11.01.2013 Views

Wholegrain cereals for coronary heart disease (Review) - Teesside's ...

Wholegrain cereals for coronary heart disease (Review) - Teesside's ...

Wholegrain cereals for coronary heart disease (Review) - Teesside's ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

This full text version, available on TeesRep, is the PDF (final version) of:<br />

Kelly, S. A. M. et al. (2007) '<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong>', The<br />

Cochrane Database of Systematic <strong>Review</strong>s, 2 (Art. no. CD005051).<br />

For details regarding the final published version please click on the following DOI link:<br />

http://dx.doi.org/10.1002/14651858.CD005051.pub2<br />

When citing this source, please use the final published version as above.<br />

This document was downloaded from http://tees.openrepository.com/tees/handle/10149/93496<br />

Please do not use this version <strong>for</strong> citation purposes.<br />

All items in TeesRep are protected by copyright, with all rights reserved, unless otherwise indicated.<br />

TeesRep: Teesside University's Research Repository http://tees.openrepository.com/tees/


<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Kelly SAM, Summerbell CD, Brynes A, Whittaker V, Frost G<br />

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library<br />

2009, Issue 1<br />

http://www.thecochranelibrary.com<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.


T A B L E O F C O N T E N T S<br />

HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1<br />

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1<br />

PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2<br />

BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2<br />

OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3<br />

METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3<br />

Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11<br />

RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13<br />

DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24<br />

AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25<br />

ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25<br />

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25<br />

CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30<br />

DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48<br />

Analysis 1.1. Comparison 1 Total cholesterol, Outcome 1 4 week outcomes. . . . . . . . . . . . . . . 50<br />

Analysis 1.2. Comparison 1 Total cholesterol, Outcome 2 All endpoint outcomes. . . . . . . . . . . . . 51<br />

Analysis 2.1. Comparison 2 LDL cholesterol, Outcome 1 4 week outcomes. . . . . . . . . . . . . . . 52<br />

Analysis 2.2. Comparison 2 LDL cholesterol, Outcome 2 All endpoint outcomes. . . . . . . . . . . . . 53<br />

Analysis 3.1. Comparison 3 HDL cholesterol, Outcome 1 4 week outcomes. . . . . . . . . . . . . . . 54<br />

Analysis 3.2. Comparison 3 HDL cholesterol, Outcome 2 All endpoint outcomes. . . . . . . . . . . . . 55<br />

Analysis 4.1. Comparison 4 Triglycerides, Outcome 1 4 week outcomes. . . . . . . . . . . . . . . . . 56<br />

Analysis 4.2. Comparison 4 Triglycerides, Outcome 2 All endpoint outcomes (Keenan data as SD). . . . . . . 57<br />

Analysis 4.3. Comparison 4 Triglycerides, Outcome 3 All endpoint outcomes (Keenan data as SEM). . . . . . . 58<br />

Analysis 5.1. Comparison 5 Body weight (kg), Outcome 1 All endpoint outcomes. . . . . . . . . . . . . 59<br />

WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59<br />

HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59<br />

CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60<br />

DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60<br />

SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60<br />

INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

i


[Intervention <strong>Review</strong>]<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Sarah AM Kelly 1 , Carolyn D Summerbell 2 , Audrey Brynes 3 , Victoria Whittaker 4 , Gary Frost 5<br />

1 School of Dental Sciences, University of Newcastle, Newcastle upon Tyne, UK. 2 School of Medicine and Health, Wolfson Research<br />

Institute, Durham University, Stockton-on-Tees, UK. 3 Department of Dietetics, Hammersmith Hospital, London, UK. 4 School of<br />

Health and Social Care, University of Teesside, Middlesbrough, UK. 5 Department of Nutrition and Dietetics, Hammersmith Hospital,<br />

London, UK<br />

Contact address: Sarah AM Kelly, School of Dental Sciences, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK.<br />

s.a.m.kelly@newcastle.ac.uk.<br />

Editorial group: Cochrane Heart Group.<br />

Publication status and date: Edited (no change to conclusions), published in Issue 1, 2009.<br />

<strong>Review</strong> content assessed as up-to-date: 14 January 2007.<br />

Citation: Kelly SAM, Summerbell CD, Brynes A, Whittaker V, Frost G. <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong>. Cochrane<br />

Database of Systematic <strong>Review</strong>s 2007, Issue 2. Art. No.: CD005051. DOI: 10.1002/14651858.CD005051.pub2.<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Background<br />

A B S T R A C T<br />

There is increasing evidence from observational studies that wholegrains can have a beneficial effect on risk factors <strong>for</strong> <strong>coronary</strong> <strong>heart</strong><br />

<strong>disease</strong> (CHD).<br />

Objectives<br />

The primary objective is to review the current evidence from randomised controlled trials (RCTs) that assess the relationship between<br />

the consumption of wholegrain foods and the effects on CHD mortality, morbidity and on risk factors <strong>for</strong> CHD, in participants<br />

previously diagnosed with CHD or with existing risk factors <strong>for</strong> CHD.<br />

Search strategy<br />

We searched CENTRAL (Issue 4, 2005), MEDLINE (1966 to 2005), EMBASE (1980 to 2005), CINAHL (1982 to 2005), ProQuest<br />

Digital Dissertations (2004 to 2005). No language restrictions were applied.<br />

Selection criteria<br />

We selected randomised controlled trials that assessed the effects of wholegrain foods or diets containing wholegrains, over a minimum<br />

of 4 weeks, on CHD and risk factors. Participants included were adults with existing CHD or who had at least one risk factor <strong>for</strong><br />

CHD, such as abnormal lipids, raised blood pressure or being overweight.<br />

Data collection and analysis<br />

Two of our research team independently assessed trial quality and extracted data. Authors of the included studies were contacted <strong>for</strong><br />

additional in<strong>for</strong>mation where this was appropriate.<br />

Main results<br />

Ten trials met the inclusion criteria. None of the studies found reported the effect of wholegrain diets on CHD mortality or CHD<br />

events or morbidity. All 10 included studies reported the effect of wholegrain foods or diets on risk factors <strong>for</strong> CHD. Studies ranged in<br />

duration from 4 to 8 weeks. In eight of the included studies, the wholegrain component was oats. Seven of the eight studies reported<br />

lower total and low density lipoproteins (LDL) cholesterol with oatmeal foods than control foods. When the studies were combined in<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

1


a meta-analysis lower total cholesterol (-0.20 mmol/L, 95% confidence interval (CI) -0.31 to -0.10, P = 0.0001 ) and LDL cholesterol<br />

(0.18 mmol/L, 95% CI -0.28 to -0.09, P < 0.0001) were found with oatmeal foods. However, there is a lack of studies on other<br />

wholegrains or wholegrain diets.<br />

Authors’ conclusions<br />

Despite the consistency of effects seen in trials of wholegrain oats, the positive findings should be interpreted cautiously. Many of<br />

the trials identified were short term, of poor quality and had insufficient power. Most of the trials were funded by companies with<br />

commercial interests in wholegrains. There is a need <strong>for</strong> well-designed, adequately powered, longer term randomised controlled studies<br />

in this area. In particular there is a need <strong>for</strong> randomised controlled trials on wholegrain foods and diets other than oats.<br />

P L A I N L A N G U A G E S U M M A R Y<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

<strong>Wholegrain</strong> foods encompass a range of products and examples are wholegrain wheat, rice, maize and oats. The term wholegrain also<br />

includes milled wholegrains such as oatmeal and wholemeal wheat. The evidence found by this review is limited to wholegrain oats,<br />

and to changes in lipids as an outcome. There is a lack of studies on other wholegrain foods or diets. There is some evidence from this<br />

review that oatmeal foods can beneficially lower lipid levels such as low density lipoproteins (LDL) cholesterol and total cholesterol in<br />

those previously diagnosed with risk factors <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (CHD) even with relatively short interventions. However, the<br />

results should be interpreted with caution because the trials found are small, of short duration and many were commercially funded.<br />

No studies were found that reported the effect of wholegrain foods or diets on deaths from, or occurrence of CHD.<br />

B A C K G R O U N D<br />

In western society, <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (CHD) is the major<br />

cause of death. The prevalence of CHD is increasing worldwide<br />

(Murray 1997). In the UK, there were 238 deaths from CHD<br />

per 100,000 population in 1997 (OHE 1999). In England and<br />

Wales in 1996 one fifth of men and one eighth of women aged<br />

over 65 were treated <strong>for</strong> CHD in general practice (Carter 1999).<br />

CHD costs the health care system in the UK about £1750 million<br />

a year (BHF 2004). In addition, CHD costs the UK economy<br />

about £5300 million because of days lost due to death, illness and<br />

in<strong>for</strong>mal care of people with the <strong>disease</strong> (BHF 2004).<br />

A wholegrain contains the entire edible parts of a natural grain<br />

kernel. The structure of all whole grains is similar and includes<br />

the endosperm, germ and bran. <strong>Wholegrain</strong>s are rich in dietary<br />

fibre, anti-oxidants, resistant starch, phyto-oestrogens and other<br />

important micronutrients such as vitamins and folic acid (Slavin<br />

2003). In the grain-refining process, most of the bran and some<br />

of the germ is removed, resulting in the loss of dietary fibre, vitamins,<br />

minerals, lignans, phyto-oestrogens, phenolic compounds<br />

and phytic acid. The remaining starchy endosperm is ground to<br />

produce refined white flours.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Important grains in the western diet include wheat, rice, maize<br />

and oats. Wholemeal foods are made from wholegrains which have<br />

been milled to a finer texture rather than leaving them whole in<br />

the final product. Both wholegrain and wholemeal cereal foods<br />

are those grain foods which include the outer layers of the grain,<br />

including the bran and germ. Food Standards Australia and New<br />

Zealand (FSANZ 2004) have proposed a definition of the term<br />

wholegrain as ’ wholegrain means the intact grain, or the dehulled,<br />

ground, milled, cracked or flaked grain where the constituents -<br />

endosperm, germ and bran are present in such proportions that<br />

represent the typical ratio of those fractions occurring in the whole<br />

cereal, and includes wholemeal’ (FSANZ 2004). The proposal also<br />

relates to foods made from whole grains such as breads, breakfast<br />

<strong>cereals</strong>, pasta, biscuits, oats, rice and grain-based snack foods. Research<br />

has shown that such processing of whole-grains does not<br />

remove biologically important compounds (Slavin 2001). Nutritionally<br />

wholegrain and wholemeal foods are similar.<br />

The US Food and Drug Administration (FDA 1999) does not<br />

define wholegrain however since 1999 has permitted food manufacturers<br />

to make the following health claim on wholegrain food<br />

2


products ’diets rich in wholegrain foods and other plant foods and<br />

low in total fat, saturated fat and cholesterol may reduce the risk<br />

of <strong>heart</strong> <strong>disease</strong> and some cancers’ (FDA 1999). For the purpose<br />

of this health claim, wholegrain foods must contain 51% or more<br />

of wholegrain ingredients (bran, germ and endosperm) by weight<br />

per reference amount.<br />

A few epidemiological studies have specifically examined the association<br />

between intakes of wholegrain foods and risk of CHD.<br />

The Iowa Women’s Health study (Jacobs 1998; Jacobs 1999) of<br />

34,492 post menopausal women followed <strong>for</strong> 6 years found that<br />

a greater intake of wholegrain was associated with a reduced risk<br />

of death from CHD. The 10-year Nurses’ Health Study, a large<br />

prospective study of 75,521 women aged 38-63, found that increased<br />

wholegrain intake was associated with decreased risk of<br />

CHD (Liu 1999). The lower risk associated with higher wholegrain<br />

intake was not fully explained by its contribution to intakes<br />

of dietary fibre, folate, vitamin B6 and vitamin E. More recent<br />

trials have not demonstrated beneficial effects of vitamin E or folate/B6<br />

on CHD outcomes. The Atherosclerosis Risk in Communities<br />

(ARIC) study found a beneficial relationship between<br />

wholegrain consumption and the risk of total mortality and the<br />

incidence of <strong>coronary</strong> artery <strong>disease</strong> but not with the risk of ischaemic<br />

stroke (Steffen 2003). The study followed 15,792 people<br />

aged 45-64 <strong>for</strong> 11 years. A recent review of wholegrains and CVD<br />

risk (Seal 2006) concluded that there is an increasing body of evidence,<br />

including that from prospective population studies and<br />

epidemiological observational studies suggesting a strong inverse<br />

relationship between increased consumption of wholegrain foods<br />

and CVD risk.<br />

Other studies have focused on the relationship between cereal fibre<br />

and CHD. The amount of cereal fibre is much greater in wholegrain<br />

foods than in refined grain products so the cereal fibre intake<br />

is likely to mirror the wholegrain intake. In a large prospective<br />

study of 43,757 male health professionals, cereal fibre was associated<br />

with reduced risk <strong>for</strong> myocardial infarction, with a 29%<br />

decrease in risk <strong>for</strong> each 10 g increase in cereal fibre (Rimm 1996).<br />

A significant inverse association between total dietary fibre intake<br />

and risk of CHD was found in another large study of 68,782<br />

women who were followed <strong>for</strong> 10 years (Wolk 1999). There was a<br />

significant inverse relation with cereal fibre but not with vegetable<br />

or fruit fibre. Women in the highest quintile of cereal fibre had a<br />

34% lower risk of CHD events compared with those in the lowest<br />

quintile. However, the DART trial (Burr 1989) did not show any<br />

benefit of dietary fibre.<br />

Associations between wholegrain consumption and risk factors<br />

<strong>for</strong> CHD have also been reported. In the Framingham Offspring<br />

study, diets rich in wholegrains were inversely associated with total<br />

cholesterol, LDL cholesterol and body mass index (McKeown<br />

2002). A prospective study of women aged 38 to 63 years found<br />

that substituting wholegrain products <strong>for</strong> refined grain products<br />

may decrease the risk of Type 2 diabetes mellitus (Liu 2000). Di-<br />

abetes is a major risk factor <strong>for</strong> CHD. Reduced insulin sensitivity<br />

is another factor that contributes to the metabolic environment<br />

that predisposes to CHD (Reaven 1993). There is evidence of<br />

reduced insulin sensitivity at diagnosis in 60% of patients with<br />

CHD (Feranninni 1991). <strong>Wholegrain</strong> consumption improves insulin<br />

sensitivity in overweight and obese adults (Pereira 2002).<br />

O B J E C T I V E S<br />

The aim of this systematic review was to assess the effect of wholegrain<br />

foods or diets on total mortality from <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

(CHD), CHD events and morbidity or changes in major risk<br />

factors <strong>for</strong> CHD, in people with existing CHD or risk factors<br />

<strong>for</strong> CHD. All available randomised controlled trials and concurrent<br />

controlled trials and meta-analytic techniques were included<br />

where appropriate.<br />

M E T H O D S<br />

Criteria <strong>for</strong> considering studies <strong>for</strong> this review<br />

Types of studies<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Randomised controlled studies (RCTs). Cross-over studies and<br />

parallel studies were included. Trials were only included if outcome<br />

data could be collected (by communication with authors where<br />

necessary).<br />

Types of participants<br />

Non-institutionalised adults of either sex with diagnosed CHD,<br />

or with at least one major risk factor <strong>for</strong> CHD, were included.<br />

Major risk factors <strong>for</strong> CHD include overweight (body mass index<br />

>25), abdominal obesity, diabetes, family history of CHD, raised<br />

blood pressure, hypertension, abnormal lipid levels (high density<br />

lipoproteins (HDL) and low density lipoproteins (LDL) cholesterol,<br />

triglycerides or total cholesterol), impaired glucose tolerance,<br />

reduced insulin sensitivity, insulin resistance, hyperinsulinaemia,<br />

hyperglycaemia or abnormal clotting factors. Participants with<br />

myocardial infarction, <strong>coronary</strong> artery bypass graft, percutaneous<br />

transluminal <strong>coronary</strong> angioplasty, or who had angina pectoris or<br />

<strong>coronary</strong> artery <strong>disease</strong> defined by angiography were included.<br />

Participants without previously diagnosed CHD or risk factors <strong>for</strong><br />

CHD were excluded. Children (age


Participants on medication were included with the exception of<br />

those on weight loss medication who were excluded.<br />

Types of interventions<br />

Studies were included if they compared the effect of individual<br />

wholegrain foods, or diets high in wholegrain foods, with other<br />

diets or foods with lower levels or no wholegrains. Comparisons<br />

were between diets with similar overall carbohydrate, fat, protein<br />

and energy levels. For the purpose of this review the term wholegrain<br />

includes foods based on milled wholegrains, such as wholemeal<br />

or oatmeal, where the components of the endosperm, bran<br />

and germ have not been removed. Studies had to have a minimum<br />

of four weeks intervention period (or follow-up period following<br />

dietary advice).<br />

Studies were not included if they were multiple component interventions,<br />

or interventions which incorporated factors other than<br />

wholegrain foods or diets, unless the effect of wholegrain foods<br />

or diets could be separated out from the other factors. Studies on<br />

foods which were based only on individual components (e.g. bran,<br />

germ or other components) of the grain were not included e.g.<br />

oat bran or wheatgerm. Studies that examined the effect of high<br />

fibre, dietary fibre or cereal fibre, but where the specific effect of<br />

wholegrain foods or diets could not be distinguished, were not<br />

included.<br />

Types of outcome measures<br />

Primary outcomes<br />

(1) Total CHD mortality.<br />

(2) Combined CHD events and morbidity (including fatal and<br />

non fatal myocardial infarction, angina, unplanned <strong>coronary</strong><br />

artery bypass graft or percutaneous transluminal <strong>coronary</strong> angioplasty).<br />

(3) Changes in major risk factors <strong>for</strong> CHD including overweight,<br />

lipids (HDL and LDL cholesterol levels, triglycerides and total<br />

cholesterol), blood pressure, measures of diabetic control including<br />

changes in medication, glycosylated haemoglobin, glucose tolerance<br />

and control), insulin resistance, insulin sensitivity, clotting<br />

factors, hyperinsulinaemia, hyperglycaemia.<br />

Secondary outcomes<br />

(1) Measures of quality of life and attitudes to diets.<br />

Where reported, harms such as bloating, nausea, weight gain, difficulty<br />

in eating out, difficulty in preparing meals, excessive weight<br />

loss, intention to continue diet, were noted.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Search methods <strong>for</strong> identification of studies<br />

(1) Electronic databases searched included the Cochrane Central<br />

Register of Controlled Trials (CENTRAL) on The Cochrane Library<br />

(Issue 3, 2005), MEDLINE (1966 to December Week 1<br />

2005), EMBASE (1980 to December Week 1 2005), CINAHL<br />

(1982 to December Week 1 2005) and ProQuest Digital Dissertations<br />

(2005). The searches were limited to studies in humans.<br />

There were no language restrictions <strong>for</strong> either searching or trial<br />

inclusion.<br />

(2) The reference lists of all included studies were searched in order<br />

to find other potentially eligible studies.<br />

(3) Known experts in the field were consulted to ensure completeness.<br />

(4) Published reviews in the area of wholegrains and CHD were<br />

also sought as a source of RCTs.<br />

The following search strategy was developed <strong>for</strong> MEDLINE, and<br />

used with a highly sensitive filter <strong>for</strong> identifying RCTs (Dickersin<br />

1994):<br />

1. wholegrain$.ab,ti.<br />

2. wholemeal$.ab,ti.<br />

3. wholewheat$.ab,ti.<br />

4. (whole adj3 grain$).ab,ti.<br />

5. (whole adj3 meal).ab,ti.<br />

6. (whole adj3 wheat).ab,ti.<br />

7. (whole adj3 food$).ab,ti.<br />

8. (wheat adj3 meal).ab,ti.<br />

9. cereal$.ab,ti.<br />

10. bread$.ab,ti.<br />

11. wheat$.ab,ti.<br />

12. oat$.ab,ti.<br />

13. rye$.ab,ti.<br />

14. barley$.ab,ti.<br />

15. maize$.ab,ti.<br />

16. corn.ab,ti.<br />

17. cornmeal.ab,ti.<br />

18. popcorn.ab,ti.<br />

19. sorghum$.ab,ti.<br />

20. (bulgar or bulghar).ab,ti.<br />

21. couscous$.ab,ti.<br />

22. grain$.ab,ti.<br />

23. porridge.ab,ti.<br />

24. rice$.ab,ti.<br />

25. millet$.ab,ti.<br />

26. exp CEREALS/<br />

27. exp BREAD/<br />

28. exp Dietary Fiber/<br />

29. exp Coronary Disease/<br />

30. exp Cardiovascular Diseases/<br />

31. <strong>heart</strong> <strong>disease</strong>$.tw.<br />

32. <strong>coronary</strong> <strong>disease</strong>$.tw.<br />

33. chd.tw.<br />

34. cardiovascular.tw.<br />

4


35. angina.tw.<br />

36. cvd.tw.<br />

37. exp CHOLESTEROL/<br />

38. exp Blood Pressure/<br />

39. exp Obesity/<br />

40. exp Insulin Resistance/<br />

41. exp Diabetes Mellitus/<br />

42. exp LIPIDS/<br />

43. insulin resistance.ab,ti.<br />

44. insulin sensitivity.ab,ti.<br />

45. (glyc?emic adj3 control).ab,ti.<br />

46. or/1-28<br />

47. or/29-45<br />

48. 46 and 47<br />

The above search was adapted as necessary <strong>for</strong> the searches of<br />

EMBASE (Table 1), CINAHL (Table 2) The CENTRAL (Table<br />

3), and ProQuest Digital Dissertations (Table 4).<br />

Table 1. EMBASE search strategy<br />

Search terms<br />

1. wholegrain$.ab,ti.<br />

2. wholemeal$.ab,ti.<br />

3. wholewheat$.ab,ti.<br />

4. (whole adj3 grain$).ab,ti.<br />

5. (whole adj3 meal$).ab,ti.<br />

6. (whole adj3 wheat).ab,ti.<br />

7. (whole adj3 food$).ab,ti.<br />

8. (wheat adj3 meal).ab,ti.<br />

9. cereal$.ab,ti.<br />

10. bread.ab,ti.<br />

11. breads.ab,ti.<br />

12. wheat$.ab,ti.<br />

13. oat$.ab,ti.<br />

14. rye$.ab,ti.<br />

15. barley$.ab,ti.<br />

16. maize.ab,ti.<br />

17. corn.ab,ti.<br />

18. cornmeal.ab,ti.<br />

19. popcorn.ab,ti.<br />

20. sorghum$.ab,ti.<br />

21. (bulgar or bulghar).ab,ti.<br />

22. couscous.ab,ti.<br />

23. grain.ab,ti.<br />

24. grains.ab,ti.<br />

25. porridge.ab,ti.<br />

26. exp Cereal/<br />

27. exp BREAD/<br />

28. exp Dietary Fiber/<br />

29. exp Coronary Artery Disease/<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

