MENTAL ILLNESS AND MENTAL DISORDERS

After years of empirical research and theoretical activity, social scientists still do not agree about what mental illness actually is, let alone about what its primary causes are or about the efficacy of various treatments. Sociologists disagree about whether or not mental disorder is truly a disease that some people have and other people do not have, thus fitting a medical model of health and illness. They disagree about the relative importance of genetics, biochemical abnormalities, personality characteristics, and stress in the onset and course of psychiatric impairment. Most sociologists do, however, agree that definitions of mental illness are shaped by the historical, cultural, and interpersonal contexts within which they occur. They argue that the significance of any particular set of psychological or behavioral symptoms to a diagnosis of mental disorder lies in part with the actor and in part with the audience. Given this understanding of mental illness, sociologists are often as interested in understanding the consequences of being labeled mentally ill as they are in understanding the causes. Sociologists do, indeed, study the social distribution and determinants of mental disorder. However, they also study social reactions to mental illness and the mentally ill and investigate ways in which mental health professionals and institutions can come to serve as agents of social control.

CLASSIFICATION AND DIAGNOSIS

Although psychiatrists themselves have difficulty defining mental illness, the official system for classifying and diagnosing mental disorder in the United States is produced by the American Psychiatric Association (APA). It is known as the Diagnostic and Statistical Manual of Mental Disorder (DSM) and was first published in 1952. In its earliest form, DSM-I included a list of 60 separate mental illnesses. By the second edition in 1968, psychiatric definitions of mental illness had changed so markedly that 145 different types of mental disorder were included. Despite the attempt in DSM-II to define the parameters of mental illness more precisely, critics from inside and outside psychiatry pointed out that diagnoses of mental disorder were extremely unreliable. When different psychiatrists independently used DSM-II to diagnose the same patients, they did so with substantially different results. Studies conducted during the 1960s and 1970s indicated that there was poor agreement about what disease classification was appropriate for any given patient; studies also found that clinicians had difficulty in differentiating normal persons from mental patients and that they frequently disagreed about prognosis and the clinical significance of particular symptom patterns (Loring and Powell 1988).

After years of debate, some of which was quite heated, the APA published a third edition of DSM in 1980. Mental disorder was defined in DSM-III as "a clinically significant behavioral or psychological syndrome or pattern that occurs in an individual and that is associated with either a painful symptom (distress) or impairment in one or more areas of functioning (disability)" (APA 1980, p. 6); deviant behaviors and conflicts between individuals and society were specifically excluded from this definition unless they were symptoms of another diagnosable disorder. DSM-III took a purely descriptive approach to diagnosis, outlining the essential and associated features of each disorder but making no attempt to explain the etiology of either symptoms or illnesses. At the time of the most recent revision (DSM-IV) in 1994, there was considerable debate about whether or not to actually retain the phrase "mental disorder" in the title of the DSM. The argument was that doing so implied a false distinction between mind and body, and between mental disorders and physical or general medical conditions. For lack of a better term, however, DSM-IV retained the same terminology and definition of mental disorder that had been used in DSM-III. DSM-IV contains hundreds of mental diagnoses, including such disorders as caffeine intoxication, circadian rhythm sleep disorder, and hypoactive sexual desire disorder. It also contains a section on "other conditions that may be a focus of clinical attention" but that are not mental disorders themselves. Included here are such conditions as relational problems connected with a family member's mental disorder, noncompliance with treatment, religious or spiritual, academic, occupational, acculturative, and phase-of-life problems.

The use of DSM-III and DSM-IV criteria has vastly improved the overall reliability of psychiatric diagnoses, thereby enabling psychiatry to meet one of the major criticisms of the medical model of mental disorder. The inclusion of more and more categories of illness in each succeeding version of DSM has led to more precise and consequently more reliable diagnoses. However, some scholars have argued that this expansion of mental diagnoses has less to do with problems of disease classification than with "problems" of third-party reimbursement (Kirk and Kutchins 1992; Mirowsky and Ross 1989). Each increase in the number of disorders listed in DSM has increased the scope of psychiatric practice. As the number of patients with recognized illnesses increases, so too does the amount of compensation that psychiatrists receive from insurance companies.

Even the firmest supporters of DSM-IV recognize that the classification of mental disorder is influenced by nonmedical considerations. In fact, DSM-IV itself includes a discussion of specific culture, age, and gender features which should be taken into account for each diagnosis. Pressures from outside psychiatry also influence diagnostic classifications. In order to reduce their payment liabilities, insurance companies have lobbied the APA to reduce the number of diagnoses. Changes in public attitudes toward sexual preference issues led, in 1974, to dropping homosexuality from the list of mental disorders, and veterans' groups successfully pressed for the inclusion of posttraumatic stress syndrome (Scott 1990). The storm of controversy that surrounds the issue of whether premenstrual syndrome is a medical or a psychological condition, or whether it is socially unacceptable behavior, is another example of the intersection of political, social, economic, and diagnostic concerns (Figert 1994); at the insistence of feminists inside and outside psychiatry, premenstrual dysphoric disorder has been relegated to an appendix of DSM-IV on the basis that there is currently little scholarly evidence to support such a diagnosis and that its social implications dangerously feed "the prejudice that women's hormones are a cause of mental illness" (Tavris 1993, p. 172). In sum, there is a less than perfect correspondence between some disease-producing entity or syndrome and the diagnosis of mental disorder; psychiatric diagnosis is based partly in the reality of disordered behavior and emotional pain, and partly in the evaluations that society makes of that behavior and pain. Thus, questions about the validity of psychiatric diagnosis are as troubling for DSM-IV as they were for DSM-I. As one observer has noted, "We have learned how to make reliable diagnoses, but we still have no adequate criterion of their validity" (Kendell 1988, p. 374). Given the problems scholars have in defining "mental disorder" and given the validity problems that ensue, it is not surprising that epidemiologists have used a number of different strategies to estimate rates of psychiatric impairment. These different research methodologies often have led to quite different interpretations of the role of social factors in the etiology of mental illness.

