NATILONAL INSTITUTE OF MENTAL HEALTH- - Stressful Life: Event Theory and Research: .Im pzlizcatioins; for; anary Prevention us DEPARTMENT OF HEALTH AND HUMAN SERVICES h ‘ > _ Public Hiea; Serwce Alsatian D’F‘ugyAbUSeu and, Mental Health: Administration Stressful Life Event Theory and Research: Implications for Primary Prevention by Bernard L. Bloom, Ph.D. Department of Psychology University of Colorado Boulder, Colorado US. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Alcohol, Drug Abuse, and Mental Health Administration National Institute of Mental Health 5600 Fishers Lane Rockville, Maryland 20857 This monograph is the ninth issue in the Prevention Publication Series. The purpose of this series is to disseminate and exchange informtion on prevention activities. especially primary prevention. in the mental health field; stimulate development of prevention projects in mental health. public health. and other human service facilities; same the raining of mental health workers in prevention; and promote prevemion research. This series includes scientific monographs. cmference proceedinas. com— missioned papers. and other materials that meet a need for information about prevention. Publications in this series will be issued as materials and manuscripts are developed. The views expressed in these publications are those of the author-(s) and do not necessarily reflect the official posi— tion of the National lmtitute of Mental Health of any other part of the [1.5. Department of Health and Human Services. Stephen E. Goldston, Bd.D., M.S.P.H. Director. Office of Prevention National Institute of Mental Health and Editor, Prevention Publication Series /, Previous Prevention Publication Series issues: (1) Primary Prevention: An idea Whose Time Has Come (2) Mutual Help Groups: A Guide for Mental Health Workers (3) Preventive intervention in Schizophrenia: Are We Ready? (4) New Directions in Prevention Among American indian and Alaska Native Communities (5) Preventing the Harmful Consequences of Severe and Persistent Loneliness (6) A Guide to Evaluating Prevention Programs in Mental Health (7) Preventing Stress-Related Psychiatric Disorders (8) Psychiatric Epidemiology and Primary Prevention: The Possibilities This document was prepared under contract number 81MO49268201D from the National institute of Mental Health. Dr. Stephen E. Goldston. Direc— tor. Office of Prevention. served as the NIMH Project Officer. The figure appearing on page 8 is copyrighted and is reproduced herein with permission of the copyright holder. Further reproduction of this copyrighted material is prohibited without specific permission of the copyright holder. All other material contained in this volume except quoted passages from copyrighted moss is in the public domain andmay be used or reproduced without permission from the lmtitute or the authors. Citation of the source is appreciated. DHHS Publication No. (ADM)8$—1385 Printed 1985 For sale by the Superintendent of Documents, US, Government Printing Office Washington, DC. 20402 FOREWORD Stressful life events and their effects on well—being have emerged over the past two decades as fruitful lines of inquiry. Al— though what we now call stress has long been recognized as a pre— cipitant of mental disorders, research into stressful life events has developed into a field of study only recently. Knowledge has ex— panded to the point where it can significantly improve intervention strategies that may prevent mental and physical illness. Thus. a review of the theory and research of stressful life events is timely. While this body of literature has already proven of value, Dr. Bloom highlights the questions that still need answering: How should the magnitude of a stressful event be assessed? What cir— cumstances increase the risk of a deterioration in health? What are the mechanisms by which external stress becomes internalized? Finding answers to these questions poses musually complex meth— odological demands. This monograph reviews both the substantive and methodological literature in the field of stressful life events. A major portion of it describes some of the prevention programs that have grown out of research during the past decade. Within each of its themes. the monograph distinguishes between experimental programs that are soundly described and evaluated and the larger number that have not yet been fully assessed. The monograph emphasizes the need for prospective studies. The preventive intervention programs modeled on the stressful life event paradigm extend across the entire developmental life span and deal with a very wide variety of life events. Indeed, one of the most remarkable aspects of this emerging field is the inventive— ness of practitioners interested in prevention program development. Many intervention programs are directed at reducing the stress accompanying treatment for physical illness. These programs are among the most carefully evaluated. and they have demonstrated remarkable success. Thus. the monograph can be useful to practitioners as well as researchers. Its critical approach has the potential for significantly improving both the quality of research and the effectiveness of intervention programs in the entire mental health field. Stephen E. Goldston. Ed.D., M.S.P.H. Director. Office of Prevention National Institute of Mental Health iii PREFACE This monograph examines stressful life events, those contem— porary experiences that may require very high levels of personal adaption. Understanding how stressful life events may affect well—being is particularly important to social scientists and mental health professionals who are interested in the prevention of psy— chological disorders. This importance is due to accumulating evi— dence that stressful life events can precipitate a wide variety of illnesses in people who are vulnerable but otherwise healthy, that is. in people who are unable to adapt successfully. In the past decade, there has been an enormous expansion in the number of prevention—oriented intervention programs that have been based on the stressful life event paradigm; many of these programs are reviewed in this monograph. The theoretical. empirical, and programmatic literature on stressful life events is so voluminous. however, that it has not been possible to include all or even most of it in this review. I can only hope that what is in— cluded is sufficiently representative so that the issues that must be considered in examining the field of stressful life event research have been identified and usefully discussed. Some of the material in this monograph has appeared in the second edition of Community Mental Health: A General Intro— duction. I appreciate the willingness of Brooks/Cole Publishing Company to allow that material to be used in this context. While I am. of course. responsible for all of the decisions that went into the final version of this monograph, I want to acknowledge the contributions of Elyse Morgan who helped prepare the summaries of the preventive intervention programs that are described. and Gerald Caplan and Herbert C. Schulberg who read the entire document critically prior to its final revisions and publication. Bernard L. Bloom. Ph.D. CONTENTS Foreword Preface Chapter 1. The Conceptualization of Stressful Life Events—— The Research Paradigm The Study of Stressful Life Events ————————————————————— Conceptualizing the Role of Stressful Life Events ——————— Sm'mnary Chapter 2. The Consequences of Stressful Life Events—"— Stressful Life Events and Psychopathology——____________ Stressful Life Events and Physical Illness ——————————— The Effects of Specific Stressful Life Events____.___-____ Summary Chapter 3. Moderating Factors in Reactions to Stressful Life Events Personal Resources Social Resources Summary Chapter 4. Methodological Issues in the Study of Stressful Life Events Stressful Life ENents in Children Desirable and Undesirable Stressful Life Events ————————— Subjective Judgments of Stress The Need for Prospective Studies ———————————————————— Summary Page iii H HafiéN Chapter 5. Examples of Preventive Intervention Programs— 55 Stresses Associated With Parenthood——————-—- 56 Physical Illness in Children 63 Adolescent Role Transitions 66 Specific Adult Stressors 70 Bereavement 83 Community—Wide Stressful Life Events—————--———— 91 Summary 94 Chapter 6. Concluding Summary 96 References 100 Chapter 1 THE CONCEPTUALIZATION OF STRESSFUL LIFE EVENTS This monograph presents a review and analysis of recent re— search and conceptualization in the field of stressful life events and examines the implications of that body of scholarly work for developing primary prevention programs. ’i ‘7 Baumlet al. 981 have de stress as a " ocess in ' environmental events or orces, called stressorsLJtllmaI-enan‘gr— gWflgfl-fiem' gj'“(p.4; see also Hefferin 1980): Stress__hashalso been‘definediasnércgnditionjnnwhichlhere is a marked discEp’hth mad or anisim‘ and [the organism's capability to respond. The consequences of sue a discrepancy cme organism's future in re— spect to conditions essential to its well—being (McGrath 1970; Cap— lan 1981). Stressful life events. then. are those external events that make adaptive demands on a person. These demands may be suc— cessfully met or might inaugurate a process of internal psycho— logical or physiological straining that could culminate in some form of illness. Stressful life event research and its associated conceptualization and more formal theory building attempt to understand one of the important ways that psychological and sociocultural factors can play a role in the development of mental disorders. Identification of stressful life events in a community. combined with the devel— opment of appropriate intervention programs, may be one of the keys to the effective prevention of much emotional disorder. By definition, stressful life events are contemporary. 3.5. and B.P. Dohrenwend distinguish between stressful life events and personal dispositions. W life events are those that areiproximate to. rather than? remote from, the "on'set'of a disorder. WGW recent, death of a friend or relative but not the fact that an adult's father died When he'Tf‘she‘was‘adcmld. The latter event is not irrelevant to life stress but is subsumed under personal dispositions, since we assume that the early death of a parent can affect an adult's behavior only insofar as its \ ? impact was internalized (1981a, p. 131). / 1 I This monograph does not concern itself directly with what might be called chronic debilitating life circumstances. such as maternal deprivation. incurable illness. chronic parental mental illness, or poverty. These sources of stress. undoubtedly as important as con- temporary stressful life events, are represented by a large theo— retical and empirical literature, and the field is sufficiently im— portant to warrant a monograph of its own. Included among stressful life events but somewhat more removed in time from the more proximate events identified by the Dohren— wends are those episodes that produce what are called delayed stress responses. These syndromes, according to Horowitz and Solomon. ”often begin only after termination of real environmental stress events and after a latency period of apparent relief" (1978, p. 268). Interest in delayed stress reactions has intensified within the last several years as a consequence of the growing realization that Vietnam veterans appear to be surprisingly vulnerable to such reactions (Figley 1978; Williams 1980). The Research Paradigm Inquiry into the role of stressful life events in human welfare is based upon a very different paradigm from the one that has been employed in the analysis of specific psychiatric diseases. This paradigm does not begin with the assumption that every specific disorder has a single or even a multiple necessary precondition. Rather, this paradigm is based upon the clearly established asso— ciation of stress with increased risk of illness and assumes that we are all more or less vulnerable to stressful life experiences and that "almost any disease or disability may be associated with these events" (Dohrenwend and Dohrenwend 1974, p. 314). Regarding stressful life events. Caplan (1981) has commented, Years ago we used to think that particular sets of such events in association with certain personality patterns would cause specific bodily or mental illnesses. Nowadays many of us be— lieve that individuals exposed to such circumstances may suffer an increase in nonspecific vulnerability to a wide range of bodily and mental illnesses (p. 413). Cassel (1976) made a similar case when he wrote that "it is most unlikely that any given psychosocial process or stressor will be etiologically specific for any disease. at least as currently clas— sified" (p. 109; see also Cassel 1973, 1974a. 1974b). Cassel's own research, as well as his reviews of the studies of others, led him to conclude that. although psychosocial processes enhance suscepti— bility to disease, the clinical manifestations of this enhanced sus— ceptibility would not be a function of the particular psychosocial stressor. 0n the basis of the stress paradigm. preventive intervention programs can be organized for the purpose of reducing the inci— dence of particular stressful life events, whenever possible. or facilitating their mastery once they occur. In either case. one need not have undue regard for the prediction of the specific disorders that will be prevented. That is, this new paradigm begins by aban— doning at the outset the search for a unique cause, or set of causes, for each unique disorder. Rather, the paradigm far more commonly examines the precursors of what Frank (1973) has called demor— alization and Dohrenwend and his colleagues (1980) have more recently called nonspecific psychological distress. The new paradigm has the following sequence of steps: 1. Identify a stressful life event. or set of such events, that appears to have undesirable consequences. Develop proce— dures for reliably identifying persons who have undergone or who are undergoing those stressful experiences. 2. By traditional epidemiological and laboratory methods, study the consequences of those events and develop hypotheses related to how one might go about reducing or eliminating their negative consequences. 