HEALTH AND ILLNESS BEHAVIOR

Health and illness behaviors are associated with level of disability, quality of life, patterns of illness, and risk of death. It is tempting to view such health-related outcomes solely through the lenses provided by the biomedical sciences; however, the behaviors that importantly shape individuals' experiences of sickness or wellness, and life or death, are more completely understood from a sociological perspective. The confluence of individuals' life histories, their personality characteristics and social experiences, and their social positions influences health and illness behaviors and tells us much about how to enhance health and well-being, and mitigate disability and sickness. An examination of health and illness behaviors, therefore, has important public health implications.

HEALTH BEHAVIORS

Health behavior usually refers to preventive orientations and the positive steps people take to enhance their physical well-being and vitality. Traditionally, work in health behavior has focused on the use of preventive services such as immunizations, medical checkups, hypertension screening, and prophylactic dentistry (Becker 1974). It also includes research on such behaviors as cigarette smoking, seat-belt use, medication adherence, substance abuse, nutritional practices, and exercise (Janz and Becker 1994).

The conventional approach to health behavior has been limited, focusing on the origins of particular behaviors damaging to health and strategies to modify them. The most widely used general model—the health belief model—conceptualizes preventive health action within a psychological cost-benefit analysis (Rosenstock 1974). The health belief model conceptualizes decisions to take positive health actions as motivated by perceived threat (either susceptibility to a particularly condition or perceptions that the condition is severe) and judgments about the barriers and benefits associated with specific changes in behavior. Behavior change is seen as following motives that are salient, in situations where people have conflicting motives, following those that are perceived as yielding valuable benefits. An important component of the model involves cues to action, since an activating stimulus often appears to be necessary in the initiation of a new behavioral sequence. Both internal (e.g., feelings of symptoms) and external (e.g., suggestions from doctors, peers, or the media) stimuli may act as cues motivating change. Over the years, this model has been expanded (Becker and Maiman 1983), but it serves more as an organizing framework for the study of preventive health behavior than as a successful predictive model. An analysis of studies that have used the health belief model to explain a variety of health behaviors indicates that the predictive value of the model is, at best, modest; the average variance in health behaviors explained is approximately 20 percent (Harrison, Mullen and Green 1992).

A second commonly used model to explain health decisions is the theory of planned behavior, originally developed as the theory of reasoned action (Ajzen 1991; Ajzen and Fishbein 1977). Like the health belief model, the theory of planned behavior conceptualizes changes in behavior as products of the saliency of individuals' beliefs about the potential costs and benefits associated with an outcome or action. The theory of planned behavior, however, places greater weight on persons' intentions, arguing that behavior is centrally motivated by intentions that are shaped by normative beliefs, feelings of control, and judgments about the barriers and benefits associated with potential change. Again, however, the model has had only modest success in predicting an array of health behavior; the association between intentions and behavior is typically about .40 (Conner and Norman 1994).

There are many other models and theories proposed to predict health behaviors, and much has been written comparing the relative efficacy of each for predicting health behaviors (Conner and Norman 1994; Mullen et al. 1987; Weinstein 1993). It appears that efforts to develop a general theory are limited by the fact that behavior conducive to health derives from diverse and sometimes conflicting motives. Consistently, research indicates that health behavior, or a healthy lifestyle, is not a unitary construct (Johnson et al. 1998; Sobal et al. 1992). One study of health behaviors among a nationally representative sample of adult Americans examined the clustering of four health behaviors: diet quality, alcohol use, tobacco use, and physical exercise (Patterson et al. 1994). The results (based on data from the late 1980s) suggest that approximately 10 percent of Americans live a "healthy" lifestyle, defined by a good diet, low use of tobacco and alcohol, and engaging in regular physical exercise. In comparison, about 2 percent of the population practice unhealthy behavior on all four of these dimensions. Thus, most Americans fall somewhere between the two extremes, practicing some positive health behaviors while neglecting others. Research that has attempted to establish specific clusters of related health behaviors has proved inconclusive. Some studies have identified many apparently interrelated clusters of behaviors—for example, smoking/drinking and exercise/diet—and others have identified few (cf.Johnson et al. 1998; Sobal et al. 1992). In short, the research indicates that there is no simple identifiable positive health orientation that can serve as a basis for promoting risk aversion and health maintenance.