MEASUREMENT

The earliest sociological research on mental disorder relied on data from individuals receiving psychiatric care. In a classic epidemiological study, Faris and Dunham (1939) reviewed the records of all patients admitted to Chicago's public and private mental hospitals between 1922 and 1934. They found that admission rates for psychosis were highest among individuals living in the inner city. Several years later, researchers used a similar design to study the social class distribution of mental disorder in New Haven. In contrast to the Chicago study, which focused only on individuals who had been hospitalized, Hollingshead and Redlich (1958) included individuals receiving outpatient care from private psychiatrists in their study. Results from the New Haven study confirmed the earlier findings; the lower the social class, the higher the rate of mental disorder. More recent studies have also used information on treated populations. Studies of patient populations provide useful information to be sure; findings shed light on the social factors that influence the course of mental health treatment. Individuals receiving treatment for mental disorder, however, are not a random subset of the population of individuals experiencing psychological distress. Everybody who has potentially diagnosable mental disorder does not receive treatment. Furthermore, pathways to mental health care may be systematically different for individuals with different social characteristics. Consequently, research based on treated rates of mental disorder seriously underestimates the true rate of mental illness in a population. (In 1994 the Institute of Medicine Committee on the Prevention of Mental Disorders estimated, for instance, that only 10–30 percent of those with a mental disorder receive any treatment.) Furthermore, such research may confuse the effects of variables such as social class, gender, place of residence, and age on psychiatric treatment with the impact of those same variables on the development of psychiatric impairment.

An alternative strategy for studying the epidemiology of mental disorder is the community survey. Early studies such as the Midtown Manhattan study (Srole et al. 1962) used symptom checklists with large random samples to estimate the amount of psychiatric impairment in the general population. Although such studies provided less biased estimates of the prevalence of psychological distress than did research on patient populations, they were subject to a different set of criticisms. The most serious limitation of the early community studies was that they used impairment scales that measured global mental health. Not only did the scales fail to distinguish different types of disorders, they confounded symptoms of physical and psychological disorder (Crandall and Dohrenwend 1967), measured relatively minor forms of psychiatric impairment, and frequently failed to identify the most serious forms of mental illness (Dohrenwend and Crandall 1970). Since it was not clear what relationship psychological symptom scales bore to cases of actual psychiatric disorder, it was also not clear how results from those studies contributed to an understanding of the social causes of mental illness.

Since the early 1980s, symptom scales that measure specific forms of impairment have largely replaced the early global scales with the consequence that the reliability and validity of community survey research has been vastly improved. The CES-D, for instance, is a twenty-item depression scale that can accurately distinguish clinical from normal populations and depression from other psychiatric diagnoses (Weissman et al. 1977). Consistently with the overall improvement in diagnostic reliability that has accompanied the development of DSM-III and DSM-IV, methods have been developed that provide reliable psychiatric diagnoses of many disorders among community residents. The most widely used diagnostic instrument of this sort was developed by a team of researchers at Washington University as part of the National Institute of Mental Health (NIMH) Division of Biometry and Epidemiology's Catchment Area Program. Called the Diagnostic Interview Schedule (DIS), the instrument can be administered by nonpsychiatrists doing interviews with the general population. Using DSM-IV criteria, it provides both current and lifetime diagnoses for many adult psychiatric disorders (Eaton et al. 1985). Enormous amounts of time and money have been devoted to the development of the DIS, and research that makes use of it promises to provide a vital link between studies of clinical and community populations. Nevertheless, even instruments like the DIS have shortcomings.

Mirowsky and Ross (1989) have challenged the DIS and the DSM upon which it is based on the grounds that psychiatric diagnosis is a weak form of measurement and that it is of questionable validity. These authors claim that psychiatric disorders are dimensional, not categoric. By collapsing a pattern of symptoms into a single diagnostic case, valuable information is lost about the nature of the disorder. As a result, the causes of mental, emotional, and behavioral problems are obscured. Mirowsky and Ross go on to suggest that the reliance on diagnosis does not give a true, that is, a valid, picture of psychiatric distress. Instead, psychiatrists use diagnosis because it allows them to receive payment from insurance companies who will pay only for cases and because it establishes mental distress as a problem that can be treated only by a physician. Although their criticisms are harsh, these authors reestablish the important distinction between the social construction of psychiatric diagnoses and the social causes of psychological pain. It is the latter issue, however, that most sociological research has addressed.

THE EPIDEMIOLOGY OF MENTAL DISORDER

Socioeconomic Status. The inverse relationship between socioeconomic status and mental disorder is now so well established that it has almost acquired the status of a sociological law. The relationship is surprisingly robust; it holds for most forms of mental disorder, no matter how socioeconomic status is measured, and for both patient populations and community samples. The relationship is strongest and most consistent for schizophrenia, personality disorders, and medically based syndromes. Findings for the major affective disorders are somewhat less consistent. Studies tend to report weak to moderate inverse relationships between social class and the incidence of major affective disorders such as anxiety and depression (Kessler et al. 1994). However, studies sometimes report no relationship (Weissman et al. 1991) or a positive class gradient (Weissman and Myers 1978). Evidence on the class distribution of minor depression is more clear-cut, with studies almost universally showing higher levels of depressive symptomatology among the lower strata. Similarly, research consistently shows that the highest levels of general distress are also found among those with the lowest income, education, or occupational status.