5


Table 1. EMBASE search strategy (Continued)<br />

30. exp Cardiovascular Disease/<br />

31. <strong>heart</strong> <strong>disease</strong>$.tw.<br />

32. <strong>coronary</strong> <strong>disease</strong>$.tw.<br />

33. chd.tw.<br />

34. cardiovascular.tw.<br />

35. angina.tw.<br />

36. cvd.tw.<br />

37. exp CHOLESTEROL/<br />

38. exp Blood Pressure/<br />

39. exp OBESITY/<br />

40. exp Insulin Resistance/<br />

41. (glyc?emic adj3 control).ab,ti.<br />

42. exp Diabetes Mellitus/<br />

43. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24<br />

or 25 or 26 or 27 or 28<br />

44. 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42<br />

45. 43 and 44<br />

46. Controlled Study/<br />

47. Clinical Trial/<br />

48. random$.tw.<br />

49. compar$.ab,ti.<br />

50. control$.ab,ti.<br />

51. study.ab,ti.<br />

52. follow$ up.ab,ti.<br />

53. clinic$.ab,ti.<br />

54. blind$.ab,ti.<br />

55. Double Blind Procedure/<br />

56. double$.ab,ti.<br />

57. 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56<br />

58. 45 and 57<br />

59. limit 58 to human<br />

Table 2. CINAHL search strategy<br />

Search terms<br />

1. wholegrain$.ab,ti.<br />

2. wholemeal$.ab,ti.<br />

3. wholewheat$.ab,ti.<br />

4. (whole adj3 grain$).ab,ti.<br />

5. (whole adj3 meal).ab,ti.<br />

6. (whole adj3 wheat).ab,ti.<br />

7. (whole adj3 food$).ab,ti.<br />

8. (wheat adj3 meal).ab,ti.<br />

9. cereal$.ab,ti.<br />

10. bread$.ab,ti.<br />

11. wheat$.ab,ti.<br />

12. oat$.ab,ti.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

6


Table 2. CINAHL search strategy (Continued)<br />

13. rye$.ab,ti.<br />

14. barley$.ab,ti.<br />

15. maize$.ab,ti.<br />

16. corn.ab,ti.<br />

17. cornmeal.ab,ti.<br />

18. popcorn.ab,ti.<br />

19. sorghum$.ab,ti.<br />

20. (bulgar or bulghar).ab,ti.<br />

21. couscous$.ab,ti.<br />

22. grain$.ab,ti.<br />

23. porridge.ab,ti.<br />

24. rice$.ab,ti.<br />

25. millet$.ab,ti.<br />

26. exp CEREALS/<br />

27. exp BREAD/<br />

28. exp Dietary Fiber/<br />

29. exp Coronary Disease/<br />

30. exp Cardiovascular Diseases/<br />

31. <strong>heart</strong> <strong>disease</strong>$.tw.<br />

32. <strong>coronary</strong> <strong>disease</strong>$.tw.<br />

33. chd.tw.<br />

34. cardiovascular.tw.<br />

35. angina.tw.<br />

36. cvd.tw.<br />

37. exp CHOLESTEROL/<br />

38. exp Blood Pressure/<br />

39. exp Obesity/<br />

40. exp Insulin Resistance/<br />

41. exp Diabetes Mellitus/<br />

42. exp LIPIDS/<br />

43. insulin resistance.ab,ti.<br />

44. insulin sensitivity.ab,ti.<br />

45. (glyc?emic adj3 control).ab,ti.<br />

46. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 26<br />

or 27 or 28<br />

47. 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 43 or 44 or 45<br />

48. 46 and 47<br />

49. clinical trial.pt.<br />

50. exp Clinical Trials/<br />

51. (clin$ adj25 trial$).tw.<br />

52. ((singl$ or doubl$ or trebl$ or tripl$) adj (blind$ or mask$)).tw.<br />

53. exp PLACEBOS/<br />

54. placebo$.tw.<br />

55. random$.tw.<br />

56. exp Evaluation Research/<br />

57. exp Prospective Studies/<br />

58. exp Random Assignment/<br />

59. exp Random Sample/<br />

60. exp Crossover Design/<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

7


Table 2. CINAHL search strategy (Continued)<br />

61. exp Comparative Studies/<br />

62. 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61<br />

63. 48 and 62<br />

Table 3. CENTRAL search strategy<br />

Search terms<br />

1. wholegrain$.af.<br />

2. wholemeal$.af.<br />

3. wholewheat$.af.<br />

4. (whole adj3 grain$).af.<br />

5. (whole adj3 meal$).af.<br />

6. (whole adj3 wheat$).af.<br />

7. (whole adj3 food$).af.<br />

8. (wheat adj3 meal$).af.<br />

9. <strong>cereals</strong>.af.<br />

10. bread$.af.<br />

11. wheat$.af.<br />

12. oat$.af.<br />

13. rye$.af.<br />

14. barley$.af.<br />

15. maize$.af.<br />

16. corn.af.<br />

17. cornmeal.af.<br />

18. popcorn.af.<br />

19. sorghum$.af.<br />

20. (bulgar or bulghar).af.<br />

21. couscous$.af.<br />

22. grain$.af.<br />

23. porridge$.af.<br />

24. (rice$:ti or rice$:ab).af.<br />

25. millet$.af.<br />

26. <strong>cereals</strong>.sh.<br />

27. bread.sh.<br />

28. dietary fiber.sh.<br />

29. (diet$ adj fiber$).af.<br />

30. (diet$ adj fiber$).af.<br />

31. or/1-30<br />

32. cardiovascular <strong>disease</strong>s.sh.<br />

33. (<strong>heart</strong> adj <strong>disease</strong>$).af.<br />

34. chd.af.<br />

35. (<strong>coronary</strong> adj3 <strong>disease</strong>$).af.<br />

36. cardiovascular.af.<br />

37. angina.af.<br />

38. cvd.af.<br />

39. cholesterol.sh.<br />

40. cholesterol.tw.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

8


Table 3. CENTRAL search strategy (Continued)<br />

41. (blood adj pressure).af.<br />

42. blood pressure.sh.<br />

43. hypertension.sh.<br />

44. hypertension.tw.<br />

45. obesity.sh.<br />

46. obesity.tw.<br />

47. obese.af.<br />

48. insulin resistance.af.<br />

49. (insulin adj resistance).af.<br />

50. (metabolic adj syndrome).af.<br />

51. diabetes mellitus.sh.<br />

52. diabetes.af.<br />

53. (insulin adj sensitivity).af.<br />

54. (glycemic adj3 control).af.<br />

55. (glycaemic adj3 control).af.<br />

56. hyperlipidemia.af.<br />

57. hyperlipidaemia.af.<br />

58. hyperlipidemia.sh.<br />

59. or/32-58<br />

60. 31 and 59<br />

1. wholegrain$.af.<br />

2. wholemeal$.af.<br />

3. wholewheat$.af.<br />

4. (whole adj3 grain$).af.<br />

5. (whole adj3 meal$).af.<br />

6. (whole adj3 wheat$).af.<br />

7. (whole adj3 food$).af.<br />

8. (wheat adj3 meal$).af.<br />

9. <strong>cereals</strong>.af.<br />

10. bread$.af.<br />

11. wheat$.af.<br />

12. oat$.af.<br />

13. rye$.af.<br />

14. barley$.af.<br />

15. maize$.af.<br />

16. corn.af.<br />

17. cornmeal.af.<br />

18. popcorn.af.<br />

19. sorghum$.af.<br />

20. (bulgar or bulghar).af.<br />

21. couscous$.af.<br />

22. grain$.af.<br />

23. porridge$.af.<br />

24. (rice$:ti or rice$:ab).af.<br />

25. millet$.af.<br />

26. <strong>cereals</strong>.sh.<br />

27. bread.sh.<br />

28. dietary fiber.sh.<br />

29. (diet$ adj fiber$).af.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

9


Table 3. CENTRAL search strategy (Continued)<br />

30. (diet$ adj fiber$).af.<br />

31. or/1-30<br />

32. cardiovascular <strong>disease</strong>s.sh.<br />

33. (<strong>heart</strong> adj <strong>disease</strong>$).af.<br />

34. chd.af.<br />

35. (<strong>coronary</strong> adj3 <strong>disease</strong>$).af.<br />

36. cardiovascular.af.<br />

37. angina.af.<br />

38. cvd.af.<br />

39. cholesterol.sh.<br />

40. cholesterol.tw.<br />

41. (blood adj pressure).af.<br />

42. blood pressure.sh.<br />

43. hypertension.sh.<br />

44. hypertension.tw.<br />

45. obesity.sh.<br />

46. obesity.tw.<br />

47. obese.af.<br />

48. insulin resistance.af.<br />

49. (insulin adj resistance).af.<br />

50. (metabolic adj syndrome).af.<br />

51. diabetes mellitus.sh.<br />

52. diabetes.af.<br />

53. (insulin adj sensitivity).af.<br />

54. (glycemic adj3 control).af.<br />

55. (glycaemic adj3 control).af.<br />

56. hyperlipidemia.af.<br />

57. hyperlipidaemia.af.<br />

58. hyperlipidemia.sh.<br />

59. or/32-58<br />

60. 31 and 59<br />

Table 4. ProQuest Digital Dissertations search strategy<br />

Search terms<br />

AB (wholegrain*)or AB (wholemeal)or AB (grain*) or AB (wheat*) or AB (cereal*) or AB ( bread* ) or AB (oat*) or AB (rye*) or AB<br />

(barley*) or AB (maize*) or AB (corn*) or AB ( sorghum*) or AB (bulgar*) or AB (bulghar*) or AB (couscous*)or AB (grain*) or AB<br />

(porridge*)or AB ( rice*) or AB (millet*)<br />

Data collection and analysis<br />

Trials selection<br />

For full details of the searches and selection of trials <strong>for</strong> the review<br />

see the Figure 1 (Moher 1999).<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

10


The MEDLINE search identified 1937 potentially relevant articles,<br />

the EMBASE search 2286 articles and the CINAHL search<br />

338 articles. The potentially relevant articles from these three<br />

databases were imported into Endnote reference management software<br />

and combined. The combined articles were then de-duplicated<br />

using the Endnote software leaving 3976 combined hits. A<br />

further 968 search hits were identified from the CENTRAL search<br />

and 77 hits from ProQuest Digital Dissertations and screened separately.<br />

Figure 1. QUOROM flow chart.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

One reviewer scanned the titles and abstracts of each record retrieved<br />

from the searches. Articles were only rejected on initial<br />

screen if the reviewer could determine from the title and abstract<br />

that the article definitely did not meet the inclusion criteria <strong>for</strong><br />

the review.<br />

When a title/abstract could not be rejected with certainty, the full<br />

text of the article was obtained <strong>for</strong> further evaluation and the entire<br />

article assessed. An in/out <strong>for</strong>m was used to assess studies inclusion<br />

(or otherwise) into the review. If a trial was excluded at any time<br />

11


after this point, a record of both the article and exclusion reason<br />

was kept.<br />

A total of 92 potentially relevant full papers were identified:-<br />

81 from the combined MEDLINE, EMBASE and CINAHL<br />

searches; a further two full papers were identified from the CEN-<br />

TRAL search; a further 9 potentially relevant papers were identified<br />

from handsearching of relevant reviews.<br />

Of the 92 potentially relevant full papers, 81 were excluded and<br />

11 papers were included into the review. Two papers were from<br />

the same study (Pins 2002), so 10 studies in total were included<br />

in the review.<br />

Data extraction<br />

Original reports of trial results were extracted by two reviewers independently.<br />

Any differences between reviewers’ extraction results<br />

were resolved by discussion and, when necessary, in consultation<br />

with a third reviewer.<br />

The following data were extracted.<br />

1. General in<strong>for</strong>mation: Published/unpublished, title, authors,<br />

source, country, year of publication, duplicate publications.<br />

2. Trial characteristics: Design, duration, randomisation (and<br />

method), allocation concealment (and method), blinding (outcome<br />

assessors), check of blinding.<br />

3. Intervention: Dietary in<strong>for</strong>mation/foods provided, length of<br />

intervention, comparison interventions.<br />

4. Participants: Sampling (random/convenience), exclusion criteria,<br />

total number and number in comparison groups, gender/age,<br />

diagnosis of CHD or risk factors, similarity of groups at baseline,<br />

withdrawals/losses to follow-up, assessment of compliance, medications<br />

used, smoking status where provided.<br />

5. Outcomes: Outcomes as specified above including what was<br />

the main outcome assessed in the study, other events, length of<br />

follow-up.<br />

6. Results: For outcomes and times of assessment.<br />

Quality assessment of trials<br />

The quality of each trial was assessed based mainly on the quality<br />

criteria specified in the Cochrane <strong>Review</strong>ers’ Handbook (Section<br />

4.1.6) (Higgins 2005). In particular the following factors were<br />

examined.<br />

1. Method of randomisation - each factor was marked as ’done’,<br />

’not done’ or ’unclear’.<br />

Could the study be described as randomised (including use of<br />

words such as ’random’, ’randomly’ and ’randomisation’)?<br />

Did the study describe the method of randomisation and was it<br />

an appropriate method? A method to determine the sequence of<br />

randomisation will be regarded as appropriate if it allows <strong>for</strong> each<br />

study participant to have the same chance of receiving each intervention<br />

and <strong>for</strong> the investigators to be unable to predict which<br />

treatment will be next. Methods of allocation using date of birth,<br />

date of admission, hospital numbers, or alternation were not regarded<br />

as appropriate.<br />

2. Concealment of allocation - this was scored A (adequate), B<br />

(unclear), C (inadequate), following criteria adopted from the<br />

Cochrane <strong>Review</strong>er’s Handbook (Section 4.1.6) (Higgins 2005)<br />

and Schulz et al (Schulz 1995).<br />

A - Adequate measures to conceal allocations such as central randomisation;<br />

serially numbered, opaque, sealed envelopes; or other<br />

descriptions with convincing concealment.<br />

B - Unclearly concealed trials, in which the authors either did not<br />

report allocation concealment at all, or reported an approach that<br />

did not fall into one of the categories in (A).<br />

C - Inadequately concealed trials, in which the method of allocation<br />

was not concealed, such as alteration methods or use of case<br />

record numbers.<br />

3. Blinding - each factor was marked as ’done’, ’not done’ or ’unclear’<br />

With lifestyle interventions, such as the topic <strong>for</strong> this review, it is<br />

difficult to blind participants and those providing dietary advice.<br />

However, it is possible to blind outcome assessors and this were<br />

marked as ’done’, ’not done’ or ’unclear’.<br />

4. Intention-to-treat analysis - marked as ’done’, ’not done’ or<br />

’unclear’<br />

Whether an intention-to-treat analysis was possible on all patients<br />

from the published data (i.e. whether there were any exclusions<br />

from the trial after randomisation) and the number of patients who<br />

were lost to follow-up. If there were no withdrawals this should<br />

have been stated in the article. An intention-to-treat analysis was<br />

considered adequate if outcome data was analysed <strong>for</strong> all participants<br />

randomised.<br />

Based on these criteria, studies were broadly subdivided into the<br />

following three categories, see Cochrane Handbook, section 6.7.1<br />

(Higgins 2005):<br />

A - all quality criteria met: low risk of bias.<br />

B - one or more of the quality criteria only partly met: moderate<br />

risk of bias.<br />

C - one or more criteria not met: high risk of bias.<br />

Data conversion<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

For the purposes of pooling data, serum and plasma cholesterol<br />

measurements were converted to units of millimoles per litre<br />

(mmol/L). In some studies outcome data were reported in milligrams<br />

per decilitre (mg/dL). This was converted to mmol/L by<br />

multiplying by conversion factors of 0.0259 <strong>for</strong> total, HDL and<br />

LDL cholesterol; 0.0113 <strong>for</strong> triglycerides; and 0.0555 <strong>for</strong> glucose<br />

(JAMA 2001). When insulin was reported in µIU/mL(microunits<br />

of insulin per millilitre) it was converted to pmol/L (picomoles of<br />

insulin per litre) by multiplying by a conversion factor of 6.945 (<br />

JAMA 2001). Where body weight was reported in pounds it was<br />

converted to kilograms by multiplying by a conversion factor of<br />

0.45.<br />

One study reported results <strong>for</strong> men and women separately (<br />

Leinonen 2000). These results were combined <strong>for</strong> meta-analysis.<br />

12


Data analysis<br />

No data on CHD mortality or cardiovascular events were found<br />

in the studies included in the review. Relevant outcome data identified<br />

in the included trials were:<br />

(1) total cholesterol (mmol/L) Outcome 01;<br />

(2) LDL cholesterol (mmol/L) Outcome 02;<br />

(3) HDL cholesterol (mmol/L) Outcome 03;<br />

(4) triglycerides (mmol/L) Outcome 04;<br />

(5) body weight (kg) Outcome 05;<br />

Some data was also found on fasting glucose, fasting insulin, insulin<br />

resistance and blood pressure but there was insufficient data<br />

to combine in a meta-analysis. A number of studies also reported<br />

whether there were any side effects as a secondary outcome.<br />

When data were available, sufficiently similar and of sufficient<br />

quality, statistical analyses were per<strong>for</strong>med using the RevMan software.<br />

Heterogeneity between trial results was tested <strong>for</strong> using a<br />

standard chi-squared test. A P value < 0.1 was used to indicate that<br />

significant heterogeneity was present. Heterogeneity was also examined<br />

using the I 2 statisitic (Higgins 2003). I 2 values over 50%<br />

indicate a substantial level of heterogeneity. Tests of heterogeneity<br />

were used <strong>for</strong> examining whether the observed variation in study<br />

results was compatible with variation expected with chance alone.<br />

If heterogeneity was found, then data were not pooled and we have<br />

attempted to determine potential reasons <strong>for</strong> heterogeneity by examining<br />

the study characteristics and populations. Where appropriate,<br />

recorded post treatment values were pooled using weighted<br />

mean differences in a fixed effects meta-analysis. No trials were<br />

found where participants were randomised by cluster/groups.<br />

Data has been pooled at the 4-week and end-of-study periods. In<br />

analyses using end of study data, data reported at the end of the<br />

study has been combined irrespective of the length of the study.<br />

Meta-analysis has not been per<strong>for</strong>med when less than three studies<br />

reported outcome data, except <strong>for</strong> some sensitivity analyses where<br />

the exclusion of a study from the analysis results in the pooling of<br />

only two studies.<br />

Sensitivity analysis was used to take into account the influence of<br />

various factors e.g. (a) study quality, (b) exclusion of particularly<br />

small and underpowered trials. The results of sensitivity analysis<br />

are reported separately within each comparison section.<br />

Crossover studies<br />

One crossover study was incorporated into the meta-analysis (<br />

Leinonen 2000). There was insufficient data published in the<br />

source paper to conduct paired analysis as recommended in the<br />

Cochrane Handbook (section 8.11.3) (Higgins 2005). The data<br />

incorporated into the meta-analysis was the final mean and standard<br />

deviation (SD) <strong>for</strong> the whole group at the end of each arm of<br />

the trial. Crossover and parallel studies have been presented separately<br />

within the meta-analysis and also combined.<br />

R E S U L T S<br />

Description of studies<br />

See: Characteristics of included studies; Characteristics of excluded<br />

studies; Characteristics of ongoing studies.<br />

Overview<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Details of the studies included in the review are shown in the<br />

table of characteristics of included studies. The column headed<br />

’methods’ presents the results of the quality assessment.<br />

Ten different studies were identified (from eleven papers) that met<br />

all the inclusion criteria <strong>for</strong> this review. No studies were found that<br />

reported the effect of wholegrain foods or diets on CHD mortality<br />

or CHD events and morbidity. All ten included studies report the<br />

effect of wholegrain foods or diets on major risk factors <strong>for</strong> CHD<br />

according to the inclusion criteria <strong>for</strong> this review outlined in the<br />

section on ’Types of outcome measures’. Eight of the studies were<br />

parallel randomised controlled trials (Davidson 1991; Johnston<br />

1998; Karmally 2005; Keenan 2002; Pins 2002; Reynolds 2000;<br />

Van Horn 1988; Van Horn 1991) and two studies were of randomised<br />

crossover design (Leinonen 2000; Pereira 2002). The total<br />

number of participants in the ten included studies was 914.<br />

One study (Pereira 2002) examined the effects of a varied diet<br />

containing wholegrain breakfast cereal, bread, rice, pasta, muffins,<br />

cookies and snacks and one study (Leinonen 2000) investigated the<br />

effects of wholegrain rye bread as part of the usual diet. However,<br />

in eight of the ten included studies the wholegrain component of<br />

the diet was oats (Davidson 1991; Johnston 1998; Karmally 2005;<br />

Keenan 2002; Pins 2002; Reynolds 2000; Van Horn 1988; Van<br />

Horn 1991; ). Three of the studies were based on a commercial<br />

wholegrain oat cereal (Cheerios) (Johnston 1998; Reynolds 2000;<br />

Karmally 2005).<br />

All studies included were over 4 weeks in length. In seven of the<br />

studies the intervention period was 6 weeks or less (Davidson<br />

1991;Johnston 1998; Karmally 2005; Keenan 2002; Leinonen<br />

2000; Pereira 2002; ; Reynolds 2000). Two studies continued <strong>for</strong><br />

8 weeks (Van Horn 1988; Van Horn 1991) and one <strong>for</strong> 12 weeks (<br />

Pins 2002) . All of the studies reported outcomes at the end of the<br />

intervention period. Some of the studies also reported outcomes<br />

at intermediate timescales (<strong>for</strong> example at 4 weeks). None of the<br />

studies reported follow-up outcomes beyond the end of the intervention<br />

period.<br />

Population and setting<br />

The inclusion criteria <strong>for</strong> this review specified that the participants<br />

should have at least one major risk factor <strong>for</strong> CHD. In six<br />

studies, participants were diagnosed with elevated cholesterol levels<br />

(Davidson 1991; Johnston 1998; Karmally 2005; Leinonen<br />

2000; Reynolds 2000; Van Horn 1991). Additionally, one study<br />

recruited a range of participants described as healthy but reports<br />

13


subgroup results <strong>for</strong> those with elevated cholesterol above the median<br />

at baseline (Van Horn 1988) and has also been included in<br />

the review because results <strong>for</strong> those with raised cholesterol levels<br />

can be separated out from the general population. In one study<br />

participants were diagnosed with hypertension (Pins 2002) and<br />

in another study, participants were both hypertensive and hyperinsulinemic<br />

(Keenan 2002). In one study, participants were both<br />

overweight and hyperinsulinemic (Pereira 2002).<br />

Most of the studies (nine) were conducted in the US and one<br />

in Finland (Leinonen 2000). Of the studies carried out in the<br />

US, four were carried out at the same University (University of<br />

Minnesota). However, it is not possible to determine from the<br />

published papers whether all four studies were conducted in the<br />

same clinic setting.<br />

Intervention<br />

The study by Leinonen et al (Leinonen 2000) determined the effects<br />

of wholemeal rye compared to white refined wheat bread in<br />

people with elevated serum cholesterol concentrations. The subjects<br />

were 18 men and 22 women with baseline serum cholesterol<br />

concentration of 6.4 +/- 0.2 mmol/L. The study design was a 2<br />

x 4 week crossover trial during which participants randomly consumed<br />

rye and refined wheat breads (20% of daily energy) as part<br />

of their usual diet <strong>for</strong> 4 weeks. There was a 4 week washout period<br />

between each intervention period.<br />

Pereira et al (Pereira 2002) compared diets (55% carbohydrate,<br />

30% fat, 15% protein) that included 6-10 servings of breakfast<br />

cereal, bread, rice, pasta , muffins, cookies and snacks of either<br />

whole or refined grains and measured the effect on insulin sensitivity<br />

in eleven overweight and obese adults aged 25 to 56 years.<br />

The study design was a crossover trial with two 6-week feeding<br />

periods with a washout of 6 weeks between feeding periods. The<br />

wholegrain diet was created by substitution of an equal volume of<br />

whole-grain food items <strong>for</strong> the refined grain products.<br />

Johnston et al (Johnston 1998) evaluated the effects of a wholegrain<br />

oat ready-to-eat cereal (Cheerios) in 135 men and women,<br />

aged 40-70 years, with elevated blood cholesterol levels (mean<br />

LDL-cholesterol levels of 130-190 mg/dL and triglycerides 130 mmHg, diastolic blood pressure >85 mmHg)<br />

and hyperinsulinaemic (>10uU/mL) men and women aged 27 to<br />

59 years. The study design was a randomised, controlled parallel<br />

group study with a 6 week intervention period.<br />

In a population of 88 men and women being treated <strong>for</strong> hypertension<br />

Pins et al (Pins 2002) compared the effect of wholegrain<br />

oat-based <strong>cereals</strong> with refined grain wheat-based <strong>cereals</strong>. Individuals<br />

in the wholegrain oats group received a daily serving of 60<br />

g of oatmeal and 77 g of oat squares. The cereal treatments were<br />

isocaloric and individuals in the control group consumed 65 g hot<br />

wheat cereal and 81 g of Kellog’s Crispix. The study was a randomised<br />

controlled parallel-group study of 12 weeks intervention<br />

in total, designed to determine the effect of wholegrain oat <strong>cereals</strong><br />

on the need <strong>for</strong> antihypertensive medications. The study consisted<br />

of three 4-week phases: a baseline feeding phase, a medication reduction<br />

phase and a maintenance phase. The wholegrain oat and<br />

control <strong>cereals</strong> were administered during each of the four phases<br />

of the study.<br />

Van Horn (Van Horn 1988) examined the effect of an oatmealenhanced<br />

diet compared to a fat-modified diet recommended by<br />

the American Heart Association (AHA). All participants initially<br />

followed the AHA diet <strong>for</strong> 4 weeks and were then randomly assigned<br />

