnal hypo: BG thresholds forsymptoms and CR responseselevatedElevated BG thresholds for symptomsand CR responsesRob<strong>in</strong>son et al. (1995) 10 Clamp (3.0) 2 h 24 h and 6 days Clamp (2.5) Reduced adrenal<strong>in</strong>e and sweat<strong>in</strong>g at 6daysPeters et al. (1995) 10 iv. bolus <strong>in</strong>jection of<strong>in</strong>sul<strong>in</strong> (< 28)< 15 m<strong>in</strong>4 consecutive dailyhyposiv. bolus (< 28)X4No effect on symptoms or CRresponsesGeorge et al. (1995) 8 Clamp (2.9) 2 h 2 and 5 days Clamp (2.5) Physiological responses still reducedafter 1 weekDavis et al. (1997) 8 Stepped clamp (2.9) 2 h < 24 h Clamp (2.9) Magnitude of reduced CR responserelated to depth of AHDavis et al. (2000a) 31 Clamp (2.9) 2 h (n = 15);30 m<strong>in</strong> (n = 16); 5 m<strong>in</strong>(n = 10)24 h Clamp (2.9) Reduced CR responses <strong>in</strong> all studies.Symptom scores unaffected byshort duration hypoglycaemiaDavis et al. (2000b) 16 Clamp (2.9) 2 h 24 h Exercise (90 m<strong>in</strong>) Blunt<strong>in</strong>g of CR responses to exercise.Attenuated metabolic responses(reduced endogenous glucoseproduction)Fruehwald-Schulteset al. (2000)30 Stepped clamp (2.6) 2.5 h 18–24 h Clamp (3.1) Auditory-evoked bra<strong>in</strong> potentials andshort-term memory less affectedBG = blood glucose; AH = antecedent hypoglycaemia; CR = counterregulatory (hormonal); iv = <strong>in</strong>travenous

Table 7.3 Studies of antecedent hypoglycaemia <strong>in</strong> people with <strong>in</strong>sul<strong>in</strong>-treated diabetesReference No. of subjectsMethod of <strong>in</strong>duction,nadir BG (mmol/l)and duration of AHInterval before testhypoglycaemiaTest hypo: methodand BG nadir(mmol/l)Effect of AH on responses tohypoglycaemiaDavis et al.(1992)Dagogo-Jacket al. (1993)L<strong>in</strong>genfelseret al. (1993)George et al.(1997)Ovalle et al.(1998)Fanelli et al.(1998)Galassettiet al. (2003)13 Clamp (3.0) 2 h 60 m<strong>in</strong> Clamp (3.0) Reduced CR responses and hepaticglucose output26 (± AN) 12(non-diabetic)18 iv. bolus <strong>in</strong>jectionof <strong>in</strong>sul<strong>in</strong> (< 2.2)m<strong>in</strong>s X 3Clamp (2.7) 2 h 15 h Stepped clamp(2.6)3 days Stepped clamp(1.7)8 Clamp (2.8) 2 h 2 days Stepped clamp(2.2)6 Clamp (2.8) 2 h,twice weekly for1 month15 Clamp (2.6) 3.5 h,nocturnal48 h Stepped clamp(2.5)5–10 h Stepped clamp(2.5)16 Clamp (2.9) 2 h 24 h Exercise for 90m<strong>in</strong>Elevated BG thresholds forsymptoms and autonomicresponsesReduced symptoms, CR andneurophysiological responsesPhysiological responses dim<strong>in</strong>ishedfor 2 daysReduced symptoms, CR andcognitive dysfunction. Fewersymptomatic episodes of cl<strong>in</strong>icalhypoglycaemia.Less deterioration <strong>in</strong> cognitivefunction after nocturnalhypoglycaemia (versuseuglycaemia). Elevated BGthreshold for cognitivedysfunctionBlunt<strong>in</strong>g of CR responses toexercise. Three fold <strong>in</strong>crease <strong>in</strong>glucose needAH = antecedent hypoglycaemia; AN = autonomic neuropathy; iv = <strong>in</strong>travenous; CR = counterregulatory (hormonal)

160 IMPAIRED AWARENESS OF HYPOGLYCAEMIAFigure 7.8 Schematic representation of the effect of antecedent hypoglycaemia on the neuroendocr<strong>in</strong>eand symptomatic responses to subsequent hypoglycaemiaBox 7.4 Mechanisms of cerebral adaptation caus<strong>in</strong>g impaired awareness ofhypoglycaemiaSymptomatic and Neuroendocr<strong>in</strong>e responses to hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treateddiabetes are dim<strong>in</strong>ished <strong>in</strong> association with:• strict glycaemic control (HbA 1c <strong>in</strong> non-diabetic range)• antecedent (episodic) hypoglycaemia• chronic (protracted) hypoglycaemiaThey may be restored by:• relaxation of glycaemic control• scrupulous avoidance of hypoglycaemiaAlthough antecedent hypoglycaemia may <strong>in</strong>duce transient impairment of awareness ofhypoglycaemia it is unclear how this mechanism would <strong>in</strong>duce chronic or prolonged loss ofsymptomatic perception. Although frequent, recurrent hypoglycaemia may have a contributoryeffect to <strong>in</strong>duc<strong>in</strong>g hypoglycaemia unawareness, presumably the hypoglycaemia has tobe relatively protracted to <strong>in</strong>duce prolonged cerebral adaptation, and the phenomenon is notlimited to patients who have strict glycaemic control. The problem rema<strong>in</strong>s of expla<strong>in</strong><strong>in</strong>gthe <strong>in</strong>duction of protracted or chronic hypoglycaemia unawareness, which often appears tobe a permanent defect. Presumably repetitive hypoglycaemic <strong>in</strong>sults to the bra<strong>in</strong> (which arenot necessarily severe) eventually ‘downregulate’ the central mechanisms that sense a lowblood glucose and activate the glucoregulatory responses with<strong>in</strong> the hypothalamus. There isevidence of a permanent redistribution of regional cerebral blood flow <strong>in</strong> diabetic patientswith a history of recurrent severe hypoglycaemia (MacLeod et al., 1994a) with, <strong>in</strong> particular,a relative <strong>in</strong>crease <strong>in</strong> blood flow to the frontal lobes. This may represent a chronic adaptive

IMPAIRED AWARENESS OF HYPOGLYCAEMIA 161response to protect vulnerable areas of the bra<strong>in</strong> from recurrent, severe neuroglycopenia.However, a further study showed that the changes <strong>in</strong> regional cerebral blood flow <strong>in</strong> responseto controlled hypoglycaemia <strong>in</strong> patients with type 1 diabetes occurred <strong>in</strong>dependently of thestate of awareness of hypoglycaemia (MacLeod et al., 1996). The EEG changes associatedwith modest hypoglycaemia are more pronounced <strong>in</strong> patients with type 1 diabetes who haveimpaired awareness of hypoglycaemia (Tribl et al. 1996). Most studies suggest that a diffusefunctional abnormality is present <strong>in</strong> the anterior part of the bra<strong>in</strong> <strong>in</strong> diabetes, and this may beimplicated <strong>in</strong> the impaired perception of hypoglycaemia. The pre-frontal areas of the cortexare closely connected to sub-cortical areas, and localised dysfunction could theoreticallyreduce the ability of the bra<strong>in</strong> to perceive symptomatic hypoglycaemia.Further clues may be provided by neuroimag<strong>in</strong>g studies. A study us<strong>in</strong>g positron emissiontomography (PET) compared changes <strong>in</strong> global and regional bra<strong>in</strong> glucose metabolism dur<strong>in</strong>geuglycaemia and hypoglycaemia <strong>in</strong> 12 men with type 1 diabetes, six of whom had impairedawareness of hypoglycaemia (B<strong>in</strong>gham et al., 2005). Bra<strong>in</strong> glucose content was reducedby hypoglycaemia <strong>in</strong> both groups with a relative <strong>in</strong>crease <strong>in</strong> tracer uptake on the prefrontalcortical regions. Differences between the groups were observed <strong>in</strong> glucose handl<strong>in</strong>g <strong>in</strong> regionsof the bra<strong>in</strong>, and whereas the cerebral metabolic rate for glucose showed a relative rise <strong>in</strong> theaware subjects, it fell <strong>in</strong> the unaware subjects. Global neuronal activation was observed withhypoglycaemia <strong>in</strong> the aware patients, but was absent <strong>in</strong> the unaware, suggest<strong>in</strong>g that corticalactivation is a necessary correlate of hypoglycaemia awareness. Further studies us<strong>in</strong>g PETscans or other forms of neuroimag<strong>in</strong>g may identify regional differences <strong>in</strong> the response tohypoglycaemia with<strong>in</strong> the bra<strong>in</strong>s of people with impaired awareness, which will help toelucidate the functional abnormalities associated with this syndrome.IMPAIRED AWARENESS OF HYPOGLYCAEMIA ANDLONG-TERM EFFECT ON COGNITIVE FUNCTIONImpaired awareness of hypoglycaemia is a major risk factor for severe hypoglycaemia, andpatients with the chronic form of this condition have a six-fold higher frequency (Gold et al.,1994). It is possible therefore that impaired hypoglycaemia awareness may be associatedwith evidence of a decl<strong>in</strong>e <strong>in</strong> cognitive function. Hepburn et al. (1991) noted that diabeticpatients with a history of impaired awareness of hypoglycaemia performed less well thanthose with normal awareness of hypoglycaemia on limited cognitive function test<strong>in</strong>g, bothat a normal blood glucose and dur<strong>in</strong>g hypoglycaemia. This suggested that an acquiredcognitive impairment may have been superimposed upon an <strong>in</strong>creased susceptibility toneuroglycopenia. A modest, but <strong>in</strong>significant, decl<strong>in</strong>e <strong>in</strong> <strong>in</strong>tellectual function was notedwith progressive loss of hypoglycaemia awareness <strong>in</strong> a population study (MacLeod et al.,1994b).Formal measurement of cognitive function dur<strong>in</strong>g controlled hypoglycaemia (bloodglucose 2.5 mmol/l) showed that patients with type 1 diabetes who had impaired hypoglycaemiaawareness exhibited more profound cognitive dysfunction dur<strong>in</strong>g acute hypoglycaemiathan patients with normal awareness, and that this persisted for longer follow<strong>in</strong>grecovery of blood glucose (Gold et al., 1995b). By contrast, a more recent study of peoplewith type 1 diabetes, which exam<strong>in</strong>ed the rate of recovery of differ<strong>in</strong>g doma<strong>in</strong>s of cognitivefunction after hypoglycaemia, showed that dur<strong>in</strong>g hypoglycaemia (blood glucose 2.5 mmol/l)cognitive function did not deteriorate <strong>in</strong> those with impaired awareness, suggest<strong>in</strong>g that

162 IMPAIRED AWARENESS OF HYPOGLYCAEMIAcerebral adaptation to hypoglycaemia had occurred <strong>in</strong> these patients (Zammitt et al., 2005).These apparently discrepant results leave this issue unresolved.HUMAN INSULINFor 60 years after its discovery, <strong>in</strong>sul<strong>in</strong> for therapeutic use was obta<strong>in</strong>ed from the pancreataof cattle and pigs. With the development of recomb<strong>in</strong>ant DNA technology it was possibleto ‘genetically eng<strong>in</strong>eer’ molecules and <strong>in</strong>sul<strong>in</strong> was the first prote<strong>in</strong> to be made <strong>in</strong> thisway, becom<strong>in</strong>g available for the treatment of humans <strong>in</strong> the 1980s. Several of the exist<strong>in</strong>ganimal <strong>in</strong>sul<strong>in</strong> formulations were withdrawn, pr<strong>in</strong>cipally for commercial reasons, and human<strong>in</strong>sul<strong>in</strong> rapidly became the most commonly prescribed form of <strong>in</strong>sul<strong>in</strong>. The structure ofhuman <strong>in</strong>sul<strong>in</strong> differs from porc<strong>in</strong>e <strong>in</strong>sul<strong>in</strong> by a s<strong>in</strong>gle am<strong>in</strong>o acid and from bov<strong>in</strong>e <strong>in</strong>sul<strong>in</strong>by three am<strong>in</strong>o acids. In <strong>in</strong>itial trials it was not expected that human <strong>in</strong>sul<strong>in</strong> would differsubstantially <strong>in</strong> potency from animal <strong>in</strong>sul<strong>in</strong>, but because human <strong>in</strong>sul<strong>in</strong> was slightly purerthan some of the animal <strong>in</strong>sul<strong>in</strong>s, patients were advised to reduce the dose by around 10%when convert<strong>in</strong>g from animal to human <strong>in</strong>sul<strong>in</strong>. Detailed pharmacok<strong>in</strong>etic studies compar<strong>in</strong>ghuman and animal <strong>in</strong>sul<strong>in</strong>s did not demonstrate any major differences, but human <strong>in</strong>sul<strong>in</strong> hasa slightly faster onset of action, a slightly shorter duration of action, and is less immunogenicthan equivalent animal <strong>in</strong>sul<strong>in</strong>s. Most cl<strong>in</strong>ical studies, conducted on a worldwide scale,showed no significant differences between human and animal <strong>in</strong>sul<strong>in</strong>s <strong>in</strong> their cl<strong>in</strong>icalapplication.In Switzerland, however, one group of cl<strong>in</strong>icians reported encounter<strong>in</strong>g serious cl<strong>in</strong>icalproblems with the use of human <strong>in</strong>sul<strong>in</strong> <strong>in</strong> patients with type 1 diabetes (Teuscher and Berger,1987). In particular, they claimed that patients experienced more frequent hypoglycaemiawith human <strong>in</strong>sul<strong>in</strong>, and that warn<strong>in</strong>g symptoms were modified by human <strong>in</strong>sul<strong>in</strong>, as a resultof which many patients were unable to detect the onset of hypoglycaemia. A pathologist <strong>in</strong> theUnited K<strong>in</strong>gdom then claimed that the number of patients dy<strong>in</strong>g from severe hypoglycaemiahad <strong>in</strong>creased s<strong>in</strong>ce the <strong>in</strong>troduction of human <strong>in</strong>sul<strong>in</strong> (see Chapter 12). The evidence forthis irresponsible statement did not withstand scrut<strong>in</strong>y, but <strong>in</strong> the UK anecdotal reportsemerged of problems experienced by patients with human <strong>in</strong>sul<strong>in</strong>, and solicitors act<strong>in</strong>g onbehalf of over 400 patients tried to br<strong>in</strong>g a legal action aga<strong>in</strong>st the <strong>in</strong>sul<strong>in</strong> manufacturers,alleg<strong>in</strong>g that human <strong>in</strong>sul<strong>in</strong> gave less warn<strong>in</strong>g of hypoglycaemia. Additional claims <strong>in</strong>cludedallegations that human <strong>in</strong>sul<strong>in</strong> may have caused personality changes <strong>in</strong> <strong>in</strong>dividuals and evenother disease states such as multiple sclerosis. This group action was abandoned <strong>in</strong> 1993because of the lack of robust scientific evidence for these claims.However, this issue generated much controversy and heated debate and stimulated severalstudies compar<strong>in</strong>g human with animal <strong>in</strong>sul<strong>in</strong>s, which are not reviewed here. A systematicreview of the extensive literature on this topic exam<strong>in</strong>ed whether published evidencesupported a difference <strong>in</strong> the frequency and awareness of hypoglycaemia <strong>in</strong>duced by humanand animal <strong>in</strong>sul<strong>in</strong>s (Airey et al., 2000). A total of 52 randomized controlled trials wereidentified, 37 of which were of double-bl<strong>in</strong>d design, whereas others reported hypoglycemicoutcomes as a secondary or <strong>in</strong>cidental outcome dur<strong>in</strong>g comparative <strong>in</strong>vestigations of efficacyor immunogenicity. Seven of the double-bl<strong>in</strong>d studies reported differences <strong>in</strong> frequency ofhypoglycaemia or of symptomatic awareness, and four of the unbl<strong>in</strong>ded trials reported differences<strong>in</strong> hypoglycaemia. None of the four population time trend studies found any relationshipbetween the <strong>in</strong>creas<strong>in</strong>g use of human <strong>in</strong>sul<strong>in</strong> and hospital admission for hypoglycaemia

TREATMENT STRATEGIES 163or unexpla<strong>in</strong>ed death among people with <strong>in</strong>sul<strong>in</strong>-treated diabetes. The authors observed thatthe least rigorous scientific studies gave the greatest support to the premise that treatmentwith human <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fluences the frequency, severity or symptoms of hypoglycaemia. Thereport concluded that the published evidence did not support the contention that human<strong>in</strong>sul<strong>in</strong> is responsible either for the alleged problems with impaired awareness of hypoglycaemiaor for a higher risk of severe hypoglycaemia, but were unable to decide whether someof the phenomena associated with the use of human <strong>in</strong>sul<strong>in</strong> resulted from stricter glycaemiccontrol (Airey et al., 2000).In cl<strong>in</strong>ical practice there are undoubtedly a small number of people with <strong>in</strong>sul<strong>in</strong>-treateddiabetes <strong>in</strong> whom the use of human <strong>in</strong>sul<strong>in</strong> has not been satisfactory, be<strong>in</strong>g associated withfrequent and unpredictable hypoglycaemia and a dim<strong>in</strong>ished sense of well-be<strong>in</strong>g. Whetherthis is related to the different pharmacok<strong>in</strong>etics of human <strong>in</strong>sul<strong>in</strong> or represents an idiosyncraticresponse <strong>in</strong> <strong>in</strong>dividuals is unclear, but such patients clearly wish to reta<strong>in</strong> the option touse animal species of <strong>in</strong>sul<strong>in</strong>, and it is to be hoped that the availability of the animal speciesof <strong>in</strong>sul<strong>in</strong>s will be ma<strong>in</strong>ta<strong>in</strong>ed. No problem with symptomatic awareness of hypoglycaemiahas been reported with the use of the newer <strong>in</strong>sul<strong>in</strong> analogues.TREATMENT STRATEGIESWhen impaired awareness of hypoglycaemia is therapy-related, that is, result<strong>in</strong>g from strictglycaemic control, the approach to management is relatively simple. The total <strong>in</strong>sul<strong>in</strong> doseshould be reduced, attention paid to the appropriateness of the <strong>in</strong>sul<strong>in</strong> regimen, and overallglycaemic control should be relaxed. Liu et al. (1996) reported an improvement <strong>in</strong> symptomaticand counterregulatory hormonal responses to hypoglycaemia after three months ofless strict glycaemic control <strong>in</strong> a small group of <strong>in</strong>sul<strong>in</strong>-treated patients, <strong>in</strong> whom the meanHbA 1c rose from 6.9% to 8.0%.It has been claimed that impaired awareness of hypoglycaemia (and to some extent counterregulatoryhormonal deficiency) can be reversed by scrupulous avoidance of hypoglycaemiathrough meticulous attention to diabetic management (Cranston et al., 1994; Dagogo-Jacket al., 1994; Fanelli et al., 1994). The effect that this had on glycaemic thresholds for cognitivedysfunction and the recovery of counterregulatory hormonal secretion to hypoglycaemiadiffered between these studies, but all demonstrated an improved symptomatic responsefollow<strong>in</strong>g avoidance of hypoglycaemia for periods vary<strong>in</strong>g from three weeks to one year.However, the studies can be criticised for the follow<strong>in</strong>g reasons:• Only a small number of patients were studied.• The def<strong>in</strong>ition of hypoglycaemia unawareness was based on an <strong>in</strong>creased frequency ofasymptomatic biochemical hypoglycaemia, and with the exception of the study by Dagogo-Jack et al. (1994), was not based on hav<strong>in</strong>g a history of hypoglycaemia unawareness.• In all studies there was a small but def<strong>in</strong>ite rise <strong>in</strong> glycated haemoglob<strong>in</strong>, suggest<strong>in</strong>g thatthe improved symptomatic awareness was related primarily to relaxation of glycaemiccontrol.Although the scrupulous avoidance of hypoglycaemia is clearly desirable, and may bebeneficial to reduc<strong>in</strong>g the severity of hypoglycaemia unawareness, it is very difficult to

164 IMPAIRED AWARENESS OF HYPOGLYCAEMIAFigure 7.9 Augmentation of the normal secretory response of ep<strong>in</strong>ephr<strong>in</strong>e (adrenal<strong>in</strong>e) to, andawareness of, acute hypoglycaemia (blood glucose 2.8 mmol/l) by the prior <strong>in</strong>gestion of caffe<strong>in</strong>e <strong>in</strong><strong>in</strong>sul<strong>in</strong>-treated diabetic patients. Derived from data <strong>in</strong> Debrah et al. (1996)achieve as it is extremely time-consum<strong>in</strong>g and labour-<strong>in</strong>tensive both for patients and healthprofessionals. The use of cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion overnight <strong>in</strong>stead ofisophane (NPH) <strong>in</strong>sul<strong>in</strong> at bedtime has been shown to be beneficial <strong>in</strong> diabetic patientswith impaired awareness of hypoglycaemia, improv<strong>in</strong>g warn<strong>in</strong>g symptoms and counterregulatoryresponses to hypoglycaemia, presumably by reduc<strong>in</strong>g the frequency of nocturnalhypoglycaemia (Kanc et al., 1998).The <strong>in</strong>gestion of caffe<strong>in</strong>e uncouples the relationship between cerebral blood flow andglucose utilisation via antagonism of adenos<strong>in</strong>e receptors, caus<strong>in</strong>g relative neuroglycopeniaand earlier release of counterregulatory hormones dur<strong>in</strong>g moderate hypoglycaemia. Theprior consumption of caffe<strong>in</strong>e augments the symptomatic and counterregulatory hormonalresponses to a modest reduction of blood glucose <strong>in</strong> non-diabetic subjects (Kerr et al., 1993),and a similar phenomenon occurs <strong>in</strong> people with type 1 diabetes follow<strong>in</strong>g the <strong>in</strong>gestion of adose of caffe<strong>in</strong>e equivalent to two or three cups of coffee (Debrah et al., 1996). The reduction<strong>in</strong> cerebral blood flow is susta<strong>in</strong>ed, the counterregulatory response is augmented (Figure 7.9)and greater awareness of hypoglycaemia occurs (see Chapter 5). This raises the prospectof identify<strong>in</strong>g some form of therapeutic <strong>in</strong>tervention, which utilises a similar mechanism toheighten the residual symptomatic response <strong>in</strong> people with type 1 diabetes who have impairedawareness of hypoglycaemia. The adenos<strong>in</strong>e-receptor antagonist, theophyll<strong>in</strong>e, stimulatesthe secretion of catecholam<strong>in</strong>es and reduces cerebral blood flow, and a s<strong>in</strong>gle <strong>in</strong>travenousdose has been shown to enhance counterregulatory hormone responses to hypoglycaemiaand partially restore perception of hypoglycaemic symptoms <strong>in</strong> patients with type 1 diabeteswith impaired awareness of hypoglycaemia (de Galan et al., 2002). Glycaemic thresholdsfor haemodynamic and symptomatic responses were restored to normal. It is not knownwhether oral theophyll<strong>in</strong>e would be as effective, and whether the effects can be susta<strong>in</strong>ed,as the development of tolerance to these drugs is common.It is clearly desirable to avoid severe hypoglycaemia at all costs, and treatment strategiesshould be adopted to achieve this aim (Box 7.5). Frequent blood glucose monitor<strong>in</strong>g isessential <strong>in</strong> affected patients, and may require occasional nocturnal measurements to detect

CONCLUSIONS 165Box 7.5 Treatment strategies for patients with impaired awareness ofhypoglycaemia• Frequent blood glucose monitor<strong>in</strong>g (<strong>in</strong>clud<strong>in</strong>g nocturnal measurements).• Avoid blood glucose values < 40 mmol/l.• Set target range of blood glucose higher than for ‘aware’ patients (e.g. preprandialbetween 6.0–12.0 mmol/l; bedtime > 80 mmol/l)• Avoid HbA 1c <strong>in</strong> non-diabetic range.• Use predom<strong>in</strong>antly short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>s (e.g. basal-bolus regimen; <strong>in</strong>sul<strong>in</strong>analogues).• Regular snacks between meals and at bedtime, conta<strong>in</strong><strong>in</strong>g unref<strong>in</strong>ed carbohydrate.• Appropriate additional carbohydrate consumption and/or <strong>in</strong>sul<strong>in</strong> dose adjustmentfor premeditated exercise.• Learn to identify subtle neuroglycopenic cues to low blood glucose.low blood glucose dur<strong>in</strong>g the night. Blood glucose awareness tra<strong>in</strong><strong>in</strong>g has been developed<strong>in</strong> the USA, with re-education of affected patients to recognise neuroglycopenic cues (Coxet al., 1995), but this also requires facilities and resources that are not available <strong>in</strong> mostcentres. Intensive <strong>in</strong>sul<strong>in</strong> therapy is contra<strong>in</strong>dicated <strong>in</strong> patients who have impaired awarenessof hypoglycaemia and treatment goals have to be considered <strong>in</strong>dividually. The avoidanceof severe hypoglycaemia is paramount as this may exacerbate the problem, and the use ofmostly short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> (and possibly <strong>in</strong>sul<strong>in</strong> analogues) <strong>in</strong> basal-bolus regimens maybe particularly useful <strong>in</strong> avoid<strong>in</strong>g biochemical and symptomatic hypoglycaemia withoutcompromis<strong>in</strong>g overall glycaemic control.CONCLUSIONS• An <strong>in</strong>adequate symptomatic warn<strong>in</strong>g to hypoglycaemia is common <strong>in</strong> people with<strong>in</strong>sul<strong>in</strong>-treated diabetes and is described as impaired awareness of hypoglycaemia orhypoglycaemia unawareness. It <strong>in</strong>creases <strong>in</strong> prevalence with duration of <strong>in</strong>sul<strong>in</strong>-treateddiabetes.• In people who report impaired awareness of hypoglycaemia, asymptomatic hypoglycaemiaoccurs more frequently dur<strong>in</strong>g rout<strong>in</strong>e blood glucose monitor<strong>in</strong>g. This may alert thecl<strong>in</strong>ician to the possibility that an <strong>in</strong>dividual is develop<strong>in</strong>g this problem.• Impaired awareness of hypoglycaemia may be associated with strict glycaemic control;significant modification of the symptomatic response occurs when the HbA 1c concentrationis with<strong>in</strong> the non-diabetic range.

166 IMPAIRED AWARENESS OF HYPOGLYCAEMIA• The mechanisms underly<strong>in</strong>g impaired awareness of hypoglycaemia may be multifactorial.Possible mechanisms <strong>in</strong>clude chronic exposure to a low blood glucose, antecedenthypoglycaemia, and central autonomic and glucoregulatory failure.• Antecedent hypoglycaemia has a significant <strong>in</strong>fluence on the magnitude of the symptomaticand counterregulatory responses to subsequent hypoglycaemia occurr<strong>in</strong>g with<strong>in</strong>the follow<strong>in</strong>g 48 hours.• When impaired awareness of hypoglycaemia results from strict glycaemic control, thetotal <strong>in</strong>sul<strong>in</strong> dose should be reduced, attention paid to the suitability of the <strong>in</strong>sul<strong>in</strong> regimen,and overall glycaemic control should be relaxed.• Impaired awareness of hypoglycaemia, and to some extent counterregulatory hormonaldeficiency, can probably be reversed by scrupulous avoidance of hypoglycaemia throughmeticulous attention to diabetic management.• Intensive <strong>in</strong>sul<strong>in</strong> therapy is contra<strong>in</strong>dicated <strong>in</strong> patients who have impaired awareness ofhypoglycaemia. The use of mostly short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>s (<strong>in</strong>clud<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues) <strong>in</strong>basal-bolus regimens may be particularly useful <strong>in</strong> avoid<strong>in</strong>g biochemical and symptomatichypoglycaemia without compromis<strong>in</strong>g overall glycaemic control.REFERENCESAirey CM, Williams DRR, Mart<strong>in</strong> PG, Bennett, CMT, Spoor PA (2000). <strong>Hypoglycaemia</strong> <strong>in</strong>duced byexogenous <strong>in</strong>sul<strong>in</strong> – ‘human’ and animal <strong>in</strong>sul<strong>in</strong> compared. Diabetic Medic<strong>in</strong>e 17: 416–32.Bacatselos SO, Karamitsos DT, Kourtoglou GI, Zambulis CX, Yovos JG, Vyzantiadis AT (1995).<strong>Hypoglycaemia</strong> unawareness <strong>in</strong> type 1 diabetic patients under conventional <strong>in</strong>sul<strong>in</strong> treatment.<strong>Diabetes</strong>, Nutrition and Metabolism 8: 267–75.Berl<strong>in</strong> I, Grimaldi A, Payan C, Sachon C, Bosquet F, Thervet F, Puech AJ (1987). Hypoglycemicsymptoms and decreased B-adrenergic sensitivity <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetic patients. <strong>Diabetes</strong>Care 10: 742–7.B<strong>in</strong>gham EM, Dunn JT, Smith D, Sutcliffe-Goulden J, Reed LJ, Marsden PK, Amiel SA (2005).Differential changes <strong>in</strong> bra<strong>in</strong> glucose metabolism dur<strong>in</strong>g hypoglycaemia accompany loss of hypoglycaemiaawareness <strong>in</strong> men with type 1 diabetes mellitus. An ( 11 C)-3-O-methyl-D-glucose PETstudy. Diabetologia 48: 2080–9.Bjork E, Palmer M, Schvarcz E, Berne C (1990). Incidence of severe hypoglycaemia <strong>in</strong> an unselectedpopulation of patients with <strong>in</strong>sul<strong>in</strong>-treated diabetes mellitus, with special reference to autonomicneuropathy. <strong>Diabetes</strong>, Nutrition and Metabolism 4: 303–9.Bott<strong>in</strong>i P, Boschetti E, Pampanelli S, Ciofetta M, Del S<strong>in</strong>daco P, Scionti L et al. (1997). Contribution ofautonomic neuropathy to reduced plasma adrenal<strong>in</strong>e responses to hypoglycemia <strong>in</strong> IDDM. Evidencefor a nonselective defect. <strong>Diabetes</strong> 46: 814–23.Boyle PJ, Schwartz NS, Shah SD, Clutter WE, Cryer PE (1988). Plasma glucose concentrations at theonset of hypoglycemic symptoms <strong>in</strong> patients with poorly controlled diabetes and <strong>in</strong> non-diabetics.New England Journal of Medic<strong>in</strong>e 318: 1487–92.Boyle PJ, Nagy RJ, O’Connor AM, Kempers SF, Yeo RA, Qualls C (1994). Adaptation <strong>in</strong> bra<strong>in</strong>glucose uptake follow<strong>in</strong>g recurrent hypoglycemia. Proceed<strong>in</strong>gs of National Academy of Science,USA 91: 9352–6.Boyle PJ, Kempers SF, O’Connor AM, Nagy RJ (1995). Bra<strong>in</strong> glucose uptake and unawarenessof hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 333: 1726–31.

REFERENCES 167Boyle PJ (1997). Alteration <strong>in</strong> bra<strong>in</strong> glucose metabolism <strong>in</strong>duced by hypoglycaemia <strong>in</strong> man,Diabetologia, 40, Suppl. S69–S74.Clarke WL, Cox DJ, Gonder-Frederick LA, Julian D, Schlundt D, Polonsky W (1995). Reducedawareness of hypoglycemia <strong>in</strong> adults with IDDM. A prospective study of hypoglycemic frequencyand associated symptoms. <strong>Diabetes</strong> Care 18: 517–22.Cox D, Gonder-Frederick L, Polonsky W, Schlundt D, Julian D, Clarke W (1995). A multicenterevaluation of blood glucose awareness tra<strong>in</strong><strong>in</strong>g – II. <strong>Diabetes</strong> Care 18: 523–8.Cranston I, Lomas J, Maran A, Macdonald I, Amiel SA (1994). Restoration of hypoglycaemia awareness<strong>in</strong> patients with long-duration <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 344: 283–7.Cryer PE (1992). Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure<strong>in</strong> IDDM. A vicious circle. <strong>Diabetes</strong> 41: 255–60.Cryer P and Frier BM (2004). <strong>Hypoglycaemia</strong>, International Textbook of <strong>Diabetes</strong>, 3rd edn,Ferrann<strong>in</strong>i E., Zimmet P., De Fronzo R. and Keen H. (eds), 1079–100.Cryer PE (2005). Mechanisms of hypoglycemia-associated autonomic failure and its componentsyndromes <strong>in</strong> diabetes. <strong>Diabetes</strong> 54: 3592–601.Dagogo-Jack SE, Craft S, Cryer PE (1993). Hypoglycemia-associated autonomic failure <strong>in</strong> <strong>in</strong>sul<strong>in</strong>dependentdiabetes mellitus. Journal of Cl<strong>in</strong>ical Investigation 91: 819–28.Dagogo-Jack S, Rattarasarn C, Cryer PE (1994). Reversal of hypoglycemia unawareness, but notdefective glucose counterregulation, <strong>in</strong> IDDM. <strong>Diabetes</strong> 43: 1426–34.Davis MR, Mellman M, Shamoon H (1992). Further defects <strong>in</strong> counterregulatory responses <strong>in</strong>ducedby recurrent hypoglycemia <strong>in</strong> IDDM. <strong>Diabetes</strong> 41: 1335–40.Davis SN, Shevers C, Mosqueda-Garcia R, Costa F (1997). Effects of differ<strong>in</strong>g antecedent hypoglycemiaon subsequent counterregulation <strong>in</strong> normal humans. <strong>Diabetes</strong> 46: 1328–35.Davis SN, Mann S, Galassetti P, Neill RA, Tate D, Ertl AC, Costa F (2000a). Effects of differ<strong>in</strong>gdurations of antecedent hypoglycemia on counterregulatory responses to subsequent hypoglycemia<strong>in</strong> normal humans. <strong>Diabetes</strong> 49: 1897–903.Davis SN, Galassetti P, Wasserman DH, Tate D (2000b). Effects of antecedent hypoglycemia onsubsequent counterregulatory responses to exercise. <strong>Diabetes</strong> 49: 73–81.de Galan BE, Tack CJ, Lenders JW, Pasman JW, Elv<strong>in</strong>g LD, Russel FG et al. (2002). Theophyll<strong>in</strong>eimproves hypoglycemia unawareness <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> 51: 790–6.Debrah K, Sherw<strong>in</strong> RS, Murphy J, Kerr D (1996). Effect of caffe<strong>in</strong>e on recognition of and physiologicalresponses to hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dependent diabetes. Lancet 347: 19–24.Fanelli C, Pampanelli S, Epifano L, Rambotti AM, Di V<strong>in</strong>cenzo A, Modarelli F et al. (1994). Long-termrecovery from unawareness, deficient counterregulation and lack of cognitive dysfunction dur<strong>in</strong>ghypoglycaemia, follow<strong>in</strong>g <strong>in</strong>stitution of rational, <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> IDDM. Diabetologia37: 1265–76.Fanelli CG, Paramore DS, Hershey T, Terkamp C, Ovalle F, Craft S, Cryer PE (1998). Impact of nocturnalhypoglycemia on hypoglycemic cognitive dysfunction <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> 47: 1920–7.Fritsche A, Stumvoll M, Grub M, Sieslack S, Renn W, Schmull<strong>in</strong>g RM et al. (1998). Effect ofhypoglycemia on beta-adrenergic sensitivity <strong>in</strong> normal and type 1 diabetic subjects. <strong>Diabetes</strong> Care21: 1505–10.Fritsche A, Stefan N, Har<strong>in</strong>g H, Gerich J, Stumvoll M (2001). Avoidance of hypoglycemia awarenessby <strong>in</strong>creas<strong>in</strong>g beta-adrenergic sensitivity <strong>in</strong> type 1 diabetes. Annals of Internal Medic<strong>in</strong>e 134: 729–36.Fruehwald-Schultes B, Born J, Kern W, Peters A, Fehm HL (2000). Adaptation of cognitive functionto hypoglycemia <strong>in</strong> healthy men. <strong>Diabetes</strong> Care 23: 1059–66.Galassetti P, Tate D, Neill RA, Morrey S, Wasserman DH, Davis SN (2003). Effect of antecedenthypoglycemia on counterregulatory responses to subsequent euglycemic exercise <strong>in</strong> type 1 diabetes.<strong>Diabetes</strong> 52: 1761–9.George E, Harris N, Bedford C, Macdonald IA, Hardisty CA, Heller SR (1995). Prolonged but partialimpairment of the hypoglycaemic physiological response follow<strong>in</strong>g short-term hypoglycaemia <strong>in</strong>normal subjects. Diabetologia 38: 1183–90.

