Promoting Health: Intervention Strategies from Social and Behavioral Research (2000)

George L.Maddox, Ph.D.

In the last half of the twentieth century the United States discovered the implications of an aging population for both social institutions and individuals. In the context of public health, aging populations were of interest because they were large and observed to be growing rapidly at mid century. At the end of the century, 13% of the population of the United States was 65 years of age or older and an estimated 4%, 80 or older. In Europe, aging populations are already much larger, rising to 20% in Germany, for example. Public health interest is also sustained by evidence that older adults have high rates of chronic conditions accompanied by functional disability, consume health and social services at a high rate, and illustrate so consistently the negative consequences for health of lower socioeconomic status (SES). The bad news from research has been that individuals could be expected to live longer but they could not depend on dying conveniently with functional capacity and financial resources intact. The social institutions through which a society does its essential business of transportation, housing, retirement at the end of a work career, personal income maintenance, and providing essential health and social care for elderly citizens have been seriously challenged. The good new from behavioral and social research on aging, however, has delivered increasingly positive evidence that aging well is possible

for the substantial majority of older adults. And the prospects of beneficial behavioral and social interventions to enhance well-being in later life continue to improve. The mind set of contemporary gerontology is increasingly interventionist and optimistic about change.

The challenges of individual and population aging are real and should not be underestimated. Demographers have documented, and often viewed with alarm, the increase in dependency ratios (the number of nonworking individuals versus those actively in the workforce) and the age gradient of risk of dependency, and have forecast problems for pension, income maintenance, and health care financing. Epidemiologists have documented both the significance and the persistent age gradient of morbidity and comorbidity and, with behavioral and social scientists, have also documented notable variation in life expectancy and risk of functional disability associated with socioeconomic status, ethnicity, and gender. The documented diversity of older populations has been a significant factor in the growing awareness of gerontologists that chronological age per se is a relatively weak explanatory variable in assessing the prospects of continuing to age well in later life, which is the experience of a substantial majority of older adults (Campbell et al., 1976; Maddox and Glass, 1999).

The established behavioral and social diversity of older adults has had important practical and theoretical consequences in gerontological research. Practically, sampling older populations for research and for generalizing about older adults, who constitute about 13% of the U.S. population currently, typically requires oversampling to insure inclusion of minorities and the very old. Theoretically, repeatedly observed diversity within age cohorts suggests both that behavioral and social factors are involved in aging processes and that such factors are potentially modifiable. The intellectual shot that was heard around the world in gerontology was fired by Stanford's James Fries (1980) in a seminal article on “the compression of morbidity. ” Fries argued that even if the life span is typically limited (he optimistically and incorrectly concluded that the limit is 85 years), the risk of morbidity and related functional disability can be delayed through interventions designed to improve physical activity, social integration, and cognitive performance. Critics initially scoffed and dismissed such optimism, but not for long. The evidence supporting the modifiability of functional disability in later life has continued to accumulate and has laid the foundation for the increasing interest of gerontologists in purposive behavioral and social interventions to improve the well-being of older adults (e.g., Rowe and Kahn, 1987; Berkman, 1988; Manton, 1988; Fries, 1995).

Unlike their clinical colleagues who view interventions at the personal level to be a natural extension of practical knowledge, behavioral and social scientists have moved cautiously in developing enthusiasm for psychosocial interventions. And they have remained particularly cautious about macrosocial legislated interventions such as Social Security and Medicare or Medicaid, both of which are viewed as important and beneficial natural experiments but outside the competence of most behavioral and social scientists to initiate and evaluate as political interventions (see, e.g., Marmor and Okma, 1998). An important turning point in

interest in psychosocial interventions in research on aging on a modest scale occurred, however, when the Institute of Medicine, responding to a request of the National Institute on Aging to propose a research agenda for the future, recommended increased emphasis on psychosocial interventions and field trials (Lonergan, 1991). Interest in psychosocial interventions in later life has increased markedly in the past decade.

What follows in this paper are three sections, beginning with an overview of what is known about aging individuals and populations. Evidence from epidemiology and medical demography provides a clear characterization of the health and well-being of older adults and populations. While the evidence is primarily from the United States, the characteristics of aging populations in other industrial societies are similar. A second section presents brief accounts of four quite varied psychosocial interventions that range from emphasis on health promotion and disease prevention through physical activity for well elderly persons through rehabilitation regimes for persons with diagnosed disease to enhancement of autonomy among institutionalized elderly. A concluding section illustrates how the convergence of theory and research on the social structuring of living and work environments, and on the reconceptualization of dependency in later life as a modifiable interpersonal strategy, suggests a distinctive psychosocial intervention for frail, nursing home-eligible older adults living in the community. The intervention is a distinctive new type of housing with services known as “assisted living.”

