A Life Course Perspective on Social Inequalities in Health

1
A Life Course Perspective on Social Inequalities
in Health

• David Blane
• with Juliet Stone, Gopalakrishnan Netuveli

Imperial College London
ESRC International Centre for Life Course Studies
in Society Health
2
Contents

• Origins
• Theory
• Methods
• Social inequalities in health
• Limits of the life course approach
• Summary

3
Origins
1990s

• The life course now is a core theme in social
  epidemiology.

4
Biological programming
Origins

• Organ development in utero (as indexed by eg.
  birth weight) and during infancy determines
  maximum function during adulthood
• Examples
• Lung development adult COPD
• Kidney development adult hypertension
• Pancreatic development adult diabetes

Barker D 1991, 1994 (BMJ Books).
5
Birth cohort studies
Origins

• 1946 birth cohort at age 36 years
• first use of accumulation
• risk of lower respiratory tract disease reduced
  lung function at 36 accumulates with chest
  infections and poor, crowded housing during early
  childhood, air pollution exposure during later
  childhood and tobacco smoking during early
  adulthood

Mann et al. 1992 J Epidemiol Com Health

6
Health inequalities
Origins

• Behavioural risk factors account for only c.1/3
  of class difference in mortality risk.
• Mortality risk is fine grained.
• Social structure disadvantages (or advantages)
  cluster cross-sectionally and accumulate
  longitudinally.

Rose Marmot 1981 Brit Heart J
Goldblatt 1990 HMSO Blane 1995 American J
Public Health
7
Theory Models of the life course

• Critical periods
• Accumulation
• Pathways

8
Model Critical periods

• Extends the idea of biological programming to
  include
• Childhood
• Psycho-social stress at the time of brain
  maturation may both inhibit child growth and
  mis-set the developing BP control mechanisms,
  producing later high BP
• Social development
• Key social transitions
• Montgomery et al. 2000 Archives Disease
  Childhood
• Bartley et al. 1997 British Medical Journal

9
Model Accumulation

• Disadvantages, or advantages, tend to cluster
  cross-sectionally
• occupation residence area of residence
  consumption
• and accumulate longitudinally.
• childhood adulthood older ages
• This social process may have a major impact on
  health through the accumulation of numerous
  relatively minor effects.
• Blane et al. 1997 European J Public Health

10
Model Pathways

• Early advantage or disadvantage sets a person on
  a pathway to a later exposure that is the
  aetiologically important event.
• Educationally successful women (pathway) tend to
  delay their first pregnancy (aetiologically
  important event), which increases their risk of
  breast cancer.
• Power Hertzman. 1997 British Medical
  Bulletin

11
Models A judgement

• Models are difficult to distinguish empirically
  and conceptually
• Perhaps best to see accumulation as the general
  social process which drives life course
  trajectories with critical periods and pathways,
  in addition to accumulation, being the biological
  processes of disease causation
• Hallqvist et al.2004 Social Science and
  Medicine
• Blane et al. 2007 Revue dEpidemiologie et de
  Sante Publique

12
Methods

• Birth cohort studies
• Linked-register data sets
• Epidemiological archaeology

historical study tracing to present-day
location retrospective data /- Lifegrid
(event history calendar)
13
Birth cohort studies
Methods

• 1946 birth cohort
• Health at 36 years and physical disability and
  handicap at 43 influenced by parental social
  class, health during childhood and own adult
  social class
• Poor diet and obesity at 36 years
  influenced by manual parental social class and
  few educational qualifications

Kuh Wadsworth 1993 Soc Sci Med
Kuh et al. 1994 J Epid Com Health Braddon
et al. 1988 J Epid Com Hlth Braddon et al.
1986 Brit Med J
14
Linked registers
Methods

• Birth weight and blood pressure at age 50 years
  in 1300 Swedish men
• weak inverse linear relationship, only for
  systolic pressure
• Birth weight during 1915-29 and all deaths among
  14600 Swedish men women to 1995
• weak inverse relationship for cardiovascular
  disease deaths, only for men

Leon et al. 1996 British Medical J Leon
et al. 1998 British Medical J
15
Epidemiological archaeology
Methods

• Boyd Orr lifegrid sub-sample
• Height measured during childhood in 1937-39
• Height and blood pressure measured during early
  old age in 1997-98
• Child growth (child height conditioned on adult
  height) predicted pulse pressure and systolic
  blood pressure 60 years later.

Montgomery et al. 2000 Arch Dis Child
16
Health inequalities and life course

• Predictive power
• Aetiological insights
• Health inequality debates
• Social policy implications

17
Predictive power
Inequalities

• West of Scotland Collaborative Study
• Life course socioeconomic position (child, first
  job and adult manual social class, range 0-3)
  predicted systolic and diastolic blood pressures,
  serum cholesterol concentration, height, body
  mass index, lung function (FEV1), symptoms of
  angina and chronic bronchitis and 21-year
  mortality risk.

Davey Smith et al. 1998 Brit Med J
18
Aetiological insights
Inequalities

• West of Scotland Collaborative Study
• cause-specific mortality and mutually adjusted
  child adult social class
• Lung cancer mostly adult class
• Coronary heart disease and respiratory disease
  accumulating child and adult class
• Stroke and stomach cancer mostly child class

Davey Smith et al. 1998 Brit Med J
19
Unresolved debates
Inequalities

• General susceptibility
• Mortality risk determined by balance of
  advantage and disadvantage
• Cause of death determined by specifics of an
  individuals social trajectory
• Gradient constraint
• Health-related social mobility constrains, rather
  than creates, widening inequalities

Blane et al. 1997 European J Pub Health
Bartley Plewis 1997 J Hlth Social Beh
20
Social policy implications
Inequalities

• Traditional welfare state safety nets assume that
  misfortune soon will be replaced by the persons
  normal, more advantaged, state.
• The accumulation model draws attention to the
  likelihood that misfortune will have been
  preceded by earlier disadvantages, requiring a
  springboard to repair previous damage.

Bartley et al. 1997 British Medical J
21
Limits of the life course approach.

• Spectrum of impact.
• Major social disruption.
• Effect diluted at older ages? (speculative)

22
Spectrum of impact
Limits

• Physiological risk factors influenced by whole
  life course behavioural risk factors influenced
  mainly by current context
• Modest life course influences on diet at older
  ages
• Quality of life at older ages influenced mainly
  by current circumstances

Blane et al. 1996 British Medical J
Maynard et al. 2006 Eur J Pub Hlth Wiggins et
al. 2004 Ageing Society
23
Major social disruption
Limits

• Many areas of health where life course
  perspective adds little
• Not an alternative to a concern with the
  cross-sectional and immediate
• Acute shortening of life expectancy that followed
  end of USSR in 1989 (life course might help
  explain particular vulnerability of single men)

Watson 1995 Social Science and Medicine
24
Dilution at older ages?
Limits

• Norwegian linked registers
• 19,000 people aged 70 years in 1990, traced back
  to 1960 and forward to all deaths during
  1990-1998
• Social gradient in mortality explained mostly by
  social position in 1990, with minor cumulative
  effects 1960-1980
• Is life course effect weakened at older ages?

Naess et al. 2006 Social Science Med
25
Summary

• Life course approach not relevant to all aspects
  of health inequalities.
• Most relevant to inequality in chronic disease
  morbidity and mortality.
• Accumulation is main social process and one of
  the three aetiological processes.

26
Follow-up contact

• E-mail d.blane_at_imperial.ac.uk
• Website of ESRC International Centre for Life
  Course Studies in Society and Health
  www.ucl.ac.uk/icls.