168 IMPAIRED AWARENESS OF HYPOGLYCAEMIAGeorge E, Marques JL, Harris ND, Macdonald IA, Hardisty CA, Heller SR (1997). Preservation ofphysiological responses to hypoglycemia 2 days after antecedent hypoglycemia <strong>in</strong> patients withIDDM. <strong>Diabetes</strong> Care 20: 1293–8.Gold AE, MacLeod KM, Frier BM (1994). Frequency of severe hypoglycemia <strong>in</strong> patients with type 1diabetes with impaired awareness of hypoglycemia. <strong>Diabetes</strong> Care 17: 697–703.Gold AE, Deary IJ, MacLeod KM, Thomson KJ, Frier BM (1995a). Cognitive function dur<strong>in</strong>g <strong>in</strong>sul<strong>in</strong><strong>in</strong>ducedhypoglycemia <strong>in</strong> humans: short-term cerebral adaptation does not occur. Psychopharmacology119: 325–33.Gold AE, MacLeod KM, Deary IJ, Frier BM (1995b). Hypoglycemia-<strong>in</strong>duced cognitive dysfunction<strong>in</strong> diabetes mellitus: effect of hypoglycemia unawareness. Physiology and Behavior 58: 501–11.Grimaldi A, Bosquet F, Davidoff P, Digy JP, Sachon C, Landault C et al. (1990). Unawareness ofhypoglycemia by <strong>in</strong>sul<strong>in</strong>-dependent diabetics. Hormone and Metabolic Research 22: 90–5.Heller SR, Cryer PE (1991). Reduced neuroendocr<strong>in</strong>e and symptomatic responses to subsequenthypoglycemia after 1 episode of hypoglycemia <strong>in</strong> non-diabetic humans. <strong>Diabetes</strong> 40: 223–6.Heller, SR, Chapman J, McCloud J, Ward J (1995). Unreliability of reports of hypoglycaemia bydiabetic patients. British Diabetic Journal 310: 440.Henderson JN, Allen KV, Deary IJ, Frier BM (2003). <strong>Hypoglycaemia</strong> <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes:frequency, symptoms and impaired awareness. Diabetic Medic<strong>in</strong>e 20: 1016–21.Hepburn DA, Patrick AW, Ead<strong>in</strong>gton DW, Ew<strong>in</strong>g DJ, Frier BM (1990). Unawareness of hypoglycaemia<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients: prevalence and relationship to autonomic neuropathy. DiabeticMedic<strong>in</strong>e 7: 711–7.Hepburn DA, Patrick AW, Brash HM. Thomson I, Frier BM (1991). <strong>Hypoglycaemia</strong> unawareness<strong>in</strong> type 1 diabetes: a lower plasma glucose is required to stimulate sympatho-adrenal activation.Diabetic Medic<strong>in</strong>e 8: 934–45.Hepburn DA, Deary IJ, Frier BM (1992). Classification of symptoms of hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>treateddiabetic patients us<strong>in</strong>g factor analysis: relationship to hypoglycaemia unawareness. DiabeticMedic<strong>in</strong>e 9: 70–5.Hepburn DA, MacLeod KM, Pell ACH, Scougal IJ, Frier BM (1993a). Frequency and symptoms ofhypoglycaemia experienced by patients with type 2 diabetes treated with <strong>in</strong>sul<strong>in</strong>. Diabetic Medic<strong>in</strong>e10: 231–7.Hepburn DA, MacLeod KM, Frier BM (1993b). Physiological, symptomatic and hormonal responsesto acute hypoglycaemia <strong>in</strong> type 1 diabetic patients with autonomic neuropathy. Diabetic Medic<strong>in</strong>e10: 940–9.Hilsted J, Madsbad S, Krarup T, Sestoft L, Christensen NJ, Tronier B, Galbo H (1981). Hormonal,metabolic, and cardiovascular responses to hypoglycemia <strong>in</strong> diabetic autonomic neuropathy.<strong>Diabetes</strong> 30: 626–33.Hilsted J, Richter E, Madsbad S, Tronier B, Christensen NJ, Hildebrandt P et al. (1987). Metabolic andcardiovascular responses to ep<strong>in</strong>ephr<strong>in</strong>e <strong>in</strong> diabetic autonomic neuropathy. New England Journal ofMedic<strong>in</strong>e 317: 421–6.Hoeldtke RD, Boden G, Shuman CR, Owen OE (1982). Reduced ep<strong>in</strong>ephr<strong>in</strong>e secretion and hypoglycemiaunawareness <strong>in</strong> diabetic autonomic neuropathy. Annals of Internal Medic<strong>in</strong>e 96: 459–62.Jorgensen HV, Pedersen-Bjergaard U, Rasmussen AK, Borch-Johnsen K (2003). The impact of severehypoglycemia and impaired awareness of hypoglycemia on relatives of patients with type 1 diabetes.<strong>Diabetes</strong> Care 26: 1106–9.Josl<strong>in</strong> EP, Gray H, Root HF. (1922). Insul<strong>in</strong> <strong>in</strong> hospital and home. Journal of Metabolic Research2: 651–99.Kanc K, Janssen MMJ, Keulen ETP, Jacobs MAJM, Popp-Snijders C, Snoek FJ, He<strong>in</strong>e RJ (1998).Substitution of night-time cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion therapy for bedtime NPH <strong>in</strong>sul<strong>in</strong><strong>in</strong> a multiple <strong>in</strong>jection regimen improves counterregulatory hormonal responses and warn<strong>in</strong>g symptomsof hypoglycaemia <strong>in</strong> IDDM. Diabetologia 41: 322–9.

REFERENCES 169Kerr D, Sherw<strong>in</strong> RS, Pavalkis F, Fayad P, Sikorski L, Rife F et al. (1993). Effect of caffe<strong>in</strong>e on therecognition of and responses to hypoglycemia <strong>in</strong> humans. Annals of Internal Medic<strong>in</strong>e 119: 799–804.K<strong>in</strong>sley BT, Widom B, Simonson DC (1995). Differential regulation of counterregulatory hormonesecretion and symptoms dur<strong>in</strong>g hypoglycemia <strong>in</strong> IDDM. Effect of glycemic control. <strong>Diabetes</strong> Care18: 17–26.Kubiak T, Hermanns HJ, Schreckl<strong>in</strong>g HJ, Kulzer B, Haak T (2004). Assessment of hypoglycaemiaawareness us<strong>in</strong>g cont<strong>in</strong>uous glucose monitor<strong>in</strong>g. Diabetic Medic<strong>in</strong>e 21: 487–90.Kumagai AK, Kang Y-S, Boado RJ, Pardridge WM (1995). Upregulation of blood–bra<strong>in</strong> barrierGLUT 1 glucose transporter prote<strong>in</strong> and mRNA <strong>in</strong> experimental chronic hypoglycemia. <strong>Diabetes</strong>44: 1399–404.Lawrence RD (1941). Insul<strong>in</strong> hypoglycaemia. Changes <strong>in</strong> nervous manifestations. Lancet ii: 602.L<strong>in</strong>genfelser T, Renn W, Sommerwerck U, Jung MF, Buettner UW, Zaiser-Kaschel H et al. (1993).Compromised hormonal counterregulation, symptom awareness, and neurophysiological functionafter recurrent short-term episodes of <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> IDDM patients. <strong>Diabetes</strong>42: 610–8.Liu D, McManus RM, Ryan EA (1996). Improved counter-regulatory hormonal and symptomaticresponses to hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus after 3 months ofless strict glycemic control. Cl<strong>in</strong>ical and Investigative Medic<strong>in</strong>e 19: 71–82.MacLeod KM, Hepburn DA, Deary IJ, Goodw<strong>in</strong> GM, Dougall N, Ebmeier KP, Frier BM (1994a).Regional cerebral blood flow <strong>in</strong> IDDM patients: effects of diabetes and of recurrent severe hypoglycaemia.Diabetologia 37: 257–63.MacLeod KM, Deary IJ, Graham KS, Hepburn DA, Frier BM (1994b). <strong>Hypoglycaemia</strong> unawareness <strong>in</strong>adult patients with type 1 diabetes: relationship to severe hypoglycaemia and cognitive impairment.<strong>Diabetes</strong>, Nutrition and Metabolism 7: 205–12.MacLeod KM, Gold AE, Ebmeier KP, Hepburn DA, Deary, IJ, Goodw<strong>in</strong> GM, Frier BM (1996). Theeffects of acute hypoglycemia on relative cerebral blood flow distribution <strong>in</strong> patients with type 1(<strong>in</strong>sul<strong>in</strong>-dependent) diabetes and impaired hypoglycemia awareness. Metabolism 45:974–80.McCall AL, Fixman LB, Flem<strong>in</strong>g N, Tornheim K, Chick W, Ruderman NB (1986). Chronic hypoglycaemia<strong>in</strong>creases bra<strong>in</strong> glucose transport. American Journal of Endocr<strong>in</strong>ology 251: E442–7.Maddock RK, Krall LP (1953). Insul<strong>in</strong> reactions. Manifestations and need for recognition of longact<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> reactions. Archives of Internal Medic<strong>in</strong>e 91: 695–703.Mellman MJ, Davis MR, Brisman M, Shamoon H (1994). Effect of antecedent hypoglycemia oncognitive function and on glycemic thresholds for counterregulatory hormone secretion <strong>in</strong> healthyhumans. <strong>Diabetes</strong> Care 17: 183–8.Mitrakou A, Ryan C, Veneman T, Mokan M, Jenssen T, Kiss I et al. (1991). Hierarchy of glycemicthresholds for counterregulatory hormone secretion, symptoms and cerebral dysfunction. AmericanJournal of Endocr<strong>in</strong>ology 260: E67–74.Mitrakou A, Fanelli C, Veneman T, Perriello G, Calderone S, Platanisiotis D et al. (1993). Reversibilityof unawareness of hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>omas. New England Journal of Medic<strong>in</strong>e329: 834–9.Mokan M, Mitrakou A, Veneman T, Ryan C, Korytkowski M, Cryer P, Gerich J (1994). Hypoglycemiaunawareness <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 17: 1397–403.Muhlhauser I, Heirnemann L, Fritsche E, von Lennep K, Berger M (1991). Hypoglycemic symptomsand frequency of severe hypoglycemia <strong>in</strong> patients treated with human and animal <strong>in</strong>sul<strong>in</strong>preparations. <strong>Diabetes</strong> Care 14: 745–9.Orchard TJ, Maser RE, Becker DJ, Dorman JS, Drash AL (1991). Human <strong>in</strong>sul<strong>in</strong> use and hypoglycaemia:<strong>in</strong>sights from the Pittsburgh Epidemiology of <strong>Diabetes</strong> Complications Study. DiabeticMedic<strong>in</strong>e 8: 469–74.Ovalle F, Fanelli CG, Poramore DS, Hersley T, Craft S, Cryer PE (1998). Brief twice-weekly episodesof hypoglycemia reduce detection of cl<strong>in</strong>ical hypoglycemia <strong>in</strong> type 1 diabetes mellitus. <strong>Diabetes</strong>47: 1472–9.

170 IMPAIRED AWARENESS OF HYPOGLYCAEMIAPampanelli S, Fanelli C, Lalli C, Ciofetta M, Del S<strong>in</strong>dorco P, Lepore M et al. (1996). Long-term<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> IDDM: effects on HbA 1c , risk for severe and mild hypoglycaemia, statusof counterregulation and awareness of hypoglycaemia. Diabetologia 39: 677–86.Peters A, Rohloff F, Kerner W (1995). Preserved counterregulatory hormone release and symptomsafter short term hypoglycemic episodes <strong>in</strong> normal men. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology andMetabolism 80: 2894–8.Pramm<strong>in</strong>g S, Thorste<strong>in</strong>sson B, Bendtson I, B<strong>in</strong>der C (1991). Symptomatic hypoglycaemia <strong>in</strong> 411 type1 diabetic patients. Diabetic Medic<strong>in</strong>e 8: 217–22.Rob<strong>in</strong>son AM, Park<strong>in</strong> HM, Macdonald IA, Tattersall RB (1995). Antecedent hypoglycaemia <strong>in</strong> nondiabeticsubjects reduces the adrenal<strong>in</strong>e response for 6 days but does not affect the catecholam<strong>in</strong>eresponse to other stimuli. Cl<strong>in</strong>ical Science 89: 359–66.Ryder REJ, Owens DR, Hayes TM, Ghatei MA, Bloom SR (1990). Unawareness of hypoglycaemia and<strong>in</strong>adequate hypoglycaemic counterregulation: no causal relation with diabetic autonomic neuropathy.British Medical Journal 301: 783–7.Sever<strong>in</strong>ghaus EL (1926). Hypoglycemic coma due to repeated <strong>in</strong>sul<strong>in</strong> overdosage. American Journalof Medical Sciences 172: 573–80.Stephenson JM, Kempler P, Cavallo Peria P, Fuller JH, EURODIAB IDDM Complications StudyGroup (1996). Is autonomic neuropathy a risk factor for severe hypoglycaemia? The EURODIABIDDM Complications Study. Diabetologia 39: 1372–6.Sussman KE, Crout JR, Marble A (1963). Failure of warn<strong>in</strong>g <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemic reactions.<strong>Diabetes</strong> 12: 38–45.Teuscher A, Berger WG (1987). <strong>Hypoglycaemia</strong> unawareness <strong>in</strong> diabetics transferred from beef/porc<strong>in</strong>e<strong>in</strong>sul<strong>in</strong> to human <strong>in</strong>sul<strong>in</strong>. Lancet ii: 382–5.The DCCT Research Group (1991). Epidemiology of severe hypoglycemia <strong>in</strong> the <strong>Diabetes</strong> Controland Complications Trial. American Journal of Medic<strong>in</strong>e 90: 450–9.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1993). The effect of <strong>in</strong>tensive treatmentof diabetes on the development and progression of long-term complications <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. The New England Journal of Medic<strong>in</strong>e 329: 977–86.Thorste<strong>in</strong>sson B, Pramm<strong>in</strong>g S, Lauritzen T, B<strong>in</strong>der C (1986). Frequency of daytime biochemical hypoglycaemia<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients: relation to daily median blood glucose concentrations.Diabetic Medic<strong>in</strong>e 3: 147–51.Tribl G, Howorka K, Heger G, Anderer P, Thoma H, Zeitlhofer J (1996). EEG topography dur<strong>in</strong>g<strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. EuropeanNeurology 36: 303–9.Vea H, Jorde R, Sager G, Vaaler S, Sundsfjord J (1992). Reproducibility of glycaemic thresholdsfor activation of counterregulatory hormones and hypoglycaemic symptoms <strong>in</strong> healthy subjects.Diabetologia 35: 958–61.Veneman T, Mitrakou A, Mokan M, Cryer P, Gerich J (1993). Induction of hypoglycemia unawarenessby asymptomatic nocturnal hypoglycemia. <strong>Diabetes</strong> 42: 1233–7.Widom B, Simonson DC (1992). Intermittent hypoglycemia impairs glucose counterregulation.<strong>Diabetes</strong> 41: 1335–40.Zammitt NN, Warren RE, Deary IJ, Frier BM (2005). Recovery of cognitive function after <strong>in</strong>sul<strong>in</strong><strong>in</strong>ducedhypoglycaemia <strong>in</strong> people with type 1 diabetes with either normal or impaired awareness ofhypoglycaemia. Diabetologia 48 (Suppl 1): A293 (abstract).

8 Risks of Strict GlycaemicControlStephanie A. AmielINTRODUCTIONThe benefit of strict glycaemic control <strong>in</strong> dim<strong>in</strong>ish<strong>in</strong>g the risks of the development of longtermcomplications of diabetes is beyond doubt, but the negative aspects of such therapiesneed to be considered, and their risks identified, understood and m<strong>in</strong>imised. Modern <strong>in</strong>tensified<strong>in</strong>sul<strong>in</strong> management need not necessarily <strong>in</strong>crease the risk of iatrogenic problems andcan deliver better glycaemic control more safely than <strong>in</strong> the past, although substantial scoperema<strong>in</strong>s for improvement. New drugs for type 2 diabetes may offer greater opportunities toachieve near-normoglycaemia but may also br<strong>in</strong>g new risks. These risks need to be expla<strong>in</strong>edcarefully to every patient, who can then make an <strong>in</strong>dividual, <strong>in</strong>formed choice about themanagement of their diabetes.The risks of <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy, the focus of this chapter, are those of <strong>in</strong>sul<strong>in</strong>itself – <strong>in</strong>tensified. Thus the major side-effects are weight ga<strong>in</strong> (The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1988) and hypoglycaemia (The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1993; 1995a; 1997). Both of these problems mayappear to be m<strong>in</strong>imised with modern strategies for patient self-management, at least <strong>in</strong>published studies (Jorgens et al., 1993; DAFNE Study Group, 2002; Plank 2004 et al.;Samann et al. 2005), yet they rema<strong>in</strong> serious issues for large numbers of people. Weight ga<strong>in</strong>,attributed primarily to the resolution of caloric loss <strong>in</strong> glycosuria (Carlson and Campbell,2003), is theoretically responsive to dietary strategies, but <strong>in</strong>sul<strong>in</strong> and peripheral <strong>in</strong>sul<strong>in</strong>sensitizers do cause lipogenesis and fluid retention, both of which contribute to a rise <strong>in</strong>weight that may be unacceptable to patients. Evidence is accumulat<strong>in</strong>g about the potentialeffects of <strong>in</strong>sul<strong>in</strong> and other anti-diabetic agents on appetite control and satiety that maymake the control of weight more difficult. Although the long-term dim<strong>in</strong>ution of risk ofvascular complications is now established beyond doubt, the sudden <strong>in</strong>stitution of strictglycaemic control after a prolonged period of hyperglycaemia, can transiently, but sometimesseriously, destabilise microvascular disease (Agardh et al., 1992; The <strong>Diabetes</strong> Control andComplications Trial Research Group, 1998). In type 1 diabetes, the long-term follow up ofthe DCCT cohorts unequivocally has extended the evidence to <strong>in</strong>clude slowed progressionof macrovascular, as well as microvascular, disease (Nathan et al., 2003; Writ<strong>in</strong>g team forthe <strong>Diabetes</strong> Control and Complications Trial/Epidemiology of <strong>Diabetes</strong> Intervention andComplications Research Group, 2002), and so the risks of <strong>in</strong>tensified therapy need to bebalanced aga<strong>in</strong>st the potentially large ga<strong>in</strong>s. A new risk, of unknown magnitude, is the<strong>in</strong>creas<strong>in</strong>g use of novel <strong>in</strong>sul<strong>in</strong>s, which have different properties from endogenous human<strong>in</strong>sul<strong>in</strong> and thus, at least <strong>in</strong> theory, may have different side-effects.<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

172 RISKS OF STRICT GLYCAEMIC CONTROLWhen cohorts of patients are studied rather than <strong>in</strong>dividuals, other potential risks ofthe demands of <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapies, and <strong>in</strong> particular the <strong>in</strong>herent psychosocialstra<strong>in</strong>s, do not emerge as a problem. Concerns have been expressed that greater use ofhome blood glucose monitor<strong>in</strong>g may <strong>in</strong>crease anxiety, particularly <strong>in</strong> type 2 patients whoare not tak<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>, and who have limited means at their disposal of respond<strong>in</strong>g tohigh blood glucose values (Franciosi et al., 2001). In contrast, <strong>in</strong> type 1 diabetes, researchsuggests that patients may prefer <strong>in</strong>tensified treatment regimens. In the <strong>Diabetes</strong> Controland Complications Trial, patients <strong>in</strong> the <strong>in</strong>tensive treatment arm of the study had an overallimprovement <strong>in</strong> their subjective feel<strong>in</strong>gs of control and well-be<strong>in</strong>g, although it was offset bya greater fear of hypoglycaemia (The <strong>Diabetes</strong> Control and Complications Trial ResearchGroup, 1996a). However, this balance may be particularly positive <strong>in</strong> people who activelychoose to use <strong>in</strong>tensified therapies. In the DAFNE Trial, all participants had selected the<strong>in</strong>tensive programme of treatment, and significant and apparently last<strong>in</strong>g benefits <strong>in</strong> qualityof life measures were demonstrated us<strong>in</strong>g the <strong>in</strong>tensified management strategy (DAFNEResearch Group, 2002). However, it must be acknowledged that when <strong>in</strong>dividual patientsare exhorted to achieve perfection <strong>in</strong> glycaemic control, they may experience difficultiesand frustration with the impossible task of try<strong>in</strong>g to elim<strong>in</strong>ate blood glucose read<strong>in</strong>gs thatlie outside the normal range, especially if they are not equipped to act upon such read<strong>in</strong>gs.In general, the ma<strong>in</strong> risk of <strong>in</strong>tensified diabetes therapy rema<strong>in</strong>s hypoglycaemia. Thischapter exam<strong>in</strong>es the problem of hypoglycaemia that is specifically associated with strictglycaemic control, an area that has aroused much concern and controversy. Most commentsrelate to patients with type 1 diabetes. The risks of severe hypoglycaemia associated withstrict control <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes are likely to be similar, but occur much later<strong>in</strong> the natural history of the disease, when <strong>in</strong>sul<strong>in</strong> deficiency is profound (see Chapter 11).DEFINITION OF HYPOGLYCAEMIAIt is difficult to determ<strong>in</strong>e a frequency of hypoglycaemia without first def<strong>in</strong><strong>in</strong>g what is meant by‘hypoglycaemia’. In many studies, hypoglycaemia is documented by self-report<strong>in</strong>g, which maybe very unreliable (Heller et al., 1995). Retrospective analyses suffer from problems of recall,and accurate documentation of hypoglycaemia is obta<strong>in</strong>ed only <strong>in</strong> prospective research studiesthat require biochemical verification of low blood glucose concentrations (see Chapter 3).<strong>Hypoglycaemia</strong> can be categorised by its symptomatology and its severity, but no realconsensus exists. ‘Mild’ hypoglycaemia is usually def<strong>in</strong>ed as an episode that a person recognisesand treats themselves and does not significantly disrupt daily liv<strong>in</strong>g; ‘severe’ hypoglycaemiais an episode <strong>in</strong> which blood glucose has fallen to a level where the patient hasbecome so disabled that assistance is required from another person (The <strong>Diabetes</strong> Controland Complications Trial Research Group, 1991). Alternatively, ‘severe’ hypoglycaemia maybe def<strong>in</strong>ed by the requirement for parenteral treatment (<strong>in</strong>tramuscular glucagon or <strong>in</strong>travenousdextrose), with or without hospital admission, or by the development of coma (The<strong>Diabetes</strong> Control and Complications Trial Research Group, 1987). A category of ‘moderate’hypoglycaemia, <strong>in</strong> which an <strong>in</strong>dividual requires external assistance but which falls shortof requir<strong>in</strong>g parenteral therapy or develop<strong>in</strong>g a coma, or the division of hypoglycaemia<strong>in</strong>to grades of severity has also been used (Limbert et al., 1993). Obviously the def<strong>in</strong>itionused will affect the estimate of <strong>in</strong>cidence, and if severe hypoglycaemia is def<strong>in</strong>edsolely as coma, rates will be lower than if all episodes requir<strong>in</strong>g assistance are <strong>in</strong>cluded.

CONTRIBUTORS TO INCREASED RISK OF SEVERE HYPOGLYCAEMIA 173Various levels of blood glucose concentration have been used to def<strong>in</strong>e mild and moderatehypoglycaemia biochemically, but there is now recognition that the blood glucose concentrationswhich have been used arbitrarily to def<strong>in</strong>e pathological ‘spontaneous’ hypoglycaemia(such as < 22 mmol/l) are unsuitable for def<strong>in</strong><strong>in</strong>g hypoglycaemia <strong>in</strong> people with diabetes.An arterialised plasma glucose concentration of around 3.6 mmol/l may be sufficient tocause physiological autonomic responses <strong>in</strong> healthy volunteers (Chapter 1). Subtle changes<strong>in</strong> cognitive function can <strong>in</strong>itially be detected by formal test<strong>in</strong>g below 4.0 mmol/l, althoughcl<strong>in</strong>ically relevant cognitive impairment does not occur until the arterialised plasma glucoseconcentration has fallen to approximately 3.0 mmol/l (Chapter 2). There is evidence thatif plasma glucose falls below 3.0 mmol/l for a period of time, this can reduce the symptomaticresponses to a further episode of hypoglycaemia occurr<strong>in</strong>g with<strong>in</strong> the follow<strong>in</strong>g 24hours (Heller and Cryer, 1991), so-called antecedent hypoglycaemia (Chapter 7). The cont<strong>in</strong>uousavoidance of low blood glucose concentrations (below 3.0 mmol/l) can restore normalhormonal and symptomatic responses to hypoglycaemia (Fanelli et al., 1993; Cranston et al.,1994), and it follows that values below 3.0 mmol/l can be considered to be hypoglycaemic.<strong>Diabetes</strong> UK co<strong>in</strong>ed the phrase ‘make four the floor’ to protect aga<strong>in</strong>st potentially dangeroushypoglycaemia and, most recently, a panel convened by the American <strong>Diabetes</strong> Associationconcluded <strong>in</strong> favour of a concentration of 3.9 mmol/l (Work<strong>in</strong>g Group on Hypoglycemia,American <strong>Diabetes</strong> Association, 2005). This was based on research data, show<strong>in</strong>g evidenceof counterregulatory responses at blood glucose concentrations (often arterial or arterialised)of 3.9 mmol/l and the demonstration of a small reduction <strong>in</strong> the glucagon response tohypoglycaemia <strong>in</strong>duced immediately afterwards. However, this def<strong>in</strong>es ‘hypoglycaemia’ atblood glucose concentrations that commonly occur <strong>in</strong> healthy non-diabetic people and maytherefore encourage over-treatment. It is probably more appropriate to differentiate betweentargets for adjust<strong>in</strong>g therapy, which may properly be above 4.0 mmol/l, and ‘hypoglycaemia’,which implies a pathological cause and requires <strong>in</strong>tervention.CONTRIBUTORS TO INCREASED RISK OF SEVEREHYPOGLYCAEMIA IN PATIENTS UNDERTAKING INTENSIFIEDINSULIN THERAPYFactors Predispos<strong>in</strong>g Patients to Severe <strong>Hypoglycaemia</strong> <strong>in</strong> IntensifiedInsul<strong>in</strong> TherapyThe relationship between impaired symptomatic awareness of hypoglycaemia and an<strong>in</strong>creased rate of severe hypoglycaemia is well established (Hepburn et al., 1990; Gold et al.,1994; Clarke et al., 1995), although affected patients <strong>in</strong> these studies were not subject to strictglycaemic control. The association between counterregulatory failure and <strong>in</strong>creased risk ofsevere hypoglycaemia is also well recognised (Ryder et al., 1990). Indeed, counterregulatoryfailure was proposed as a predictor of risk of severe hypoglycaemia <strong>in</strong> the subsequentapplication of <strong>in</strong>tensified therapy (White et al., 1983), and it was not until later that theability of <strong>in</strong>tensified therapy to cause counterregulatory failure was suggested (Simonsonet al., 1985a). It is <strong>in</strong>deed very important to appreciate that neither asymptomatic nor severehypoglycaemia are restricted to people us<strong>in</strong>g <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy.

174 RISKS OF STRICT GLYCAEMIC CONTROLApart from a previous history of severe hypoglycaemia, the greatest risk may be the degreeof <strong>in</strong>sul<strong>in</strong> deficiency, as reflected by the absence of C-peptide (Muhlhauser et al., 1998), aswell as the glycaemic control prior to embark<strong>in</strong>g upon <strong>in</strong>tensified therapy and the determ<strong>in</strong>ationto reach the glycaemic targets (Bott et al., 1994; Muhlhauser et al., 1998). Preservationof endogenous <strong>in</strong>sul<strong>in</strong> is not affected by <strong>in</strong>tensification of <strong>in</strong>sul<strong>in</strong> therapy, althoughthere is evidence to suggest that if strict glycaemic control is imposed when diabetes is diagnosed,this may result <strong>in</strong> more prolonged preservation of endogenous <strong>in</strong>sul<strong>in</strong> secretion (Shahet al., 1989, The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1998b). Otherfactors, related to the patient rather than to treatment, may <strong>in</strong>crease the risk of severe hypoglycaemia,<strong>in</strong>clud<strong>in</strong>g social class (Muhlhauser et al., 1998) and possibly genetics. In a study fromDenmark, much of the risk of severe hypoglycaemia was attributed to ACE genotype (Pedersen-Bjergaard et al., 2003), although this has not been confirmed and has aroused controversy;also, the absence of traditional risk factors <strong>in</strong> the Danish study is a cause for concern.The Effects of Intensified Insul<strong>in</strong> Therapy Upon Risk of Severe<strong>Hypoglycaemia</strong>In the DCCT, a clear l<strong>in</strong>k was demonstrated between <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy and thefrequency of severe hypoglycaemia. In that trial, a three-fold higher rate of severe hypoglycaemiawas recorded by the patients <strong>in</strong> the <strong>in</strong>tensive treatment arm when compared with thoseon conventional therapy (The <strong>Diabetes</strong> Control and Complications Trial Research Group,1991; 1993; 1997). This persisted throughout the entire study, although absolute rates decl<strong>in</strong>edgradually <strong>in</strong> both groups. Furthermore, the risk of severe hypoglycaemia was higher for anygiven HbA 1c , for the people receiv<strong>in</strong>g <strong>in</strong>tensive treatment. This phenomenon has not beenadequately expla<strong>in</strong>ed. It is now known that exposure to hypoglycaemia per se can <strong>in</strong>ducedefects <strong>in</strong> counterregulation and loss of subjective awareness of hypoglycaemia (Heller andCryer, 1991; George et al., 1995; 1997; Davis et al., 1997). It has been assumed that <strong>in</strong>tensivetherapy exposes the patient to a greater frequency of mild hypoglycaemia that is sufficientto <strong>in</strong>duce such defects and thereby <strong>in</strong>crease the risk of severe hypoglycaemia by thatmechanism. However, methods for deliver<strong>in</strong>g <strong>in</strong>tensified diabetes therapy have subsequentlyimproved. Modern methods that focus on transferr<strong>in</strong>g skills of <strong>in</strong>sul<strong>in</strong> adjustment to the patientsthemselves are reported to achieve improvements <strong>in</strong> HbA 1c with multiple daily <strong>in</strong>jectiontherapy regimens, without caus<strong>in</strong>g more episodes of severe hypoglycaemia, and <strong>in</strong> their mostsuccessful forms achieve a parallel reduction of hypoglycaemia (Jorgens et al., 1993; DAFNEStudy Group, 2002; Plank et al., 2004; Samann et al., 2005). The judicious use of <strong>in</strong>sul<strong>in</strong>analogues <strong>in</strong> <strong>in</strong>tensified regimens may be associated with slightly less risk of hypoglycaemia(Ashwell et al., 2006), whereas the use of cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion (CSII)with pumps is associated with a much lower frequency of severe hypoglycaemia, and has beenused successfully as treatment for patients with problematical hypoglycaemia (Bode et al.,1996, Rodrigues et al., 2005) and <strong>in</strong> the context of cl<strong>in</strong>ical trials (Hoogma et al., 2006).The L<strong>in</strong>k Between Intensified Insul<strong>in</strong> Therapy and Risk of Severe<strong>Hypoglycaemia</strong>Patients describe symptoms of hypoglycaemia at a wide range of blood glucose concentrations.In an <strong>in</strong>dividual patient, the ma<strong>in</strong> determ<strong>in</strong>ant of the blood glucose concentration

CONTRIBUTORS TO INCREASED RISK OF SEVERE HYPOGLYCAEMIA 175at which protective responses commence is probably the recent prevail<strong>in</strong>g range of bloodglucose concentration to which the patient has been exposed. For example, when patientswith poorly-controlled type 2 diabetes were studied with a controlled hypoglycaemic challengeafter blood glucose had been normalised overnight, their ep<strong>in</strong>ephr<strong>in</strong>e responses tohypoglycaemia were triggered at higher blood glucose values than <strong>in</strong> well-controlled patients(Korzon-Burakowska et al., 1998).As mentioned earlier, the first <strong>in</strong>dication that strict glycaemic control might cause abnormalresponses to hypoglycaemia was observed when controlled hypoglycaemia was <strong>in</strong>duced <strong>in</strong>a small group of patients with type 1 diabetes before, and after, they had been treatedwith <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy (Simonson et al., 1985a). Follow<strong>in</strong>g the improvement <strong>in</strong>glycaemic control, the magnitude of the counterregulatory hormonal response to an abruptlower<strong>in</strong>g of blood glucose to 2.8 mmol/l was significantly less than observed previously.This study had been planned to <strong>in</strong>vestigate the potential of better glycaemic control to restoresome of the defects of normal counterregulation that develop <strong>in</strong> people with type 1 diabetes(see Chapter 6), so these results were unexpected. The importance of these prelim<strong>in</strong>aryobservations was underl<strong>in</strong>ed by a subsequent study <strong>in</strong> which patients with type 1 diabetesreceiv<strong>in</strong>g <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> treatment were found to have impaired glucose counterregulation(Amiel et al., 1987). Dur<strong>in</strong>g an <strong>in</strong>travenous <strong>in</strong>fusion of <strong>in</strong>sul<strong>in</strong>, most patients were unableto ma<strong>in</strong>ta<strong>in</strong> arterialised plasma glucose above 3.0 mmol/l, <strong>in</strong> contrast with conventionallytreateddiabetic patients whose glycaemic control was not as good (as demonstrated by higherglycated haemoglob<strong>in</strong> concentrations) or non-diabetic volunteers. The <strong>in</strong>tensively-treateddiabetic patients were less symptomatic, and although the rise <strong>in</strong> their plasma ep<strong>in</strong>ephr<strong>in</strong>ewas of similar magnitude to the other groups, this occurred only when the hypoglycaemia wasmore profound. Further studies of hypoglycaemia, us<strong>in</strong>g a stepped glucose clamp to producea controlled reduction of blood glucose, confirmed that the symptomatic and hormonalresponses started at lower blood glucose concentrations <strong>in</strong> patients with strict glycaemiccontrol, and were delayed <strong>in</strong> onset and dim<strong>in</strong>ished <strong>in</strong> magnitude for any given blood glucoseconcentration (Amiel et al., 1988) (Figure 8.1).The delayed onset and dim<strong>in</strong>ished vigour of symptomatic and hormonal responses to hypoglycaemia<strong>in</strong> strictly-controlled diabetic subjects offers a partial explanation for the <strong>in</strong>creasedFigure 8.1 The effect of <strong>in</strong>tensified diabetes therapy (IRx) on ep<strong>in</strong>ephr<strong>in</strong>e responses to a slowreduction <strong>in</strong> plasma glucose over four hours. Copyright © 1988 American <strong>Diabetes</strong> Association. FromAmiel et al., 1988. Repr<strong>in</strong>ted with permission from The American <strong>Diabetes</strong> Association

176 RISKS OF STRICT GLYCAEMIC CONTROLoccurrence of asymptomatic biochemical hypoglycaemia. The risk may be particularly manifestedwhen the glycated haemoglob<strong>in</strong> concentration is reduced to with<strong>in</strong>, or just above, thenon-diabetic range (Box 8.1). This was shown <strong>in</strong> a study of 34 subjects with type 1 diabeteswho had a wide range of total HbA 1 values (K<strong>in</strong>sley et al., 1995). They were subjected to astepped glucose clamp to lower arterialised blood glucose to 2.3 mmol/l and the responses werecompared with a non-diabetic control group. Symptomatic responses (particularly autonomic)and some counterregulatory hormonal responses were dim<strong>in</strong>ished <strong>in</strong> the seven diabetic subjectswho had a total HbA 1 of 7.85% or less, i.e., glycaemic control that was with<strong>in</strong> their local nondiabeticrange of total HbA 1 . A very similar study by Pampanelli et al. (1996) produced identicalobservations <strong>in</strong> 10 of 33 subjects, whose HbA 1c was with<strong>in</strong> the local non-diabetic range, and <strong>in</strong>whom it was also noted that the onset of some aspects of cognitive dysfunction was delayed.Current evidence would suggest that it is the <strong>in</strong>creased exposure to episodic hypoglycaemia,associated with the treatment strategy that is promot<strong>in</strong>g the problem. Most importantly, a seriesof studies has shown that hypoglycaemia awareness and counterregulatory hormone responsescan be restored <strong>in</strong> well-controlled diabetic subjects by avoidance of blood glucose concentrationsbelow 3.0 mmol/l <strong>in</strong> daily life, confirm<strong>in</strong>g the circular l<strong>in</strong>k between hypoglycaemiaexposure and impaired awareness of hypoglycaemia (Fanelli et al., 1993; Cranston et al., 1994).Thus, although impaired awareness of hypoglycaemia is a major problem <strong>in</strong> cl<strong>in</strong>icalpractice, it is by no means exclusively conf<strong>in</strong>ed to <strong>in</strong>tensified therapy. Although the riskrema<strong>in</strong>s greater with lower mean glucose and glycated haemoglob<strong>in</strong> concentrations, impairedawareness is reversible, at least <strong>in</strong> the sett<strong>in</strong>g of carefully controlled research studies, byscrupulous avoidance of even modest hypoglycaemia <strong>in</strong> daily life (Fanelli et al., 1993;Cranston et al., 1994). Although this may result <strong>in</strong> a deterioration of glycaemic control asthe problem was reversed, with a rise <strong>in</strong> mean HbA 1c from 6.9% to 8.0% <strong>in</strong> one smallstudy of seven patients with impaired awareness of hypoglycaemia (Liu et al., 1996), thisis not <strong>in</strong>evitable (Cranston et al., 1994). It is possible for avoidance of hypoglycaemia toresult <strong>in</strong> an improvement of glycated haemoglob<strong>in</strong>, as post-hypoglycaemia hyperglycaemiais eradicated.Much less work has been done <strong>in</strong> type 2 diabetes, although there is <strong>in</strong>creas<strong>in</strong>g evidencethat <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-treated type 2 diabetes of long duration, the prevalence of severehypoglycaemia is not greatly different from people with type 1 diabetes (see Chapter 11).Modern trends of start<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> earlier <strong>in</strong> type 2 diabetes, when <strong>in</strong>sul<strong>in</strong> deficiency is notBox 8.1Effects of strict glycaemic control <strong>in</strong> type 1 diabetes• Reduction <strong>in</strong> microvascular and macrovascular complications.• Potential <strong>in</strong>crease <strong>in</strong> risk of severe hypoglycaemia.• Dim<strong>in</strong>ished counterregulatory and symptomatic responses to hypoglycaemia.• Altered glycaemic thresholds for activation of responses (i.e., lower blood glucoserequired).• Promotion of <strong>in</strong>creased frequency of exposure to hypoglycaemia which exacerbatesimpaired awareness of hypoglycaemia.• Tendency to weight ga<strong>in</strong>.