3. Mount and evaluate experimental preventive intervention pro- grams based on these hypotheses. This new stressful life event paradigm has turned our attention from longstanding predisposing factors in psychopathology to far more recent precipitating factors and is part of an even broader phenomenon. It has long been known that biological, psychological, and sociological factors differentially predispose persons to emo— tional disorders. With few exceptions, however, efforts to develop effective preventive services based on attempts to modify these distal predisposing factors have been unsuccessful. Eisenberg (1981) has recently commented. for example, that measurement of distant outcome places a terrible burden of proof on childhood interventions; they must be powerful indeed to be able to show a clear effect despite the vicissitudes of subsequent life experience (p. 4). Predisposing and precipitating factors are interdependent. how— ever. in that predisposing factors can serve as sources of vulner— ability. The presence or absence of predisposing factors may help identify groups of persons who are at particularly high risk of de- veloping illnesses in the face of stressful life events. Brown and Harris (1978), for example. have shown that, while contemporary stressful life events appear to have a precipitating effect upon depression in women, certain childhood characterisitics (e.g., loss 3 of the mother prior to age 11) seem to potentiate this reaction and make it more probable. There is every reason to believe that prevention programs linked to the more successful management of stressful life events can be effective. particularly when we set about to build on what is al— ready known about crisis theory and crisis intervention (Caplan 1964. pp. 34—54; Parad 1965; Parad et al. 1976; Mann 1978). Many stressful life events. such as school entrance, new parenting. sep— aration and divorce. retirement, and widowhood are common. many are becoming more common. and few sustained and comprehensive services exist within our communities to assist people in mastering any of them. The Study of Stressful life Events Early clinical interest in the role of stressful life events is as— sociated with the work of Cannon, Wolff , Meyer, and Selye. Cannon (1929) held that external stressors could invade the body through their effects upon emotions. Wolff (1950) believed that external stressors produced protective reactions that had both physiological and psychological components. Meyer (1951) taught that ordinary life events. if they were stressful, could play a role in the etiology of disease and urged that practitioners undertake systematic in— quiry into such life events as part of their clinical assessments. Selye (1946, 1956) introduced the concept of the general adaptation syndrome to characterize what he believed to be a complex but nonspecific bodily reaction to stress and was instrumental in the beginning investigation of the biochemical nature of those reac— tions. Hinkle (1974) has summarized the implications of stress— related clinical research as of the post—World War II period by noting that it seemed evident that there would probably be no aspect of human growth. development, or disease that would in theory be immune to the influence of the effect of a man's relation to his social and interpersonal environment (p. 10). An important early paper reviewing the evidence linking recent traumatic events to the occurrence of mental disorders was pre— pared by Reid (1961) as part of a conference held in 1959, spon— sored by the Milbank Memorial Fund. that sought to review the epidemiological knowledge base pertinent to mental disorders. While that paper never employed the phrase "stressful life event," clearly Reid was dealing with exactly that concept. The field of stressful life event research came into its own and was virtually transformed by the development of the Social Re— adjustment Rating Scale by Holmes and Rahe (1967; see also Holmes and Masuda 1974; Rahe 1979). Holmes had worked in 4 Wolff's laboratory and followed Meyer's admonition to search for stressful events in the recent life of patients. He and his colleagues were persuaded that stressful life events, by evoking psycho— physiologlcal reactions, could play a precipitating role in many diseases. Beginning in 1949. Holmes and his colleagues studied life stres- sors that were reported by more than 5,000 patients as occurring shortly before the onset of their illnesses. From this data pool, they assembled a list of 43 representative life events and scaled them in terms of stressfulness. In this process. they made two critical decisions. First. while most of the items in their list could be considered undesirable events. the concept of a stressful life event was not limited to such items but could include desirable events as well. Stressfulness was defined in terms of the need for readjustment by the person undergoing the event. and desirable events (an outstanding personal achievement. for example) were thought to require a measure of readjustment just as undesirable events (being fired. for example). Second, Holmes and his col— leagues believed that it was appropriate to develop a rating scale of stressfulness that would be derived from expert outside judg— ment rather than judgments made by the persons themselves. They sought to develop a scale in which each stressful event was given a single. universal weight. The methodological details of the devel— opment of the scale are described by Holmes and Masuda (1974). While research was continuing to explore the physiological con— comitants of stressful life events, there was a remarkably rapid growth in the study of stressful life events per se, stimulated in large measure by the availability of the Holmes and Rahe scale. That scale made it possible to assign weights to events. total these weights across a broad array of such events, and then examine the resulting total scores in terms of their relationships to subsequent illness. Research activity grew at such a pace that only 6 years after the publication of the Social Readjustment Rating Scale. Barbara and Bruce Dohrenwend organized an intemational con- ference to bring many of the prominent stressful life event re— searchers together. The report of that conference (Dohrenwend and Dohrenwend 1974) provides an eloquent review of the state of stressful life event research as of that time. The major thrust of that report for future research was to 1m— derline the complexity of the relationships between stressful life events and subsequent disorder. In particular, the report urged the development of more satisfactory measurement procedures and research designs, greater attention to conceptualization. and ex— amination of the factors that might serve to moderate or poten— tiate the effects of stressful life events. The field has been ex— tremely responsive to these recommendations. " Recent stressful life event literature (Aakster 1974; Rabldn and Struening 1976; Tennant and Andrews 1976; Gersten et al. 1977; Wildman and Johnson 1977; Dohrenwend 1978; Mueller et a1. 1978; S Syme and Torfs 1978; Zautra and Beier 1978; Antonovsky 1979; Barrett 1979; Blmn 1979; Dohrenwend 1979; Jenkins et a1. 1979; Crook and Eliot 1980; Dooley and Catalano 1980) suggests that the role of stressful life events in the precipitation of illnesses is not as simple as investigators had initially envisioned. These studies often, but not invariably, find that physical illnesses or psycho— logical disorders are preceded by an excess number and intensity of stressful life events, but not enough is known to predict which specific persons are at unusually high risk. As for research exploring the physiological consequences of stressful life events, that is, how stressful experiences become internalized, recent reviews of both animal and human research (Borysenko and Borysenko 1982; Locke 1982) suggest that certain forms of stress suppress cellular immune functions. Locke has noted that acute, short—term stresses have different biological effects from chronic long—term stresses. and has postulated that: The initial response to a state of mild stress is a transient activation to certain components of the immune system. When this state of arousal is prolonged, the transient enhancement may deplete the immune system. resulting in temporary. rel— ative immunodepression. There appears to be an interactive effect of the stressor magnitude and the individual‘s adaptive capacity on the immune response; exposure to a severe stressor coupled with adaptive failure due to ineffective cop— ing can lead to immunosuppressive changes, even during a short—term stress. Thus, it is not stress itself which is im- munosuppressive, but stress coupled with poor coping (1982. p. 56). Borysenko and Borysenko, after reviewing the animal research linking stress. behavior, and immunity. have come to conclusions similar to those of Locke and have suggested that multidisciplinary research in neurosciences. immunology. psy— chology. endocrinology, and psychiatry promise significant advances in our understanding of the effects of stress on im— munity and resistance to disease (1982. p. 65). Conceptualizing the Role of Stressful Life Events The pathway between a stressful life event and subsequent ill— ness is a complex one involving a number of important components. Rahe (1974) has proposed that the steps that must be taken into account in understanding how a stressful life event may precipitate an illness include: 1. Past experience, that is. how the person has traditionally 6 managed stressful life events so that they are more or less stressful than would generally be the case 2. Psychological defenses. that is, the abilities a person has to deal with stressful life events so that they have no negative consequences 3. Physiological reactions. that is. the nature of the physiologi— cal impact of those stressful life events that are not dealt with by successful psychological defenses 4. Coping, that is. the ability to attenuate or compensate for physiological reactions 5. Illness behavior, that is, how individuals come to interpret the remaining physiological reactions as symptoms or as illness, and how they decide to seek medical care Cobb (1974) has advanced a similar conceptualization. His scheme calls attention to: 1. The objective and subjective stress associated with the stressful life event 2. The physiological and behavioral straining that takes place in response to that stress 3. The subsequent development of illness and illness—related behavior Cobb thinks of personal characteristics in much the same way as Rahe views past experience. but also suggests that Imderstanding the role of stressful life events in the precipitation of illness and illness behavior requires an appreciation of the social situation within which the stressful life event is embedded. In particular, Cobb mentions the need to understand the current life situation. the nature of social supports available to the person, and the at— titudes of peers and of medical care gatekeepers. Personal char— acteristics and the nature of the social situation are important mediating factors and can serve to provide either virtual immunity from or excess susceptibility to illness in the event of the occur— rence of stressful life events. B.S. Dohrenwend (1978) and B.P. Dohrenwend (1979) have con— ceptualized the process whereby psychosocial stress induces psy— chopathology. much like Cobb. A stressful life event. usually un— desirable and always requiring change or readjustment, leads to a stress reaction that is inherently transient or self—limiting. Sit— uational and psychological factors serve as mediators in determin— ing whether these transient stress reactions will result in psycho- 7 Stressful life FIGURE 1 events Adverse health changes A. Victimization Hypothesis Stressful Psychophysiological strain or other non- life events pathological response Adverse health changes B. Stress-strain Hypothesis Social situations Stressful life events Personal dispositions C. Vulnerability Hypothesis Social situations Stressful Adverse life health events changes Personal dispositions D. Additive Burden Hypothesis Social situations Adverse health Personal / changes dispositions E. Chronic Burden Hypothesis Symptoms of adverse health change StmsSful Exacerbation life of adverse events health change F. Event Proneness Hypothesis From ”Stressful Life Events and Their Contexts edited by Barbara Snell Dohrenwend and Bruce P. Dohrenwend. Copyright ® 1981. Neale Watson Academic Publications. Inc.; copyright ® 1984, Rutgers University, The State University. Six Hypotheses About the Life Stress Process logical growth on the one hand; the development of psychopath— ology. on the other hand; or a lack of any substantial psychological change. Sarason's theoretical framework for stress research views stress as following "a call for action when one's capabilities are perceived as falling short of the needed personal resources" (1980, p. 74). Stress can be handled by appropriate orientation to the task, or conversely, by such responses as unproductive preoccupation. anx- iety. anger, depression, denial. or retreat into fantasy. Mediating factors are identified by Sarason as person or situational variables. Examples of person variables are characteristics such as the ability to anticipate danger, feel safe, or feel confident, and history of stress—arousing experiences. Situational variables include factors such as available social supports. Albee (1979) has proposed a sim— ilar analysis in his formula that suggests that the risk of developing an emotional disorder is directly related to organic vulnerabilities and stress and inversely related to coping skills. competence. and the adequacy of social supports. Conceptual models linking stressful life events to illness have a number of aspects in common. First. in the process of describing the links between stressful life events and subsequent disorders. life stressors need to be viewed within a social and psychological context. That context may contain factors that moderate the ef— fects of stressful life events. such as a strong social support net- work or personal robustness, or may contain factors that potentiate the effects of stressful life events. such as a history of poor crisis management, characteristic physiological overreaction. or a sense of external locus of control. Second. the long—term consequences of stressful life events may not necessarily be deleterious—they may have no measurable consequences at all or may have adaptive components. Finally, to evaluate the consequences of stressful life events, help—seeking patterns must be taken into account along with an understanding of the medical and psychological care sys— tem accessible to the person. ' Keeping in mind the social context or situation. personal dis— positions, and the stressful life event, we can examine the various hypotheses that have been advanced to account for that chain of events. Dohrenwend and Dohrenwend (1981b) have identified six competing hypotheses that are shown in figure 1. Model A suggests that cumulations of stressful life events cause adverse health changes. The Dohrenwends call this model the vic- timization hypothesis, and they believe that this model is appro— priate to account for studies of extreme situations, such as combat or incarceration in a concentration camp, or certain severe stresses, such as the death of a loved one, over which an individual has no control. Model B suggests that psychophysiological strain mediates the impact of life events on subsequent illness and is called the stress—strain hypothesis. Support for this hypothesis comes from 9 studies that show, for example, that if the effects of symptoms of psychophysiological straining are eliminated, correlations of stressful life event scores and measures of illness are significantly reduced. ’ Model C supposes that preexisting personal dispositions and so— cial conditions moderate the causal relation between stressful life events and psychopathology. This model is called the vulner— ability hypothesis. It is this model that forms the basis for the examination of the effects of mediating factors, such as the strength of social supports. optimism. or locus of control, on the relationship between stress and illness. Model D suggests that personal dispositions and social conditions make independent causal contributions to the occurrence of psy— chopathology. This model is called the additive burden hypothesis and represents an alternative to the vulnerability hypothesis in explaining the role of personality variables or social resources in the precipitation of illness. Model E proposes that transitory stressful life events have no role in precipitating illness, but rather, that more stable personal dispositions and social conditions alone cause the adverse health changes. This model is called the chronic burden