The lack of such a general orientation results because most behaviors with important implications for health arise from motives not related to health and are significantly programmed into the daily patterns and institutional life of communities and families (Mechanic 1990). Health-protective behaviors that are consequences of accepted, everyday, conventional activities require neither conscious motivation nor special efforts to be sustained. The favorable health experience of Mormons, for example, is a product of their belief systems and patterns of activity reinforced by the way of life of this cultural community (Mechanic 1990). To the extent that health behaviors are more the result of habits than cognitive decisions, we might expect that past health behaviors are robust predictors of current health behaviors. Yet neither the theory of reasoned action nor the health belief model adequately incorporates past behavior in its model, an omission that might partially explain their lack of predictive power (Conner and Norman 1994).

Promoting health may be more a matter of changing culture and social structure than of modifying personal motives or intentions. Patterns of behavior that depend on sustained conscious motivation are less stable than those that are a natural consequence of the accepted norms and understandings within a community. Expectations not only affect the prevalence of varying behaviors but also establish constraints on the acquired behaviors of children and adolescents. Changes in the social constraints on smoking, and the growing unacceptability of smoking in varying social contexts, may have more significance than any program to change personal behavior for explaining the dramatic decline from about 42 percent of the U.S. adult population being current smokers in 1965 to 25 percent in 1995 (National Center for Health Statistics 1998).

Although there is no evidence for a unitary health orientation, some social factors, particularly socioeconomic status (SES), predict good outcomes across a wide range of health indicators (Bunker et al. 1989; Marmot 1998; Ross and Wu 1995). Occupational status, income, and education each reflect some part of SES, and all are associated with health behaviors, whether one is comparing the health behaviors of populations or of individuals. Generally, wealthy nations show the highest rates of preventative health practices, such as child immunization, routine dental care, and the use of mammography, compared with less wealthy nations. But the importance of SES may be indirect, through social conditions. Indeed, as Caldwell (1986, 1993) has argued, mothers' educational attainment appears to be particularly important, influencing health outcomes net of its relationship to per capita income. Caldwell suggests that maternal education increases women's autonomy, enhances their ability to interact efficaciously with available health services and technology (even when such technology is not advanced), and enables women to increasingly control their own health and that of family members.

Within nations, there are also important SES differences in health behaviors. Table 1 presents some examples of SES differences in the U.S. population; the behaviors presented are meant to illustrate the gap and are not an exhaustive list of possible health behaviors. As shown in Table 1, Americans with lower SES are more likely to engage in health-risk behaviors and less likely to engage in health-promoting behaviors. The precise ways in which SES affects these outcomes are not fully understood, but the consistent findings point to an explanation of health behaviors that goes beyond personal responsibility and free choice. Research that examines the impact of SES over the life course illustrates the fallacy of relying on "choice" explanations to account for SES differences in health behaviors. In a sample of Finnish men, Lynch and his colleagues (1997) examined the effects of SES during childhood (parents' occupation), adolescence (education), and adulthood (occupation) on psychosocial characteristics, such as hostility and hopelessness, that are important to health and on health behaviors, including smoking and alcohol use. They found that lower SES in childhood and adolescence is associated with greater health-risk behaviors in adulthood, in addition to greater feelings of hopelessness and hostility. Given that childhood SES is not a matter of choice, their findings support an explanation of SES differences in health behavior rooted in persistent structural disadvantages and the accompanying differential opportunities and constraints. Others have also demonstrated that higher SES provides not only obvious economic advantages and related opportunities but also enhanced personal autonomy, increased sense of control, and greater social participation (Marmot 1998; Ross and Wu 1995, 1996), all of which also influence health outcomes.

A variety of behaviors noxious to health (smoking, drug use, and drinking) develop or increase during adolescence and young adulthood. However, young people who have a good relationship with their parents and who are attuned to parent-oriented values—as measured by school performance, attendance at religious services, and participation in meals with parents—do relatively well across a variety of health measures (Hansell and Mechanic 1990). In contrast, high engagement with peer-oriented social activities is associated with increases in behavior associated with health risk. In addition, children model their parents' health behaviors, an effect that persists at least into young adulthood (Lau et al. 1990).