<strong>for</strong> 8 weeks to either an intervention group which continued<br />

to follow the AHA diet with the addition of 56 g (2 oz dry<br />

weight) oatmeal isocalorically substituted <strong>for</strong> other carbohydrate<br />

foods, or to a control group that continued to follow the AHA<br />

diet but consumed no oat products throughout the study. Participants<br />

in the study were healthy volunteers (aged 30 to 65 years)<br />

not taking lipid-altering medication. However, the results of the<br />

study are reported by subgroups with different cholesterol levels<br />

at baseline. The study has been included in this review because it<br />

reports subgroup results <strong>for</strong> those with elevated cholesterol above<br />

the median at baseline (n = 118) so the results <strong>for</strong> this group can<br />

be separated out from the overall results.<br />

Van Horn (Van Horn 1991) also examined the effect of an oatmeal<br />

enhanced diet in a study where all participants recruited to the<br />

trial were hypercholesterolemic (serum cholesterol levels > 5.20<br />

mmol/L), aged between 22 to 76 years. Individuals (n = 80) were<br />

randomised <strong>for</strong> 8 weeks to one of two groups, stratified by sex and<br />

pre-screen cholesterol level, above or below 6.34 mmol/L, prior<br />

to baseline visit. The intervention group was asked to consume<br />

two packets (56.7 g dry weight) of oatmeal per day, isocalorically<br />

14


substituted <strong>for</strong> other carbohydrate foods in their diet. The control<br />

group was asked to maintain usual intake throughout the study.<br />

Davidson (Davidson 1991) examined the effect of oatmeal at doses<br />

of either 28 g (1 oz), 56 g (2 oz) or 84 g (3 oz) dry weight in<br />

adults aged 30 to 60 years with elevated serum cholesterol levels,<br />

compared to a farina control group (farina is a refined grain made<br />

from the endosperm of the grain only, and does not contain bran<br />

or germ), over 6 weeks. The specific aim of this study was to examine<br />

the hypocholesterolemic effects of Beta-glucan in oatmeal,<br />

however, the study does meet the inclusion criteria <strong>for</strong> this review<br />

as oatmeal is classified as wholegrain. The study also included<br />

separate groups randomised to additional oatbran, however, these<br />

groups were not included in the meta-analysis as bran products<br />

are excluded from this review. 156 adults with LDL-cholesterol<br />

levels above 4.14 mmol/L or between 3.37 and 4.14 mmol/L with<br />

multiple risk factors were randomised to 6 intervention groups or<br />

to the farina control group.<br />

The study by Karmally et al (Karmally 2005) investigated the<br />

cholesterol-lowering effect of an oat cereal versus a corn cereal<br />

in mild to moderate hypercholesterolemic Hispanic Americans<br />

over 6 weeks. The oat cereal is described as commercial Cheerios<br />

in the paper and as an oat bran cereal. However, the author has<br />

confirmed to us (W. Karmally, personal communication) that the<br />

commercial Cheerios used are a wholegrain cereal. One hundred<br />

and fifty two men and women, ages 30 to 70 years with baseline<br />

LDL-cholesterol levels between 120 and 190 mg/dL (3.1 to 4.9<br />

mmol/L) and triglycerides < 400mg/dL (< 4.5 mmol/L).<br />

Funding<br />

Based on the details provided in the published papers, most of the<br />

studies included in this review were wholly or partly funded by<br />

commercial sources with interests in cereal products.<br />

The studies by Davidson and Van Horn 1991 were supported<br />

by Quaker Oats Company (Davidson 1991; Van Horn 1991).<br />

Additionally, the Pins study (Pins 2002) was supported in part<br />

by Quaker Oats who provided test <strong>cereals</strong> and financial support<br />

and Van Horn 1988 (Van Horn 1988) acknowledges assistance<br />

from Quaker Oats personnel but provides no further details on<br />

funding.<br />

The studies by Johnston (Johnston 1998) and Pereira (Pereira<br />

2002) were supported in part by General Mills Inc. The authors of<br />

the Karmally study acknowledge the support of General Mills Inc,<br />

but emphasise that the study sponsors had no role in study design,<br />

data collection, data analysis, data interpretation or writing of the<br />

report. Leinonen (Leinonen 2000) was supported in part by the<br />

Fazer Bakeries Ltd. and Vaasan and Vaasan Ltd.<br />

Two studies provided no in<strong>for</strong>mation about funding (Keenan<br />

2002; Reynolds 2000). However, the first author of the Reynolds<br />

study gives an address based at General Mills Inc., US. The studies<br />

were both carried out at the University of Minnesota. General<br />

MIlls, the company that makes Cheerios is based in Minnesota.<br />

Risk of bias in included studies<br />

Overview<br />

All 10 studies had some methodological weaknesses according to<br />

criteria as set out in the Cochrane Handbook (Higgins 2005).<br />

None fulfilled all quality criteria. Two of the 10 studies were randomised<br />

crossover trials and 8 of the studies were parallel randomised<br />

controlled trials.<br />

Randomisation<br />

All of the trials mentioned randomisation without describing the<br />

randomisation process. None of the ten included studies reported<br />

on the method of randomisation so we could not tell if this was appropriate.<br />

Only one of the studies mentioned an attempt at allocation<br />

concealment (Reynolds 2000). Differences in baseline characteristics<br />

between intervention and control groups were discussed<br />

in all of the included studies, however in some studies randomisation<br />

does not appear to have been successful and differences in<br />

cholesterol levels between groups at baseline were bigger than the<br />

differences that the study was trying to measure (Reynolds 2000).<br />

Blinding<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Blinding of either the study participants or the providers of diets<br />

was not included in the protocol criteria <strong>for</strong> quality assessment of<br />

studies as it can be difficult to implement in lifestyle studies of the<br />

type included in the review. The original criteria <strong>for</strong> this systematic<br />

review only included blinding of outcome assessors. However,<br />

some of the included studies did report blinding of participants<br />

or providers of diets.<br />

One study reported that both the study participants and the experimental<br />

team who distributed the <strong>cereals</strong> were blinded (Reynolds<br />

2000). This was achieved by producing both the wholegrain cereal<br />

and the control cereal on the same commercial processing<br />

equipment and puffing the <strong>cereals</strong> into the same shape so that<br />

participants could not distinguish the <strong>cereals</strong> by shape. Both <strong>cereals</strong><br />

were packaged so that the experimenter giving out the <strong>cereals</strong><br />

could not tell the difference. Another study conducted by some<br />

of the same team members (Johnston 1998) may have used the<br />

same process to blind participants and the cereal providers. The<br />

study states the same in<strong>for</strong>mation about using commercial cereal<br />

processing equipment and packaging but does not explicitly state<br />

that participants and providers were blinded (we have attempted<br />

to contact the research team <strong>for</strong> this study to clarify this without<br />

success). It is not clear from the published in<strong>for</strong>mation whether<br />

outcome assessors were blinded in either of these studies.<br />

One study reported that the physician who dispensed the <strong>cereals</strong><br />

and the physician responsible <strong>for</strong> BP measurement, blood draws<br />

and patient examinations was blinded (Pins 2002).<br />

15


One study was reported to be single-blinded (Davidson 1991)<br />

but it was not clear from the published details which part of the<br />

process was blinded. One study reported that blinding of outcome<br />

assessors was done <strong>for</strong> blood pressure but did not state whether<br />

the process of sampling and measuring lipids was blinded (Keenan<br />

2002).<br />

In the other five included studies no details were given about blinding<br />

( Karmally 2005; Leinonen 2000; Pereira 2002; Van Horn<br />

1988; Van Horn 1991).<br />

Crossover studies<br />

A carryover effect may occur in a crossover trial when the effects<br />

of the treatment in the first period of the trial continue over to the<br />

second period. In such cases, the results from the second period<br />

may be inaccurate. A washout period is the time when no intervention<br />

occurs between the two treatment periods and should be<br />

long enough to allow the effects from the first treatment period to<br />

completely disappear.<br />

Both of the included studies that were crossover trials reported<br />

an adequate washout period between diet phases. The washout<br />

periods were 4 weeks (Leinonen 2000) and 6-9 weeks (Pereira<br />

2002). Meta-analysis has been carried out on crossover and parallel<br />

studies separately to determine if the study design has any influence<br />

on the overall results.<br />

Potential confounders<br />

Body weight<br />

Of the 10 included studies, nine stated that the dietary programmes<br />

were designed to maintain body weight in both the<br />

intervention and control groups (Johnston 1998; Keenan 2002;<br />

Leinonen 2000; Pereira 2002; Pins 2002; Reynolds 2000; Van<br />

Horn 1988; Van Horn 1991). One other study does not specifically<br />

state this but it is implied in the text (Davidson 1991).<br />

Nine of the 10 included studies reported no significant change in<br />

body weight throughout the study in either the intervention or<br />

control groups (Davidson 1991; Johnston 1998; Keenan 2002;<br />

Leinonen 2000; Pereira 2002; Pins 2002; Reynolds 2000; Van<br />

Horn 1991). One study reported no difference after 8 weeks, but<br />

a small increase after 4 weeks in the mean weight of the oatmeal<br />

group (+0.2 lb) compared to the control group (-0.6 lb) (P = 0.039<br />

<strong>for</strong> the intergroup difference) (Van Horn 1988).<br />

Medication status<br />

Most of the studies excluded participants on lipid-lowering medication<br />

(Davidson 1991; Karmally 2005; Leinonen 2000; Pereira<br />

2002; Van Horn 1988; Van Horn 1991 ). Three studies included<br />

participants on beta-blockers as long as the dose was kept constant<br />

throughout the study (Johnston 1998; Reynolds 2000). The study<br />

by Johnston excluded subjects taking certain drugs e.g. corticosteroids,<br />

androgens or lipid-lowering drugs but did allow subjects<br />

to take other drugs potentially affecting lipids such as beta-blockers,<br />

estrogen, estrogen progestin, thiazide diuretics and thyroid<br />

hormones as long as they were on stable doses (Johnston 1998).<br />

In one study the major outcome was reduction in medication<br />

status and improvement in blood pressure control (Pins 2002),<br />

so participants were allowed no more than one anti-hypertensive<br />

medication and/or one diuretic medication.<br />

In one study medication status was not reported (Keenan 2002).<br />

Smoking status<br />

Smokers were excluded from three of the studies (Keenan 2002;<br />

Pereira 2002; Pins 2002) and smoking status was not reported<br />

in five studies (Davidson 1991; Karmally 2005; Leinonen 2000;<br />

Van Horn 1988; Van Horn 1991). The inclusion criteria <strong>for</strong> two<br />

studies allowed participants to smoke (Johnston 1998; Reynolds<br />

2000). In both studies, smoking remained constant throughout<br />

the study <strong>for</strong> both intervention and control groups.<br />

Dietary fibre<br />

Any reported differences in dietary fibre content between the treatment<br />

and control groups have been reported in the ’Notes’ section<br />

of the table of ’Characteristics of included studies’. The focus of<br />

this review is to assess whether wholegrain diets have a beneficial<br />

effect on CHD and major risk factors <strong>for</strong> CHD in order to provide<br />

recommendations as to whether more wholegrain foods should be<br />

included in the diets of the general public. All wholegrains contain<br />

dietary fibre. Dietary fibre content may potentially be a major<br />

contributing factor to any differences between wholegrain diets<br />

and the control diets but is not the primary focus of this review.<br />

Compliance to diet<br />

None of the studies reported any problems with compliance to<br />

the diets.<br />

Power<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Power calculations were conducted in four of the 10 studies included<br />

in this review (Davidson 1991; Leinonen 2000; Pins 2002;<br />

Reynolds 2000). The sample size of the study by Leinonen (<br />

Leinonen 2000) was based on 80% power to detect a difference in<br />

serum total cholesterol concentration of 0.4 to 0.5 mmol/L based<br />

on a significance level of 0.05. The study by Davidson (Davidson<br />

1991) was designed to have an 90% chance of detecting a difference<br />

in lipid levels of 0.39 mmol/L at significance 0.05. It is not<br />

clear from the published details whether the power calculations in<br />

the Davidson study took into account that the participants were<br />

randomised to a total of seven groups. Such differences of 0.3-<br />

0.5 mmol/L are reasonable <strong>for</strong> dietary intervention studies but the<br />

16


differences measured in the studies included in this review were<br />

often lower than this. The sample size calculation in the Keenan<br />

study was based on a level of significance set at 0.05 and power of<br />

80% to detect a 15% difference in medication reduction (Keenan<br />

2002). While the study by Reynolds did report that power calculations<br />

were done be<strong>for</strong>e the study, the calculations showed only<br />

a 53% chance of detecting a 3.8% mean drop in cholesterol assuming<br />

25 subjects in each group at a significance level of 0.05.<br />

In the paper, however, the analysis was actually based on less than<br />

25 subjects in each group (21 and 22 subjects) so the power of the<br />

study is likely to have been less than 53%. None of the other five<br />

included studies reported power calculations. In two of the studies<br />

that did not report power calculations, the number of participants<br />

included in the analysis was very small. In the cross-over study<br />

by Pereira only 12 people were randomised (Pereira 2002). In the<br />

study by Keenan, there were only 10 in the treatment group and<br />

eight in the control group (Keenan 2002).<br />

Population<br />

In all 10 studies included in the review, the participants were diagnosed<br />

with risk factors <strong>for</strong> CHD at baseline. In seven studies<br />

the participants had raised cholesterol (Davidson 1991; Johnston<br />

1998; Karmally 2005; Leinonen 2000; Reynolds 2000; Van Horn<br />

1988; Van Horn 1991), in two studies they were hypertensive (<br />

Keenan 2002; Pins 2002) and in one study they were overweight<br />

(Pereira 2002). The study by Pereira was not included in the analyses<br />

as it reported a different outcome (insulin sensitivity) than<br />

the other studies. The studies on hypertensive participants both<br />

reported lower baseline cholesterol levels than most of the other<br />

studies but the results were consistent with the studies on hypercholesterolemic<br />

participants with lower cholesterol reported in the<br />

wholegrain group than the control group.<br />

Intention to treat (ITT) analysis<br />

Only one of the 10 included studies reported an intention-totreat<br />

(ITT) analysis along with an analysis of those compliant<br />

to diet (Davidson 1991). However, the reported ITT analysis is<br />

considered to be inadequate because it did not incorporate data<br />

<strong>for</strong> all subjects randomised (156 subjects were randomised, but the<br />

ITT analysis only took into account data from 148 subjects). A<br />

further two studies (Leinonen 2000; Pins 2002) did not <strong>for</strong>mally<br />

report an ITT analysis but in both cases all the subjects who were<br />

randomised were included in the analysis and no drop-outs were<br />

reported. None of the other studies conducted ITT analysis.<br />

Quality assessment<br />

Table 5. Summary of main results (comparison wholegrain versus control)<br />

None of the studies met all of the quality assessment parameters<br />

specified <strong>for</strong> this review based on the criteria outlined in the<br />

Cochrane <strong>Review</strong>er’s handbook (Higgins 2005) and listed above<br />

in the section on ’Quality Assessment of Trials’ in ’Methods of the<br />

review’. Based on these criteria all 10 of the included studies would<br />

be classified as quality C. This means that one or more quality<br />

criteria were not met and the studies have a high risk of bias. None<br />

of the studies were excluded on the basis of a low quality score<br />

as all of the 10 included studies were graded at the same level of<br />

quality (C) based on the Cochrane criteria.<br />

As all the trials were rated at the same level, sensitivity analysis on<br />

the basis of quality was not carried out.<br />

Effects of interventions<br />

Heterogeneity was not found in the pooled analyses <strong>for</strong> total<br />

cholesterol, LDL cholesterol, HDL cholesterol and triglycerides.<br />

Heterogeneity was found <strong>for</strong> the combined body weight data so<br />

the results of pooled analyses are not shown in the graphs, only<br />

the sub-totals (Comparison 05).<br />

A summary of the main results is shown in additional Table 5.<br />

Outcome Intervention No. of studies Pooled result Type of wholegrain<br />

Outcome measure (sensitivity<br />

analyses)<br />

Total cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

All end of study outcomes<br />

(mmol/L)<br />

8 Parallel studies 0.19 (95% CI<br />

-0.30 to -0.08), P = 0.0005<br />

7 Parallel -0.22 (95% CI-0.33<br />

to -0.10), P = 0.0002<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Oats<br />

Oats<br />

17


Table 5. Summary of main results (comparison wholegrain versus control) (Continued)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

Total cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

LDL cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds<br />

study)<br />

All end of study outcomes<br />

All end of study outcomes<br />

LDL cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

HDL cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

6 Parallel -0.21 (95% CI-0.32<br />

to -0.09), P = 0.0006<br />

9 Parallel and crossover studies -<br />

0.20 (95% CI-0.31 to -0.10),<br />

P = 0.0001<br />

8 Parallel and crossover studies -<br />

0.23 (95% CI-0.33 to -0.12),<br />

P < 0.0001<br />

7 Parallel and crossover studies -<br />

0.22 (95% CI-0.33 to -0.10),<br />

P = 0.0002<br />

8 Parallel studies -0.18 (95%<br />

CI-0.28 to -0.09), P = 0.0001<br />

7 Parallel studies -0.20 (95%<br />

CI-0.30 to -0.10), P < 0.0001<br />

6 Parallel studies -0.19 (95%<br />

CI-0.29 to -0.09), P = 0.0002<br />

9 Parallel and crossover studies -<br />

0.18 (95% CI-0.28 to -0.09),<br />

P < 0.0001<br />

8 Parallel and crossover studies<br />

-0.20 (95% CI-0.30 to 0.11),<br />

P < 0.0001<br />

7 Parallel and crossover studies -<br />

0.19 (95% CI-0.29 to -0.10),<br />

P < 0.0001<br />

6 Parallel studies 0.00 (95% CI-<br />

0.05 to -0.05), P = 0.95<br />

5 Parallel studies 0.00 (95% CI-<br />

0.06 to -0.05), P = 0.88<br />

4 Parallel studies -0.01 (95%<br />

CI-0.07 to -0.05), P = 0.87<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Oats<br />

Oats and rye<br />

Oats and rye<br />

Oats and rye<br />

Oats<br />

Oats<br />

Oats<br />

Oats and rye<br />

Oats and rye<br />

Oats and rye<br />

Oats<br />

Oats<br />

Oats<br />

18


Table 5. Summary of main results (comparison wholegrain versus control) (Continued)<br />

HDL cholesterol All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

Triglycerides All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

Triglycerides All end of study outcomes<br />

- (without Reynolds<br />

study)<br />

- (without Reynolds/<br />

Keenan)<br />

All end of study outcomes<br />

All end of study outcomes<br />

7 Parallel and crossover studies<br />

0.00 (95% CI-0.05 to 0.05),<br />

P = 0.95<br />

6 Parallel and crossover studies<br />

0.00 (95% CI-0.06 to 0.05),<br />

P = 0.89<br />

5 Parallel and crossover studies<br />

0.00 (95% CI-0.06 to 0.05),<br />

P = 0.87<br />

6 Parallel studies 0.01 (95% CI-<br />

0.11 to 0.13), P = 0.83<br />

5 Parallel studies -0.02 (95%<br />

CI-0.14 to 0.11), P = 0.81<br />

4 Parallel studies 0.00 (95% CI-<br />

0.13 to 0.13), P = 1.00<br />

7 Parallel and crossover studies<br />

0.01 (95% CI-0.09 to 0.11),<br />

P = 0.86<br />

6 Parallel and crossover studies<br />

-0.01 (95% CI-0.11 to 0.09),<br />

P = 0.85<br />

5 Parallel and crossover studies<br />

0.00 (95% CI-0.11 to 0.11),<br />

P = 1.00<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Oats and rye<br />

Oats and rye<br />

Oats and rye<br />

Oats<br />

Oats<br />

Oats<br />

Oats and rye<br />

Oats and rye<br />

Oats and rye<br />

19


Table 5. Summary of main results (comparison wholegrain versus control) (Continued)<br />

Comparison 01 Total cholesterol<br />

4 week outcomes Comparison 01 Outcome 01<br />

Three parallel studies (Reynolds 2000; Van Horn 1988; Van Horn<br />

1991) and one crossover study (Leinonen 2000) reported outcomes<br />

at 4 weeks. The three parallel studies were based on wholegrain<br />

oatmeal and the crossover study was on wholegrain rye.<br />

On an intra-study basis, two of the three parallel studies in oats<br />

reported lower total cholesterol at 4 weeks in the wholegrain groups<br />

than in the control groups (Van Horn 1988; Van Horn 1991).<br />

In the other study (Reynolds 2000), a bigger reduction in total<br />

cholesterol was observed in the wholegrain group when the preand<br />

post-treatment data were compared. However, this effect is<br />

not evident from the post-treatment data presented in the metaanalysis<br />

because of a large baseline difference between the groups<br />

and insufficient power to detect the small changes in cholesterol<br />

observed. The crossover study on rye also reported lower total<br />

cholesterol at 4 weeks in the wholegrain group (Leinonen 2000).<br />

The difference between the groups was not significant in any of<br />

the studies.<br />

When the 4 week results from the three parallel studies (all oats)<br />

were pooled in a meta-analysis, there was no evidence of an effect<br />

on total cholesterol in the wholegrain group compared to the<br />

control group, although the direction of the effect favoured a reduction<br />

in total cholesterol on the wholegrain diets. The weighted<br />

mean difference was -0.12 mmol/L (95% CI -0.32 to 0.09). Addition<br />

of the one crossover study (rye) to the analysis did not change<br />

the result markedly and gave weighted mean difference of -0.16<br />

mmol/L (95% CI -0.34 to 0.01).<br />

In a sensitivity analysis, the data was re-analysed without the study<br />

by Reynolds et al because of the baseline differences between the<br />

groups. With only two parallel studies in the sensitivity analysis,<br />

there remained no evidence of an effect on total cholesterol with<br />

a weighted mean difference of -0.20 mmol/L (95% CI -0.46 to<br />

-0.06). However, when both parallel and crossover studies were<br />

included in the analysis, there was some evidence <strong>for</strong> a reduction in<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

total cholesterol in the wholegrain group (effect size -0.24 mmol/L,<br />

95% CI -0.45 to -0.03, P = 0.03).<br />

All end-of-study outcomes Comparison 01 Outcome 02<br />

In total, eight parallel studies (Davidson 1991; Johnston 1998;<br />

Karmally 2005; Keenan 2002; Pins 2002; Reynolds 2000; Van<br />

Horn 1991; Van Horn 1988) and one crossover study (Leinonen<br />

2000) reported total cholesterol outcomes. All eight parallel studies<br />

were based on wholegrain oatmeal and the crossover study was on<br />

wholegrain rye.<br />

On an intra-study basis, seven of the eight parallel studies in oats<br />

reported lower total cholesterol in the wholegrain groups than<br />

in the control groups (Davidson 1991; Johnston 1998; Karmally<br />

2005; Keenan 2002; Pins 2002; Van Horn 1988; Van Horn 1991).<br />

The difference between the groups was significant in one study (<br />

Pins 2002). In the other study (Reynolds 2000), a bigger reduction<br />

in total cholesterol was observed in the wholegrain group when<br />

the pre- and post-treatment data were compared. However, this<br />

effect is not evident from the post-treatment data presented in<br />

the meta-analysis because of a large baseline difference between<br />

the groups and insufficient power to detect the small changes in<br />

cholesterol observed. The one crossover study on rye also reported<br />

lower total cholesterol in the wholegrain group (Leinonen 2000)<br />

but the difference was not significant.<br />

When the results from the eight parallel studies (all oats) were<br />

pooled in a meta-analysis, there was evidence of an effect on total<br />

cholesterol in the wholegrain group compared to the control<br />

group, and the direction of the effect favoured lower total cholesterol<br />

on the wholegrain diets. The weighted mean difference was<br />

-0.19mmol/L (95% CI -0.30 to -0.08, P = 0.0005). Addition of<br />

the one crossover study (rye) to the analysis had little effect on the<br />

combined results (weighted mean difference -0.20, 95% CI -0.31<br />

to -0.10, P = 0.0001).<br />

Sensitivity analysis was conducted without the study by Reynolds.<br />

The effect increased slightly and remained significant (-0.23 95%<br />

CI -0.33 to -0.12). Additionally, removal of the study by Keenan<br />

20


ecause of the small number of subjects had little effect on the<br />

analysis (weighted mean difference -0.22 mmol/L, 95% CI -0.33<br />

to -0.10).<br />

Total cholesterol - summary of results<br />

There is evidence that total cholesterol is lower on diets containing<br />

oatmeal compared to diets based on refined grains. Eight of nine<br />

studies that reported total cholesterol as an outcome were based on<br />

oatmeal. When the end-of-study results from the eight studies were<br />

combined, the weighted mean difference was -0.19 mmol/L (95%<br />

CI -0.30 to -0.08, P = 0.0005) <strong>for</strong> the oatmeal diets compared<br />