CEREBRAL ADAPTATION 177severe, are likely to reduce the overall risk of hypoglycaemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-treatedtype 2 diabetes. Recent studies us<strong>in</strong>g bedtime basal <strong>in</strong>sul<strong>in</strong> as the first l<strong>in</strong>e of <strong>in</strong>tensify<strong>in</strong>gdiabetes treatment for type 2 patients who are not achiev<strong>in</strong>g glycaemic targets, have reporteda low risk of severe hypoglycaemia, even when us<strong>in</strong>g conventional <strong>in</strong>sul<strong>in</strong>s (Yki-Jarv<strong>in</strong>enet al., 2006). However, caution is <strong>in</strong>dicated when patients require conversion to full <strong>in</strong>sul<strong>in</strong>therapy. In a small study of poorly-controlled patients treated with oral medication, <strong>in</strong> whichresponses to hypoglycaemia were measured before and after improv<strong>in</strong>g glycaemic controlwith <strong>in</strong>sul<strong>in</strong>, counterregulatory responses and the blood glucose thresholds at which thesewere <strong>in</strong>itiated were modified, as occurs <strong>in</strong> type 1 diabetes (Korson-Burakowska et al., 1998).CEREBRAL ADAPTATIONWhen hypoglycaemia occurs, the stimulus for counterregulation appears to be a fall <strong>in</strong> thecerebral metabolic rate of glucose. Boyle et al. (1994) measured arteriovenous differences<strong>in</strong> glucose concentration <strong>in</strong> the human bra<strong>in</strong> dur<strong>in</strong>g hypoglycaemia to show that the rate ofuptake of glucose (and by implication of metabolism) falls before most of the counterregulatoryresponses and cognitive changes occur. They also demonstrated that this fall <strong>in</strong> metabolicrate of the bra<strong>in</strong> was reduced <strong>in</strong> healthy volunteers who were made acutely hypoglycaemicfollow<strong>in</strong>g a period of 56 hours of protracted moderate hypoglycaemia, suggest<strong>in</strong>g that themetabolism of the human bra<strong>in</strong> can adapt to prolonged exposure to low blood glucose.This enables the bra<strong>in</strong> to ma<strong>in</strong>ta<strong>in</strong> its metabolism and cont<strong>in</strong>ue to function <strong>in</strong> response tosubsequent hypoglycaemia. A further study <strong>in</strong> diabetic patients showed that diabetic patientswith strict glycaemic control and impaired awareness of hypoglycaemia were able to ma<strong>in</strong>ta<strong>in</strong>the rate of cerebral uptake of glucose dur<strong>in</strong>g experimental hypoglycaemia, while otherswith normal symptomatic awareness exhibited a marked fall <strong>in</strong> cerebral uptake of glucose,associated with symptomatic and counterregulatory hormonal responses (Figure 8.2) (Boyleet al., 1995). These data led to the hypothesis that impaired awareness of hypoglycaemiaand defective glucose counterregulation may result from an adaptation <strong>in</strong> the sensitivity ofthe cerebral glucose sensor, which allows it to susta<strong>in</strong> its metabolic rate (and so not triggercounterregulation) dur<strong>in</strong>g subsequent hypoglycaemia (see Chapter 7).However, the expectation that patients with impaired awareness of hypoglycaemia willshow an <strong>in</strong>crease <strong>in</strong> bra<strong>in</strong> glucose metabolic rate at any given blood glucose concentrationhas not been supported by neuroimag<strong>in</strong>g studies. In studies utilis<strong>in</strong>g positron emissiontomography that used various tracers for glucose to measure either the metabolic rate ofbra<strong>in</strong> glucose or glucose tracer uptake <strong>in</strong> humans, several <strong>in</strong>vestigators have failed to f<strong>in</strong>ddifferences dur<strong>in</strong>g euglycaemia or hypoglycaemia that could be <strong>in</strong> accord with prevail<strong>in</strong>gglycaemic control (Cranston et al., 2001; Segal et al., 2001). One study found evidence dur<strong>in</strong>ghypoglycaemia of a difference <strong>in</strong> the change <strong>in</strong> uptake of the glucose tracer, de-oxyglucose,<strong>in</strong> the bra<strong>in</strong> region around the hypothalamus <strong>in</strong> <strong>in</strong>tensively-treated diabetic subjects who hadimpaired awareness of hypoglycaemia (Cranston et al., 2001), which is of <strong>in</strong>terest becauseanimal studies have implicated this region (among others) <strong>in</strong> sens<strong>in</strong>g hypoglycaemia. Inmore recent studies, the difference <strong>in</strong> bra<strong>in</strong> glucose metabolism <strong>in</strong> subjects with impairedawareness of hypoglycaemia was a failure of <strong>in</strong>crease of cerebral metabolic rate dur<strong>in</strong>ghypoglycaemia, associated with a failure to generate or perceive symptoms (B<strong>in</strong>gham et al.,2005). These data are compatible with the concept that cortical activation is important forperception of symptoms and that this fails <strong>in</strong> people who develop a loss of awareness of

178 RISKS OF STRICT GLYCAEMIC CONTROLFigure 8.2 Changes from basel<strong>in</strong>e (mean ± SD) <strong>in</strong> (a) glucose uptake <strong>in</strong> the bra<strong>in</strong> and (b) hypoglycaemiasymptom scores, and plasma concentrations of (c) ep<strong>in</strong>ephr<strong>in</strong>e and (d) pancreatic polypeptidedur<strong>in</strong>g hypoglycaemia <strong>in</strong> patients with type 1 diabetes with differ<strong>in</strong>g degrees of glycaemic control(black bars), and <strong>in</strong> non-diabetic subjects (grey bars). Reproduced from Boyle et al. (1995) withpermission. Copyright © 1995 Massachusetts Medical Society

CEREBRAL ADAPTATION 179hypoglycaemia. It is becom<strong>in</strong>g evident that changes <strong>in</strong> symptomatic responses and corticalfunction <strong>in</strong> hypoglycaemia are driven by complex mechanisms associated with, but notexclusively controlled directly by, the changes <strong>in</strong> the glucose metabolic rate of neurones.Some cognitive functions are better preserved than others dur<strong>in</strong>g hypoglycaemia <strong>in</strong> subjectswho have previous experience of hypoglycaemia than <strong>in</strong> hypoglycaemia-naive subjects whohave normal counterregulation (Fanelli et al., 1993; Boyle et al., 1995). This does not entirelyfit the cl<strong>in</strong>ical picture of patients becom<strong>in</strong>g significantly confused dur<strong>in</strong>g hypoglycaemiawhile rema<strong>in</strong><strong>in</strong>g asymptomatic.One measure of cognitive function, the choice reaction time, does not appear to adapt, andwhen hypoglycaemia is <strong>in</strong>duced slowly, it deteriorates at similar levels of blood glucose <strong>in</strong>most subjects, irrespective of their previous glycaemic experience and their state of hypoglycaemiaawareness (Maran et al., 1995). Other measures of cognitive function also deteriorateat similar levels of blood glucose <strong>in</strong> diabetic subjects who have had very disparate experiencesof preced<strong>in</strong>g glycaemia (Widom and Simonson, 1990; Amiel et al., 1991; Hvidberget al., 1996). The ability of the bra<strong>in</strong> to adapt its metabolic and functional capacity accord<strong>in</strong>gto previous glycaemic experience varies across different regions of the bra<strong>in</strong>. Regions of thebra<strong>in</strong> that detect hypoglycaemia, and some parts of the cerebral cortex, may be able to adaptmore effectively to antecedent hypoglycaemia than other areas, to susta<strong>in</strong> glucose metabolismdur<strong>in</strong>g subsequent exposure. As blood glucose falls this would effectively destroy the normalprotective hierarchy of corrective and symptomatic responses that precede cognitive impairment,replac<strong>in</strong>g it with the dangerous situation whereby cognitive impairment is the <strong>in</strong>itialresponse to hypoglycaemia, with autonomic responses not occurr<strong>in</strong>g until the blood glucosedecl<strong>in</strong>es to a much lower level. In this situation the patient becomes too confused and unableto recognise the warn<strong>in</strong>g symptoms and so take appropriate corrective action (Figure 8.3).Figure 8.3 The change <strong>in</strong> hierarchy of responses to hypoglycaemia (a) before and (b) after <strong>in</strong>tensified<strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> type 1 diabetes mellitus

180 RISKS OF STRICT GLYCAEMIC CONTROLThe magnitude of the change <strong>in</strong> glycaemic thresholds for various functions of the bra<strong>in</strong><strong>in</strong> response to strict control of diabetes is variable. Where glucose thresholds for cognitivedysfunction do alter <strong>in</strong> people with impaired awareness of hypoglycaemia, the differencesbetween the blood glucose thresholds for the symptomatic and autonomic responses and thosefor the onset of cognitive impairment are much smaller. As a result, the w<strong>in</strong>dow of opportunityfor the patient to recognise that hypoglycaemia is develop<strong>in</strong>g is much narrower, giv<strong>in</strong>gless time for corrective action to be taken. As described above, the molecular mechanismscontroll<strong>in</strong>g the thresholds for activation of the various components of the counterregulatoryresponses rema<strong>in</strong> the subject of <strong>in</strong>tense research.OTHER RISKS OF INTENSIFIED INSULIN THERAPYDiabetic Ketoacidosis and Hyper<strong>in</strong>sul<strong>in</strong>aemiaAlthough severe hypoglycaemia was <strong>in</strong>disputably the major metabolic side-effect of <strong>in</strong>tensive<strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> the DCCT, concerns have been expressed that some <strong>in</strong>tensive treatmentregimens may also <strong>in</strong>crease the risk of develop<strong>in</strong>g ketosis. This was primarily related tothe use of CSII (with <strong>in</strong>sul<strong>in</strong> pump therapy) and was thought to relate to the absence ofany <strong>in</strong>termediate-act<strong>in</strong>g or background <strong>in</strong>sul<strong>in</strong> <strong>in</strong> the event of pump failure. In <strong>in</strong>sul<strong>in</strong> pumptherapy, soluble or fast-act<strong>in</strong>g analogue <strong>in</strong>sul<strong>in</strong> is delivered steadily by a slow <strong>in</strong>fusion ofvery low doses throughout the day. The <strong>in</strong>sul<strong>in</strong> delivery is accelerated before meals to deliverboluses, ak<strong>in</strong> to giv<strong>in</strong>g <strong>in</strong>termittent subcutaneous <strong>in</strong>jections of short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong>. Becausethe basal <strong>in</strong>sul<strong>in</strong> is delivered <strong>in</strong> a very low volume and there is no depot of <strong>in</strong>termediateact<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> <strong>in</strong> the subcutaneous tissues to act as a reservoir, an <strong>in</strong>terruption <strong>in</strong> the deliveryof <strong>in</strong>sul<strong>in</strong> can rapidly lead to hyperglycaemia and even ketosis, especially if the patient’sblood glucose is already elevated (Castilloa et al., 1996). This may occur as a result ofdisconnection of the pump, air <strong>in</strong> the delivery system, blockage <strong>in</strong> the tub<strong>in</strong>g or morerarely, mechanical failure of the pump. The apparently high risk of diabetic ketoacidosis(DKA) with <strong>in</strong>sul<strong>in</strong> pump therapy was first described when pumps left the experimentalcentres where they had been <strong>in</strong>vented and were taken up for more general use (Knightet al., 1985), although the same centre that reported a problem with DKA also reportedsatisfactory experience overall with pump therapy (Knight et al., 1986). In 1997, a metaanalysisof trials of CSII has <strong>in</strong>dicated that the rate of DKA was significantly higher (Eggeret al., 1997) and the rate of development of DKA was also slightly greater <strong>in</strong> <strong>in</strong>tensivelytreatedpatients <strong>in</strong> the DCCT, although many of those patients used multiple <strong>in</strong>jectionsof <strong>in</strong>sul<strong>in</strong> to improve their glycaemic control (The <strong>Diabetes</strong> Control and ComplicationsTrial Research Group, 1995b). However, a more recent meta-analysis, <strong>in</strong>clud<strong>in</strong>g studiesus<strong>in</strong>g more modern equipment and up-dated algorithms for us<strong>in</strong>g pump therapy, is morereassur<strong>in</strong>g (Pickup et al., 2002). Current <strong>in</strong>tensive treatment regimens focus more on transferr<strong>in</strong>gskills of flexible <strong>in</strong>sul<strong>in</strong> dose adjustment more effectively to the patients and <strong>in</strong> thissett<strong>in</strong>g, DKA rates are not higher. Indeed, experienced centres have utilised <strong>in</strong>sul<strong>in</strong> pumptherapy to help people avoid recurrent DKA (Rodrigues et al., 2005). However, the riskis worth reiterat<strong>in</strong>g, as it can return when new technologies are applied <strong>in</strong> <strong>in</strong>experiencedsett<strong>in</strong>gs.Intensive <strong>in</strong>sul<strong>in</strong> therapy often leads to a redistribution of the times of adm<strong>in</strong>istrationof <strong>in</strong>sul<strong>in</strong> rather than a straightforward <strong>in</strong>crease <strong>in</strong> dosage, and concerns have been raised

THERAPEUTIC MANIPULATION 181Figure 8.4 Severe hypoglycaemia (events per 100 patient-years) at basel<strong>in</strong>e with multiple daily<strong>in</strong>jections (MDI) and by year on cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion (CSII). Copyright © 1996American <strong>Diabetes</strong> Association. From Bode et al., 1996. Repr<strong>in</strong>ted with permission from The American<strong>Diabetes</strong> Associationthat cont<strong>in</strong>uous peripheral hyper<strong>in</strong>sul<strong>in</strong>aemia may be deleterious. This anxiety may be moretheoretical than real, as improved glycaemic control <strong>in</strong> type 1 diabetes improves <strong>in</strong>sul<strong>in</strong>sensitivity (Simonson et al., 1985b), which ultimately should prevent hyper<strong>in</strong>sul<strong>in</strong>aemia.However, achiev<strong>in</strong>g adequate plasma concentrations of <strong>in</strong>sul<strong>in</strong> <strong>in</strong> the hepatic circulation isalways likely to be at the cost of promot<strong>in</strong>g hyper<strong>in</strong>sul<strong>in</strong>aemia <strong>in</strong> the systemic circulation,as <strong>in</strong>sul<strong>in</strong> has to be delivered by subcutaneous <strong>in</strong>jection. This potential over-<strong>in</strong>sul<strong>in</strong>isationmay contribute to the risk of hypoglycaemia, and when <strong>in</strong>sul<strong>in</strong> is delivered <strong>in</strong>to the portalsystem, as with <strong>in</strong>traperitoneal <strong>in</strong>fusion systems, hypoglycaemia is less frequent at any givenblood glucose level (Lassmann-Vague et al., 1996; Dunn et al., 1997). However, one study(Figure 8.4) demonstrated a pronounced and susta<strong>in</strong>ed reduction <strong>in</strong> the frequency of severehypoglycaemia follow<strong>in</strong>g the transfer of patients from multiple <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> to CSII(Bode et al., 1996), and so appropriate temporal distribution of the action of <strong>in</strong>sul<strong>in</strong> may,be the critical factor <strong>in</strong> prevent<strong>in</strong>g hypoglycaemia.THERAPEUTIC MANIPULATIONAvoidance of <strong>Hypoglycaemia</strong>It is important to stress that management of the potentially devastat<strong>in</strong>g syndrome of impairedawareness of hypoglycaemia and deficient counterregulation, with its high risk of severehypoglycaemia, should not be an excuse for encourag<strong>in</strong>g poor glycaemic control. However,some patients with these acquired syndromes are not suitable for <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy and

182 RISKS OF STRICT GLYCAEMIC CONTROLvery strict glycaemic control, and blood glucose targets may have to be higher when facedwith these problems. Frequent blood glucose monitor<strong>in</strong>g is essential to identify biochemicalhypoglycaemia, and the use of basal-bolus <strong>in</strong>sul<strong>in</strong> regimens (which predom<strong>in</strong>antly use shortact<strong>in</strong>g<strong>in</strong>sul<strong>in</strong>s) may be beneficial <strong>in</strong> avoid<strong>in</strong>g recurrent hypoglycaemia. However, frequentblood glucose monitor<strong>in</strong>g alone can sometimes exacerbate the problem, unless the patientis <strong>in</strong>structed <strong>in</strong> how to use the data to adjust <strong>in</strong>sul<strong>in</strong> regimens prospectively to avoidproblems, rather than to react by immediate treatment of the <strong>in</strong>evitable occasional highread<strong>in</strong>g. There have been few detailed behavioural studies <strong>in</strong> people with impaired awarenessof hypoglycaemia but it may be a particular problem where the l<strong>in</strong>ks between high bloodglucose and risk of vascular complications have been very well accepted by the patient, whomay need conv<strong>in</strong>c<strong>in</strong>g that transient hyperglycaemia is not a problem. Cl<strong>in</strong>ical experiencesuggests that many patients with such problems ‘glucose chase’, tak<strong>in</strong>g corrective actionwhenever they identify a high blood glucose concentration. Correct<strong>in</strong>g this behaviour, <strong>in</strong>particular avoid<strong>in</strong>g postprandial glucose correction, and re-tra<strong>in</strong><strong>in</strong>g patients to use glucosemonitor<strong>in</strong>g to seek patterns for future dose adjustment, can be very beneficial. Programmesthat teach patients to test blood glucose before eat<strong>in</strong>g and to use the <strong>in</strong>formation to adjust thedose of <strong>in</strong>sul<strong>in</strong> required for the immediate meal, allow people to live with greater flexibility.By record<strong>in</strong>g the blood glucose results, recurrent patterns <strong>in</strong> changes can be sought aga<strong>in</strong>stwhich prospective adjustments to <strong>in</strong>sul<strong>in</strong> regimens can be made. These measures allow HbA 1cto be improved while lower<strong>in</strong>g the risk of severe hypoglycaemia. It is important to recognisethat the preservation of physiological defences to hypoglycaemia is dependent upon theavoidance of hypoglycaemia <strong>in</strong> daily life, and not on tolerance of chronic hyperglycaemiaand an elevated glycated haemoglob<strong>in</strong>. As discussed above, the studies that have attempted torestore awareness of hypoglycaemia by avoidance of hypoglycaemia did not cause any majordeterioration of glycaemic control, although a modest <strong>in</strong>crease <strong>in</strong> HbA 1c of around 0.5–1.0%occurred <strong>in</strong> two studies (Fanelli et al., 1993; Dagogo-Jack et al., 1994). Anecdotally, averageblood glucose concentrations and HbA 1c may even improve with strategies for avoid<strong>in</strong>ghypoglycaemia, by prevent<strong>in</strong>g wide fluctuations <strong>in</strong> blood glucose.Strategies to avoid hypoglycaemia can be very time-consum<strong>in</strong>g and labour-<strong>in</strong>tensive, forthe patient as well as for the physician, and require several supportive measures, such ashav<strong>in</strong>g to ma<strong>in</strong>ta<strong>in</strong> daily telephone contact between the patient and the medical and nurs<strong>in</strong>gstaff (Fanelli et al., 1993). It took Cranston and colleagues (1994) up to 12 months for thesubjects tak<strong>in</strong>g part <strong>in</strong> their study to achieve three consecutive weeks when the home bloodglucose read<strong>in</strong>gs did not fall below 3.0 mmol/l. Two of three studies demonstrated partialrecovery of the counterregulatory responses to hypoglycaemia (Fanelli et al., 1993; Cranstonet al., 1994), but one did not (Dagogo-Jack et al., 1994), although the symptoms of hypoglycaemiawere restored. The number of subjects was small <strong>in</strong> all of these studies. However,the studies were often done <strong>in</strong> patients with an established problem of hypoglycaemia andnewer tools are at hand to help prevent or reverse the problems.Educational strategies cannot be emphasised too strongly. They rema<strong>in</strong> to be testedspecifically <strong>in</strong> people who have major problems with recurrent severe hypoglycaemia buttheir ability to achieve better glycaemic control with less frequent hypoglycaemia is amajor improvement over the outcomes of the DCCT. For patients with type 1 diabetes,the <strong>in</strong>troduction of fast-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues for <strong>in</strong>sul<strong>in</strong> replacement at meal-times hasreduced hypoglycaemia, particularly at night, because of the much shorter duration of actionof fast-act<strong>in</strong>g analogues (Brunelle et al., 1998; Home et al., 2000). Likewise, the longeract<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> analogues have been associated with a lower risk of nocturnal hypoglycaemia

PATIENTS UNSUITABLE FOR STRICT CONTROL 183(Pieber et al., 2000; Vague et al., 2003) and regimens that comb<strong>in</strong>e short and longer-act<strong>in</strong>g<strong>in</strong>sul<strong>in</strong> analogues claim to have a lower risk of hypoglycaemia <strong>in</strong> patients with type 1 diabetes(Hermansen et al., 2004; Ashwell et al., 2005), although the methods of measurement ofhypoglycaemia <strong>in</strong> many of these trials was much less robust than was discussed earlier.The potential of CSII to reduce hypoglycaemia has also been highlighted. The use offaster and less pa<strong>in</strong>ful glucose monitor<strong>in</strong>g devices will facilitate home monitor<strong>in</strong>g, althoughit rema<strong>in</strong>s critical that the data obta<strong>in</strong>ed are utilised through appropriate patient education.The advent of real-time glucose monitors may allow patients to avoid hypoglycaemia, as theycan take action to <strong>in</strong>terrupt a steady decl<strong>in</strong>e <strong>in</strong> blood glucose concentration that they wouldnot previously have had the opportunity to observe. The value of these new technologiesrema<strong>in</strong>s to be proven <strong>in</strong> rout<strong>in</strong>e cl<strong>in</strong>ical use, but they do hold promise.The ma<strong>in</strong> defence aga<strong>in</strong>st recurrent hypoglycaemia with its consequent blunt<strong>in</strong>g of subjectivesymptomatic awareness rema<strong>in</strong>s the establishment of therapeutic goals that are realisticfor <strong>in</strong>dividual patients. The physician’s tendency to concentrate on elim<strong>in</strong>at<strong>in</strong>g hyperglycaemiahas led to subnormal blood glucose values be<strong>in</strong>g ignored, a practice worsened by thebelief of some physicians and patients alike that, because an episode of biochemical hypoglycaemiais asymptomatic, it is not important. There is no doubt that a cl<strong>in</strong>ically detectabledeterioration <strong>in</strong> performance of some aspects of cognitive function occurs <strong>in</strong> human subjectsat arterialised blood glucose concentrations of 3.0 mmol/l (see Chapter 2), and an absenceof symptoms at that level should r<strong>in</strong>g alarm bells with the patient’s physician. Given thathealthy non-diabetic subjects do not commonly exhibit fast<strong>in</strong>g blood glucose concentrationsbelow 4.0 mmol/l, it seems wholly unnecessary to encourage or even permit such subnormality<strong>in</strong> patients with diabetes (one exception to this maxim be<strong>in</strong>g pregnancy where healthynon-diabetic women do exhibit lower blood glucose levels as discussed <strong>in</strong> Chapter 10). With<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy the therapeutic targets should be near-normal blood glucose levels(before meals 4.0–7.0 mmol/l, after meals 4.0–9.0 mmol/l, depend<strong>in</strong>g on time of test<strong>in</strong>g),with a slightly higher than normal glucose at bedtime (7.0–9.0 mmol/l) to reduce the risk ofhypoglycaemia occurr<strong>in</strong>g dur<strong>in</strong>g the night. Blood glucose measured dur<strong>in</strong>g the night may bea little lower (≥ 36 mmol/l), but <strong>in</strong> view of the evidence presented above, patients shouldavoid allow<strong>in</strong>g it to fall any lower than this.PATIENTS UNSUITABLE FOR STRICT CONTROLThe DCCT demonstrated that any reduction <strong>in</strong> glycated haemoglob<strong>in</strong> is associated with areduced risk of microvascular complications over time, and the benefits are greater withhigher glycated haemoglob<strong>in</strong> concentrations (The <strong>Diabetes</strong> Control and Complications TrialResearch Group, 1996b). A cross-sectional study which suggested that the risk reductionfor nephropathy is near-maximal at a glycated haemoglob<strong>in</strong> of 8% (Krolewski et al., 1995)cannot be extrapolated to other microvascular complications, because <strong>in</strong> the DCCT noglycaemic threshold (estimated by glycated haemoglob<strong>in</strong>) for the development of ret<strong>in</strong>opathywas demonstrated <strong>in</strong> a patient group whose average HbA 1c was 7%. The long-term followupof the DCCT cohort has confirmed that <strong>in</strong>tensive therapy benefits macrovascular aswell as microvascular risk and that its effects are susta<strong>in</strong>ed (Writ<strong>in</strong>g team for the <strong>Diabetes</strong>Control and Complications Trial/ Epidemiology of <strong>Diabetes</strong> Interventions and ComplicationsResearch Group, 2002). Thus, unless an <strong>in</strong>dividual already has normal glycated haemoglob<strong>in</strong>with no problematical hypoglycaemia, no patient who has diabetes is unsuitable for attempts

184 RISKS OF STRICT GLYCAEMIC CONTROLto improve their glycaemic control, especially with modern techniques that are able to deliverimproved control without <strong>in</strong>creas<strong>in</strong>g hypoglycaemia risk.In practice, however, there are patients <strong>in</strong> whom attempts to achieve a near normal glycatedhaemoglob<strong>in</strong> are not appropriate (Box 8.2). Patients with advanced complications, especiallyret<strong>in</strong>opathy, have not been shown to benefit and a sudden improvement <strong>in</strong> glycaemic controlmay cause acceleration <strong>in</strong> severity of pre-proliferative or early proliferative ret<strong>in</strong>opathy(Hanssen et al., 1986). Although some authorities claim that this should not be a contra<strong>in</strong>dicationto improv<strong>in</strong>g glycaemic control (Chantelau and Kohner, 1997), as yet there is no realevidence for benefit <strong>in</strong> advanced cases and the ret<strong>in</strong>opathy should be treated appropriatelybefore glycaemic control is <strong>in</strong>tensified. Similarly, <strong>in</strong> patients with established renal impairmentand severe macrovascular disease, attempts to treat elevated blood pressure and plasmalipids and to encourage patients to stop smok<strong>in</strong>g may be more beneficial than target<strong>in</strong>gglycaemic control alone. As <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy is aimed at achiev<strong>in</strong>g benefit over aperiod of five to ten years or more, patients with a reduced life expectancy should not beexposed to the risks and rigours associated with this treatment regimen. This applies also toelderly patients, who may be frail and physically <strong>in</strong>active.Very young patients may not be good candidates for very strict glycaemic control. Poorcontrol should not be encouraged <strong>in</strong> children, as growth may be jeopardised, and there issome evidence that pre-pubertal glycaemic control may <strong>in</strong>fluence the later risk of complications(Donaghue et al., 1997; Holl et al., 1998). However, very small children, who arevery <strong>in</strong>sul<strong>in</strong>-sensitive, may be at risk of <strong>in</strong>tellectual damage if exposed to recurrent severehypoglycaemia (see Chapters 9 and 13).Box 8.2Application of strict glycaemic control <strong>in</strong> type 1 diabetesCaution required:• Long duration of <strong>in</strong>sul<strong>in</strong>-treated diabetes (counterregulatory deficiences).• Previous history of severe hypoglycaemia.• Established impaired awareness of hypoglycaemia.• History of epilepsy.• Patient unwill<strong>in</strong>g to do home blood glucose monitor<strong>in</strong>g.Contra<strong>in</strong>dicated:• Extremes of age.• Ischaemic heart disease.• Unstable diabetic ret<strong>in</strong>opathy (can be <strong>in</strong>stituted after treatment).• Advanced diabetic complications.• Limited life expectancy (e.g. serious coexist<strong>in</strong>g disease).

REFERENCES 185It is the patient who determ<strong>in</strong>es the degree of glycaemic control that they feel is worththe effort. Patients with very erratic life styles, and those who are not prepared to committhemselves to regular self-monitor<strong>in</strong>g of blood glucose, with frequent attention to the tim<strong>in</strong>gof <strong>in</strong>jection and adjustment of dosage of <strong>in</strong>sul<strong>in</strong>, cannot safely undertake measures toachieve near-normoglycaemia. A compromise must be reached after a full discussion of therisks. Patients currently experienc<strong>in</strong>g problematical hypoglycaemia may not wish to aim forglycaemic targets near the normal range, although the regimens of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapymay still be appropriate for them if they can elim<strong>in</strong>ate hypoglycaemia from their daily lives.This may also be true of people undertak<strong>in</strong>g dangerous or physically demand<strong>in</strong>g jobs, whomay deliberately set higher blood glucose targets to protect aga<strong>in</strong>st hypoglycaemia, but whoshould be encouraged to practice regular self-monitor<strong>in</strong>g and adjustment of <strong>in</strong>sul<strong>in</strong> doses. Itis the <strong>in</strong>formed patient who must determ<strong>in</strong>e their own therapeutic aims at any given time.The doctor’s role is to try to ensure that the patient has the knowledge to make appropriatedecisions and to provide the tools to achieve these aims.CONCLUSIONS• The pr<strong>in</strong>cipal risks of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy are hypoglycaemia and weight ga<strong>in</strong>.• In earlier studies such as the DCCT, patients us<strong>in</strong>g <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimens were threetimes more likely to have an episode of severe hypoglycaemia than those on conventional<strong>in</strong>sul<strong>in</strong> regimens, but newer techniques for tra<strong>in</strong><strong>in</strong>g patients to use <strong>in</strong>sul<strong>in</strong> flexibly candeliver improved glycaemic control without any <strong>in</strong>crease <strong>in</strong> severe hypoglycaemia, andsometimes with reduced hypoglycaemia occurrence.• It is likely that exposure to hypoglycaemia dur<strong>in</strong>g <strong>in</strong>tensive therapy impairs the counterregulatoryresponses to hypoglycaemia and symptomatic awareness and this may be seenparticularly if glycated haemoglob<strong>in</strong> is with<strong>in</strong> the non-diabetic range.• Total avoidance of hypoglycaemia can restore the symptomatic response to hypoglycaemia,but achiev<strong>in</strong>g this once problematic hypoglycaemia has been established may bedemand<strong>in</strong>g and time-consum<strong>in</strong>g, both for patients and healthcare professionals. Nevertheless,the long-term benefits of good diabetic control <strong>in</strong> type 1 diabetes are unequivocaland current technologies should help more patients achieve it.• It is essential that all def<strong>in</strong>itions of good glycaemic control <strong>in</strong>clude an absence of severehypoglycaemia as well as near-normal glycated haemoglob<strong>in</strong>. Intensified treatment regimensshould be adjusted to <strong>in</strong>corporate both.REFERENCESAgardh CD, Eckert B, Agardh E (1992). Irreversible progression of severe ret<strong>in</strong>opathy <strong>in</strong> young type-1<strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus patients after improved metabolic control. Journal of DiabeticComplications 6: 96–100.Amiel SA, Tamborlane WV, Simonson DC, Sherw<strong>in</strong> RS (1987). Defective glucose counterregulationafter strict glycemic control of <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 316: 1376–83.