The basic evidence for characterizing the health and well-being of older populations is now well established (Berkman, 1988; Manton, 1988; Manton et al., 1997; Maddox and Glass, 1999):

• The distinctive age gradients of morbidity, disability, and mortality vary significantly by gender, ethnicity, and socioeconomic status, although the gradients are attenuated in very late life as a result of selective survival.

• table of chronic conditions in the mid-1990s for women were arthritis, hypertension, hearing loss, and chronic heart disease; for men the predominant conditions were arthritis, hearing loss, hypertension, and ischemic heart disease (Verbrugge, 1995).

• The expected age gradient in medical care utilization occurs, but the utilization patterns of men and women vary significantly; higher total utilization by women reflects the fact that older women are more likely to be functionally disabled than men, and they live longer.

• Comorbidity is a condition well known to geriatricians; the average older adult has two chronic conditions at any time of observation.

Older populations are typically characterized in terms of functional status in addition to health status. The usual designations of functioning are Activities of Daily Living (ADLs) and Instrumental Activities of Daily Living (IADLs). The former reference self-care activities such as bathing, toileting, and eating; the latter, activities such as ambulation, cooking, and money management. These designations have stood up well in both research and practice for use in characterizing types and levels of need in populations and as triggers for authorizing services for individuals. Manton (1988), for many years the data manager of the Health Care Financing Administration's longitudinal National Long Term Care Survey (NLTCS), observes that functional status complements medical diagnoses in a useful way: functional assessment suggests the range of services likely to be required in care management while diagnosis suggests the possible length of time services may be required. Functional status has also been a key construct in the Medical Outcome Study of patients being seen in primary care settings (Stewart et al., 1989) as an outcome measure; and the concept has been broadened to include the capacity to perform social roles.

In reviewing the evidence on the prevalence and distribution of functional disability in later life, one finds many expectations fulfilled, but also a few surprises:

• The most significant surprise is that 8 or 10 older adults report no fuctional disability at all and 95% live in the community.

• A persistent 5% of older adults reside in nursing homes at any observation, and the lifetime risk of any individual for at least some exposure to a nursing home is an estimated 20%. An additional estimated 5% of older adults in the community have functional disabilities that make them “nursing home eligible.”

• For both men and women, arthritis and circulatory problems are the most totally disabling conditions. Total disablement is the experience of 30% of older adults with circulatory problems and 26% of those with arthritis. For both conditions, women are more likely than men to be totally disabled.

• Senility, an outmoded term that persists to indicate cognitive deficit in late life, ranks fourth (15%) in the NLTCS league table of chronic conditions producing total disability.

• Depression, a common mental health problem in adulthood, has not been found to be consistently elevated in older populations. However, when present, both depression and cognitive impairment increase the risk of functional impairment from chronic conditions, complicate the management of chronic conditions, and increase the risk of extraordinary outcomes such as suicide among older males (Stewart et al., 1988; Roberts et al., 1997; Reynolds et al., 1998).

• The consequences of the socioeconomic gradient of chronic conditions (see, e.g., Evans et al., 1994) is dramatically underscored by the distribution of active life expectancy in later life (Guralnick et al., 1993). Active life expectancy designates disability-free years of living post-age 65. Higher educational attainment, for example, predicts for men an 18% increase in disability-free

years and for women, 25%. The SES gradient of health and illness is observed worldwide, although the gradient can be affected by sociocultural factors.

• Ethnicity as well as socioeconomic status affects functioning and well-being in later life (Clark et al., 1993; Markides and Wallace, 1996). These effects are observed in distinctive patterns of disability and dependency, need for long-term care services, and cultural expectations about responsibility for providing care for family members.

• Longitudinal trajectories of disability over a 4-year period in the NLTCS document stability of functional status for a substantial majority (80%), and for those whose functional status changes, the change is positive for 30%.

• Healthy life-styles are observed as frequently among older adults as among adults generally and have the expected positive effects (see, e.g., Berkman and Breslow, 1983; Rowe and Kahn, 1997). But commitment to healthy life-styles among adults in the United States remains suboptimal.