to the refined grain diets. A similar effect was seen from three<br />

studies that provided data at 6 weeks intervention (weighted mean<br />

difference -0.23, 95% CI -0.40 to -0.05, P = 0.01), but not at 4<br />

weeks (except in one sensitivity analysis).<br />

There is insufficient evidence to make any conclusions about<br />

wholegrain diets other than oatmeal. Only one other study on<br />

other wholegrains was found. When this study on rye was pooled<br />

with the other data <strong>for</strong> oatmeal, it made little difference to the<br />

results.<br />

Comparison 02 LDL cholesterol<br />

4 week outcomes Comparison 02 Outcome 01<br />

Three parallel studies (Reynolds 2000; Van Horn 1988; Van Horn<br />

1991) and one crossover study (Leinonen 2000) reported LDL<br />

cholesterol at 4 weeks. The three parallel studies were based on<br />

wholegrain oatmeal and the crossover study was on wholegrain<br />

rye.<br />

On an intra-study basis, two of the three parallel studies in oats reported<br />

lower LDL cholesterol at 4 weeks in the wholegrain groups<br />

than in the control groups (Van Horn 1988; Van Horn 1991). The<br />

difference between the groups was significant in one of the studies<br />

(Van Horn 1988). In the other study by Reynolds et al, a bigger reduction<br />

in LDL cholesterol was observed in the wholegrain group<br />

when the pre- and post-treatment data were compared. However,<br />

this effect is not evident from the post-treatment data presented<br />

in the meta-analysis because of a large baseline difference between<br />

the groups and insufficient power to detect the small changes in<br />

cholesterol observed. In the one crossover study on rye, lower LDL<br />

cholesterol was also reported in the wholegrain group (Leinonen<br />

2000), but the difference was not significant.<br />

When the 4 week results from the three parallel studies (all oats)<br />

were pooled, there was no evidence of an effect on LDL cholesterol<br />

in the wholegrain group compared to the control group, although<br />

the direction of the effect favoured a reduction in LDL cholesterol<br />

on the wholegrain diets and was close to significance (P = 0.07).<br />

The weighted mean difference was -0.19 mmol/L (95% CI -0.39<br />

to 0.02). However, the combined analysis did become significant<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

when the one crossover study (rye) was included in the analysis.<br />

The weighted mean difference was -0.19 mmol/L (95% CI -0.37<br />

to -0.01, P = 0.04).<br />

In a sensitivity analysis, the data was re-analysed without the study<br />

by Reynolds et al because of differences between groups at baseline.<br />

With only two parallel studies in the sensitivity analysis, there was<br />

evidence of an effect on LDL cholesterol with a weighted mean<br />

difference of -0.31 mmol/L (95% CI -0.56 to -0.05). Addition of<br />

the crossover study to the analysis had little effect (effect size -0.28<br />

mmol/L, 95% CI -0.49 to -0.06).<br />

All end-of-study outcomes Comparison 02 Outcome 02<br />

In total, eight parallel studies (Davidson 1991; Johnston 1998;<br />

Karmally 2005; Keenan 2002; Pins 2002; Reynolds 2000; Van<br />

Horn 1988; Van Horn 1991) and one crossover study (Leinonen<br />

2000) reported LDL cholesterol outcomes. All eight parallel studies<br />

were based on wholegrain oatmeal and the crossover study was<br />

on wholegrain rye.<br />

On an intra-study basis, seven of the eight parallel studies in oats<br />

reported lower LDL cholesterol in the wholegrain groups than<br />

in the control groups (Davidson 1991; Johnston 1998; Karmally<br />

2005; Keenan 2002; Pins 2002; Van Horn 1988; Van Horn 1991).<br />

The difference between the groups was significant in three studies<br />

(Johnston 1998; Karmally 2005; Pins 2002). In the other study<br />

by Reynolds et al, a bigger reduction in LDL cholesterol was observed<br />

in the wholegrain group when the pre- and post-treatment<br />

data were compared. However, this effect is not evident from the<br />

post-treatment data presented in the meta-analysis because of a<br />

large baseline difference between the groups and insufficient power<br />

to detect the small changes in cholesterol observed. In the one<br />

crossover study in rye, LDL cholesterol was also lower in the wholegrain<br />

group (Leinonen 2000) but the difference was not significant.<br />

When the results from the eight parallel studies (all oats) were<br />

pooled in a meta-analysis, there was evidence of an effect on<br />

LDL cholesterol in the wholegrain group compared to the control<br />

group, and the direction of the effect favoured lower LDL cholesterol<br />

on the wholegrain diets. The weighted mean difference was<br />

-0.18 mmol/L (95% CI -0.28 to -0.09, P = 0.0001). Addition of<br />

the one crossover study (rye) to the analysis had little effect on the<br />

combined results (weighted mean difference -0.18, 95% CI -0.28<br />

to -0.09, P < 0.0001).<br />

Sensitivity analysis was conducted without the study by Reynolds.<br />

When all the studies were included, the effect increased slightly and<br />

remained significant. Effect size -0.20 (95% CI -0.30 to -0.11).<br />

Similarly, removal of the study by Keenan, because of the small<br />

number of subjects, had little effect on the analysis (weighted mean<br />

difference -0.19 mmol/L (95% CI -0.29 to -0.10, P < 0.0001).<br />

21


LDL cholesterol - summary of results<br />

Eight of nine studies that reported LDL cholesterol as an outcome<br />

were based on oatmeal. There is evidence that LDL cholesterol is<br />

lower on diets containing oatmeal compared to diets based on refined<br />

grains. When the end-of-study results from the eight studies<br />

were combined, the weighted mean difference was -0.18 mmol/L<br />

(95% CI -0.28 to -0.09, P < 0.0001) <strong>for</strong> the oatmeal diets compared<br />

to the refined grain diets. A similar effect was seen from<br />

three studies that provided data at 6 weeks intervention (weighted<br />

mean difference -0.25, 95% CI -0.39 to -0.10, p= 0.0008), but<br />

not at 4 weeks (except in one sensitivity analysis).<br />

There is insufficient evidence to make any conclusions about<br />

wholegrain diets other than oatmeal. Only one other study on<br />

other wholegrains was found. When this study on rye was pooled<br />

with the other data <strong>for</strong> oatmeal, it made little difference to the<br />

overall results.<br />

Comparison 03 HDL cholesterol<br />

4 week outcomes Comparison 03 Outcome 01<br />

Two parallel studies ( Reynolds 2000; Van Horn 1991) and one<br />

crossover study (Leinonen 2000) reported outcomes at 4 weeks.<br />

The two parallel studies were based on wholegrain oatmeal and<br />

the crossover study was on wholegrain rye.<br />

On an intra-study basis, the two parallel studies reported slightly<br />

higher HDL cholesterol in the wholegrain group (Reynolds 2000;<br />

Van Horn 1991) and the crossover study (Leinonen 2000) showed<br />

no difference between groups, but the difference between the<br />

wholegrain and control groups was not significant in any of the<br />

studies.<br />

When the 4 week results from all three studies were pooled in a<br />

meta-analysis, there was no evidence of an overall effect on HDL<br />

cholesterol. Parallel and crossover studies are reported together<br />

because there are only three studies. The weighted mean difference<br />

between the wholegrain and control groups was 0.03 mmol/L<br />

(95% CI -0.06 to 0.11), slightly higher in the wholegrain group.<br />

The result did not differ markedly in a sensitivity analysis, without<br />

the study by Reynolds et al.<br />

All end-of-study outcomes Comparison 03 Outcome 02<br />

In total six parallel studies (Davidson 1991; Johnston 1998;<br />

Karmally 2005; Keenan 2002; Reynolds 2000; Van Horn 1991)<br />

and one crossover study (Leinonen 2000) reported HDL cholesterol<br />

outcomes. All six parallel studies were based on wholegrain<br />

oatmeal and the crossover study was on wholegrain rye.<br />

On an intra-study basis, four of the six parallel studies reported<br />

higher HDL cholesterol in the wholegrain groups than in the<br />

control groups (Davidson 1991; Johnston 1998; Keenan 2002;<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Reynolds 2000), two studies reported lower HDL cholesterol in<br />

the wholegrain group ( Karmally 2005; Van Horn 1991). The<br />

one crossover study in rye reported no difference between the<br />

groups (Leinonen 2000). The difference between the groups was<br />

not significant in any of the studies.<br />

When the results from the six parallel studies (all oats) were<br />

pooled in a meta-analysis, there was no evidence of an effect on<br />

HDL cholesterol in the wholegrain group compared to the control<br />

group. The weighted mean difference was 0.00 mmol/L (95% CI<br />

-0.05 to 0.05, P = 0.95). Addition of the one crossover study (rye)<br />

to the analysis had no effect on the combined results. Also, sensitivity<br />

analyses without either the Reynolds study and additionally<br />

without the Keenan study had no effect on the combined result.<br />

HDL cholesterol - summary of results<br />

There is no evidence of a difference in HDL cholesterol on diets<br />

containing oatmeal compared to diets based on refined grains from<br />

outcome data from pooling of data at 4 weeks, 6 weeks or from<br />

the pooling of all end-of-study data.<br />

There is insufficient evidence to make any conclusions about<br />

wholegrain diets other than oatmeal. Only one other study on<br />

other wholegrains was found. When this study on rye was pooled<br />

with the other data <strong>for</strong> oatmeal, it made no difference to the results.<br />

Comparison 04 Triglycerides<br />

4 week outcomes Comparison 04 Outcome 01<br />

Two parallel studies (Reynolds 2000; Van Horn 1991) and one<br />

crossover study (Leinonen 2000) reported outcomes at 4 weeks.<br />

The two parallel studies were based on wholegrain oatmeal and<br />

the crossover study was on wholegrain rye.<br />

On an intra-study basis, one of the two parallel studies in oats<br />

reported higher triglycerides at 4 weeks in the wholegrain groups<br />

than in the control groups (Van Horn 1991), one reported lower<br />

mean triglycerides (Reynolds 2000). The one crossover study in<br />

rye reported no difference between the groups (Leinonen 2000).<br />

The difference between the groups was not significant in any of<br />

the studies.<br />

When the 4 week results from all three studies were pooled in<br />

a meta-analysis, there was no evidence of an effect on triglycerides<br />

in the wholegrain group compared to the control group.<br />

The weighted mean difference was 0.01 mmol/L (95% CI -0.13<br />

to 0.16, P = 0.84).<br />

In a sensitivity analysis, the data was re-analysed without the study<br />

by Reynolds et al. With only two studies in the sensitivity analysis<br />

(one crossover and one parallel), there remained no evidence of<br />

an effect on triglycerides with a weighted mean difference of 0.05<br />

mmol/L (95% CI -0.11 to 0.20).<br />

22


All end-of-study outcomes (Comparison 04 Outcomes 02-<br />

03)<br />

In total, six parallel studies (Davidson 1991; Karmally 2005;<br />

Keenan 2002; Pins 2002; Reynolds 2000; Van Horn 1991) and<br />

one crossover study (Leinonen 2000) reported triglycerides as an<br />

outcome. All six parallel studies were based on wholegrain oatmeal<br />

and the crossover study was on wholegrain rye.<br />

On an intra-study basis, three of the parallel studies in oats reported<br />

lower triglycerides in the wholegrain groups than in the control<br />

group (Davidson 1991; Pins 2002; Reynolds 2000), three studies<br />

reported higher triglycerides in the wholegrain group (Karmally<br />

2005; Keenan 2002; Van Horn 1991). The one parallel study in<br />

rye reported no difference between the groups (Leinonen 2000).<br />

The difference between the groups was not significant in any of<br />

the studies.<br />

When the results from the six parallel studies (all oats) were pooled<br />

in a meta-analysis, there was no evidence of an effect on triglycerides<br />

in the wholegrain group compared to the control group<br />

(Comparison 04 Outcome 02). The weighted mean difference was<br />

0.01 mmol/L (95% CI -0.11 to 0.13, P = 0.83). Addition of the<br />

one crossover study (rye) to the analysis had little effect on the<br />

combined results (weighted mean difference 0.01, 95% CI -0.09<br />

to 0.11, P = 0.86). The study by Keenan reports the data as mean<br />

triglycerides and standard error of the mean (SEM). However, the<br />

data appears to be SD rather than SEM. The pooled data was analysed<br />

assuming the reported data was SD (Comparison 04 Outcome<br />

02) which gave weighted mean difference 0.01 (95% CI -0.09 to<br />

0.11) and then assuming the reported data was SEM (Comparison<br />

04 Outcome 03) which gave weighted mean difference -0.01 (95%<br />

CI -0.11 to 0.09). Little difference was seen in the pooled results<br />

using SD or SEM <strong>for</strong> this study.<br />

Sensitivity analysis was conducted without the studies by Reynolds<br />

and Keenan. There was little effect on the combined results and<br />

the weighted mean difference between the wholegrain and control<br />

diets was 0.00 mmol/L (95% CI -0.11 to 0.11).<br />

Triglycerides - summary of results<br />

There is no evidence of a difference in triglycerides on diets containing<br />

oatmeal compared to diets based on refined grains from<br />

outcome data from pooling of data at 4 weeks, 6 weeks or from<br />

the pooling of all end-of- study data.<br />

There is insufficient evidence to make any conclusions about<br />

wholegrain diets other than oatmeal. Only one other study on<br />

other wholegrains was found. When this study on rye was pooled<br />

with the other data <strong>for</strong> oatmeal, it made no difference to the results.<br />

Comparison 05 Body weight<br />

All end-of-study outcomes (Comparison 05, Outcome 01)<br />

In total, five parallel studies (Johnston 1998; Karmally 2005;<br />

Keenan 2002; Pins 2002; Reynolds 2000 ) and two crossover studies<br />

(Leinonen 2000; Pereira 2002) reported body weight at the end<br />

of the study. All five parallel studies were based on wholegrain oatmeal<br />

and the crossover studies were on wholegrain rye and a general<br />

wholegrain diet. In all of the studies the dietary programmes<br />

were designed to maintain body weight in both the wholegrain<br />

and control participants.<br />

On an intra-study basis, three of the parallel studies (oats) reported<br />

lower final body weight in the wholegrain groups than in the control<br />

group (Keenan 2002; Pins 2002; Reynolds 2000) and two<br />

studies reported higher body weight (Johnston 1998; Karmally<br />

2005). The difference between the groups was significant in only<br />

one study (Karmally 2005). However, in the paper, a big difference<br />

between the groups was apparent at baseline so this effect appears<br />

to be due to poor randomisation. Neither of the two crossover<br />

studies (rye, general wholegrain diet) reported a significant difference<br />

in body weight between the wholegrain diets and control<br />

diets.<br />

Body weight data was not pooled because there was significant<br />

heterogeneity (parallel studies on oats, P =0.02, I 2 =64.2%; parallel<br />

and crossover studies P=0.04. I 2 =54.1%). Graphs are shown <strong>for</strong><br />

in<strong>for</strong>mation only (Comparison 05, Outcome 01).<br />

Body weight - summary of results<br />

There is no evidence of a difference in body weight on diets containing<br />

oatmeal compared to diets based on refined grains from<br />

individual studies. Data was not pooled in meta-analysis because<br />

significant heterogeneity was present.<br />

There is insufficient evidence to make any conclusions about<br />

wholegrain diets other than oatmeal. Two other studies on other<br />

wholegrains were found.<br />

Other outcomes<br />

Insufficient evidence was found to make any conclusions about<br />

the effect of wholegrains on any other risk factors <strong>for</strong> CHD. A<br />

brief summary of findings is given below.<br />

Fasting glucose<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Two studies reported fasting glucose outcomes (Pereira 2002; Pins<br />

2002). In the Pereira study, there was a tendency <strong>for</strong> fasting glucose<br />

to be lower with the wholegrain diet (5.2 +/-(SE) 0.08 mmol/L)<br />

than with the refined grain diet (5.3 +/- (SE) 0.08 mmol/L).<br />

Keenan reported a significant difference in fasting glucose groups<br />

(change score). At baseline, the oats group was 118.4 +/- (SE)<br />

4.1 mg/dL and post treatment 106.1 +/- (SE) 4.2. In the control<br />

group it was 117.1 +/- (SE) 5.2 at baseline and 119.8 +/- (SE) 5.5<br />

at post-treatment.<br />

23


Fasting insulin<br />

Two studies reported fasting insulin (Keenan 2002; Pereira 2002).<br />

In the study by Keenan, plasma insulin increased from 12.2 +/-<br />

(SE) 0.24 uU/mL at baseline in the oats group to 12.8 +/- (SE)<br />

1.6 post-treatment. In the control group, it was 12.9 +/- (SE)<br />

2.5 at baseline and 12.9 +/- (SE) 3.2 post-treatment. In the other<br />

study, fasting insulin was significantly lower during the follow-up<br />

period with the wholegrain (141 +/- (SE) 3.9 pmol/L) than with<br />

the refined grain (156 +/-(SE) 3.9 pmol/L) diet, P = 0.03.<br />

Blood pressure<br />

Three studies reported blood pressure data (Pins 2002; Reynolds<br />

2000; Van Horn 1991). In the Van Horn study, systolic blood<br />

pressure decreased in the oats group from 127.9/80.3 +/- (SD)<br />

13/10 mmHg to 123.5/77.2 +/- (SD) 14/11. In the control group<br />

blood pressure also decreased from 129.1/79.1 +/- (SD) 20/8 to<br />

124.2/77.2 +/- (SD) 13/9. Blood pressure changes were not significant.<br />

In the other two studies SD data was missing.<br />

Insulin resistance<br />

One study reported a homeostasis model <strong>for</strong> insulin resistance<br />

which showed that insulin resistance was lower with the wholegrain<br />

diet (5.4+/-0.18 U) than with the refined grain diet (6.2 +/-<br />

(SE) 0.18 U) (Pereira 2002).<br />

Secondary Outcomes<br />

Of the 10 included studies, six studies reported on side effects<br />

associated with the consumption of wholegrain foods. The other<br />

four studies made no comments about side effects. In general,<br />

there was little difference in side effects between the intervention<br />

and control groups and no serious side effects.<br />

Pins (Pins 2002) administered a 42-question side-effect questionnaire<br />

to participants at the beginning of the baseline phase and at<br />

the end of the intervention. Participants reported the frequency<br />

with which they experienced side effects such as flatulence, loose<br />

stools, headaches, dizziness using a scale of 1 to 5 and values were<br />

tallied across all 42 questions. A final score was assigned to each<br />

participant <strong>for</strong> both time points. Mean scores by group were used<br />

in the analyses. The oatmeal group had a mean side effects score of<br />

58.2+/- 7.2 at baseline and a mean score of 47.6+/- 6.9 post-study.<br />

The control group had a mean side effects score of 56.7 +/- 8.1 at<br />

baseline and a mean score of 53.4 +/- 7.2 post-study. Side effects<br />

decreased by 22% in the intervention group and this change was<br />

significantly different from the change in the control group (at P<br />

< 0.05).<br />

Keenan (Keenan 2002) assessed side-effects by a similar method<br />

to Pins using a 21-question side-effect questionnaire. Mean scores<br />

by group were used in the analyses. The oatmeal group had a<br />

mean side effects score of 29.4+/- 3.4 at baseline and a mean score<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

of 23.3+/- 3.8 post-study. The control group had a mean side<br />

effects score of 28.4 +/- 4.3 at baseline and a mean score of 24.9<br />

+/- 3.7 post-study. There was no statistically significant difference<br />

between the change in the side effects score in the intervention<br />

group and the change in the control group (at P < 0.05).<br />

The study by Pereira (Pereira 2002) included a questionnaire about<br />

44 side effects using a scale of 1 to 5, including gastrointestinal<br />

symptoms and general symptoms of acute or chronic illnesses.<br />

Unusual symptoms were reported more often during the refined<br />

grain than the wholegrain period and the authors of the study attributed<br />

the difference to common upper respiratory illnesses in<br />

two individuals and to chronic low grade abdominal pain in another<br />

subject. There<strong>for</strong>e, the incidence of prescription medication<br />

was higher with the refined grain than with the wholegrain diet.<br />

Dry cough (1.4 +/- 0.13 compared with 1.1 +/- 0.13) and sweating<br />

(1.3 +/- 0.13 compared with 1.0 +/- 0.13) tended to be scored<br />

higher with the refined grain diet. Bowel movements were more<br />

frequent with the wholegrain than the refined grain diet (1.8+/-<br />

0.17 compared with 1.4 +/-0.17 movements/day; P < 0.001). No<br />

significant differences between treatments were reported <strong>for</strong> <strong>heart</strong>burn,<br />

indigestion, diarrhoea, or loose stools.<br />

Johnston (Johnston 1998) reported that five out of the total of 124<br />

participants reported mild gastrointestinal side effects. Three of the<br />

five were in the placebo group (one reported flatus, two reported<br />

increased constipation). Of the two participants in the wholegrain<br />

group who reported gastrointestinal complaints, one reported increased<br />

flatus and one reported increased stool frequency.<br />

Reynolds (Reynolds 2000) reported that both the wholegrain oat<br />

cereal and the control cereal were well tolerated with no serious<br />

side effects. The non serious adverse reactions reported equally in<br />

both groups were mild flatus. No further in<strong>for</strong>mation was given.<br />

Van Horn (Van Horn 1991) reported that there were no reports<br />

of side effects by the participants.<br />

D I S C U S S I O N<br />

While there is growing evidence from observational studies that<br />

wholegrains have benefits <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong>, there is insufficient<br />

evidence available from randomised controlled trials to<br />

make any conclusions about wholegrains in general and <strong>coronary</strong><br />