hypothesis. Finally, Model F proposes that the presence of a disorder leads to stressful life events which, in turn. exacerbate the disorder. This model is called the event proneness hypothesis in the sense that proneness to stressful life events is thought to characterize persons who are already ill. It is possible to identify the principal activities in the field of stressful life event research within these conceptual models. The psychological or physiological consequences of a particular stressful life event or of cmnulative stressful life events are con— tinuing to be examined in a large number of studies. A growing number of studies are assessing the effects of mediating factors, including the characteristics both of the person and of the social environment. Fewer studies are seeking to identify positive con— sequences of stressful life events or are contrasting specific stressful life events in terms of their actual consequences. Finally, papers making general contributions to the methodology of stressful life event research have been numerous in recent years. It is perhaps an indication of the continuing progress in the field of stressful life event research that Holmes has identified 10 pre- ventive measures that derive from the recent research that has employed the Social Readjustment Rating Scale. These measures include: Become familiar with the life events and the amount of change they require. . . . Put the Scale where you and the family can see it easily several times a day. . . . With practice you can recognize when a life event happens. . . . Think about the meaning of the event for you and try to identify some of 10 the feelings you experience. . . . Think about the different ways you might best adjust to the event. . . . Take your time in arriving at decisions. . . . Anticipate life changes and plan for them well in advance, if possible (1979. p.51). Summary Stimulated by the seminal work of Holmes and Rahe (1967). the past 15 years have seen a tremendous growth of scholarly activity in examining the role of stressful life events in the precipitation of physical as well as emotional disorders. There seems little doubt that stressful life events have the capacity to make healthy but vulnerable people ill. Many salient variables in the link between stressful life events and subsequent illness have been identified, at least theoretically. and general conceptual models for viewing the steps in the process between stressful life events and illness are in place. In contrast to the disease prevention paradigm that seeks causes and ways of preventing specific diseases, the stressful life event paradigm is nonspecific and seeks to understand and control increased vulnerability to illness in general. The process by which external stressors cause changes in internal bodily functioning is only partially understood. It seems clear. however, that reducing the incidence of such life stressors and increasing the capacity of persons subject to such stressors to deal with them more competently represent promising strategies for continued careful research—research that may lead to more successful prevention of a wide variety of disorders. 11 Chapter 2 THE CONSEQUENCES OF STRESSFUL LIFE EVENTS In this chapter, studies that have examined the effects of stress— ful life events on both physical and psychological well—being will be selectively reviewed. This review will be divided into three sec— tions—the association of stressful life events and psychopatholo— gy; the association of stressful life events and physical illness: and the eff ects of specific stressful life events. Stressful life Events and Psychopathology Schizophrenia In one of the earliest studies of the temporal relationship be— tween the occurrence of life events and the onset of schizophrenia. Brown and Birley (1968) conducted indepth interviews with 50 psychiatric patients who had had an acute onset of schizophrenia and separate interviews with their relatives. Brown and Birley were interested in obtaining accurate data regarding the frequency of certain kinds of crisis and life changes during the 4 months before the onset of the disorder. The authors conducted essentially identi— cal interviews with a general population sample of 325 persons for comparison purposes. After having ascertained the date of onset (for the control popu— lation, the date of interview). the authors determined the occur— rence of events "which on common sense grounds are likely to produce emotional disturbance in many people" (p. 204). They also judged each event reported as either clearly independent or pos— sibly independent of the disorder (that is. how completely the event was outside the patient's control). They found that the patient group experienced substantially more events overall and. most important, that this difference oc— curred in the 3 weeks immediately prior to onset. The frequency of stressful events for the patient group in this period was three times greater than in the general population—60 percent and 20 percent. respectively, experiencing at least one stressful life event. Par- 12 thermore. outside these 3 weeks, the rate in the two groups was virtually the same. These results were obtained whether clearly independent events were considered together with or separately from possibly independent events. The authors concluded that taking all into account, we believe that there is reasonably sound evidence that environmental factors can precipitate a schizophrenic attack and that such events tend to cluster in the 3 weeks before onset. We do not regard the events as sufficient causes. We believe that a number of factors must contribute and perhaps coincide to produce the conditions necessary for an acute schizophrenic attack. and that we have demonstrated one of these—some sort of crisis or life change (p. 211). In another early study. Steinberg and Durell (1968) reviewed the military records of every noncommissioned soldier in the U.S. Army who was hospitalized with a diagnosis of schizophrenia during 1956—60. They found that the rate of hospitalization was markedly increased in the early months of military service compared with the rate of hospitalization during the second year of military serv— ice. Steinberg and Durell found that only a very small proportion of the early hospitalizations was due to the detection of chronic cases, and thus concluded that there was "a genuine increase in the rate of onset of acute schizophrenic symptoms during the early months of service" (p. 1104). These authors believed that their findings were consistent with the hypothesis that a situation producing an intense need for social adaptation is an effective precipitant of acute schizophrenic symptoms in individuals predisposed to develop the schizophrenic syndrome (p. 1104). Several similar studies have subsequently been reported; Rabkin (1980) examined a number of them to determine if stressful life events were related to the onset of schizophrenia. These studies fall into three types according to the comparison group used. First, some studies contrast the frequency of stressful life events in schizophrenics and normals. Other studies contrast schizophrenics with other types of psychiatric patients. Finally, some studies have contrasted the frequency of stressful life events in relapsing and nonrelapsing schizophrenics. In the two studies contrasting schizophrenics with normal con— trols, some support was found for the hypothesis that patients re— ported more recent stressful life events than did the controls. In the case of the five studies contrasting schizophrenics with other psychiatric patients. stressful life events were not found to be more frequent in the case of schizophrenics. If anything. such events appeared to occur more commonly in the case of depressed 13 patients. Finally, in the case of the four studies contrasting schizo— phrenics who relapsed with those who did not. stressful life events were reported more frequently in the case of the relapsing pa— tients. Rabkin's conclusion based on her critical review was that "life events may contribute incrementally to an already inflated stress level and so may influence the timing, if not the probability. of illness onset" (p. 424). Affective Disorders A number of studies have been reported in which the incidence of affective disorders was examined as a function of the presence or absence of prior stressful life events. These studies have con— sistently found that such life events place a person at higher risk of depression than persons who do not undergo such life stressors. Paykel (1978, 1979; see also Paykel et a1. 1969) has been examin— ing the relationship of life events to depression and has accumulat— ed considerable evidence pointing to the importance of such events in the precipitation of clinical depression. Stressful life events, particularly those that are judged to be undesirable and those that reflect losses. have been found to be significantly more frequent in the 6 months prior to symptomatic onset of depression than in a comparable 6—month period in a control population. In general, Paykel found that stressful life events were implicated in the genesis of most depressions except for a small minority that ap— peared endogenous in character. While the vast majority of persons who underwent stressful life events did not become depressed, people who underwent these events were clearly at excess risk of developing depressions. Specifically. a stressful life event in— creased the risk of depression five— to sixfold during the 6 months following the event. Furthermore, the risk was particularly high immediately after the event and tended to diminish with passing time. Relative risk for depression increased tenfold during the first month after a stressful life event when contrasted with the risk in a group in which the stressful life event had not occurred. Similarly. Dunner et a1. (1979) found a high incidence of reported stressful life events in the 3 months prior to the episode in the case of 79 patients in treatment for bipolar manic—depressive disorders. For patients whose first episode was manic, more men than women reported stressful life events prior to onset. For patients whose first episode was depressive. more women than men reported stressful life events. Among these events. the most common was childbirth. Jacobs et a1. (1974) contrasted matched samples of 50 schizo— phrenics and 50 depressives admitted for the first time for inpa— tient care, in terms of the reported stressful life events in the 6 months prior to onset of illness. Depressives reported significantly more stressful life events than schizophrenics during the same 6—month period. particularly increases in the number of arguments 14 with family members not living in the house and with closely related members of the opposite sex, e.g., steady girl friends. An excess number of undesirable stressful life events was found in the case of the depressives, but no such difference was found in the case of desirable events. Similarly. an excess number of events referring to exits from the social field (e.g., death. divorce, leaving home) was found for the depressives, while no such excess of events representing entrances into the social field (e.g., marriage. birth of a child. a new person moving into the home) was found. Depressives reported significantly more stressful life events in the area of finances and health than did schizophrenics. The authors view their findings as consistent with the position that "lif e events are involved in the genesis of depression and schizophrenia" (p. 452). The relationship appears to be stronger for depressives than for schizophrenics, however. and appears to increase with the independently judged stressfulness of the event (see also Rahe 1979) In summarizing the finding of this series of studies, Paykel (1978) wrote: The occurrence of any of the spectrum of events included in these studies increases the risk of developing a schizophrenic illness in the next 6 months by something of the order of two or three times; of depression by two—five, and of a suicide attempt by about six times, with a particularly high risk of the last in the month after the event. . . . Occurrence of the more stressful classes of events increases the risk of depression at least sixfold. . . . For all disorders risks seem to diminish with time. as reflected by declining life event rates assessed ret— rospectively (p. 251). Brown and Harris (1978) studied the life event precursors of de— pression in a clinical sample of 114 female inpatients and outpa— tients and in a comparison population of 458 women living in the same community-—a borough within Inner London. In addition to examining life events, Brown and Harris assessed both predisposing and current vulnerability factors. Their hypothesis was that "clinical depression is an understandable response to adversity" (p. 46). They found that while relatively minor stressful life events were no more common among the patients than the controls. severely stressful life events involving long—term threat were four times more common among patients than among the control group. What is perhaps the area of greatest uncertainty regarding these studies is that while stressful life events appear to occur more commonly among persons who are ill than among health comparison groups, within the group of ill persons. stressful life events are not dramatically more common in the case of one diagnosis than another. Hudgens et a1. (1967) found, for example, in contrasting the incidence of prior stressful life events of hospitalized de— 15 pressed patients with the incidence of matched persons hospital— ized on medical and surgical wards who had no history of psychiat— ric illness in their lives. there were only trivial differences in the frequency of such events. Neurotic Disorders Stressful life events appear to be implicated in the development of neurotic disorders just as in the case of psychotic disorders. Cooper and Sylph (1973) found many more stressful life events during the preceding 3 months in a sample of 34 patients with new cases of neurotic illness seeking medical care in London than in a matched normeurotic sample. Serious threatening events (such as unexpected crises and news of failure) were particularly common among the new neurotics and uncommon among the controls. Cooper and Sylph analyzed their data in such a way that they were able to calculate the proportion of neurotic illness attributable to stressful events and the average time by which the stressful life events advanced the onset of the neuroses (Brown et a1. 