In sum, the demography and epidemiology of later life have reinforced some of the basic maxims of contemporary gerontology: Older populations are and remain substantially diverse; the gradients of chronic conditions and related functional disability vary significantly by gender, ethnicity, and socioeconomic status; and the mutability of functional status and well-being in later life is well established. In my 40 years in gerontology, probably the most significant event has been increasing evidence-based interest in psychosocial interventions. Gerontology has rediscovered the venerable maxim of clinical and scientific research: If you want to understand something (aging), try to change it. This maxim has increasingly dominated my own thinking about the feasibility of beneficial psychosocial interventions (Maddox, 1987) and my research (Maddox and Clark, 1992; Maddox et al., 1994). The Social Security Administration's Longitudinal Retirement History Survey, which was designed to explore retirement as a social process for a large population (11,000) over the period of a decade, sensitized me to the SES and gender gradients in disability among older adults. What we found in my research was the following:

• The expected age gradient of functional disability was confirmed for both men and women in the observed trajectories of functional impairment over the period of a decade; women were and remained consistently more impaired than men; and the expected socioeconomic gradient of disability was found.

• But two significant surprises were revealed in the analysis. (1) At the upper ranges of age, the trajectories of disability converged for men and women, a finding anticipated from research on African Americans that indicated a convergence and reversal of disability rates at upper age ranges, probably reflecting selective survival. (2) The more important surprise was that when socioeconomic status was controlled, the initially observed gender difference disappeared.

With such information one does not leap to the conclusion that gender does not have a biological component. But the evidence clearly suggests that gender

differences in functional status and well-being have psychosocial components, particularly income and education, known to be modifiable. As noted earlier, the most obvious political interventions at the societal level focused on access to education and higher income have obvious merit but are beyond the scope of this paper. There is merit, however, in keeping in mind that the modest psychosocial interventions we typically choose to consider do not exhaust the possibilities. As noted above, the enactment of Social Security in 1933 and of Medicare and Medicaid in 1965 has had far-reaching benefits for the welfare of older adults in the United States. Similarly, a brief note about the Black report (Townsend and Davidson, 1982) will remind us just how much we prefer modest nonpolitical interventions in the interest of improving the public's health and well-being.

In Great Britain at the end of 25 years of the National Health Services, a Labour government asked a commission to review how well this free at the point of access health care system had done in improving the health of the British people. Not enough, came the answer from Sir Douglas Black and his colleagues; pronounced social class differences in health persist. What should be done to improve this discrepancy? Do not increase the number of physicians and hospital beds came the answer. Instead, came an answer that might have been and quite possibly was copied from an epidemiology textbook: Concentrate public policy initiatives on redistributing income and increasing access of the young to education. This interesting advice, as luck would have it, had to be offered to the Conservative government of Mrs. Thatcher who came into power in 1980, and it was promptly shelved. Mrs. Thatcher had a different view of the role of government in beneficial interventions that concentrated on private-sector initiatives.

Evidence of the modifiability of aging processes and the experience of aging has stimulated an increasing number of psychosocial interventions in later life. Strategies for design and implementation of such interventions and the variety of their objectives are conveniently summarized in a recent volume edited by Schulz et al. (1998). The substance of the interventions summarized ranges broadly over topics that include uses of pharmacologic agents in combination with psychosocial therapy in control of depression; physical activity regimes to improve physical health and psychological functioning; improvement of memory; home modification to improve ambulation; incontinence control; improving motivation in nursing homes; and a comparative review of what one learns from societal policy initiatives about the modifiability of aging processes. In each instance, extensive bibliographies permit a review of a broad range of research on each topic.

The sampler of psychosocial interventions in later life presented here features four illustrative initiatives beginning with a macroanalysis of a very large number of activity regimes designed to improve physical health and sense of well-being. Two illustrations follow that address interventions to improve man-

agement of patients diagnosed with chronic conditions, specifically, osteoporosis and diabetes. One of these interventions addresses the issue of psychosocial interventions with minority and low-income patients. The fourth illustration is from a comparative study of interventions to reduce excessive dependency among very frail nursing home residents. Although these four interventions were chosen to illustrate variety in objectives and strategies, their findings proved to have two dominant themes in common: (1) emphasis on the benefits of social support groups and social network integration for health and well-being; and (2) self-efficacy as a key predictor variable in successful intervention outcomes.

