Themes

Epidemiology of ageing:
Healthy old age only for the
socially well-off? well off?
Prof. Dorly J.H. Deeg, PhD
Department of Psychiatry and
Institute for Research in Extramural Medicine
VU University Medical Centre, Amsterdam, The Netherlands

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

%
Survival to age 1, birth cohorts 1900-1937
Males and females, The Netherlands
100
90
80
70
60
50
1900 '10 '20 '30
Year of birth
Source: Nethderlands Interdisciplinary Demographic Institute
Men
Women

%
100
Survival to age 50, birth cohorts 1907-1937
Males and females, The Netherlands
90
80
70
60
50
1910 1925
Year of birth
Source: Netherlands Interdisciplinary Demographic Institute
Men
Women

Remaining years of life
25
20
15
10
5
0
Bron: CBS Statline 2003
Development of the life expectancy
from age 65 and 80 years
'50 '60 '70 '80 '90 '00
Jaar
Men 65
Women 65
Men 80
Women 80

Change over time
Population trends: compression of mortality across time
Before/after 1950: reduction mortality ~ social/medical

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and selfperceived
health
Use of health care services
Shifts (trends) in functional limitations and in
care services supply

Longitudinal Aging Study Amsterdam
Random sample
3107 men and women
Ages 55-85
Start 1992
3-year intervals
North Sea
Amsterdam
Belgium
Oss
Zwolle
Germany

Independent
variables
Coping
The LASA-model
Objective
functioning
Standards / aspirations
Subjective
functioning
Contributions /
needs:
- participation
- use of care
General
well-being
Mortality

Development of health problems
Disability
Limitation
Impairment
Disease

Chronic physical diseases and Age
80
70
60
50
40
30
20
10
0
% no chronic
diseases
% > 1 chronic
diseases
55-64 65-74 75-84 85-94
Source: Longitudinal Aging Study Amsterdam, 2002

Main chronic conditions
Heart diseases (Myocardial Infarction,
Angina Pectoris, Heart Failure,
Atherosclerosis)
Diabetes (type II)
Lung Diseases (Asthma, COPD)
Cancer
Joint disorders (Osteo-arthritis)
Psychiatric complaints (Depression,
Anxiety, Dementia)

%
60
50
40
30
20
10
0
*
Heart
Chronic diseases 65+
*
Diabetes
*
Lung
Cancer
Joints
Psych
2001
Source: Longitudinal Aging Study Amsterdam * expected increase >10% to 2020 (NPHP)
?

Female/male difference in disaease prevalence 65+
f / m
Heart 0,6
Diabetes 1,5
Lung 1,0
Cancer 1,0
Joints 2,4
Psychiatric 1,7

35
30
25
20
15
10
5
0
Life expectancy at age 65, 1993
Average
Lung/Cancer
Heart diseases
Joint disorders
Source: Longitudinal Aging Study Amsterdam
Healthy
Men
Women

Why sex differences?
Biological differences: genetic, hormonal
(estrogens)
Risks during the life course
(labour, smoking)
Use of health care
Coping with disease (disease behaviour)
Reporting behaviour

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

• Definition:
Disability
√ Inability to carry out the usual tasks of daily living
√ Gap between capacity of individual and demands of
environment
• Powerful measure of health in older age
√ Associated with most chronic conditions (e.g., stroke,
arthritis, diabetes)
√ Independent predictor of mortality

80
60
40
20
Change in functional limitations with aging
0
70-74 75-59 80-84 85-89
Source: Longitudinal Aging Study Amsterdam
Men
Women

3-year changes in functional limitations
55-64 65-74 75-85 yrs
Worse 18% 30% 50%
Better 12% 9% 7%
Source: Longitudinal Aging Study Amsterdam

Chronic conditions: consequences
0,2
0,15
0,1
0,05
0
Heart
Diabetes
Lung
Cancer
Joints
Psych
Physical Limitations Mortality in 1,5 years
Source: Longitudinal Aging Study Amsterdam

Special groups
• Low education/low income
• Living alone
• Urban areas
↔
• Higher mortality
• More health problems

20
18
16
14
12
10
8
Life expectancy at 65 years
Elementary
by level of education
Lower-middle
Source: TNO-PG (ERGO, GLOBE, LASA)
Higher-middle
Higher
Men
Women

100
90
80
70
60
50
% disability-free life expectancy
at 65 years, by level of education
Elementary
Lower-middle
Source: TNO-PG (ERGO, GLOBE, LASA)
Higher-middle
Higher
Men
Women

Social determinants
Associated with lower risk of disability:
• Higher education: life-long asset;
knowledge, living standard
• Having a partner: life-long asset but can be lost;
support, self-esteem, living standard, information
… independent of age, gender, chronic conditions

Cluster analysis
Trajectories of disability
In surviving subjects (n=1988):
limitation scores T 1 , T 2 , T 3 --> 5 course types
In subjects who died between T 2 , T 3 (n=325):
limitation scores T 1 , T 2 --> 2 course types
In subjects who survived between T 1 , T 2 (n=417):
limitation score T 1 --> 1 course type