186 RISKS OF STRICT GLYCAEMIC CONTROLAmiel SA, Sherw<strong>in</strong> RS, Simonson DC, Tamborlane WV (1988). Effect of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapyon glycemic thresholds for counterregulatory hormone release. <strong>Diabetes</strong> 37: 901–7.Amiel SA, Pott<strong>in</strong>ger RC, Archibald HR, Chusney G, Cunnah DTF, Prior PF, Gale EAM (1991). Effectof antecedent glucose control on cerebral function dur<strong>in</strong>g hypoglycemia. <strong>Diabetes</strong> Care 14: 109–18.Ashwell SG, Amiel SA, Bilous RW, Dashora U, Heller SR, Hepburn DA et al. (2006). Improvedglycaemic control with <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e plus <strong>in</strong>sul<strong>in</strong> lispro: a multicentre, randomized, cross-overtrial <strong>in</strong> people with type 1 diabetes. Diabetic Medic<strong>in</strong>e 23: 285–92.B<strong>in</strong>gham EM, Dunn JT, Smith D, Sutcliffe-Goulden J, Reed LJ, Marsden PK, Amiel SA (2005).Differential changes <strong>in</strong> bra<strong>in</strong> glucose metabolism dur<strong>in</strong>g hypoglycaemia accompany loss of hypoglycaemiaawareness <strong>in</strong> men with type 1 diabetes mellitus. An [11C]-3-O-methyl-D-glucose PETstudy. Diabetologia 48: 2080–9.Bode BW, Steed RD, Davidson PC (1996). Reduction <strong>in</strong> severe hypoglycemia with long-term cont<strong>in</strong>uoussubcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong> Care 19: 324–7.Bott U, Jorgens V, Grusser M, Bender R, Muhlhauser I, Berger M (1994). Predictors of glycaemiccontrol <strong>in</strong> type 1 diabetic patients after participation <strong>in</strong> an <strong>in</strong>tensified treatment and teach<strong>in</strong>gprogramme. Diabetic Medic<strong>in</strong>e 11: 362–71.Boyle PJ, Nagy RJ, O’Connor AM, Kempers SF, Yeo RA, Qualls C (1994). Adaptation <strong>in</strong> bra<strong>in</strong>glucose uptake follow<strong>in</strong>g recurrent hypoglycemia. Proceed<strong>in</strong>gs of the National Academy of Sciencesof the United States of America 91: 9352–6.Boyle PJ, Kempers SF, O’Connor AM, Nagy RJ (1995). Bra<strong>in</strong> glucose uptake and unawarenessof hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 333: 1726–31.Brunelle BL, Llewelyn J, Anderson JH Jr, Gale EA, Koivisto VA (1998). Meta-analysis of the effect of<strong>in</strong>sul<strong>in</strong> lispro on severe hypoglycemia <strong>in</strong> patients with type 1 diabetes. <strong>Diabetes</strong> Care 21: 1726–31.Carlson MG, Campbell PJ (1993). Intensive <strong>in</strong>sul<strong>in</strong> therapy and weight ga<strong>in</strong> <strong>in</strong> IDDM. <strong>Diabetes</strong> 42:1700–7.Castilloa MJ, Sheen AJ, Lefebvre PJ (1996). The degree/rapidity of the metabolic deteriorationfollow<strong>in</strong>g <strong>in</strong>terruption of a cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion is <strong>in</strong>fluenced by the prevail<strong>in</strong>gblood glucose level. Journal of Cl<strong>in</strong>ical Endocr<strong>in</strong>ology and Metabolism 81: 1975–8.Chantelau E, Kohner EM (1997). Why some cases of ret<strong>in</strong>opathy worsen when diabetic controlimproves. British Medical Journal 315: 1105–6.Clarke WL, Cox DJ, Gonder-Frederick LA, Julian D, Schlundt D, Polonsky W (1995). Reducedawareness of hypoglycemia <strong>in</strong> adults with IDDM. A prospective study of hypoglycemic frequencyand associated symptoms. <strong>Diabetes</strong> Care 18: 517–22.Cranston I, Lomas J, Maran A, Macdonald IA, Amiel SA (1994). Restoration of hypoglycaemiaawareness <strong>in</strong> patients with long-duration <strong>in</strong>sul<strong>in</strong>-dependent diabetes. Lancet 344: 283–7.Cranston I, Reed LJ, Marsden PK, Amiel SA (2001). Changes <strong>in</strong> regional bra<strong>in</strong> (18)Ffluorodeoxyglucoseuptake at hypoglycemia <strong>in</strong> type 1 diabetic men associated with hypoglycemiaunawareness and counter-regulatory failure. <strong>Diabetes</strong> 50: 2329–36.DAFNE Study Group (2002). Tra<strong>in</strong><strong>in</strong>g <strong>in</strong> flexible, <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> management to enable dietaryfreedom <strong>in</strong> people with type 1 diabetes: Dose Adjustment For Normal Eat<strong>in</strong>g (DAFNE) randomisedcontrolled trial. British Medical Journal 325: 746.Dagogo-Jack S, Rattarasarn C, Cryer PE (1994). Reversal of hypoglycemia unawareness, but notdefective glucose counterregulation, <strong>in</strong> IDDM. <strong>Diabetes</strong> 43: 1426–34.Davis SN, Shavers C, Mosqueda-Garcia R, Costa F (1997). Effects of differ<strong>in</strong>g antecedent hypoglycemiaon subsequent counterregulation <strong>in</strong> normal humans. <strong>Diabetes</strong> 46: 1328–35.Donaghue KC, Fung AT, H<strong>in</strong>g S, Fairchild J, K<strong>in</strong>g J, Chan A et al. (1997). The effect of prepubertaldiabetes duration on diabetes microvascular complications <strong>in</strong> early and late adolescence. <strong>Diabetes</strong>Care 20: 77–80.Dunn FL, Nathan DM, Scav<strong>in</strong>i M, Selam JL, W<strong>in</strong>grove TG (1997). Long term therapy of IDDM with animplantable <strong>in</strong>sul<strong>in</strong> pump. The Plantable Insul<strong>in</strong> Pump Trial Study Group. <strong>Diabetes</strong> Care 20: 59–63.

REFERENCES 187Egger M, Davey Smith G, Stettler C, Diem P (1997). Risk of adverse effects of <strong>in</strong>tensified treatment<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: a meta-analysis. Diabetic Medic<strong>in</strong>e 14: 919–28.Fanelli CG, Epifano L, Rambotti AM, Pampanelli S, Di V<strong>in</strong>cenzo A, Modanelli F et al. (1993).Meticulous prevention of hypoglycemia normalizes the glycemic thresholds and magnitude of mostof neuroendocr<strong>in</strong>e responses to, symptoms of, and cognitive function dur<strong>in</strong>g hypoglycemia <strong>in</strong><strong>in</strong>tensively treated patients with short-term IDDM. <strong>Diabetes</strong> 42: 1683–9.Franciosi M, Pellegr<strong>in</strong>i F, De Berardis G, Belfiglio M, Cavaliere D, Di Nardo B et al., QuED StudyGroup (2001). Diabetic patients: an urgent need for better educational strategies. <strong>Diabetes</strong> Care 24:1870–7.George E, Harris N, Bedford C, Macdonald IA, Hardisty CA, Heller SR (1995). Prolonged but partialimpairment of the hypoglycaemic physiological response follow<strong>in</strong>g short-term hypoglycaemia <strong>in</strong>normal subjects. Diabetologia 38: 1183–90.George E, Marques JL, Harris ND, Macdonald IA, Hardisty CA, Heller SR (1997). Preservation ofphysiological responses to hypoglycemia 2 days after antecedent hypoglycemia <strong>in</strong> patients withIDDM. <strong>Diabetes</strong> Care 20: 1293–8.Gold AE, MacLeod KM, Frier BM (1994). Frequency of severe hypoglycemia <strong>in</strong> patients with type 1diabetes with impaired awareness of hypoglycemia. <strong>Diabetes</strong> Care 17: 697–703.Hanssen KF, Dahl-Jorgensen K, Lauritzen T, Feldt-Rasmussen B, Br<strong>in</strong>chmann-Hansen O, DeckertT (1986). Diabetic control and microvascular complications: the near-normoglycaemic experience.Diabetologia 29: 677–84.Heller SR, Cryer PE (1991). Reduced neuroendocr<strong>in</strong>e and symptomatic responses to subsequenthypoglycemia after 1 episode of hypoglycemia <strong>in</strong> non-diabetic humans. <strong>Diabetes</strong> 40: 223–6.Heller S, Chapman J, McLoud J, Ward J (1995). Unreliability of reports of hypoglycaemia by diabeticpatients. British Medical Journal 310: 440.Hepburn DA, Patrick AW, Ead<strong>in</strong>gton DW, Ew<strong>in</strong>g DJ, Frier BM (1990). Unawareness of hypoglycaemia<strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetic patients: prevalence and relationship to autonomic neuropathy. DiabeticMedic<strong>in</strong>e 7: 711–7.Hermansen K, Fonta<strong>in</strong>e P, Kukolja KK, Peterkova Y, Leth G, Gall MA (2004). Insul<strong>in</strong> analogues(<strong>in</strong>sul<strong>in</strong> detemir and <strong>in</strong>sul<strong>in</strong> aspart) versus traditional human <strong>in</strong>sul<strong>in</strong>s (NPH <strong>in</strong>sul<strong>in</strong> and regularhuman <strong>in</strong>sul<strong>in</strong>) <strong>in</strong> basal-bolus therapy for patients with type 1 diabetes. Diabetologia 47: 622–9.Holl RW, Lang GE, Grabert M, He<strong>in</strong>ze E, Lang GK, Debat<strong>in</strong> KM (1998). Diabetic ret<strong>in</strong>opathy <strong>in</strong>pediatric patients with type 1 diabetes: effect of diabetes duration, prepubertal and pubertal onsetof diabetes, and metabolic control. Journal of Pediatrics 132: 790–4.Home PD, L<strong>in</strong>dholm A, Riis A, European Insul<strong>in</strong> Aspart Study Group (2000). Insul<strong>in</strong> aspart versushuman <strong>in</strong>sul<strong>in</strong> <strong>in</strong> the management of long-term blood glucose control <strong>in</strong> type 1 diabetes mellitus: arandomized controlled trial. Diabetic Medic<strong>in</strong>e 17: 762–70.Hoogma RP, Hammond PJ, Gomis R, Kerr D, Bruttomesso D, Bouter KP et al. 5 Nations StudyGroup (2006). Comparison of the effects of cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion (CSII) andNPH-based multiple daily <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jections (MDI) on glycaemic control and quality of life: resultsof the 5-nations trial. Diabetic Medic<strong>in</strong>e 23: 141–7.Hvidberg A, Fanelli CG, Hershey T, Terkamp C, Craft S, Cryer PE (1996). Impact of recent antecedenthypoglycemia on hypoglycemic cognitive dysfunction <strong>in</strong> non-diabetic humans. <strong>Diabetes</strong> 45: 1030–6.Jorgens V, Grusser M, Bott U, Mulhauser I, Berger M (1993). Effective and safe translation of<strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy to general <strong>in</strong>ternal medic<strong>in</strong>e departments. Diabetologia 36: 99–105.K<strong>in</strong>sley BT, Widom B, Simonson DC (1995). Differential regulation of counterregulatory hormonesecretion and symptoms dur<strong>in</strong>g hypoglycemia <strong>in</strong> IDDM. Effect of glycemic control. <strong>Diabetes</strong> Care18: 17–26.Knight G, Jenn<strong>in</strong>gs AM, Boulton AJ, Toml<strong>in</strong>son S, Ward JD (1985). Severe hyperkalaemia andketoacidosis dur<strong>in</strong>g rout<strong>in</strong>e treatment with an <strong>in</strong>sul<strong>in</strong> pump. British Medical Journal 291: 371–2.Knight G, Boulton AJ, Drury J, Ward JD (1986). Long term glycaemic control by alternative regimens<strong>in</strong> a feasibility study of cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion. <strong>Diabetes</strong> Research 3: 335–8.

188 RISKS OF STRICT GLYCAEMIC CONTROLKorzon-Burakowska A, Hopk<strong>in</strong>s D, Matyka K, Lomas J, Pernet A, Macdonald I, Amiel S (1998).Effects of glycemic control on protective responses aga<strong>in</strong>st hypoglycemia <strong>in</strong> type 2 diabetes.<strong>Diabetes</strong> Care 21: 283–90.Krolewski AS, Laffel LM, Krolewski M, Qu<strong>in</strong>n M, Warram JH (1995). Glycosylated hemoglob<strong>in</strong>and the risk of microalbum<strong>in</strong>uria <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New EnglandJournal of Medic<strong>in</strong>e 332: 1251–5.Lassmann-Vague V, Belicar P, Alessis C, Raccah D, Vialettes B, Vague P (1996). Insul<strong>in</strong> k<strong>in</strong>etics <strong>in</strong>type I diabetic patients treated by cont<strong>in</strong>uous <strong>in</strong>traperitoneal <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion: <strong>in</strong>fluence of anti-<strong>in</strong>sul<strong>in</strong>antibodies. Diabetic Medic<strong>in</strong>e 13: 1051–5.Limbert C, Schw<strong>in</strong>gshandl J, Haas J, Roth R, Borkenste<strong>in</strong> M (1993). Severe hypoglycemia <strong>in</strong> childrenand adolescents with IDDM: frequency and associated factors. Journal of <strong>Diabetes</strong> Complications7: 216–20.Liu D, McManus RM, Ryan EA (1996). Improved counter-regulatory hormonal and symptomaticresponses to hypoglycemia <strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus after 3 months ofless strict glycemic control. Cl<strong>in</strong>ical and Investigative Medic<strong>in</strong>e 19: 71–82.Maran A, Lomas J, Macdonald IA, Amiel SA (1995). Lack of preservation of higher bra<strong>in</strong> functiondur<strong>in</strong>g hypoglycaemia <strong>in</strong> patients with <strong>in</strong>tensively-treated IDDM. Diabetologia 38: 1412–18.Muhlhauser I, Overmann H, Bender R, Bott U, Berger M (1998). Risk factors of severe hypoglycaemia <strong>in</strong>adult patients with type I diabetes – a prospective population based study. Diabetologia 41: 1274–82.Nathan DM, Lach<strong>in</strong> J, Cleary P, Orchard T, Brillon DJ, Backlund JY et al. (2003). Intensive diabetestherapy and carotid <strong>in</strong>tima-media thickness <strong>in</strong> type 1 diabetes mellitus. New England Journal ofMedic<strong>in</strong>e 348: 2294–303.Pampanelli S, Fanelli C, Lalli C, Gofetta M, Del S<strong>in</strong>daco P, Lepore M et al. (1996). Long-term<strong>in</strong>tensive therapy <strong>in</strong> IDDM: effects on HbA 1c , risk for severe and mild hypoglycaemia, status ofcounterregulation and awareness of hypoglycaemia. Diabetologia 39: 677–86.Pedersen-Bjergaard U, Agerholm-Larsen B, Pramm<strong>in</strong>g S, Hougaard P, Thorste<strong>in</strong>sson B (2003). Predictionof severe hypoglycaemia by angiotens<strong>in</strong>-convert<strong>in</strong>g enzyme activity and genotype <strong>in</strong> type 1diabetes. Diabetologia 46: 89–96.Pickup J, Mattock M, Kerry S (2002). Glycaemic control with cont<strong>in</strong>uous subcutaneous <strong>in</strong>sul<strong>in</strong><strong>in</strong>fusion compared with <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jections <strong>in</strong> patients with type 1 diabetes: meta-analysisof randomised controlled trials. British Medical Journal 324: 705–8.Pieber TR, Eugene-Jolch<strong>in</strong>e I, Derobert E (2000). Efficacy and safety of HOE 901 versus NPH <strong>in</strong>sul<strong>in</strong><strong>in</strong> patients with type 1 diabetes. The European Study Group of HOE 901 <strong>in</strong> type 1 diabetes. <strong>Diabetes</strong>Care 23: 157–62.Plank J, Kohler G, Rakovac I, Semlitsch BM, Horvath K, Bock G et al. (2004). Long-term evaluationof a structured outpatient education programme for <strong>in</strong>tensified <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> patients with type1 diabetes: a 12-year follow-up. Diabetologia 47: 1370–75.Rodrigues I, Reid HA, Ismail K, Amiel SA (2005). Indications and efficacy of cont<strong>in</strong>uous subcutaneous<strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion (CSII) therapy <strong>in</strong> type 1 diabetes mellitus: a cl<strong>in</strong>ical audit <strong>in</strong> a specialist service.Diabetic Medic<strong>in</strong>e: 22: 842–9.Ryder RE, Owens DR, Hayes TM, Ghatei MA, Bloom SR (1990). Unawareness of hypoglycaemia and<strong>in</strong>adequate hypoglycaemic counterregulation: no causal relation with diabetic autonomic neuropathy.British Medical Journal 301: 783–7.Samann A, Muhlhauser I, Bender R, Kloos C, Muller UA (2005). Glycaemic control and severehypoglycaemia follow<strong>in</strong>g tra<strong>in</strong><strong>in</strong>g <strong>in</strong> flexible, <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy to enable dietary freedom <strong>in</strong>people with type 1 diabetes: a prospective implementation study. Diabetologia 48: 1965–70.Segal SA, Fanelli CG, Dence CS, Markham J, Videen TO, Paramore DS et al. (2001). Blood-to-bra<strong>in</strong>glucose transport, cerebral glucose metabolism, and cerebral blood flow are not <strong>in</strong>creased afterhypoglycemia. <strong>Diabetes</strong> 50: 1911–7.Shah SC, Malone JL, Simpson NE (1989). A randomized trial of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy <strong>in</strong> newlydiagnosed <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. New England Journal of Medic<strong>in</strong>e 320: 550–4.

REFERENCES 189Simonson DC, Tamborlane WV, DeFronzo RA, Sherw<strong>in</strong> RS (1985a). Intensive <strong>in</strong>sul<strong>in</strong> therapy reducescounterregulatory hormone responses to hypoglycemia <strong>in</strong> patients with type 1 diabetes. Annals ofInternal Medic<strong>in</strong>e 103: 184–90.Simonson DC, Tamborlane WV, Sherw<strong>in</strong> RS, Smith JD, DeFronzo RA (1985b). Improved <strong>in</strong>sul<strong>in</strong>sensitivity <strong>in</strong> patients with type 1 diabetes mellitus after CSII. <strong>Diabetes</strong> 34 (suppl 3): 80–6.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1987). The <strong>Diabetes</strong> Control andComplications Trial (DCCT). Results of a feasibility study. <strong>Diabetes</strong> Care 10: 1–19.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1988). Weight ga<strong>in</strong> associated with<strong>in</strong>tensive therapy <strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. <strong>Diabetes</strong> Care 11: 567–73.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1991). Epidemiology of severe hypoglycemia<strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. American Journal of Medic<strong>in</strong>e 90:450–9.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1993). The effect of <strong>in</strong>tensive treatmentof diabetes on the development and progression of long-term complications <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. New England Journal of Medic<strong>in</strong>e 329: 977–86.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1995a). Adverse events and theirassociation with treatment regimens <strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. <strong>Diabetes</strong> Care18: 1415–27.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1995b). Implementation of treatmentprotocols <strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. <strong>Diabetes</strong> Care 18: 361–76.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1996a). Influence of <strong>in</strong>tensive diabetestreatment on quality-of-life outcomes <strong>in</strong> the <strong>Diabetes</strong> Control and Complications Trial. <strong>Diabetes</strong>Care 19: 195–203.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1996b). The absence of a glycemicthreshold for the development of long-term complications: the perspective of the <strong>Diabetes</strong> Controland Complications Trial. <strong>Diabetes</strong> 45: 1289–98.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1997). Hypoglycemia <strong>in</strong> the <strong>Diabetes</strong>Control and Complications Trial. <strong>Diabetes</strong> 46: 271–86.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1998). Effect of <strong>in</strong>tensive therapy onresidual beta-cell function <strong>in</strong> patients with type 1 diabetes <strong>in</strong> the <strong>Diabetes</strong> Control and ComplicationsTrial. A randomized, controlled trial. Annals of Internal Medic<strong>in</strong>e 128: 517–23.Vague P, Selam JL, Skeie S, De Leeuw I, Elte JW, Haahr H et al. (2003). Insul<strong>in</strong> detemir is associatedwith more predictable glycemic control and reduced risk of hypoglycemia than NPH <strong>in</strong>sul<strong>in</strong> <strong>in</strong>patients with type 1 diabetes on a basal-bolus regimen with premeal <strong>in</strong>sul<strong>in</strong> aspart. <strong>Diabetes</strong> Care26: 590–6.White NH, Skor DA, Cryer PE, Levandoski LA, Bier DM, Santiago JV (1983). Identification of type 1diabetic patients at <strong>in</strong>creased risk for hypoglycemia dur<strong>in</strong>g <strong>in</strong>tensive therapy. New England Journalof Medic<strong>in</strong>e 308: 485–91.Widom B, Simonson DC (1990). Glycemic control and neuropsychologic function dur<strong>in</strong>g hypoglycemia<strong>in</strong> patients with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Annals of Internal Medic<strong>in</strong>e 112:904–12.Work<strong>in</strong>g Group on Hypoglycemia, American <strong>Diabetes</strong> Association (2005). Def<strong>in</strong><strong>in</strong>g and report<strong>in</strong>ghypoglycemia <strong>in</strong> diabetes: a report from the American <strong>Diabetes</strong> Association Workgroup on Hypoglycemia.<strong>Diabetes</strong> Care 28: 1245–9.Writ<strong>in</strong>g team for the <strong>Diabetes</strong> Control and Complications Trial/ Epidemiology of <strong>Diabetes</strong> Interventionsand Complications Research Group (2002). Effect of <strong>in</strong>tensive therapy on the microvascularcomplications of type 1 diabetes mellitus. Journal of the American Medical Association 287: 2563–9.Yki-Jarv<strong>in</strong>en H, Kaupp<strong>in</strong>en-Makel<strong>in</strong> R, Tiikka<strong>in</strong>en M, Vahatalo M, Virtamo H, Nikkila K et al.(2006). Insul<strong>in</strong> glarg<strong>in</strong>e or NPH comb<strong>in</strong>ed with metform<strong>in</strong> <strong>in</strong> type 2 diabetes: the LANMET study.Diabetologia 49: 442–51.

9 <strong>Hypoglycaemia</strong> <strong>in</strong> Childrenwith <strong>Diabetes</strong>Krystyna A. MatykaINTRODUCTIONSub-optimal care of children with type 1 diabetes mellitus carries devastat<strong>in</strong>g consequences.Young children, previously thought to be protected from the early development of microvascularcomplications, have been found to be at significant risk of these complications that canpresent <strong>in</strong> adolescence (Danne et al., 1994; Solders et al., 1997; Schultz et al., 1999). Yetthey are also at risk of detrimental neuropsychological sequelae, which are thought to berelated to recurrent episodes of hypoglycaemia that may accompany <strong>in</strong>tensification of <strong>in</strong>sul<strong>in</strong>therapy aimed at decreas<strong>in</strong>g the risk of these microvascular complications. The exaggeratedmetabolic demands of the grow<strong>in</strong>g child, comb<strong>in</strong>ed with a lifestyle that is unpredictableeven on a day to day basis, make children very vulnerable to both repeated and severeepisodes of hypoglycaemia (Allen et al., 2001). This chapter exam<strong>in</strong>es the aetiology, physiology,consequences and management of episodes of hypoglycaemia dur<strong>in</strong>g this dynamictime of life.DEFINITION OF HYPOGLYCAEMIA IN CHILDHOODThe def<strong>in</strong>ition of hypoglycaemia <strong>in</strong> childhood has been extremely controversial. It hasbeen suggested that children can tolerate lower levels of blood glucose, especially as thedevelop<strong>in</strong>g bra<strong>in</strong> can use alternative substrates for cerebral metabolism. This has beensupported by the cl<strong>in</strong>ical f<strong>in</strong>d<strong>in</strong>g that some children with diabetes appear to be ‘normal’ whentheir blood glucose concentrations are low as demonstrated by home blood glucose monitor<strong>in</strong>g.However, difficulties arise as young children are not expected to perform complexpsychomotor tasks and cl<strong>in</strong>ical detection of mild changes <strong>in</strong> performance can be difficult.This subject is not easy to study because of the ethical difficulties of perform<strong>in</strong>g studies ofnormal glucose homeostasis dur<strong>in</strong>g fast<strong>in</strong>g <strong>in</strong> young children, which could be used to def<strong>in</strong>ethe limits of normality of blood glucose concentrations.Fast<strong>in</strong>g glucose requirements will, <strong>in</strong> part, be determ<strong>in</strong>ed by bra<strong>in</strong> glucose uptake andthe central nervous system has a pivotal role <strong>in</strong> carbohydrate metabolism throughout life.The bra<strong>in</strong> of <strong>in</strong>fants and children can use glucose at a rate of 3–5 mg/kg/m<strong>in</strong>, equivalent toalmost all endogenous production as def<strong>in</strong>ed by stable isotope studies of cerebral glucosemetabolism, and a l<strong>in</strong>ear correlation exists between glucose production and estimated bra<strong>in</strong><strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

192 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESsize, from premature <strong>in</strong>fants to adult life (Bier et al., 1977). There is a marked change<strong>in</strong> the correlation between body weight and hepatic glucose production correspond<strong>in</strong>g tomid-puberty (approximately 40 kg) <strong>in</strong>dicat<strong>in</strong>g that bra<strong>in</strong> growth is virtually complete (Bieret al., 1977). However, the develop<strong>in</strong>g bra<strong>in</strong> also has the ability to use alternative substratesfor cerebral metabolism, and dur<strong>in</strong>g fast<strong>in</strong>g young children have higher concentrations ofketones and lactate when compared to adults (Haymond et al., 1982).A number of studies have exam<strong>in</strong>ed metabolic parameters, <strong>in</strong>clud<strong>in</strong>g blood glucoseconcentrations, follow<strong>in</strong>g fasts of vary<strong>in</strong>g duration <strong>in</strong> children. On the whole these studieshave been used to provide normative data for use <strong>in</strong> the cl<strong>in</strong>ical evaluation of children withpotential disorders of metabolism; as a result the fasts have been prolonged and sampl<strong>in</strong>g<strong>in</strong>frequent. There is little detailed <strong>in</strong>formation on metabolic variables dur<strong>in</strong>g a short fast asis experienced by children with diabetes, for example when they go to sleep. One study ofnocturnal glucose homeostasis <strong>in</strong> 39 children, subsequently found to have a constitutionalshort stature, demonstrated a cyclical variation <strong>in</strong> blood glucose concentrations dur<strong>in</strong>g thenight with a periodicity of 80–120 m<strong>in</strong>utes (Stirl<strong>in</strong>g et al., 1991). At some stage of thenight blood glucose levels fell transiently to below 3 mmol/l <strong>in</strong> 5% of the children, but it isdifficult to know if these children are representative of those who grow normally.Other studies have exam<strong>in</strong>ed glucose and metabolite profiles <strong>in</strong>termittently dur<strong>in</strong>gprolonged fasts of 24–36 hour duration (Chaussa<strong>in</strong>, 1973; Chaussa<strong>in</strong> et al., 1974; Chaussa<strong>in</strong>et al., 1977; Saudubray et al., 1981; Haymond et al., 1982; Kerr et al., 1983; Lamerset al., 1985a; Lamers et al., 1985b). Fast<strong>in</strong>g glucose concentrations varied depend<strong>in</strong>g on theduration of the fast and age of child studied. After 24 hours, blood glucose was found torange from 3.0–3.5 mmol/l. Some of these studies suggested a positive correlation betweenfast<strong>in</strong>g blood glucose concentrations and age (Chaussa<strong>in</strong> et al., 1977; Saudubray et al., 1981;Lamers et al., 1985a).Def<strong>in</strong>ition of <strong>Hypoglycaemia</strong> <strong>in</strong> Childhood <strong>Diabetes</strong>Although there has been great controversy regard<strong>in</strong>g the biochemical def<strong>in</strong>ition of hypoglycaemiafor the diagnosis of pathological states <strong>in</strong> the paediatric population (Koh et al., 1988),these arguments should not apply to type 1 diabetes. The management of type 1 diabetes<strong>in</strong>volves striv<strong>in</strong>g towards the ma<strong>in</strong>tenance of glucose concentrations well with<strong>in</strong> the physiologicalrange rather than merely outside of the pathological range. The only study from thosediscussed above to exam<strong>in</strong>e glucose concentrations after a duration of fast<strong>in</strong>g that wouldrepresent that occurr<strong>in</strong>g as part of normal daily liv<strong>in</strong>g (14 hours) found a fast<strong>in</strong>g glucoseconcentration of 43 ± 01 mmol/l <strong>in</strong> children aged 3–15 years old (Lamers et al., 1985b).Guidel<strong>in</strong>es now suggest that diabetes management should aim to keep plasma glucose above4 mmol/l (‘four should be the floor’ was recommended <strong>in</strong> a <strong>Diabetes</strong> UK report <strong>in</strong> 1996),and this is probably a reasonable recommendation for children.Both <strong>in</strong> research studies and <strong>in</strong> cl<strong>in</strong>ical care, hypoglycaemia is often subdivided <strong>in</strong>todegrees of severity based on the <strong>in</strong>tervention required. An example of such a classificationthat was suggested by the International Society for Paediatric and Adolescent Diabetologists(ISPAD) is shown <strong>in</strong> Table 9.1. It should be noted that the usual adult def<strong>in</strong>ition of ‘mild’(i.e., self-treated) hypoglycaemia cannot be applied <strong>in</strong> young children who rely on theiradult carers for their diabetes management, because they are unlikely to be able to treat theepisodes themselves.

SIGNS AND SYMPTOMS OF HYPOGLYCAEMIA 193Table 9.1 Example of classification of hypoglycaemia (data sourced from ISPAD Consensus Guidel<strong>in</strong>es,2000)GradeMild (Grade 1)Moderate (Grade 2)Severe (Grade 3)DescriptionAware of, responds to and self-treats the hypoglycaemiaCannot respond to hypoglycaemia and requires help fromsomeone else, but oral treatment is successfulSemi-conscious or unconscious or <strong>in</strong> coma ± convulsions andmay require parenteral therapy (glucagon or IV glucose)PREVALENCE OF HYPOGLYCAEMIAAs the def<strong>in</strong>ition of hypoglycaemia has been controversial, studies of prevalence have oftenused variable def<strong>in</strong>itions, both for daytime hypoglycaemia and for that occurr<strong>in</strong>g dur<strong>in</strong>gsleep. <strong>Hypoglycaemia</strong> is notoriously under-reported, as episodes, particularly mild ones, arequickly forgotten. Even severe episodes may be overlooked. The <strong>in</strong>dividual themselves mayhave amnesia for the event and, if it occurred away from their normal environment, nobodymay document what took place.A number of studies have exam<strong>in</strong>ed the prevalence of severe hypoglycaemia <strong>in</strong> thepaediatric population (Table 9.2). Some studies <strong>in</strong>cluded only episodes of coma/convulsionwhereas others also <strong>in</strong>cluded those events <strong>in</strong> which neurological impairment was severeenough to require <strong>in</strong>tervention. <strong>Hypoglycaemia</strong> has been shown to be a significant problembut, given the methodological complexities, prevalence rates of severe hypoglycaemia havevaried greatly from 3.1 to 85.7 episodes per 100 patient-years. It is likely that less severeepisodes are much more common and under-reported.Nocturnal <strong>Hypoglycaemia</strong>Studies have also exam<strong>in</strong>ed the prevalence of nocturnal hypoglycaemia. In these studiesa biochemical rather than a symptomatic def<strong>in</strong>ition of hypoglycaemia was used and hasbeen very variable, rang<strong>in</strong>g from 3.0 to 3.8 mmol/l. The prevalence of hypoglycaemia hasvaried from 10 to 55% but it is noticeable that the more recent studies have detecteda higher prevalence (Beregszaszi et al., 1997; Lopez et al., 1997; Porter et al., 1997;Matyka et al., 1999a). The majority of these episodes do not wake the patient and, <strong>in</strong>everyday life, a great number of episodes of nocturnal hypoglycaemia will be completelyundetected.SIGNS AND SYMPTOMS OF HYPOGLYCAEMIADaytime <strong>Hypoglycaemia</strong>The classical symptoms of hypoglycaemia, as described <strong>in</strong> adults (Chapter 2), are classified<strong>in</strong>to three dist<strong>in</strong>ct groups: autonomic, neuroglycopenic and non-specific (Deary et al.,1993). In adults the sympathoadrenal response dur<strong>in</strong>g hypoglycaemia is primarily responsiblefor the classical autonomic symptoms which alert the <strong>in</strong>dividual to the fall<strong>in</strong>g glucose

Table 9.2 Summary of studies exam<strong>in</strong><strong>in</strong>g prevalence of severe hypoglycaemia s<strong>in</strong>ce the <strong>Diabetes</strong> Control and Complications trial <strong>in</strong> 1993StudyAge(years)Number ofsubjects Def<strong>in</strong>ition Study typeNo. episodes/ 100patient years CorrelationsDCCT (1994) 13–17 195 coma, seizure or requir<strong>in</strong>gassistanceprospective 85.7 lower HbA 1cBognetti et al. (1997) 3.2–25.5 187 retrospective 14.9 younger ageMortensen et al.(1997)Nordfelt andLudvigsson (1997)0–18 2897 coma ± seizures cross-sectional 22 younger age;lower HbA 1c1–18 146 coma ± seizures;prospective 15–19lower age at onset;requir<strong>in</strong>g assistanceprospective 10.1–12.6 longer disease duration;lower HbA 1cprospective 4.8younger age;13.1lower HbA 1cprospective 7.8younger age15.4lower HbA 1cDavis et al. (1997) 0–18 657 coma ± seizures;requir<strong>in</strong>g assistanceDavis et al. (1998) 0–18 709 coma ± seizures;requir<strong>in</strong>g assistanceRosilio et al. (1998) 1–19 2579 coma ± seizures orglucagon <strong>in</strong>jectionTupola et al. (1998b) 1–24 376 coma ± seizures orglucagon <strong>in</strong>jectioncross-sectional 45 lower HbA 1c ;more exercise;more daily bloodglucose measurementsprospective 3.1 lower HbA 1c ;higher <strong>in</strong>sul<strong>in</strong> doseThomsett et al. (1999) 1–19 268 coma ± seizures retrospective 5 younger age;lower HbA 1c ;number of cl<strong>in</strong>ic visitsAllen et al. (2001) 0–34 415 coma prospective 7% of patients4 yrs afterdiagnosis;4% after6.5 yearslower HbA 1c ;older age

Lev<strong>in</strong>e et al. (2001) 7–16 300 episodes requir<strong>in</strong>g outsideassistanceprospective 62 lower HbA 1c ;younger ageRewers et al. (2002) 0–19 1243 coma ± seizures prospective 19 young age;<strong>in</strong>creased diabetesduration;lower HbA 1c ;under<strong>in</strong>suranceCraig et al. (2002) 1.2–15.8 1190 coma ± seizures prospective 36 younger age;males;> 3 <strong>in</strong>sul<strong>in</strong><strong>in</strong>jections/day;longer diabetes durationHoll et al. (2003) 11–18 872 coma ± seizures cross-sectional 17.9 not evaluatedVanelli et al. (2005) 1.6–17.1 3560 coma ± seizures ;episodes requir<strong>in</strong>gparenteral therapy;episodes requir<strong>in</strong>g outsideassistancecross-sectional 17.6 none apparentWagner et al. (2005) 0–9 6309 episodes requir<strong>in</strong>g outsideassistanceprospective 22.6 younger age;longer diabetes duration;higher <strong>in</strong>sul<strong>in</strong> dose;<strong>in</strong>sul<strong>in</strong> regimen;centre experience

196 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESconcentration so that corrective action can be taken. The classical autonomic symptomsoccur <strong>in</strong> adults between 3.0 and 3.6 mmol/l and <strong>in</strong>clude: sweat<strong>in</strong>g, palpitations, hunger andshak<strong>in</strong>g. If glucose concentrations fall further, neuroglycopenia will develop, usually at aglucose concentration around 2.8 mmol/l, and if it falls further the <strong>in</strong>dividual may not be ableto correct the hypoglycaemia themselves. The most common neuroglycopenic symptomsare: confusion, drows<strong>in</strong>ess, odd behaviour, speech difficulty and <strong>in</strong>coord<strong>in</strong>ation. If bloodglucose cont<strong>in</strong>ues to fall, coma or convulsion could ensue although the glycaemic thresholdat which this occurs is not certa<strong>in</strong>.The symptoms of hypoglycaemia <strong>in</strong> childhood differ from adults. In one study, childrenand parents were asked which symptoms and signs alerted them to an episode of hypoglycaemia(McCrimmon et al., 1995). The authors reported that symptoms did not separate <strong>in</strong>todist<strong>in</strong>ct autonomic and neuroglycopenic categories and behavioural symptoms were prom<strong>in</strong>ent.Another study exam<strong>in</strong>ed the frequency of hypoglycaemia <strong>in</strong> a group of children andadolescents us<strong>in</strong>g a three-month diary (Tupola et al., 1998a). Episodes of hypoglycaemia(def<strong>in</strong>ed as a blood glucose ≤ 3 mmol/l) were documented along with symptoms. Of thepatients (aged 2.5–21 years), 52% had a total of 287 episodes of hypoglycaemia, the majorityof which (77%) were mild. The most common present<strong>in</strong>g symptoms were weakness, tremor,hunger and drows<strong>in</strong>ess; 39% of symptoms were classified as ‘adrenergic’ (autonomic) and61% as neuroglycopenic or non-specific behavioural (Table 9.3). The dom<strong>in</strong>ant symptomswere different <strong>in</strong> different age groups of children. Those children less than six years ofage had fewer autonomic symptoms than adolescents, with the commonest symptom be<strong>in</strong>gdrows<strong>in</strong>ess <strong>in</strong> young children and tremor <strong>in</strong> the older children (Tupola et al., 1998a).Nocturnal <strong>Hypoglycaemia</strong>The majority of episodes of nocturnal hypoglycaemia do not awaken the child from sleepand thus go undetected, even <strong>in</strong> those who have normal awareness of hypoglycaemiadur<strong>in</strong>g wak<strong>in</strong>g hours (Gale and Tattersall, 1979; Bendtson et al., 1993; Porter et al., 1996;Table 9.3 Reported symptoms dur<strong>in</strong>g 221 episodes of mild hypoglycaemia(data sourced from Tupola et al., 1998b)Symptom Prevalence (%)autonomic tremor 22hunger 14sweat<strong>in</strong>g 4neuroglycopenic drows<strong>in</strong>ess 34irritability/aggressiveness 4dizz<strong>in</strong>ess 1poor concentration 1blurred vision 1non-specific weakness 7nausea 7abdom<strong>in</strong>al pa<strong>in</strong> 2headache 2tearfulness 1