There are two general qualifications to the pattern outlined above. First, even though socioeconomic status is negatively associated with most types of mental disorder, the relationship is probably not linear. Extremely high rates of disorder are typically found in the lowest stratum. Higher strata do have progressively lower rates, but variation is considerably less between them than between the lowest and next-to-lowest tier. Some scholars have claimed, therefore, that serious mental illness is primarily an underclass phenomenon. Second, the inverse relationship between social class and mental disorder may be stronger in urban than in rural areas and is probably stronger in the United States than in other societies. (For a comprehensive review of this literature, see Ortega and Corzine 1990.)

Most, if not all, of the major sociological theories of mental illness begin with the empirical observation that psychological disorder is most prevalent among those individuals with the fewest resources and the least social power. Until recently, researchers focused almost exclusively on one dimension of inequality—social class. Indeed, the dominant paradigms in the sociology of mental health have derived primarily from the attempt to explain this relationship; hypotheses regarding the effects of gender, age, or marital status on psychological distress are often simple elaborations of models derived from the study of social class and mental disorder. Three general models of the relationship between social resources and mental illness have been suggested. These are (1) the "social causation" hypothesis; (2) the "social selection" or "drift" hypothesis; and (3) the "labeling" or "societal reaction" approach.

Social Causation. Social causation is a general term used to encompass a number of specific theories about the class-linked causes of mental disorder. Perhaps the most common version of social causation explains the higher rates of mental disorder among the lowest socioeconomic strata in terms of greater exposure to stress. According to this perspective, members of the lower class experience more stressful life events and more chronic strains (Turner et al. 1995). In addition, they are more likely to experience physical hazards in the environment, blocked aspirations, and status frustration (Cockerham 1996). Taken together, these stresses produce elevated rates of psychiatric impairment. In another version of social causation, scholars have argued that class differences in coping resources and coping styles are at least as important in the etiology of mental disorder as are class differences in exposure to stress (Pearlin and Schooler 1978). In this view, poverty increases the likelihood of mental illness because it (1) disrupts precisely those social networks that might effectively buffer the effects of stressful events and (2) inhibits the development of an active, flexible approach to dealing with problems. For both social and psychological reasons, then, the lower classes make use of less effective coping strategies. Finally, part of the class difference in mental disorder, especially rates of treated disorder, may stem from class differences in attitudes toward mental illness and psychiatric care. Because of more negative attitudes toward mental illness and because of inadequate access to appropriate psychiatric care, the lower classes may be more seriously ill when they first come in contact with the mental health care system, and thus they may be more likely to be hospitalized (Rushing and Ortega 1979).

Social Selection and Drift. This perspective implies that, rather than causing mental disorder, low socioeconomic status is a result of psychological impairment. Two mobility processes can be involved. According to the drift hypothesis, the onset of mental disorder adversely affects an individual's ability to hold a job and generate income. As a result of psychological disorder, then, individuals experience downward intragenerational mobility and physical relocation to less socially desirable neighborhoods (Eaton 1980). Social selection, on the other hand, occurs when premorbid characteristics of the mentally disordered individual prevent him or her attaining as high a social status as would be expected of similar individuals in the general population. Here, the focus is on intergenerational mobility (Kendler et al. 1995).

Labeling or Societal Reaction. Based on the work of Thomas Scheff (1966), this approach holds that much of the class difference in mental disorder stems not from any real difference in mental illness but rather from a tendency to diagnose or label a disproportionate number of lower-class individuals as psychologically impaired. According to Scheff, the process works as follows. Psychiatric symptoms have many different causes and many people experience them. Only a few individuals, however, are ever labeled as mentally ill. People who are so labeled are drawn from the ranks of those least able to resist the imputation of deviance. Once an individual is identified as mentally ill, a number of forces work to reinforce and solidify a mentally ill self-identity. Once labeled, individuals are encouraged by family and mental health professionals to acknowledge their illness. They are rewarded for behaving as "good" patients should, a task made easier by virtue of the fact that individuals learn the stereotypes of mental illness in early childhood. When individuals are discharged from the mental hospital, or when they otherwise terminate treatment, they may be rejected by others. This rejection has psychological consequences that simply reinforce a mentally ill identity. The process is self-fulfilling, leading Scheff to conclude that attachment of the mentally ill label is the single most important factor in the development of chronic mental disorder.

An Assessment. After two decades or more of acrimonious debate, the search for unitary explanations of the relationship between mental disorder and social class has largely been abandoned. Sociologists seldom claim that mental illness is derived only from medical factors, is caused only by features of the social environment, or stems purely from societal reaction. Most scholars now believe that different types of disorders require different types of explanations. Genetic and other biomedical factors are clearly involved in schizophrenia, and certain forms of depression. However, genetics, brain chemistry, and other medical factors do not provide the entire answer since, even among identical twins, concordance rates for mental illness fall only in the range of 30 to 50 percent. Thus, the causes of mental disorder must also be sought in the social environment. Research does suggest a modest relationship between the social stressors attendant to lower-class status and the onset of some forms of mental disorder. The evidence is clearest, however, for anxiety, substance abuse, and relatively minor forms of depression or psychological distress. For the more severe forms of mental illness and for conduct disorders, the drift and selection hypothesis appears to have the most empirical support. (See Miech et al. 1999 for a comprehensive review and data bearing on these points.) Although labeling is not the only cause of chronic mental illness, it is clear that the mental illness label does have negative consequences. In what has come to be called modified labeling theory, researchers have demonstrated that the status of ex-mental patient and the discrimination that follows from it, coupled with the ex-patient's expectation of rejection by others, adversely affects earnings, work status, and subsequent mental health (Link 1987; Markowitz 1998). Thus, labeling is one of the processes through which drift occurs. As researchers continue to refine the definitions and measurement of various mental disorders and as they more clearly delineate the processes of social causation, drift, and labeling, it is likely that further theoretical convergences will be identified.