Summary of trajectories
Compression of disability
‘So far so good’
Stable no disability 53%
‘Terminal drop’
No limitations, died 18%
Late increase in disability 5%
Expansion of disability
All others 24%

70
60
50
40
30
20
10
0
%
Education
high
Social determinants
Stable no disability Terminal drop Expansion
* *
Education
low
Partner No partner
* Distribution of trajectories significantly different (p < 0.05); adjusted for age, sex and other factor

70
60
50
40
30
20
10
0
Interaction of social determinants
Education
high
Stable no disability Terminal drop Expansion
Partner No partner
* *
Education
low
* Distribution of trajectories significantly different (p < 0.05)
Education
high
Education
low

Conclusions (1)
• Increase in life expectancy since 1950 largely due to
medical advances
• Multimorbidity is rule more than exception in older age
• Joint disorders most prevalent, especially in women
• Psychiatric disorders second in prevalence, first in
disabling consequences
• Gender- and educational differences in (healthy) life
expectancy
• Most unfavourable trajectories of disability among
lower educated persons without partner

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

Change over time
• Population trends: compression versus expansion of
disability across time:
√ Compression: disability decreases faster than mortality or
disability increases slower than mortality
√ Expansion: disability decreases slower than mortality or
disability increases faster than mortality

Health status in 1993 compared to 1956
1956 (reference year)
Mortality
>2 diseases
Physical limitations
Fair/poor self-rated ealth
Men Women
1.00
0.80
1.63
1.23
2.25
1.00
0.40
1.42
0.78
1.46
Sources:
Dutch Longitudinal Study among the Elderly, Longitudinal Aging Study Amsterdam
(Deeg et al 1994)

1992/93
1995/96
1998/99
LASA time schedule (1)
Baseline cycle T 1
Second cycle T 2
Third cycle T 3
n=3107
n=2545
n=2076
2001/02 Fourth cycle T4 n=1691
Longitudinal change

1992/93
1995/96
1998/99
LASA time schedule (2)
Baseline cycle T 1
Second cycle T 2
Third cycle T 3
n=3107
n=2545
n=2076
2001/02 Fourth cycle T4 n=1691
Trend shift

Number
2,5
2
1,5
1
0,5
0
Chronic diseases 65+, 1993-2002
Source: Longitudinal Aging Study Amsterdam
65-74 75-85
1993
1996
1999
2002

%
25
20
15
10
5
0
National Long Term Care Survey (Spillman 2004)
Trend in disability 65+,
United States, 1984-1999
1984 1989 1994 1999
Total
limitations
Severe
limitations

%
40
35
30
25
20
15
10
5
0
Trend in disability 65+,
The Netherlands, 1993-2002
1992-93 1995-96 1998-99 2001-02
Longitudinal Aging Study Amsterdam (Portrait 2003)
Total
limitations
Severe
limitations

1992/93
1995/96
1998/99
LASA time schedule (3)
Baseline cycle T 1
Second cycle T 2
Third cycle T 3
n=3107
n=2545
n=2076
2001/02 Fourth cycle T4 n=1691
2002/03 Baseline new n=1002
Trend shift

12
10
8
6
4
2
0
Trends in lung diseases, 55-64-years
Men
1992/93
* adjusted for age
Brinkkemper 2005
Men
2002/03
Women
1992/93
Women
2002/03
Men 1992/93
Men 2002/03
Women 1992/93
Women 2002/03

12
10
8
6
4
2
0
Men
1992/93
* adjusted for age
Brinkkemper 2005
Trends in diabetes, 55-64-years
Men
2002/03
Women
1992/93
Women
2002/03
Men 1992/93
Men 2002/03
Women 1992/93
Women 2002/03

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Trends in body weight, 55-64-years
Men
1992/93
* adjusted for age, education
Visser et al. 2005
Men
2002/03
Women
1992/93
Women
2002/03
Normal weight
Overweight
Obesity

Mi
nu
te
n
pe
r
da
Trends in physical activity, 55-64-years
250
200
150
100
50
0
61
45
17
* adjusted for age, education
Visser et al. 2005
58
33
176
42
138
29
9 12 11
23 18 18 15
Mannen
1992/93
Mannen
2002/03
*
*
*
Vrouwen
1992/93
Vrouwen
2002/03
*
*
sports
biking
walking
household
activities

Ages: 65-84
Trends 1992-2002: Education
Education:
Low Middle High
Year: * * *
. 1992-3 50 39 11
. 1995-6 44 44 12
. 1998-9 40 47 14
. 2001-2 36 48 15
Note: Prevalences are weighted to the general population
* Significant trend

Trends 1992-2002: Education
Ages: 55-64
Education:
Low Middle High
Year: * * *
. 1992-3 32 51 17
. 2002-3 21 55 24
Note: Prevalences are weighted to the general population
* Significant trend

Explanations
Increasing level of education =>
Healthier cohorts

Explanations for mixed changes over time
• Earlier diagnosis of chronic conditions
• Better disease management
• Increase in mild, decrease in severe
limitations
• Higher expectations of medical science
• Higher norms of good health
• Poorer life style

