Elsevier

Social Science & Medicine

Volume 53, Issue 12, December 2001, Pages 1575-1585
Social Science & Medicine

Using an interactive framework of society and lifecourse to explain self-rated health in early adulthood

https://doi.org/10.1016/S0277-9536(00)00437-8Get rights and content

Abstract

This paper presents an integrated model of the determinants of adult health combining lifecourse factors and contemporary circumstances. Using the 1958 British Birth Cohort, it operationalises lifecourse influences in terms of factors from birth to age 33, which might act through latent, pathway, or cumulative effects. Contemporary circumstances are represented by variables at different levels of social aggregation: macro (socio-economic circumstances); meso (involvement in civil society functions); micro (personal social support); and intersecting (job insecurity and life control). Multiple regression models were fitted, using self-rated health at age 33 as the health outcome. To allow for temporal ordering of events, early life factors were entered first in the final model, followed by later childhood factors and, finally current factors. Self-rated health was predicted by variables representing both early and later stage of the lifecourse and also contemporary societal-level factors. The effects of childhood factors were not removed by including contemporary factors, and conversely, contemporary factors contributed to the prediction of self-rated health over and above lifecourse factors. The factors were not collinear; supporting the notion that each dimension was distinct from the others. Although the model accounted for only 9% of the variance in self-rated health, the general conclusion is that both lifecourse and contemporary circumstances should be considered together in explaining adult health.

Introduction

The study of the social determinants of health involves two broad perspectives: one based upon the lifecourse and the other emphasising contemporary life circumstances. Within these broad perspectives, alternative models have been proposed connecting specific factors with adult health. What is needed is a unified framework that consolidates these models and integrates the two perspectives.

With regard to the lifecourse perspective, investigators have postulated three processes whereby early life environments may affect adult health: first, latent effects by which early life environment affects adult health independent of intervening experience; second, pathway effects, through which early life environment sets individuals onto life trajectories that in turn affect health status over time; and, third, cumulative effects whereby the intensity and duration of exposure of unfavourable environments adversely affects health status, according to a dose–response relationship (Kuh & Ben-Shlomo, 1997; Marmot & Wadsworth, 1997; Power & Hertzman, 1997; Davey-Smith, Hart, Blane, Gillis, & Hawthorne, 1997; Lynch, Kaplan, & Shema, 1997). The essence of the latency model is that specific biological (e.g. low birthweight) or developmental (e.g. visual acuity) factors at sensitive periods in (early) life have a lifelong impact on health and well-being, regardless of subsequent living conditions. It has been argued that the link between low birthweight and cardiovascular disease in adulthood is evidence of a latency effect (Barker, 1992). Similarly, the results from early childhood stimulation programs for disadvantaged children are consistent with a latency effect, given their effectiveness in improving adult outcomes even without any attempt to provide them with special help in the intervening years (Palmer, 1979; Schweinhart, Barnes, & Weikart, 1993).

In practice, latency effects can be difficult to disentangle from pathway effects. This is because the pathways model acknowledges that differences in early life environment may direct children onto different life courses. To illustrate, stimulation, stability and security in early childhood affect the child's readiness for schooling (Case & Griffin, 1991). In turn, lack of school readiness leads to an increased risk of behavioural problems and, also, to school failure (Pulkkinen & Tremblay, 1992). Behavioural problems and failure in school lead to low levels of mental well-being in early adulthood (Power, Manor, & Fox, 1991). Meanwhile, the status of one's parents helps to determine the community where one grows up, which, by the early school years, starts to influence the child's life chances through the social networks, community values, and opportunities which present themselves (Haan, Kaplan, & Camacho, 1987).