The importance of social support groups and social integration in a variety of natural settings has been a major theme in most psychosocial research on aging in recent decades. In particular, Berkman (1988) and colleagues (Berkman and Syme, 1979; Berkman and Breslow, 1983; see also House et al., 1988) have consistently documented the positive relationships between social integration and support and favorable morbidity and mortality outcomes. Such a transactional view of the person as an active agent negotiating within a structured social context continues to be a central theme in a social ecological view of development over the life course. Further, a social learning perspective on development in later life that stresses the importance of self-efficacy is a central theme in a wide range of health intervention research in the last decade.

• Objective—McAuley and Katula (1998) provide a meta-analysis of 1,300 studies of “physical activities interventions” in later life intended to improve physical health and cognitive or emotional functioning (mood, morbidity, and sense of well-being). The straightforward hypothesis typically advanced is that activity is beneficial.

• Design Issues—Overall the studies reviewed are given low marks regarding their weak conceptualization of causation, inadequate measurement of key variables, and poor sample maintenance and motivation.

• Findings and Conclusions—Consistent evidence of a direct causal link between physical activity and beneficial outcomes across the various intervention studies reviewed was “surprisingly weak and equivocal.” The evidence was particularly unclear that physical activity per se, as distinct from social activity, generally was the primary causal factor in outcomes (see, e.g., Glass et al., 1999). Further, in many studies, favorable outcomes appeared to require a sense of social efficacy and a favorable outcome expectancy as intervening variables (see Bandura, 1997; also Antonovsky, 1987, for a discussion of the meaning of activities as a component of motivation).

• Objective—Osteoporosis, a common chronic condition among women, results in considerable pain and the possibility of significant skeletal deformity that often produces negative emotional outcomes such as depression, hostility, stress, and reduced self-esteem. To explore effective ways to counter these negative outcomes a multidisciplinary team of clinicians in a specialty clinic at Duke Medical Center created a controlled trial of education and exercise to reduce these unwanted outcomes (Gold et al., 1993).

• Design—Subjects in their mid-60s, primarily women, were recruited from surrounding life care communities and offered individual evaluations by a physician; personal tutoring regarding best ways to achieve pain reduction; and group exercise regimens in structured support groups. Subjects were interviewed 6 months later and compared with controls from the same clinic.

• Findings and Conclusion—While the intervention had no significant effect on pain reduction, psychiatric symptoms such as depression and stress were significantly reduced as were obsessive-compulsive behavior and anxiety. In subsequent research, investigators followed up what appeared to be higher levels of anxiety about self-image resulting from deformity than clinical assessment would suggest. This anticipatory anxiety appeared to be reduced by participation in a structured group-supported exercise regime that participants described as offering an “element of hope ” and a feeling of being better in control of their lives even in the absence of pain reduction. This finding is reminiscent of many intervention studies indicating that structured social support groups are a key variable in achieving beneficial effects from activity and that a sense of self-efficacy is a critical component of this effectiveness.

• Objective—While the intervention was initially prompted by interest in the benefits of exercise in management of diabetes in older adults, a greater interest was found in exploring how poor and minority populations can be attracted and motivated to continue in such structured activities (Clark, 1997; Clark and Nothwehr, 1999).

• Design—A large population of inner-city residents in a metropolitan area was invited to participate in a straightforward exposure to exercise as a component in managing diabetes. From a review of literature on successful exercise interventions, investigators concentrated on the more basic question of motivating patients to participate at all. Motivation to participate was found to be affected substantially by a sense of self-efficacy and outcome expectancy. These were the key intervening variables in achieving beneficial outcomes.

• Findings and Conclusions—The expectation that sense of efficacy and outcome expectance would be the key modifiable variables in whether patients made themselves available to beneficial exercise interventions was confirmed.

Further, the importance of involving local neighborhood institutions such as churches in interventions in the community was also confirmed.

• Objective—Using social-ecological theory to understand how various forms of dependency are created in later life, Margaret Baltes (1996; see also Baltes and Baltes, 1990) explored the dynamic transaction between frail older adults and those who care for them in nursing homes, primarily in Germany but also in the United States. She asked four questions: (1) Is behavioral dependency modifiable? (2) What kinds of social environments foster dependence? (3) Why do caregivers foster dependency? (4) Can dependency-creating environments be modified?

• Design—An important conceptual distinction was made at the outset between “learned helplessness,” in which any sense of self-efficacy and favorable outcome expectancy is abandoned, and dependency as an interpersonal strategy that a nursing home resident can use to receive attention and parlay a social exchange advantageously. Seven observational studies provided evidence to document how “scripts” in different care settings typically promote dependency if not helplessness, particularly in the more individualistically oriented United States.