<strong>heart</strong> <strong>disease</strong>, except <strong>for</strong> wholegrain oats. There is some evidence<br />

from RCTs that wholegrain oats can lower LDL- and total cholesterol<br />

levels in those with pre-existing risk factors <strong>for</strong> CHD. This<br />

effect was seen at 4 weeks <strong>for</strong> total cholesterol so may be effective<br />

even with short interventions. However the results of this review<br />

should be treated with caution because many studies were short<br />

term, underpowered and funded by companies with commercial<br />

interests in wholegrains.<br />

The long-term effects of oats have not been evaluated. Well-designed,<br />

adequately powered studies are needed to evaluate the ef-<br />

24


fects of continued consumption of oats. This review found no<br />

studies on the effect of wholegrains on CHD mortality or events.<br />

There is also a lack of evidence on outcomes other than lipids.<br />

This review found a few studies that reported outcomes other than<br />

lipids such as fasting glucose, fasting insulin, blood pressure and<br />

insulin resistance but there was not enough evidence to make any<br />

conclusions about the effect of wholegrains on these outcomes.<br />

Studies were included in this review if there was a comparison<br />

between a wholegrain food or diet and a diet or food containing<br />

no wholegrains or fewer wholegrains. The review did not aim to<br />

assess the effects of dietary fibre, although the beneficial effects of<br />

wholegrains may well be associated with the dietary fibre content<br />

(Brown 1999). For in<strong>for</strong>mation, the dietary fibre content of the<br />

foods or diets has been reported in the notes section of the Table of<br />

Included Studies and is consistently higher in the wholegrain diets<br />

than the control diets. Many of the studies on oats that have been<br />

included in the review were aimed at assessing the effect of Betaglucan<br />

soluble fibre in oatmeal. It has previously been reported<br />

that the B-glucans in oats can produce a modest reduction in blood<br />

cholesterol concentration in normocholesterolemic subjects and a<br />

greater reduction in hypercholesterolemic subjects (Ripsin 1992).<br />

It is possible that weight loss may account <strong>for</strong> the observed reductions<br />

in lipid levels. While weight data has been reported in<br />

this review, most of the studies had insufficient power to measure<br />

small changes in weight. Also, studies were only included in the<br />

review if the only difference in dietary composition was substitution<br />

of refined grains by wholegrains and there<strong>for</strong>e the macronutrient<br />

contents of the diets were similar. However, it is possible that<br />

small differences in saturated fat intake between the wholegrain<br />

and control diets may account <strong>for</strong> the reduction in lipid levels<br />

found with wholegrain diets.<br />

There is surprisingly little evidence available from controlled trials<br />

about the effects of wholegrain foods and diets other than oats and<br />

there is a need <strong>for</strong> studies in this area. One observation from this<br />

References to studies included in this review<br />

Davidson 1991 {published data only}<br />

Davidson MH, Dugan LD, Burns JH, Bova J, Story K, Drennan<br />

KB. The hypocholesterolemic effects of beta-glucan in oatmeal and<br />

oat bran. JAMA 1991;265(14):1833–9.<br />

Johnston 1998 {published data only}<br />

Johnston L, Reynolds HB, Hunninghake DB, Schultz K,<br />

Westereng B. Cholesterol-lowering benefits of a whole grain oat<br />

ready to eat cereal. Nutrition Clinical Care 1998;1(1):6–12.<br />

Karmally 2005 {published data only}<br />

∗ Karmally W, Montez MG. Cholesterol-lowering benefits of oatcontaining<br />

cereal in Hispanic Americans. Journal of the American<br />

Dietetic Association 2005;105(6):967–70.<br />

R E F E R E N C E S<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

review is that there is some inconsistency in the use of the term<br />

wholegrain. For example, some reviews of wholegrain products<br />

have included studies on dietary fibre or oatbran and have used the<br />

term wholegrain interchangeably. There is a need <strong>for</strong> some clarity<br />

and consistency in this area.<br />

A U T H O R S ’ C O N C L U S I O N S<br />

Implications <strong>for</strong> practice<br />

There is some evidence from controlled trials that wholegrain oats<br />

can reduce LDL cholesterol and total cholesterol risk factors <strong>for</strong><br />

<strong>coronary</strong> <strong>heart</strong> <strong>disease</strong>. However, there is a lack of trials on other<br />

wholegrain foods and diets. Of the trials that met the review criteria,<br />

there were many poor quality studies. Many of the trials identified<br />

were short-term and conducted on small sample sizes. There<br />

is enough evidence <strong>for</strong> wholegrain oats to suggest that healthcare<br />

professionals could recommend oats as part of cholesterol reduction<br />

programmes.<br />

Implications <strong>for</strong> research<br />

There is a need <strong>for</strong> well-designed, adequately powered studies to<br />

evaluate the effect of wholegrains other than oats on <strong>coronary</strong> <strong>heart</strong><br />

<strong>disease</strong> and risk factors. There is a need <strong>for</strong> well designed trials to<br />

evaluate the long-term effects of oats. There is a need <strong>for</strong> clarity<br />

and consistency in the use of the terms wholegrain, bran, germ<br />

and fibre in published research.<br />

A C K N O W L E D G E M E N T S<br />

Margaret Burke, Cochrane Heart Group <strong>for</strong> advice on the search<br />

strategy and <strong>for</strong> assistance with translation of a paper.<br />

Lone Gale <strong>for</strong> assistance with translation of a paper.<br />

Keenan 2002 {published data only}<br />

Keenan JM, Pins JJ, Frazel C, Moran A, Turnquist L. Oat ingestion<br />

reduces systolic and diastolic blood pressure in patients with mild<br />

or borderline hypertension: A pilot trial. The Journal of Family<br />

Practice (Online) www.jfponline.com 2002;51(4):369.<br />

Leinonen 2000 {published data only}<br />

Leinonen KS, Poutanen KS, Mykkanen HM. Rye bread decreases<br />

serum total and LDL cholesterol in men with moderately elevated<br />

serum cholesterol. Journal of Nutrition 2000;130:164–70.<br />

Pereira 2002 {published data only}<br />

∗ Pereira MA, Jacobs JR, Pins JJ, Raatz SK, Gross MD, Slavin JL,<br />

Seaquist ER. Effect of whole grains on insulin sensitivity in<br />

overweight hyperinsulinemic adults. American Journal of Clinical<br />

25


Nutrition 2002;75:848–55.<br />

Pins 2002 {published data only}<br />

Keenan JM, Pins JJ, Geleva D, Frazel C, O’Connor PJ, Cherney<br />

LM. Whole-grain oat cereal consumption reduces antihypertensive<br />

medication need: a cost analysis. Preventive Medicine in Managed<br />

Care 2002;3(1):9–17.<br />

∗ Pins JJ, Geleva DRD, Keenan JM, Frazel C, O’Connor PJ,<br />

Cherney LM. Do whole-grain oat <strong>cereals</strong> reduce the need <strong>for</strong><br />

antihypertensive medications and improve blood pressure control?.<br />

The Journal of Family Practice 2002;51(4):353–9.<br />

Reynolds 2000 {published data only}<br />

Reynolds HR, Quiter E, Hunninghake DB. Whole grain oat cereal<br />

lowers serum lipids. Topics in Clinical Nutrition 2000;15(4):74–83.<br />

Van Horn 1988 {published data only}<br />

Van Horn L, Emidy LA, Liu K, Liao Y, Ballew C, King J, et<br />

al.Serum lipid response to a fat-modified, oatmeal-enhanced diet.<br />

Preventive Medicine 1988;17:377–86.<br />

Van Horn 1991 {published data only}<br />

Van Horn L, Moag-Stahlberg A, Liu K, Ballew C, Ruth K, Hughes<br />

R, et al.Effects on serum lipids of adding instant oats to usual<br />

American diets. American Journal of Public Health 1991;81(2):<br />

183–8.<br />

References to studies excluded from this review<br />

Anderson 1978 {published data only}<br />

Anderson JW, Ward K. Long-term effects of high-carbohydrate,<br />

high-fiber diets on glucose and lipid metabolism: A preliminary<br />

report on patients with diabetes. Diabetes Care 1978;1(2):77–82.<br />

Anderson 1979 {published data only}<br />

Anderson JW, Ward KW. High-carbohydrate, high-fiber diets <strong>for</strong><br />

insulin-treated men with diabetes mellitus. American Journal of<br />

Clinical Nutrition. 1979;32:2312–21.<br />

Asp 1981 {published data only}<br />

Asp N-G, Agardh C-D, Ahren B, Dencker I, Johansson, C-G, et<br />

al.Dietary fibre in Type II diabetes. Acta. Med. Scand. 1981;Suppl<br />

656:47–50.<br />

Behall 2004a {published data only}<br />

Behall KM, Scholfield DJ, Hallfrisch J. Diets containing barley<br />

significantly reduce lipids in mildly hypercholesterolemic men and<br />

women. American Journal of Clinical Nutrition 2004;80(5):<br />

1185–93.<br />

Behall 2004b {published data only}<br />

Behall KM, Scholfield DJ, Hallfrisch J. Lipids significantly reduced<br />

by diets containing barley in moderately hypercholesterolemic men.<br />

Journal of the American College of Nutrition 2004;23(1):55–62.<br />

Birkeland 1991 {published data only}<br />

Birkeland KI, Gullestad L, Torsvik H. Cholesterol-lowering effect of<br />

oats. Tidsskrift <strong>for</strong> Den Norske Laege<strong>for</strong>ening 1991;111(17):2081–5.<br />

Birketvedt 2000 {published data only}<br />

Birketvedt GS, Auseth J, Florholmen JR, Ryttig K. Long term effect<br />

of fibre supplement and reduced energy intake on body weight and<br />

blood lipids in overweight subjects. Acta Medica 2000;43(3):<br />

129–132.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Booyens 1966 {published data only}<br />

Booyens J, De Waal VM, Rademeyer LJ. The effect of dietary maize<br />

meal supplementation on the levels of serum cholesterol and<br />

magnesium. South African Medical Journal. 1966;40(11):237–9.<br />

Bourdon 1999 {published data only}<br />

Bourdon I, Yokoyama W, Davis P, Hudson C, Backus R, Richter D,<br />

et al.Postprandial lipid, glucose, insulin, and cholecystokinin<br />

responses in men fed barley pasta enriched with beta-glucan.<br />

American Journal of Clinical Nutrition 1999;69:55–63.<br />

Bruttomesso 1989 {published data only}<br />

Bruttomesso D, Briani G, Bilardo G, Vitale E, Lavagnini T,<br />

Marescotti C, et al.The medium-term effect of natural or extractive<br />

dietary fibres on plasma amino acids and lipids in type 1 diabetics.<br />

Diabetes Research and Clinical Practice 1989;6:149–55.<br />

Burr 1989 {published data only}<br />

Burr ML, Gilbert JF, Holliday RM, Elwood PC, Fehily AM, Rogers<br />

S, et al.Effects of changes in fat, fish, and fibre intakes on death and<br />

myocardial reinfarction: Diet and reinfarction trial (DART). Lancet<br />

1989;2:757–61.<br />

Cairella 1995 {published data only}<br />

Cairella G, Cairella M, Marchini G. Effect of dietary fibre on<br />

weight correction after modified fasting. European Journal of<br />

Clinical Nutrition 1995;49(Suppl 3):S325–7.<br />

Chandalia 2000 {published data only}<br />

Chandalia M, Garg A, Lutjohann D, Von Bergman K, Grundy SM,<br />

Brinkley LJ. Beneficial effects of high dietary fiber intake in patients<br />

with type 2 diabetes mellitus. New England Journal of Medicine<br />

2000;342:1392–8.<br />

Collier 1982 {published data only}<br />

Collier G, O’Dea K. Effect of physical <strong>for</strong>m of carbohydrate on the<br />

postprandial glucose, insulin, and gastric inhibitory responses in<br />

type 2 diabetes. American Journal of Clinical Nutrition 1982;36:<br />

10–14.<br />

Comi 1995 {published data only}<br />

Comi D, Brugnani M, Gianino A. Metabolic effects of hypocaloric<br />

high-carbohydrate/high-fibre diet in non-insulin dependent<br />

diabetic patients. European Journal of Clinical Nutrition 1995;49<br />

(Suppl 3):S242–4.<br />

Crapo 1981 {published data only}<br />

Crapo PA, Insel J, Sperling M, Kolterman OG. Comparison of<br />

serum glucose, insulin, and glucagon responses to different types of<br />

complex carbohydrate in noninsulin-dependent diabetic patients.<br />

American Journal of Clinical Nutrition 1981;34:184–90.<br />

Data 1980 {published data only}<br />

Data PG, Cacchio M, Sergiacomo P, Di Tano G. Investigation of<br />

the importance of dietetic fibers in the regulation of<br />

cholesterolemia. Bollettino della Societa Italiana di Biologia<br />

Sperimentale 1980;16:1545–50.<br />

Davy 2002a {published data only}<br />

Davy BM, Melby CL, Beske SD, Ho RC, Davrath LR, Davy KP.<br />

Oat consumption does not affect resting casual and ambulatory 24h<br />

arterial blood pressure in men with high-normal blood pressure to<br />

stage I hypertension. Journal of Nutrition. 2002;132(3):394–8.<br />

26


Davy 2002b {published data only}<br />

Davy BM, Davy KP, Ho RC, Beske SD, Davrath LR, Melby CL.<br />

High-fiber oat cereal compared with wheat cereal consumption<br />

favorably alters LDL-cholesterol subclass and particle numbers in<br />

middle-aged and older men. American Journal of Clinical Nutrition<br />

2002;76:351–8.<br />

Eliasson 1992 {published data only}<br />

Eliasson K, Ryttig KR, Hylander B, Rossner S. A dietary fibre<br />

supplement in the treatment of mild hypertension. A randomized,<br />

double-blind, placebo-controlled trial. Journal of Hypertension<br />

1992;10:195–9.<br />

Fehily 1986 {published data only}<br />

Fehily AM, Burr M, Butland BK, Eastham RD. A randomised<br />

controlled trial to investigate the effect of a high fibre diet on blood<br />

pressure and plasma fibrinogen. Journal of Epidemiology &<br />

Community Health. 1986;40:334–7.<br />

Fischer 2000 {published data only}<br />

Fischer J. Fiber-rich food improves the day-time blood glucose<br />

profile in patients with overweight and diabetes mellitus type 2.<br />

Schweizerische Rundschau fur Medizin Praxis 2000;89:1975–6.<br />

Fordyce-Baum 1989 {published data only}<br />

Fordyce-Baum MK, Langer L, Mantero-Atienza E, Crass R, Beach<br />

RS. Use of an expanded whole-wheat product in the reduction of<br />

body weight and serum lipids in obese females. American Journal of<br />

Clinical Nutrition 1989;50(1):30–6.<br />

Fraser 1981 {published data only}<br />

Fraser GE, Jacobs DR, Anderson JT, Foster N, Palta M, Blackburn<br />

H. The effect of various vegetable supplements on serum<br />

cholesterol. American Journal of Clinical Nutrition 1981;34:<br />

1272–7.<br />

Fung 2002 {published data only}<br />

Fung TT, Hu FB, Pereira MA, Liu S, Stampfer MJ, Colditz GA, et<br />

al.Whole-grain intake and the risk of type 2 diabetes: A prospective<br />

study in men. American Journal of Clinical Nutrition 2002;76:<br />

535–40.<br />

Golay 1992 {published data only}<br />

Golay A, Koellreutter B, Bloise D, Assal J-P, Wursch P. The effect of<br />

muesli or cornflakes at breakfast on carbohydrate metabolism in<br />

type 2 diabetic patients. Diabetes Research & Clinical Practice.<br />

1992;15:135–42.<br />

Guzic 1994 {published data only}<br />

Guzic B, Sundell IB, Keber I, Keber D. The effect of oat husk<br />

supplementation in diet on plasminogen activator inhibitor type 1<br />

in diabetic survivors of myocardial infarction. Fibrinolysis 1994;8<br />

(Suppl 2):44–6.<br />

Hagander 1985 {published data only}<br />

Hagander B, Bjorck I, Asp N-G, Lunquist I, Nilsson-Ehle P,<br />

Schrezenmeir J, et al.Hormonal and metabolic responses to<br />

breakfast meals in NIDDM: Comparison of white and whole-grain<br />

wheat bread and corresponding extruded products. Human<br />

Nutrition Applied Nutrition 1985;39A:114–23.<br />

Hagander 1988 {published data only}<br />

Hagander B, Asp N-G, Efendic S, Nilsson-Ehle P, Schersten B.<br />

Dietary fiber decreases fasting blood glucose levels and plasma LDL<br />

concentration in non-insulin-dependent diabetes mellitus patients.<br />

American Journal of Clinical Nutrition 1988;47:852–8.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

He 1995 {published data only}<br />

He J, Klag MJ, Whelton PK, Mo J-P, Qian M-C, Mo P-S, et<br />

al.Oats and buckwheat intakes and cardiovascular <strong>disease</strong> risk<br />

factors in an ethnic minority of China. American Journal of Clinical<br />

Nutrition 1995;61:366–72.<br />

Heaton 1976 {published data only}<br />

Heaton KW, Manning AP, Hartog M. Lack of effect on blood lipid<br />

and calcium concentrations of young men on changing from white<br />

to wholemeal bread. British Journal of Nutrition. 1976;35:55–60.<br />

Hoffman 1982 {published data only}<br />

Hoffman CR, Fineberg SE, Howey DC, Clark CM, Pronsky Z.<br />

Short-term effects of a high-fiber, high carbohydrate diet in very<br />

obese diabetic individuals. Diabetes Care 1982;5(6):605–11.<br />

Hollenbeck 1986 {published data only}<br />

Hollenbeck CB, Coulston AM, Reaven GM. To what extent does<br />

increased dietary fiber improve glucose and lipid metabolism in<br />

patients with noninsulin-dependent diabetes mellitus (NIDDM)?.<br />

American Journal of Clinical Nutrition. 1986;43:16–24.<br />

Hunninghake 1994 {published data only}<br />

Hunninghake DB, Miller VT, LaRosa JC, Kinosian B, Brown V,<br />

Howard WJ, et al.Hypocholesterolemic effects of a dietary fiber<br />

supplement. American Journal of Clinical Nutrition. 1994;59:<br />

1050–4.<br />

Jacobs 2002 {published data only}<br />

Jacobs DR, Pereira MA, Stumpf K, Pins JJ, Adlercreutz H. Whole<br />

grain food intake elevates serum enterolactone. British Journal of<br />

Nutrition 2002;88:111–6.<br />

Jang 2001 {published data only}<br />

Jang Y, Lee JH, Kim OY, Park HY, Lee SY. Consumption of whole<br />

grain and legume powder reduces insulin demand, lipid<br />

peroxidation, and plasma homocysteine concentrations in patients<br />

with <strong>coronary</strong> artery <strong>disease</strong>: Randomized controlled clinical trial.<br />

Arteriosclerosis Thrombosis & Vascular Biology 2001;21:2065–71.<br />

Jenkins 1985 {published data only}<br />

Jenkins DJA, Wolever TMS, Kalmusky J, Giudici S, Giordano C,<br />

Wong GS, et al.Low glycemic index carbohydrate foods in the<br />

management of hyperlipidemia. American Journal of Clinical<br />

Nutrition 1985;42:604–17.<br />

Jenkins 1993 {published data only}<br />

Jenkins D, Wolever T, Rao V, Hegele RA, Mitchell SJ, Ransom T,<br />

et al.Effect on blood lipids of very high intakes of fiber in diets low<br />

in saturated fat and cholesterol. New England Journal of Medicine<br />

1993;329:21–6.<br />

Judd 1981 {published data only}<br />

Judd PA, Truswell AS. The effect of rolled oats on blood lipids and<br />

fecal steroid excretion in man. American Journal of Clinical<br />

Nutrition 1981;34:2061–7.<br />

Juntunen 2002 {published data only}<br />

Juntunen KS, Niskanen LK, Liukkonen KH, Poutanen KS, Holst<br />

JJ, Mykkanen HM. Postprandial glucose, insulin, and incretin<br />

responses to grain products in healthy subjects. American Journal of<br />

Clinical Nutrition 2002;75:254–62.<br />

Juntunen 2003 {published data only}<br />

Juntunen KS, Laaksonen DE, Poutanen KS, Niskanen LK,<br />

Mykkanen HM. High-fiber rye bread and insulin secretion and<br />

27


sensitivity in healthy menopausal women. American Journal of<br />

Clinical Nutrition 2003;77:385–91.<br />

Kabir 2002 {published data only}<br />

Kabir M, Oppert J-M, Vidal H, Bruzzo F, Fiquet C, Wursch P,<br />

Slama G, Rizkalla SW. Four-week low-glycemic index breakfast<br />

with a modest amount of soluble fibers in type 2 diabetic men.<br />

Metabolism: Clinical & Experimental 2002;51(7):819–26.<br />

Karlstrom 1984 {published data only}<br />

Kalstrom B, Vessby B, Asp N-G, Boberg M, Gustafsson I-B, Lithell<br />

H, et al.Effects of increased content of cereal fibre in the diet of<br />

Type 2 (non-insulin-dependent) diabetic patients. Diabetologia<br />

1984;26:272–7.<br />

Katz 2001a {published data only}<br />

Katz DL, Nawaz H, Boukhalil J, Chan W, Ahmadi R, Giannamore<br />

V, et al.Effects of oat and wheat <strong>cereals</strong> on endothelial responses.<br />

Preventive Medicine. 2001;33:476–84.<br />

Katz 2001b {published data only}<br />

Katz DL, Nawaz H, Boukhalil J, Giannamore V, Chan W, Ahmadi<br />

R, et al.Acute effects of oats and vitamin E on endothelial responses<br />

to ingested fat. American Journal of Preventive Medicine 2001;20(2):<br />

124–9.<br />

Kay 1977 {published data only}<br />

Kay RM, Truswell AS. The effect of wheat fibre on plasma lipids<br />

and faecal steroid excretion in man. British Journal of Nutrition<br />

1977;37:227–35.<br />

Kay 1981 {published data only}<br />

Kay RM, Grobin W, Track NS. Diets rich in natural fibre improve<br />

carbohydrate tolerance in maturity-onset, non-insulin dependent<br />

diabetics. Diabetologia 1981;20:18–21.<br />

Kesaniemi 1990 {published data only}<br />

Kesaniemi YA, Tarpila S, Miettinen TA. Low vs high dietary fiber<br />

and serum, biliary, and fecal lipids in middle-aged men. American<br />

Journal of Clinical Nutrition. 1990;51:1007–12.<br />

Kleemola 1999 {published data only}<br />

Kleemola P, Puska P, Vartiainnen, Roos E, Luoto R, Ehnholm C.<br />

The effect of breakfast cereal on diet and serum cholesterol: A<br />

randomized trial in North Karelia, Finland. European Journal of<br />

Clinical Nutrition 1999;53:716–21.<br />

Kris-Etherton 2002 {published data only}<br />

Kris-Etherton PM, Shaffer-Taylor D, Smicklas-Wright H, Mitchell<br />

DC, Bekhuis TC, Olson BH, et al.High-soluble-fiber foods in<br />

conjunction with a telephone-based, personalized behavior change<br />

support service result in favorable changes in lipids and lifestyles<br />

after 7 weeks. Journal of the American Dietetic Association. 2001;<br />

102:503–10.<br />

Lakshmi 1996 {published data only}<br />

Lakshmi KB, Vimala V. Hypoglycemic effect of selected sorghum<br />

recipes. Nutrition Research 1996;16(10):1651–58.<br />

Leinonen 1999 {published data only}<br />

Leinonen K, Liukkonen K, Poutanene K, Uusitupa M, Mykkanen<br />

H. Rye bread decreases postprandial insulin response but does not<br />

alter glucose response in healthy Finnish subjects. European Journal<br />

of Clinical Nutrition 1999;53(4):262–7.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Liese 2003 {published data only}<br />