Gender. It is not yet clear whether there are significant gender differences in overall rates of mental illness. There is little doubt, however, that certain types of disorders occur more frequently among women than among men. Research clearly shows that women are more likely to suffer from major and minor depression and anxiety than are men (Kessler et al. 1994); men, however, are usually found to have higher rates of antisocial personality disorders and the various forms of substance abuse and dependence (Aneshensel et al. 1991). The sex ratio for some forms of mental disorder may be age dependent; males have higher rates of schizophrenia prior to adolescence and females have higher rates in later adulthood (Loring and Powell 1988). Studies also find that male-female differences in levels of depression are most pronounced among young adults (Dean and Ensel 1983). Furthermore, gender effects appear to interact with those of marital, occupational, and parental roles.

Scholars continue to disagree about the precise form of the interaction effects of gender and marital status on mental illness. Virtually all studies report that gender differences are most pronounced among married persons; married women consistently show higher levels of depression and anxiety than married men. Evidence on the unmarried, however, is mixed. Research based on treated populations often finds higher rates of disorder among single men. Studies based on community samples more frequently report higher distress levels among unmarried women. The interaction between gender and marital status is further complicated by the presence of children, work outside the home, or both.

Some research on married persons finds that gender differences are reduced when both husbands and wives are employed. Studies comparing groups of women often find that employment has modest, positive effects on mental health. However, other studies report no difference between employed women and housewives (Carr 1997) and a few report that married, full-time homemakers with children have fewer worries and more life satisfaction (Veroff et al. 1981). These apparently contradictory findings stem, in part, from the different measures of mental health and illness used. It is possible, for instance, that small children can simultaneously increase their mothers' life satisfaction and their overall levels of anxiety and distress. However, two substantive factors also appear to be involved. First, it is the demands created by children and employment, rather than by parental or employment status per se, that cause elevated levels of distress among married women (Rosenfeld 1989). The level of demands varies, of course, depending upon the level of male responsibility for child care and housework. Second, employment decreases gender differences in distress only when it is consistent with both the husband's and the wife's desires. Married men's distress levels may, in fact, surpass married women's when wives work but their husbands prefer them not to (Ross et al. 1983).

As is true for social class, explanations for gender differences in mental disorder fall into three broad classes: social causation, social selection, and labeling. Because of the consistency of gender effects (at least for depression and anxiety) and improvements in the reliability of psychiatric diagnoses, most recent work has focused on the ways in which the social and psychological correlates of male and female roles cause variation in rates of mental disorder. Some have argued that differences in sex-role socialization make females more likely to direct frustration inward, toward themselves, rather than towards others, as males might. Thus, women are more likely to develop intropunitive disorders, whereas men are more likely to behave in antisocial ways (Loring and Powell 1988). Others have argued that women are more attached to others and are more sensitive to others' needs than are men. As a result, not only is women's mental health influenced by their own experiences but, also in contrast to men, they are more psychologically vulnerable to the stresses or losses of loved ones (Kessler and McLeod 1984). Although the empirical literature is far from clear on this point (cf. Umberson et al. 1996), according to this perspective, women experience more stressful events and are more psychologically reactive to them than are men. Other explanations—for both direct and interactive effects—of gender on mental illness have focused on male-female differences in power, resources, demands, and personal control. Insofar as employment increases women's power and resources, it is likely to have positive effects on mental health. Well-educated employed women have fewer mental symptoms than nonworking women; among working-class and lower-class women, however, employment may actually increase anxiety and depression because it elevates demands at the same time that it produces only marginal increases in resources (Sales and Hanson Frieze 1984). Since employed women generally retain full responsibility for children, the demands of caring for children, particularly those under the age of 6, exacerbate work-related stress. Thus, male-female differences in power and resources produce differences in ability to control demands. Gender differences in control, in turn, shape perceptions of personal mastery; personal mastery is the psychological mechanism that connects gender differences in resources and demands to gender differences in mental illness (Rosenfield 1989).

The social selection perspective is valid only for explaining male-female differences in the relationship between marital status and mental disorder. The argument is that mental illness is more likely to select men out of marriage than women. (See Rushing 1979 for a related discussion.) According to this perspective, male forms of mental disorder—psychosis and antisocial personality, for example—prevent impaired men from satisfactorily discharging the traditional male obligation to be good economic providers, making them ineligible as marriage partners. In contrast, female forms of psychiatric impairment may go undetected for long periods of time and may not seriously interfere with a woman's ability to fulfill the traditional housekeeping role. Thus, the higher rates of female disorder among the married may be a partial artifact of the differing probabilities of marriage for mentally disordered men and women.