Although it is apparent that adolescence is a time of life where there are likely to be changes in health behaviors, we know very little about other stages of the life course or the life transitions that may be especially important (Prohaska and Clark 1997). Prohaska and his colleagues propose a "stages of change" model that recognizes important transitions in the life course as explanations for changes in health practices (Prohaska and Clark 1997; Prohaska et al. 1994). They argue that individuals go through a number of steps, from not thinking there is a need for change to maintaining the new behavior after change. Public health efforts, therefore, could benefit from understanding what motivates or hinders a person's progression through the steps. For instance, life transitions such as motherhood may motivate progression toward positive health practices, while transitions such as death of a spouse may make it difficult to maintain health practices and, therefore, may explain deterioration of positive health behaviors among persons recently widowed. Moreover, according to this model, persons may be differently prepared to progress through the steps depending

on their stage in the life course; for example, older persons may be better prepared to contemplate the health risks of smoking or heavy alcohol use, while adolescents are not. The important point for our purposes is that understanding how to modify health practices requires appreciation of how stages in the life course and transitions may influence when people are willing or able to make changes.

ILLNESS BEHAVIOR

The study of illness behavior, in contrast to health behavior, is concerned with the way people monitor their bodies, define and interpret bodily indications, make decisions about needed treatment, and use informal and formal sources of care (Mechanic 1986, 1995). Like other behavior, illness behavior is learned through socialization in families and peer groups and through exposure to the mass media and education. There is great diversity of attitudes, beliefs, knowledge, and behavior, all of which affect the definitions of problematic symptoms, the meanings and causal attributions that explain them, socially anticipated responses, and the definition of appropriate remedies and sources of care. Motivation and learning affect the initial recognition of symptoms, reactions to pain, the extent of stoicism and hypochondriasis, and the readiness to seek release from work, school, and other obligations and to seek help (Mechanic 1978).

Illness behavior begins prior to the use of services with the recognition of illness or sickness. While a complex array of variables might explain variations in interpretation of sickness, they can be summarized in ten general categories: (1) the visibility, recognizability, or perceptual salience of deviant signs and symptoms; (2) the extent to which the person perceives the symptoms as serious (that is, the person's estimate of the present and future probabilities of danger; (3) the extent to which symptoms disrupt family, work, and other social activities; (4) the frequency of the appearance of deviant signs and symptoms, or their persistence, or their frequency of recurrence; (5) the tolerance threshold of those who are exposed to and evaluate the deviance signs of symptoms; (6) the information available to, the knowledge of, and the cultural assumptions and understanding of the evaluator; (7) the degree to which autistic psychological processes (perceptual processes that distort reality) are present; (8) the presence of needs that conflict with the recognition of illness or the assumptions of the sick role; (9) the possibility that competing interpretations can be assigned to the symptoms once they are recognized; and (10) the availability of treatment resources, their physical proximity, and the psychological and monetary costs of taking action (including not only physical distance and costs of time, money, and effort, but also stigmatization, resulting social distance, and feelings of humiliation resulting from a particular illness decision) (Mechanic 1978).

In short, illness appraisal is a two-step process (Mechanic 1972). In the initial step, persons monitor their bodies to assess the location, duration, intensity, and persistence of discomfort. In the second stage, which may occur almost concurrently, they seek explanations for perceived changes. If an obvious explanation is not available, or is disconfirmed by further checking, individuals look to their environment for new cues and explanations. These interpretations, in light of knowledge and other beliefs, may then play a role in the formal initiation of care.

Individuals' appraisals of symptoms vary importantly. Processes of symptom appraisal are influenced by the manner in which symptoms occur and their characteristics, by knowledge, and by past experiences with illness (Leventhal 1986; Mechanic 1972). Some symptoms are so painful and incapacitating that they inevitably lead to intervention without significant inquiry. Others are so familiar and generally understood as self-limited that they also are dealt with routinely. Many symptoms, however, are neither familiar nor easily understood, resulting in a process of interpretation within the context of personality, situational cues and stressors, and environmental influences. Only a small proportion of symptoms lead to formal consultation or care. The vast majority are denied, normalized, or evaluated as having little significance

Much research has focused on how persons come to make judgments of their own health status. One approach is to study health appraisals among children and adolescents who have little serious illness. In a prospective study of adolescents, those who were more competent and more engaged in age-related activities, as measured by school performance and participation in sports and other exercise, rated their health more highly (Mechanic and Hansell 1987). Adolescents' health assessments are shaped by their overall sense of functioning, and they do not seem to differentiate among physical and psychological aspects of well-being in making general assessments of how they feel.