RISK FACTORS FOR HYPOGLYCAEMIA 197Beregszaszi et al., 1997; Lopez et al., 1997; Matyka et al., 1999a). The reasons for thislack of awareness of nocturnal episodes of hypoglycaemia are unclear, but are discussed<strong>in</strong> Chapter 4. As symptom generation depends on <strong>in</strong>tact autonomic and counterregulatorydefence mechanisms, it has been postulated that dim<strong>in</strong>ished counterregulation overnight mayresult <strong>in</strong> lack of arousal when hypoglycaemia occurs dur<strong>in</strong>g sleep (Bendtson et al., 1993;Jones et al., 1998).RISK FACTORS FOR HYPOGLYCAEMIAThe discussion of risk factors naturally overlaps with that of prevention of hypoglycaemia,and a fuller review of specific <strong>in</strong>terventions is provided <strong>in</strong> the section entitled ‘Prevention’ onpage 207. Table 9.3 presents the results of studies that have been performed to exam<strong>in</strong>e theprevalence of severe hypoglycaemia <strong>in</strong> childhood s<strong>in</strong>ce the <strong>Diabetes</strong> Control and ComplicationsTrial (DCCT) was published (The <strong>Diabetes</strong> Control and Complications Trial ResearchGroup, 1993). Although a number of studies have suggested correlations between youngerage and strict glycaemic control, it is important to note that many <strong>in</strong>dividual episodes ofhypoglycaemia may be expla<strong>in</strong>ed by missed meals or unplanned exercise, but this wouldnot have been addressed <strong>in</strong> epidemiological studies.Glycaemic ControlInsul<strong>in</strong> requirements vary with age and are approximately 0.5–1 U/kg/day before puberty and1.5–2 U/kg/day dur<strong>in</strong>g adolescence, reflect<strong>in</strong>g the <strong>in</strong>sul<strong>in</strong> resistance that is present dur<strong>in</strong>g thisperiod of rapid growth and development (Dunger, 1992). Despite numerous developments<strong>in</strong> terms of novel <strong>in</strong>sul<strong>in</strong> preparations and modes of delivery, people with type 1 diabetesstill experience vary<strong>in</strong>g states of <strong>in</strong>sul<strong>in</strong> deficiency or excess that are difficult to controland predict. This is probably most evident <strong>in</strong> adolescents with type 1 diabetes <strong>in</strong> whomperipheral hyper<strong>in</strong>sul<strong>in</strong>aemia is achieved <strong>in</strong> an attempt to replace adequate levels of <strong>in</strong>sul<strong>in</strong><strong>in</strong> the portal circulation dur<strong>in</strong>g the pubertal growth spurt (Dunger, 1992).The DCCT highlighted the dilemma faced by all patients with type 1 diabetes (The<strong>Diabetes</strong> Control and Complications Trial Research Group, 1993). Attempts at improv<strong>in</strong>gglycaemic control, by <strong>in</strong>tensify<strong>in</strong>g diabetes management, <strong>in</strong> an effort to decrease the likelihoodof the long-term microvascular complications of diabetes led to a significant <strong>in</strong>crease<strong>in</strong> the risk of severe hypoglycaemia (SH). In the DCCT, subjects <strong>in</strong> the <strong>in</strong>tensified treatmentgroup had a three-fold higher risk of SH (The <strong>Diabetes</strong> Control and Complications TrialResearch Group, 1993). A group of 195 adolescents, aged between 13 and 17 years, took part<strong>in</strong> this trial (The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1994). Althoughthe benefits of improved glycaemic control <strong>in</strong> terms of microvascular complications werestill significant, the adolescents found it more difficult to achieve the low HbA 1c concentrationthan adults (806 ± 013 versus 712 ± 003%; p

198 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESWe do not have rights to reproduce thisfigure electronicallyFigure 9.1 Rates of (a) Severe hypoglycaemia, and (b) average HbA 1c by calender year. Reproducedfrom Bulsara et al. (2004) with permission from The American <strong>Diabetes</strong> Associationhypoglycaemia was as common <strong>in</strong> those centres where metabolic control was poor, as <strong>in</strong>those centres that achieved better control as judged by HbA 1c , suggest<strong>in</strong>g that research doesnot always reflect cl<strong>in</strong>ical experience (Mortensen et al., 1997).S<strong>in</strong>ce 1993, ample opportunity has been present to ref<strong>in</strong>e the approaches to <strong>in</strong>tensive<strong>in</strong>sul<strong>in</strong> therapy and to improve education both for patients and physicians. Longitud<strong>in</strong>alstudies of the <strong>in</strong>cidence of hypoglycaemia are unusual but one audit study from a largepaediatric cl<strong>in</strong>ic <strong>in</strong> Western Australia demonstrated an <strong>in</strong>terest<strong>in</strong>g trend <strong>in</strong> <strong>in</strong>cidence of SHover a period of ten years (Bulsara et al., 2004). Over the first five years of the study,the <strong>in</strong>cidence <strong>in</strong>creased by 29% <strong>in</strong> conjunction with a decl<strong>in</strong>e <strong>in</strong> the average HbA 1c ofabout 0.2% per year. Despite a cont<strong>in</strong>ued improvement <strong>in</strong> glycaemic control, the <strong>in</strong>cidenceof SH appeared to plateau at this cl<strong>in</strong>ical centre suggest<strong>in</strong>g that improved diabetesmanagement, from more effective <strong>in</strong>sul<strong>in</strong> regimens or better education, can improve bloodglucose concentrations without a concomitant <strong>in</strong>crease <strong>in</strong> <strong>in</strong>cidence of hypoglycaemia(Figure 9.1).Nocturnal Insul<strong>in</strong> RequirementsThe mismatch of <strong>in</strong>sul<strong>in</strong> delivery and <strong>in</strong>sul<strong>in</strong> requirements on standard <strong>in</strong>sul<strong>in</strong> regimensis particularly evident dur<strong>in</strong>g the night and most episodes of severe hypoglycaemia occurdur<strong>in</strong>g sleep (Edge et al., 1990a; The <strong>Diabetes</strong> Control and Complications Trial ResearchGroup, 1997). Current <strong>in</strong>sul<strong>in</strong> replacement regimens tend to result <strong>in</strong> hyper<strong>in</strong>sul<strong>in</strong>aemia <strong>in</strong>the early part of the night, although physiological <strong>in</strong>sul<strong>in</strong> requirement is at its nadir between24:00–03:00 hours, and so exacerbates the risk of hypoglycaemia at this time (Matykaet al., 1999a; Mohn et al., 1999; Ford-Adams et al., 2003). Insul<strong>in</strong> requirements then peakbetween 04:00–08:00 hours and a ‘dawn phenomenon’ occurs which can lead to fast<strong>in</strong>ghyperglycaemia (Bolli and Gerich, 1984; Edge et al., 1990b) and is thought to result fromGH secretion dur<strong>in</strong>g the later part of the night (De Feo et al., 1990; Edge et al., 1990b)

RISK FACTORS FOR HYPOGLYCAEMIA 199exacerbated further by the delayed effects of daytime physical activity on muscle glucosemetabolism and the prolonged period of fast<strong>in</strong>g that occurs overnight, especially <strong>in</strong> youngchildren. This suggests that the overnight period is the time of greatest hypoglycaemia risk(The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1997).Intensive Insul<strong>in</strong> RegimensFew studies have exam<strong>in</strong>ed the impact of <strong>in</strong>sul<strong>in</strong> regimen on the risk of hypoglycaemia<strong>in</strong> children. The DCCT did f<strong>in</strong>d a significantly higher risk of hypoglycaemia among the195 adolescents who participated <strong>in</strong> the study although this was a comparison of overallglycaemic control and not of specific regimens (The <strong>Diabetes</strong> Control and ComplicationsTrial Research Group, 1994). A number of studies exam<strong>in</strong><strong>in</strong>g prevalence of hypoglycaemiahave found an <strong>in</strong>verse correlation between hypoglycaemia risk and glycated haemoglob<strong>in</strong>(Table 9.2). The Hvidore Study Group has formed a collaboration between 21 <strong>in</strong>ternationalpaediatric centres from 18 countries (Holl et al., 2003). The group surveyed paediatricdiabetes management of 2873 children aged up to 18 years <strong>in</strong> 1995 and restudied 872 ofthese children <strong>in</strong> 1998. Although the use of multiple <strong>in</strong>jection regimens <strong>in</strong>creased from 42%to 71% this did not lead to an improvement <strong>in</strong> glycaemic control as judged by glycatedhaemoglob<strong>in</strong> concentrations. Although there was a tendency towards an <strong>in</strong>crease <strong>in</strong> thefrequency of severe hypoglycaemia <strong>in</strong> the group of children/adolescents who had had an<strong>in</strong>tensification of <strong>in</strong>sul<strong>in</strong> regimen, this did not reach statistical significance, perhaps becauseof the low number of events recorded (Holl et al., 2003).Another study of more than 6000 children has suggested that <strong>in</strong>jection regimen and centreexperience, as judged by the size of the cl<strong>in</strong>ic, may be significant risk factors for severehypoglycaemia (Wagner et al., 2005). In this study of children aged up to n<strong>in</strong>e years, an<strong>in</strong>creased risk of hypoglycaemia was observed <strong>in</strong> those children tak<strong>in</strong>g four <strong>in</strong>sul<strong>in</strong> <strong>in</strong>jectionsdaily or on <strong>in</strong>sul<strong>in</strong> pump therapy, compared to those children tak<strong>in</strong>g one to three <strong>in</strong>jectionsdaily.It is important to note that even the more recent studies do not <strong>in</strong>clude data acquired s<strong>in</strong>cethe <strong>in</strong>troduction of the <strong>in</strong>sul<strong>in</strong> analogues or the use of more physiological and <strong>in</strong>tensive<strong>in</strong>sul<strong>in</strong> regimens. Small studies of a few children <strong>in</strong> which <strong>in</strong>sul<strong>in</strong> analogues have beencompared with human <strong>in</strong>sul<strong>in</strong>s suggest that the risk of hypoglycaemia may be lower withanalogues without compromis<strong>in</strong>g glycaemic control (see later section on hypoglycaemiaprevention on page 209).DietChildren with type 1 diabetes have the same nutritional requirements as children withoutdiabetes. However, meals and snacks conta<strong>in</strong><strong>in</strong>g a high proportion of carbohydrate, have tobe regularly distributed throughout the day to avoid the extremes of hypo- and hyperglycaemia(Magrath et al., 1993). This can be an issue for children who do not want to bedifferent from their peers and do not want to eat at times when there friends are play<strong>in</strong>g.Toddlers can present a special problem as many do not eat regular meals but graze dur<strong>in</strong>gthe day.Surpris<strong>in</strong>gly, there has been little systematic study of the role of both quantity andquality of dietary components on the risk of hypoglycaemia. However, some studies of the

200 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESprevalence of hypoglycaemia have found that a number of episodes can be attributed tomissed meals (Daneman et al., 1989; Davis et al., 1997; Tupola et al., 1998b). The impactof dietary <strong>in</strong>terventions <strong>in</strong> the avoidance of hypoglycaemia, ma<strong>in</strong>ly dur<strong>in</strong>g sleep, have beenexam<strong>in</strong>ed, and this is discussed later <strong>in</strong> the section on hypoglycaemia prevention.Physical ActivityAs early as 1926, it was found that exercise could potentiate the hypoglycaemic effect of<strong>in</strong>sul<strong>in</strong> <strong>in</strong> patients with type 1 diabetes (Lawrence, 1926). Dur<strong>in</strong>g the first 5–10 m<strong>in</strong>utes ofexercis<strong>in</strong>g, muscle glycogen is used as the primary source of energy (Price et al., 1994).Subsequently, fuel is provided <strong>in</strong>creas<strong>in</strong>gly by circulat<strong>in</strong>g glucose, through gluconeogenesisand free fatty acids (FFAs), the release of which are under hormonal control and dependentpredom<strong>in</strong>antly on the portal glucagon : <strong>in</strong>sul<strong>in</strong> ratio (Ahlborg et al., 1974). The acute effectsof exercise are followed by restoration of the metabolic milieu. Muscle glucose uptakerema<strong>in</strong>s elevated as glycogen stores are replenished and although <strong>in</strong>sul<strong>in</strong> sensitivity isenhanced <strong>in</strong> the period after exercise, <strong>in</strong>creased glucose uptake by skeletal muscle can occureven <strong>in</strong> the absence of <strong>in</strong>sul<strong>in</strong> (Cartee and Holloszy, 1990). The time taken to restore muscleglycogen to pre-exercise levels will depend on the <strong>in</strong>tensity and duration of the exerciseperformed, and the tim<strong>in</strong>g and amount of dietary carbohydrate <strong>in</strong>take, but it can take severalhours – typically 6–20 hours (Ivy and Holloszy, 1981; Richter, 1996).Until recently there has been little systematic study of the impact of exercise <strong>in</strong> childhoodon glucose homeostasis. One study used cont<strong>in</strong>uous glucose monitor<strong>in</strong>g to study a standardisedexercise protocol <strong>in</strong> a group of children who were us<strong>in</strong>g cont<strong>in</strong>uous subcutaneous<strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion (CSII). Glucose profiles were exam<strong>in</strong>ed both dur<strong>in</strong>g and after exercise ona cycle ergometer with the <strong>in</strong>fusion pump either switched on or off (Admon et al., 2005).<strong>Hypoglycaemia</strong> was more common after exercise than dur<strong>in</strong>g it, and this was true whetherCSII was on or off. All subjects had one to three episodes of symptomatic hypoglycaemiawith<strong>in</strong> 2.5 to 12 hours after exercise and four subjects had asymptomatic hypoglycaemiadur<strong>in</strong>g exercise, only one of whom had consumed extra carbohydrate because their preexerciseblood glucose had been below 5.5 mmol/l. Another study exam<strong>in</strong>ed the impact ofdaytime exercise on overnight blood glucose profiles <strong>in</strong> 50 subjects aged 10–18 years on<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimens (Tsalikian et al., 2005). On one occasion they were studied dur<strong>in</strong>ga day of afternoon exercise, <strong>in</strong>volv<strong>in</strong>g four periods of 15 m<strong>in</strong>utes each on a treadmill at aheart rate estimated to be 55% of maximum effort for this age group, and on a separateoccasion dur<strong>in</strong>g a rest day. In this study, 22% of subjects developed hypoglycaemia dur<strong>in</strong>gexercise; overnight hypoglycaemia was more common dur<strong>in</strong>g the night after the afternoonexercise than dur<strong>in</strong>g a night after the rest day (Tsalikian et al., 2005).AgeStudies of prevalence of hypoglycaemia have consistently found that younger children,especially those under the age of five years, are at <strong>in</strong>creased risk of hypoglycaemia. Thismay be a consequence of <strong>in</strong>creased <strong>in</strong>sul<strong>in</strong> sensitivity, irregular eat<strong>in</strong>g patterns or impairedsymptomatic awareness.

COUNTERREGULATION IN CHILDHOOD 201GeneticsRecent studies suggest that molecular markers may <strong>in</strong>fluence hypoglycaemia risk. A polymorphism<strong>in</strong> the gene encod<strong>in</strong>g for angiotens<strong>in</strong> convert<strong>in</strong>g enzyme (ACE) has been described<strong>in</strong>dicat<strong>in</strong>g the presence (<strong>in</strong>sertion, I) or absence (deletion, D) of a 287 base pair sequencewith<strong>in</strong> <strong>in</strong>tron 16 result<strong>in</strong>g <strong>in</strong> three genotypes: II, ID and DD. These genotypes are stronglyrelated to serum ACE concentration with the highest values <strong>in</strong> DD and the lowest values<strong>in</strong> II genotypes (Rigat et al., 1990). Danish studies of adults with type 1 diabetes observedthat patients with an II (<strong>in</strong>sertion) genotype, who had a low serum ACE activity, had asignificantly lower frequency of severe hypoglycaemia (Pedersen-Bjergaard et al., 2001;Pedersen-Bjergaard et al., 2003). A Swedish study of children with type 1 diabetes hasreported a six-fold lower frequency of severe hypoglycaemia <strong>in</strong> those patients who had lowserum ACE activity (Nordfelt and Samuelsson, 2003). However, a study of 585 children andadolescents <strong>in</strong> an Australian centre has not confirmed this association (Bulsara et al., 2007),so this rema<strong>in</strong>s controversial.Cl<strong>in</strong>ic ExperienceRecent therapeutic developments, with the availability of novel <strong>in</strong>sul<strong>in</strong> analogues and thegreater use of <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimens, have required re-education not only of patientsbut also the multidiscipl<strong>in</strong>ary team. Few studies have exam<strong>in</strong>ed the impact of the cl<strong>in</strong>icstructure on risk of hypoglycaemia. One large multicentre study <strong>in</strong> Germany did show al<strong>in</strong>k between small cl<strong>in</strong>ic size (< 50 children) and an <strong>in</strong>creased risk of severe hypoglycaemia(Wagner et al., 2005). Another longitud<strong>in</strong>al study of prevalence of severe hypoglycaemia<strong>in</strong> Australia showed an <strong>in</strong>crease <strong>in</strong> hypoglycaemia risk as the mean glycatedhaemoglob<strong>in</strong> <strong>in</strong> the cl<strong>in</strong>ic decl<strong>in</strong>ed. However, over the f<strong>in</strong>al five years of the study the riskof hypoglycaemia plateaued while the HbA 1c cont<strong>in</strong>ued to decrease, suggest<strong>in</strong>g the possiblebenefit of familiarity among patients, healthcare professionals or both (Bulsara et al., 2004)(Figure 9.1).COUNTERREGULATION IN CHILDHOODThe physiology of counterregulation is the subject of Chapters 1 and 6, but a brief overviewof studies of counterregulation <strong>in</strong> childhood will be presented here. Despite the ethical andpractical problems of <strong>in</strong>duc<strong>in</strong>g hypoglycaemia <strong>in</strong> children for research purposes, a numberof studies have been performed (Amiel et al., 1987; Brambilla et al., 1987; S<strong>in</strong>ger-Granicket al., 1988; Hoffman et al., 1991; Jones et al., 1991; Bjorgaas et al., 1997a; Ross et al.,2005). These studies have all exam<strong>in</strong>ed experimentally-<strong>in</strong>duced hypoglycaemia, either us<strong>in</strong>gan <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion method or a hyper<strong>in</strong>sul<strong>in</strong>aemic, hypoglycaemic glucose clamp technique.The results of <strong>in</strong>dividual hormonal responses are discussed briefly.GlucagonAs <strong>in</strong> adults the glucagon response dur<strong>in</strong>g hypoglycaemia is lost <strong>in</strong> children with diabetes(Amiel et al., 1987; Jones et al., 1991; Ross et al., 2005). This is also the case <strong>in</strong> toddlers,

202 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESaged 18–57 months old, who have a very short duration of diabetes (Brambilla et al., 1987).This means that <strong>in</strong>dividuals with diabetes are more reliant on adequate ep<strong>in</strong>ephr<strong>in</strong>e responsesto correct hypoglycaemia.Ep<strong>in</strong>ephr<strong>in</strong>eThe majority of studies suggest that children have exaggerated ep<strong>in</strong>ephr<strong>in</strong>e responses tohypoglycaemia, with peak values that are two-fold higher than those found <strong>in</strong> adults (Amielet al., 1987). Data from prepubertal children have been analysed separately from pubertalchildren and no significant differences were found <strong>in</strong> ep<strong>in</strong>ephr<strong>in</strong>e responses (Amiel et al.,1987; Ross et al., 2005). One study, us<strong>in</strong>g an <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion as opposed to a hyper<strong>in</strong>sul<strong>in</strong>aemicclamp, did suggest that ep<strong>in</strong>ephr<strong>in</strong>e responses were blunted <strong>in</strong> a group ofpoorly-controlled adolescents (Bjorgaas et al., 1997a). The reason for the discrepancy <strong>in</strong>results is not clear.Glucose thresholds for counterregulatory responses have received very little attention.In a study of poorly-controlled adolescents (average total HbA 1 : 15%) glucose thresholdsfor ep<strong>in</strong>ephr<strong>in</strong>e secretion were significantly higher, with the poorly-controlled adolescentsreleas<strong>in</strong>g ep<strong>in</strong>ephr<strong>in</strong>e at a glucose concentration of 4.7 mmol/l compared to 3.9 mmol/l <strong>in</strong>healthy adolescents (Jones et al., 1991).Growth HormoneNone of the reported studies have identified defects <strong>in</strong> GH release dur<strong>in</strong>g hypoglycaemia.Generally GH has been found to <strong>in</strong>crease <strong>in</strong> response to hypoglycaemia both <strong>in</strong> childrenwith diabetes and <strong>in</strong> non-diabetic controls (Amiel et al., 1987; Brambilla et al., 1987; Joneset al., 1991).CortisolCortisol, like growth hormone, becomes more important as hypoglycaemia becomesprolonged. Studies of hypoglycaemia <strong>in</strong> children have shown variable results. Brambillafound no <strong>in</strong>crease <strong>in</strong> cortisol <strong>in</strong> either the diabetic or control group of toddlers studied(Brambilla et al., 1987), whereas others have documented an <strong>in</strong>crease <strong>in</strong> cortisol both <strong>in</strong>children with and without diabetes (Amiel et al., 1987; Jones et al., 1991).Effect of Sleep Stage on CounterregulationOne study of nocturnal hypoglycaemia <strong>in</strong> prepubertal children on conventional <strong>in</strong>sul<strong>in</strong>regimens, found that the median glucose nadir dur<strong>in</strong>g episodes of nocturnal hypoglycaemiawas 1.9 mmol/l (range: 1.1–3.3 mmol/l) and the median duration was 270 m<strong>in</strong>utes (range:30–630 m<strong>in</strong>utes) (Matyka et al., 1999a). This is similar to adults <strong>in</strong> whom the averageduration of hypoglycaemia (glucose below 2 mmol/l) dur<strong>in</strong>g the night was found to be threehours <strong>in</strong> a group of adults with poorly-controlled diabetes (Gale and Tattersall, 1979).Conventional wisdom would argue that prolonged episodes of hypoglycaemia are unusualas hypoglycaemia is promptly corrected by counterregulatory defence mechanisms. However,

CONSEQUENCES OF HYPOGLYCAEMIA 203studies of nocturnal hypoglycaemia suggest that counterregulatory responses may be blunteddur<strong>in</strong>g sleep (see Chapter 4). A study of spontaneous nocturnal hypoglycaemia <strong>in</strong> prepubertalchildren with diabetes demonstrated blunted and delayed counterregulatory hormoneresponses dur<strong>in</strong>g sleep (Matyka et al., 1999b). Peak ep<strong>in</strong>ephr<strong>in</strong>e response was only 0.9 nmol/land there was a marked delay between the glucose fall<strong>in</strong>g below 3.5 mmol/l and a significantrise <strong>in</strong> ep<strong>in</strong>ephr<strong>in</strong>e; the mean delay was 170 m<strong>in</strong>utes. Another study exam<strong>in</strong>ed counterregulatoryhormone responses dur<strong>in</strong>g hypoglycaemia that was experimentally-<strong>in</strong>duced dur<strong>in</strong>gthe time of night when slow wave sleep predom<strong>in</strong>ates, and these responses were comparedwith those to hypoglycaemia <strong>in</strong>duced dur<strong>in</strong>g the day with the subjects awake and then aga<strong>in</strong>when they were awake dur<strong>in</strong>g the night (Jones et al., 1998). Adolescents with diabetesand healthy controls participated <strong>in</strong> the study. Ep<strong>in</strong>ephr<strong>in</strong>e responses dur<strong>in</strong>g hypoglycaemiawere blunted when hypoglycaemia was <strong>in</strong>duced dur<strong>in</strong>g slow wave sleep compared to whenhypoglycaemia was <strong>in</strong>duced when subjects were awake dur<strong>in</strong>g the day or dur<strong>in</strong>g the night(Jones et al., 1998). Studies of the physiology of sleep have demonstrated both variations<strong>in</strong> autonomic tone and cerebral glucose metabolism go<strong>in</strong>g from slow wave sleep through torapid eye movement sleep which may <strong>in</strong>fluence the counterregulatory response dur<strong>in</strong>g sleep(Maquet et al., 1990; Parmeggiani and Morrison, 1990).CONSEQUENCES OF HYPOGLYCAEMIACognitive ImpairmentSevere hypoglycaemia can cause catastrophic cerebral damage when it is profound andprolonged, and <strong>in</strong> very young children this may be a risk associated with a variety of causes(Lucas et al., 1988). Glucose is critical not only as the major fuel for cerebral metabolism butalso as a precursor of essential substrates which are essential for normal bra<strong>in</strong> development(Glazer and Weber, 1971). Concerns have been raised that recurrent hypoglycaemia couldaffect long-term academic achievement <strong>in</strong> children with type 1 diabetes, from the effectsof hypoglycaemia disrupt<strong>in</strong>g school performance to the possibility of damage accumulat<strong>in</strong>gover time.Acute effectsFew studies have studied the impact of acute hypoglycaemia on cognitive performance <strong>in</strong> childrenor adolescents. One study of the effects of experimentally-<strong>in</strong>duced mild hypoglycaemia(3.1–3.6 mmol/l) found decrements <strong>in</strong> tests of mental flexibility and on measures that requiredplann<strong>in</strong>g and decision mak<strong>in</strong>g and a rapid response, although the results were variable with<strong>in</strong>subjects (Ryan et al., 1990). The learn<strong>in</strong>g ability of children, who spend much of their day atschool assimilat<strong>in</strong>g <strong>in</strong>formation, could be seriously compromised if they experience frequentepisodes of even mild hypoglycaemia dur<strong>in</strong>g their time <strong>in</strong> class. Children can also be affectedif they miss lessons because of severe hypoglycaemic events and could be compromised byasymptomatic episodes of hypoglycaemia that go undetected and therefore untreated. Thereis some evidence to suggest that children may be particularly susceptible to mild episodes ofhypoglycaemia – studies exam<strong>in</strong><strong>in</strong>g P300 evoked potentials and EEG changes <strong>in</strong> response to

204 HYPOGLYCAEMIA IN CHILDREN WITH DIABETEShypoglycaemia have found that abnormalities commence at a higher blood glucose level <strong>in</strong>children than <strong>in</strong> adults (Jones et al., 1995; Bjorgaas et al., 1998).The effect of nocturnal hypoglycaemia on cognitive function has received little attention.Cognitive performance could be directly affected by hypoglycaemia, but also by sleepdisturbance. The few studies of nocturnal hypoglycaemia <strong>in</strong> children and adolescents withtype 1 diabetes have shown no deleterious effect on overall sleep physiology (Bendtsonet al., 1992; Porter et al., 1996; Matyka et al. 2000, Pillar et al., 2003) or on cognitivefunction the follow<strong>in</strong>g morn<strong>in</strong>g.Long-term effectsEarly studies of children with diabetes suggested that they were ‘mentally superior’ (Westet al., 1934). However, <strong>in</strong> recent years concern has been expressed about the potential impactof recurrent episodes of hypoglycaemia on <strong>in</strong>tellectual performance. It is beyond the scopeof this chapter to provide a comprehensive review on this topic and only more recent studiesare reviewed.The develop<strong>in</strong>g bra<strong>in</strong> is extremely vulnerable to all types of cerebral trauma. Studies ofchildren who have experienced closed head <strong>in</strong>juries suggest that the consequences may bedelayed, with subtle cerebral damage becom<strong>in</strong>g evident with time as normal developmentalmilestones are delayed. A large number of studies have exam<strong>in</strong>ed the impact of recurrenthypoglycaemia <strong>in</strong> childhood (Ryan et al., 1985; Golden et al., 1989; Bjorgaas et al., 1997b;Hershey et al., 1999; Rovet and Ehrlich, 1999; Northam et al., 2001; Wysocki et al.,2003). Almost without exception, the results have shown a possible l<strong>in</strong>k between severehypoglycaemia and decrements <strong>in</strong> cognitive performance and that those children most atrisk of cognitive impairment have been those diagnosed early <strong>in</strong> life – usually less than fiveyears of age (Ryan et al., 1985; Golden et al., 1989; Bjorgaas et al., 1997b; Northam et al.,2001). Deficiencies have been found <strong>in</strong> several cognitive doma<strong>in</strong>s but are more likely <strong>in</strong>those orig<strong>in</strong>at<strong>in</strong>g <strong>in</strong> the frontal lobe. One of the most impressive longitud<strong>in</strong>al studies hasbeen that of Northam and colleagues. Cognitive performance was assessed <strong>in</strong> a large groupof children (123 at basel<strong>in</strong>e) with newly-diagnosed diabetes, aged 3–14 years, who werecompared to healthy controls at three months, two years and six years follow<strong>in</strong>g diagnosis(Northam et al., 2001). At six years, children with diabetes performed more poorly <strong>in</strong>measures of <strong>in</strong>telligence, attention, process<strong>in</strong>g speed, long-term memory and executive skills.Attention, process<strong>in</strong>g speed and executive skills were especially affected <strong>in</strong> those childrenwho had developed diabetes when less than four years of age (Northam et al., 2001). Severehypoglycaemia was associated with lower verbal and full-scale <strong>in</strong>telligence quotient (IQ)scores. The authors concluded that recurrent hypoglycaemia was a potential explanation forthese cognitive deficits but could not exclude an effect of chronic hyperglycaemia.Another study has performed a cognitive test battery and structural neuroimag<strong>in</strong>g us<strong>in</strong>gMagnetic Resonance Imag<strong>in</strong>g <strong>in</strong> a group of young adults with type 1 diabetes and comparedthe f<strong>in</strong>d<strong>in</strong>gs between those with either early onset disease, def<strong>in</strong>ed as less than sevenyears, and late onset disease (Ferguson et al., 2005). Physiological risk variables such asdiabetes duration, evidence of microvascular disease and retrospective report<strong>in</strong>g of preced<strong>in</strong>gexposure to severe hypoglycaemia were also assessed. The patients with early onset diabeteswere found to have deficits <strong>in</strong> non-verbal <strong>in</strong>telligence, <strong>in</strong>formation process<strong>in</strong>g ability andpsychomotor speed. The authors also found higher ventricular volumes and higher frequency

CONSEQUENCES OF HYPOGLYCAEMIA 205of mild ventricular atrophy <strong>in</strong> those with early onset diabetes. None of the f<strong>in</strong>d<strong>in</strong>gs wererelated to the presence of microvascular disease or diabetes duration, suggest<strong>in</strong>g that thecumulative effects of hyperglycaemia were unlikely to be causative. However, no def<strong>in</strong>ite l<strong>in</strong>kwas confirmed between exposure to severe hypoglycaemia and any of these defects, althoughthe study was limited by retrospective report<strong>in</strong>g of the history of preced<strong>in</strong>g hypoglycaemiaover a long period of time (Ferguson et al., 2005). Another study did not f<strong>in</strong>d a difference<strong>in</strong> tests of cognitive function over a period of 18 months <strong>in</strong> 142 patients tak<strong>in</strong>g part <strong>in</strong> astudy exam<strong>in</strong><strong>in</strong>g the impact of <strong>in</strong>tensive therapy versus usual care (Wysocki et al., 2003).The study group had a wide age range from 6–15 years and data from younger childrenwere not analysed separately.Although there are methodological problems <strong>in</strong> design<strong>in</strong>g studies to assess long-termcognitive function <strong>in</strong> children who have an ongo<strong>in</strong>g chronic disorder, these studies do raiseanxieties. It is felt that until hypoglycaemia can be reliably avoided, glycaemic control shouldbe less <strong>in</strong>tensively managed <strong>in</strong> younger children to avoid the risk of severe hypoglycaemia.This would put younger children at greater risk of develop<strong>in</strong>g the long-term microvascularcomplications of diabetes <strong>in</strong> an attempt to avoid the possible cognitive defects that may beassociated with recurrent hypoglycaemia.Hypoglycaemic HemiplegiaThis is a rare complication of acute hypoglycaemia <strong>in</strong> which the patient recovers fromthe hypoglycaemia with a transient hemiparesis last<strong>in</strong>g no more than 24 hours. Whenneuroimag<strong>in</strong>g is performed it is rare to f<strong>in</strong>d an abnormality. There is no evidence of anysevere sequelae to this neurological manifestation of severe neuroglycopenia (Pocecco andRonfani, 1998).Fear of <strong>Hypoglycaemia</strong>Both the children with diabetes and their parents worry about the prospect of a severe episodeof hypoglycaemia (Gold et al., 1997; Clarke et al., 1998; Gonder-Frederick et al., 1997;Marrero et al., 1997; Nordfelt and Ludvigson, 2005) (see Chapter 14). In one study severehypoglycaemia caused more fear than the prospect of an episode of diabetic ketoacidosis(Nordfelt and Ludvigson, 2005). Although there is little evidence that this modifies behaviourto attempt hypoglycaemia avoidance, such as relax<strong>in</strong>g glycaemic control or eat<strong>in</strong>g moresnacks, the adverse effects on quality of life should not be underestimated (Gold et al.,1997).Prediction of Nocturnal <strong>Hypoglycaemia</strong>Overnight glucose profiles have been shown to be extremely variable. As a result studies ofovernight hypoglycaemia have been unable to provide a ‘safe’ glucose value with which togo to bed. What is more useful is the fast<strong>in</strong>g blood glucose concentration on the follow<strong>in</strong>gmorn<strong>in</strong>g. One study showed that the median fast<strong>in</strong>g blood glucose at 07:00 hours wassignificantly lower follow<strong>in</strong>g hypoglycaemia than a night with no hypoglycaemia (3.7[1.4–10.6] versus 8.5 [3.8–19.2] mmol/l, p = 000001) (Matyka et al., 1999a) (Figure 9.2).

206 HYPOGLYCAEMIA IN CHILDREN WITH DIABETES2015Glucose (mmol / l)1050hypoglycaemiano hypoglycaemiaFigure 9.2 Glucose values at 07.00 hours follow<strong>in</strong>g a night of hypoglycaemia () and a night of nohypoglycaemia (△). Reproduced by permission of K. Matyka, PhD thesis ‘Nocturnal hypoglycaemia<strong>in</strong> prepubertal children with type 1 diabetes mellitus’, University of LondonThe occurrence of a ‘Somogyi phenomenon’, whereby overnight hypoglycaemia promotesglucose counterregulation and causes fast<strong>in</strong>g hyperglycaemia, has not been demonstrated <strong>in</strong>studies of nocturnal hypoglycaemia <strong>in</strong> children (Porter et al., 1996; Beregszaszi et al., 1997;Matyka et al., 1999a).MANAGEMENT OF HYPOGLYCAEMIAThe management of acute hypoglycaemia will depend on the severity of the episode. TheInternational Society for Paediatric and Adolescent Diabetologists has provided guidel<strong>in</strong>esfor the management of acute episodes of hypoglycaemia based on severity (ISPAD, 2000).Blood glucose measurement is the only way to confirm hypoglycaemia if the diagnosisis uncerta<strong>in</strong> and also confirm the return of the blood glucose to normal after treatment.Figure 9.3 shows the flow diagram for the management of hypoglycaemia.PreventionWhen faced <strong>in</strong> cl<strong>in</strong>ic with a child who is hav<strong>in</strong>g recurrent episodes of hypoglycaemia, adetailed history should be obta<strong>in</strong>ed regard<strong>in</strong>g the tim<strong>in</strong>g of hypoglycaemia, <strong>in</strong>sul<strong>in</strong> regimen,dietary <strong>in</strong>take and the relation to periods of physical activity. This will enable an assessmentto be made of possible risk factors and <strong>in</strong>form how these may be avoided. If no obviouscause is found then other pathology should be sought, such as co<strong>in</strong>cidental coeliac diseaseor the possibility of Addison’s disease, although these are relatively rare causes of recurrenthypoglycaemia (see Chapter 3).When contemplat<strong>in</strong>g preventive management the follow<strong>in</strong>g aspects should be considered.