The third process linking early life environment and adult health recognizes the importance of cumulative effects, wherein the focus is on the accumulation of advantage or disadvantage over time, based upon the duration and intensity of exposure to the factor(s) of interest. For instance, a cumulative effect of income is suggested by the stronger association with mortality found for earnings over several years than for single-year earnings (McDonough, Duncan, Williams, & House, 1997). With respect to socio-economic circumstances, it was shown that mortality risk in a prospective study of Scottish men was graded by cumulative social class, comprising class of origin, at labour market entry, and in later adulthood (Davey-Smith et al., 1997). Cumulative and pathway effects have also been established in the 1958 cohort (Power, Manor, & Mathews, 1999): socio-economic conditions from birth to age 33 had a cumulative effect on self-rated health in addition to that of level of education achieved (a pathway effect). Simultaneous examination of education and cumulative soico-economic conditions showed independent associations, though with slightly diminished effects after adjusting for the other factor.

As explained above, there is in addition to the lifecourse perspective, a major focus on contemporary circumstances as determinants of adult health. These can be organized according to three levels of social aggregation. At the most “macro” or broadest level there is the national socioeconomic environment. The principal determinants of health at this level are income per capita and how equitably it is distributed (Wilkinson, 1997; Kaplan, Pamuk, Lynch, Cohen, & Balfour, 1996). Among countries with per capita income below $10–15,000 US, increasing national income strongly correlates with increasing health status (World Bank, 1993); above this level, equity of income distribution matters more than variations in national income per se.

At the “meso” or civil society level, there are a series of factors that include voluntarism; social affiliation, trust, and cohesion; and the capacity of important social institutions to respond to current and changing human needs. Variables as diverse as psychosocial work characteristics and the level of membership in service organizations can be regarded as aspects of civil society. This construct has been studied by different investigators (Putnam, 1993; Rose, 1995; Kaplan et al., 1996; Kawachi, Kennedy, Lochner, & Prothrow-Stith, 1997), often under the label “social capital”. We avoid this term because it is not clear whether the characteristics of “capital”, as understood by economists, applies here. Finally, at the most “micro” level, there are the determinants of health associated with private life (Berkman, 1995): the quality of intimate relationships, access to social support, and the availability of informal help to solve the problems of daily life.

To date, it has not been resolved as to how the lifecourse perspective fits with models of the determinants of health that focus exclusively on contemporary adult circumstances. Our purpose here is to propose an integrated model of the determinants of adult health that combines lifecourse and contemporary circumstances. Fig. 1 presents our conceptual framework. It represents contemporary circumstances at three levels of social aggregation (macro, meso and micro) in the form of a bullseye, while representing the individual lifecourse as an arrow passing through the bullseye at an arbitrary lifestage. The arrow runs left to right, from gestation to old age, subsuming latent, pathway and cumulative effects. Latent factors are, by definition, found towards the beginning of the arrow, whilst pathway and cumulative factors unfold along it. The underlying assumption of this framework is that adult health is determined by both lifecourse and contemporary factors, and not exclusively one or the other.

We propose to test the validity of this framework using data from the 1958 British birth cohort; which is one of the few studies with sufficient data to model both lifecourse and contemporary circumstances, using self-rated health at age 33 as the outcome. To our knowledge, this is the first time that such an exercise has been attempted with detailed longitudinal data from birth.

Section snippets

Study sample and overview of determinants of health

The 1958 birth cohort includes all children born in England, Wales and Scotland during the 3–9 March 1958. The study originated in the Perinatal Mortality Study, which gathered information on 16,964 live births (97% of all births, after excluding deaths and refusals). Information was obtained from parents, teachers, medical officers and individuals themselves at subsequent follow-up at ages 7, 11, and 16, and from a personal interview with study subjects at ages 23 and 33. Despite sample

Data analysis

A series of multiple linear regression models were fitted, using self-rated health (scored 1,2,3 and 4) at age 33 as the dependent health outcome. We selected multiple linear regression because it provides a simple method with which to identify the strongest correlates of health within each factor, and by adding factors in blocks, it afforded a simple way to impose the framework on the data. Although linear regression is more appropriate for a continuous outcome variable, it is recognised as a

Results

Table 3 presents the regression coefficients and the associated level of significance for the two groups of lifecourse factors. Several early life factors were significantly associated with self-rated health at age 33. The greatest effects were seen for reading ability at age 7 and the percentage of adult height attained by age 7. Intermediate effects were found for parental frequency of reading to their child, height, socio-emotional status and parental interest in education all at age 7.