1973). They concluded that "major events act as aetiological agents in the strict sense of provoking illnesses which otherwise, in terms of probability. would not occur until years later" (Cooper and Sylph, p. 427). Tenth and Andrews (1978) surveyed a representative suburban sample of 863 persons in Sydney, Australia, and examined the re— lationship between a self—report measure of neurotic impairment (Goldberg 1972) and three measures of stressful life events in— curred during the previous yeaIHtotal number of events, events weighted for emotional distress. and cumulated score weighted for life change. Although the three measures of stressful life events were highly correlated (correlation coefficients ranged from .79 to .89) and all three measures were modestly but significantly corre— lated with the measure of neurotic impairment,partial correlations suggested that only the measure of stressful life events weighted for distress was independently related to neurotic impairment. Tennant and Andrews concluded that "lif e events are pathogenic in neurosis because of the emotionally distressing impact and not simply because they produce life change" (p. 863). Barre“; (1979) studied stressful life events in the 6 months prior to onset in a group of 203 outpatients with neurotic disorders. His work was concerned with neurotic depressives and with patients whose chief symptom was anxiety. In contrasting the depressed patients with those whose primary symptoms were anxiety. Barrett fotmd significant differences in the frequency of reports of unde— sirable events and events that represented a change in the life setting. Such events not only occurred more frequently but were generally rated as more distressing in the case of neurotically depressed patients than those with anxiety. particularly when the event involved losses or changes in interpersonal relationships. By 16 way of contrast, events that involved the need to carry out some instrumental behavior (e.g., take an examination. change to a different line of work. financial difficulties, arguments) were seen as more stressful by patients with anxiety neuroses than by depressives. In another study, Henderson (1981) examined the interaction of adversity, as assessed by the presence of stressful life events, and the availability and sense of satisfaction with social relationships in a multiwave survey of the general population in Canberra. Scores on an index of neurotic disability were found to be related both to the presence of stressful life events and to the adequacy of social relationships, with the relationship considerably stronger in the case of stressful life events. With regard to the interaction of these two causal factors, Henderson concluded that "a deficiency in social relationships is more strongly associated with subsequent symptoms if there is also high adversity" (p. 393). Childhood Disorders Findings linking stressful life events in childhood with subsequent emotional difficulties are more mixed. Sandler and Block (1979) examined the relationship between life stress and maladjustment in a group of 99 maladjusting children and 44 matched nonmaladjust— ing children in four inner—city schools. The children were generally poor and of minority status. Most were from families on welfare. Life stress was assessed using a modification of the stressful life event scale developed for elementary school children (Coddington 1972a) and covered the preceding year. The scale was completed by parents. and items selected for inclusion in the study were all judged to be events beyond the child's control; that is. events thought not to be confounded with the behavioral adjustment of the child. In addition. all items were independently rated as desirable. undesirable. or ambiguous. Nature of maladjustment was assessed by teachers and by parents. Maladapting children were found to have suffered significantly more stressful life events. particularly undesirable ones, in the preceding year than adapting children. but the difference was found only in the case of children not on welfare. That is, when maladapting children on welfare were contrasted with adapting children on welfare, no difference in the frequency of stressful life events was found. The authors asserted that children on welfare had generally gone through fewer stressful life events than children not on welfare and suggested that some factors associated with being on welfare (perhaps including the availability of stronger social support systems) appear to moderate the effects of stressful life events. The fact that the stressful life events studied in this sample were not under the control of children "strengthens the implication of stress as a contributory cause of the children's ad— justment problems" (p. 436). 17 Stressful Life Events and Physical Illness Fully as much interest has been expressed in the possible role of stressful life events in precipitating physical illness as in the pre— cipitation of emotional disorders (Cohen 1979). Evidence linking stressful life events with subsequent physical disorder seems very persuasive. Hinkle (1974) has been examining the effects of changes in interpersonal relationships and of changes in the social and cultural milieu on health in a series of studies over the past 30 years. In a study of more than 500 telephone operators who were part of a comprehensive health insurance program, and who had worked for 20 years or more, Hinkle found a strong connection be— tween stressful interpersonal and environmental difficulties and illness. He concluded that the longitudinal and retrospective life histories obtained from the frequently ill and moderately ill usually suggested that a cluster of illness coincided with a period when the individual was experiencing many demands and frustrations arising from his social environment or his interpersonal relations. These histories suggested that significant changes in the relations of ill people to their social group and in their relations to other important people in their lives were significantly associated with changes in their health (p. 21). In Hinlde's studies of Chinese—born persons who immigrated to the United States soon after World War II, of Hungarian refugees who came to the United States in 1956, and of American prisoners of war in North Korea, excess illness was generally found. In the case of the refugees from the Hungarian revolution, Hinkle wrote, Among these Hungarians . . . we found that individuals dif— fered markedly in their general susceptibility to illness. . . . The frequently ill people . . . had experienced a greater num— ber of disturbances of mood, thought, and behavior (p. 29). Aakster's survey of more than 1,500 adults in Holland (1974) examined the interrelationships of psychosocial stresses and health disturbances. Aakster's analysis indicates that while most of the variation in health disturbances is accounted for by generalized psychosocial dysfunction, specific disorders seem to be associated, in addition, with certain specific psychological difficulties. Aakster has concluded that illness is the more or less automatic result of a failure to adjust to stress. Stress was defined as a discrepancy between the ideal state of the individual in relation to his desired goals and the actual position in which he finds himself. We therefore infer that there is a close relationship between all kinds of 18 failures of the individual in realizing his existential goals. and the development of illness (pp. 88—89). Gallin (1980) examined the use of medical services in a group of 124 relatively healthy women, ages 20 to 40, in an ambulatory health care center. She found that those women who were poor, less educated, and separated or divorced were likely to define the physician's authority very broadly and to feel that "their social resources were only nominally effective in mediating between them and the pressures of life—strains" (p. 262). Gallin found that "these women visited the health center more frequently for vague, unorganized symptoms than women whose situations were less bur— dened" and that "the disadvantaged women's high use of medical services was an attempt to deal with their distress—social, psych— ic, and somatic—and consistent with and an adjustment to their deleterious life situations" (p. 262). Cooke and Greene (1981) examined the relationship between stressful life events and the report of both psychological and physi— cal symptoms in a group of 131 premenopausal and menopausal women. The authors found that subjects who developed psycho- logical symptoms were reacting to general stress rather than to internal changes alone. In addition. these subjects appeared to be less able to cope with general stresses than had formerly been the case. Totman (1979) examined the responses to stressful life events in a sample of 25 males who were admitted to a coronary care unit because of a first myocardial infarction and a comparison group of males attending the ear. nose. and throat clinic at the same hospi— tal. Totman fomd support for his hypothesis that during the year prior to seeking medical care the heart patients would have ex— hibited a significant decrease in involvement in goal—directed activities and in social interaction when contrasted with the ear. nose, and throat patients. Both groups had had about the same number and variety of stressful life events during the preceding year. but their reactions appeared to be quite different. Totman's conclusion is that "being cut off from customary pursuits and from familiar social contacts is potentially harmful to health—espe- cially when no substitutes are formed" (p. 198). With regard to the specificity issue in the link between stressful life events and vari— ous disease responses, Totman concludes that growing evidence that a similar class of life event is to be found in the premorbid histories of patients who develop psy— chiatric symptoms and patients who develop physical syrup— toms . . . brings to the forefront the possibility that physical and psychiatric illness are alternative responses to the same underlying ‘problem' (p. 198). In another study, Totman et a1. (1980) examined the relationship 19 of stressful life events to vulnerability to the common cold in a sample of 52 persons who were given colds by nasal inoculation with an infecting rhinovirus. These subjects ranged in age from 18 to 49 and included people from many walks of life. Prior to their inoculation with the rhinovirus, they were assessed on a variety of measures of recent life stress and with a number of personality inventories. The severity of the cold was measured by the amount of virus present in nasal washings. The authors obtained "clear evi- dence of a psychosomatic component in colds" (p. 155). Introverts deVeloped significantly more severe colds than extroverts. Perhaps more importantly, these authors found that certain stressful life events, particularly those associated with changes in general activity level, were significantly related to the magnitude of the subsequent infection. Selzer and Vinokur (1974) studied a sample of 274 automobile drivers (of whom nearly half had been involved in an automobile accident within the past year) and 258 automobile drivers who were in either inpatient or outpatient treatment for alcoholism (of whom about 20 percent had been involved in an automobile accident within the past year). While demographic and personality variables were very modestly associated with accident history in both groups of drivers. the frequency of reported stressful life events during the past year and measures of current physical and subjective stress in a variety of life contexts (e.g.. marital and family life. working conditions. financial status, and health concerns) were much more highly related to automobile accident history. Selzer and Vinolcur concluded that life change and subjective stress measures appear to be statistically more important than the demo— graphic. personality, and social maladjustment variables that have previously been the focus of behavioral scientists . relatively transitory life events, subjective stress phenomena. and the resultant changes they impose are an important factor in the traffic accident process (p. 906). Cassel (1974b) has concluded that one of the psychosocial factors that may play a role in reducing vulnerability to disease is the presence of other members of the same species. His conclusion derives from a review of animal studies as well as studies with human subjects. He has proposed four hypotheses for consideration. First, the social process linking high population density to enhanced susceptibility to disease is not the crowding per se but the disordered relationships that, in animals, are inevitable con- sequences of such crowding. . . . In human populations the circumstances in which increased susceptibility to disease would occur would be those in which there is some evidence of social disorganization (p. 1041). 20 Second, persons who occupy subordinate podtions in the hierarchy of power or prestige will be particularly vulnerable to disease. Third, biological and social buffers exist (e.g., coping capacities and social supports) that can cushion the individual from the phys— iological or psychological consequences of social disorganization. Finally, Cassel has suggested that Imder conditions of social change and social disorganization susceptibility to disease is in general increased. A number of these hypotheses were examined in the work of Rowland (1977). who reviewed the literature linking death among the elderly with three specific stressful life events—death of a significant other. relocation, and retirement—events identified as being particularly stressful among the elderly. In the case of death of a spouse. Rowland concludes that evidence argues for the loss of a spouse being detrimental. particularly in the case of the surviving male and particularly within the first year of bereavement. Re— location, specifically from a private home to some type of group home for the aged or from one group setting to another. appears to predict death for those elderly persons who are already in poor health. Retirement appears to be a much more equivocal stressful life event. It has not yet been possible to design a study in which retirement has been independent of prior health status. Rowland has also examined the principal theories that have been advanced to link these three stressful life events to premature death, including crisis theory, reinforcement theory. and learned helplessness theory. Each model appears to generate important and researchable propositions. Crisis theory would suggest, for exam— ple, that prior coping styles and coping skills might play a role in determining whether relocation might lead to premature death. Reinforcement theory would lead one to examine patterns of reinforcement before and after relocation and changes in these patterns as a function of the event. Learned helplessness theory would postulate, for example, a differential effect of forced versus voluntary relocation. In addition, each theory would emphasize somewhat different approaches to reducing the impact of relocation. The Effects of Specific Stressful Life Events In Dohrenwend's (1979) recent assessment of the literature link- ing specific stressful life events to subsequent psychopathology, he suggests that: with respect to . . . whether stressful life events are important. in the causation of . . . psychopathology in the general popu— lation. the best evidence is indirect. The reason is that this evidence comes not from epidemiological studies of commu— nity populations under ordinary conditions but rather from 21 studies of individuals and groups under extraordinary condi— tions imposed by natural and man—made disasters, especially the disaster of war. . . . Studies of such extreme situations have provided compelling evidence that the stressful events involved can produce psychopathology in previously normal persons. . . . Certainly at the extreme of the exposure to the brutalities of Nazi concentration camps, there is strong evi— dence that not only does severe stress—induced psycho— pathology persist in the survivors . . . but also that the survi— vors are more prone to physical illness and early death (pp. 3—4; see also Kinstron and Rosser 1974). Ursano and his colleagues (Ursano 1981; Ursano et a1. 