This finding is consistent with the body of research examining the impact of global judgments of health on mortality and disability. Several longitudinal studies of the elderly have found that subjective self-assessments of health predict future mortality after taking account of known risk factors and sociodemographic measures (Idler and Angel 1990; Idler and Benyamini 1997; Idler and Kasl 1995; Mossey and Shapiro 1982). Likewise, some research suggests that global assessments of health predict future level of functioning or disability (Farmer and Ferraro 1997; Idler and Kasl 1995). It is remarkable that simple self-assessments of health prospectively predict longevity and disability better than physicians' assessments or known health-risk factors. Many studies suggest that judgments of health and well-being and interpretations of sickness are shaped by factors beyond those traditionally captured by biomedical conceptions of illness. Individuals' appraisals of their health appear to depend as much on their global sense of well-being as they do on specific patterns of illness. The identification of the exact psychological and social factors responsible for self-assessed health, however, remains elusive. Some have suggested that individuals take into account important social and psychological resources, such as social support, feelings of control, and optimism, when making judgments of their own health—and that these psychosocial resources provide protection against morbidity and mortality (Idler and Benyamini 1997; Kaplan and Camacho 1983).

One of the most consistent research findings has been that perceptions of physical illness depend on the person's psychological well-being or level of psychological distress (Aneshensel et al. 1984; Farmer and Ferraro 1997; Tessler and Mechanic 1978). Individuals tend to assess their health holistically in terms of vitality and capacity to perform their social activities and roles. Psychological distress often diminishes vitality and functioning as much as serious medical conditions. The RAND Medical Outcomes Study found that depressive symptoms were more disabling than many chronic physical conditions that physicians view as extremely serious (Wells et al. 1989).

Persons seeking medical care commonly express their distress and lowered sense of well-being through many diffuse physical complaints, such as fatigue, insomnia, and aches and pains in different bodily systems, a process referred to as somatization (Kleinman 1986). Although much of the existing literature focuses on somatization as a problematic process, it is by far the predominant pattern for expressing distress. Until the last forty or fifty years, it was uncommon to conceptualize distress in psychological terms, and even now such expressions are used primarily among well-educated populations receptive to psychological interpretations. General distress has both physical and psychological concomitants; the language that people use to characterize distress depends on the cultural context, the perceived appropriateness of psychological complaints, and the stigma attached to emotional disorder.

There is controversy as to whether psychological idioms are inaccessible to many people due to cultural factors or limited schooling, or whether somatization represents a choice among alternative idioms because such presentations are seen as more consistent with the medical care context. Rates of reported depression in Chinese cultures, for example, are extremely low, although "neurasthenia" is a common diagnosis in Chinese medical care settings (Kleinman 1986). Psychiatrists in China routinely view neurasthenia as a "disorder of brain function involving asthenia of cerebral cortical activity," but the symptoms reported are strikingly similar to the physical manifestations of depressive disorders more commonly seen in Western countries. It remains unclear whether these diagnoses characterize the same underlying disorders that are expressed differently in varying cultural groups or whether they are fundamentally different. Kleinman (1986) treated Chinese patients diagnosed as neurasthenic who met the American diagnostic criteria for major depression with antidepressant drugs but found on follow-up that while a majority showed significant improvements in clinical psychiatric symptoms, they continued to be impaired, to function badly, and to seek help for their condition. They also remained skeptical of the drug treatment they received. Kleinman links these responses to the patients' needs for the medical legitimization of their "illness" to explain past failures and to justify continuing difficulties in meeting social expectations.