The labeling explanation for male-female differences in psychiatric impairment begins by challenging the notion that women actually experience more symptoms and disorders than men do. Labeling theorists argue that women are overdiagnosed and overmedicated because of biases on the part of predominantly male psychiatrists and because of the male biases inherent in psychiatric nomenclature. Coupled with the greater willingness of females to admit their problems and to seek help for them, these biases simply produce the illusion that women are more likely to be disordered than men. Scholars using the labeling–societal reaction–critical perspective argue that the effects of gender biases are not benign and that they have consequences at two levels. First, individual women are unlikely to receive appropriate services for their real mental health problems. Second, and at a societal level, critics argue that psychiatry simply legitimates traditional gender roles, thereby buttressing the status quo (Chesler 1973).

A Theoretical Assessment. With the development of DSM-IV and with increases in the number of female mental health professionals, concern over the issues raised by labeling theorists has diminished somewhat. Trusting that the most blatant instances of sexism have been eliminated, researchers have turned their attention toward specifying the social psychological dynamics of the gender–mental health equation; considerable progress has been made in elucidating the circumstances under which women are most likely to experience symptoms of mental disorder. Nevertheless, it may be premature to close the question of gender bias in psychiatric disorders. In one study, male clinicians appeared to overestimate the prevalence of depressive disorders among women, a tendency that is certainly consistent with gender stereotypes. In the same study, black males were most likely to be diagnosed as paranoid schizophrenics, a view consistent with both gender and racial stereotypes (Loring and Powell 1988). In yet another study, male and female psychiatrists made similar diagnoses of male and female patients presenting severe Axis I conditions but made significantly different diagnoses for male and female patients with Axis II conditions, such as personality disorders (Dixon et al. 1995). Thus, advances of DSM-III (and IV) notwithstanding, the authors of these studies conclude that sex and race of client and psychiatrist continue to influence diagnosis even when psychiatric criteria appear to be clear-cut.

Age. Among adults, and with the exception of some types of dementia and other syndromes due to general medical conditions, rates of mental illness decrease with age. Rates of schizophrenia, manic disorder, drug addiction, and antisocial personality all peak between the ages of 25 and 44 (Robins et al. 1984). Furthermore, an older person with a serious mental disorder is likely to have had a first psychiatric episode in young or middle adulthood. At least 90 percent of older schizophrenics experienced the onset of the disorder in earlier life. Similarly, about two-thirds of older alcoholics have a long history of alcohol abuse or dependence (Hinrichsen 1990). Depression is the disorder most likely to occur among the elderly, and a substantial proportion of older community residents do report some of its symptoms. In general, the relationship between age and depression appears to be curvilinear, with depression lowest among the middle aged, higher among younger and older adults, and highest among the oldest (Mirowsky and Ross 1992). Nevertheless, relatively few of these older individuals meet criteria for clinical depression (Blazer et al. 1987), and rates of major depression are lower among older adults than in younger age groups. Some older persons, however, are more vulnerable to depression than others. As is true throughout the life cycle, women, individuals with health problems, the unmarried, and those with lower socioeconomic status are at greater risk of depression in late life than their peers. Estimating the true prevalence of depression among the elderly is especially problematic because its symptoms are frequently confused with Alzheimer's disease or other forms of dementia.

According to some estimates, two to four million older Americans suffer some form of mental disorder due to a general medical condition. Of these, roughly half are diagnosed with Alzheimer's, a disease that involves an irreversible, progressive deterioration of the brain. Approximately half of all nursing home residents are estimated to suffer from some form of dementia. Because there is no known treatment for most of these disorders, older mental patients receive little psychiatric care. Critics suggest, however, that many older persons are improperly diagnosed as having disorders of general medical origin. A sizable minority may actually be depressed; others may have treatable forms of dementia caused by medications, infection, metabolic disturbances, alcohol, or brain tumors. In some instances, then, the stereotype that senility is a concomitant of the aging process prevents appropriate diagnosis, intervention, and treatment.

Most explanations of the age–mental health relationship have focused on specific age groups. Clinicians suggest, for instance, that anxiety and depression in middle age are a consequence of hormonal change or of changes in family and occupational roles. The personality disorders of young adulthood are often explained in terms of the stresses produced by the transition from adolescence to full adult roles. Among the elderly, explanations have focused on either organic or environmental factors. The dementias have recognized organic causes. Although neither is a normal part of the aging process, the two major causes of these disorders are (1) the deterioration of the brain tissue that is associated with Alzheimer's disease and (2) cerebral arteriosclerosis. However, environmental factors also contribute to the onset of the dementias. They do so, in part, by increasing the likelihood of stroke or heart attack. In contrast, primary mental disorders, such as depression, personality disorders, and anxiety, depend more directly upon environmental factors. Some types of depression appear to have a genetic component, but the genetic link appears to be stronger in early- than in late-onset cases. Individuals who have their first episode of clinical depression prior to the age of 50, for instance, are more likely to have relatives with depression than those who become depressed in later years (Hinrichsen 1990). Consequently, losses typical of late life—losses of health, occupation, income, and loved ones—appear to be the primary causes of mental health problems among older adults.

Clearly, no single theory can adequately explain the etiology of mental disorder; at each stage of the life cycle, variables that are relevant to the onset of one type of disorder may be insignificant in the onset of other illnesses. Similarly, no single variable or set of variables is likely to explain age differences in overall rates of mental disorder. Nevertheless, efforts are under way to systematically explain the inverse relationship between age and primary psychiatric impairment. Gove and his associates have suggested that psychological distress decreases with age because individuals are able, over time, to find and settle into an appropriate social niche; as individuals move through life, they become less emotional and less self-absorbed, function more effectively in their selected roles, and generally become more content with themselves and with others. As a result, rates of mental disorder decrease from late adolescence through late life (Gove 1985; Gove et al. 1989).