1981) have been monitoring the long—term consequences of one particular stressful life event—being a prisoner during the Vietnam War. Their research leaves no doubt that repatriated prisoners of war had to face a significant degree of psychological stress and an increased risk of psychiatric illness. Loss of Employment Kasl (1979) has examined job loss and retirement as two particu— lar stressful life events and has reviewed studies linking these events to changes in mental health status. Job loss and retirement must be understood within a context that takes into account the meaning of work to the individual and the social, psychological, and physical setting in which the work takes place. that is, the work environment. Kasl concludes there is little evidence to suggest that the transi— tion from work to retirement is accompanied by a generally ad— verse impact on the physical or mental health of the person. Varia— tions in postretirement adjustment seem to reflect continuities with preretirement status and adjustment, particularly in the areas of health, social and leisure time activities, and general well— being. In contrast. loss of a job, particularly involuntarily. does cause stress, but of relatively short duration. Persons who become unemployed appear to adapt to that state and its consequences rather quickly, whether or not they find new employment, except for the stresses associated with economic deprivation. Oliver and Pomicter (1981) examined the effects of lmemploy— ment on depression in a sample of 182 automotive assembly line workers following an economic recession responsible for extended layoffs. Because these authors treated job layoffs as a stressful life event, they hypothesized that employment insecurity would be significantly and positively correlated with depression. Employ— ment insecurity was measured objectively (by whether or not the employee had been laid off and by their seniority) and subjectively (by perceived job instability). In the analysis of their data, Oliver and Pomicter first found that 22 they had to distinguish between those employees who had been laid off and those who had not. In the group who had not been laid off , depression scores were correlated with variables previously found to be related to depression—race. marital status, and education. In addition. judged optimism of the future of the economy was found to be related to level of depression. Among the group that had been laid off , depression scores were significantly related to judgments about the employee's future job and financial security rather than to any demographic variables, such as marital status. The stressful character of being laid off could be seen in the ensuing high depression scores when there was little optimism that they would soon be rehired. Marital Disruption In investigations such as those of Holmes and Rahe (1967), sepa— ration and divorce are consistently judged by samples of prof es— sionals or community residents as among the most stressful life events that can occur to an adult. In the stressful life event scales developed by Coddington and his colleagues (1972a. 1972b; Heisel et al. 1973) for use with children. only death in the family ranks higher than separation and divorce of the parents in judged stressfulness (also see Gersten et al. 1974). 'IWo other types of studies serve to corroborate and supplement that assertion. First, studies assessing morbidity and mortality rates as a function of marital status consistently show that persons who are separated and divorced are at substantially higher risk than are persons in intact marriages (Bloom et al. 1978). This statement holds true for psychopathology, many physical disorders. motor vehicle accidents, alcoholism. suicide, and homicide. Admission rates into both inpatient and outpatient psychiatric facilities are far higher for separated and divorced persons than for persons in intact marriages particularly in the case of males (Thomas and Locke 1963; Grad and Sainsbury 1966; Crago 1972; Redick and Johnson 1974; Robertson 1974; Bloom 1975; Bloom et al. 1978). Community surveys have found that mental status is substantially poorer in the case of adults whose marriages have been disrupted than persons in intact marriages (Srole et al. 1962; Blumenthal 1967; Tauss 1967; Briscoe et al. 1973; Radloff 1975; Pearlin and Johnson 1977). Related studies yielding similar findings have been reported by Paykel et al. (1969). Leff et al. (1970). Smith (1971), Udell and Hornstra (1975), Zautra and Beier (1978), Belfer et al. (1979), and Kessler (1979). Another type of study that provides evidence as to the stressful nature of marital disruption is based on self—reports of persons undergoing marital disruption. These studies identify the specific nature of the perceived stress. They also converge on identifying a small but important number of problems faced by persons under— going marital disruption. Included in such a list would be a general— 23 1y weakened social support system. the need to work through the variety of psychological reactions to the disruption, problems with child rearing, resocialization, finances, education and employment planning. housing and homemaking, and protection of legal rights. There is some consistent evidence that the most critical period in the marital disruption process is around the time of separation. Within the past several years, an additional group of studies has been published continuing the identification of problems faced by persons undergoing marital disruption (Hetherington et a1. 1976. 1977; Weiss 1976, 1979; Chiriboga and Cutler 1977; Herman 1977; Kitson and Sussman 1977; Raschke 1977; Chiriboga et a1. 1978; Bane 1979; Chiriboga et al. 1979; Granvold et al. 1979; Spanier and Casto 1979; Kitson et a1. 1980; Wallerstein and Kelly 1980; Kelly 1982). Kelly (1982) has summarized these findings—stress asso— ciated with marital disruption is reported in terms of anger, de— pression, loneliness, economic difficulties, disequilibrimn, regres— sion, and ambivalent, but persistent, attachment to the spouse. These stress responses are invariably multifaceted and surprisingly long lasting. Community-Wide Stressful Life Events Certain types of events, such as economic recession or major disasters, may constitute life stressors for an entire community. Viewing economic downturns as stressful life events, however. does run the risk of committing the ecological fallacy. that is, the risk of assmning that an event that affects a community at large affects each community resident equally. Stated differently. the ecological fallacy would be committed by showing a relationship between economic downturns and community—wide well—being without showing that any specific individual suffered after experiencing economic stress. In spite of these problems in interpreting the findings of studies examining stressful life events that have an impact on entire com— munities. a number of researchers have felt it useful to assess the extent to which such community—wide stressful events are 'associ— ated with increases in various disorders. For example. Brermer (1973) found that increasing unemployment rates was highly asso— ciated with subsequent increases in mental hospitalization rates. Catalano and Dooley (1977). examining field survey data collected in Kansas City, found that economic change was followed in 1 to 3 months by significant increases in self—reported life change and depressed mood. In a more detailed subsequent analysis of the Kansas City data, Dooley and Catalano (1979) were able to examine the relationship between both economic and noneconomic life events and reported psychiatric symptoms in groups disaggregated by age, sex, and socioeconomic status. Data for the Kansas City study were ob— tained by interviewing one adult in each of 28 different households 24 chosen weekly for 16 months. Economic life events included start— ing to work. graduating from school, getting laid off, and acquiring property. Noneconomic life events included getting married or divorced, pregnancy, getting arrested. and taking a vacation. These two types of events were also rated as to whether they were gen- erally seen as desirable or undesirable. Economic indicators at the community level included such variables as monthly unemployment rate and monthly inflation rate. Psychiatric symptoms were as— sessed by means of a standard short self—report measure (Langner 1962). Individually reported stressful life events were not found to be predictive of subsequent self—reported psychiatric symptoms. Eco— nomic indicators at the community level were significantly related to subsequent economic life event variables at the individual level, however, and to subsequent reports of psychiatric symptoms. Re- lationships tended to be stronger in men than women. There were no remarkable differences in the obtained relationships as a ftmc— tion of age. but the low socioeconomic status group was more responsive to economic change than the middle income group. Dooley and Catalano (1979) concluded that "changes in a communi— ty's economy are associated with later changes in noneconomic . . . and economic events" (p. 392). Furthermore, they concluded that "increases in unemployment and absolute economic indicators precede. by 4 to 8 weeks, increases in psychophysiological symp— toms reported by samples of a metropolitan population" (p. 393). In their recent review of the literature linking economic change and subsequent behavior disorder. Dooley and Catalano (1980) have conceptualized the links between economic change and treated disorder rates in a series of discrete steps. First. they postulated a link between environmental economic change (e.g.. increasing un— employment rate) and reported increased stressful life events. both economic and noneconomic, at an individual level. Second, these individual stressful life events may be linked to subsequent symp— tomatology. Third. this increased symptomatology may be linked to subsequent increases in demand for health—related services resulting in an increased treated incidence rate. Brenner (1979) has reviewed his own work linking economic characteristics of the social environment with physical and psy— chological disability. He has concluded that ”major economic life changes . . . are associated with severe pathology in the areas of mental and physical health and criminal aggression” (p. 174). In addition, further analyses of his data have suggested that while "abrupt economic changes . . . are stress provoking . . . undesirable changes, such as unemployment and income loss. are substantially more generative of pathology" (p. 175). The impact of major disasters that may constitute stressful life events for entire communities has been examined for a number of years. Perhaps the most carefully studied such event was the Buffalo Creek disaster. On February 26, 1972. a dam constructed 25 479-088 0 - 85 — 2 by the Buffalo Mining Company in West Virginia gave way, 1m— leashing more than a million gallons of water and mud into Buffalo Creek Valley. The water's impact lasted no more than 15 minutes at any one point in the 18—mile-long valley, and within 3 hours the water and mud had washed into the Guyandotte River at the foot of the valley. Yet everything in the path of the flow was destroyed. A total of 125 persons were killed and 4,000 left homeless. A group of 65 social scientists and mental health professionals, mainly from the University of Cincinnati, were retained by the attorneys for 625 persons who brought legal action against the conglomerate owning the mining operation. In 1974 the plaintiffs agreed to a settlement of $13.5 million, of which more than $6 million was for "psychic impairment"——a term devised to describe the social, vocational, and personal disability from psychological causes arising from the disaster and its aftermath (Erikson 1976; Lifton and Olson 1976; Newman 1976; Stern 1976; Titchener and Kapp 1976). A team of raters assessed the degree of psychic impairment of the 625 plaintiffs and concluded that 2 percent had no impairment, more than 20 percent had minimal impairment, more than 70 per— cent had moderate to severe impairment, and 1 percent were con— sidered totally disabled. Thus, these judges found that negative psychological consequences were nearly universal and that they were often persistent. still affecting many survivors 2 years after the disaster. More recently, reports of the effects of the 1979 Three Mile Island accident have begun to appear. Three groups of persons thought to be at unusual risk have been studied—mothers of pre— school children living within 10 miles of the reactor. mental health agency clients living within the same radius, and workers at the nuclear power plant. Subjects at Three Mile Island were inter— viewed twice—8—9 months after the accident and. 3 months later, on the first anniversary of the accident. Comparison groups of mothers of preschool children, mental health agency clients. and nuclear power plant workers who lived near a neighboring nuclear power plant were also interviewed twice within equivalent time intervals. Important differences were found to exist between the two sam— ples of mothers of preschool children. Mothers at Three Mile Island were functioning under more stress and reported significantly more psychiatric symptoms during the year after the accident than com- parison group mothers, particularly in measures of anxiety and de— pression. Far smaller differences were found in the case of power plant employees and mental health agency clients (Bromet et a1. 1982). That is, while the Three Mile Island nuclear reactor accident appears to have had significant negative consequences for mothers of young children. it has not had any demonstrable effect upon mental health agency clients or upon nuclear reactor plant employees. 26 Summary A sampling of studies examining the consequences of stressful life events has been reviewed in this chapter. As can be noted. stressful life events play a significant role in the precipitation of physical as well as psychiatric disorders, in both children and adults. Among psychiatric disorders, depression appears to be the most common consequence of stressful life events. Otherwise. the effects of stressful life events appear quite nonspecific, that is, characteristics of the vulnerable person appear to be more impor— tant in determining the exact nature of the resulting illness than characteristics of the stressful life event. 