Similar examples are provided within the American context when "new" disorders become part of our popular understanding of what is (and is not) physical illness. For example, in recent years we have seen much debate around the legitimacy of chronic fatigue syndrome and Gulf War illness as physical disorders (Abbey and Garfinkel 1991; NIH 1994; Presidential Advisory Committee 1996). The case definitions associated with these conditions are not well specified, and the etiology of each remains undetermined. Those suffering from the symptoms of each condition have benefited greatly from the efforts of interest groups who have lobbied to have the conditions treated as unique clinical disorders. The status of disorder often brings access to insurance coverage and other entitlement programs. Equally important, recognition of a group of symptoms as a specific disorder or distinct illness acts to reduce stigma for those suffering and provides some legitimacy to persons disabled by the symptoms. Studies of patients diagnosed with chronic fatigue syndrome highlight the important role of attribution in the experience of illness. Some research indicates that patients who view their illness as essentially physical have less favorable outcomes than those who attribute the cause of their illness to social or psychological factors (Joyce et al. 1997).

In situations where there are no obvious explanations for the occurrence of symptoms, individuals seek meanings for changes in their feeling states. The commonsense theories they apply may either be idiosyncratic or drawn from socially prevalent conventional explanations such as stress, lack of sleep, overwork, and overeating. These lay explanations are influential on subsequent behavior such as care seeking and use of medication (Kleinman 1980; Leventhal et al. 1980; Leventhal et al. 1985). For example, it is commonly believed that stress increases blood pressure and that relaxation reduces it. Most individuals, however, cannot assess whether their blood pressure is high or low on the basis of available cues, yet many believe they can. Persons with hypertension, an asymptomatic condition, commonly use self-assessments of their stress levels or relaxation as an indicator of their blood pressure levels and adjust their medication accordingly, despite medical advice to the contrary (Leventhal et al. 1985). Similarly, many patients with limited understanding of the biological processes through which drugs such as antibiotics or antidepressants act increase or decrease medication in relation to changes in how they feel and environmental cues.

The decision to seek medical care in response to symptoms also depends on prevailing norms about the conditions that are within the purview of the medical field. Such norms differ across cultures and time. In recent years, the role of the popular media in influencing help-seeking appears to have increased. In the late 1980s, Prozac was widely advertised and discussed as a treatment for a wide range of symptoms ranging from clinical depression to shyness, self-criticism, low self-esteem, or just feeling blue (Barondes 1994; Kramer 1993). It became the most quickly accepted drug in U.S. history (Kramer 1993). In April 1998 Viagra was introduced to the market, and the popular press was inundated with stories of the potential of the drug to increase sexual pleasure. It replaced Prozac as the fastest-selling pharmaceutical in history, and by the end of the year approximately 7 million prescriptions had been written for some 3 million patients in the United States (Pfizer Inc. 1999). Unmet need partially explains the rapid acceptance of both drugs. However, it is also likely that variations in human feelings or behavior that would previously have been normalized were more likely to be viewed as medical conditions requiring treatment.

Sickness is an accepted role in society, bringing sympathetic attention and legitimate release from expected performance (Parsons 1951). Determinations of what is (and what is not) illness may involve intense negotiations about individuals' claims that, when legitimized, may justify failure to meet expectations or allow escape from onerous obligations (Mechanic 1978). In some situations, the sick role becomes a point of tension and conflict between the claimant, who seeks legitimation of sickness with its special privileges, and other interested parties, including families, employers, and welfare administrators, who may seek to limit release from social obligations or diminish special privileges granted to the sick and disabled (Field 1957). Most illness situations are neither problematic nor sources of conflict, but the contested cases make evident the social assumptions and expectations around which illness is organized.

Certification of illness becomes a public issue when physicians, government bureaucrats, or managed care companies have moral and legal authority to define illness and disability and to sanction the sick role. Such influence is found in certifying justified absenteeism for employers, in litigation, and in decisions on eligibility in insurance and disability entitlement programs. Efforts are often made to maintain the illusion that these are objective decisions based solely on medical expertise and clinical experience, but judgments often depend on whom the physician represents. The state or other formal organizations may thus attempt to control physicians by limiting their discretion, as happened in the Soviet Union when physicians were viewed as allowing people excuses to escape work too easily (Field 1957).