Liese AD, Roach AK, Sparks KC, Marquart L, D’Agostino RB,<br />

Mayer-Davis E. Whole-grain intake and insulin sensitivity: the<br />

Insulin Resistance Atherosclerosis Study. American Journal of<br />

Clinical Nutrition 2003;78(5):965–71.<br />

Lousley 1984 {published data only}<br />

Lousley SE, Jones DB, Slaughter P, Carter RD, Jelfs R, Mann JI.<br />

High carbohydrate-high fibre diets in poorly controlled diabetes.<br />

Diabetic Medicine 1984;1(1):21–5.<br />

MacMahon 1998 {published data only}<br />

Macmahon M, Carless J. Ispaghula husk in the treatment of<br />

hypercholesterolaemia: a double blind controlled study. Journal of<br />

Cardiovascular Risk 1998;5(3):167–72.<br />

Manhire 1981 {published data only}<br />

Manhire A, Henry CL, Hartog M, Heaton KW. Unrefined<br />

carbohydrate and dietary fibre in treatment of diabetes mellitus.<br />

Journal of Human Nutrition 1981;35:99–101.<br />

Mathur 1968 {published data only}<br />

Mathur KS, Khan MA, Sharma RD. Hypocholesterolaemic effect<br />

of Bengal gram: a long-term study in man. British Medical Journal.<br />

1968;1:30–1.<br />

McIntosh 1991 {published data only}<br />

McIntosh GH, Whyte J, McArthur R, Nestel PJ. Barley and wheat<br />

foods: influence on plasma cholesterol concentrations in<br />

hypercholesterolemic men. American Journal of Clinical Nutrition<br />

1991;53:1205–9.<br />

Montonen 2003 {published data only}<br />

Montonen J, Knekt P, Jarvinen R, Aromaa A, Reunanen A. <strong>Wholegrain</strong><br />

and fiber intake and the incidence of type 2 diabetes.<br />

American Journal of Clinical Nutrition 2003;77:622–9.<br />

Nielsen 1988 {published data only}<br />

Nielsen GL, Thuesen H. Blood glucose responses to sweetcorn and<br />

potato meals. Diabetic Medicine 1988;5(6):598–9.<br />

O’Kell 1988 {published data only}<br />

O’Kell RT, Duston AA. Lack of effect of dietary oats on serum<br />

cholesterol. Missouri Medicine 1988;85(11):726–8.<br />

Odes 1993 {published data only}<br />

Odes HS, Lazovski H, Stern I, Madar Z. Double-blind trial of a<br />

high dietary fiber, mixed grain cereal in patients with chronic<br />

constipation and hyperlipidemia. Nutrition Research 1993;13:<br />

979–85.<br />

Pacy 1986 {published data only}<br />

Pacy PJ, Dodson PM, Taylor MP. The effect of a high fibre, low fat,<br />

low sodium diet on diabetics with intermittent claudication. British<br />

Journal of Clinical Practice 1986;40(8):313–7.<br />

Poulter 1993 {published data only}<br />

Poulter N, Chang CL, Cuff A, Poulter C, Sever P, Thom S. Lipid<br />

profiles after the daily consumption of an oat-based cereal: a<br />

controlled crossover trial. American Journal of Clinical Nutrition<br />

1993;58:66–9.<br />

Reynolds 1989 {published data only}<br />

Reynolds HR, Lindeke E, Hunninghake DB. Effect of oat bran on<br />

serum lipids. Journal of the American Dietetic Association 1989;89<br />

(Suppl):A112.<br />

28


Rigaud 1990 {published data only}<br />

Rigaud D, Ryttig KR, Angel LA, Apfelbaum M. Overweight<br />

treated with energy restriction and a dietary fibre supplement: a 6month<br />

randomized, double-blind, placebo-controlled trial.<br />

International Journal of Obesity. 1990;14:763–9.<br />

Roth 1985 {published data only}<br />

Roth G, Leitzmann C. Long-term influence of breakfast <strong>cereals</strong> rich<br />

in dietary fibres on human blood lipid values [Langzeieinfluss<br />

ballast–stoffreicher Fruhstuck–cerealien auf die Blutlipide beim<br />

Menschen]. Aktuelle Ernahringsmedizin Klinik und Praxis 1985;10<br />

(3):106–9.<br />

Russ 1985 {published data only}<br />

Russ CS, Atkinson RL. Use of high fiber diets <strong>for</strong> the outpatient<br />

treatment of obesity. Nutrition Reports International 1985;32(1):<br />

193–8.<br />

Rytter 1996 {published data only}<br />

Rytter E, Erlanson-Albertsson C, Lindahl L, Lundquist I, Viberg U,<br />

Akesson B, et al.Changes in plasma insulin, enterostatin, and<br />

lipoprotein levels during an energy-restricted dietary regimen<br />

including a new oat-based liquid food. Annals of Nutrition &<br />

Metabolism. 1996;40:212–20.<br />

Saltzman 2001a {published data only}<br />

Saltzman E, Moriguti JC, Das SK, Corrales A, Fuss P, Greenberg<br />

AS, et al.Effects of a cereal rich in soluble fiber on body composition<br />

and dietary compliance during consumption of a hypocaloric diet.<br />

Journal of the American College of Nutrition 2001;20(1):50–7.<br />

Saltzman 2001b {published data only}<br />

Saltzman E, Das SK, Lichtenstein AH, Dallal GE, Corrales A,<br />

Schaefer EJ, et al.An oat-containing hypocaloric diet reduces<br />

systolic blood pressure and improves lipid profile beyond effects of<br />

weight loss in men and women. Journal of Nutrition. 2001;131:<br />

1465–70.<br />

Schlamowitz 1987 {published data only}<br />

Schlamowitz P, Halberg T, Warnoe O, Wilstrup F, Ryttig K.<br />

Treatment of mild to moderate hypertension with dietary fibre.<br />

Lancet. 1987;2:622–3.<br />

Turnbull 1987 {published data only}<br />

Turnbull WH, Leeds AR. Reduction of total and LDL-C<br />

cholesterol in plasma by oats. Journal of Clinical Nutrition<br />

Gastroenterology 1987;2:177–81.<br />

Turnbull 1989 {published data only}<br />

Turnbull WH, Leeds AR. The effect of rolled oats and a reduced/<br />

modified fat diet on apolipoproteins A1 and B. J Clin Nutr<br />

Gastroenterol 1989;1:15–19.<br />

Turpeinen 2000 {published data only}<br />

Turpeinen AM, Juntunen K, Mutanen M, Mykkanen H. Similar<br />

responses in hemostatic factors after consumption of wholemeal rye<br />

bread and low-fiber wheat bread. European Journal of Clinical<br />

Nutrition 2000;54:418–23.<br />

Van Horn 1986 {published data only}<br />

Van Horn LV, Liu K, Parker D, Emidy L, Liao Y, Pan WH, et<br />

al.Serum lipid response to oat product intake with a fat-modified<br />

diet. Journal of the American Dietetic Association 1986;86(6):<br />

759–64.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

Van Horn 2001 {published data only}<br />

Van Horn L, Liu K, Gerber J, Garside D, Schiffer L, Gernhofer N,<br />

et al.Oats and soy in lipid-lowering diets <strong>for</strong> women with<br />

hypercholesterolemia: is there synergy?. Journal of the American<br />

Dietetic Association. 2001;101:1319–1325.<br />

Willms 1987 {published data only}<br />

Willms B, Arends J. Comparison of isolated (guar) and natural<br />

(Musli) dietary fiber in the treatment of type II diabetes [Verglech<br />

von isolierten (Guar) und naturlichen (Musli) Ballastoffen in der<br />

therapie des Typ–II–Diabetes]. Medizinische Klinik. 1987;12/13:<br />

429–31.<br />

Wolever 2003 {published data only}<br />

Wolever TMS, Tshilias EB, McBurney MI, Le N-A. Long-term<br />

effect of reduced carbohydrate or increased fiber intake on LDL<br />

particle size and HDL composition in subjects with type 2 diabetes.<br />

Nutrition Research 2003;23:15–26.<br />

Wolffenbuttel 1992 {published data only}<br />

Wolffenbuttel BHR, Sels J-PJE, Heesen BJ, Menheere PPCA,<br />

Nieuwenhuijzen-Kruseman AC. The effects of dietary fibre and<br />

insulin treatment on the serum levels of lipids and lipoprotein (a) in<br />

patients with diabetes mellitus type II. Nederlands Tijdschrift voor<br />

Geneeskunde 1992;136(15):739–42.<br />

Wursch 1991 {published data only}<br />

Wursch P, Koellreutter B, Haesler E, Felber JP, Golay A. Metabolic<br />

effects of slow release starch in non-insulin dependent diabetic<br />

patients. Diabetes, Nutrition & Metabolism 1991;4(3):195–9.<br />

References to ongoing studies<br />

FSA 2005a {unpublished data only}<br />

Food Standards Agency. FSA research project list. Available at:<br />

http://www.food.gov.uk/multimedia/pdfs/reslistnov05.pdf<br />

[accessed 13 2 2007] 1995.<br />

FSA 2005b {unpublished data only}<br />

Food Standards Agency. FSA research projects list. Available at:<br />

http://www.food.gov.uk/multimedia/pdfs/reslistnov05.pdf<br />

[accessed 13 2 2007] 2005.<br />

Additional references<br />

BHF 2004<br />

Petersen S, Peto V, Rayner M. Coronary <strong>heart</strong> <strong>disease</strong> statistics.<br />

London: British Heart Foundation, 2004.<br />

Brown 1999<br />

Brown L, Rosner B, Willett WW, Sacks FM. Cholesterol-lowering<br />

effects of dietary fibre: a meta-analysis. American Journal of Clinical<br />

Nutrition 1999;69:30–42.<br />

Carter 1999<br />

Carter M, Moser K, Kelly S. Health of older people: <strong>disease</strong><br />

prevalence, prescription and referral rates, England and Wales<br />

1996. Health Statistics Quarterly 1999;Winter:9–15.<br />

Dickersin 1994<br />

Dickersin K, Scherer R, Lefebvre C. Identifying relevant studies <strong>for</strong><br />

systematic reviews. British Medical Journal 1994;309:1286–91.<br />

FDA 1999<br />

FDA (US Food, Drug Administration). Health claim notification<br />

<strong>for</strong> wholegrain foods. Available at: http://www.cfsan.fda.gov/~dms/<br />

flgrains.html [accessed 13 2 2007] 1999.<br />

29


Feranninni 1991<br />

Ferranninni E, Haffner SM, Mitchell BD, Stern MP.<br />

Hyperinsulinaemia: the key feature of a cardiovascular and<br />

metabolic syndrome. Diabetologia 1991;34:416–22.<br />

FSANZ 2004<br />

Food Standards Australia New Zealand (FSANZ). Draft assessment<br />

report. Definition of wholegrain. Application A464. Available at:<br />

http://www.foodstandards.gov.au/˙srcfiles/<br />

A464˙<strong>Wholegrain</strong>˙DAR%20˙FINAL.pdf#search=%22wholegrain%22<br />

[accessed 13 2 2007] 2004.<br />

Higgins 2003<br />

Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring<br />

inconsistency in meta-analyses. BMJ 2003;327(7414):557–60.<br />

Higgins 2005<br />

Higgins JPT, Green S, editors. Cochrane Handbook <strong>for</strong> Systematic<br />

<strong>Review</strong>s of Interventions 4.2.5 [updated May 2005]. The Cochrane<br />

Library, Issue 3, 2005. Chichester, UK: John Wiley & Sons, Ltd.<br />

Jacobs 1998<br />

Jacobs DR, Meyer KA, Kushi LH, Folsom AR. Whole grain intake<br />

may reduce the risk of ischemic <strong>heart</strong> <strong>disease</strong> death in<br />

postmenopausal women: The Iowa Women’s Health Study.<br />

American Journal of Clinical Nutrition 1998;68:248–57.<br />

Jacobs 1999<br />

Jacobs DR, Meyer KA, Kushi LH, Folsom AR. Is whole-grain<br />

intake associated with reduced total and cause-specific death rates<br />

in older women? the Iowa women’s health study. American Journal<br />

of Clinical Nutrition 1999;89:322–9.<br />

JAMA 2001<br />

JAMA. Author instructions: Systeme International (SI) conversion<br />

factors <strong>for</strong> selected laboratory components. Available at: http://<br />

jama.ama-assn.org/misc/auinst˙si.dtl [accessed 1 8 2005] 2001.<br />

Liu 1999<br />

Liu S, Stampfer MJ, Hu FB, Giovannucci E, Rimm E, Manson J, et<br />

al.Whole-grain consumption and risk of <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong>:<br />

results from the Nurses Health Study. American Journal of Clinical<br />

Nutrition 1999;70:412–9.<br />

Liu 2000<br />

Liu S, Manson JE, Stampfer MJ, Hu FB, Giovannucci EMD,<br />

Colditz GA, et al.A prospective study of whole-grain intake and risk<br />

of type 2 diabetes mellitus in US women. American Journal of<br />

Public Health 2000;90(9):1409–15.<br />

McKeown 2002<br />

McKeown NM, Bessesen DH, Hamman RF. Whole-grain intake is<br />

inversely associated with metabolic risk factors <strong>for</strong> type 2 diabetes<br />

and cardiovascular <strong>disease</strong> in the Framingham Offspring study.<br />

American Journal of Clinical Nutrition 2002;76:390–8.<br />

Moher 1999<br />

Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF.<br />

Improving the quality of reports of meta-analyses of randomised<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

controlled trials: the Quorom Statement. Lancet 1999;354:<br />

1896–1900.<br />

Murray 1997<br />

Murray CLJ, Lopez AD. Mortality by cause <strong>for</strong> eight regions of the<br />

world: Global burden of <strong>disease</strong>. Lancet 1997;349:1269–76.<br />

OHE 1999<br />

Office of Health Economics. OHE compendium of health statistics.<br />

11th Edition. London: Office of Health Economics, 1999.<br />

Reaven 1993<br />

Reaven GM. Role of insulin resistance in human <strong>disease</strong> (syndrome<br />

X): an expanded definition. Annual <strong>Review</strong> of Medicine 1993;44:<br />

121–31.<br />

Rimm 1996<br />

Rimm EB, Ascherio A, Giovannucci E, Spiegelman D, Stampfer M,<br />

Willet W. Vegetable, fruit and cereal fibre intake and risk of<br />

<strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> among men. Journal of the American Medical<br />

Association 1996;59:1386–94.<br />

Ripsin 1992<br />

Ripsin CM, Keenan JM, Jacobs DR, Elmer PJ, Welch RR, Van<br />

Horn L, et al.Oat products and lipid-lowering. JAMA 1992;267:<br />

3317–25.<br />

Schulz 1995<br />

Schulz KF, Chalmers I, Hays RJ, Altman DG. Empirical evidence<br />

of bias. Dimensions of methodological quality associated with<br />

estimates of treatment effects in controlled trials. JAMA 1995;273:<br />

408–12.<br />

Seal 2006<br />

Seal CJ. <strong>Wholegrain</strong>s and CVD risk. Proceedings of the Nutrition<br />

Society 2006;65:1–12.<br />

Slavin 2001<br />

Slavin JL, Jacobs D, Marquart L, Wiemer K. The role of whole<br />

grains in <strong>disease</strong> prevention. Journal of the American Dietetic<br />

Association 2001;101:780–5.<br />

Slavin 2003<br />

Slavin J. Why wholegrains are protective: biological mechanisms.<br />

Proceedings of the Nutrition Society 2003;62:129–34.<br />

Steffen 2003<br />

Steffen LM, Jacobs DR, Stevens J, Shahar E, Carithers T, Folsom A.<br />

Associations of whole-grain, refined grain, and fruit and vegetable<br />

consumption with risks of all-cause mortality and incident <strong>coronary</strong><br />

artery <strong>disease</strong> and ischemic stroke: the Atherosclerosis Risk in<br />

Communities (ARIC) study. American Journal of Clinical Nutrition<br />

2003;78:383–90.<br />

Wolk 1999<br />

Wolk A, Manson JE, Stampfer MJ, Colditz GA. Long-term intake<br />

of dietary fiber and increased risk of <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> among<br />

women. JAMA 1999;281:1998–2004.<br />

∗ Indicates the major publication <strong>for</strong> the study<br />

30


C H A R A C T E R I S T I C S O F S T U D I E S<br />

Characteristics of included studies [ordered by study ID]<br />

Davidson 1991<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear - reported to be a single-blind study but not clear who was blinded.<br />

Length of intervention: 6 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Reported but inadequate (only 148 of 156 patients randomised included in<br />

ITT analysis).<br />

Participants Diagnosis of CHD or risk factors: Hypercholesterolemic: LDL cholesterol >4.14 mmol/L or between 3.37<br />

and 4.14 mmol/L with multiple risk factors.<br />

Exclusion criteria: LDL cholesterol outwith the above values, women who were pregnant or lactating,<br />

subjects weighing more than 150% of their ideal body weight or with serious metabolic disorders, including<br />

alcoholism and diabetes.<br />

Medications used: No lipid-lowering medication used <strong>for</strong> 6 weeks prior to trial.<br />

Smoking status: Not reported.<br />

n = 156 randomised to 7 groups, 148 completed. For this review only those in the OM-28 oatmeal group<br />

have been included: OM-28 n = 20 (baseline and follow-up); control: n = 15 (baseline and follow-up)<br />

Mean age: 51.1 years (no SD).<br />

Male/female ratio: 7/13<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

Interventions Intervention: This study was a dose-response study but only the OM-28 group (28g oatmeal group) has<br />

been included in meta-analysis <strong>for</strong> this review as a direct w/w comparison with the control.<br />

The intervention was a low fat, low cholesterol diet plus 28 g of dry oatmeal daily consumed as hot cereal,<br />

muffins or shakes as a substitute <strong>for</strong> matching servings of complex carbohydrates in the diet. (8 week runin<br />

on low-fat, low cholesterol diet only prior to intervention).<br />

Control: Low fat, low cholesterol diet as above replacing oatmeal with 28g of dry farina placebo.<br />

Assessment of dietary compliance: 4-day food records<br />

Comparison of total energy and macronutrient intakes: Reported<br />

Outcomes Main outcomes: Total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides<br />

Other outcomes: None<br />

Notes Dietary fibre intake - the study reports that a significant group difference and a time-by group interaction<br />

were detected <strong>for</strong> total dietary fibre, insoluble fibre and water-soluble fibre. This study was a dose response<br />

study and reports that the differences were expected and proportionally responded to the supplementation<br />

of oat cereal and beta-glucan. However, no specific in<strong>for</strong>mation as to whether differences in fibre content<br />

were significant between the OM-28 group and control.<br />

Risk of bias<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

31


Davidson 1991 (Continued)<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Johnston 1998<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear.<br />

Blinding of participants and cereal dispensers may have been done but unclear from published details.<br />

Study participants and the providers of the <strong>cereals</strong> were blinded.<br />

Length of intervention: 6 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant<br />

Protection against contamination: Unclear<br />

Intention-to-treat analysis: Not done<br />

Participants Diagnosis of CHD or risk factors: Mild to moderate primary hypercholesterolemia. LDL cholesterol 3.37<br />

- 4.9 mmol/L prior to study.<br />

Exclusion criteria: LDL cholesterol outwith the above values; participants outwith 40-70 years old; baseline<br />

triglycerides >3.39 mmol/L; body weight > 140% of ideal; history of major surgery in the previous 3<br />

months; clinically significant metabolic, renal, hepatic, gastrointestinal, pulmonary, hemapoietic, thyroid<br />

or cardiovascular <strong>disease</strong>s; history of allergic reactions to corn, wheat, oats or rice products; subjects taking<br />

certain corticosteroids, androgens or lipid-lowering drugs.<br />

Medications used: Subjects taking some drugs potentially affecting lipids i.e. estrogen, estrogen/progestin,<br />

thiazide diuretics, beta blockers and thyroid hormones were allowed as long as they were on stable doses.<br />

Smoking status: Remained constant throughout the study n = 135 randomised to 2 groups, 124 completed.<br />

Control group: n = 62 baseline, n = 62 end. Intervention group (Cheerios): n = 62 baseline, n = 62 end.<br />

Mean age: control group 57.3 years (no SD); intervention group 56.7 years.<br />

Male/female ratio: 38/24 (control); 40/22 (intervention).<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

Interventions Intervention: <strong>Wholegrain</strong> ready-to-eat cereal (Cheerios) composed of wholegrain oat flour 3 ounces daily<br />

as part of ’Step One’ diet (6 week run-in on ’Step One’ diet prior to intervention).<br />

Control: Commercial cornflakes 3 ounces daily as part of ’Step One’ diet.<br />

Assessment of dietary compliance: subject interviews, daily log of cereal intake, counting returned unopened<br />

boxes of cereal.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, triglycerides, HDL cholesterol, LDL cholesterol, body weight.<br />

Other outcomes: Apo A-1, Apo B, participants report of adverse reactions.<br />

Notes Dietary fibre intake - the study reports that minor but statistically significant differences between control<br />

and treatment groups were detected <strong>for</strong> posttreatment intake of total dietary fibre (20.2 g control; 24.8 g<br />

treatment P


Johnston 1998 (Continued)<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Karmally 2005<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear.<br />

Length of intervention: 6 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Unclear.<br />

Participants Diagnosis of CHD or risk factors: Mild to moderate primary hypercholesterolemia. LDL-cholesterol 3.1<br />

to 4.9 mmol/L prior to study and triglycerides < 4.52 mmol/L.<br />

Exclusion criteria: LDL cholesterol outwith the above values; participants outwith 30-70 years old; baseline<br />

triglycerides >4.52 mmol/L; body mass index > 38, history of <strong>heart</strong> <strong>disease</strong>, stroke, liver <strong>disease</strong>, abnormal<br />

blood count and renal function, use of lipid-lowering drugs/ botanicals/corticosteroids, androgens and<br />

food allergies. Medication was allowed if subjects were on stable doses of estrogen, estrogen/progestin,<br />

thiazide, beta-blockers and thyroid hormones.<br />

Smoking status: Unclear<br />

n = 152 randomised to 4 groups, 146 completed. Two arms <strong>for</strong> each intervention group <strong>for</strong> each location<br />

i.e. Columbia University(CU)/University of Texas (UoT). In total 79 in corn cereal group (39 CU/40<br />

UoT); in total 73 in oat-group (35 UC/ 38 UoT).<br />

Mean age: Corn group: CU 46.3+/- 9.7; UoT 51.5+/- 10.5. Oat group: CU 48.4+/- 11.9; UoT 49.7+/-<br />

10.5 years.<br />

Male/female ratio: Corn group: CU 9/3; UoT 12 /28. Oat group: CU 14/21; UoT 14/24.<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

Interventions Intervention: 45g of wholegrain ready to eat oat cereal (Cheerios) as part of National Cholesterol Education<br />

Program (NCEP) ’Step One’ diet.<br />

Control: 45 ounces of corn cereal daily as part of ’Step One’ diet. Both groups ate the NCEP ’Step One’<br />

diet <strong>for</strong> 5 weeks be<strong>for</strong>e intervention.<br />

Assessment of dietary compliance: 3 day food records and unannounced telephone calls <strong>for</strong> 24 hour recalls.<br />

Compliance with cereal consumption was determined from the number of unopened and empty packages<br />

at each visit, daily cereal intake records and participant interviews.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, triglycerides, HDL cholesterol, LDL cholesterol, body weight.<br />

Other outcomes: Apo A-1, Apo B, participants report of adverse reactions.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

33


Karmally 2005 (Continued)<br />

Notes Dietary fibre intake - the study reports that the wholegrain oat cereal provided 3.0 g of soluble fibre daily<br />

and the control cereal contained no soluble fibre. There was no difference in the soluble fibre content of<br />

the background ’Step One’ diets in the two groups.<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Keenan 2002<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Done <strong>for</strong> blood pressure but not clear if done <strong>for</strong> lipids.<br />

Length of intervention: 6 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done.<br />

Participants Diagnosis of CHD or risk factors: Hypertensives with average SBP of 130 to 160 mmHg and DBP of 85<br />

to 100 mmHg and with at least one reading greater than 140/190 as well as moderately elevated levels of<br />

fasting insulin (>10 µU/mL).<br />

Exclusion criteria: SBP, DBP and fasting insulin outwith the specified levels <strong>for</strong> inclusion. History of<br />

complications of hypertension, intestinal surgery or gastrointestinal <strong>disease</strong>s, chronic use of medications<br />

affecting the gastrointestinal tract, excessive alcohol use, smoking, diabetes, obesity, high soluble fibre diet,<br />

estrogen replacement therapy dosage unstable or greater than 2mg. Medications used: Not reported.<br />

Smoking status: Smokers excluded.<br />

n = 22 randomised to 2 groups, 18 completed.<br />

Treatment n=10 (baseline and follow-up); control: n=8 (baseline and follow-up)<br />

Mean age: 44 years (SD 18)<br />

Male/female ratio: 50% M/50% F in both groups.<br />

Baseline characteristics: Reported<br />

Geographical location: USA<br />

Interventions Intervention: Whole oat cereal.<br />

Control: Low fibre cereal (type not specified).<br />

Assessment of dietary compliance: Self-report in a daily cereal calendar, 3-day food records at baseline and<br />

at 6 weeks.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Systolic blood pressure, diastolic blood pressure, total cholesterol, triglycerides, HDL<br />

cholesterol, LDL cholesterol, body weight, fasting insulin.<br />

Other outcomes: Apo A-1, Apo B, participants report of adverse reactions.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