Discussion

The framework presented here, performed well in identifying distinct groups of lifecourse and contemporary factors that influence self-rated health by age 33. In so doing, the notion that lifecourse and contemporary factors ought to be considered together has been validated. However, the proportion of variance explained by this specific model, using the variables available to us, was not very high. Thus, we must regard this analysis as a starting point and further refinements of the framework

Acknowledgements

The research was supported by the Canadian Institute for Advanced Research (who support C. Hertzman and C. Power) and the UK Economic and Social Research Council under the Health Variations Programme (L128251021).

Data acknowledgement

Centre for Longitudinal Studies, Institute of Education, National Child Development Study Composite File including selected Perinatal Data and sweeps one to five [computer file]. National Birthday Trust Fund, National Children's Bureau, City University Social

References (44)

  • G Davey Smith et al.

    Lifetime socioeconomic position and mortalityProspective observational study

    British Medical Journal

    (1997)
  • Drever, F., & Whitehead, M. (1997). Health inequalities. Office for National Statistics Decennial (Suppl. 15, pp. 2–6)....
  • J Elander et al.

    Use and development of the Rutter parents’ and teachers scales

    International Journal of Methods in Psychiatric Research

    (1995)
  • Ferri, E. (ed.), (1993). Life at 33: The fifth follow-up of the national child development study. London: National...
  • M Ghodsian et al.

    Changes in behaviour ratings of a national sample of children

    British Journal of Social and Clinical Psychology

    (1980)
  • M.D Gliksman et al.

    Childhood socioeconomic status and risk of cardiovascular disease in middle aged US womenA prospective study

    Journal of Epidemiology and Community Health

    (1995)
  • M Haan et al.

    Poverty and healthprospective evidence from the Alameda Country study

    American Journal of Epidemiology

    (1987)
  • E.L Idler et al.

    Self rated health and mortality in the NHANES-1 epidemiologic follow up study

    American Journal of Public Health

    (1990)
  • E.L Idler et al.

    Self-rated health and mortalityA review of twenty-seven community studies

    Journal of health and Social Behaviour

    (1997)
  • G.A Kaplan et al.

    Perceived health and mortalityA nine-year follow-up of the human population laboratory cohort

    American Journal of Epidemiology

    (1983)
  • G.A Kaplan et al.

    Inequality in income and mortality in the United Statesanalysis of mortality and potential pathways

    British Medical Journal

    (1996)
  • I Kawachi et al.

    Social capital, income inequality, and mortality

    American Journal of Public Health

    (1997)
  • Cited by (187)

    • Residential history and changes in perceived health—The Northern Finland Birth Cohort 1966 study

      2022, Health and Place
      Citation Excerpt :

      However, aside from moving around to undertake their daily activities, people also move around and change their residential locations over time. Residence duration and the accumulation of certain environmental circumstances can be important to adult health, as both past and contemporary living environments may contribute to health outcomes (Clarke et al., 2014; Glymour et al., 2010; Hertzman et al., 2001; Macintyre et al., 2002; Pearce, 2018; Pearce et al., 2018; Sharp et al., 2015; Spring 2018; Yang and South, 2020). In longitudinal studies conducted in the United States and Scotland, neighbourhood disadvantage has been found to predict poor perceived health (Glymour et al., 2010) and has been associated negatively with mental health outcomes (Pearce et al., 2018), as well as functional decline and mortality (Clarke et al., 2014).

    View all citing articles on Scopus
    View full text