MANAGEMENT OF HYPOGLYCAEMIA 207Check blood glucose – Glucose meter test and formal laboratory glucose if possible<strong>Hypoglycaemia</strong>MildSevereCONSCIOUSi.e. gag reflex <strong>in</strong>tactUNCONSCIOUSOral glucosee.g. 5–15 grams of glucoseor 100 mls sweet dr<strong>in</strong>kGlucogelOne ampoule orallyI.M. Glucagon5 yrs 1.0 mgSuccessNo successIf no response <strong>in</strong>15 m<strong>in</strong>s give 1–2mls/kg of 10%dextrose I.V.Repeat until thereis a cl<strong>in</strong>icalresponse.As symptoms improve or normoglycaemia is restored add oral complex carbohydratee.g. biscuit, bread and so on.If unable to tolerate oral carbohydrate may need a glucose <strong>in</strong>fusione.g. 5–10% glucose at ma<strong>in</strong>tenance rateFigure 9.3Management of hypoglycaemiaEducationGreat importance is placed on education <strong>in</strong> the management of type 1 diabetes and structurededucation programmes are now an essential part of any diabetes service provision. Despitethis, little validation has been made of the use of structured educational programmes <strong>in</strong>children. One study from Scand<strong>in</strong>avia has shown the benefits of a focused education package(Nordfelt et al., 2003). In this study families were given both written and video <strong>in</strong>formationon diabetes. One group were given general <strong>in</strong>formation on diabetes management and theother was given material to educate about the importance and means of hypoglycaemiaavoidance. Although no differences <strong>in</strong> glycated haemoglob<strong>in</strong> were found between the two

208 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESgroups, those who had the targeted <strong>in</strong>tervention had a significantly reduced rate of severehypoglycaemia (Nordfelt et al., 2003).Insul<strong>in</strong>As noted earlier, the DCCT suggested that attempts to <strong>in</strong>tensify <strong>in</strong>sul<strong>in</strong> regimens may <strong>in</strong>creasethe risk of severe hypoglycaemia. Recent <strong>in</strong>troductions of analogue <strong>in</strong>sul<strong>in</strong>s, however,<strong>in</strong>dicate that improved glycaemic control does not always lead to hypoglycaemia, althoughfew studies have been designed specifically to exam<strong>in</strong>e the benefits of different <strong>in</strong>sul<strong>in</strong>regimens on the risk of hypoglycaemia. One study that compared <strong>in</strong>sul<strong>in</strong> lispro with soluble(short-act<strong>in</strong>g) <strong>in</strong>sul<strong>in</strong> as part of a basal bolus regimen, showed small but statistically significantreductions <strong>in</strong> the prevalence of hypoglycaemia over a 30 day period when <strong>in</strong>sul<strong>in</strong>lispro was be<strong>in</strong>g used (Holcombe et al., 2002). Other studies have found benefits of <strong>in</strong>sul<strong>in</strong>analogue regimens on nocturnal hypoglycaemia. One randomised cross-over study <strong>in</strong> adolescentscompared <strong>in</strong>sul<strong>in</strong> lispro and glarg<strong>in</strong>e, as part of a daily multiple <strong>in</strong>jection regimen,to human soluble and isophane <strong>in</strong>sul<strong>in</strong>s. Nocturnal hypoglycaemia was 43% lower withthe analogue regimen, although no difference was observed <strong>in</strong> self-reported symptomatichypoglycaemia (Murphy et al., 2003). Another study <strong>in</strong> prepubertal children exam<strong>in</strong>ed thebenefits of a thrice daily <strong>in</strong>sul<strong>in</strong> regimen, where the even<strong>in</strong>g dose of mixed <strong>in</strong>sul<strong>in</strong> wasreplaced by a rapid-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogue with the even<strong>in</strong>g meal and isophane <strong>in</strong>sul<strong>in</strong> beforebed (Ford-Adams et al., 2003). Although there was no difference <strong>in</strong> glycated haemoglob<strong>in</strong>between the two treatment arms, the prevalence of hypoglycaemia was lower <strong>in</strong> the earlypart of the night (22.00–04.00 hours) when the analogue was used.Although not every patient is suited to us<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> pump therapy, cl<strong>in</strong>ic-based studiesof CSII therapy have shown that more stable blood glucose control can be achieved withoutan <strong>in</strong>creased risk of hypoglycaemia. In an American study describ<strong>in</strong>g the experience ofus<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> pumps <strong>in</strong> a paediatric cl<strong>in</strong>ic, it was found that 50 adolescents on multipledaily <strong>in</strong>jections experienced 134 episodes of severe hypoglycaemia per 100 patient-yearscompared to 76 episodes per 100 patient-years <strong>in</strong> the 25 adolescents who opted for pumptherapy (Boland et al., 1999).DietFew studies have exam<strong>in</strong>ed the impact of dietary <strong>in</strong>terventions on hypoglycaemia risk exceptfor nocturnal hypoglycaemia. The major dietary modification has been that of the <strong>in</strong>troductionof a larger proportion of starch, as a form of long-act<strong>in</strong>g carbohydrate, as part of the even<strong>in</strong>gsnack (Ververs et al., 1993; Kaufman et al., 1995; Detlofson et al., 1999; Matyka et al.,1999a). In these studies the benefits of starch have been <strong>in</strong>consistent. One study found alower frequency of nocturnal hypoglycaemia, although capillary sampl<strong>in</strong>g was performedonly <strong>in</strong>termittently dur<strong>in</strong>g the night and some episodes of hypoglycaemia may have beenundetected (Kaufman et al., 1995). Others found no beneficial effect of cornstarch on theprevention of nocturnal hypoglycaemia although blood glucose concentrations fell moreslowly, but <strong>in</strong> one study this occurred at the expense of promot<strong>in</strong>g hyperglycaemia (Ververset al., 1993; Matyka et al., 1999a). Although not designed to exam<strong>in</strong>e the impact of diet onhypoglycaemia, a study of a low glycaemic <strong>in</strong>dex diet has been shown to improve glycaemic

CONCLUSIONS 209control without an <strong>in</strong>crease <strong>in</strong> rate of hypoglycaemia, and appeared to have enhanced qualityof life (Gilbertson et al., 2001)ExerciseWhen adequate plasma <strong>in</strong>sul<strong>in</strong> concentrations are available, exercise can lead to acutehypoglycaemia. However if exercise is performed at a time of relative <strong>in</strong>sul<strong>in</strong> deficiency,hyperglycaemia with ketosis can occur. In addition, delayed hypoglycaemia may occur asmuscle glycogen stores recover ma<strong>in</strong>ly overnight (Admon et al., 2005; Tsalikian et al., 2005).Although few data are currently available regard<strong>in</strong>g the most appropriate management ofplanned periods of physical activity, a number of guidel<strong>in</strong>es have been proposed. The InternationalSociety for Paediatric and Adolescent Diabetologists has published guidel<strong>in</strong>es onthe Internet (www.ispad.org). These recommend that careful monitor<strong>in</strong>g of blood glucose isessential to match food and <strong>in</strong>sul<strong>in</strong> to the <strong>in</strong>tensity of exercise and that a reduction of <strong>in</strong>sul<strong>in</strong>should be considered. Additional slowly absorbed carbohydrate will be necessary, especiallyat bedtime, if exercise has been performed <strong>in</strong> the afternoon or early even<strong>in</strong>g. From the dataavailable so far (Admon et al., 2005; Tsalikian et al., 2005), these guidel<strong>in</strong>es do seem areasonable approach to the avoidance of both exercise related hypo- and hyperglycaemia.It is important, however, to work with the child and family to provide <strong>in</strong>dividually-tailoredrecommendations that are tried and tested for the child. The management of unpredictableepisodes of physical activity are likely to rema<strong>in</strong> a problem until a cure for diabetesis found.CONCLUSIONS• <strong>Hypoglycaemia</strong> is a common problem for children with type 1 diabetes, especially youngchildren, and for their families.• Behavioural symptoms are more common <strong>in</strong> childhood than more typical autonomicsymptoms. The majority of episodes of nocturnal hypoglycaemia are totally asymptomatic.• Episodes of hypoglycaemia rema<strong>in</strong> a significant barrier when striv<strong>in</strong>g for a degree ofglycaemic control that will delay or prevent the development of the microvascular complicationsof diabetes.• Glucagon responses dur<strong>in</strong>g hypoglycaemia are lost early <strong>in</strong> the course of type 1 diabetes.Ep<strong>in</strong>ephr<strong>in</strong>e responses dur<strong>in</strong>g overnight hypoglycaemia are blunted <strong>in</strong> both healthy childrenand those with type 1 diabetes and may contribute to the lack of symptoms of manyepisodes of hypoglycaemia overnight.• Concerns rema<strong>in</strong> about the possible long-term implications of hypoglycaemia <strong>in</strong> terms ofcognitive dysfunction although hyperglycaemia may also be important.• More recent data suggest that novel <strong>in</strong>sul<strong>in</strong> analogues and regimens may enable improvements<strong>in</strong> glycaemic control to be achieved without a concomitant <strong>in</strong>crease <strong>in</strong> the risk ofhypoglycaemia.

210 HYPOGLYCAEMIA IN CHILDREN WITH DIABETES• It is important to consider both the child and their family when assess<strong>in</strong>g the causes ofrecurrent episodes of hypoglycaemia and when suggest<strong>in</strong>g possible avoidance strategies.• All children with diabetes should have a personalised action plan for hypoglycaemiaavoidance.REFERENCESAdmon G, We<strong>in</strong>ste<strong>in</strong> Y, Falk B, We<strong>in</strong>trob N, Benzaquen H, Ofan R et al. (2005). Exercise with andwithout an <strong>in</strong>sul<strong>in</strong> pump among children and adolescents with type 1 diabetes mellitus. Pediatrics116: 348–55.Ahlborg G, Felig P, Hagenfeldt L, Hemdler R, Wahren J (1974). Substrate turnover dur<strong>in</strong>g prolongedexercise <strong>in</strong> man. Journal of Cl<strong>in</strong>ical Investigation 53: 1080–90.Allen C, LeCaire T, Palta M, Daniles K, Meredith M, D’Alessio DJ, for the Wiscons<strong>in</strong> <strong>Diabetes</strong>Registry Project (2001) Risk factors for frequent and severe hypoglycemia <strong>in</strong> type 1 diabetes.<strong>Diabetes</strong> Care 24: 1878–81.Amiel SA, Simonson DC, Sherw<strong>in</strong> RS, Lauritano AA, Tamborlane WV (1987). Exaggeratedep<strong>in</strong>ephr<strong>in</strong>e responses to hypoglycemia <strong>in</strong> normal and <strong>in</strong>sul<strong>in</strong>-dependent diabetic children. Journalof Pediatrics 110: 832–7.Bendtson I, Gade J, Thomsen CE, Rosenfalck A, Wildschiodtz G (1992). Sleep disturbances <strong>in</strong> IDDMpatients with nocturnal hypoglycemia. Sleep 15: 74–81Bendtson I, Rosenfalck AM, B<strong>in</strong>der C (1993). Nocturnal versus diurnal counterregulation to hypoglycemia<strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong>-dependent) diabetic patients. Acta Endocr<strong>in</strong>ologica 128: 109–15.Beregszaszi M, Tubiana-Rufi N, Benali K, Noel M, Bloch J, Czernichow P (1997). Nocturnal hypoglycemia<strong>in</strong> children and adolescents with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: Prevalence and riskfactors. Journal of Pediatrics 131: 27–33.Bier DM, Leake RD, Haymond MW, Arnold KJ, Gruenke LD, Sperl<strong>in</strong>g MA, Kipnis DM (1977).Measurement of ‘true’ glucose production rates <strong>in</strong> <strong>in</strong>fancy and childhood with 6,6-dideuteroglucose.<strong>Diabetes</strong> 26: 1016–23.Bjorgaas M, Vik T, Sand T, Birkeland K, Sager G, Vea H, Jorde R (1997a). Counterregulatory andsymptom responses to hypoglycaemia <strong>in</strong> diabetic children. Diabetic Medic<strong>in</strong>e 14: 433–41.Bjorgaas M, Gimse R, Vik T, Sand T (1997b). Cognitive function <strong>in</strong> type 1 diabetic children with andwithout episodes of severe hypoglycaemia. Acta Paediatrica 86: 148–53.Bjorgaas M, Sand T, Vik T, Jorde R (1998). Quantitative EEG dur<strong>in</strong>g controlled hypoglycaemia <strong>in</strong>diabetic and non-diabetic children. Diabetic Medic<strong>in</strong>e 15: 30–7.Bognetti F, Brunelli A, Meschi F, Viscardi M, Bonfanti R, Chiumello G (1997). Frequency andcorrelates of severe hypoglycaemia <strong>in</strong> children and adolescents with diabetes mellitus. EuropeanJournal of Paediatrics 156: 589–91.Boland EA, Grey M, Oesterle A, Fredrickson L, Tamborlane WV (1999). Cont<strong>in</strong>uous subcutaneous<strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion. A new way to lower risk of severe hypoglycemia, improve metabolic control, andenhance cop<strong>in</strong>g <strong>in</strong> adolescents with type 1 diabetes. <strong>Diabetes</strong> Care 22: 1779–84.Bolli G, Gerich J (1984). The dawn phenomenon – a common occurrence <strong>in</strong> both non-<strong>in</strong>sul<strong>in</strong> dependentand <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus. New England Journal of Medic<strong>in</strong>e 310: 746–50.Brambilla P, Bougneres PF, Santiago JV, Chauss<strong>in</strong> JL, Pouplard A, Castano L (1987). Glucose counterregulation<strong>in</strong> pre-school-age children with recurrent hypoglycemia dur<strong>in</strong>g conventional treatment.<strong>Diabetes</strong> 36: 300–4.Bulsara MK, Holman CDJ, Davis EA, Jones TW (2004). The impact of a decade of chang<strong>in</strong>g treatmenton rates of severe hypoglycemia <strong>in</strong> a population-based cohort of children with type 1 diabetes.<strong>Diabetes</strong> Care 27: 2293–8.

REFERENCES 211Bulsara MK, Holman CDJ, van Bockxmeer FM, Davis EA, Gallego PH, Beilby JP et al. (2007). Therelationship between ACE genotype and risk of severe hypoglycaemia <strong>in</strong> a large population-basedcohort of children and adolescents with type 1 diabetes. Diabetologia 50: 965–71.Cartee GD, Holloszy JO (1990). Exercise <strong>in</strong>creases susceptibility of muscle glucose transport toactivation by various stimuli. American Journal of Endocr<strong>in</strong>ology 258: E390–3.Chaussa<strong>in</strong> J-L (1973). Glycemic response to 24-hour fast <strong>in</strong> normal children and children with ketotichypoglycemia. Journal of Pediatrics 82: 438–43.Chaussa<strong>in</strong> J-L, Georges P, Olive G, Job J-C (1974). Glycemic response to 24-hour fast <strong>in</strong> normalchildren and children with ketotic hypoglycemia: II Hormonal and metabolic changes. Journal ofPediatrics 85: 776–81.Chaussa<strong>in</strong> J-L, Georges P, Calzada L, Job J-C (1977). Glycemic response to 24-hour fast <strong>in</strong> normalchildren: III Influence of age. Journal of Pediatrics 91: 711–4.Clarke WL, Gonder-Frederick A, Snyder AL, Cox DJ (1998). Maternal fear of hypoglycemia <strong>in</strong>their children with <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus. Journal of Pediatric Endocr<strong>in</strong>ology andMetabolism 11: 189–94.Craig ME, Handelsman P, Donaghue KC, Chan A, Blades B, La<strong>in</strong>a R et al., NSW/ACTHbA(1c) Study Group (2002). Predictors of glycaemic control and hypoglycaemia <strong>in</strong> childrenand adolescents with type 1 diabetes from NSW and ACT. Medical Journal of Australia 177:235–8.Daneman D, Frank M, Perlman K, Tamm J, Ehrlich R (1989). Severe hypoglycemia <strong>in</strong> children with<strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: frequency and predispos<strong>in</strong>g factors. Journal of Pediatrics 5:681–5.Danne T, Weber B, Hartmann R, Enders I, Burger W, Hovener G (1994). Long-term glycemic controlhas a nonl<strong>in</strong>ear association to the frequency of background ret<strong>in</strong>opathy <strong>in</strong> adolescents with diabetes.Follow-up of the Berl<strong>in</strong> Ret<strong>in</strong>opathy Study. <strong>Diabetes</strong> Care 17: 1390–6.Davis EA, Keat<strong>in</strong>g B, Byrne GC, Russell M, Jones TW (1997). Hypoglycemia: <strong>in</strong>cidence and cl<strong>in</strong>icalpredictors <strong>in</strong> a large population based sample of children and adolescents with IDDM. <strong>Diabetes</strong>Care 20: 20–5.Davis EA, Keat<strong>in</strong>g B, Byrne GC, Russell, Jones TW (1998). Impact of improved glycaemic controlon rates of hypoglycaemia <strong>in</strong> IDDM. Archives of Disease <strong>in</strong> Childhood 78: 111–5.Deary IJ, Hepburn DA, MacLeod KM, Frier BM (1993). Partition<strong>in</strong>g the symptoms of hypoglycaemiaus<strong>in</strong>g multi-sample confirmatory factor analysis. Diabetologia 36: 771–7.De Feo P, Torlone E, Fanelli C, Santeusanio F, Brunetti P, Bolli GB (1990). Nocturnal spikes of growthhormone secretion cause the dawn phenomenon <strong>in</strong> type 1 (<strong>in</strong>sul<strong>in</strong> dependent) diabetes mellitusby decreas<strong>in</strong>g hepatic (and extrahepatic) sensitivity to <strong>in</strong>sul<strong>in</strong> <strong>in</strong> the absence of <strong>in</strong>sul<strong>in</strong> wan<strong>in</strong>g.Diabetologia 33: 52–9.Detlofson I, Kroon M, Aman J (1999). Oral bedtime cornstarch supplementation reduces the risk fornocturnal hypoglycaemia <strong>in</strong> young children with type 1 diabetes. Acta Paediatrica 88: 595–7.Dunger DB (1992). <strong>Diabetes</strong> <strong>in</strong> puberty. Archives of Disease <strong>in</strong> Childhood 67: 569–70.Edge JA, Matthews DR, Dunger DB (1990a). Failure of current <strong>in</strong>sul<strong>in</strong> regimes to meet the overnightrequirements of adolescents with <strong>in</strong>sul<strong>in</strong>-dependent diabetes. <strong>Diabetes</strong> Research 15: 109–12.Edge JA, Matthews DR, Dunger DB (1990b). The dawn phenomenon is related to overnight growthhormone release <strong>in</strong> adolescent diabetics. Cl<strong>in</strong>ical Endocr<strong>in</strong>ology 33: 729–37.Ferguson SC, Blane A, Wardlaw J, Frier BM, Perros P, McCrimmon RJ, Deary IJ (2005). Influenceof an early-onset age of type 1 diabetes on cerebral structure and cognitive function. <strong>Diabetes</strong> Care28: 1431–7.Ford-Adams ME, Murphy NP, Moore EJ, Edge JA, Ong KL, Watts AP et al. (2003). Insul<strong>in</strong> lispro:a potential role <strong>in</strong> prevent<strong>in</strong>g nocturnal hypoglycaemia <strong>in</strong> young children with diabetes mellitus.Diabetic Medic<strong>in</strong>e 20: 656–60.Gale EAM, Tattersall RB (1979). Unrecognised nocturnal hypoglycaemia <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-treated diabetics.Lancet i: 1049–52.

212 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESGilbertson HR, Brand-Miller JC, Thorburn AW, Evans S, Chondros P, Werther GA (2001). The effectof flexible low glycemic <strong>in</strong>dex dietary advice versus measured carbohydrate exchange diets onglycemic control <strong>in</strong> children with type 1 diabetes. <strong>Diabetes</strong> Care 24: 1137–43.Glazer RI, Weber G (1971). Incorporation of [6- 3 H] glucose <strong>in</strong>to lipid, prote<strong>in</strong>, RNA and DNA slicesof differentiat<strong>in</strong>g rat cerebral cortex. Journal of Neurochemistry 18: 1569–76.Gold AE, Deary IJ, Frier BM (1997). <strong>Hypoglycaemia</strong> and non-cognitive aspects of psychologicalfunction <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependent (type 1) diabetes mellitus (IDDM). Diabetic Medic<strong>in</strong>e 14: 111–8.GoldenMP,IngersollGM,BrackCJ,RussellBA,WrightJC,HubertyTJ(1989).Longitud<strong>in</strong>alrelationshipof asymptomatic hypoglycemia to cognitive function <strong>in</strong> IDDM. <strong>Diabetes</strong> Care 12: 89–93.Gonder-Frederick L, Cox D, Kovatchev B, Julian D, Clarke W (1997). The psychosocial impact ofsevere hypoglycemic episodes on spouses of patients with IDDM. <strong>Diabetes</strong> Care. 20: 1543–6.Haymond MW, Karl IE, Clarke WL, Pagliara AS, Santiago JV (1982). Differences <strong>in</strong> circulat<strong>in</strong>ggluconeogenic substrates dur<strong>in</strong>g short-term fast<strong>in</strong>g <strong>in</strong> men, women and children. Metabolism 31:33–42.Hershey T, Bhargava N, Sadler M, White NH, Craft S (1999). Conventional versus <strong>in</strong>tensive diabetestherapy <strong>in</strong> children with type 1 diabetes – effects on memory and motor speed. <strong>Diabetes</strong> Care 22:1318–24.Hoffman RP, S<strong>in</strong>ger-Granick C, Drash AL, Becker DJ (1991). Plasma catecholam<strong>in</strong>e responses tohypoglycemia <strong>in</strong> children and adolescents with IDDM. <strong>Diabetes</strong> Care 14: 81–8.Holcombe JH, Zalani S, Arora VK, Mast CJ (2002). Lispro <strong>in</strong> Adolescents Study Group. Comparisonof <strong>in</strong>sul<strong>in</strong> lispro with regular human <strong>in</strong>sul<strong>in</strong> for the treatment of type 1 diabetes <strong>in</strong> adolescents.Cl<strong>in</strong>ical Therapeutics 24: 629–38.Holl RW, Swift PG, Mortensen HB, Lynggaard H, Hougaard P, Aanstoot HJ et al., on behalf of theHvidore Study Group (2003). Insul<strong>in</strong> <strong>in</strong>jection regimens and metabolic control <strong>in</strong> an <strong>in</strong>ternationalsurvey of adolescents with type 1 diabetes over 3 years: results from the Hvidore Study Group.European Journal of Paediatrics 162: 22–9.ISPAD Consensus Guidel<strong>in</strong>es for the Management of type 1 <strong>Diabetes</strong> Mellitus <strong>in</strong> Children andAdolescents. Swift P, ed. Medical Forum International, The Netherlands. 2000.Ivy JL, Holloszy JO (1981). Persistent <strong>in</strong>crease <strong>in</strong> glucose uptake by rat skeletal muscle follow<strong>in</strong>gexercise. American Journal of Physiology 241: C200–3.Jones TW, Boulware SD, Kraemer DT, Caprio S, Sherw<strong>in</strong> RS, Tamborlane WV (1991). Independenteffects of youth and poor diabetes control on responses to hypoglycemia <strong>in</strong> children. <strong>Diabetes</strong> 40:358–63.Jones TW, Borg WP, Boulware SD, McCarthy G, Sherw<strong>in</strong> RS, Tamborlane WV (1995). Enhancedadrenomedullary response and <strong>in</strong>creased susceptibility to neuroglycopenia: Mechanisms underly<strong>in</strong>gthe adverse effects of sugar <strong>in</strong>gestion <strong>in</strong> healthy children. Journal of Pediatrics 126: 171–7.Jones TW, Porter P, Sherw<strong>in</strong> RS, Davis EA, O’Leary P, Frazer F et al. (1998). Decreased ep<strong>in</strong>ephr<strong>in</strong>eresponses to hypoglycemia dur<strong>in</strong>g sleep. New England Journal of Medic<strong>in</strong>e 338: 1657–62.Kaufman FR, Halvorson M, Kaufman ND (1995). A randomized, bl<strong>in</strong>ded trial of uncooked cornstarchto dim<strong>in</strong>ish nocturnal hypoglycemia at diabetes camp. <strong>Diabetes</strong> Research and Cl<strong>in</strong>ical Practice. 30:205–9.Kerr DS, Hansen IL, Levy MM (1983). Metabolic and hormonal responses of children and adolescentsto fast<strong>in</strong>g and 2-deoxyglucose. Metabolism 22: 951–9.Koh THHG, Eyre JA, Aynsley-Green A (1988). Neonatal hypoglycaemia – the controversy regard<strong>in</strong>gdef<strong>in</strong>ition. Archives of Disease <strong>in</strong> Childhood 63: 1386–8.Lamers KJB, Doesburg WH, Gabreels FJM, Lemmens WAJG, Romson AC, Wevers RA, Renier WO(1985a). The concentration of blood components related to fuel metabolism dur<strong>in</strong>g prolonged fast<strong>in</strong>g<strong>in</strong> children. Cl<strong>in</strong>ica Chimica Acta 152: 155–63.Lamers KJB, Doesburg WH, Gabreels FJM, Romson AC, Renier WO, Wevers RA, Lemmens WAJG(1985b). Reference values of blood components related to fuel metabolism <strong>in</strong> children after anovernight fast. Cl<strong>in</strong>ica Chimica Acta 145: 17–26.

REFERENCES 213Lawrence RD (1926). The effects of exercise on <strong>in</strong>sul<strong>in</strong> action <strong>in</strong> diabetes. British Medical Journal 1:648–52.Lev<strong>in</strong>e B, Anderson BJ, Butler DA, Antisdel JE, Brackett J, Laffel LMB (2001). Predictors of glycemiccontrol and short-term adverse outcomes <strong>in</strong> youths with type 1 diabetes. Journal of Pediatrics 139:197–203.Lopez MJ, Oyarzabal M, Barrio R, Hermoso F, Lopez JP, Rodriguez M et al. (1997). Nocturnalhypoglycaemia <strong>in</strong> IDDM patients younger than 18 years. Diabetic Medic<strong>in</strong>e 14: 772–7.Lucas A, Morley R, Cole TJ (1988). Adverse neurodevelopmental outcome of moderate neonatalhypoglycaemia. British Medical Journal 297: 1304–8.McCrimmon RJ, Gold AE, Deary IJ, Kelnar CJ, Frier BM (1995). Symptoms of hypoglycemia <strong>in</strong>children with IDDM. <strong>Diabetes</strong> Care 18: 858–61.Magrath G, Hartland BV, the Nutrition Subcommittee of the British Diabetic Association’s ProfessionalAdvisory Committee (1993). Dietary recommendations for children with diabetes: an implementationpaper. Journal of Human Nutrition and Dietetics 6: 491–507.Maquet P, Dive D, Salmon E, Sadzot B, Franco G, Poirrier R et al. (1990). Cerebral glucose utilizationdur<strong>in</strong>g the sleep-wake cycle <strong>in</strong> man determ<strong>in</strong>ed by positron emission tomography and 18F-2-fluoro-2-deoxy-D-glucose method. Bra<strong>in</strong> Research 513: 136–43.Marrero DG, Guare JC, Vandagriff JL, F<strong>in</strong>eberg NS (1997). Fear of hypoglycemia <strong>in</strong> the parents ofchildren and adolescents with diabetes: maladaptive or healthy response? <strong>Diabetes</strong> Educator 23:281–6.Matyka KA, Wiggs L, Pramm<strong>in</strong>g S, Stores G, Dunger DB (1999a). Cognitive function and moodfollow<strong>in</strong>g profound nocturnal hypoglycaemia dur<strong>in</strong>g conventional therapy of diabetes. Archives ofDisease <strong>in</strong> Childhood 81: 138–42.Matyka KA, Crowne EC, Havel PJ, Macdonald IA, Matthews D, Dunger DB (1999b). Counterregulationdur<strong>in</strong>g spontaneous nocturnal hypoglycemia <strong>in</strong> prepubertal children with <strong>in</strong>sul<strong>in</strong> dependentdiabetes mellitus. <strong>Diabetes</strong> Care 22: 1144–50.Matyka KA, Crawford C, Wiggs L, Dunger DB, Stores G (2000). Alterations <strong>in</strong> sleep physiology <strong>in</strong>young children with <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus: relationship to nocturnal hypoglycemia.Journal of Pediatrics 137: 233–8.Mohn A, Matyka KA, Harris DA, Ross KM, Edge JA, Dunger DB (1999). Lispro or regular <strong>in</strong>sul<strong>in</strong> formultiple <strong>in</strong>jection therapy <strong>in</strong> adolescence. Differences <strong>in</strong> free <strong>in</strong>sul<strong>in</strong> and glucose levels overnight.<strong>Diabetes</strong> Care 22: 27–32.Mortensen HB, Hougaard P, for the Hvidore Study Group on Childhood <strong>Diabetes</strong> (1997). Comparisonof metabolic control <strong>in</strong> a cross-sectional study of 2,873 children and adolescents with IDDM from18 countries. <strong>Diabetes</strong> Care 20: 714–20.Murphy NP, Keane SM, Ong KK, Ford-Adams M, Edge JA, Acer<strong>in</strong>i C, Dunger DB (2003).Randomized cross-over trial of <strong>in</strong>sul<strong>in</strong> glarg<strong>in</strong>e plus lispro or NPH <strong>in</strong>sul<strong>in</strong> plus regular human<strong>in</strong>sul<strong>in</strong> <strong>in</strong> adolescents with type 1 diabetes on <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimens. <strong>Diabetes</strong> Care 26:799–804.Nordfelt S, Ludvigsson J (1997). Severe hypoglycemia <strong>in</strong> children with IDDM: a prospective populationstudy 1992–1994. <strong>Diabetes</strong> Care 20: 497–503.Nordfelt S, Samuelsson U (2003). Serum ACE predicts severe hypoglycemia <strong>in</strong> children and adolescentswith type 1 diabetes. <strong>Diabetes</strong> Care 26: 274–8.Nordfelt S, Johansson C, Carlsson E, Hammersjo J-A (2003). Prevention of severe hypoglycaemia <strong>in</strong>type 1 diabetes: a randomised controlled population study. Archives of Disease <strong>in</strong> Childhood 88:240–5.Nordfelt S, Ludvigsson J (2005). Fear and other disturbances of severe hypoglycaemia <strong>in</strong> children andadolescents with type 1 diabetes. Journal of Pediatric Endocr<strong>in</strong>ology 18: 83–91.Northam EA, Anderson PJ, Jacobs R, Hughes M, Warne GL, Werther GA (2001). Neuropsychologicalprofiles of children with type 1 diabetes 6 years after disease onset. <strong>Diabetes</strong> Care24: 1541–6.

214 HYPOGLYCAEMIA IN CHILDREN WITH DIABETESParmeggiani PL, Morrison AR (1990). Alterations <strong>in</strong> autonomic functions dur<strong>in</strong>g sleep. In: CentralRegulation of Autonomic Functions. Loewy AD and Spyer KM, eds. Oxford University Press,Oxford: 367–86.Pedersen-Bjergaard U, Agerholm-Larsen, Pramm<strong>in</strong>g S, Hougaard P, Thorste<strong>in</strong>sson B (2001). Activityof angiotens<strong>in</strong> convert<strong>in</strong>g enzyme and risk of severe hypoglycaemia <strong>in</strong> type 1 diabetes mellitus.Lancet 357: 1248–53.Pedersen-Bjergaard U, Agerholm-Larsen, Pramm<strong>in</strong>g S, Hougaard P, Thorste<strong>in</strong>sson B (2003). Predictionof severe hypoglycaemia by angiotens<strong>in</strong> convert<strong>in</strong>g enzyme activity and genotype <strong>in</strong> type 1 diabetes.Diabetologia 46: 89–96.Pillar G, Schuscheim G, Weiss R, Malhotra A, McCowen KC, Shlitner A et al. (2003). Interactionsbetween hypoglycemia and sleep architecture <strong>in</strong> children with type 1 diabetes mellitus. Journal ofPediatrics 142: 163–8.Pocecco M, Ronfani L (1998). Transient focal neurologic deficits associated with hypoglycaemia<strong>in</strong> children with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Italian Collaborative Paediatric DiabetologicGroup. Acta Paediatrica 87: 542–4.Porter P, Byrne G, Stick S, Jones TW (1996). Nocturnal hypoglycaemia and sleep disturbances <strong>in</strong>young teenagers with <strong>in</strong>sul<strong>in</strong> dependent diabetes mellitus. Archives of Disease <strong>in</strong> Childhood 75:120–3.Porter PA, Keat<strong>in</strong>g B, Byrne G, Jones TW (1997). Incidence and predictive criteria of nocturnalhypoglycemia <strong>in</strong> young children with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Journal of Pediatrics130: 366–72.Price TB, Rothman DL, Taylor R, Avison MJ, Shulman GI, Shulman RG (1994). Human muscleglycogen resynthesis after exercise: <strong>in</strong>sul<strong>in</strong>-dependent and <strong>in</strong>dependent phases. Journal of AppliedPhysiology 76: 104–11.Rewers A, Chase HP, Mackenzie T, Walravens P, Roback M, Rewers M et al. (2002). Predictors ofacute complications <strong>in</strong> children with type 1 diabetes. Journal of the American Medical Association287: 2511–8.Richter EA (1996). Glucose utilization. In: Exercise: Regulation and Integration of Multiple Systems.Rowell LB and JT Shepherd JT, eds. Oxford University Press, New York: 912–51.Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F (1990). An <strong>in</strong>sertion/deletionpolymorphism <strong>in</strong> the angiotens<strong>in</strong>-1 convert<strong>in</strong>g enzyme gene account<strong>in</strong>g for half the variance ofserum enzyme levels. Journal of Cl<strong>in</strong>ical Investigation 86: 1343–6.Rosilio MR, Cotton JB, Wieliczko MC, Genrault B, Carel JC, Couvaras O et al., on behalf of theFrench Pediatric <strong>Diabetes</strong> Group (1998). Factors associated with glycemic control. <strong>Diabetes</strong> Care21: 1146–53.Ross LA, Warren RE, Kelnar CJH, Frier BM (2005). Pubertal stage and hypoglycaemia counterregulation<strong>in</strong> type 1 diabetes. Archives of Disease <strong>in</strong> Childhood 90: 190–4.Rovet JF, Ehrlich RM (1999). The effect of hypoglycemic seizures on cognitive function <strong>in</strong> childrenwith diabetes: a seven year prospective study. Journal of Pediatrics 134: 503–6.Ryan C, Vega A, Drash A (1985). Cognitive deficits <strong>in</strong> adolescents who developed diabetes early <strong>in</strong>life. Pediatrics 75: 921–7.Ryan CM, Atchison J, Puczynski S, Puczynski M, Arslanian S, Becker D (1990). Mild hypoglycemiaassociated with deterioration of mental efficiency <strong>in</strong> children with <strong>in</strong>sul<strong>in</strong>-dependent diabetesmellitus. Journal of Pediatrics 117: 32–8.Saudubray JM, Marsac C, Limal JM, Dumurgier E, Charpentier C, Ogier H, Coude FX (1981).Variation <strong>in</strong> plasma ketone bodies dur<strong>in</strong>g a 24-hour fast <strong>in</strong> normal and hypoglycemic children:Relationship to age. Journal of Pediatrics 98: 904–8.Schultz CJ, Konopelska-Bahu T, Dalton RN, Carroll TA, Stratton I, Gale EA et al. (1999). Microalbum<strong>in</strong>uriaprevalence varies with age, sex, and puberty <strong>in</strong> children with type 1 diabetes followedfrom diagnosis <strong>in</strong> a longitud<strong>in</strong>al study. Oxford Regional Prospective Study Group. <strong>Diabetes</strong> Care22: 495–502.