Place of Residence. Sociologists have commonly assumed that rates of mental disorder are higher in urban than in rural areas. However, this assumption is based more on the antiurban bias of much sociological theory than it is on empirical research. In a thoughtful and systematic review, Wagenfeld (1990) has argued that there is little evidence in the mental health literature to suggest the superiority of rural life. In several of the rural community studies Wagenfeld cites, researchers report a "probable" case rate of depression and anxiety of 12 to 20 percent. Studies that explicitly compare rural and urban communities generally find that rates of psychosis are higher in rural communities and that rates of depression are somewhat higher in urban areas. Residents of metropolitan communities also appear more likely to have multiple diagnoses than do rural residents, leading Kessler et al. (1994) to conclude that urban-rural differences in the prevalence of mental disorder probably reflect differences in comorbidity rather than differences in rates of individuals having a psychiatric condition. Differences in case definition and diagnosis, differences in how "rural place of residence" is defined and measured, and differences in the time period during which studies were conducted make it difficult, overall, to assess whether rural communities have significantly higher overall rates of pathology than urban areas. Results are sufficient, however, to suggest that rural life is not as blissful as it is often claimed to be. Recent declines in the rural economy, the out-migration of the young and upwardly mobile, and the relative paucity of mental health services are likely to be major contributing factors in the etiology of rural mental health problems.

Other factors. Epidemiologists have also explored the relationships between the incidence or prevalence of mental disorder and such variables as race and ethnicity, migration, social mobility, and marital status. In each case, results generally support the view that individuals with the fewest resources—both economic and social—are most likely to experience psychiatric impairment. However, most research has adopted a rather static view; few studies have assessed the extent to which relationships between each of these variables and mental disorder have changed over time. Given the significant changes in diagnostic practices and in the mental health professions over the last decades, this is a striking omission.

AN AGENDA FOR FUTURE RESEARCH

Since the early 1960s, psychiatric sociology has undergone enormous changes. During the 1960s and 1970s, much of the literature was sharply critical of psychiatry and of medical models of madness. Although sociologists were divided about the relative importance of labeling processes in the etiology of mental illness, most agreed that psychiatric diagnoses were unreliable and were influenced by social status and social resources, that long-term institutionalization had detrimental effects, and that at least some patients were hospitalized inappropriately. Such criticisms provided one impetus for the substantial change that took place in psychiatric care during the same period; laws were changed to make involuntary commitment more difficult; steps were taken to deinstitutionalize many mental patients; and a major effort was made to improve the reliability of mental diagnoses. By the time DSM-III was published in 1980, the most flagrant abuses and the sharpest criticism of psychiatry seemed to have disappeared. Consequently, many sociologists shifted their attention from concerns about the lives of people with serious mental disorder to the social correlates of psychological distress among the general population (Cook and Wright 1995). Using what is basically a medical model of impairment, researchers have focused on delineating the relationship between social variables (such as gender, age, race, social class, place of residence, life events, and stress) and specific diagnoses (most often depressive symptoms, anxiety, and substance abuse). Indeed, the psychiatric view of mental disorder is so well established in sociology that the growing literature on homelessness has generally accepted the assertion of mental health professionals that most of the homeless are simply individuals who have fallen through the cracks of the mental health care system. (For notable exceptions, see Bogard et al. 1999; Snow, Baker, and Anderson 1986.) It is surprising that sociologists have been so uncritical in their acceptance of this position; it is also surprising that in the decade of the 1990s, declared by the National Institute of Mental Health to be the "Decade of the Brain," they have been so ready to accept the view that mental illness is primarily a problem of genetics or brain chemistry and that it can be treated just like any other disease. It is certainly true that enormous strides have been made in the diagnosis and psychopharmacological treatment of mental disorder. It is also certainly true that biomedical factors are causally involved in some types of mental illness. Sociologists must, therefore, continue their efforts to develop a model of mental disorder that integrates medical, psychological, and social factors.

As some critics point out, however, the current emphasis on diagnoses, cases, and the medical model of mental illness has limitations. Acceptance of the psychiatric view of mental disorder leads to the acceptance of policy recommendations that are not yet firmly grounded in empirical research. It is far from clear, for instance, that deinstitutionalization of the mentally ill is the primary cause of homelessness in America. As Bogard et al. (1999) point out, conventional wisdom notwithstanding, very few homeless mothers are mentally ill; it can be reasonably argued, then, that the enormous resources that have been directed toward providing them with mental health care might more appropriately and effectively be used to provide safe, affordable housing. In a similar vein, Link and Phelan (1995) note that current attention to individual risk factors in disease gives rise to "personal policy" recommendations that leave totally unaddressed the fundamental social conditions that cause differential exposure to risk.

Furthermore, few studies have assessed the extent to which changes in psychiatric diagnosis or changes in the civil rights guarantees of mental patients have affected the delivery and quality of mental health services. Consumers and families have voiced concern that the powerful new psychopharmacological drugs are being inappropriately used as forms of social control and chemical restraint at the same time that research continues to show that it is racial and ethnic minority consumers who are most likely to be so restrained (Cook and Wright 1995). Aside from the field trials used in their formulation, few studies have assessed the reliability and validity of DSM-IV diagnoses. However, results from several studies show that nonclinical factors such as gender, race, the availability of viable community housing and the presence of reliable caretakers significantly affect not only diagnosis but treatment protocols and outcomes. (See Cook and Wright [1995] for a review of these studies and these concerns.) It is far from clear, then, that lower-class women are any more likely to receive appropriate care in 1999 than they were in 1950 or 1970. It is unclear whether urban-rural differences in rates of mental disorder have changed over time and, if so, to what extent changes in diagnostic systems or service availability are implicated. Evidence that rural residents may actually experience mental illness at approximately the same rates as urban residents coupled with an acute shortage of rural mental health providers suggest the importance of understanding the diagnostic practices of primary-care physicians and of providing appropriate training to them.