34


Keenan 2002 (Continued)<br />

Notes Dietary fibre intake - the study reports that the treatment oat cereal was standardised to 5.52 g/day betaglucan<br />

and to a low fibre cereal control (less than 1.0 g/day total fibre).<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Leinonen 2000<br />

Methods Study design: Randomised crossover study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear.<br />

Length of intervention: 4 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done <strong>for</strong>mally but all 40 subjects randomised appear to have completed<br />

the study.<br />

Participants Diagnosis of CHD or risk factors: Elevated serum cholesterol concentrations. Total serum cholesterol 5.5<br />

- 7.5 mmol/L.<br />

Exclusion criteria: Serum total cholesterol outwith above values; serum total triglycerides > 2.5 mmol/L,<br />

body mass index outwith 20-32; subjects taking lipid-lowering medication. Medications used: None<br />

reported other than those on lipid-lowering medication were excluded.<br />

Smoking status: Not reported.<br />

n = 40 randomised (crossover), 40 completed.<br />

Mean age: men 43+/- 2.0 years (SEM); women 43 +/- 1.6 years (SEM)<br />

Male/female ratio: 18/22<br />

Baseline characteristics: Reported<br />

Geographical location: Finland<br />

Interventions Intervention: Wholemeal rye bread based on finely milled wholemeal rye flour replaced customarily used<br />

breads and baked products in the usual diet.<br />

Control: Wheat breads made from refined wheat flour replaced customarily used breads and baked products<br />

in the usual diet.<br />

Washout interval (<strong>for</strong> crossover studies): 4 weeks.<br />

Assessment of dietary compliance: Daily records and 4-day food records.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, body weight<br />

Other outcomes: Values not reported but no significant changes were found in the fasting plasma glucose<br />

and insulin concentrations during the study.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

35


Leinonen 2000 (Continued)<br />

Notes Dietary fibre intake - the study reports significant differences <strong>for</strong> both men and women between the<br />

wholemeal rye bread and refined wheat bread <strong>for</strong> total dietary fibre, soluble fibre and insoluble fibre (P <<br />

0.05).<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Pereira 2002<br />

Methods Study design: Randomised crossover study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear.<br />

Length of intervention: 6 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done.<br />

Participants Diagnosis of CHD or risk factors: overweight and obese adults with hyperinsulinaemia.<br />

Exclusion criteria: outwith age 21-65 years; body mass index outwith 26-36; body weight fluctuation over<br />

past 6 months > 10%; smoking; consumption > 2 alcoholic drinks/day; diagnosed with diabetes, cancer<br />

CVD or other chronic medical conditions; those taking medications that would affect glucose, insulin,<br />

lipids or blood pressure; those engaging in a high level of physical activity; those following a special diet;<br />

those allergic to any foods.<br />

Medications used: None of the participants were taking medications that would affect glucose, insulin,<br />

lipids or blood pressure.<br />

Smoking status: Smokers excluded from study.<br />

n = 12 randomised (crossover), 11 completed.<br />

Mean age: 41.6+/- 2.67 years (SEM).<br />

Male/female ratio: 5/6<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

Interventions Intervention: <strong>Wholegrain</strong> diet including 6-10 servings daily of wholegrain breakfast cereal, bread, rice,<br />

pasta, muffins, cookies and snacks of which approx. 80% were wheat and the remainder oats, rice, corn,<br />

barley, rye on a 6-day menu rotation.<br />

Control: As above but 6-10 servings of refined grain foods daily instead of wholegrain foods based on<br />

white bread and refined wheat, rice and corn products.<br />

Washout interval (<strong>for</strong> crossover studies) : 6-9 weeks.<br />

Assessment of dietary compliance: Daily records and all food provided to participants. All participants<br />

were renumerated on completion of the study.<br />

Comparison of total energy and macronutrient intakes: Reported<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

36


Pereira 2002 (Continued)<br />

Outcomes Main outcomes: Fasting insulin, fasting glucose, insulin sensitivity (insulin clamp), body weight.<br />

Other outcomes: Side effects reported by participants<br />

Notes Dietary fibre intake - the study reports higher intakes of dietary fibre in the wholegrain group than<br />

the refined grain group <strong>for</strong> total dietary fibre (28.0 g wholegrain; 17.8 g control); soluble fibre (7.7 g<br />

wholegrain; 6.7 g control) and insoluble fibre (19.7 g wholegrain; 10.8 g control) but does not report<br />

whether the differences were statistically significant.<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Pins 2002<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Done .<br />

The cereal dispenser was also blinded.<br />

Length of intervention: 12 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done <strong>for</strong>mally but all 88 subjects randomised appear to have completed<br />

the study.<br />

Participants Diagnosis of CHD or risk factors: Mild or moderate hypertension.<br />

Exclusion criteria: History of systolic BP >160mmHg or diastolic BP > 115 mmHg; existing complications<br />

of hypertension; history of major intestinal surgeries; malabsorption of the gastrointestinal tract or biliary<br />

<strong>disease</strong>; use of beta-blockers; diabetes mellitus; body mass index > 35; history of excessive use of alcohol;<br />

current smoking; high soluble fibre intake; clinical use of antacids, bulk laxatives or other medications<br />

affecting gastrointestinal tract; continuous treatment with estrogen replacements, participation in another<br />

study 3 months be<strong>for</strong>e randomisation.<br />

Medications used: No more than one anti-hypertensive medication and/or one diuretic medication. 80<br />

participants were on a single anti-hypertensive medication, 8 were on an anti-hypertensive drug plus a<br />

diuretic medication.<br />

Smoking status: No smokers included.<br />

n = 88 randomised to 2 groups, 88 completed.<br />

Control group: n = 43 baseline, n = 43 end. Intervention group: (Cheerios) n = 45 baseline, n = 45end.<br />

Mean age: Control group 46.4 years (+/- 15.3 SD); Intervention group 48.7 years (+/- 16.9 SD)<br />

Male/female ratio: 45/43<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

37


Pins 2002 (Continued)<br />

Interventions Intervention: Whole oat <strong>cereals</strong> (oatmeal and oat squares) as part of usual diet.<br />

Control: Refined grain wheat-based <strong>cereals</strong> (hot wheat cereal and Kellogg’s Crispix) as part of usual diet.<br />

Assessment of dietary compliance: 3-day food records at baseline and end of 12 weeks, and cereal diaries<br />

kept.<br />

Comparison of total energy and macronutrient intakes: Reported<br />

Outcomes Main outcomes: Total cholesterol, LDL cholesterol, triglycerides, bodyweight.<br />

Other outcomes: Proportion of participants reducing anti-hypertensive medication (no SD/SEM), mean<br />

blood pressure (no SD/SEM), plasma glucose (not clear if this is fasting glucose)<br />

Notes Dietary fibre intake - **still awaiting table W1 from full paper**.<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Reynolds 2000<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Done.<br />

Blinding of outcome assessors: Not clear.<br />

Length of intervention: 4 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done.<br />

Participants Diagnosis of CHD or risk factors: mild to moderate hypercholesterolemia (mean baseline value 5.94+/-<br />

0.65 mmol/L).<br />

Exclusion criteria: Baseline triglyceride > 3.39 mmol/L; body weight > 30% over ideal; history of recent<br />

myocardial infarction; major surgery; unstable angina; congestive <strong>heart</strong> failure; significant metabolic, renal,<br />

hepatic or gastrointestinal <strong>disease</strong>; concurrent use of corticosteroids, thiazides, estrogens, progesterones,<br />

antibiotics or lipid-lowering agents.<br />

Medications used: Beta-adrenergic blockers were allowed if dose was kept constant during the study.<br />

Smoking status: Smokers included but proportion not reported.<br />

n = 46 randomised. Participants were paired by gender and cholesterol levels be<strong>for</strong>e randomisation, 43<br />

completed. Whole oat cereal treatment group n = 22 (baseline and follow-up); Control: n = 21 (baseline<br />

and follow-up).<br />

Mean age: 51.6 years (no SD).<br />

Male/female ratio: 21/22<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

38


Reynolds 2000 (Continued)<br />

Interventions Intervention: American Heart Association (AHA) ’Step One’ diet plus 2 portions of 42.5g wholegrain oat<br />

puffs daily.<br />

Control: AHA ’Step One’ diet plus two portions of 42.5g commercial cornflakes daily.<br />

Assessment of dietary compliance: Subject interviews, counting returned unopened boxes of cereal.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, blood pressure, bodyweight.<br />

Notes Dietary fibre intake - the study reports that differences between control and treatment groups were detected<br />

<strong>for</strong> mean posttreatment intake of total dietary fibre (16.8+/-1.1 g control; 25.5 +/- 1.5 g treatment P<br />

< 0.001); posttreatment soluble fibre intake (5.2 +/- 0.3 g control; 8.8+/-0.4 g whole oat treatment P<br />

198 mg/dL (5.12 mmol/L)<br />

Exclusion criteria: Those taking lipid-lowering medication, pregnancy, more than 1.5 times desirable<br />

weight <strong>for</strong> height.<br />

Medications used: No lipid-lowering medication used.<br />

Smoking status: Not reported.<br />

For subgroup with total cholesterol > 198 mg/dL number randomised not reported. Results reported <strong>for</strong><br />

118 who completed. Control group: n = 59 baseline, n = 59 end. Intervention group (oatmeal) n = 59<br />

baseline, n = 59 end.<br />

Mean age: Mean age: 42.2 years (no SD) <strong>for</strong> whole group.<br />

Male/female ratio: 63.5:/36.5 <strong>for</strong> whole group.<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

39


Van Horn 1988 (Continued)<br />

Interventions Intervention: Inclusion of 56 g dry weight of oatmeal in the American Heart Association (AHA) fatmodified<br />

diet, isocalorically substituted <strong>for</strong> other foods.<br />

Control: AHA fat-modified diet with no oat products.<br />

Assessment of dietary compliance: 3-day food records every 4 weeks.<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, LDL cholesterol.<br />

Other outcomes: None.<br />

Notes Dietary fibre intake - not specifically reported <strong>for</strong> subgroup with baseline total cholesterol > 198mg/dL .<br />

For whole group the increase in soluble fibre intake was statistically significantly greater at weeks 4 and 8<br />

<strong>for</strong> the oatmeal group (P < 0.005). There was no significant difference in insoluble fibre intake.<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Van Horn 1991<br />

Methods Study design: Randomised parallel study.<br />

Random allocation: Unclear.<br />

Allocation concealment: Unclear.<br />

Blinding of outcome assessors: Unclear.<br />

Length of intervention: 8 weeks.<br />

Unit of allocation: Participant.<br />

Unit of analysis: Participant.<br />

Protection against contamination: Unclear.<br />

Intention-to-treat analysis: Not done.<br />

Participants Diagnosis of CHD or risk factors: Hypercholesterolemic individuals previously found to have serum total<br />

cholesterol levels above 5.20 mmol/L.<br />

Exclusion criteria: Those on lipid-lowering drugs or anti-hypertensive medication; diabetic, pregnant or<br />

lactating; body weight > 1.5 times desirable weight; on a weight loss diet.<br />

Medications used: No lipid-lowering medication used <strong>for</strong> 6 weeks prior to trial.<br />

Smoking status: Not reported.<br />

n = 156 randomised to 7 groups, 148 completed. For this review only those in the OM-28 oatmeal group<br />

have been included: OM-28 n = 20 (baseline and follow-up); control: n = 15 (baseline and follow-up).<br />

Mean age: 51.1 years (no SD)<br />

Male/female ratio: 7/13<br />

Baseline characteristics: Reported.<br />

Geographical location: USA.<br />

Interventions Intervention: Two packets (56.7 g dry weight) of instant oats per day, substituting oats <strong>for</strong> other carbohydrate<br />

foods in the usual diet.<br />

Control: The control group was aked to maintain usual intake throughout the study.<br />

Assessment of dietary compliance: 3-day food records at baseline, 4 and 8 weeks.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

40


Van Horn 1991 (Continued)<br />

Comparison of total energy and macronutrient intakes: Reported.<br />

Outcomes Main outcomes: Total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, blood pressure, bodyweight,<br />

body mass index.<br />

Other outcomes: VLDL cholesterol.<br />

Notes Dietary fibre intake - the study reports statistically significant differences between the oatmeal (Group<br />

1) and control groups (Group 2) <strong>for</strong> intake of total fibre (difference = 4.2 g P < 0.001); soluble fibre (<br />

difference = 1.8 g P < 0.001) and insoluble fibre (difference = 2.3 g P < 0.01)<br />

Risk of bias<br />

Item Authors’ judgement Description<br />

Allocation concealment? Unclear B - Unclear<br />

Karmally 2005 - the authors (W. Karmally) have confirmed to us that the wholegrain oat cereal was a wholegrain cereal even though it<br />

is described as an oat bran cereal in the text of the review.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

41


Characteristics of excluded studies [ordered by study ID]<br />

Anderson 1978 Not specifically wholegrain<br />

Not RCT or CCT<br />

Anderson 1979 Not specifically wholegrain<br />

Participants not free-living<br />

Not RCT or CCT<br />

Intervention < 4 weeks<br />

Asp 1981 Not specifically wholegrain<br />

Cannot isolate the effect of wholegrain<br />

Intervention < 4 weeks<br />

Behall 2004a Cannot isolate the effect of wholegrain. (Intervention groups are barley, barley and wholegrain or wholegrain.<br />

Not clear if all the barley based foods are wholegrain but any comparison would be wholegrain versus wholegrain).<br />

Behall 2004b Cannot isolate the effect of wholegrain. (Intervention groups are barley, barley and wholegrain or wholegrain.<br />

Not clear if all the barley based foods are wholegrain but any comparison would be wholegrain versus wholegrain).<br />

Birkeland 1991 From translation of paper, both the intervention and control groups were given products containing oat bran<br />

which does not meet the definition of wholegrain <strong>for</strong> this review.<br />

Birketvedt 2000 Not wholegrain<br />

Booyens 1966 Participants not diagnosed with CHD or risk factors<br />

Bourdon 1999 Not wholegrain<br />

Intervention < 4 weeks<br />

Participants not diagnosed with CHD or risk factors<br />

Bruttomesso 1989 Not specifically wholegrain<br />

Burr 1989 Not specifically wholegrain<br />

Cairella 1995 Not specifically wholegrain<br />

Chandalia 2000 Not specifically wholegrain<br />

Collier 1982 Intervention < 4 weeks<br />

Comi 1995 Not specifically wholegrain<br />

Crapo 1981 Not wholegrain<br />

Intervention < 4 weeks<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

42


(Continued)<br />

Data 1980 Intervention < 4 weeks<br />

Davy 2002a Comparison is wholegrain versus wholegrain<br />

Cannot isolate effect of wholegrain<br />

Davy 2002b Cannot isolate the effect of wholegrain<br />

Eliasson 1992 Not wholegrain<br />

Fehily 1986 Participants not diagnosed with CHD or risk factors<br />

Fischer 2000 Not specifically wholegrain<br />

Fordyce-Baum 1989 Product is described as a whole wheat protein isolate. We contacted the authors but were unable to obtain<br />

any further in<strong>for</strong>mation on the nature of the product.<br />

Fraser 1981 Participants not diagnosed with CHD or risk factors<br />

Intervention < 4 weeks<br />

Fung 2002 Not RCT or CCT<br />

Golay 1992 Not specifically wholegrain<br />

Intervention < 4 weeks<br />

Guzic 1994 Not wholegrain<br />

Hagander 1985 Intervention < 4 weeks<br />

Hagander 1988 Not specifically wholegrain<br />

He 1995 Not RCT or CCT<br />

Heaton 1976 Participants not diagnosed with CHD or risk factors<br />

Not concurrent control?<br />

Hoffman 1982 Not specifically wholegrain<br />

Participants not free-living<br />

Not RCT or CCT<br />

Intervention < 4 weeks<br />

Hollenbeck 1986 Not specifically wholegrain<br />

Cannot isolate effect of wholegrain<br />

Hunninghake 1994 Not wholegrain<br />

Jacobs 2002 Outcome is serum enterolactone which was not a specified outcome <strong>for</strong> this review<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

43


(Continued)<br />

Jang 2001 Not specifically wholegrain<br />

Cannot isolate the effect of wholegrain<br />

Jenkins 1985 Not specifically wholegrain<br />

Not RCT or CCT<br />

Jenkins 1993 Not specifically wholegrain<br />

Judd 1981 Participants not diagnosed with CHD or risk factors<br />

Cannot isolate effect of wholegrains<br />

Not RCT or CCT<br />

Intervention < 4 weeks<br />

Juntunen 2002 Participants not diagnosed with CHD or risk factors<br />

Intervention < 4 weeks<br />

Juntunen 2003 Intervention group consumed wholemeal rye bread enriched with rye bran - cannot isolate the effect of<br />

wholegrain.<br />

Confirmation of composition of bread received from authors (K. Juntunen).<br />

Kabir 2002 Cannot isolate the effect of wholegrain<br />

Karlstrom 1984 Not specifically wholegrain<br />

Participants not free-living<br />

Intervention < 4 weeks<br />

Katz 2001a Cannot isolate the effect of wholegrain<br />

Participants not diagnosed with CHD or risk factors<br />

Katz 2001b Participants not diagnosed with CHD or risk factors<br />

Intervention < 4 weeks<br />

Kay 1977 Not wholegrain<br />

Participants not diagnosed with CHD or risk factors<br />

Not RCT or CCT<br />

Intervention < 4 weeks<br />

Kay 1981 Not specifically wholegrain<br />

Participants not free-living<br />

Intervention < 4 weeks<br />

Kesaniemi 1990 Not specifically wholegrain<br />

Participants not diagnosed with CHD or risk factors<br />

Kleemola 1999 Not wholegrain<br />

Kris-Etherton 2002 Not specifically wholegrain<br />

Lakshmi 1996 Intervention < 4 weeks<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

44


(Continued)<br />

Leinonen 1999 Participants not diagnosed with CHD or risk factors<br />

Intervention < 4 weeks<br />

Liese 2003 Not RCT or CCT<br />

Lousley 1984 Not specifically wholegrain<br />

Different macronutrient (carbohydrate)compositions<br />

MacMahon 1998 Not wholegrain<br />

Manhire 1981 Intervention is the effect of wholegrain plus the withdrawal of refined sugars - cannot isolate the effect of<br />

wholegrain.<br />

Mathur 1968 Not specifically wholegrain<br />

Participants not diagnosed with CHD or risk factors<br />

Not RCT or CCT<br />

McIntosh 1991 Cannot specifically isolate the effect of wholegrain<br />

Montonen 2003 Not RCT or CCT<br />

Nielsen 1988 Not wholegrain<br />

Intervention < 4 weeks<br />

O’Kell 1988 Participants not diagnosed with CHD or risk factors<br />

Odes 1993 Not wholegrain cereal<br />

Pacy 1986 Not RCT or CCT<br />

Poulter 1993 Not all participants diagnosed with CHD or risk factors<br />

Cannot isolate the effect of wholegrain<br />

Reynolds 1989 Abstract only - no full paper found<br />

Attempted to contact authors but no further in<strong>for</strong>mation could be obtained.<br />

Results quoted appear to be average of 2 and 4 week results not end results after 4 weeks.<br />

Rigaud 1990 Not wholegrain<br />

Roth 1985 Participants not diagnosed with CHD or at risk of CHD<br />

Russ 1985 High fibre versus low fibre - not specifically wholegrain<br />

Rytter 1996 Not specifically wholegrain<br />

Not RCT or CCT<br />

Saltzman 2001a Participants not diagnosed with CHD or risk factors<br />

Not all participants free-living<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

45


(Continued)<br />

Saltzman 2001b Participants not diagnosed with CHD or risk factors<br />

Schlamowitz 1987 Not wholegrain<br />

Turnbull 1987 Comparison is rolled oats versus wheat but it is not clear from the paper whether the wheat products used<br />

in the comparison were wholegrain or refined grain. The authors were contacted but no further details were<br />

obtained.<br />

Turnbull 1989 Comparison is rolled oats versus wheat but it is not clear from the paper whether the wheat products used<br />

in the comparison were wholegrain or refined grain. The authors were contacted but no further details were<br />

obtained.<br />

Turpeinen 2000 Participants not diagnosed with CHD or risk factors<br />

Van Horn 1986 Participants not dignosed with CHD or risk factors<br />

Van Horn 2001 Cannot isolate the effects of wholegrain (intervention is oats and oat bran)<br />

Willms 1987 Participants not free-living<br />

Intervention < 4 weeks<br />

Wolever 2003 Not specifically wholegrain<br />

Wolffenbuttel 1992 Intervention is not specifically wholegrain<br />

Wursch 1991 Not specifically wholegrain<br />

Intervention < 4 weeks<br />

Characteristics of ongoing studies [ordered by study ID]<br />

FSA 2005a<br />

Trial name or title Randomised controlled trial to test the impact of increased consumption of wholegrain foods on cardiovascular<br />

<strong>disease</strong> risk (the WHOLE<strong>heart</strong> study). University of Newcastle and MRC Cambridge Human Nutrition<br />

Research Centre). Funded by the UK Food Standards Agency (project no. N02036). Parallel RCT study in<br />

free-living subjects with three treatment arms recruited in 2 cohorts in two different centres.<br />

Methods<br />

Participants 300 men and women, 150 at each of the participating centre, aged 30-65 years with a BMI>25 kg/m3.<br />

Interventions Treatment A - control, no intervention.<br />

Treatment B - Consuming 3 portions of wholegrain food per day <strong>for</strong> 3 months.<br />

Treatment C - Consuming 3 portions of wholegrain food per day <strong>for</strong> 2 months increasing to 6 portions per<br />

day <strong>for</strong> a further 2 months.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

46


FSA 2005a (Continued)<br />

Outcomes The primary outcome measure will be LDL cholesterol. Other outcome measures to be assessed are:<br />

Body composition, fasting lipid profile, insulin sensitivity, inflammatory status and markers of endothelial<br />

function, blood pressure, dietary change.<br />

Starting date Start date 01/01/2005. End date 31/12/2007<br />

Contact in<strong>for</strong>mation Dr. Chris Seal, University of Newcastle, School of Agriculture, Food and Rural Development.<br />

(chris.seal@ncl.ac.uk)<br />

Notes http://www.food.gov.uk/multimedia/pdfs/reslistnov05.pdf. Also details from personal communication from<br />

Chris Seal and from website.<br />

FSA 2005b<br />

Trial name or title Comparison of effects of increased wholegrain foods on markers of cardiovascular <strong>disease</strong> risk. Funded by<br />

Food Standards Agency, UK (Project no. N02035). Collaboration between the University of Aberdeen, the<br />

Rowett Research Institute and The Robert Gordon University.<br />

Methods<br />

Participants Volunteers aged between 40-65<br />

Interventions Comparison of wheat-based wholegrain diet with an oat and wheat-based diet and a control diet.<br />

Outcomes Weight, blood pressure, cholesterol and other outcomes.<br />

Starting date Start date 01/01/2005. End date 30/06/2009.<br />

Contact in<strong>for</strong>mation Dr. Frank Thies, Dept. of Medicine and Therapeutics, Polwarth Building, Foresterhill, University of Aberdeen<br />

AB25 2ZD.<br />

(f.thies@abdn.ac.uk)<br />

Notes http://www.food.gov.uk/multimedia/pdfs/reslistnov05.pdf. Details also from http://www.abdn.ac.uk/mediareleases/release.php?id=289<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

47


D A T A A N D A N A L Y S E S<br />

Comparison 1. Total cholesterol<br />

Outcome or subgroup title<br />

No. of<br />

studies<br />

No. of<br />

participants Statistical method Effect size<br />

1 4 week outcomes 4 321 Mean Difference (IV, Fixed, 95% CI) -0.16 [-0.34, 0.01]<br />

1.1 Parallel studies (all oats) 3 241 Mean Difference (IV, Fixed, 95% CI) -0.12 [-0.32, 0.09]<br />

1.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) -0.30 [-0.65, 0.05]<br />

