as timely and useful in helping us to understand the dynamics of adaptive life transitions in varied groups. Identifying factors that enhance active life expectancy of ethnic minorities is an essential step in developing effective policy to assist minority elderly persons. Research on population dynamics needs to give special attention to at-risk populations—in particular, those subpopulations at high risk for poverty, social isolation, underemployment, inadequate education, illness, and inaccessibility to health care. Higher-priority attention should be given to such subpopulations as the poor, women, and minorities.
The physical and mental health of men and women in the United States differ in significant but paradoxical ways. Women have a higher incidence of transitory illness and a higher prevalence of nonfatal chronic diseases, they report more symptoms of illness, and they are major consumers of health care services. Women under age 65 experience more injuries, bed disability, and restricted activity days than men; this gender difference becomes even more pronounced at age 65 (Minkler and Stone, 1983). Maddox (1987a) found that women were more likely to be impaired and to be impaired earlier than men; when socioeconomic status was equalized, however, this difference disappeared. Yet, on average, women live 8 years longer than men (Verbrugge, 1985, 1988). These differences provide an obvious opportunity for cohort-sequential longitudinal research to document whether observed gender differences in morbidity and well-being are persisting or changing over time.
It is important that research also strives to identify and explain psychological variables both as predictors of health, longevity, and functional disability and as ends in themselves. Psychological and behavioral variables may not only contribute to biological aging, but may also be the behavioral representation of physiological dysfunction. In addition, certain psychological states need to be identified as conditions that must be managed in the aging process. Equally, this research necessitates detailed analyses of existing databases and the creation of new databases that contain information on health histories and psychological functioning to allow testing of the reciprocal effects of illness, lifestyles, personality variables, and psychological competence.
Research should be undertaken to study the manner in which social structures and changes in those structures affect aging.
Just as individuals change, so too do social structures; observations of old age today show neither what old age was like in the past nor what it will be like in the future (Bengtson et al., 1985; Riley et