REFERENCES 215S<strong>in</strong>ger-Granick C, Hoffman RP, Kerensky K, Drash AL, Becker DJ (1988). Glucagon responses tohypoglycemia <strong>in</strong> children and adolescents with IDDM. <strong>Diabetes</strong> Care 11: 643–9.Solders G, Thalme B, Aguirre-Aqu<strong>in</strong>o M, Brandt L, Berg U, Persson A (1997). Nerve conductionand autonomic nerve function <strong>in</strong> diabetic children. A 10 year follow-up study. Acta Paediatrica 86:361–6.Stirl<strong>in</strong>g HF, Darl<strong>in</strong>g JAB, Kelnar CJH (1991). Nocturnal glucose homeostasis <strong>in</strong> normal children.Hormone Research 35: 54 (abstract).The <strong>Diabetes</strong> Control and Complications Trial Research Group (1993). The effect of <strong>in</strong>tensive treatmentof diabetes on the development and progression of long-term complications <strong>in</strong> <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus. New England Journal of Medic<strong>in</strong>e 329: 977–86.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1994). Effect of <strong>in</strong>tensive diabetestreatment on the development and progression of long-term complications <strong>in</strong> adolescents with<strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus: <strong>Diabetes</strong> Control and Complications Trial. Journal of Pediatrics125: 177–88.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1997). Hypoglycemia <strong>in</strong> the <strong>Diabetes</strong>Control and Complications Trial. <strong>Diabetes</strong> 46: 271–86.Thomsett M, Shield G, Batch J, Cotterill A (1999). How well are we do<strong>in</strong>g? Metabolic control <strong>in</strong>patients with diabetes. Journal of Paediatrics and Child Health 35: 479–82.Trumper BG, Reschke K, Moll<strong>in</strong>g J (1995). Circadian variation of <strong>in</strong>sul<strong>in</strong> requirement <strong>in</strong> IDDM: therelationship between circadian change <strong>in</strong> <strong>in</strong>sul<strong>in</strong> demand and diurnal patterns of growth hormone,cortisol and glucagon dur<strong>in</strong>g euglycaemia. Hormone and Metabolic Research 27: 141–7.Tsalikian E, Mauras N, Beck RW, Tamborlane WV, Janz KF, Chase HP et al., <strong>Diabetes</strong> Research InChildren Network Direcnet Study Group (2005). Impact of exercise on overnight glycemic control<strong>in</strong> children with type 1 diabetes. Journal of Pediatrics 147: 528–34.Tupola S, Rajantie J (1998a). Documented symptomatic hypoglycaemia <strong>in</strong> children and adolescentsus<strong>in</strong>g multiple daily <strong>in</strong>jection therapy. Diabetic Medic<strong>in</strong>e 15: 492–6.Tupola S, Rajantie J, Maenpaa J (1998b). Severe hypoglycaemia <strong>in</strong> children and adolescents dur<strong>in</strong>gmultiple-dose <strong>in</strong>sul<strong>in</strong> therapy. Diabetic Medic<strong>in</strong>e 15: 695–9.Vanelli M, Cerutti F, Chiarelli F, Lor<strong>in</strong>i R, Meschi F, MCDC – Italy Group (2005). Nationwide crosssectionalsurvey of 3560 children and adolescents with diabetes <strong>in</strong> Italy. Journal of Endocr<strong>in</strong>ologicalInvestigation 28: 692–9.Ververs MT, Rouwe C, Smit GP (1993). Complex carbohydrates <strong>in</strong> the prevention of nocturnalhypoglycaemia <strong>in</strong> diabetic children. European Journal of Cl<strong>in</strong>ical Nutrition 47: 268–73.Wagner VM, Grabert M, Holl RW (2005). Severe hypoglycaemia, metabolic control and diabetesmanagement <strong>in</strong> children with type 1 diabetes <strong>in</strong> the decade after the <strong>Diabetes</strong> Control and ComplicationsTrial – a large-scale multicentre study. European Journal of Paediatrics 164: 73–9.West H, Richey A, Eyre MB (1934). Study of <strong>in</strong>telligence levels of juvenile diabetics. PsychologicalBullet<strong>in</strong> 31: 598.Wysocki T, Harris MA, Mauras N, Fox L, Taylor A, Jackson SC, White NH (2003). Absence ofadverse effects of severe hypoglycemia on cognitive function <strong>in</strong> school-aged children with diabetesover 18 months. <strong>Diabetes</strong> Care 26: 1100–5.

10 <strong>Hypoglycaemia</strong> <strong>in</strong> PregnancyAnn E. Gold and Donald W.M. PearsonINTRODUCTION<strong>Diabetes</strong> mellitus is one of the most common medical conditions affect<strong>in</strong>g women dur<strong>in</strong>gtheir reproductive years. A successful outcome of diabetic pregnancy can usually be anticipatedwith current management strategies, although an adverse outcome is still more commonthan <strong>in</strong> the non-diabetic population (Casson et al., 1997; Penney et al., 2003a; Evers et al.,2004; Jensen et al., 2004; Confidential Enquiry <strong>in</strong>to Maternal and Child Health, 2005).Meticulous control of blood glucose before conception and throughout gestation is the cornerstoneof management to reduce congenital anomalies, neonatal morbidity and mortality.However, striv<strong>in</strong>g for cont<strong>in</strong>uous normoglycaemia comes at a cost. Many women experiencean <strong>in</strong>creased frequency of hypoglycaemia, accompanied by impaired awareness ofhypoglycaemia or modification of their hypoglycaemic symptoms. This chapter describeswhy hypoglycaemia is a recognised problem dur<strong>in</strong>g pregnancy and how this <strong>in</strong>fluences themanagement of diabetic pregnancies.Population studies have shown that <strong>in</strong> many countries the average age of mothers withdiabetes (type 1 and type 2 diabetes) dur<strong>in</strong>g pregnancy is around 30 years (Penney et al.,2003a). At the time of their first pregnancy, women with type 1 diabetes have on average haddiabetes for over ten years, whereas some will have been exposed to the long-term effectsof chronic hyperglycaemia for much longer when they conceive. S<strong>in</strong>ce the microvascularcomplications of diabetes are associated with the duration of the condition, many women haveestablished microangiopathy at the time of conception. Careful preparation for pregnancyand regular obstetrical and medical surveillance throughout pregnancy and delivery aremandatory (SIGN, 2001; American <strong>Diabetes</strong> Association, 2003), along with rapid access tospecialist paediatric facilities for the neonate who may be heavy for dates and prematuredelivery, often by caesarean section.METABOLIC CHANGES DURING PREGNANCYFundamental changes occur <strong>in</strong> maternal metabolism and physiology dur<strong>in</strong>g pregnancy. Over280 days the mother’s weight <strong>in</strong>creases on average by 12.5 kg. The ma<strong>in</strong> <strong>in</strong>crease <strong>in</strong> weightoccurs <strong>in</strong> the second half of pregnancy and is caused by the growth of the conceptus, theenlargement of maternal organs, maternal storage of fat and prote<strong>in</strong> and an <strong>in</strong>crease <strong>in</strong>maternal blood volume and <strong>in</strong>terstitial fluid. An <strong>in</strong>crease <strong>in</strong> the basal metabolic rate results <strong>in</strong>the need for <strong>in</strong>creased energy <strong>in</strong>take. In addition throughout pregnancy maternal metabolism<strong>Hypoglycaemia</strong> <strong>in</strong> Cl<strong>in</strong>ical <strong>Diabetes</strong>, 2nd Edition.© 2007 John Wiley & Sons, LtdEdited by B.M. Frier and M. Fisher

218 HYPOGLYCAEMIA IN PREGNANCYadapts to ensure an adequate supply of nutrients to the grow<strong>in</strong>g fetus and develop<strong>in</strong>g placenta.A normal pregnancy is characterised by major alterations of glucose homeostasis. Fast<strong>in</strong>gglucose decl<strong>in</strong>es and, although the plasma glucose is elevated after an oral glucose tolerancetest, the mean plasma glucose level is around 4 mmol/l dur<strong>in</strong>g the third trimester of a normalnon-diabetic pregnancy on a normal diet (Paretti et al., 2001).Development of the placenta <strong>in</strong> the uterus dur<strong>in</strong>g the first trimester of pregnancy occurs<strong>in</strong> a low oxygen environment when maternal blood supply is restricted. Dur<strong>in</strong>g this timefetal metabolism is heavily anaerobic, which may serve to protect the develop<strong>in</strong>g embryofrom oxygen free radical-mediated teratogenesis. At the start of the second trimester, whenorganogenesis is complete, the maternal circulation develops to support fetal growth.In the second and third trimester the development of <strong>in</strong>sul<strong>in</strong> resistance leads to <strong>in</strong>creased<strong>in</strong>sul<strong>in</strong> secretion to avoid abnormal <strong>in</strong>creases <strong>in</strong> glucose, free fatty acids and am<strong>in</strong>o acids.In normal pregnancy <strong>in</strong>sul<strong>in</strong> sensitivity is decreased by between 30 and 60%. Chang<strong>in</strong>ghormonal levels make a major contribution to <strong>in</strong>sul<strong>in</strong> resistance. Human placental lactogen(HPL) has actions similar to growth hormone. It <strong>in</strong>creases lipolysis with a rise <strong>in</strong> freefatty acids which are a steady source of energy for the mother and fetus dur<strong>in</strong>g periodsof starvation. Progesterone is also associated with <strong>in</strong>sul<strong>in</strong> resistance. Maternal lipid stores<strong>in</strong>crease dur<strong>in</strong>g pregnancy and adipok<strong>in</strong>es and cytok<strong>in</strong>es may play a role <strong>in</strong> the developmentof <strong>in</strong>creas<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> resistance. The cytok<strong>in</strong>e tumour necrosis factor-alpha (TNF-) risesas the fat mass <strong>in</strong>creases and can be related to <strong>in</strong>sul<strong>in</strong> resistance. In pregnant women adecrease of adiponect<strong>in</strong> has been shown to relate to <strong>in</strong>creas<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> resistance <strong>in</strong> the thirdtrimester. In women with type 1 diabetes the physiological development of <strong>in</strong>sul<strong>in</strong> resistancedur<strong>in</strong>g pregnancy poses challenges to the expectant mother who is attempt<strong>in</strong>g to ma<strong>in</strong>ta<strong>in</strong>normoglycaemia.Many other changes <strong>in</strong> physiology occur <strong>in</strong> pregnancy. The complex process of placentaldevelopment is mostly complete by the end of the second trimester though the placentacont<strong>in</strong>ues to expand with the grow<strong>in</strong>g fetus. In the third trimester maternal metabolismswitches from anabolism to catabolism, permitt<strong>in</strong>g an enhanced transfer of nutrients acrossthe placenta to susta<strong>in</strong> rapid fetal growth. The placenta is an active organ <strong>in</strong> this process. Inaddition to synthesis<strong>in</strong>g various hormones the placenta regulates the transfer of maternal fuelsto the fetus and facilitates maternal metabolic adaptation at different stages of pregnancy.Cells <strong>in</strong> contact with the maternal circulation and fetal circulation have a range of receptors,transporters and channels on both placental surfaces.At term the placenta of the mother with diabetes shows a number of differences fromthose <strong>in</strong> women who do not have diabetes. These <strong>in</strong>clude changes <strong>in</strong> morphology, bloodflow, transport and metabolism. This is important s<strong>in</strong>ce transplacental transport of glucoseis a facilitated process and net transfer is strongly dependent on the concentration gradientof glucose between the maternal and fetal blood. However, the correlation between various<strong>in</strong>dices of glucose control – e.g. HbA 1c and fetal growth – is poor, suggest<strong>in</strong>g that factorsother than maternal hyperglycaemia contribute to accelerated fetal growth (Penney et al.,2003b). Up-regulation of placental glucose transporters <strong>in</strong> type 1 diabetes may contributeto <strong>in</strong>creased placental glucose transfer and stimulate fetal growth even if the mother hasexcellent glycaemic control. Transport of am<strong>in</strong>o acids across the human placenta is an activeprocess result<strong>in</strong>g <strong>in</strong> am<strong>in</strong>o acid concentrations <strong>in</strong> the fetal circulation that are substantiallyhigher than those <strong>in</strong> the maternal circulation.Management strategies <strong>in</strong> women with type 1 diabetes need to take the metabolic adaptationsof pregnancy <strong>in</strong>to account. Although <strong>in</strong>sul<strong>in</strong> resistance is the characteristic feature

FREQUENCY OF HYPOGLYCAEMIA IN DIABETIC PREGNANCY 219of the later stage of pregnancy, <strong>in</strong> the first trimester a modest <strong>in</strong>crease <strong>in</strong> <strong>in</strong>sul<strong>in</strong> sensitivityoccurs. The <strong>Diabetes</strong> In Early Pregnancy study (DIEP) reported decl<strong>in</strong><strong>in</strong>g <strong>in</strong>sul<strong>in</strong>requirements <strong>in</strong> the middle of the first trimester of pregnancy <strong>in</strong> women with type1 diabetes (Jovanovic et al., 2001). Over-<strong>in</strong>sul<strong>in</strong>isation at this stage may be an issues<strong>in</strong>ce women will be striv<strong>in</strong>g for optimal glycaemic control dur<strong>in</strong>g the crucial periodof organogenesis. Hyperemesis gravidarum may also contribute to an <strong>in</strong>creased risk ofhypoglycaemia.FREQUENCY OF HYPOGLYCAEMIA IN DIABETIC PREGNANCY<strong>Hypoglycaemia</strong> is a frequent problem experienced by women with diabetes dur<strong>in</strong>g pregnancy.A number of studies have reported the frequency of hypoglycaemia dur<strong>in</strong>g pregnancy <strong>in</strong>women with diabetes but comparison is made difficult by the variations used <strong>in</strong> the def<strong>in</strong>itionof hypoglycaemia and the methods used to collect the data. Table 10.1 summarises thefrequency of hypoglycaemia <strong>in</strong> some studies of women with pre-gestational diabetes. In all ofthe studies, with the exception of that by Persson and Hanson (1993), severe hypoglycaemiawas common dur<strong>in</strong>g pregnancy <strong>in</strong> women with pre-gestational diabetes. In the study byPersson and Hanson (1993), a lower <strong>in</strong>cidence of hypoglycaemia was reported us<strong>in</strong>g an<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimen comb<strong>in</strong>ed with very frequent self-monitor<strong>in</strong>g, which may partlyaccount for the difference <strong>in</strong> frequency as compared with the other studies.Most studies have demonstrated that the peak <strong>in</strong>cidence of hypoglycaemia occurs dur<strong>in</strong>gthe first and second trimesters. Kimmerle et al. (1992) observed that 84% of hypoglycaemicepisodes that resulted <strong>in</strong> impaired consciousness occurred before week 20. A peak <strong>in</strong>cidenceof hypoglycaemia was observed dur<strong>in</strong>g weeks 10–15 <strong>in</strong> a study of women receiv<strong>in</strong>g<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy (Rosenn et al., 1995). In the <strong>Diabetes</strong> Control and ComplicationsTrial (DCCT), a similar number of episodes of hypoglycaemia was recorded <strong>in</strong> the firstand second trimesters and fewer episodes were reported dur<strong>in</strong>g the third trimester. In alarger study of 323 women, severe hypoglycaemia was almost 2.5 times more frequent <strong>in</strong>the first trimester compared with the third trimester (Evers et al., 2002a; 2004). However,the reported <strong>in</strong>cidence varies considerably between studies, which may represent differences<strong>in</strong> patient groups and management strategies as well as vary<strong>in</strong>g def<strong>in</strong>itions of severehypoglycaemia.Nocturnal hypoglycaemia is particularly common dur<strong>in</strong>g pregnancy (Kimmerle et al.,1992). The advent of cont<strong>in</strong>uous blood glucose monitor<strong>in</strong>g (CBGM) has confirmed this high<strong>in</strong>cidence of nocturnal hypoglycaemia dur<strong>in</strong>g pregnancy <strong>in</strong> mothers with type 1 diabetes(Yogev et al., 2003). In this study, 34 women were monitored for a 72-hour period betweenweeks 16 and 32 of pregnancy. Dur<strong>in</strong>g this short time period, nocturnal hypoglycaemic eventswere recorded <strong>in</strong> 26 (76%) women but only 17 of the patients experienced symptoms. In allof the affected patients an <strong>in</strong>terval of 1–4 hours elapsed before any cl<strong>in</strong>ical manifestationsof hypoglycaemia were apparent.Comparison of two national audits of diabetic pregnancy <strong>in</strong> Scotland has shown thatsignificantly more women experienced severe hypoglycaemia <strong>in</strong> 1998–1999 than <strong>in</strong> 2003–2004 (41 versus 30%). This difference may be expla<strong>in</strong>ed by a number of factors, such asmore women attend<strong>in</strong>g for pre-pregnancy care <strong>in</strong> 2003–2004 and the application of newer<strong>in</strong>sul<strong>in</strong> regimens that employed short-act<strong>in</strong>g analogues (Scottish <strong>Diabetes</strong> <strong>in</strong> Pregnancy StudyGroup, 2004).

220 HYPOGLYCAEMIA IN PREGNANCYTable 10.1pregnancySummary of studies of the <strong>in</strong>cidence of ‘severe’ hypoglycaemia <strong>in</strong> pre-gestational diabeticReferenceType ofdiabetesDef<strong>in</strong>ition of severehypoglycaemiaWomen experienc<strong>in</strong>gseverehypoglycaemiadur<strong>in</strong>g pregnancynKimmerleet al., 1992Persson et al.,1993Rosenn et al.,1995Type 1Type 1Type 1Impaired consciouslevel respond<strong>in</strong>g toglucose/glucagonRequir<strong>in</strong>g externalhelp for recoverySubdivided <strong>in</strong>to:a) requir<strong>in</strong>gexternal helpb) coma/seizureDCCT, 1996 Type 1 Seizure or loss ofconsciousnessMasson et al.,2003Garg et al.,2003Evers et al.2002a; 2004Scottish<strong>Diabetes</strong> <strong>in</strong>PregnancyStudy Group,2004Type 1 (alltak<strong>in</strong>g lispro<strong>in</strong>sul<strong>in</strong>)Type 1 (alltak<strong>in</strong>g lispro<strong>in</strong>sul<strong>in</strong>)Type 1Type 1Requir<strong>in</strong>g externalhelp for recoveryRequir<strong>in</strong>g externalhelp for recoveryRequir<strong>in</strong>g externalhelp for recoveryRequir<strong>in</strong>g externalhelp for recovery41% (77% episodesoccurred dur<strong>in</strong>gsleep)4.4% 11371% 34% 8417% <strong>in</strong> <strong>in</strong>tensivegroup19.8% <strong>in</strong>conventionalgroup27% 7623% overall 6241% 1st trimester17% 3rd trimester30% overall20% 1st trimester17% 2nd trimester10% 3rd trimester77(85pregnancies)180(270pregnancies)278 (2002)323 (2004)155<strong>Hypoglycaemia</strong> also occurs <strong>in</strong> women with type 2 diabetes, who often require <strong>in</strong>sul<strong>in</strong> <strong>in</strong>the pre-pregnancy period <strong>in</strong> order to optimise glycaemic control, and also <strong>in</strong> women whodevelop gestational diabetes (typically <strong>in</strong> the late second and third trimesters). However,the frequency of hypoglycaemia has not been documented with accuracy <strong>in</strong> these groups; itis the authors’ impression that hypoglycaemia occurs much less frequently than <strong>in</strong> type 1diabetes.Why are Women with Pre-gestational <strong>Diabetes</strong> at Greater Risk of<strong>Hypoglycaemia</strong> dur<strong>in</strong>g Pregnancy?Great importance is placed upon ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g good glycaemic control throughout diabeticpregnancy when women are highly motivated and often achieve levels of glycated

FREQUENCY OF HYPOGLYCAEMIA IN DIABETIC PREGNANCY 221haemoglob<strong>in</strong> that are close to the non-diabetic range. In addition, because of the changes<strong>in</strong> <strong>in</strong>sul<strong>in</strong> sensitivity <strong>in</strong> the first trimester, their risk of severe hypoglycaemia is <strong>in</strong>creased.It has been shown that women with a previous history of hypoglycaemia are at <strong>in</strong>creasedrisk of experienc<strong>in</strong>g severe hypoglycaemia dur<strong>in</strong>g pregnancy (Kimmerle et al., 1992; Everset al., 2002a). Women who demonstrate a greater fluctuation <strong>in</strong> blood glucose dur<strong>in</strong>g pregnancyare also at greater risk of experienc<strong>in</strong>g severe hypoglycaemia (Rosenn et al., 1995)and women who are not us<strong>in</strong>g an ‘<strong>in</strong>tensive’ <strong>in</strong>sul<strong>in</strong> regimen have an enhanced risk. Everset al. (2002a) observed that the risks of experienc<strong>in</strong>g severe hypoglycaemia dur<strong>in</strong>g the firsttrimester were significantly greater <strong>in</strong> women who had duration of diabetes greater than tenyears, HbA 1c less than 6.5% or an <strong>in</strong>sul<strong>in</strong> dose of greater than 0.1 units of <strong>in</strong>sul<strong>in</strong> per kgbody weight.It is possible that impaired counterregulation could contribute to the <strong>in</strong>creased risk ofhypoglycaemia. A rat model demonstrated that the glucagon and ep<strong>in</strong>ephr<strong>in</strong>e (but notnorep<strong>in</strong>ephr<strong>in</strong>e) responses to hypoglycaemia (plasma glucose 3.4 mmol/l) were suppresseddur<strong>in</strong>g pregnancy (Rossi et al., 1993); these data suggest that the counterregulatoryresponses to hypoglycaemia <strong>in</strong> rats may be impaired. A few studies have exam<strong>in</strong>edhormonal counterregulatory responses <strong>in</strong> pregnant women with type 1 diabetes dur<strong>in</strong>gexperimental hypoglycaemia dur<strong>in</strong>g pregnancy. All studies utilised a hyper<strong>in</strong>sul<strong>in</strong>aemichypoglycaemic clamp technique, except that by Nisell et al. (1994), which used an<strong>in</strong>sul<strong>in</strong> bolus <strong>in</strong>jection to <strong>in</strong>duce hypoglycaemia. The results of these studies are shown <strong>in</strong>Table 10.2.The studies vary <strong>in</strong> terms of the time of gestation at the time of study and theglucose nadir reached and it is difficult to determ<strong>in</strong>e whether counterregulation is impaireddur<strong>in</strong>g pregnancy as a result of pregnancy per se or because of differences <strong>in</strong> glycaemiccontrol and the presence of diabetes. The study by Rosenn et al. (1996) probably hadthe most appropriate study design to answer this question. One criticism, however, isthat the glucose nadir achieved was only 3.3 mmol/l, which may not have been sufficientto stimulate a counterregulatory response <strong>in</strong> the non-diabetic control group. The studiesprovide conflict<strong>in</strong>g evidence as to whether the counterregulatory response to hypoglycaemiais deficient dur<strong>in</strong>g pregnancy. No studies have exam<strong>in</strong>ed counterregulatory effectsdur<strong>in</strong>g the first trimester at the time when the frequency of, and exposure to, hypoglycaemiaare at a peak. However, to conduct such studies would raise major ethicalconcerns.Few studies have exam<strong>in</strong>ed the development of impaired hypoglycaemia awarenessdur<strong>in</strong>g pregnancy although most cl<strong>in</strong>icians would agree that this is particularly problematicaldur<strong>in</strong>g the first trimester. Evers et al. (2002a) observed that severe hypoglycaemia<strong>in</strong> the first trimester was more likely to occur <strong>in</strong> women with reduced symptomaticawareness of hypoglycaemia. In laboratory-<strong>in</strong>duced hypoglycaemia Björklundet al. (1998a) did measure symptomatic responses to hypoglycaemia dur<strong>in</strong>g the thirdtrimester, and also postnatally, and found that symptoms such as ‘<strong>in</strong>ability to concentrate’,‘headache’ and ‘pound<strong>in</strong>g heart’ were less prom<strong>in</strong>ent dur<strong>in</strong>g pregnancy comparedwith dur<strong>in</strong>g the postnatal period. However, it is difficult to ascerta<strong>in</strong> whether thisis a consequence of differences <strong>in</strong> glycaemic control or the <strong>in</strong>cidence of hypoglycaemiadur<strong>in</strong>g these two time periods. Fear of hypoglycaemia is also greater <strong>in</strong> womenwho have experienced severe hypoglycaemia (Evers et al., 2002a) and this is animportant problem for the diabetic mother which should be addressed dur<strong>in</strong>g antenatalcare.

222 HYPOGLYCAEMIA IN PREGNANCYTable 10.2 Summary of studies exam<strong>in</strong><strong>in</strong>g counterregulatory responses to hypoglycaemia <strong>in</strong> womenwith pre-gestational diabetesReferenceGestationat time ofstudyControl groupBloodglucose nadirdur<strong>in</strong>g studyResultsnDiamondet al.,1992Nisellet al.,1994Rosennet al.,1996Björklundet al.,1998aBjörklundet al.,1998b21–37 weeks Non-pregnant,non-diabeticage matchedwomenLasttrimester,and 8–12weekspost-partumNon-pregnant(1);24–28 (2);32–34 (3)30–34 weeks(1);5–13 monthspostnatal(2)Acted as owncontrolsNon-diabetic,age-matched.Cases werealso studiedon 3occasionsi.e. acted asown controlsActed as owncontrols2.5 mmol/l No glucagon response <strong>in</strong>cases.Ep<strong>in</strong>ephr<strong>in</strong>e releasesuppressed <strong>in</strong> cases; cf.controls.Lower blood glucose forep<strong>in</strong>ephr<strong>in</strong>e and growthhormone release <strong>in</strong>cases; cf. controls3.2 mmol/l No glucagon or cortisolresponse at either timepo<strong>in</strong>t.Similar <strong>in</strong>creases <strong>in</strong>ep<strong>in</strong>ephr<strong>in</strong>e andnorep<strong>in</strong>ephr<strong>in</strong>e onboth occasions3.3 mmol/l Reduced ep<strong>in</strong>ephr<strong>in</strong>eresponse <strong>in</strong> cases; cf.controls at all studytimes.Reduced ep<strong>in</strong>ephr<strong>in</strong>eresponse <strong>in</strong> casesdur<strong>in</strong>g pregnancy; cf.pre-pregnancy.Growth hormone responsesreduced <strong>in</strong> pregnancy <strong>in</strong>cases and controls; cf.pre-pregnancy2.3 mmol/l Ep<strong>in</strong>ephr<strong>in</strong>e responsesimilar <strong>in</strong> pregnancy andpostnatal.Dehydroepiandrosterone<strong>in</strong>creased more rapidlyand less susta<strong>in</strong>eddur<strong>in</strong>g pregnancy.Growth hormone responsesreduced <strong>in</strong> pregnancy30–34 weeks No controls 2.3 mmol/l Increase <strong>in</strong> placentalgrowth hormone, but nochanges <strong>in</strong> otherplacental hormonesdur<strong>in</strong>g hypoglycaemia9 type 1cases;7 controls8 pregestationaltype 1patients;1 gestational17 type 1cases;10 controls1010

CLINICAL MANAGEMENT BEFORE AND DURING PREGNANCY 223CLINICAL MANAGEMENT BEFORE AND DURING PREGNANCYPre-conception CareThe advantages of planned pregnancy should be regularly emphasised to women withdiabetes dur<strong>in</strong>g their reproductive years and effective contraceptive advice should be part ofrout<strong>in</strong>e cl<strong>in</strong>ical care. After menarche, all girls with type 1 diabetes should be aware of theimportance of pregnancy plann<strong>in</strong>g, because the <strong>in</strong>fants of mothers who have attended forpre-pregnancy care have fewer major congenital malformations and require shorter periods<strong>in</strong> special care facilities than <strong>in</strong>fants of mothers who do not attend for pre-conception care(Fuhrmann et al., 1983; Steel et al., 1989; 1990; Kitzmiller et al., 1991; Ray et al., 2001).Attendance for structured pre-pregnancy care is also associated with a reduction <strong>in</strong> the rate ofspontaneous abortion. Pre-pregnancy counsell<strong>in</strong>g should address ways of m<strong>in</strong>imis<strong>in</strong>g the riskof severe hypoglycaemia both before and dur<strong>in</strong>g pregnancy. Ideally this should be discusseddur<strong>in</strong>g the pre-pregnancy period. However, if the pregnancy has not been planned, womenshould be made aware of their <strong>in</strong>creased risk of hypoglycaemia dur<strong>in</strong>g pregnancy and howto avoid and manage potential episodes, particularly as the greatest risk is dur<strong>in</strong>g the firsttrimester.Organisation of Cl<strong>in</strong>ical CareIn many centres cl<strong>in</strong>ical care is delivered by a multidiscipl<strong>in</strong>ary comb<strong>in</strong>ed obstetric/diabeticteam with very regular out-patient reviews to assess metabolic control and obstetric progress(Figure 10.1). Home blood glucose monitor<strong>in</strong>g results are assessed and <strong>in</strong>sul<strong>in</strong> regimenand dietary <strong>in</strong>take modified to optimise glycaemic control and HbA 1c (Figure 10.2). Mostwomen present for book<strong>in</strong>g at around eight weeks gestation when an early scan will providean accurate estimate of gestational age. This is important to allow the optimal time ofdelivery to be determ<strong>in</strong>ed. Screen<strong>in</strong>g is performed rout<strong>in</strong>ely for Down’s syndrome andneural-tube defects. Although the prevalence of congenital anomaly has decl<strong>in</strong>ed follow<strong>in</strong>gthe <strong>in</strong>troduction of pre-conceptional counsell<strong>in</strong>g, the <strong>in</strong>cidence of congenital malformationis still higher than <strong>in</strong> the non-diabetic population. A detailed ultrasound scan at around 20weeks is performed to detect severe congenital anomalies, particularly to identify majormalformation of the heart and the central nervous system. Frequent scann<strong>in</strong>g is performedlater <strong>in</strong> pregnancy to monitor fetal growth. In the third trimester regular cardiotocography,Doppler ultrasound and fetal movement charts are used to monitor fetal progress. Figure 10.3demonstrates the measurement of abdom<strong>in</strong>al circumference (AC). Sequential measurementsof AC are used to monitor fetal growth.Good glycaemic control reduces stillbirth rate, neonatal hypoglycaemia and respiratorydistress syndrome. Women should strive to ma<strong>in</strong>ta<strong>in</strong> blood glucose levels as near to the nondiabeticrange as possible without an excessive risk of hypoglycaemia. This usually meansblood glucose target levels between 4 and 7 mmol/l. The diabetes team, but <strong>in</strong> particularthe diabetes specialist nurses and specialist midwives, have an important role <strong>in</strong> educat<strong>in</strong>gwomen on the need for home blood glucose monitor<strong>in</strong>g (usually four to six times a day) and<strong>in</strong> <strong>in</strong>troduc<strong>in</strong>g <strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> regimens if the women are not already on these programmes.Maternal issues <strong>in</strong> pregnancy are shown <strong>in</strong> Table 10.3.

224 HYPOGLYCAEMIA IN PREGNANCYType 1 and type 2 diabetes <strong>in</strong> pregnancy;Gestational diabetes on <strong>in</strong>sul<strong>in</strong>Blood tests Other Tests Scann<strong>in</strong>g Miscellaneous Counsell<strong>in</strong>gPre-pregnancy HbA 1c Hb U&E TFT Rubella Eyes MSU/MA Folic acidDieticianMidwife

Figure 10.2 Example of home blood glucose monitor<strong>in</strong>g chart used <strong>in</strong> the comb<strong>in</strong>ed antenatal diabetic cl<strong>in</strong>ic at Aberdeen Maternity Hospital, Scotland

226 HYPOGLYCAEMIA IN PREGNANCYFigure 10.3 Ultrasound scan of at 32 weeks gestation demonstrat<strong>in</strong>g measurement of abdom<strong>in</strong>alcircumference (AC)Optimis<strong>in</strong>g Insul<strong>in</strong> RegimensData is limited with regard to studies of different <strong>in</strong>sul<strong>in</strong> regimens dur<strong>in</strong>g pregnancy andmany studies are small and observational <strong>in</strong> nature. The effects of conventional versus<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy were studied <strong>in</strong> the women participat<strong>in</strong>g <strong>in</strong> the DCCT (The <strong>Diabetes</strong>Control and Complications Trial Research Group, 1996). In women who had <strong>in</strong>itially been

CLINICAL MANAGEMENT BEFORE AND DURING PREGNANCY 227Table 10.3Maternal issues <strong>in</strong> pregnancy<strong>Diabetes</strong>-related<strong>Hypoglycaemia</strong>Increased risk of ketoacidosisRet<strong>in</strong>opathy may deteriorateNephropathy and fluid retentionPre-eclampsia/hypertensionIncreased caesarean section rateObstetricPreterm deliveryIncreased caesarean section rateMacrosomia and shoulder dystociaassigned to conventional treatment, the occurrence of hypoglycaemia result<strong>in</strong>g <strong>in</strong> impairedconscious level was 4.0 per 100 pregnant-months compared with 2.8 <strong>in</strong> those receiv<strong>in</strong>g<strong>in</strong>tensive treatment. Those women <strong>in</strong> the conventional group who had changed to <strong>in</strong>tensivetherapy before conception experienced significantly fewer episodes of hypoglycaemia dur<strong>in</strong>gpregnancy than those who had not <strong>in</strong>tensified their regimen until after conception (2.0 versus5.0 episodes per 100 patient years). This suggests that it may be preferable, where possible,to <strong>in</strong>tensify the <strong>in</strong>sul<strong>in</strong> regimen dur<strong>in</strong>g the pre-pregnancy period rather than mak<strong>in</strong>g thechange once conception has occurred.Another study of 118 women with pre-gestational diabetes compared twice daily freemix<strong>in</strong>g of soluble and isophane <strong>in</strong>sul<strong>in</strong>s with a basal-bolus regimen of soluble <strong>in</strong>sul<strong>in</strong>before meals and bedtime isophane <strong>in</strong>sul<strong>in</strong> (Nachum et al., 1999). The authors reported nodifference between the regimens <strong>in</strong> the <strong>in</strong>cidence of hypoglycaemia, although those tak<strong>in</strong>gfour <strong>in</strong>jections of <strong>in</strong>sul<strong>in</strong> per day achieved better glycaemic control. One <strong>in</strong>terpretation ofthe data is that the greater the <strong>in</strong>sul<strong>in</strong> dosage, the better the glycaemic control. However, theoverall <strong>in</strong>cidence of recorded hypoglycaemia was remarkably low (10%) and almost half thewomen had type 2 diabetes.Short-act<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> analogues have received some attention and one randomised trial of<strong>in</strong>sul<strong>in</strong> lispro <strong>in</strong> pregnancies complicated by type 1 diabetes has been reported (Perssonet al., 2002). Thirty-three women were studied with 16 be<strong>in</strong>g randomised to lispro and 17 tosoluble <strong>in</strong>sul<strong>in</strong> at gestational week 15. Most of the women had very good glycaemic controlat entry to the study (mean HbA 1c 6.5%, range 4.5–8.6%). Only two patients experiencedsevere hypoglycaemia dur<strong>in</strong>g the rema<strong>in</strong>der of their pregnancies and both were on humansoluble <strong>in</strong>sul<strong>in</strong>, although it is important to remember that the study was undertaken <strong>in</strong>the latter half of pregnancy when the risk of hypoglycaemia is usually lower. However, asignificantly higher rate of biochemical hypoglycaemia (< 30 mmol/l) was documented <strong>in</strong>the women us<strong>in</strong>g lispro (who were monitor<strong>in</strong>g blood glucose six times daily). No differenceswere observed between the groups with regard to glycated haemoglob<strong>in</strong>, although womentak<strong>in</strong>g lispro had lower postprandial glucose excursions. In an observational study <strong>in</strong> whichwomen received lispro before and dur<strong>in</strong>g pregnancy, the prevalence of severe hypoglycaemiawas 23%, with no greater <strong>in</strong>cidence of adverse outcomes compared to other regimens(Garg et al., 2003). A similar prospective study that was designed to assess progressionof ret<strong>in</strong>opathy dur<strong>in</strong>g pregnancy <strong>in</strong> women on lispro, demonstrated that the number ofsubjective hypoglycaemic events did not differ between those tak<strong>in</strong>g lispro and those tak<strong>in</strong>gsoluble <strong>in</strong>sul<strong>in</strong>, although the overall rates of hypoglycaemia were relatively low (Loukovaaraet al., 2003).