American examples of the processes involved in legitimization of illness come from observations of legislative changes affecting definitions of what is (and is not) disability. For example, in 1996 Congress passed legislation that removed substance use as a disability qualifying for Supplemental Insurance Income (SSI). Approximately 140,000 people lost their official status as "disabled" under the Social Security Act and concomitantly lost the right to income and medical benefits (Gresenz et al. 1998). Similarly, substance-use disorders associated with illegal drugs are explicitly excluded from the classes of disorders covered by the Americans with Disability Act (Mechanic 1998a). The removal or exclusion of a class of illness from the definition of disability highlights the importance of societal norms and values in sanctioning which groups of persons may legitimately occupy the sick role. Moreover, much of the debate surrounding implementation of both acts concerned judgments of personal responsibility and "badness" and "goodness," rather than the clinical significance of substance-use disorders as disabling conditions. In contrast, veterans of the Persian Gulf War have been awarded medical and disability entitlements based on Gulf War illness, despite the weight of medical evidence suggesting that there is no such unique clinical entity (NIH 1994; Presidential Advisory Committee 1996). Both examples illustrate the point that the certification of persons as legitimately sick (and thus deserving of public benefits) is not always the result of clinical judgments. Definition of illness and disability is the "rope" in a tug of war in which competing parties seek determinations in their own interest.

The study of illness behavior has many applications in research, clinical care, public health, and social policy (Mechanic 1998b). Such patterns of behavior substantially affect pathways into care and selectively shape the samples studied in clinical contexts. Failure to understand these selection effects and how they operate leads to erroneous conclusions about the nature of basic disease processes. A common error is to attribute to the etiology of the illness influences triggering use of services, a problem that plagues much research on stress and illness. At the clinical level, awareness of how people construe illness, present symptoms, and respond to care can improve understanding and communication and help health professionals to guide the treatment regimen more effectively.

Health and illness behavior studies make clear that the forces affecting health and treatment outcomes transcend medical care and the transactions that take place between doctor and patient. In recent decades there have been increasing tendencies toward the medicalization of social problems and a failure to address the complicated attitudinal needs of patients with serious chronic conditions and disabilities. Moreover, the problems in managing the frailties of old age, characterized by a combination of medical, instrumental, and social needs, increasingly challenge assumptions of treatment focused on narrow definitions of disease and care. Studies of health and illness behavior teach the importance of moving beyond initial complaints and narrow definitions of problems and toward examining the broad context of individuals' lives and the factors that affect social functioning and quality of life. They point to the diverse adaptations among persons with comparable social and physical debility and potential. They reinforce the need to take account of the environmental and social context of people's lives, their potential assets, and their disease. A medical care system responsive to these broad concerns would be better prepared for the impending health care challenges of the new millennium.

references

Abbey, S. E., and P. E. Garfinkel 1991 "Neurasthenia and Chronic Disease Syndrome: The Role of Culture in the Making of a Diagnosis." American Journal ofPsychiatry 148:1638–1646.

Ajzen, I. 1991 "The Theory of Planned Behavior." Organizational Behavior and Human Decision Processes 50:179–211.

——, and M. Fishbein 1977 "Attitude-Behavior Relations: A Theoretical Analysis and Review of Empirical Research." Psychological Bulletin 84:888–918.

Aneshensel, Carol S, Ralph R. Frerichs, and George Huba 1984 "Depression and Physical Illness: A Multiwave, Nonrecursive Causal Model." Journal ofHealth and Social Behavior 25:350–371.

Barondes, Samuel H. 1994 "Thinking About Prozac." Science 263:1102–1103.

Becker, Marshall, (ed.) 1974 The Health Belief Model andPersonal Health Behavior. Thorofare, N.J.: Slack.

——, and Lois Maiman 1983 "Models of Health Related Behavior." In D. Mechanic, ed., Handbook ofHealth, Health Care and the Health Professions. New York: Free Press.

Bunker, John, Deanna Gomby, and Barbara Kehrer, (eds.) 1989 Pathways to Health: The Role of SocialFactors. Menlo Park, Calif.: Kaiser Family Foundation.

Caldwell, John 1986 "Routes to Low Mortality in Poor Countries." Population and Development Review 12:171–220.

——1993 "Health Transition: The Cultural, Social, and Behavioural Determinants of Health in The Third World." Social Science and Medicine 36:125–135.