• Findings and Conclusions—Evidence answered the four questions asked. The promotion of dependency is reinforced by a stereotypic association by caregivers of dependency in later life with incompetence. Through training, frail individuals could selectively increase their capacity for self-care and institutional caregivers could be trained to change their “script” so that they could attribute competence to frail residents and promote greater autonomy of action. As in the other interventions illustrated above, Bandura 's concepts of “sense of efficacy” and outcome expectancy are central. Also, dependency was found to have “many faces,” including being a device for increasing the sense of social contact and control on the part of a frail resident living in an environment scripted to assume incompetence. The behavior of both residents and caregivers in institutional settings, being learned, can be modified. Baltes' emphasis on the consequences of organizational structuring of the everyday contexts of living parallels Kohn's (1995) extensive research on the behavioral effects of work environments. Varied, challenging environments, Kohn found, promote satisfaction, a sense of efficacy, and the willingness and ability to respond to subsequent environmental change. And Bandura's complementary concept of self-efficacy and its key role in a theory of social learning appear with notable frequency in literature searches for reports of beneficial psychosocial interventions across the life course. Self-efficacy is a likely product of exposure to a supportive, socially integrated environment.

Evidence has established that the debilitating behavioral dependence of a substantial minority of older adults can be countered beneficially with appropriate psychosocial interventions. Some beneficial interventions directed primarily toward individuals, small groups of older adults, or those with identified clinical conditions have been illustrated. Major federal interventions at the societal level benefiting the functioning and well-being of older adults, such as Social Security and Medicare, have been noted but only briefly because the conceptualization and implementation of such interventions are essentially complex political undertakings beyond the scope of this paper. The final section of this paper, however, discusses a timely and potentially consequential intervention that shifts from interventions primarily with individuals or small groups to an intervention focused on the structural, policy, and program enhancements of a type of organization of increasing importance to frail older adults in the United States. The particular type of organization of interest is assisted living housing.

Assisted living housing is of interest as a strategic intervention site in aging for both theoretical and practical reasons. Kohn (1995), in an overview of his research on the effect of work environments on personality and behavior, has documented that the complexity of work environments and opportunities for personal involvement and mastery not only tends to produce a sense of well-being but also enhances the prospects for subsequent personal initiative and the willingness to take on and master the challenges of environmental change with confidence. Further, Moos and Lemke (1994) have documented that group living environments that maximize policies emphasizing personal control and involvement of residents in group decision making tend to enhance the sense of well-being of residents and reduce their inappropriate use of facility care services. From the perspective of Bandura's theory of social cognition, care environments can be designed to maximize social support resources that, rather than encouraging behavioral dependence, produce a sense of self-efficacy and outcome expectancy of independent living among frail older adults at both the personal and the group levels. Environments designed to maximize appropriate behavioral independence are not a recommendation of rugged individualism and the denial of appropriate care but a recognition that Baltes and Baltes (1990; see also Baltes, 1996) are on strong theoretical ground in describing the process of successful aging as “selective optimization with compensation.” That is, realistically, while later life is a season of increased risk of chronic conditions and functional disability, functional competence is not lost at the same rate in all areas of functioning; as the risks of disability increase, successful agers are observed to strive to do a smaller number of things but to do them well and seek environments in which compensatory help can be found in areas of decline.

Assisted living housing has emerged as a naturally occurring strategic site in which housing becomes the reasonable counterpart of a work site whose

structure is consequential for the competence and well-being of its participants. And for a subset of assisted living housing facilities in the United States, an emergent philosophy is identifiable that stresses the intention to maximize the residential quality of assisted living environments and at the same time enhance the sense of self and group efficacy among residents (Kane and Wilson, 1993; Wilson, 1996). Regnier et al., (1995) noted four basic concepts of a distinctive assisted living philosophy: (1) creating a place of one's own; (2) matching services to individual need; (3) sharing responsibility among caretakers, family, and resident; and (4) providing residents with choice and control of their lives.

Good theory has converged with some practical considerations to highlight assisted living housing as a strategic site for psychosocial intervention on a relatively large scale. Historically, housing, much less housing with services, has not been considered an integral component of health and health care policy or provision in the United States. Without much enthusiasm, nursing homes may call their inhabitants residents, but they are really patients in what operates very much like a hospital. Federal housing policy until recently has forced a separation between housing and service provision in minimally regulated and supervised settings variously known as old-age, retirement, or board-and-care homes. For an aging population in this country that clearly has expressed its preference to age at home if possible, the prospect of a purpose-built facility that invites residents at an affordable price to buy housing and at the same time purchase levels of service appropriate to their assessed needs has been just what they wanted. Private-sector developers got the picture quickly, and by 1998 there were 28,000 assisted living housing facilities in the United States with provision for more than 612,000 residents, and the level of investment in this market continues to be high. Twenty-five percent of the available places for residents are in California, Florida, and Pennsylvania alone (Mollica, 1998).

