of evidence for the factors. The final section considers the critical and often overlooked issue of selection, which affects all studies and analyses of differences in health in late life.

Genetic predispositions underlie the mechanisms involved in health and disease, but whether these vary sufficiently by racial and ethnic groups is a question considerably complicated by multiple factors, such as the fact that racial and ethnic groups are not strict genetic groupings and the dependence of gene action on environmental factors. While some genetic variability across races and ethnic groups is tied to vulnerability to particular diseases (e.g., Neel, 1997), the variability within groups is considerable. Single-gene disorders make a trivial contribution to overall health differences (Cooper and David, 1986). Nevertheless, gene-gene and gene-environment interactions could play an important role in differences in specific diseases.

Among other factors, it is clear that socioeconomic status is implicated in racial and ethnic health differences. We have noted multiple dimensions of status recognized in the literature, chief among them education and income and wealth. The complications in the effects of status involve not only its reciprocal relationship with health, but also nonlinearities in its effects. For older persons, socioeconomic status over their lifetimes may be relevant for current health.

One possible route of influence of socioeconomic status is through health risk behavior, which covers such important factors as diet, smoking, exercise, sexual risk taking, and drug abuse. These behaviors may have an impact on health at specific stages in the life course, such as in adolescence, or may have a cumulative effect, in either case extending their influence into later life and affecting differences at older ages.

Possibly offsetting health risk behaviors may be positive factors: personal resources and social support may provide various ways of coping with unfavorable circumstances, such as avoiding physical or mental illness or mitigating its severity. For instance, religious involvement or a sense of personal control may contribute to psychological resilience and help avoid depression. To take another example, support from relatives or other caregivers is helpful in illness or disability, although the effects on differences are not necessarily straightforward.

While other people may contribute to an individual’s health through providing social support, they may also contribute to disease through prejudice and discriminatory behavior. Such behavior may also be characterized as racism, though we assume that it could be triggered not only by race but also by immigrant status, nativity, or religion. The roles of prejudice and discrimination may be broader than represented in Figure 2-1, since they can also affect such other factors as socioeconomic status, neighborhood conditions, and access to appropriate health care (e.g., Byrd and Clayton, 2002; Williams, 2001b). Researchers have focused more recently on the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement