Aging, Health Behaviors, and Health Outcomes

K. Warner Schaie, Dan Blazer, James S. House, James A. House

• Publisher: Psychology Press
• Year: 2013
• eBook ISBN: 9781317844211
• Topic: Psicología
• Subtopic: Psicología del desarrollo

This volume is the fourth in a series designed to facilitate inter-disciplinary communication between scientists concerned with the description of societal phenomena and those investigating adult development. As such, it contains a compilation of papers presented at an annual conference held at the Pennsylvania State University. These essays by sociologists and epidemiologists deal with the impact of disease and health outcomes with advancing age and are critiqued by members of related disciplines. In addition, there are overviews as well as specific discussions about the impact of cancer, depression, and cardiovascular diseases upon psychosocial functions.

1
Social Stratification, Age, and Health

James S. House
Ronald C. Kessler
A. Regula Herzog
Richard P. Mero
Ann M. Kinney
Martha J. Breslow
University of Michigan

Introduction

This chapter integrates and summarizes a line of research developed in somewhat more detail in several other papers (House et al., 1990a, 1990b, 1991). The research derives from an ongoing program project funded by the National Institute on Aging, entitled “Stress, Health and Productive Activity in Middle and Later Life,” which seeks to identify psychosocial factors that maintain and enhance health and effective functioning in middle and later life. The work presented here suggests that such psychosocial factors are both macrosocial and microsocial in nature. At the macrosocial level we focus on the system of social stratification in our society. We then try to show how position in the stratification system shapes exposure to microsocial risk factors that are the more proximate determinants of health over the life course.

In this chapter we argue that our society can neither understand nor deal with the related problems of aging and health without coming to grips with what we have come to term the social stratification of aging and health (House et al., 1990b, 1991). By this we mean that the way in which health varies by age within and across individuals is heavily influenced by people’s socioeconomic status (House et al., 1990a). People of higher socioeconomic status (SES) generally experience high levels of health until quite late in life, whereas people of lower SES are much more likely to manifest significant declines in health by early middle age. Thus, the relation of age to health varies markedly by SES. Furthermore, the relation of SES to health is very different at different points in the life course: SES differences in health are small in early adulthood, increase steadily during middle age and early old age, and then diminish again in advanced old age.

We first suggest theoretically why the relation of age to health should vary across SES levels due mainly to: (a) the differential exposure of socioeconomic strata to major psychosocial health-risk factors and (b) the differential impact of these risk factors by age. Specifically, we argue that lower SES groups are likely to be more exposed to all psychosocial risk factors, especially in middle and early old age, and the impact of these risk factors is likely to increase with age up through middle age and early old age. Consequently, socioeconomic differentials in health should be greatest in middle age and early old age (when SES differences in exposure to risk factors are greatest and their impact on health is also large) and relatively small in both early adulthood (when SES differences in exposure to risk factors may be sizable but their impact is muted) and advanced old age (when SES differences in exposure to risk factors diminish though their impact remains substantial).

Second, we show empirically in national cross-sectional survey data that the relationship of age to health does vary substantially by SES in the expected way. Further, longitudinal data over 2% years suggest in a preliminary way that within-individual changes in self-report indicators of health vary by age and socioeconomic status in ways that are consistent with interpreting the cross-sectional data as reflecting a causal impact of socioeconomic status on the way health changes as people age. Third, we show that variations in both the exposure to, and the impact of, psychosocial risk factors across SES and age are generally as we expect if these factors are to explain the observed social stratification of aging and health. Fourth, we demonstrate that adjusting for this pattern of differential exposure to, and impact of, psychosocial risk factors can largely account for or explain the observed social stratification of aging and health. Finally, we discuss the implications of these ideas and data for what we and others take to be the fundamental agenda for research and policy on aging and health; that is, to understand what factors determine the relation of age to morbidity, functional capacity, and mortality, and the extent to which these factors make it possible to postpone morbidity and functional limitations into an increasingly brief final phase of the finite, although not perhaps fixed, human life span (Gerontologica Perspecta, 1987).

Why Should the Process of Aging and Health be Stratified Socioeconomically?

Human life expectancy has increased more in the last century than in all prior history (Preston, 1977). At the same time, two other aspects of the biology and sociology of human health and aging have remained remarkably invariant. First, the maximal human life span has not increased commensurately with life expectancy, if at all (Fries, 1980). Second, socioeconomic differentials in mortality, disability, and morbidity have remained remarkably persistent in the United States and other developed countries (Fingerhut, Wilson, & Feldheim, 1980; Kitigawa & Hauser, 1973; Marmot, Kogevinas, & Elston, 1987; Syme & Berkman, 1976; Williams, 1990). Each of these persisting verities raises important scientific and policy issues regarding human aging and health, yet socioeconomic status (or indicators of it such as education, occupation, or income) is not even considered in recent overviews of government statistics on aging and health (e.g., Brody, Brock, & Williams, 1987; National Center for Health Statistics, 1987), and is treated only cursorily in major reviews of the literature on aging and health (e.g., Shanas & Maddox, 1985).

The dramatic increase of life expectancy in the face of a finite human life span (of about 85 to 90 years on average) has suggested both utopian and dystopian scenarios for the future. The utopian scenario argues that continuing improvements in health and life expectancy will increasingly postpone or “compress” morbidity and disability into a relatively brief period at the end of the life span, thus improving the quality of life and perhaps even reducing the need for medical care (Fries, 1980). The dystopian scenario suggests that recent and future gains in life expectancy largely add years to life in which people are chronically ill and disabled, and thus high consumers of health care (Gruenberg, 1977; Manton, 1982; Schneider & Brody, 1983; Verbrugge, 1984).

Which scenario proves correct depends on the extent to which currently observed declines in physical and psychological functioning with age are intrinsically linked to the biological process of aging (and thus, not readily preventable or postponable), or are determined by factors extrinsic to that process (and hence more preventable or postponable). Such extrinsic factors are increasingly likely to be social, psychological, and behavioral in nature because morbidity, disability, and mortality in middle and older age are largely a function of chronic diseases, the etiology and course of which are increasingly determined by social, psychological, and behavioral factors (Rowe & Kahn, 1987).

Both theory and data now indicate that SES is a major determinant of the degree to which mortality, morbidity, and disability are postponed into the last phase of the finite human life span (House et al., 1990a). That is, people in the upper socioeconomic strata increasingly approximate the utopian scenario of longer life and better health, with not only mortality but also morbidity and disability or functional limitations generally postponed to quite late in life. In contrast, people in the lower socioeconomic strata experience significant levels of mortality, morbidity, and disability beginning relatively early in mid-life. Thus, with increasing age, levels of health become increasingly differentiated by SES, until the final period of the normal human life span when both biological and social forces tend again to reduce socioeconomic differences.

We believe this social stratification of aging and health is produced by social and biological mechanisms that determine both exposure to, and the impact of, a set of psychosocial variables that have been increasingly recognized as major determinants or risk factors for chronic disease mortality, morbidity, and disability. These include: (a) health behaviors such as smoking and immoderate eating (leading to under or overweight) or drinking (Berkman & Breslow, 1983); (b) lack of social relationships and supports (House, Landis, & Umberson, 1988); (c) lack of what is variously termed self-efficacy, self-directedness, competence, mastery, or control (Rodin, 1986; Rodin, Schooler, & Schaie, 1990; Rowe & Kahn, 1987); and (d) chronic and acute stress (House, 1987; Pearlin, Lieberman, Menaghan, & Mullan, 1981; Theorell, 1982; Thoits, 1983). A small but growing body of theory and evidence suggests that all of these risk factors are more prevalent in lower socioeconomic groups (Williams, 1990). For some variables this may become increasingly true with age as SES becomes more fixed and more cumulative in its effects through middle and early old age. These differences may, however, be muted in later old age. In recent decades our society has directly invested more heavily in improving the social, economic, and health-care status of the older population, as compared to children and younger adults (Duncan & Smith, 1989; Preston, 1984). These social welfare policies, as well as other changes in the biological, social, and psychological status of people as they age may serve to attenuate socioeconomic differences in exposure to psychosocial risk factors in older age.

Additionally, there is reason to believe that the impact of many or most of these psychosocial risk factors increases with age, at least until early old age (House & Robbins, 1983). Biologically, people become more vulnerable to a wide range of diseases with age. Socially and psychologically, issues like the maintenance and loss of social relationships and supports or of self-efficacy and control may become more problematic and hence more consequential in older age (e.g., House & Robbins, 1983; Rodin, 1986).

In summary, scattered but growing bodies of theory and data suggest that the process by which health changes with age may be importantly stratified by SES. On average, we should see the largest socioeconomic differentials in health in middle and early old age because these age groups are most likely to be characterized by both sizable SES differentials in exposure to risk factors and substantial impact of the risk factors. In contrast, in early adulthood, SES differences in exposure may be sizable but their health impact is muted, whereas in later old age SES differences in exposure become somewhat muted even if their impact remains strong. Let us now consider how well available data accord with these expectations.

Methods

Data Sources

Our principal data source is an ongoing longitudinal survey, entitled Americans’ Changing Lives (or ACL), carried out by the Survey Research Center of the University of Michigan on a multi-stage, stratified, area probability sample of noninstitutionalized persons 25 years of age or older and living in the coterminous United States, with oversampling of Blacks and persons age 60 +. Initial face-to-face interviews (known as ACL 1) lasting 86 minutes on average were carried out in mid-1986 in the homes of 3,617 respondents, reflecting a response rate of 67% of all designated sample respondents (and 70% of designated households and of designated individuals who spoke English or Spanish and were physically and mentally capable of being interviewed). A total of 2,867, or over 83% of the surviving 1986 respondents participated in an 83-minute follow-up interview (known as ACL 2) in early 1989, and constitute our current longitudinal sample with about 2.5 years between waves. Some parallel analyses were conducted for persons age 25 and over in the 1985 National Health Interview Survey ...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Preface
• Aging, Health Behaviors, and Health Outcomes: Goals of the Conference
• 1 Social Stratification, Age, and Health
• Social Stratification and Aging: Contemporaneous and Cumulative Effects
• 2 Socioeconomic Status, Health Behaviors, and Health Status Among Blacks
• Social Structure and the Health Behaviors of Blacks
• Discussion of Socioeconomic Status, Health Behaviors, and Health Status Among Blacks
• 3 Health and Aging in the Alameda County Study
• Aging and the Public Health: Reflections on Kaplan’s Report of Health and Aging in the Alameda County Study
• The Laudables and Limits of Large Epidemiologic Studies of Mortality
• 4 Living Arrangements and Problems With Daily Tasks for Older Women With Breast Cancer
• The Content and Context of Effective Spousal Support
• Reducing Disability in Research
• 5 Social Factors and the Onset and Outcome of Depression
• Cohort Experiences, Support Versatility, Depressive Traits, and Theory
• Statistical and Causal Interaction in the Diagnosis and Outcome of Depression
• 6 Aging, Health Behaviors, and Health Outcomes: Some Concluding Comments
• Author Index
• Subject Index