The potential size of this obviously attractive market is not known. Some portion of persons currently served in nursing homes, say at least 10%, might be accommodated more appropriately and economically in assisted living facilities. The size of the community-dwelling population that is assessed as having functional disability at a level qualifying individuals as “nursing home equivalent” is at least as large, possible larger, than the 5% already in nursing homes. A population of more than 1.5 million persons could easily constitute the market for assisted living housing as an effective way to deal with chronic health problems and functional disability. For the evidence that underlies such an estimate, see the population projections, including demographic characteristics of income, disability, and living arrangements, in Health, United States, 1999 (U.S.DHHS, 1999), and the multivariate analyses of how demographic and health characteristics of older adults relate to both institutionalization and management of disability in the community (Manton, 1988; see also Wiener et al., 1990). While special housing with services is neither a desirable nor a realistic option for all disabled older adults, it is an attractive option for many.

The research of Moos and Lemke (1994) has applied a carefully evaluated methodology (the Multiphasic Environmental Assessment Procedure [MEAP])

for characterizing the structural and resident characteristics of group living facilities and generated normative data on how these characteristics relate to outcomes. This methodology is just what is needed to ask two really important questions about assisted living housing: (1) When a philosophy of housing is implemented to maximize autonomy, self-efficacy, social support, and participation in decisions about community, does assisted living housing achieve promised outcomes of quality residential care with economy? (2) Can this philosophy be implemented more widely through feasible interventions?

Bandura (1997) has extended the personal variable self-efficacy to suggest a social and organizational variable collective efficacy in a way that is particularly relevant for understanding the potential of assisted living housing. Individuals can and do experience collectively in particular settings a shared sense of mission, cooperative behavior, and resilience in the face of challenge to one's group. The empowerment or enablement implied in one's sense of self-efficacy is not bestowed by edict but is an observed outcome based on experience that conjoint behavior can be organized and implemented to attain group objectives. The concept of collective efficacy has been used in a variety of community research settings to explore how a neighborhood, for example, can develop a sense of social cohesion and unity of purpose that promotes not only networks of social support but also willingness to intervene in behalf of attaining some shared social good (see, e.g., Sampson et al., 1997). The concept clearly has relevance to explore how, in an assisted living setting, a sense of collective efficacy in support of compensatory care role modeling could arise and be sustained.

A feasible intervention cannot be spelled out in detail here, but what is suggested meets the essential tests this observer learned to apply over a quarter century of “study sectioning” to ask about proposals under review at the National Institutes of Health. Is the theory behind a proposal sound and will theory be advanced? Is methodology available that will make what is proposed feasible? And is the proposal significant, certainly in the sense of being compatible with the priorities of the research community and, if possible, also recognizable within the priorities of the policy and attractive to programming persons responsible for advancing the public's health.

On all of these counts the recommended intervention would receive high marks. The theoretical background is very strong. The experience of Moos and Lemke has obvious applicability to assisted living housing as a tested systems analytic strategy for identifying the structuring of organizations likely to improve the functionality, sense of well-being, and adaptability of frail older adults. The population likely to benefit is large, as large—probably larger—than the current nursing home population. And one final note: An intervention intended to benefit assisted living housing residents would necessarily deal with a key issue in the implementation of information from the initiative proposed —the intersection of public and private interests. In the best of all possible worlds in the United States, the aging care business will discover that “good care is good business.” It is just such a discovery that is most likely to promote the public's health in an aging society and generate a political will reinforced by private eco-

nomic incentives to promote careful study of assisted living as one promising feasible option for dealing with older adults at risk for a level of dependency requiring some level of compensatory care.

The Moos and Lemke research that lays the groundwork for the recommended intervention has included more than 300 group residences—congregate, residential, sheltered environments, and nursing homes. But regardless of the type of shelter, the MEAP strategy has proved to be applicable in characterizing (1) resident and staff; (2) physical and architectural features; (3) the basic objectives of institutional policy and programs; and (4) the social climate of the housing community.

