Intended for healthcare professionals

Education And Debate

Socioeconomic determinants of health: Health and the life course: why safety nets matter

BMJ 1997; 314 doi: (Published 19 April 1997) Cite this as: BMJ 1997;314:1194
  1. Mel Bartley, principal research fellowa,
  2. David Blane, senior lecturerb,
  3. Scott Montgomery, research fellowc
  1. a Department of Epidemiology and Public Health, University College London Medical School, London WC1E 6BT
  2. b Academic Department of Psychiatry, Charing Cross and Westminster Medical School, London W6 8RP
  3. c University Department of Medicine, Royal Free Hospital Medical School, London NW3 2QG


    This article argues that a life course approach is necessary to understand social variations in health. This is needed in order to take into account the complex ways in which biological risk interacts with economic, social, and psychological factors in the development of chronic disease. Such an approach reveals biological and social “critical periods” during which social policies that will defend individuals against an accumulation of risk are particularly important. In many ways, the authors of modern welfare states were implicitly addressing these issues, and the contribution of these policies to present day high standards of health in developed countries should not be ignored.

    Health, inequality, and the life course

    Fig 1
    Fig 1

    Manual labourers are more likely to experience job insecurity and hazards at work


    In the pioneering days of public health the inhabitants of urban slums were subject to malnutrition, diarrhoeal diseases, and endemic tuberculosis, and infectious diseases were the major causes of early death. Today the main causes of death are different, yet their social distribution has changed surprisingly little.1 Whereas the early public health physicians had to struggle to increase the spatial coverage of reforms such as clean water and safe disposal of sewerage, today's challenge is to persuade policy makers that health in middle and old age depends as much on past circumstances as on present ones.2 Living conditions cannot simply be left to fluctuate as people pass through childhood and their reproductive and working years and into old age, because health and quality of life at any one stage is affected by prior circumstances and events.

    Among researchers there is growing acceptance that health and its social distribution need to be studied over the whole of the life course.3 4 This is partly because most adult diseases have long courses of development and complex aetiologies, which in some cases may begin in utero.5 Most of the prevalent environmental hazards, such as tobacco smoke and atmospheric pollution, cause their damage slowly and usually require decades of exposure to produce disability and premature death. Also, the social distribution of mortality is too finely graded to be produced solely by short term factors, whether biological or socioeconomic. The step by step increase in risk of early death shown by the registrar general's social classes, a relatively crude measure of social circumstances, is well known. More recent reports have shown finer differences in risk, for example, between civil service grades,6 managers in small and large firms,7 and members of households with two cars rather than one.8 Similar stepwise gradations are seen in geographical areas classified according to the average income of inhabitants in the United States9 or the income levels of mortgage holders in the United Kingdom.10 These patterns do not indicate a sharp “health divide” between rich and poor but point to differences in risk between people with average incomes, those who are moderately well off, and very affluent groups. Whereas short term differences in material and social conditions are important–and obviously decisive in relation to certain causes of death such as accidents–other differences do seem to require an understanding of patterns of exposure to both physical and psychosocial hazards over the longer term. The nature of these processes has already begun to be illustrated by work on the oldest of the British cohort studies.11

    Long serving clinicians will have often observed the ways in which health and social circumstances interact in the lives of their patients and the processes by which advantages or disadvantages accumulate over time. Systematic study of these processes has become possible only recently. Several excellent longitudinal studies have collected a range of information about health and socioeconomic conditions. These include a 1% sample of the population of England and Wales followed over more than 20 years; representative samples of all British births in the years 1946, 1958, and 1970 followed up at regular intervals between birth and the present time; and a number of smaller longitudinal studies.12 13 14 Advances in computing and in statistical methods make it possible to store, access, and analyse continuities and change in health and social circumstances in these large and complex longitudinal datasets.

    Social structure and the life course

    The life course may be regarded as combining biological and social elements which interact with each other. Individuals' biological development takes place within a social context which structures their life chances, so that advantages and disadvantages tend to cluster cross sectionally and accumulate longitudinally. For example, children of less affluent families are more likely to experience failure at school,15 to find work in the more disadvantaged sectors of the labour market,16 and to experience unemployment early in their working lives.17 In addition, less affluent families are more likely to produce babies of lower birth weight.18 Even compared with others in the same social class, low birthweight children have been found to have an increased risk, which is socially structured, of socioeconomic disadvantage during childhood and adolescence19 and an increased risk of chronic disease in middle age, which may be biologically programmed.5

    There may be “critical periods” both for the development of organ systems and for psychological and social development, though their existence and nature are still controversial. The development of the neural tube in very early pregnancy is an accepted example. Although this is a biological process, the difficulty in a low income household of maintaining an intake of vitamins adequate not just for daily life but also for this critical period during pregnancy produces sharp social gradients in neural tube defects.20 The transition from school to work may be regarded as having similar importance for social development.21 22 People who enter less well paid employment are also more likely to encounter work insecurity and physical and chemical hazards at work (table 1),27 to live in less well constructed housing in more polluted neighbourhoods, and to retire on no more than the basic state pension.28 At each stage, social and economic disadvantage can push the individual another step down an aetiological pathway towards established chronic illness.