Research in the next century must adopt a more dynamic or process view of mental health issues. The consequences of changes in psychiatric diagnosis, of the increased reliance on drug therapies, of changes in mental health law and policy, and in the availability of mental health services must be assessed. Changes in the mental health system must be linked to changes in the composition of the pool of "potential clients" and to issues regarding the development of gender, age, class, and culturally appropriate systems of care.

references

American Psychiatric Association 1980 Diagnostic and Statistical Manual of Mental Disorders III. Washington: APA.

——1994 Diagnostic and Statistical Manual of Mental Disorders IV. Washington: APA.

Aneshensel, Carol S., Carolyn M. Rutter, and Peter A. Lachenbruch 1991 "Social Structure, Stress, and Mental Health." American Sociological Review 56:166–178.

Blazer, Dan, Dana C, Hughes, and Linda K. George 1987 "The Epidemiology of Depression in an Elderly Community Population." Gerontologist 27:281–287.

Bogard, Cynthia J., J. Jeff McConnell, Naomi Gerstel, and Michael Schwartz 1999 "Homeless Mothers and Depression: Misdirected Policy." Journal of Health and Social Behavior 40:46–62.

Carr, Deborah 1997 "The Fulfillment of Career Dreams at Midlife: Does It Matter for Women's Mental Health?" Journal of Health and Social Behavior 38:331–344.

Chesler, Phyllis 1973 Women and Madness. New York: Avon.

Cockerham, William 1996 Sociology of Mental Disorder, 4th ed. Upper Saddle River, N.J.: Prentice Hall.

Cook, Judith A., and Eric R. Wright 1995 "Medical Sociology and the Study of Severe Mental Illness: Reflections on Past Accomplishments and Directions for Future Research." Journal of Health and Social Behavior (Extra Issue):95–114.

Crandall, D. L., and B. P. Dohrenwend 1967 "Some Relations among Psychiatric Symptoms, Organic Illness, and Social Class." American Journal of Psychiatry 1527–1538.

Dean, Alfred, and Walter M. Ensel 1983 "Socially Structured Depression in Men and Women." In James Greenley, ed., Research in Community Mental Health, vol. 3. Greenwich, Conn.: JAI.

Dixon, Jo, C. Gordon, and T. Khomusi 1995 "Sexual Symmetry in Psychiatric Diagnosis." Social Problems 42:429–448.

Dohrenwend, Bruce P., and D. L. Crandall 1970 "Psychiatric Symptoms in Community, Clinic, and Mental Hospital Groups." American Journal of Psychiatry 126:1611–1621.

Eaton, William W. 1980 The Sociology of Mental Disorder. New York: Praeger.

——, M. M. Weissman, J. Anthony, L. Robins, D. Blazer, and M. Karno 1985 "Problems in the Definition and Measurement of Prevalence and Incidence of Psychiatric Disorder." In W. Eaton and R. Kessler, eds., Epidemiologic Field Methods in Psychiatry: The NIMH Epidemiologic Catchment Area Program. Orlando, Fla.: Academic.

Faris, Robert E., and H. Warren Dunham 1939 Mental Disorders in Urban Areas. Chicago: University of Chicago Press.

Figert, Anne 1994 Women and the Ownership of PMS: The Structuring of a Psychiatric Diagnosis. Hawthorne, N.Y.: Aldine de Gruyter.

Gove, Walter R. 1985 "The Effect of Age and Gender on Daviant Behavior: A Biopsychosocial Perspective." In Alice Rossi, ed., Gender and the Life Course. New York: Aldine de Gruyter.

——, Suzanne T. Ortega, and Carolyn Briggs Style 1989 "The Maturational and Role Perspectives on Aging and Self Through the Adult Years: An Empirical Evaluation." American Journal of Sociology 94:1117–1145.

Hinrichsen, Gregory A. 1990 Mental Health Problems andOlder Adults. Santa Barbara, Calif.: ABC-CLIO.

Hollingshead, A. B., and F. C. Redlich 1958 Social Classand Mental Illness. New York: John Wiley.

Institute of Medicine, Committee on Prevention of Mental Disorders, Division of Biobehavioral Sciences and Mental Disorders 1994 Reducing Risks for MentalDisorders: Frontiers for Preventive Intervention Research. Washington: National Academy Press.

Kendell, R. E. 1988 "What is a Case? Food for Thought for Epidemiologists." Archives of General Psychiatry 45:374–376.

Kendler, Kenneth S., Ellen E. Walters, Michael C. Neale, Ronald C. Kessler, Andrew C. Health, and Lindon J. Eaves 1995 "The Structure of the Genetic and Environmental Risk Factors for Six Major Psychiatric Disorders in Women: Phobia, Generalized Anxiety Disorder, Panic Disorder, Bulimia, Major Depression, and Alcoholism." Archives of General Psychiatry 52:374–383.

Kessler, Ronald C., Katherine A. McGonagle, Shanyang Zhao, Christopher B. Nelson, Michael Hughes, Suzanne Eshleman, Hans-Ulrich Wittchen, and Kenneth S, Kendler 1994 "Lifetime and 12-Month Prevalence of DSM-III-R Psychiatric Disorders in the United States: Results from the National Comorbidity Study." Archives of General Psychiatry 51:8–19.