In the initial research, structural features such as ownership, size, and level of care required were found to be important. But more important was the factor of social climate, particularly a climate in which autonomy and choice are legitimate and promoted; residents are involved in decisions about their community; and staff are committed to a climate in which privacy, choice, and resident involvement are encouraged.

Initial research has already provided the preliminary documentation that a pro-choice, pro-involvement social climate promotes well-being, better adaptation to change, and reduction of inappropriate use of facility care resources. These are all evidence of positive outcomes for frail older adults that theory would predict.

In their final chapter, Moos and Lemke themselves point to the next task: Now that we have identified the favorable outcomes deliverable in assisted living housing dedicated to enhancing a pro-choice and pro-involvement social climate, we should demonstrate how to promote such a social climate in assisted living facilities more widely. This is a challenge of one significant and promising intervention in the interest of the public's health in later life likely to enhance the functioning and well-being of a large number of frail older adults while reducing unnecessary use of care resources. Assisted living housing is not the only feasible intervention to improve the public 's health in our aging society. But it is an opportunity with the possibility of significant positive effect.

Antonovosky, A. ( 1987). Unraveling the Mysteries of Health: How People Manage Stress and Stay Well. San Francisco: Jossey Bass.

Baltes, M. ( 1996). The Many Faces of Dependency in Old Age. Cambridge, UK: Cambridge University Press.

Baltes, P., and Baltes, M. ( 1990). Psychological perspectives on successful aging: The model of selective optimization with compensation. In P.Baltes and M.Baltes (eds), Successful Aging: Perspectives from Behavioral Sciences. Cambridge, UK: Cambridge University Press.

Bandura, A. ( 1997). Self-Efficacy: The Exercise of Control. New York: Freeman.

Berkman, L. ( 1988). The changing and heterogeneous nature of aging and longevity: A social and biomedical perspective. In G.Maddox and P.Lawton (Eds.). Varieties of

Aging. Annual Review of Gerontology and Geriatrics, Volume 8, pp. 37–67. New York: Springer Publishing Company.

Berkman, L., and Breslow, S. ( 1983). Health and Ways of Living: The Alameda County Study. New York: Oxford University Press.

Berkman, L., and Syme, L.S. ( 1979). Social networks, host resistance and mortality: Anine year follow-up of Alameda County residents. American Journal of Epidemiology, 109, 684–694.

Campbell, A., Converse, P., and Rogers, W. ( 1976). The Quality of American Life. New York: Russell Sage Foundation.

Clark, D., ( 1997). Physical activity efficacy and its correlates among socioeconomically Disadvantaged older adults. Diabetes Care, 20(7), 1176–1182.

Clark, D., Maddox, G., and Steinhauser, K. ( 1993). Race, aging, and functional health. Journal of Aging and Health, 5(4), 536–553.

Clark, D., and Nothwehr, F. ( 1999). Exercise self-efficacy and its correlates among socioeconomically disadvantaged older adults. Health Education and Behavior, 26(4), 535–546.

Evans, R., Barer, M., and Marmor, T. ( 1994) Why Are Some People Healthy and Others Not? The Detrements of Health Populations. New York: Aldine de Gruyter.

Fries, J. ( 1980). Aging, natural death and the compression of morbidity. New England Journal of Medicine, 303, 130–36.

Fries, J. ( 1995). Compression of morbidity/disease postponement. In G.Maddox (Ed.), The Encyclopedia of Aging, 2nd edition, pp. 213–216. New York: Springer Publishing Company.

Glass, T., Menzes de Leon, C., and Berkman, L. ( 1999). Population based study of productive activities as predictors of survival among elderly Americans. British Medical Journal, 319, 478–482.

Gold, D., Stegmaier, K., Bales, C., et al. ( 1993). Psychosocial functioning and osteoporosis in late life: Results from a multidisciplinary intervention. Journal of Women's Health, 2(2), 149–155.

Guralnik, J., Land, K., Blazer, D., et al. ( 1993). Educational status and active life expectancy in older blacks and whites. New England Journal of Medicine, 329, 110–116.

House, J., Landis, K., and Umberson, D. ( 1988). Social relationships and health. Science, 241, 540–545.

Kane, R., and Wilson, K. ( 1993). Assisted Living in the United States: A New Paradigm for Residential Care for Frail Older Persons? Washington, DC: American Association of Retired Persons.

Kohn, M. ( 1995). Social structure and personality through time and space. In: P.Moen, G.Elder, and K.Luscher (Eds.). Lives in Context: Perspectives on the Ecology of Human Development, pp. 141–169. Washington, DC: American Psychological Association.