    Table 1

    Cross sectional accumulation of labour market disadvantage: men's occupations 1979-87. Values are percentages

    View this table:
    Fig 2
    Fig 2

    Height at age 7 and unemployment at age 22-32. Adjusted for social class at birth, crowding, qualifications, region, Bristol social adjustment guide score, and parental height

    Another closely related process is that by which social and economic disadvantage increase the impact of illness, regardless of how the illness is acquired. For example, Dutch children with poor health were found to do less well educationally only if they came from less privileged social groups.29 At a time of low unemployment (1973) nearly 90% of British professionals and managers with limiting chronic illness were nevertheless in paid employment, as were 70% of people with chronic illness in less skilled manual work; in the economic recession of 1993, the employment rate among professionals and managers with chronic illness fell to 78%, but in semiskilled and unskilled manual workers it plummeted to 42%.30

    Cohort data allow different “life trajectories” to be distinguished and related to disease risk.31 This is beginning to show concretely the ways in which the accumulation of disadvantage over the life course is linked to health in adulthood.11 32 Risk factors for cardiovascular disease in adult life have been found to be linked in varying strengths to both childhood and adulthood socioeconomic position.33 The risk of mortality from all major causes has been shown to increase in a stepwise fashion with the amount of time spent in manual (as opposed to non-manual) occupations34 and in residential conditions of a low standard.35 Poor growth during early childhood, which itself may be more likely where there is material or psychosocial adversity,36 has been linked to poorer health in adult life.37 Slow growth to the age of 7 years has been associated with an increased risk of unemployment in young men regardless of their adult stature,38 thus producing a complex pattern of continuity between biological and psychosocial disadvantage (fig 2).

    Policy implications

    Policy makers have not been entirely immune to warnings of rising health inequality. In May 1994 the British government established an interdepartmental working group chaired by the deputy chief medical officer for England. Its brief was to investigate the ways in which biological, social, environmental, cultural and behavioural factors interrelate to produce social variation in health and to explore the possibilities for effective intervention.39 The working group's report set a research agenda which lays emphasis on identifying and accurately measuring those factors which combine over the life course to produce the observed variations.4

    How does a life course approach help in understanding the health implications of social and economic policies, and how does it relate to recent findings regarding the relation of income distribution to population life expectancy?40 41 A plausible mechanism involves the ways in which social and economic policies affect the transition through “socially critical periods” (box). During such periods as the entry into parenthood and the transitions from the parental home to the outside world, from school to work, from one job to another, and into retirement, levels of income support and availability of publicly funded services influence the degree of insecurity and uncertainty experienced by individuals and families.42 This can have effects of both material and psychosocial kinds: by preventing dramatic falls in living standards and by a wider effect on the degree to which citizens experience a sense of control over their lives.

    Critical periods in human development

    • Transition from primary to secondary school

    • School examinations

    • Entry to labour market

    • Leaving parental home

    • Establishing own residence

    • Transition to parenthood

    • Job insecurity, change, or loss

    • Onset of chronic illness

    • Exit from labour market

    The Swedish policy analyst Gosta Esping-Anderson has examined national variations in the approach to these critical periods. He classifies policies according to their “degrees of commodification”–that is, the extent to which policies protect individuals at times when they are unable to earn an adequate wage in the open labour market. One measure he uses is the extent to which supplementary income such as unemployment, sickness, or disability benefits available during various periods of non-earning approximates the average wage for those in work.43 The obvious effect of more generous benefit levels on the overall distribution of income would be to even it out; the effect on individuals is to make it less likely that periods of inability to earn are accompanied by high risks of serious material shortfall in terms of diet, heating, and housing quality. Even for those enjoying periods of employment and material sufficiency, redistributive policies create a more stable psychosocial environment. Research indicates that it is not only adverse life events themselves which affect health,44 but the anticipation of adversity,45 46 47 48 and that this effect is mitigated where life changes do not have adverse financial implications.49 More equitable social and economic policies may therefore be effective in preventing an accumulation of disadvantage, the situation where having had one lot of bad luck increases both the risk and the fear of other types of misfortune.

    Keeping public health on the political agenda is made more difficult by the fact that chronic illness is not epidemic and does not seem to spread out from poor people to rich people. The research reviewed here implies that this is a misapprehension exemplifying a form of biological “short termism.” The redistributive policies implemented as part of the Beveridge reforms in postwar Britain were not explicitly designed with public health in mind. Yet during this period there have been major improvements in life expectancy.1 These improvements have, however, been far greater among non-manual and skilled manual groups than less skilled groups. Social and health policy analysts from Richard Titmuss onwards have repeatedly shown the ways in which, paradoxically, the welfare state was of most benefit to the middle class and “labour aristocracy.” It provided stable jobs with good pensions in the new welfare bureaucracies, as well as free services such as education and health care which these groups were best able to use during their own critical life transitions.50 Life course research is showing us the extent to which health at older ages has been affected by the impact of policy measures on different sections of the population, as cohorts move through time and encounter the life transitions which, at some stage, affect them all.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    21. 21.
    22. 22.
    23. 23.
    24. 24.
    25. 25.
    26. 26.
    27. 27.
    28. 28.
    29. 29.
    30. 30.
    31. 31.
    32. 32.
    33. 33.
    34. 34.
    35. 35.
    36. 36.
    37. 37.
    38. 38.
    39. 39.
    40. 40.
    41. 41.
    42. 42.
    43. 43.
    44. 44.
    45. 45.
    46. 46.
    47. 47.
    48. 48.
    49. 49.
    50. 50.
    View Abstract