——, and Jane McLeod 1984 "Sex Differences in Vulnerability to Undesirable Life Events." AmericanSociological Review 49:620–631.

Kirk, Stuart A., and Herb Kutchins 1992 The Selling ofDSM. Hawthorne, N.Y.: Aldine de Gruyter.

Link, Bruce 1987 "Understanding Labeling Effects in the Area of Mental Disorders: An Assessment of Expectations of Rejection." American Sociological Review 52:96–112.

——, and Jo Phelan 1995 "Social Conditions as Fundamental Causes of Disease." Journal of Health andSocial Behavior (Extra Issue):80–94.

Loring, Marti, and Brian Powell 1988 "Gender, Race, and DSM-III: A Study of the Objectivity of Psychiatric Behavior." Journal of Health and Social Behavior 29:1–22.

Markowitz, Fred E. 1998 "The Effects of Stigma on the Psychological Well-Being and Life Satisfaction of Persons with Mental Illness." Journal of Health andSocial Behavior 39:335–347.

Miech, Richard A., Avshalom Caspi, Terrie E. Moffitt, Bradley R. Entner Wright, and Phil A. Silva 1999 "Low Socioeconomic Status and Mental Disorders: A Longitudinal Study of Selection and Causation during Young Adulthood." American Journal of Sociology 104:1096–1131.

Mirowsky, John, and Catherine E. Ross 1992 "Age and Depression." Journal of Health and Social Behavior 33:187–205.

——, and Catherine E. Ross 1989 "Psychiatric Diagnosis as Reified Measurement." Journal of Health andSocial Behavior 30:11–25.

Ortega, Suzanne T., and Jay Corzine 1990 "Socioeconomic Status and Mental Disorders." In James Greenley, ed., Research in Community and Mental

Health, vol. 6. Greenwich, Conn.: JAI.

Pearlin, Leonard, and Carmi Schooler 1978 "The Structure of Coping." Journal of Health and Social Behavior 19:2–21.

Robins, Lee N., J. E. Helzer, M. M. Weissman, H. Orvaschel, E. Gruenberg, J. D. Burke, Jr., and D. A. Regier 1984 "Lifetime Prevalence of Specific Disorders in Three Sites." Archives of General Psychiatry 41:949–958.

Rosenfield, Sarah 1989 "The Effects of Women's Employment: Personal Control and Sex Differences in Mental Health." Journal of Health and Social Behavior 30:77–91.

Ross, Catherine E., John Mirowsky, and Joan Huber 1983 "Dividing Work, Sharing Work, and In-Between: Marriage Patterns and Depression." AmericanSociological Review 48:809–823.

Rushing, William A. 1979 "The Functional Importance of Sex Roles and Sex-Related Behavior in Societal Reactions to Residual Deviants." Journal of Healthand Social Behavior 20:208–217.

——, and Suzanne T. Ortega 1979 "Socioeconomic Status and Mental Disorder: New Evidence and a Sociomedical Formulation." American Journal of Sociology 84:1175–1200.

Sales, Esther, and Irene Hanson Frieze 1984 "Women and Work: Implications for Mental Health." In L. E. Walker, ed., Women and Mental Health Policy. Beverly Hills, Calif.: Sage.

Scheff, Thomas J. 1966 Being Mentally Ill: A SociologicalTheory. Chicago: Aldine de Gruyter.

Scott, Wilbur J. 1990 "PTSD in DSM-III: A Case in the Politics of Diagnosis and Disease." Social Problems 37:294–310.

Snow, David, Susan Baker, and Leon Anderson 1986 "The Myth of Pervasive Mental Illness among the Homeless." Social Problems 33:407–423.

Srole, L., T. S. Langner, S. T. Michel, M. D. Opler, and T. C. Rennie 1962 Mental Health in the Metropolis: TheMidtown Manhattan Study. New York: McGraw-Hill.

Tavris, Carol 1993 "Do You Menstruate? If So, Psychiatrists Think You May Be Nuts." Glamour (November):172.

Townsend, John M. 1980 "Psychiatry versus Societal Reaction: A Critical Analysis." Journal of Health andSocial Behavior 21:268–278.

Turner, R. Jay, Blair Wheaton, and Donald A. Lloyd 1995 "The Epidemiology of Social Stress." AmericanSociological Review 60:104–125.

Umberson, Debra, Meichu D. Chen, James S. House, Kristine Hopkins, and Ellen Slaten 1996 "The Effect of Social Relationships on Psychological Well-Being: Are Men and Women Really So Different?" AmericanSociological Review 61:837–857.

Veroff, Joseph, Elizabeth Douvan, and Richard A. Kulka 1981 The Inner American. New York: Basic.

Wagenfeld, Morton O. 1990 "Mental Health and Rural America: A Decade Review." Journal of Rural Health 6:507–522.

Weissman, Myrna M., Martha L. Bruce, Phillip J. Leaf, Louis P. Florio, and Charles E. Holzer 1991 "Affective Disorders." In Lee N. Robins and Darrel A. Regier, eds., Psychiatric Disorders in America: TheEpidemiological Catchment Area Study. New York: Free Press.

——, and Jerome K. Myers 1978 "Affective Disorders in a U.S. Urban Community." Archives of GeneralPsychiatry 35:1304–1311.

——, D. Scholomskas, M. Pottenger, B. Prusoff, and B. Locke 1977 "Assessing Depressive Symptoms in Five Psychiatric Populations: A Validation Study." American Journal of Epidemiology 106:203–214.

Suzanne T. Ortega
Sharon L. Larson