27 Chapter 3 MODERATING FACTORS IN REACTIONS TO STRESSFUL LIFE EVENTS Increasing interest in both personal and social resources that can moderate reactions to stressful life events has been expressed over the past several years. Procedures for assessing the presence and extent of these moderating factors are not yet fully articulated, although there is considerable agreement as to the most important moderating factors to be assessed. Even where the strongest relationships have been found, corre— lations between the number of stressful life events and onset of illness are quite modest. According to Rahe (1979), improved un— derstanding of the relationships of stressful life events to illness onset will require closer examination of such moderating factors as the individual's perception of the event, social supports, psycho— logical defenses. coping capabilities. and typical behavior during times of illness. Dohrenwend and Dohrenwend (1974, 1978) comment on how lit— tle is known about the factors that moderate the impact of stress— ful life events. Such moderating factors may be physiological. psychological, or sociocultural in character. They indicate that further research on stressful life events must investigate the role of moderating situational factors. and in particular. past experi- ences with the same event. the availability of social supports. the predictability of the event, and the amount of control a person has over whether the event takes place. Rabkin and Struening (1976) also note the need to analyze both internal and external moder— ating variables. such as personality characteristics or the avail— ability of social support systems in developing a more complete understanding of the stressful life event—illness relationship. Johnson and Sarason (1979), in their review of moderators of life stress, discuss social support, locus of control and perceived con— trol. and level of arousability. They believe that high levels of social support may play a stress—buffering role and to some degree protect the individual from the effects of cumulative life changes. In addition. they suggest that persons are more adversely affected 28 by life stress if they perceive themselves as having little control over their environment. Finally, these authors have concluded that level of arousability. that is. stimulation seeking, may determine the extent to which persons are affected by life changes—the higher the level of arousability, the greater the reaction to life stress. Other moderating variables they have identified include previous history of dealing with stressors; coping skills (often higher in persons who have previously experienced stress); certain behavioral styles, e.g.. Type A and Type B behavior (Friedman and Rosenman 1974); habitual use of certain defense mechanisms, e.g.. repression or denial; and strategies for the appraisal of life events (Lazarus 1966). Personal Resources At the level of personal resources, the most commonly studied moderating factors are coping ability. certain personality vari- ables, particularly vulnerability, and social competence (Crandall and Lehman 1977). CofingAbmw Coping is defined as "efforts, both action—oriented and intra- psychic, to manage (that is, master, tolerate, reduce, minimize) environmental and internal demands, and conflicts among them. which tax or exceed a person's resources" (Cohen and Lazarus 1979, p. 219). Pearlin and Schooler (1978) have made a very useful conceptual and empirical analysis of coping behavior—a response that serves to "prevent. avoid, or control emotional distress" (p. 3). These authors distinguish between social resources, psychological re— sources. and coping responses. While resources represent what people have or are. responses refer to what people do. Social re- sources "are represented in the interpersonal networks of which people are a part and which are a potential source of crucial sup— ports—family. friends, fellow workers, neighbors. and voltmtary associations" (p. 5). Psychological resources "are the personality characteristics that people draw upon to help them withstand threats posed by events and objects in their environment" (p. S). Coping responses constitute the "behaviors, cognitions, and percep— tions in which people engage when actually contending with their life—problems" (p. 5). Pearlin and S’chooler conducted open—ended interviews with more than 100 adults in Chicago and asked them to describe the prob— lems they faced and how they dealt with them. Standardized ques— tions developed after analyzing the responses to the open—ended interviews were administered to a sample of 2,300 Chicago resi- dents. Statistical analysis of the coping—style questions revealed 17 29 coping factors that could be viewed not only in terms of problem areas (marriage, parenting, household economics, and occupation) but also in terms of their apparent fimctions or strategies. One type of coping, for example, seeks to modify the situation that represents the source of stress—negotiation in marital stress, punishment in dealing with parenting stress, seeking advice, etc. A second type of coping seeks to modify the meaning or sig— nificance of the stressful situation—making comparisons with the stresses endured by others, ignoring or minimizing the significance of the stressor, etc. A third type of coping seeks to minimize the straining that constitutes the response to the external stress‘ acceptance, avoidance, trying not to worry, relaxing, etc. Pearlin and Schooler have been able to assess the relative effi— cacy of various coping strategies in various settings and to contrast the relative importance of coping strategies and psychological re— sources in moderating the relationship between external stress and internal straining. In addition, they have examined the relationship of demographic factors to coping strategy preferences and to the strength of various psychological resources. While their specific findings will require replication and further elaboration, their work constitutes a highly significant beginning to the study of coping behaviors. Cohen and Lazarus (1979) have identified five forms of coping— information seeking, direct action, inhibition of action, intra— psychic processes, and turning to others—and have applied this analysis to the examination of how persons cope with the stresses of illness. When engaged in information seeking, the person tries to find out what problems exist and what, if anything. must be done. Direct action includes anything that is done about the problem. Since the skillful coper does not engage in action impulsively or ill—advisedly, inhibition of action constitutes a mode of coping. Intrapsychic processes include all forms of def ense-denial, avoidance, or intellectualization, for example. These defenses are particularly useful when the patient can do little about the problem except to allow therapeutic procedures to be applied. Finally, turning to others constitutes a form of coping, in that there is evi— dence that persons who are ill do better if they can maintain and use supportive social relationships. Cohen and Lazarus (1979) sug— gest that "the more helpless the person is, the more he or she must depend on cognitive or intrapsychic modes of coping" (p. 221). Personality Variables Kobasa (1979) examined a number of personality variables in terms of their moderating effects on the consequences of stressful life events. She studied a sample of 837 executives who. as a group, had undergone considerable stress during the preceding 3 years and was able to isolate two groups of high stress executives who dif— fered on their reported illnesses—126 subjects above the median 30 in total stress and below the median for total illness and a. second group of 200 subjects above the median for total stress and above the median for total illness. These two samples were subsequently reduced to 100 subjects in each group, of whom all were males, and most were 40—59 years old, married. parents, and college educated. Kobasa studied three personality variables related to coping style—the degree to which participants felt that they could con— trol or influence life events, involvement in or commitment to life activities, and the extent to which change was viewed as an excit— ing challenge to further development. As expected. in the total group (including those executives below the median in stress) there was a significant but modest correlation between total stress and total illness scores. Kobasa found numerous personality differences between the two groups of high stress executives. Kobasa writes, This study provides a basis for understanding how persons can encounter great stress and remain healthy nonetheless. In order to do so, one must have a clear sense of one's values, goals, and capabilities, and a belief in their importance . . . a strong tendency toward active involvement in one's environ— ment . . . an ability to evaluate the impact of any life event in terms of a general life plan with its established priorities . . . a belief that one can control and transform the events of one's experience . . . and finally, an ability to deal with external life stresses without their becoming threats to one's private sphere and causes of subjective strain (p. 420). Zubin and Spring (1977) have suggested that the single most important concept underlying the various models that have been advanced to understand schizophrenia is the concept of vulner— ability. They postulate two major components of vulnerability— inborn and acquired—and suggest that vulnerability is an endur— ing, relatively permanent trait. Inbom vulnerability is that which is "laid down in the genes and reflected in the internal environment and neurophysiology of the organism" (p. 109). Acquired vulner— ability is due to the influence of traumas, specific diseases, perinatal complica— tions, family experiences, adolescent peer interactions, and other life events that either enhance or inhibit the develop— ment of subsequent disorder (p. 109). Life event stressors make up a major component of acquired vulnerability in that they tax the organism's adaptive capacities. Among the variables that should be taken into account in pre— dicting a person's response to a stressful life event, Zubin and Spring include normatively perceived severity of the event, indiv— idually judged severity of the event, general competence, motiva— 31 tion to cope. and vulnerability. Zubin and Spring's review of the literature does not lead them to conclude that persuasive evidence of a link between preepisode competence and vulnerability exists. They do speculate, however, that reduced coping ability, charac— terized by frequent disequilibrium, particularly in the face of relatively mild threats, combined with high vulnerability, may significantly increase the risk of psychiatric disorder. Antonovsky (1979) refers to those personal assets that result in increased resistance to disease or disorder as generalized resist- ance resources. Generalized resistance resources are those charac— teristics of persons, groups, or social environments that are effec— tive in avoiding and combating a wide variety of stressors so that tension (the straining that is the organismic response to an external stressor) is not converted into subjective stress that is, in turn, a contributing factor in the development of disease. While Antonov— slqr acknowledges the existence of specific resistance resources, that is, personal or social resources linked to reduced vulnerability to specific diseases, he believes that generalized resistance re— sources are far more important, in part, because their presence creates the possibility that persons can locate disease—specific resources. Perhaps the most fundamental generalized resistance resource, according to Antonovsky (1979). is the sense of coherence— a global orientation that expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence that one's internal and external environments are predictable and that there is a high probability that things will work out as well as can reasonably be expected (p. 123). Social Competence A final important factor that can moderate the effects of stressful life events is competence, particularly in social situa— tions. Much human misery appears to be the result of a lack of competence. that is. a lack of control over one's life, of effective coping strategies, and the lowered self—esteem that accompanies these deficiencies. This opinion is emerging out of an analysis of a substantial body of research from a variety of domains that ap— pears to converge on competence building as one of the most persuasive preventive strategies for dealing with individual and social issues in many communities. According to Iscoe (1974). the development of competence in— volves obtaining and using information and other resources "so that the members of the community may make reasoned decisions about issues confronting them. leading to the most competent coping with these problems" (pp. 607—609). As to the components of indi— vidual and community competence, Iscoe suggests, first, a reper— toire of possibilities and alternatives; second, the knowledge of 32 where and how to acquire resources; third. clout; and finally. as a consequence—hope. self-esteem. and power. Not power in the sense of the ability to control others. but rather in the sense employed by the Spanish author, Salvador de Madariaga, when he wrote, He is free who knows how to keep in his own hands the power to decide at each step the course of his life and who lives in a society which does not block the exercise of that power (Brad— ley et a1. 1969). Ryan (1967) used the term "power" in the same sense when he suggested that self-esteem is to some extent an essential requirement to the very survival of the human organism . . . and is partially de— pendent on the inclusion of a sense of power within the self— concept. . . . A mentally healthy person must be able to per— ceive himself as at least minimally powerful, capable of in— fluencing his environment to his own benefit, and further . . this sense of minimal power has to be based on the actual experience and exercise of power (p. 50). These authors contend that competence, power, and self—esteem are inextricably linked and that their loss may constitute a power— ful stressor. Empirical study of coping skills, social competency, and vulner— ability in terms of their roles in moderating the effects of stressful life events has only recently begun. The concepts of vulnerability and of coping styles and strengths have not yet been satisfactorily operationalized but considerable progress can be expected in the next several years. There is no question about the shared belief in the importance of the study of this set of personal characteristics for the more complete understanding of stressful life events and their consequences. Social Resources At the level of social or environmental resources. enormous activity is taking place aromd the concept of social support sys— tems and social networks (Dean and Lin 1977; Caplan 1981). Cobb (1976) has recently reviewed the role of social support as a moder— ator of stressful life events. Cobb views social support as falling into one of three classes. Persons can be said to have social support when they believe that 0 They are cared for and loved. 0 They are esteemed and valued. 