228 HYPOGLYCAEMIA IN PREGNANCYIn gestational diabetes, when lispro was compared with human soluble (regular) <strong>in</strong>sul<strong>in</strong>,a significant difference was demonstrated <strong>in</strong> the <strong>in</strong>cidence of hypoglycaemia (Jovanovicet al., 1999). However, the recorded <strong>in</strong>cidence of hypoglycaemia was aga<strong>in</strong> very low and itis not appropriate to extrapolate this data to pregnancy <strong>in</strong> type 1 diabetes. In a recent largestudy of diabetic pregnancy from the Netherlands (Evers et al., 2004), the rate of severehypoglycaemia <strong>in</strong> a subgroup of 35 patients who were us<strong>in</strong>g lispro, did not differ fromthat observed <strong>in</strong> patients us<strong>in</strong>g other <strong>in</strong>sul<strong>in</strong> regimens. The congenital malformation rates<strong>in</strong> those women us<strong>in</strong>g lispro have not differed from those on soluble <strong>in</strong>sul<strong>in</strong> (Wyatt et al.,2004). A recently reported multi-centre study by Mathies on et al. (2007) has demonstratedthe safe use of <strong>in</strong>sul<strong>in</strong> aspart <strong>in</strong> Type 1 diabetic pregnancy, with a non-significantly lower<strong>in</strong>cidence of severe hypoglycaemia compared to human soluble <strong>in</strong>sul<strong>in</strong> at comparable levelsof glycaemic control (Hb A 1c mostly

CLINICAL MANAGEMENT BEFORE AND DURING PREGNANCY 229Table 10.4 Insul<strong>in</strong> treatment <strong>in</strong> Gestational <strong>Diabetes</strong> Mellitus (GDM) (SIGN 2001)• Intensive <strong>in</strong>sul<strong>in</strong> regimens are occasionally recommended to achieve normal bloodglucose levels and fetal growth• Women requir<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> require a comprehensive education package regard<strong>in</strong>g allaspects of <strong>in</strong>sul<strong>in</strong> treatmentConsider <strong>in</strong>sul<strong>in</strong> treatment for GDM if:• The fast<strong>in</strong>g blood glucose is over 6 mmol/l or the 2 hr postprandial glucose is over7 mmol/l, despite appropriate lifestyle adjustments• The ultrasound scans show excessive growth with an abdom<strong>in</strong>al circumference greaterthan the 95 percentile for gestation (some authorities recommend > 75 percentile)pregnancy and labour. In some European countries, a significant percentage of womenuse this type of treatment dur<strong>in</strong>g pregnancy but <strong>in</strong> many countries only a relatively smallpercentage use CSII.Although an appropriate <strong>in</strong>sul<strong>in</strong> regimen is very important many other factors can <strong>in</strong>fluence<strong>in</strong>sul<strong>in</strong> absorption and effectiveness. Pregnancy is a good time to review all factors that couldcontribute to suboptimal glycaemic control. All practical aspects of <strong>in</strong>sul<strong>in</strong> adm<strong>in</strong>istrationshould be reviewed and <strong>in</strong>jection sites exam<strong>in</strong>ed.Some women with gestational diabetes may require <strong>in</strong>sul<strong>in</strong> dur<strong>in</strong>g the last few weeks ofpregnancy. The management of <strong>in</strong>sul<strong>in</strong> treatment <strong>in</strong> gestational diabetes is summarised <strong>in</strong>Table 10.4.Dietary and Lifestyle ManagementDietary advice is also essential before, dur<strong>in</strong>g and after pregnancy. Such advice willencourage the <strong>in</strong>take of foods with a high level of complex carbohydrates, soluble fibreand vitam<strong>in</strong>s. Balanc<strong>in</strong>g of carbohydrate <strong>in</strong>take and <strong>in</strong>sul<strong>in</strong> dose is important and manywomen may f<strong>in</strong>d it helpful to count carbohydrate portions and learn appropriate correctionfactors for high blood glucose levels. Dietetic review is therefore appropriate and a reviewof ‘sick-day rules’ should be undertaken. In the early weeks of pregnancy, where nauseaand vomit<strong>in</strong>g are common, this is particularly important. Dur<strong>in</strong>g the later weeks of pregnancy,women may prefer to take smaller but more frequent meals and their <strong>in</strong>sul<strong>in</strong> regimenshould be adjusted accord<strong>in</strong>gly. In general, women are also advised to reduce their <strong>in</strong>take ofsaturated fat.Neural tube defects <strong>in</strong> high risk pregnancies are associated with low levels of folate. A largestudy of non-diabetic women has shown that the prescription of 4 mg of folate supplement,pre- and peri-conception, provides protection aga<strong>in</strong>st neural tube defects <strong>in</strong> women at highrisk. In Scotland, all women are advised to take 5 mg of folic acid pre-conception, andcont<strong>in</strong>ue this preparation dur<strong>in</strong>g pregnancy until around 12 weeks of gestation. Smok<strong>in</strong>gshould be discouraged. If women are used to tak<strong>in</strong>g regular physical activity this should beencouraged, along with advice about reduc<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> dosages before exercise and ensur<strong>in</strong>gappropriate carbohydrate <strong>in</strong>take.

230 HYPOGLYCAEMIA IN PREGNANCYCorrection of <strong>Hypoglycaemia</strong>Advice concern<strong>in</strong>g the correction of hypoglycaemia by appropriate dietary <strong>in</strong>take shouldbe given and is similar to that for the non-pregnant patient with diabetes (see Chapter 5).Patients, their partners and other family members should be advised on the use of glucosegels, such as GlucoGel, for emergency use, and should keep a supply of this available.Instruction about the use of glucagon should also be given to partners or close familymembers. This should be done at the first review, s<strong>in</strong>ce hypoglycaemia is more common <strong>in</strong>the first trimester.Management of DeliveryThe <strong>in</strong>fant of a diabetic mother is at <strong>in</strong>creased risk dur<strong>in</strong>g labour and delivery. Mothers withdiabetes must be recognised to be a high risk obstetric patient and delivered <strong>in</strong> a unit thatcan provide experienced obstetric care and immediate access to neonatal <strong>in</strong>tensive support.Tim<strong>in</strong>g of DeliveryThere is no evidence to support the cont<strong>in</strong>uation of a type 1 diabetic pregnancy beyond40 weeks and, provided glycaemic control has been satisfactory and fetal growth has notbeen excessive, <strong>in</strong>duction of labour and delivery at around 39 weeks appears to provide acompromise between deliver<strong>in</strong>g the baby too early and avoid<strong>in</strong>g late unexpla<strong>in</strong>ed <strong>in</strong>trauter<strong>in</strong>edeath.Management of <strong>Diabetes</strong> dur<strong>in</strong>g LabourManagement of diabetes <strong>in</strong> labour is summarised <strong>in</strong> Table 10.5. The aim is to achievematernal normoglycaemia dur<strong>in</strong>g labour us<strong>in</strong>g an <strong>in</strong>travenous dextrose/<strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion.The <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion rate must be halved immediately after delivery to prevent postnatalhypoglycaemia, s<strong>in</strong>ce there is a rapid <strong>in</strong>crease <strong>in</strong> <strong>in</strong>sul<strong>in</strong> sensitivity after separation of theplacenta. After delivery, the <strong>in</strong>sul<strong>in</strong> dose should be reduced to the dosages used beforepregnancy. This will usually necessitate halv<strong>in</strong>g the doses required dur<strong>in</strong>g pregnancy. If themother decides on breast feed<strong>in</strong>g, a further reduction of the <strong>in</strong>sul<strong>in</strong> dosage may be requiredto avoid hypoglycaemia.MATERNAL COMPLICATIONS OF DIABETES DURINGPREGNANCYThe complications experienced dur<strong>in</strong>g pregnancy <strong>in</strong>clude those result<strong>in</strong>g from hypoglycaemiaand from the microvascular complications of diabetes.

MATERNAL COMPLICATIONS OF DIABETES DURING PREGNANCY 231Table 10.5 Management of diabetes dur<strong>in</strong>g labour and delivery (cl<strong>in</strong>icalcircumstances may lead to modifications <strong>in</strong> this regimen)• Fast the patient on the day of delivery for an elective section; otherwisefast from the onset of established labour.• Set up an <strong>in</strong>travenous <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion us<strong>in</strong>g a syr<strong>in</strong>ge pump, e.g. 50 unitsshort-act<strong>in</strong>g analogue <strong>in</strong> 50 ml sal<strong>in</strong>e. Start at 1–2 units per hour.• Simultaneously commence a 10% dextrose +01% KCI <strong>in</strong>travenous <strong>in</strong>fusionat a rate of 100 ml/hr.• Monitor bedside glucose hourly and set a target blood glucose of4–6 mmol/l.• Use the algorithm (figure) to adjust the <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion.• Halve the rate of the <strong>in</strong>sul<strong>in</strong> <strong>in</strong>fusion on separation and delivery of theplacenta. Cont<strong>in</strong>ue the dextrose <strong>in</strong>fusion as before.• Revert to the pre-pregnancy dosages of subcutaneous <strong>in</strong>sul<strong>in</strong> when anormal eat<strong>in</strong>g pattern is established.Insul<strong>in</strong> <strong>in</strong>fusion scaleStart <strong>in</strong>sul<strong>in</strong> at 2 units per hour.Monitor capillary bedside glucose hourly.Capillary blood glucose (mmol/l)< 4 4–6 > 6Reduce <strong>in</strong>sul<strong>in</strong> by Cont<strong>in</strong>ue current rate Increase by 0.5 unit/hr0.5 unit/hrThe Risks of Maternal <strong>Hypoglycaemia</strong> to the MotherThe morbidity and mortality of hypoglycaemia <strong>in</strong> diabetes is discussed <strong>in</strong> Chapter 12.Dur<strong>in</strong>g pregnancy the risks are similar but compounded by the potential risk to the fetuscaused by <strong>in</strong>jury. A maternal death was reported <strong>in</strong> one series (Rayburn et al., 1986).One particular concern <strong>in</strong> the pre-pregnancy and pregnancy cl<strong>in</strong>ics is the risk of hypoglycaemiawhile driv<strong>in</strong>g. One woman was <strong>in</strong>volved <strong>in</strong> a motor vehicle accident after experienc<strong>in</strong>ga hypoglycaemia-<strong>in</strong>duced convulsion and fractur<strong>in</strong>g her tibia and fibula. Whenplann<strong>in</strong>g pregnancy, advice on the avoidance of hypoglycaemia while driv<strong>in</strong>g should bere<strong>in</strong>forced for patients treated with <strong>in</strong>sul<strong>in</strong> (see Chapter 14), particularly with respectto test<strong>in</strong>g blood glucose before every journey and ensur<strong>in</strong>g appropriate management ofhypoglycaemia if it occurs while driv<strong>in</strong>g. They should also be warned that they willbe advised to cease driv<strong>in</strong>g if their warn<strong>in</strong>g symptoms of hypoglycaemia are significantlyreduced. Women who develop gestational diabetes and require <strong>in</strong>sul<strong>in</strong> treatmentshould be given similar advice although the risk of develop<strong>in</strong>g severe hypoglycaemiais lower.

232 HYPOGLYCAEMIA IN PREGNANCYMicrovascular Complications of PregnancyRet<strong>in</strong>opathy is common and may progress dur<strong>in</strong>g pregnancy. Therefore ret<strong>in</strong>al screen<strong>in</strong>gshould be undertaken dur<strong>in</strong>g each trimester (Figure 10.1). In the longer term, parous womenwith type 1 diabetes have significantly lower levels of all types of ret<strong>in</strong>opathy comparedwith non-parous women. The associated significant differences <strong>in</strong> HbA 1c suggest that theimproved glycaemic control that is associated with pregnancy may be susta<strong>in</strong>ed over timewith beneficial effects on long-term complications. Thus women should be reassured thatstrict glycaemic control dur<strong>in</strong>g, and immediately after, pregnancy can effectively reducethe long-term risk of ret<strong>in</strong>opathy. In pregnancy, diabetic nephropathy of any degree is lesscommon than ret<strong>in</strong>opathy, and requires specialist management.COMPLICATIONS IN THE INFANT OF THE DIABETIC MOTHERThe potential complications for the <strong>in</strong>fant of a mother with diabetes are shown <strong>in</strong> Table 10.6.Newly born <strong>in</strong>fants with diabetic mothers are at <strong>in</strong>creased risk of develop<strong>in</strong>g hypoglycaemia.To avoid this risk, maternal blood glucose should be kept as close to normal as possibledur<strong>in</strong>g pregnancy and particularly dur<strong>in</strong>g labour and delivery. Early feed<strong>in</strong>g of the newlyborn <strong>in</strong>fant is essential and careful monitor<strong>in</strong>g is mandatory.The Risks of Maternal <strong>Hypoglycaemia</strong> to the Fetus/<strong>in</strong>fantThe potential risks of hypoglycaemia to the offspr<strong>in</strong>g can be considered <strong>in</strong> three ma<strong>in</strong> ways:the teratogenic effects of hypoglycaemia, other immediate effects on the fetus and delayedeffects.Table 10.6 Potential problems affect<strong>in</strong>g the <strong>in</strong>fant ofa mother with diabetesNeonatal and per<strong>in</strong>atal• <strong>Hypoglycaemia</strong>• Stillbirth• Increased per<strong>in</strong>atal mortality• Congenital anomalies• Preterm delivery• Macrosomia• Birth trauma at delivery• Polycythaemia• Respiratory distress syndrome

COMPLICATIONS IN THE INFANT OF THE DIABETIC MOTHER 233Data <strong>in</strong> animals suggest that hypoglycaemia <strong>in</strong> the first trimester could be teratogenic(Buchanan et al., 1986; Akazawa et al., 1987, 1989; Ell<strong>in</strong>gton, 1987; Smoak and Sadler,1990). However, it is very difficult to extrapolate these results to human pregnancy as theexposure to hypoglycaemia <strong>in</strong> these animal studies would equate to susta<strong>in</strong>ed hypoglycaemia<strong>in</strong> humans of at least five hours. There also appears to be a differential effect betweenspecies mak<strong>in</strong>g it <strong>in</strong>appropriate to extrapolate results from one species to another (Erikssonet al., 1986).In non-diabetic women <strong>in</strong> whom hypoglycaemic coma was <strong>in</strong>duced as a treatment forpsychiatric illness, the fetus was affected more frequently when hypoglycaemia was <strong>in</strong>duceddur<strong>in</strong>g the first trimester (Impastato et al., 1964). However many potential confound<strong>in</strong>gvariables were present, and the blood glucose profiles <strong>in</strong> women with type 1 diabetes arecompletely different from the non-diabetic woman.When view<strong>in</strong>g the outcomes of diabetic pregnancy <strong>in</strong> humans, the overwhelm<strong>in</strong>g evidencesupports the achievement of strict glycaemic control <strong>in</strong> early pregnancy, which is usuallyassociated with a higher risk of hypoglycaemia but greatly reduces the risks of congenitalmalformation (Kitzmiller et al., 1991; Kimmerle et al., 1992; Evers et al., 2004; Rosennet al., 1995; The <strong>Diabetes</strong> Control and Complications Trial Research Group, 1996). Thissuggests that the risk of any teratogenic effect of hypoglycaemia is greatly outweighed bythe benefits of good glycaemic control. A large follow-up study of 329 <strong>in</strong>fants born tomothers with diabetes demonstrated a high per<strong>in</strong>atal fetal mortality rate of 21%, but noconclusive evidence could implicate hypoglycaemia or any other iatrogenic factor (Farquhar,1969). In another study of women who had relatively good glycaemic control (Mills et al.,1988), all episodes of documented hypoglycaemia between gestational weeks 5 and 12 wererecorded and no association with fetal malformations was observed. Similarly, <strong>in</strong> a study byKitzmiller et al. (1991), all episodes of symptomatic hypoglycaemia were documented andno association with fetal malformation was found.In addition to the reduced risk of congenital malformations, other complications of pregnancyappear to be fewer <strong>in</strong> women who experience hypoglycaemia dur<strong>in</strong>g pregnancy. The<strong>in</strong>cidence of macrosomia has been shown to be significantly lower <strong>in</strong> women affected bysevere hypoglycaemia dur<strong>in</strong>g the third trimester (relative risk 0.66) (Evers et al., 2002b,Evers et al., 2004).With respect to the immediate effects of maternal hypoglycaemia on the fetus <strong>in</strong> utero,some data are available from studies that have <strong>in</strong>duced experimental hypoglycaemia dur<strong>in</strong>gpregnancy. Early studies suggested that fetal heart rate variability may be decreased dur<strong>in</strong>gmaternal hypoglycaemia (Stangenberg et al., 1983). However, later studies did not demonstratechanges <strong>in</strong> fetal movements, breath<strong>in</strong>g or heart rate dur<strong>in</strong>g maternal hypoglycaemia(Nisell et al., 1994, Reece et al., 1995). Björklund et al. (1996) also studied the effectsof maternal blood glucose of 2.2 mmol/l on fetal heart rate. They observed an <strong>in</strong>crease <strong>in</strong>frequency and amplitude of fetal heart rate accelerations, but no potentially harmful effectson fetal heart rate or on umbilical Doppler waveform analysis were seen. In a more recentstudy of 116 women who experienced ‘hypoglycaemia’ rather than hyperglycaemia whileundergo<strong>in</strong>g a glucose tolerance test to screen for gestational diabetes, no adverse effects onfetal growth or other per<strong>in</strong>atal outcomes were observed, although the level of blood glucosethat was designated as ‘hypoglycaemia’ (4.9 mmol/l) may well be considered too high torepresent this state (Calfee et al., 1999).The later developmental effects <strong>in</strong> the <strong>in</strong>fant that has been subjected to maternal hypoglycaemiamay not become apparent until later <strong>in</strong> childhood. However, there are many

234 HYPOGLYCAEMIA IN PREGNANCYpotential confound<strong>in</strong>g factors when study<strong>in</strong>g this aspect. For example, a woman who hasbeen exposed to hypoglycaemia may also have experienced hyperglycaemia, which couldalso have a potentially deleterious effect on cognition and it is not possible to separatethese effects. Early studies did demonstrate that the children of diabetic mothers developedcerebral dysfunction (Yss<strong>in</strong>g, 1974) and one study reported an association between the IQof offspr<strong>in</strong>g and the presence of acetonuria, but not hypoglycaemia, dur<strong>in</strong>g pregnancy <strong>in</strong> themother (Churchill et al., 1969).A Japanese study of 33 children born to mothers with diabetes demonstrated significantlylower <strong>in</strong>telligence scores than a control group when tested at three years of age (Yamashitaet al., 1996). However, no correlation was observed between IQ and maternal HbA 1c scoresdur<strong>in</strong>g pregnancy. An electrophysiological study of 60 <strong>in</strong>fants (34 controls and 26 childrenfrom mothers with diabetes) performed at six months of age, also demonstrated deficits <strong>in</strong>recognition memory (suggest<strong>in</strong>g hippocampal damage) <strong>in</strong> the offspr<strong>in</strong>g from mothers withdiabetes that was unrelated to maternal glycaemic control, although the <strong>in</strong>cidence of maternalhypoglycaemia was not measured (Nelson et al., 2000).In summary, there are no data that suggest that hypoglycaemia is teratogenic <strong>in</strong> humansor hypoglycaemia has an immediate adverse effect on the fetus. Some studies suggest thatthe offspr<strong>in</strong>g of diabetic mothers may have some differences <strong>in</strong> cerebral function, althoughthese differences have not been attributed to maternal hypoglycaemia.CONCLUSIONS• The outlook for the offspr<strong>in</strong>g of women with type 1 diabetes has improved dramaticallyover the last 30 years, and the benefit to the <strong>in</strong>fant from optimal glycaemic control dur<strong>in</strong>gpregnancy is evident.• The risk of maternal hypoglycaemia is considerable, particularly <strong>in</strong> the first and secondtrimesters, and severe hypoglycaemia is common.• Factors that may contribute to <strong>in</strong>creased risk of hypoglycaemia <strong>in</strong>clude the requirementfor strict glycaemic control, possible effects of pregnancy to cause counterregulatory deficiencyand altered symptomatic awareness, differ<strong>in</strong>g <strong>in</strong>sul<strong>in</strong> regimens and lifestyle/dietaryfactors.• Detailed education is essential to m<strong>in</strong>imise the risk of hypoglycaemia.• There is no evidence to suggest that hypoglycaemia has an adverse effect on the humanfetus or the <strong>in</strong>fant of a diabetic mother, although significant maternal morbidity may occur.• Good organisation of pre-pregnancy and pregnancy services is required to reduce the risksof maternal and <strong>in</strong>fant morbidity dur<strong>in</strong>g diabetic pregnancy.REFERENCESAkazawa M, Akazawa S, Hashimoto M, Yamaguchi Y, Kuriya N, Toyama K et al. (1987).Effects of hypoglycaemia on early embryogenesis <strong>in</strong> rat embryo organ culture. Diabetologia30: 791–6.

REFERENCES 235Akazawa M, Akazawa S, Hashimoto M, Akashi M, Yamazaki H, Tahara D et al. (1989). Effects ofbrief exposure to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemic serum dur<strong>in</strong>g organogenesis <strong>in</strong> rat embryo culture.<strong>Diabetes</strong> 38: 1573–8.American <strong>Diabetes</strong> Association (2003). Preconception care of women with diabetes. <strong>Diabetes</strong> Care26 (Suppl 1): S91–3.Björklund AO, Adamson UKC, Almström NHH, Enocksson EA, Gennser GM, L<strong>in</strong>s PE, Westgren LM(1996). Effects of hypoglycaemia on fetal heart activity and umbilical artery Doppler velocity waveforms<strong>in</strong> pregnant women with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. British Journal of Obstetricsand Gynaecology 103: 413–20.Björklund A, Adamson U, Andreasson K, Carlstrom K, Hennen G, Ogout A, L<strong>in</strong>s PE, Westgren M(1998a). Hormonal counterregulation and subjective symptoms dur<strong>in</strong>g <strong>in</strong>duced hypoglycemia <strong>in</strong><strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus patients dur<strong>in</strong>g and after pregnancy. Acta Obstetrica GynecologicaScand<strong>in</strong>avica 77: 625–34.Björklund AO, Adamson UKC, Carlström KAM, Hennen G, Igout A, L<strong>in</strong>s PE, Westgren LM (1998b).Placental hormones dur<strong>in</strong>g <strong>in</strong>duced hypoglycaemia <strong>in</strong> pregnant women with <strong>in</strong>sul<strong>in</strong>-dependentdiabetes mellitus: evidence of an active role for placenta <strong>in</strong> hormonal counter-regulation. BritishJournal of Obstetrics and Gynaecology 105: 649–55.Buchanan TA, Schemmer JK, Fre<strong>in</strong>kel N (1986). Embryotoxic effects of brief maternal <strong>in</strong>sul<strong>in</strong>hypoglycemiadur<strong>in</strong>g organogenesis <strong>in</strong> the rat. Journal of Cl<strong>in</strong>ical Investigation 78: 643–9.Calfee EF, Rust OA, Bofill JA, Ross EL, Morrison JC (1999). Maternal hypoglycemia: is it associatedwith adverse per<strong>in</strong>atal outcome? Journal of Per<strong>in</strong>atology 19: 379–82.Casson IF, Clarke CA, Howard CV, McKendrick O, Pennycook S, Pharoah PO et al. (1997). Outcomesof pregnancy <strong>in</strong> <strong>in</strong>sul<strong>in</strong> dependent diabetic women: results of a five year population cohort study.British Medical Journal 315: 275–8.Churchill JA, Berendes HW, Nemore J (1969). Neuropsychological deficits <strong>in</strong> children of diabeticmothers. American Journal of Obstetrics and Gynecology 105: 257–68.Confidential Enquiry <strong>in</strong>to Maternal and Child Health (2005). Pregnancy <strong>in</strong> women with type 1 andtype 2 diabetes <strong>in</strong> 2002–2003. England, Wales and Northern Ireland. CEMACH, London.Diamond MP, Reece EA, Caprio S, Jones TW, Amiel SA, DeGennaro N, Laudano A et al. (1992).Impairment of counterregulatory hormone responses to hypoglycemia <strong>in</strong> pregnant women with<strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. American Journal of Obstetrics and Gynecology 166: 70–7.Ell<strong>in</strong>gton SK (1987). Development of rat embryos cultured <strong>in</strong> glucose-deficient media. <strong>Diabetes</strong> 36:1372–8.Eriksson UT, Dahlström VE, Lithell HO (1986). <strong>Diabetes</strong> <strong>in</strong> pregnancy: <strong>in</strong>fluence of background andmaternal diabetic state on the <strong>in</strong>cidence of skeletal malformations <strong>in</strong> the fetal rat. Acta Endocr<strong>in</strong>ologica(Copenhagen) 112 (Suppl 277): 66–73.Evers IM, ter Braak EWMT, de Valk HW, der Schout BV, Janssen N, Visser GA (2002a). Risk<strong>in</strong>dicators predictive of severe hypoglycemia dur<strong>in</strong>g the first trimester of type 1 diabetic pregnancy.<strong>Diabetes</strong> Care 25: 554–9.Evers IM, de Walk HW, Mol BWJ, ter Braak EWMT, Visser GHA (2002b). Macrosomia despite goodglycaemic control <strong>in</strong> type 1 diabetic pregnancy; results of a nationwide study <strong>in</strong> The Netherlands.Diabetologia 45: 1484–9.Evers IM, de Valk HW, Visser GHA (2004). Risk of complications of pregnancy <strong>in</strong> women with type1 diabetes: nationwide prospective study <strong>in</strong> the Netherlands. British Medical Journal 328: 915–8.Farquhar JW (1969). Prognosis for babies born to diabetic mothers <strong>in</strong> Ed<strong>in</strong>burgh. Archives of Disease<strong>in</strong> Childhood 44: 36–47.Fuhrmann K, Reiker H, Semmler K, Glockner E (1983). Prevention of congenital malformation <strong>in</strong><strong>in</strong>fants of <strong>in</strong>sul<strong>in</strong> dependent diabetic mothers. <strong>Diabetes</strong> Care 6: 219–23.Garg SK, Frias JP, Anil S, Gottlieb PA, MacKenzie T, Jackson WE. (2003). Insul<strong>in</strong> lispro therapy<strong>in</strong> pregnancies complicated by type 1 diabetes: glycemic control and maternal and fetal outcomes.Endocr<strong>in</strong>e Practice 9: 187-93.

236 HYPOGLYCAEMIA IN PREGNANCYImpastato DJ, Gabriel AR, Lardaro EH (1964). Electric and <strong>in</strong>sul<strong>in</strong> shock therapy dur<strong>in</strong>g pregnancy.Diseases of the Nervous System 25: 542–6.Jensen DM, Damm P, Moelsted-Pedersen L, Ovesen P, Westergaard JG, Moeller M, Beck-Nielsen H(2004). Outcomes <strong>in</strong> type 1 diabetic pregnancies: a nationwide, population-based study. <strong>Diabetes</strong>Care, 27: 2819–23.Jovanovic L, Ilic S, Pettitt DJ, Hugo K, Gutierrez M, Bowsher RR, Bastyr EJ (1999).Metabolic and immunologic effects of <strong>in</strong>sul<strong>in</strong> lispro <strong>in</strong> gestational diabetes. <strong>Diabetes</strong> Care 22:1422–7.Jovanovic L, Knopp RH, Brown, Conleg MR, Park E, Mills JL et al. (2001). Decl<strong>in</strong><strong>in</strong>g <strong>in</strong>sul<strong>in</strong>requirement <strong>in</strong> the late first trimester of diabetic pregnancy. <strong>Diabetes</strong> Care 24: 1130–6.Kimmerle R, He<strong>in</strong>emann L, Delecki A, Berger M (1992). Severe hypoglycemia <strong>in</strong>cidence and predispos<strong>in</strong>gfactors <strong>in</strong> 85 pregnancies of type 1 diabetic women. <strong>Diabetes</strong> Care 15: 1034–7.Kitzmiller JL, Gav<strong>in</strong> LA, G<strong>in</strong> GD, Jovanovic-Petersen L, Ma<strong>in</strong> EK, Z<strong>in</strong>gran WD (1991). Preconceptioncare of diabetes. Glycemic control prevents congenital abnormalities. Journal of the AmericanMedical Association 265: 731–6.Lapolla A, Dalfrà MG, Mas<strong>in</strong> M, Bruttomesso D, Piva I, Crepaldi C et al. (2003). Analysis of outcomeof pregnancy <strong>in</strong> type 1 diabetics treated with <strong>in</strong>sul<strong>in</strong> pump or conventional <strong>in</strong>sul<strong>in</strong> therapy. ActaDiabetologica 40: 143–9.Loukovaara S, Immonen I, Teramo KA, Kaaja R (2003). Progression of ret<strong>in</strong>opathy dur<strong>in</strong>g pregnancy<strong>in</strong> type 1 diabetic women treated with <strong>in</strong>sul<strong>in</strong> lispro. <strong>Diabetes</strong> Care 26: 1193–8.Masson EA, Patmore JE, Brash PD, Baxter M, Caldwell G, Gallen IW et al. (2003). Pregnancyoutcome <strong>in</strong> type 1 diabetes mellitus treated with <strong>in</strong>sul<strong>in</strong> lispro (Humalog). Diabetic Medic<strong>in</strong>e 20:46–50.Mathieson ER, K<strong>in</strong>sley B, Amiel SA, Heller S, McCance D, Duran S, Bellaire S, Raben A on behalfof the Insul<strong>in</strong> Aspart Pregnancy Study Group (2007). Maternal glycemic control and hypoglycaemia<strong>in</strong> Type 1 diabetic pregnancy. <strong>Diabetes</strong> Care 30: 771–6.Mills JL, Knopp RH, Simpson JL, Jovanovic-Peterson L, Metzger BE, Holmes, LB et al. (1988). Lackof relation of <strong>in</strong>creased malformation rates <strong>in</strong> <strong>in</strong>fant of diabetic mothers to glycemic control dur<strong>in</strong>gorganogenesis. New England Journal of Medic<strong>in</strong>e 318: 671–6.Nachum Z, Ben-Shlomo I, We<strong>in</strong>er E, Shalev E (1999). Twice versus four times daily <strong>in</strong>sul<strong>in</strong> doseregimens for diabetes <strong>in</strong> pregnancy: randomised controlled trial. British Medical Journal 319:1223–7.Nelson CA, Wewerka S, Thomas KM, Tribby-Walbridge S, deRegnier R-A, Georgieff M (2000).Neurocognitive sequelae of <strong>in</strong>fants of diabetic mothers. Behavioral Neuroscience 114: 950–6.Nisell H, Persson B, Hanson U, Lunell NO, Nylund L, Sarby B, Thronstrom S (1994). Hormonal,metabolic, and circulatory responses to <strong>in</strong>sul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> pregnant and nonpregnantwomen with <strong>in</strong>sul<strong>in</strong>-dependent diabetes. American Journal of Per<strong>in</strong>atology 11: 231–6.Paretti E, Mecacci I, Pap<strong>in</strong>i M, Cloni R, Carignani L, Mignosa M et al. (2001). Third trimester maternalglucose levels from diurnal profiles <strong>in</strong> non-diabetic pregnancies. Correlation with sonographicperameters of fetal growth. <strong>Diabetes</strong> Care 24: 1319–23.Penney GC, Mair G, Pearson DW (2003a). Outcomes of pregnancies <strong>in</strong> women with type 1 diabetes<strong>in</strong> Scotland: a national population-based study. British Journal of Obstetrics and Gynaecology 110:315–18.Penney GC, Mair G, Pearson DW (2003b). The relationship between birth weight and maternalglycated haemoglob<strong>in</strong> (HbA 1c ) concentration <strong>in</strong> pregnancies complicated by type 1 diabetes. DiabeticMedic<strong>in</strong>e 20: 162–6.Persson B, Hansson U (1993). <strong>Hypoglycaemia</strong> <strong>in</strong> pregnancy. Bailliere’s Cl<strong>in</strong>ical Endocr<strong>in</strong>ology andMetabolism 7: 731–9.Persson B, Swahn M-L, Hjertberg R, Hanson U, Nord E, Nordlander E, Hansson L-O (2002). Insul<strong>in</strong>lispro therapy <strong>in</strong> pregnancies complicated by type 1 diabetes mellitus. <strong>Diabetes</strong> Research andCl<strong>in</strong>ical Practice 58: 115–21.

REFERENCES 237Ray JG, O’Brien TE, Chan WS (2001). Preconception care and the risk of congenital anomalies <strong>in</strong>the offspr<strong>in</strong>g of women with diabetes mellitus: a meta-analysis. Quarterly Journal of Medic<strong>in</strong>e 94:435–44.Rayburn W, Piehl E, Jacober S, Schork A, Ploughman L (1986). Severe hypoglycemia dur<strong>in</strong>g pregnancy:its frequency and predispos<strong>in</strong>g factors <strong>in</strong> diabetic women. International Journal of Obstetricsand Gynecology 24: 263–8.Reece EA, Hagay Z, Roberts AB, DeGennaro N, Homko CJ, Connolly-Diamond M, Sherw<strong>in</strong> R et al.(1995). Fetal Doppler and behavioral responses dur<strong>in</strong>g hypoglycemia <strong>in</strong>duced with the <strong>in</strong>sul<strong>in</strong> clamptechnique <strong>in</strong> pregnant diabetic women. American Journal of Obstetrics and Gynecology 172: 151–5.Rosenn BM, Miodovnik M, Holcberg G, Khoury JC, Siddiqi TA (1995). Hypoglycemia: the price of<strong>in</strong>tensive <strong>in</strong>sul<strong>in</strong> therapy for pregnant women with <strong>in</strong>sul<strong>in</strong>-dependent diabetes mellitus. Obstetricsand Gynecology 85: 417–22.Rosenn BM, Miodovnik M, Khoury JC, Siddiqi TA (1996). Counterregulatory hormonal responses tohypoglycemia dur<strong>in</strong>g pregnancy. Obstetrics and Gynecology 87: 568–74.Rossi G, Lapaczewski P, Diamond MP, Jacob RJ, Shulman GI, Sherw<strong>in</strong> RS (1993). Inhibitory effectof pregnancy on counterregulatory hormone responses to hypoglycemia <strong>in</strong> awake rat. <strong>Diabetes</strong> 42:1440–5.Scottish <strong>Diabetes</strong> <strong>in</strong> Pregnancy Study Group (2004). Audit of pregnancies <strong>in</strong> women with diabetes.F<strong>in</strong>d<strong>in</strong>gs from a one year audit. Report to Scottish Programme for Cl<strong>in</strong>ical Effectiveness In ReproductiveHealth (SPCERH).SIGN Executive (2001). SIGN 55 Management of <strong>Diabetes</strong>. Scottish Intercollegiate Guidel<strong>in</strong>esNetwork, Ed<strong>in</strong>burgh: 34–8.Smoak IW, Sadler TW (1990). Embryopathic effects of short-term exposure to hypoglycemia <strong>in</strong> mouseembryos <strong>in</strong> vitro. American Journal of Obstetrics and Gynecology 63: 619–24.Stangenberg M, Persson B, Stange L, Carlstrom K (1983). Insul<strong>in</strong>-<strong>in</strong>duced hypoglycemia <strong>in</strong> pregnantdiabetics. Maternal and fetal cardiovascular reactions. Acta Obstetrica Gynecologica Scand<strong>in</strong>avica62: 249–52.Steel JM, Johnstone FD, Smith AF (1989). Pre-pregnancy preparation. In: Carbohydrate Metabolism<strong>in</strong> Pregnancy and the Newborn: IV. Sutherland HW, Stowers JM and Pearson DWM, eds. Spr<strong>in</strong>ger-Verlag, Berl<strong>in</strong>: 129–39.Steel JM, Johnstone FD, Hepburn D, Smith AF (1990). Can pre-pregnancy care of diabetic womenreduce the risk of abnormal babies? British Medical Journal 301: 1070–3.The <strong>Diabetes</strong> Control and Complications Trial Research Group (1996). Pregnancy outcomes <strong>in</strong> the<strong>Diabetes</strong> Control and Complications Trial. American Journal of Obstetrics and Gynecology 174:1343–53.Wyatt JW, Frias JL, Hoymet HE, Jovanovic L, Kaaja R, Brown F et al., the IONS study group (2004).Congenital anomaly rate <strong>in</strong> offspr<strong>in</strong>g of mothers with diabetes treated with <strong>in</strong>sul<strong>in</strong> lispro dur<strong>in</strong>gpregnancy. Diabetic Medic<strong>in</strong>e 22: 803–7.Yamashita Y, Kawano Y, Kuriya N, Murakami Y, Matsuishi T, Yoshimatsu K, Kato H (1996).Intellectual development of offspr<strong>in</strong>g of diabetic mothers. Acta Paediatrica 85: 1192–6.Yss<strong>in</strong>g M (1974). Oestriol excretion <strong>in</strong> pregnant diabetes related to long term prog