Conner, Mark, and Paul Norman 1994 "Comparing the Health Belief Model and the Theory of Planned Behavior in Health Screening." In D.R. Rutter and Lyn Quine, eds., Social Psychology and Health: European Perspectives. Brookfield, Vt.: Avebury.

Farmer, Melissa M., and Kenneth F. Ferraro 1997 "Distress and Perceived Health: Mechanisms of Health Decline." Journal of Health and Social Behavior 39:298–311.

Field, Mark G. 1957 Doctor and Patient in Soviet Russia. Cambridge, Mass.: Harvard University Press.

Gresenz, Carole Roan, Katherine Watkins, and Deborah Podus 1998 "Supplemental Security Income (SSI), Disability Insurance (DI) and Substance Abusers." Community Mental Health Journal 34(4): 337–346.

Hansell, Stephen, and David Mechanic 1990 "Parent and Peer Effects on Adolescent Health Behavior." In Klaus Hurrelman and Friedrich Losel, eds., HealthHazards in Adolescence. New York: DeGruyter.

Harrison, J. A., P. D. Mullen, and L. W. Green 1992 "A Meta-Analysis of Studies of the Health Belief Model with Adults." Health Education Research 7:107–116.

Idler, Ellen L., and Ronald J. Angel 1990 "Self-Rated Health and Mortality in the NHANES-I Epidemiologic Follow-Up Study." American Journal of Public Health 80:446–452.

Idler, Ellen L., and Yael Benyamini 1997 "Self-Rated Health and Mortality: A Review of Twenty-Seven Community Studies." Journal of Health and SocialBehavior 38:21–37.

Idler, Ellen L., and Stanislav V. Kasl 1995 "Self-Ratings of Health: Do They Also Predict Change in Functional Ability?" Journal of Gerontology: Social Sciences 50B:S344–S353.

Janz, Nancy K, and Marshall H. Becker 1984 "The Health Belief Model: A Decade Later." Health Education Quarterly 11:1–47.

Johnson, Marilyn F., Jeanne F. Nichols, James F. Sallis, Karen J. Calfas, and Melbourne F. Hovell 1998 "Interrelationships Between Physical Activity and Other Health Behaviors Among University Women and Men." Preventative Medicine 27:536–544.

Joyce, J., M. Hotopf, and S. Wessely 1997 "The Prognosis of Chronic Fatigue and Chronic Fatigue Syndrome. "Quarterly Journal of Medicine 90:223–233.

Kaplan, George, and Terry Camacho 1983 "Perceived Health and Mortality: A Nine-Year Follow-Up of the Human Population Laboratory Cohort." AmericanJournal of Epidemiology 117:292–504.

Kleinman, Arthur 1980 Patients and Healers in the Contextof Culture. Berkeley: University of California Press.

——, 1986 Social Origins of Distress and Disease: Depression, Neurasthenia, and Pain in Modem China. New Haven, Conn.: Yale University Press.

Kramer, Peter D. 1993 Listening to Prozac. New York: Viking.

Lau, Richard R., Marilyn Jacobs Quadrel, and Karen A. Hartman 1990 "Development and Change of Young Adults' Preventative Health Beliefs: Influences from Parents and Peer." Journal of Health and Social Behavior 31:240–259.

Leventhal, Howard 1986 "Symptom Reporting: A Focus on Process." In Sean McHugh and T. Michael Vallis, eds., Illness Behavior: A Multidisciplinary Model. New York: Plenum.

——, Daniel Meyer, and David Nerenz 1980 "The Common-Sense Representation of Illness Danger." In Stanley Rachman, ed., Contributions to MedicalPsychology. New York: Pergamon.

——, Thomas Prohaska, and Robert Hirschman 1985 "Preventive Health Behavior Across the Life Span." In James Rosen and Laura Solomon, eds., Preventionin Health Psychology. Hanover, N.H.: University Press of New England.

Lynch, J. W., G. A. Kaplan, and J. T. Salonen 1997 "Why Do Poor People Behave Poorly? Variation in Adult Health Behaviours and Psychosocial Characteristics by Stages of the Socioeconomic Lifecourse." SocialScience and Medicine 44(6):809–819.

Marmot, Michael G. 1998 "Contributions of Psychosocial Factors to Socioeconomic Differences in Health." The Milbank Quarterly 76:403–448.