33 0 They belong to a network of communication and mutual obligation. Cobb has shown that the presence of adequate social support is associated with the lack of complications of pregnancy, particular— ly among persons high in life stress, with recovery from a variety of illnesses and response to a variety of medical procedures, with success in alcoholism recovery programs, with favorable manage— ment of the stress of involuntary \memployment. with successful coping with bereavement. and with better management of the illnesses and infirmities associated with aging. Cobb (1976) comments that what is new about these studies is the accumulation of evidence that adequate social support can protect people in crisis from a wide variety of pathological states: from low birth weight to death, from arthritis through tuberculosis to depression, al— coholism, and other psychiatric illness. Furthermore, social support can reduce the amount of medication required and accelerate recovery and facilitate compliance with prescribed medical regimens (p. 310). As for the processes by which social supports enhance health. Cobb believes that social support encourages coping (manipulation of the environment in the service of the self) and adaption (change in the self in order to enhance person—environment fit). The concept of social support is central to Caplan's (1981) con— ceptualizations of stress mastery. Caplan suggests that responses to stressful life events go through four interdependent phases. Phase I refers to behavior that enables the individual to escape from the stressful environment, and, according to Caplan, a supportive social network complements and supplements those specific aspects of the individual's functioning which are weakened by the effect of the stressful experience (p. 415). Phase 2 refers to behavior that results in the acquisition of new capabilities to change the stressful circumstances and their conse— quences. Caplan suggests that a support system can act as a "set of auxiliary ego functions for the stressed individ " (p. 416), and notes that persons undergoing stress usually exhibit spontaneous increases in their affiliative needs and initiate contacts designed to activate social support systems as they are needed. Phase 3 in the complex response to stressful life events refers to those behaviors that defend against anxiety. anger, and other dys— phoric affects. Caplan sees the availability of social supports as facilitating the mastery of affect, in part by strengthening such defenses as denial and isolation. Finally, Phase 4 refers to the internal readjustments that serve 34 to come to terms with the stressful life event and its sequelae. The activation of social support systems has been institutionalized in varying cultural condolence traditions. for example. and Caplan underlines the general importance of long—lasting social support systems in helping people who undergo stressful life events master the complex feelings associated with the life stress. Social Resources and Psychological Well-Being The role of social resources in the maintenance of psychological well—being in times of stress has been amply documented in the empirical literature. Andrew et al. (1978) provided information on a very brief measure of coping style (immature vs. mature) and social support (including availability of support in times of emer— gency, neighborhood interaction, and community participation). Neurotic impairment was significantly more common among im— mature copers than mature copers, and among persons with few social supports available in times of emergency. Of the group of 226 persons low in stressful life events, high in coping, and high in social support, only 12.8 percent were found to be at risk of being psychologically impaired. Of the 67 persons high in stress, low in coping. and low in social support, 43.3 percent were found to be at risk of psychological impairment. In this study, which the authors think of as only a first step, coping style and social support did not serve to moderate the effect of stressful life events upon psycho- logical impairment; rather. each appeared to have an independent and additive effect upon impairment. Burke and Weir (1977) studied 189 couples in Ontario, measuring life stresses, satisfaction with the spouse's helping efforts, job and marital satisfaction, and mental and physical well—being. Measures of well—being were significantly positively correlated with satis— faction with spouse’s helping efforts and negatively correlated with levels of stress in the case of both husbands and wives. The signifi— cance of the relationship between well—being and satisfaction with the spouse's helping efforts persisted even when the authors con- trolled for level of stress. But at the same time, the reported help— fulness of spouses was found to be significantly greater under con— ditions of high experienced stress. The authors concluded that marital helping can be viewed as a moderator of the relation— ship between stress and well-being. influencing the degree to which stress will be translated into psychophysical symptom— atology (p. 129). Eaton (1978) reanalyzed some of the data previously collected by Myers et a1. (1975) in New Haven. Connecticut, particularly re— garding the relationship between the availability of social supports and psychiatric symptoms. He identified a number of variables that might indicate the extent of social supports available to each re— 35 spondent (currently married, belonging to clubs or a church, having friends or relatives visit often, going out to visit others often, having a very close friend, and not living alone). These eight items did not constitute a coherent scale, however, and Eaton thus ex- amined their individual relationships to the psychiatric symptom measure. Two variables, being married and living with others, were found to be associated with the relationship between reported stressful life events and psychiatric symptoms. In the case of those persons who were unmarried or who lived alone (the variables are, of course. not independent of each other). that relationship was significantly stronger than in the case of persons who were married or who lived with others. That is, among persons with weaker social supports, stressful life events appeared to have a greater impact on subsequent psychiatric symptoms than among persons with stronger social supports. In another examination of the role of social resources as a moderator variable, Warheit (1979) examined the role of loss- related stressful life events in depressive symptomatology by sur— veying a sample of 517 persons in the southeastern United States twice, with 3 years between interviews. In addition to assessing the extent of depression at the time of both interviews and evidence of stressful life events associated with losses (particularly important theoretically in conceptualizations regarding the development of depression) in the time interval between the two interviews, Warheit also assessed help—seeking behaviors, personal, familial, and social resources. and a number of demographic variables, in— cluding socioeconomic status. While the single most powerful pre— dictor of depression scores at the time of the second interview was depression scores at the time of the first interview, extent of loss— related stressful life events and availability of social resources were also important. Warheit described these relationships as follows: In every instance those in the high—loss category, regardless of the presence or absence of personal resources, had signifi— cantly higher depression scores than those in the low/ moderate—loss group. The data also suggest that the presence of resources significantly mitigates the impact of losses as— sociated with life events (p. 504). Lin et a1. (1979) examined the independent contributions of stressful life events and social support strength to self—reports of psychiatric symptoms in a sample of 170 Chinese Americans in the District of Columbia. These authors viewed social support as sup— port available to an individual through his or her social ties to other individuals, to groups, and to the larger community. The authors used a measure of weighted stressful life events, a self— report psychiatric symptom checklist, and a short scale for the as— sessment of the availability of various forms of social support. In 36 addition, they recorded occupational level and marital status of each respondent. Lin et al. found that the social support measure was the most powerful predictor of psychiatric symptoms. They noted that marital status, occupational prestige, and stressful life events combined to explain about 8 percent of the variance. . . . The variance explained in the dependent variable increases to 12 percent when social support is incorporated into the model (p. 113). Psychiatric symptoms were more commonly reported among 1m— married people, people of lower occupational prestige, those with more reported stressful lii‘ e events, and those with poorer social supports. Availability of social supports and incidence of stressful life events were independent of each other, and thus each could, in theory, be examined in terms of its potential for control or en— hancement as part of a preventive intervention program. Sarason (1980) has examined the nature of social supports as a moderating factor in understanding stress responses in an innova- tive laboratory setting where it was possible to manipulate the nature of social supports experimentally. During a 20—minute period before a difficult anagram task, members of the experi— mental group participated in a discussion designed to create a social support network and a sense of sharing. Subjects who par— ticipated in the group—building discussion and who were high in test anxiety scored significantly higher on the anagram task than high— test—anxiety subjects who were not part of the group—building discussion. In a second experiment in which social support was gen— erated by a brief empathic comment regarding a subject‘s inability to solve the anagrams. similar findings were obtained. Sarason has concluded that "under certain conditions social supports function as a moderator by counteracting undesirable consequences of high anxiety" (p. 26). Social Resources and Physical Illness Well—developed and well—functioning social resources appear to play an important role in moderating the negative effects of stressful life events on physical well—being. Medalie and Goldbourt (1976) examined the role played by anxiety and psychosocial prob— lems in the, development of angina pectoris (a chronic but not life—threatening heart disorder) in a longitudinal 5—year study of 10.000 Israeli male civil servants age 40 and above. The incidence of angina pectoris was fomid to vary by age. by area of birth. and by a number of physiological variables, but at the same time was found to be higher among persons reporting high anxiety. a high level of family problems, and psychosocial difficulties. Anxiety and psychosocial and family problems appear to potentiate the rela— 37 tionships found between physiological factors (such as electro- cardiographic changes, cholesterol levels, systolic blood pressure) and angina. The risk of developing angina pectoris was found to be about 20 times higher among persons with all risk factors present than among those with none present. One additional variable was assessed—the perceived love and support of one‘s wife. The authors found that with high levels of anxiety, the incidence of angina was significantly reduced in the case of those persons who reported having a loving and supportive wife. Medalie and Goldbourt concluded that preventive measures, like antismoking, reducing cholesterol and blood pressure levels and weight, will probably help to reduce the incidence of myocardial infarction and, to a lesser extent, angina pectoris. But . . . no matter how well this is done, the major sources of risk for angina pectoris will be missed, unless it is accompanied by a detailed investigation of the subject's personal, family, and occupational life situations. The latter will help the physician to assess the strengths and stresses in order to help the patient solve or adjust in a more satisfactory way to recurrent problems and thus reduce his anxiety (p. 918). Lynch (1977) has gathered an impressive array of both statistical and clinical data to support his assertion that the lack of social support systems can lead to a variety of disorders, particularly to coronary artery disease. In his words, the lack of human companionship or the loss of a loved—one can have serious effects on our physical and mental well— being. In terms of cardiovascular disease, the evidence is remarkably consistent . . . human companionship appears to play a vital role in the healthy functioning of the heart (p. 87). Social Resources Among the Elderly A number of studies have pointed to the special importance of social resources in the case of the aging population. Lowenthal and Haven (1968) examined the role of strong personal relationships in the maintenance of health and psychological well—being during middle age and senescence. They postulated that having an inti— mate relationship, a confidant, can serve as a buffer against age— linked social losses. The sample in this study consisted of 280 per— sons in San Francisco, aged 60 or above, interviewed yearly on three occasions. Assessments were made of number of social roles. level of social interaction, morale, and psychiatric impairment. Low social interaction was found to be associated with poor morale; reported social losses in the past year were generally asso— ciated with poor morale as well. The presence or absence of confi— 38 dant had a dramatic attenuating effect on these relationships. Lowenthal and Haven f ound, for example, that if you have a confidant. you can decrease your social inter— action and run no greater risk of becoming depressed than if you had increased it. Further, if you have no confidant and retrench in your social life, the odds for depression become overwhelming (p. 26). An additional finding was that an individual who has been widowed within 7 years, and who has a confidant, has even higher morale than a person who remains married, but lacks a confidant . . . . Among those having confidants, only 10 percent more of the widowed than of the married are depressed, but nearly three—fourths of the widowed who have no confidant are depressed, compared with only about half among the married who have no confidant (p. 27). In addition to these findings, the authors provide some very in— formative data on demographic characteristics of persons who do have confidants (e.g., women are more likely than men to have confidants) and on the characteristics of these confidants (e.g., most often spouses, friends, or children; rarely siblings or other relatives). These authors clearly view having a confith as perhaps the most important type of social support that might be available to an aging person. and speculate that part of the reason for the greater life expectancy among older women in contrast to older men is their greater capacity and willingness to develop and sustain intimate relationships. In another study that focused on the role of social supports in an elderly population, Abrahams and Patterson (1978—79) surveyed a sample of 445 persons age 65 and above in a New England town, viewing aging as a stressful life event. Their survey was based on a structured interview conducted in the respondent's home. In the survey, they sought to assess the prevalence of various forms of psychological impairment and to contrast psychologically healthy and impaired persons in terms of demographic characteristics and social interactions. On the basis of these analyses, the authors sought to identify some of the factors that result in excess vulnerability to the stresses of aging. A total of 8 percent of the study population reported being severely depressed. A total of 17 percent of the sample had one or more symptoms of impairment. Impairment was not found to be associated with age or sex, but was found more commonly among persons living in deteriorated neighborhoods and among less well— educated persons. Healthy aged persons were found to have signifi— cantly stronger social relationships outside of the household, to be 39 significantly higher on social initiative, and to have significantly more varied daily lives. As for the factors that appeared to make the aged more vulner— able to that stress, the authors identified loss of significant others, reduced physical vitality, and unhappiness about retirement from work. Both impaired and healthy persons in this study tended to make very little use of community services and did not differ in their help—seeking behavior. The authors suggest that preventive measures such as increasing community support systems and in— creasing lcnowledge about the need for neighborhood or peer sup— port groups could be helpful in reducing vulnerability to the aging process. Summary This chapter has examined some of the research concerned with factors that appear to attenuate the effects of stressful life events. Unquestionably, the single most important factor is the availability of social supports or a helping social network for persons undergoing life stressors. Second in importance is what might be called social competence or coping ability and style. Both factors lend themselves to preventive programing. In the first instance, it would be important to examine the consequences of strengthening the extent or availability of social support networks for people undergoing life stressors. In the latter instance. it would be important to examine the consequences of an educational ap- proach designed to increase the social competence and improve the coping styles of at—risk persons. 