Conclusions (2)
• Cohort explanation unlikely to account for full
trends in prevalence of diseases and
limitations
• Period effects seem larger part of explanation
• Either explanation leads to very different
projections

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

Source: CBS
Hospital admission by age

Distribution of diseases in persons 65+ without (red)
and with (blue) recent hospital admission
45
40
35
30
25
20
15
10
5
0
%
0 1 2 >2 diseases 0 1 2 >2 diseases
number of diseases
Source: Deeg & Broese van Groenou 2007

Longitudinal Aging Study Amsterdam
Hospital admission: reasons
Most important reasons for admission:
Cardiovascular diseases 25% L+
Musculoskeletal disorders 17% L+
Eye disorders 13%
Genital organs (men) 10%
Malignancies 7% L+
Accidents 7% L+
Digestive organs 6%
* in past 6 months
L+: relatively severely limiting

%
70
60
50
40
30
20
10
0
1992-93
Longitudinal Aging Study Amsterdam
Post-hospital disability
1995-96
1998-99
2001-02
Total
limitations
Severe
limitations

%
Trend in care services supply, Netherlands
20
18
16
14
12
10
8
6
4
2
0
1992 1993 1995 1996 1998 1999
Statistics Netherlands (Portrait 2003)
Home care
workers /
1000
Hospital
days / 1000

%
50
40
30
20
10
0
Longitudinal Aging Study Amsterdam
*
1992-93
Post-hospital use of care (1)
*
1995-96
*
1998-99
*
*
2001-02
Care by partner
if partner
Care by partner
total
Subsidised
home care
* significant difference with non-hospital admissions

Longitudinal Aging Study Amsterdam
Post-hospital use of care (2)
No change in:
• Informal care by others, stayed 13%
• Private care, stayed 18%
Change in:
• Subsidised home care: 23% – 14%
• Care by partner: 37% – 47%
• Dissatisfaction about care received
increased: 16% – 27%

Conclusions (3)
• At population level, increase in mild (not severe)
limitations
• Happens at the same time as decrease in care supply
• Most expensive care service, hospital days,
decreased
• Less subsidised home care; more care by partner
after hospital admission
• Older couples are group at risk

Population ageing
Chronic diseases
Themes
Consequences: functional limitations and
disability
Use of health care services and equipment
Shifts (trends) in functional limitations and in
care services supply

Background
• United States a.o.: Decreasing trends in
disability rates in older age
• Decreases in disability partly attributed to
increases in use of assistive devices
• The Netherlands: Contradictory evidence on
trends in disability rates at older ages
• LASA: increasing disability rates, despite
increases in use of assistive devices

Research question
Seemingly contradictory trends -
explanation?
Different trends for disability and device
use across subpopulations?

1992/93
1995/96
1998/99
LASA time schedule (3)
Baseline cycle T 1
Second cycle T 2
Third cycle T 3
n=3107
n=2545
n=2076
2001/02 Fourth cycle T4 n=1691
2002/03 Baseline new n=1002
Trend shift

For each disability item if:
Assistance
no difficulty / difficulty / only with help
Assistive device use: no - yes
Human assistance: no / seldom / sometimes - often / always

% Device use in 1992 and 2002, ages 55-64
1992 2002
No disability 2.6 2.8
Mild disability 19.3 27.5
Severe disability 24.1* 33.5*
Men 5.5 8.3
Women 6.2 12.6*
Low education 8.5* 13.8*
High education 2.8 8.2
Partner 5.3 9.3
No partner 8.4 16.8*
Longitudinal Aging Study Amsterdam ; sample 2002 weighted to 1992; * association with covariate p < 0.05
age and chronic diseases not independently associated

Compative analysis
Logistic regression analysis -> OR
• Dummy for Cohort: 0 = 1992, 1 = 2002
test of OR(Cohort) -> trend
• Adjustment for covariates
• Interaction Cohort*Covariate one by one
test of Interaction -> differential trend

Trend in disability & device use 1992-02,
adjusted OR
Mild Severe Device
Total 1.85 1.71 1.71
Men 1.36 1.05
Women 2.39 2.48
Partner 2.38 1.21
No partner 0.84 2.53
Low education 1.02
High education 2.57
Longitudinal Aging Study Amsterdam; bold italic: trend significant; bold italic: interaction significant

Conclusion on trends
• Increase in disability, mild more than severe
• Increase in assistive device use
• Increase in disability in women and partnered
• Increase in device use in higher educated and
partnerless

Explanations (?)
Education/gender discrepancy in device use:
• Higher educated better manage their disability
- if disabled
• Higher educated have higher norms of good
health
• Lower educated and women have less money
to spend than higher educated and men

Policy on ageing
A valid policy on ageing should be based
on:
•a correct understanding of the various
groups of older persons at risk in our society,
and
•a correct understanding of the macroinfluences
on the chain from disease to care
utilisation

Hendrikje Van Andel,
died 115 years old
Two oldest ladies
Jeanne Calment,
died 122 years old