2 All endpoint outcomes 9 738 Mean Difference (IV, Fixed, 95% CI) -0.20 [-0.31, -0.10]<br />

2.1 Parallel studies (all oats) 8 658 Mean Difference (IV, Fixed, 95% CI) -0.19 [-0.30, -0.08]<br />

2.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) -0.30 [-0.65, 0.05]<br />

Comparison 2. LDL cholesterol<br />

Outcome or subgroup title<br />

No. of<br />

studies<br />

No. of<br />

participants Statistical method Effect size<br />

1 4 week outcomes 4 321 Mean Difference (IV, Fixed, 95% CI) -0.19 [-0.37, -0.01]<br />

1.1 Parallel studies (all oats) 3 241 Mean Difference (IV, Fixed, 95% CI) -0.19 [-0.39, 0.02]<br />

1.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) -0.20 [-0.59, 0.19]<br />

2 All endpoint outcomes 9 738 Mean Difference (IV, Fixed, 95% CI) -0.18 [-0.28, -0.09]<br />

2.1 Parallel studies (all oats) 8 658 Mean Difference (IV, Fixed, 95% CI) -0.18 [-0.28, -0.09]<br />

2.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) -0.20 [-0.59, 0.19]<br />

Comparison 3. HDL cholesterol<br />

Outcome or subgroup title<br />

No. of<br />

studies<br />

No. of<br />

participants Statistical method Effect size<br />

1 4 week outcomes 3 203 Mean Difference (IV, Fixed, 95% CI) 0.03 [-0.06, 0.11]<br />

1.1 Parallel studies (all oats) 2 123 Mean Difference (IV, Fixed, 95% CI) 0.04 [-0.06, 0.14]<br />

1.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) Not estimable<br />

2 All endpoint outcomes 7 532 Mean Difference (IV, Fixed, 95% CI) 0.00 [-0.05, 0.05]<br />

2.1 Parallel studies (all oats) 6 452 Mean Difference (IV, Fixed, 95% CI) 0.00 [-0.05, 0.05]<br />

2.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) Not estimable<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

48


Comparison 4. Triglycerides<br />

Outcome or subgroup title<br />

No. of<br />

studies<br />

No. of<br />

participants Statistical method Effect size<br />

1 4 week outcomes 3 203 Mean Difference (IV, Fixed, 95% CI) 0.01 [-0.13, 0.16]<br />

1.1 Parallel studies (all oats) 2 123 Mean Difference (IV, Fixed, 95% CI) 0.05 [-0.21, 0.30]<br />

1.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) Not estimable<br />

2 All endpoint outcomes (Keenan<br />

data as SD)<br />

7 496 Mean Difference (IV, Fixed, 95% CI) 0.01 [-0.09, 0.11]<br />

2.1 Parallel studies (all oats) 6 416 Mean Difference (IV, Fixed, 95% CI) 0.01 [-0.11, 0.13]<br />

2.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) Not estimable<br />

3 All endpoint outcomes (Keenan<br />

data as SEM)<br />

7 496 Mean Difference (IV, Fixed, 95% CI) -0.01 [-0.11, 0.09]<br />

3.1 Parallel studies (all oats) 6 416 Mean Difference (IV, Fixed, 95% CI) -0.02 [-0.14, 0.11]<br />

3.2 Crossover studies (rye) 1 80 Mean Difference (IV, Fixed, 95% CI) Not estimable<br />

Comparison 5. Body weight (kg)<br />

Outcome or subgroup title<br />

No. of<br />

studies<br />

No. of<br />

participants Statistical method Effect size<br />

1 All endpoint outcomes 7 Mean Difference (IV, Fixed, 95% CI) Subtotals only<br />

1.1 Parallel studies (all oats) 5 425 Mean Difference (IV, Fixed, 95% CI) 1.10 [-0.73, 2.93]<br />

1.2 Crossover studies (rye) 2 104 Mean Difference (IV, Fixed, 95% CI) -0.30 [-1.08, 0.48]<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

49


Analysis 1.1. Comparison 1 Total cholesterol, Outcome 1 4 week outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 1 Total cholesterol<br />

Outcome: 1 4 week outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Reynolds 2000 22 5.87 (0.56) 21 5.83 (0.6) 26.5 % 0.04 [ -0.31, 0.39 ]<br />

Van Horn 1988 59 5.45 (0.88) 59 5.79 (1.06) 25.9 % -0.34 [ -0.69, 0.01 ]<br />

Van Horn 1991 42 6.28 (0.92) 38 6.32 (0.84) 21.5 % -0.04 [ -0.43, 0.35 ]<br />

Subtotal (95% CI) 123 118 74.0 % -0.12 [ -0.32, 0.09 ]<br />

Heterogeneity: Chi 2 = 2.48, df = 2 (P = 0.29); I 2 =19%<br />

Test <strong>for</strong> overall effect: Z = 1.10 (P = 0.27)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 6.2 (0.8) 40 6.5 (0.8) 26.0 % -0.30 [ -0.65, 0.05 ]<br />

Subtotal (95% CI) 40 40 26.0 % -0.30 [ -0.65, 0.05 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 1.68 (P = 0.094)<br />

Total (95% CI) 163 158 100.0 % -0.16 [ -0.34, 0.01 ]<br />

Heterogeneity: Chi 2 = 3.26, df = 3 (P = 0.35); I 2 =8%<br />

Test <strong>for</strong> overall effect: Z = 1.80 (P = 0.072)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.78, df = 1 (P = 0.38), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

50


Analysis 1.2. Comparison 1 Total cholesterol, Outcome 2 All endpoint outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 1 Total cholesterol<br />

Outcome: 2 All endpoint outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Davidson 1991 20 6.54 (0.68) 15 6.79 (0.85) 3.9 % -0.25 [ -0.77, 0.27 ]<br />

Johnston 1998 62 6.04 (0.61) 62 6.23 (0.54) 26.1 % -0.19 [ -0.39, 0.01 ]<br />

Karmally 2005 73 5.14 (0.65) 79 5.21 (0.7) 23.3 % -0.07 [ -0.28, 0.14 ]<br />

Keenan 2002 10 4.73 (0.47) 8 5.11 (0.48) 5.5 % -0.38 [ -0.82, 0.06 ]<br />

Pins 2002 45 4.66 (0.87) 43 5.34 (1.11) 6.2 % -0.68 [ -1.10, -0.26 ]<br />

Reynolds 2000 22 5.87 (0.56) 21 5.83 (0.6) 8.9 % 0.04 [ -0.31, 0.39 ]<br />

Van Horn 1988 59 5.47 (0.84) 59 5.78 (1.03) 9.3 % -0.31 [ -0.65, 0.03 ]<br />

Van Horn 1991 42 6.15 (0.86) 38 6.3 (0.82) 7.9 % -0.15 [ -0.52, 0.22 ]<br />

Subtotal (95% CI) 333 325 91.3 % -0.19 [ -0.30, -0.08 ]<br />

Heterogeneity: Chi 2 = 9.45, df = 7 (P = 0.22); I 2 =26%<br />

Test <strong>for</strong> overall effect: Z = 3.48 (P = 0.00050)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 6.2 (0.8) 40 6.5 (0.8) 8.7 % -0.30 [ -0.65, 0.05 ]<br />

Subtotal (95% CI) 40 40 8.7 % -0.30 [ -0.65, 0.05 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 1.68 (P = 0.094)<br />

Total (95% CI) 373 365 100.0 % -0.20 [ -0.31, -0.10 ]<br />

Heterogeneity: Chi 2 = 9.78, df = 8 (P = 0.28); I 2 =18%<br />

Test <strong>for</strong> overall effect: Z = 3.82 (P = 0.00013)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.33, df = 1 (P = 0.57), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

51


Analysis 2.1. Comparison 2 LDL cholesterol, Outcome 1 4 week outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 2 LDL cholesterol<br />

Outcome: 1 4 week outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Reynolds 2000 22 3.99 (0.52) 21 3.96 (0.6) 28.4 % 0.03 [ -0.31, 0.37 ]<br />

Van Horn 1988 59 3.18 (0.79) 59 3.64 (1.14) 25.6 % -0.46 [ -0.81, -0.11 ]<br />

Van Horn 1991 42 4.24 (0.87) 38 4.39 (0.75) 25.4 % -0.15 [ -0.51, 0.21 ]<br />

Subtotal (95% CI) 123 118 79.4 % -0.19 [ -0.39, 0.02 ]<br />

Heterogeneity: Chi 2 = 3.93, df = 2 (P = 0.14); I 2 =49%<br />

Test <strong>for</strong> overall effect: Z = 1.81 (P = 0.070)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 4.3 (0.9) 40 4.5 (0.9) 20.6 % -0.20 [ -0.59, 0.19 ]<br />

Subtotal (95% CI) 40 40 20.6 % -0.20 [ -0.59, 0.19 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.99 (P = 0.32)<br />

Total (95% CI) 163 158 100.0 % -0.19 [ -0.37, -0.01 ]<br />

Heterogeneity: Chi 2 = 3.93, df = 3 (P = 0.27); I 2 =24%<br />

Test <strong>for</strong> overall effect: Z = 2.06 (P = 0.039)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.00, df = 1 (P = 0.95), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

52


Analysis 2.2. Comparison 2 LDL cholesterol, Outcome 2 All endpoint outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 2 LDL cholesterol<br />

Outcome: 2 All endpoint outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control diet Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Davidson 1991 20 4.47 (0.61) 15 4.79 (0.73) 4.0 % -0.32 [ -0.78, 0.14 ]<br />

Johnston 1998 62 3.93 (0.49) 62 4.15 (0.42) 32.5 % -0.22 [ -0.38, -0.06 ]<br />

Karmally 2005 73 3.43 (0.56) 79 3.49 (0.58) 25.5 % -0.06 [ -0.24, 0.12 ]<br />

Keenan 2002 10 2.94 (0.63) 8 3.37 (0.45) 3.4 % -0.43 [ -0.93, 0.07 ]<br />

Pins 2002 45 3.03 (0.74) 43 3.41 (0.79) 8.2 % -0.38 [ -0.70, -0.06 ]<br />

Reynolds 2000 22 3.99 (0.52) 21 3.96 (0.6) 7.4 % 0.03 [ -0.31, 0.37 ]<br />

Van Horn 1988 59 3.25 (0.76) 59 3.42 (1.16) 6.7 % -0.17 [ -0.52, 0.18 ]<br />

Van Horn 1991 42 4.16 (0.83) 38 4.44 (0.77) 6.8 % -0.28 [ -0.63, 0.07 ]<br />

Subtotal (95% CI) 333 325 94.6 % -0.18 [ -0.28, -0.09 ]<br />

Heterogeneity: Chi 2 = 6.55, df = 7 (P = 0.48); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 3.82 (P = 0.00013)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 4.3 (0.9) 40 4.5 (0.9) 5.4 % -0.20 [ -0.59, 0.19 ]<br />

Subtotal (95% CI) 40 40 5.4 % -0.20 [ -0.59, 0.19 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.99 (P = 0.32)<br />

Total (95% CI) 373 365 100.0 % -0.18 [ -0.28, -0.09 ]<br />

Heterogeneity: Chi 2 = 6.56, df = 8 (P = 0.58); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 3.95 (P = 0.000080)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.01, df = 1 (P = 0.94), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

53


Analysis 3.1. Comparison 3 HDL cholesterol, Outcome 1 4 week outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 3 HDL cholesterol<br />

Outcome: 1 4 week outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control diet Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Reynolds 2000 22 1.29 (0.33) 21 1.19 (0.27) 22.9 % 0.10 [ -0.08, 0.28 ]<br />

Van Horn 1991 42 1.36 (0.28) 38 1.35 (0.26) 52.9 % 0.01 [ -0.11, 0.13 ]<br />

Subtotal (95% CI) 64 59 75.9 % 0.04 [ -0.06, 0.14 ]<br />

Heterogeneity: Chi 2 = 0.67, df = 1 (P = 0.41); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 0.74 (P = 0.46)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 1.4 (0.4) 40 1.4 (0.4) 24.1 % 0.0 [ -0.18, 0.18 ]<br />

Subtotal (95% CI) 40 40 24.1 % 0.0 [ -0.18, 0.18 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.0 (P = 1.0)<br />

Total (95% CI) 104 99 100.0 % 0.03 [ -0.06, 0.11 ]<br />

Heterogeneity: Chi 2 = 0.80, df = 2 (P = 0.67); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 0.64 (P = 0.52)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.13, df = 1 (P = 0.72), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

54


Analysis 3.2. Comparison 3 HDL cholesterol, Outcome 2 All endpoint outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 3 HDL cholesterol<br />

Outcome: 2 All endpoint outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Davidson 1991 20 1.48 (0.33) 15 1.34 (0.45) 3.3 % 0.14 [ -0.13, 0.41 ]<br />

Johnston 1998 62 1.29 (0.33) 62 1.25 (0.28) 20.8 % 0.04 [ -0.07, 0.15 ]<br />

Karmally 2005 73 0.93 (0.26) 79 0.97 (0.23) 39.5 % -0.04 [ -0.12, 0.04 ]<br />

Keenan 2002 10 1.07 (0.32) 8 1.06 (0.23) 3.7 % 0.01 [ -0.24, 0.26 ]<br />

Reynolds 2000 22 1.29 (0.33) 21 1.19 (0.27) 7.5 % 0.10 [ -0.08, 0.28 ]<br />

Van Horn 1991 42 1.28 (0.29) 38 1.3 (0.25) 17.3 % -0.02 [ -0.14, 0.10 ]<br />

Subtotal (95% CI) 229 223 92.1 % 0.00 [ -0.05, 0.05 ]<br />

Heterogeneity: Chi 2 = 3.87, df = 5 (P = 0.57); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 0.07 (P = 0.95)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 1.4 (0.4) 40 1.4 (0.4) 7.9 % 0.0 [ -0.18, 0.18 ]<br />

Subtotal (95% CI) 40 40 7.9 % 0.0 [ -0.18, 0.18 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.0 (P = 1.0)<br />

Total (95% CI) 269 263 100.0 % 0.00 [ -0.05, 0.05 ]<br />

Heterogeneity: Chi 2 = 3.87, df = 6 (P = 0.69); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 0.06 (P = 0.95)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.00, df = 1 (P = 0.99), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

55


Analysis 4.1. Comparison 4 Triglycerides, Outcome 1 4 week outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 4 Triglycerides<br />

Outcome: 1 4 week outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Reynolds 2000 22 1.3 (0.47) 21 1.48 (0.78) 13.9 % -0.18 [ -0.57, 0.21 ]<br />

Van Horn 1991 42 1.48 (0.92) 38 1.26 (0.61) 18.1 % 0.22 [ -0.12, 0.56 ]<br />

Subtotal (95% CI) 64 59 32.1 % 0.05 [ -0.21, 0.30 ]<br />

Heterogeneity: Chi 2 = 2.32, df = 1 (P = 0.13); I 2 =57%<br />

Test <strong>for</strong> overall effect: Z = 0.36 (P = 0.72)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 1.1 (0.4) 40 1.1 (0.4) 67.9 % 0.0 [ -0.18, 0.18 ]<br />

Subtotal (95% CI) 40 40 67.9 % 0.0 [ -0.18, 0.18 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.0 (P = 1.0)<br />

Total (95% CI) 104 99 100.0 % 0.01 [ -0.13, 0.16 ]<br />

Heterogeneity: Chi 2 = 2.41, df = 2 (P = 0.30); I 2 =17%<br />

Test <strong>for</strong> overall effect: Z = 0.20 (P = 0.84)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.09, df = 1 (P = 0.77), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

56


Analysis 4.2. Comparison 4 Triglycerides, Outcome 2 All endpoint outcomes (Keenan data as SD).<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 4 Triglycerides<br />

Outcome: 2 All endpoint outcomes (Keenan data as SD)<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Davidson 1991 20 1.28 (0.61) 15 1.42 (0.54) 6.8 % -0.14 [ -0.52, 0.24 ]<br />

Karmally 2005 73 1.71 (0.71) 79 1.63 (0.75) 18.4 % 0.08 [ -0.15, 0.31 ]<br />

Keenan 2002 10 2.11 (0.66) 8 1.7 (0.24) 5.1 % 0.41 [ -0.03, 0.85 ]<br />

Pins 2002 45 1.95 (0.47) 43 2.08 (0.52) 23.1 % -0.13 [ -0.34, 0.08 ]<br />

Reynolds 2000 22 1.3 (0.47) 21 1.48 (0.78) 6.6 % -0.18 [ -0.57, 0.21 ]<br />

Van Horn 1991 42 1.54 (1.04) 38 1.22 (0.53) 7.8 % 0.32 [ -0.04, 0.68 ]<br />

Subtotal (95% CI) 212 204 67.7 % 0.01 [ -0.11, 0.13 ]<br />

Heterogeneity: Chi 2 = 9.66, df = 5 (P = 0.09); I 2 =48%<br />

Test <strong>for</strong> overall effect: Z = 0.22 (P = 0.83)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 1.1 (0.4) 40 1.1 (0.4) 32.3 % 0.0 [ -0.18, 0.18 ]<br />

Subtotal (95% CI) 40 40 32.3 % 0.0 [ -0.18, 0.18 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.0 (P = 1.0)<br />

Total (95% CI) 252 244 100.0 % 0.01 [ -0.09, 0.11 ]<br />

Heterogeneity: Chi 2 = 9.68, df = 6 (P = 0.14); I 2 =38%<br />

Test <strong>for</strong> overall effect: Z = 0.18 (P = 0.86)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.02, df = 1 (P = 0.90), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

57


Analysis 4.3. Comparison 4 Triglycerides, Outcome 3 All endpoint outcomes (Keenan data as SEM).<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 4 Triglycerides<br />

Outcome: 3 All endpoint outcomes (Keenan data as SEM)<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Davidson 1991 20 1.28 (0.61) 15 1.42 (0.54) 7.1 % -0.14 [ -0.52, 0.24 ]<br />

Karmally 2005 73 1.71 (0.71) 79 1.63 (0.75) 19.3 % 0.08 [ -0.15, 0.31 ]<br />

Keenan 2002 10 2.11 (2.09) 8 1.7 (0.68) 0.5 % 0.41 [ -0.97, 1.79 ]<br />

Pins 2002 45 1.95 (0.47) 43 2.08 (0.52) 24.2 % -0.13 [ -0.34, 0.08 ]<br />

Reynolds 2000 22 1.3 (0.47) 21 1.48 (0.78) 6.9 % -0.18 [ -0.57, 0.21 ]<br />

Van Horn 1991 42 1.54 (1.04) 38 1.22 (0.53) 8.2 % 0.32 [ -0.04, 0.68 ]<br />

Subtotal (95% CI) 212 204 66.2 % -0.02 [ -0.14, 0.11 ]<br />

Heterogeneity: Chi 2 = 6.68, df = 5 (P = 0.25); I 2 =25%<br />

Test <strong>for</strong> overall effect: Z = 0.24 (P = 0.81)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 1.1 (0.4) 40 1.1 (0.4) 33.8 % 0.0 [ -0.18, 0.18 ]<br />

Subtotal (95% CI) 40 40 33.8 % 0.0 [ -0.18, 0.18 ]<br />

Heterogeneity: not applicable<br />

Test <strong>for</strong> overall effect: Z = 0.0 (P = 1.0)<br />

Total (95% CI) 252 244 100.0 % -0.01 [ -0.11, 0.09 ]<br />

Heterogeneity: Chi 2 = 6.70, df = 6 (P = 0.35); I 2 =10%<br />

Test <strong>for</strong> overall effect: Z = 0.19 (P = 0.85)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 0.02, df = 1 (P = 0.89), I 2 =0.0%<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-1 -0.5 0 0.5 1<br />

Favours treatment Favours control<br />

58


Analysis 5.1. Comparison 5 Body weight (kg), Outcome 1 All endpoint outcomes.<br />

<strong>Review</strong>: <strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong><br />

Comparison: 5 Body weight (kg)<br />

Outcome: 1 All endpoint outcomes<br />

Study or subgroup <strong>Wholegrain</strong> Control Mean Difference Weight Mean Difference<br />

1 Parallel studies (all oats)<br />

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI<br />

Johnston 1998 62 83.4 (13.8) 62 79.2 (12) 16.2 % 4.20 [ -0.35, 8.75 ]<br />

Karmally 2005 73 77.58 (12.74) 79 72.72 (11.97) 21.6 % 4.86 [ 0.92, 8.80 ]<br />

Keenan 2002 10 84.6 (18) 8 94.9 (17.2) 1.3 % -10.30 [ -26.63, 6.03 ]<br />

Pins 2002 45 83 (5.9) 43 83.4 (5.8) 56.1 % -0.40 [ -2.84, 2.04 ]<br />

Reynolds 2000 22 69.8 (13.1) 21 75.5 (14.7) 4.8 % -5.70 [ -14.04, 2.64 ]<br />

Subtotal (95% CI) 212 213 100.0 % 1.10 [ -0.73, 2.93 ]<br />

Heterogeneity: Chi 2 = 11.16, df = 4 (P = 0.02); I 2 =64%<br />

Test <strong>for</strong> overall effect: Z = 1.18 (P = 0.24)<br />

2 Crossover studies (rye)<br />

Leinonen 2000 40 72.6 (8.1) 40 72.9 (8.4) 4.7 % -0.30 [ -3.92, 3.32 ]<br />

Pereira 2002 12 84.8 (1) 12 85.1 (1) 95.3 % -0.30 [ -1.10, 0.50 ]<br />

Subtotal (95% CI) 52 52 100.0 % -0.30 [ -1.08, 0.48 ]<br />

Heterogeneity: Chi 2 = 0.00, df = 1 (P = 1.00); I 2 =0.0%<br />

Test <strong>for</strong> overall effect: Z = 0.75 (P = 0.45)<br />

Test <strong>for</strong> subgroup differences: Chi 2 = 1.90, df = 1 (P = 0.17), I 2 =47%<br />

W H A T ’ S N E W<br />

Last assessed as up-to-date: 14 January 2007.<br />

9 September 2008 Amended Converted to new review <strong>for</strong>mat.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

-10 -5 0 5 10<br />

Favours treatment Favours control<br />

59


H I S T O R Y<br />

Protocol first published: Issue 4, 2004<br />

<strong>Review</strong> first published: Issue 2, 2007<br />

15 January 2007 New citation required and conclusions have changed Substantive amendment<br />

C O N T R I B U T I O N S O F A U T H O R S<br />

Dr S Kelly<br />

Prepared and designed the protocol <strong>for</strong> the review. Developed and ran the search strategy. Organised the retrieval of papers and screened<br />

papers <strong>for</strong> inclusion and exclusion. Extracted data from included papers and took the primary role in writing the review.<br />

Dr G Frost<br />

Conceived the review and provided a methodological, policy and clinical perspective on the data.<br />

Dr C Summerbell<br />

Provided a methodological, policy and clinical perspective on the data.<br />

Dr A Brynes<br />

For purposes of dual data extraction extracted data from papers and contributed to quality assessment of studies.<br />

Mrs V Whittaker<br />

Advised on statistics, meta-analysis and quality assessment of studies.<br />

D E C L A R A T I O N S O F I N T E R E S T<br />

Gary Frost is a qualified dietitian.<br />

Carolyn Summerbell is a qualified dietitian.<br />

Audrey Brynes is a qualified dietitian<br />

S O U R C E S O F S U P P O R T<br />

Internal sources<br />

• University of Teesside, Middlesbrough, UK.<br />

• Hammersmith Hospital, UK.<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

60


External sources<br />

• No sources of support supplied<br />

I N D E X T E R M S<br />

Medical Subject Headings (MeSH)<br />

∗ Cereals; Cholesterol [blood]; Cholesterol, LDL [blood]; Coronary Disease [blood; ∗ diet therapy]; Randomized Controlled Trials as<br />

Topic; Risk Factors<br />

MeSH check words<br />

Humans<br />

<strong>Wholegrain</strong> <strong>cereals</strong> <strong>for</strong> <strong>coronary</strong> <strong>heart</strong> <strong>disease</strong> (<strong>Review</strong>)<br />

Copyright © 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.<br />

61

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!