Lonergan, E., (Ed.). ( 1991) Extending Life, Enhancing Life: A National Research Agenda. Washington, DC: Institute of Medicine, National Academy of Sciences.

Maddox, G. ( 1987). Aging differently. The Gerontologist, 27(5), 557–564.

Maddox, G., and Clark, D. ( 1992). Trajectories of functional impairment in later life. Journal of Health and Social Behavior, 33(2), 114–125.

Maddox, G., Clark, D., and Steinhauser. K. ( 1994). Dynamics of functional impairment in late adulthood. Social Science and Medicine, 38(7), 925–936.

Maddox, G., and Glass, T. ( 1999). Sociology of aging. In W.Hazzard et al., (Eds.). Principles of Geriatric Medicine and Gerontology, pp. 191–202. New York: McGraw Hill.

Manton, K. ( 1988). Planning long term care for heterogeneous older populations. In G. Maddox and P.Lawton (Eds.), Varieties of Aging. Annual Review of Gerontology and Geriatrics, Volume 8. New York: Springer Publishing Company.

Manton, K., Corder, L., and Stollard, E. ( 1997). Chronic disability trends in the U.S. elderly population 1982 to 1984. Proceedings of the National Academy of Sciences, 94:2593–2598.

Markides, K., and Wallace, S. ( 1996). Health and long-term care needs of minority elderly. In J.Romeis, R.Coe, and J.Morley (Eds.). Applying Health Services Research to Long-Term Care. New York: Springer Publishing Company.

Marmor, T., and Okma, G. ( 1998). Societal intervention affecting elderly citizens: A comparative approach . In R.Schulz, G.Maddox, and P.Lawton (Eds.), Interventions Research with Older Adults. Annual Review of Gerontology and Geriatrics, Volume 18, pp. 321–338. New York: Springer Publishing Company.

McAuley, J., and Katula, J. ( 1998). Physical activity interventions in the elderly: Influence on physical health and psychological function. In R.Schulz, G.Maddox, and P. Lawton (Eds.). Interventions Research with Older Adults. Annual Review of Gerontology and Geriatrics, Volume 18, pp. 111–155. New York: Springer Publishing Company.

Mollica, R. ( 1998) State Assisted Living Policy: 1998. Portland, ME: National Academy of State Health Policy.

Moos, R., and Lemke, S. ( 1994). Group Residences for Older Adults: Physical Features, Policies, and Social Climate. New York: Oxford University Press.

Regnier, V., Hamilton, J., and Yatabe, S. ( 1995). Assisted Living for the Aged: Innovations in Design, Management, and Financing. New York: Columbia University Press.

Reynolds, C., et al. ( 1998). Behavioral and pharmacologic interventions for depression in later life. In R.Schulz, G.Maddox and P.Lawton (Eds.). Focus on Interventions Research with Older Adults. Annual Review of Gerontology and Geriatrics , Volume 18, pp. 48–73. New York: Springer Publishing Company.

Roberts, E., Kaplan, G. et al. ( 1997). Does growing old increase the risk for depression? American Journal of Psychiatry, 154(10), 1384–1390.

Rowe, J., and Kahn, R. ( 1987). Human aging: Usual and successful. Science, 237, 143– 149.

Sampson, R., Rudenbush, S., and Earls, F. ( 1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science 27, 918–925.

Schulz, R., Maddox, G., and Lawton, P. ( 1998). Interventions Research with Older Adults. Annual Review of Gerontology and Geriatrics, Volume 18. New York: Springer Publishing Company.

Stewart, A., et al. ( 1989). Functional status and well-being of patients with chronic conditions . Journal of the American Medical Association, 262(7), 907–913.

Townsend, P., and Davidson, N. ( 1982). Inequalities in Health: The Black Report. London: Penguin.

U.S.Department of Health and Human Services. ( 1999). Health, United States, 1999: Health and Aging Chartbook. Washington, DC: DHHS.

Verbrugge, L. ( 1995). Sex differences in health. In G.Maddox (Ed.), Encyclopedia of Aging, 2nd edition, pp. 850–854. New York: Springer Publishing Company.

Wells, K., et al. ( 1987). The functioning and well-being of depressed patients: Results from the MOS study. Journal of the American Medical Association, 261(7), 914–919.

Wiener, J., et al. ( 1990). Measuring the activities of daily living across national surveys. Journal of Gerontology: Social Sciences, 45(6), 229–237.

Wilson, K. ( 1996). Assisted Living. Washington, DC: AARP.