Does childhood health affect chronic morbidity in later life?

Keywords

• Childhood health;
• Socioeconomic status;
• Adult health;
• Health and retirement study;
• USA

Introduction

Mounting evidence suggests that childhood life circumstances have an enduring effect on late life chronic morbidity. Although the roles of specific etiologic agents are topics of debate, research suggests that these effects may accrue from a range of factors including nutrition, exposure to infectious disease and environmental toxins, the in utero environment, and features of social and economic deprivation. Some scholars have hypothesized that chronic diseases are programmed during gestation or early childhood (e.g., the Barker hypothesis), while other scholars emphasize the role of accumulation of effects from exposure to adverse conditions over the life course (Forsdahl, 1977; Kuh & Davey Smith, 1997). Increasingly it is clear that a life course approach is important to understand how chronic morbidity comes about in later life. Such an approach encompasses the idea that chronic disease may be the long-term outcome of a range of childhood conditions and experiences, beginning as early as in utero combined with the cumulative “insults” experienced during adulthood (Kuh & Ben-Shlomo, 1997).

Our investigation adds to this growing body of literature by examining the association between self-reported childhood health experiences and the presence of a number of major chronic health problems among a nationally representative sample of Americans aged 55 to 65. Several issues guide our analysis. First, we assess whether self-reported childhood health experiences are associated with chronic health problems in later life, while controlling for socioeconomic deprivation in both childhood and adulthood. That is, we determine whether childhood health per se has long-term and direct consequences for chronic health conditions experienced decades after childhood. Second, we examine whether adult height is a surrogate for childhood health experiences and deprivation in modeling chronic disease in later life. Adult height is often viewed as an indicator of childhood health and deprivation in analyses lacking direct measures of childhood life circumstances. Lastly, we assess whether the effects of adult socioeconomic status on chronic disease are overestimated when childhood conditions are not taken into account.

The analysis is based on a new population health survey representative of Americans born from 1931 to 1941, the Health and Retirement Survey (HRS). HRS respondents were interviewed in 1992, 1994, 1996, and 1998. We use the third wave of the HRS (1996) in this analysis because at this interview a randomly chosen subset of respondents (N=654) recollected their health as children and reported their parental education and father's occupation, along with family living arrangements, and their family's financial well-being. At the same interview they reported details of their current adult health state. Measurement of both childhood health conditions and adult health is based on self-report data. Our analysis, therefore, focuses on associations between childhood conditions and health conditions at age 55–65 using self-reported information from a nationally representative survey. Previewing our results, despite the relatively small sample size of the HRS module, our statistical analysis reveals a strong association between childhood health and the occurrence of a variety of chronic health problems among middle-aged Americans.

Background

Social science and social epidemiological studies of chronic health problems have focused principally on the association between adult life circumstances, especially socioeconomic status, and disease prevalence or mortality at older ages. The general presumption is that socioeconomic status, as an enduring feature of adult life, gives rise to chronic health conditions, which by themselves develop slowly over the life cycle (Adler et al., 1994; Marmot, Kogevinas, & Elston, 1987).

In fact, evidence indicates that socioeconomic status may have a more pervasive, long-term effect, if one considers the possible ramifications of childhood deprivation. A growing number of studies support the idea that childhood life circumstances have enduring effects on late life chronic morbidity (Kuh & Ben-Shlomo, 1997). That is, rather than simply setting the stage for adult achievement which then affects health, childhood conditions may play a more direct role in influencing chronic disease decades later in old age. These effects may accrue from nutrition and dietary factors (Gunnell, Frankel, Nanchahal, Braddon & Davey Smith, 1996), in utero conditions (Barker, 1998; Barker, Bull, Osmond, & Simmonds, 1990; Barker et al., 1991; Martyn, Barker, & Osmond, 1996), and other features of deprivation (Arnesen & Forsdahl, 1985; Gliksman et al., 1995; Lundberg, 1993; Notkola, Punsar, Karvonen & Haapakoski, 1985; Peck, 1992; Peck, 1994). Behavioral factors may also play a role: conflicts or dissension in the childhood home and divorce of the parents have both been linked to illness in later life (Dahl & Birkelund, 1997; Lundberg (1993) and Lundberg (1997)). There is also evidence of the direct effect of childhood exposure to infectious disease, viruses, and environmental toxins on adult health outcomes (Brunner et al., 1996; Kuh & Wadsworth, 1993; Power & Peckham, 1990).

A growing body of evidence documents the effects of early life health infections on cardiovascular disease, cancer, diabetes, and respiratory disease, and increasingly, models of biological processes are being developed to explain these associations (see Kuh & Ben-Shlomo, 1997). Using data from US Civil War veterans, Costa (2000) found that infectious illness during young adulthood was related to respiratory problems, heart problems and joint and back problems when the cohort was aged 50 to 64. An ecological analysis in the United States showed that areal levels of infection in childhood were related to higher areal levels of heart disease and respiratory cancer later in life (Buck & Simpson, 1982). Infectious disease in childhood is thought to be related to heart disease through its effect on autoimmune complexes and consequent development of atherosclerotic lesions, resulting in plaque accumulation over the life span (Buck & Simpson, 1982; Matthews, Whittingham, & Mackay, 1974). Early respiratory infections have been linked to later lung conditions (Barker, 1998). Higher levels of musculoskeletal conditions in older age have been linked to lasting joint problems resulting from earlier infectious diseases (Costa, 2000). Infectious disease in early life may be related to higher levels of some kinds of cancer through the interaction of vitamin metabolism in early life and later life exposure to carcinogens (Buck & Simpson, 1982; Hall & Peekham, 1997).

While the above discussion links infectious disease to higher levels of adult conditions, some research has raised the possibility that higher levels of early life infection could have positive health effects later in life. Higher levels of infection early in life, for example, may be related to lower levels of autoimmune conditions later in life (Paunio Patja et al., 2000). Following this reasoning, some have argued that the reduction in childhood infectious diseases over time could be linked to altered immune system development and the subsequent rise in conditions such as asthma.