Mechanic, David 1972 "Social Psychological Factors Affecting the Presentation of Bodily Complaints." The New England Journal of Medicine 286:1132–1139.

——1978 Medical Sociology, 2nd ed. New York: Free Press.

——1986 "Illness Behavior: An Overview." In S McHugh and T. M. Vallis, eds., Illness Behavior: AMultidisciplinary Model. New York: Plenum.

——1990 "Promoting Health." Society 27(2): 16–22.

——1995 "Sociological Dimensions of Illness Behavior." Social Science and Medicine 34:1345–1350.

——1998a Mental Health and Social Policy: The Emergence of Managed Care, 4th ed. Allyn and Bacon.

——1988b "Cultural and Organizational Aspects of Application of the Americans with Disabilities Act to Persons with Psychiatric Disabilites." The MilbankQuarterly 76:5–23.

——, and Stephen Hansell 1987 "Adolescent Competence, Psychological Well-Being and Self-Assessed Physical Health." Journal of Health and Social Behavior 28:364–374.

Mossey, Jana M, and Evelyn Shapiro 1982 "Self-Rated Health: A Predictor of Mortality Among the Elderly." American Journal of Public Health 72:800–808.

Mullen, Patricia D., James C. Hersey, and Donald C. Iverson 1987 "Health Behavior Models Compared." Social Science and Medicine 11:974–981.

National Center for Health Statistics 1998. Health, United States, 1998, with Socioeconomic Status Health Chartbook. Hyattsville, Md.: Author.

NIH Technology Assessment Workshop Panel 1994 "The Persian Gulf Experience and Health." Journalof the American Medical Association 274:391–396.

Parsons, Talcott 1951 The Social System. New York: Free Press.

Patterson, Ruth E., Pamela. S. Haines, and Barry M. Popkin 1994 "Health Lifestyle Patterns of U.S. Adults." Preventative Medicine 23:453–460.

Pfizer Inc 1999 "Pfizer Inc Fourth-Quarter Total Revenues Increased 26 Percent; Net Income from Continuing Operations, Excluding Unusual Items, Increased 42 Percent." Company Press Release, January 19.

Presidential Advisory Committee on Gulf War Veteran's Illness 1996 Presidential Advisory Committee onGulf War Veteran's Illness: Final Report. Washington, D.C.: U.S. Government Printing Office.

Prohaska, Thomas R., and Melissa A. Clark 1997 "Health Behavior and the Human Life Cycle." In David S. Gochman, ed., Handbook of Health Behavior Research.Vol. III. Demography, Development, and Diversity New York: Plenum.

Prohaska, Thomas R., W. Velicer, J. Rossi, M. Goldstein, B. Marcus, W. Rakowshi, C. Fiore, L. Harlow, C. Redding, D. Rosenbloom, and S. Rossi 1994 "Stages of Change and Decisional Balance for 12 Problem Behaviors." Health Psychology 13:39–46.

Rosenstock, Irvin 1974 "The Health Belief Model and Preventive Health Behavior." In Marshall Becker, ed., The Health Belief Model and Personal Health Behavior. Thorofare, N.J.: Slack.

Ross, Catherine E., and Chia-ling Wu 1995 "The Links Between Education and Health." American Sociological Review 60:719–745.

——1996 "Education, Age and the Cumulative Advantage in Health." Journal of Health and Social Behavior 37:104–120.

Sobal, Jeffery, Dennis Revicki, and Bruce R. DeForge 1992 "Patterns of Interrelationships Among Health-Promotion Behaviors." American Journal of Preventative Medicine 8:351–359.

Tessler, Richard, and David Mechanic 1978 "Psychological Distress and Perceived Health Status." Journalof Health and Social Behavior 19:254–262.

Weinstein, Neil D. 1993 "Testing Four Competing Theories of Health-Protective Behavior." Health Psychology 12:324–333.

Wells, Kenneth, Anita Stewart, Ron D. Hays, Audrey Burnam, William Rogers, Marcia Daniels, Sandra Berry, Sheldon Greenfield, and John Ware 1989 "The Functioning and Well-Being of Depressed Patients: Results from the Medical Outcomes Study." Journal of the American Medical Association 262:914–919.

David Mechanic

Donna D. McAlpine