40 Chapter 4 METHODOLOGICAL ISSUES IN THE STUDY OF STRESSFUL LIFE EVENTS In this chapter, we will identify the major questions that have been raised by social scientists who have attempted to put the study of stressful life events on a methodologically sound founda— tion. We will examine a number of specific methodological issues in detail and. in addition. will try to provide a general analytic over— view of criticisms that have been raised about research in the field of stressful life events. Rabkin (1980) has reviewed some of the major methodological difficulties in the study of the possible effects of stressful life events. Among these difficulties, she mentions: 1. The use of retrospective study designs 2. Insufficient attention to the selection of subjects and control groups 3. Lists of stressful life events that are inadequately defined or conceptualized or insufficient in scope 4. Inability to distinguish the onset of illness from help—seeking behavior 5. Inadequate statistical procedures Rabldn noted that stressful life events may vary not only in their perceived severity, but that they may be anticipated or unantici— pated, familiar or novel. desirable or undesirable. sudden or gradu— al, discrete or prolonged. variously under the control of the person being studied. and part of or separated from the prodromal signs of impending illness (see also Miller and Ingham 1979. pp. 318—324). In a parallel paper. Cleary (1980) has identified a number of methodological problems in the design and reporting of stressful life event research. These problems include: 41 1. Failure to describe important characteristics of the sample being studied so that replications are possible 2. Failure to employ a carefully chosen list of items so that no major events are missed and events are disentangled from their presumed consequences 3. Failure to standardize the conditions for the collection of stressful life event data in order to increase reliability 4. Failure to check on the validity of stressful life event reports 5. Failure to specify how scale values have been calculated 6. Failure to specify time intervals within which stressful life events are to be identified and time intervals between the alleged stressful life events and the subsequent dependent measures of adjustment 7. Failure to use an identified and appropriate rationale in the selection of dependent measures 8. Failure to employ an undistorted base rate for life events or for undesired outcomes in the study of stressful life event predictors of maladjustment 9. Failure to employ appropriate statistical procedures in de— scribing and analyzing data that have been collected Considerable criticism has been directed at the methodology that was employed in the original Holmes and Rahe Social Read— justment Rating Scale. Dohrenwend et a1. (1978) have examined some of the criticisms and have described some proposed solutions. Regarding the issue of weighting stressful life events versus simply counting them, Dohrenwend et a1. believe that the preponderance of research to date has found more significant results with using weighted rather than unweighted measures of stressful life events. A second issue that has generated some criticism has been the practice of according each stressful life event its own specific and invariable importance rather than soliciting judgments of importance by the person being studied. Dohrenwend et a1. believe this to be an appropriate methodology in view of the research interest in assessing the importance of stressful life events as environmental inputs unconfounded by predispositions or outcomes. In their efforts to improve the study of the effects of stressful life events, Dohrenwend et al. have enlarged the list of such events. have defined each of them less ambiguously, and have dimensionalized them according to whether the events may or may not be part of the prodromal signs of illness, according to their 42 desirability, and according to the frequency with which they might be expected to occur throughout a variety of sociocultural settings. In addition, Dohrenwend and her colleagues (1978) have developed a procedure for scaling such events in terms of required readjust— ment of change. While their results are admittedly imperfect, Dohrenwend at al. hope that their proposed instrument, the Psychi— atric Epidemiology Research Interview Life Events Scale, will lead to a "general and permanent methodological gain in studies of stressful life events" (p. 228). Four specific methodological issues that are included in the critiques just reviewed have been examined in enough detail to warrant special considerations. These include: 0 The assessment of stressful life events in children 0 The interpretation of desirable and undesirable stressful life events 0 The evaluation of subjective judgments of stressfulness O The importance of prospective studies Stressful life Events in Children Coddington (1972a) had undertaken a series of studies designed to examine the clinical belief that children's illnesses frequently follow stressful life events. In order to evaluate this hypothesis. a children's version of the life events inventory needed to be devel— oped. Coddington chose to model his version after the methodology developed for adults by Holmes and Rahe (1967). Coddington pre— pared four versions of the questionnaire—one for preschool chil— dren. a second for elementary school children, a third for junior high school children, and a fourth for those in senior high school. Ratings of each stressful life event item in terms of the judged amount and duration of change that would result were made by a sample of teachers, pediatricians, and mental health workers. Coddington found very high agreement among the three groups of judges, with rank order correlations never lower than .85. and he was able to construct four age—specific stressful life events inven— tories, in which each event was weighted in terms of its judged severity. In a subsequent study. Coddington (1972b) applied the scales to a representative sample of 3,620 children in Columbus, Ohio. to establish expected norms and to examine questionnaire responses in relation to the demographic factors of sex. race, and socioeconom— ic class. Coddington found no difference in stressful life event scores by sex or by socioeconomic class. However, scores were 43 significantly related to age. generally increasing with increasing age. The regular increase in scores with increasing age was found in every demographic subgroup. In a third study (Heisel et al. 1973), the life events inventory was applied to a sample of 34 children with juvenile rheumatoid arthri— tis, 35 hemophilic children. 32 general pediatric inpatients. 31 children undergoing surgery for hernias or appendicitis. and 88 children seen at a child psychiatric outpatient clinic. In comparison with age—matched normal controls, these children. with the excep— tion of the hemophilics. had had two or three times as many and as severe stressful life events during the year prior to the onset of their medical conditions. There was no difference among the vari— ous patient groups. These authors concluded that stressful life events are diagnostically nonspecific and that "the child required to face major changes in his environment must adjust his internal milieu both psychologically and physiologically, or. more likely. both" (p. 122). Sandler and Ramsay (1980) factor analyzed the responses of a group of 10 clinical child psychologists who had been asked to judge the degree of similarity between pairs of stressful life event items in a list of 32 such items. From this factor analysis, Sandler and Ramsay were able to identify seven sets of items that appeared to tap different content categories—losses (e.g., death. divorce, or hospitalization of parent or friend). entrances (e.g., addition to the family, moving to a new house), family troubles (e.g., worsening of parents' mood. loss of job by parent, arguments between parents). positive events. primary environment change. sibling problem. and physical harm. Subsequent study of the relationship between scores on each of these categories of stressful life events and adjustment ratings as assessed by parents revealed that the items comprising the entrance and family troubles clusters were significantly related to adjustment ratings in a sample of 104 inner city children attend— ing kindergarten through third grade. In both cases. maladapting children scored significantly higher than control children on the stressful life event cluster scores. These studies serve to demonstrate how a new field of social inquiry can develop in a reasonably orderly manner. and can serve as a meful model for how similar studies of life stress in other population-s, e.g., the elderly, college students, newcomers to a community. can be undertaken. Desirable and Undesirable Stressful Life Events A immber of studies have dealt specifically with the question of desirability or mdesirability of streast life events as an approach to the determination of What type of event is most closely related to negative consequences. Klassen et al. (1974) were able to deter— 44 mine that there was generally good agreement in a community sample regarding the desirability or undesirability of various stressful life events. A total of 190 persons in a probability sample of adults in Kansas City, Missouri, judged whether each of 40 events in a stressful life event inventory was "a good thing or a bad thing for most people." A total of 70 percent or more of the par— ticipants judged 19 of the items to be desirable events; 70 percent or more of the participants judged 16 of the items to be undesir— able. Thus, according to the authors' definition of 70 percent as the cutoff criterion of unambiguity of judgment. only 5 items were viewed as ambiguous in terms of their desirability. These items included: retiring from work; being released from prison or acquitted of other than a minor traffic off ense; entering the armed services; having a new person move into the house, other than a newborn child; and having a family member leave home. Mueller et a1. (1977) examined their own survey data collected in Sacramento to consider the question of whether all stressful life events or just tmdesirable ones are related to self—reported meas- ures of well—being. Their analysis suggested that the frequency of undesirable stressful life events was substantially more highly correlated with well—being than were measures that combined desirable and undesirable stressful life events. In addition, the authors noted that regardless of whether desirability is defined by respondents or independently, the relationship of desirable events to psycho- logical status appears to be weak or negligible. Undesirable events, by comparison. are strongly linked with measures of psychological impairment (pp. 314—315). Finally. the reported relationships appeared very similar whether stressful life events were summated in weighted or unweighted form. Vinokur and Selzer (1975) examined the question of whether undesirable stressful life events have greater consequences than desirable events in terms of subsequent physical and psychological well—being. The original conceptualization of a stressful life event was one that required significant readjustment or change regard— less of its desirability. Vinokur and Selzer argued. however, that desirable events might actually reduce stress. Furthermore. since many events (e.g., moving to another city) may vary among people in terms of their perceived desirability, it is necessary to deter- mine the degree of desirability for each stressful life event on an individual basis. These authors studied a sample of over 1,000 male automobile and truck drivers to learn more about the effects of stress and personality variables on traffic accidents. A traditional stressful life events scale was modified so that desirable and undesirable events could be distinguished from each other. For example, the 4S original item "Changes in working conditions" was modified into two items—"Improvement in working conditions" and "Deteriora— tion in working conditions." Furthermore, respondents were asked to rate the degree of stress or pressure evoked by each event and to indicate whether each event was desirable or undesirable. Dependent measures included assessments of aggression, paranoid thinking. depression. suicidal proclivity, physical anxiety and tension responses, amount of drinking, and involvement in traffic accidents. Finally, subjects completed a scale that was designed to determine the extent to which perceived social desirability of questions had an influence on their answers. Vinokur and Selzer calculated the number of events checked by each participant as having occurred during the preceding year, the weighted total of these events based on the traditional stressful life events scale. and the perceived impact of these events as judged by each participant. These totals were calculated separately for the events judged desirable and for the events judged undesir— able. Finally. a series of difference scores were calculated for each subject, by subtracting the total score for desirable events from the same score for undesirable events. The three types of scores (number, weighted total, and perceived impact) turned out to be so highly correlated with each other (all correlations were above .90) that there were no differences be— tween them in terms of their relationships with the dependent variables. The authors found that the scores based only on undesir— able events were consistently more closely related to their depend— ent measures than were the scores based on desirable events or on the difference scores. The measure of social desirability had no effect on their obtained relationships. Finally, the authors indi— cated that weighted scores that reflect the amotmt of change or readjustment judged to be required by each stressful life event were insensitive to the judged undesirability of the event. Since it is their undesirability that is a crucial component in the assessment of stressful life events. self —ratings of readjustment appeared to be more suitable than fixed measures of readjustment derived by others. The authors believe that these relationships are genuine and of a causal nature—undesirable life events contribute to physical ill— ness and psychological impairment. In the authors' words, "the con— tribution of life events to psychological impairment is mediated by stress that is evoked by some undesirable aspect of the events rather than by change per se" (pp. 333—334). Similarly. Crandall and Lehman (1977) found in a sample of 81 undergraduates that report— ed undesirable stressful life events ratings were significantly correlated with reported psychiatric symptoms. In summary, the research seems to suggest that undesirable stressful life events have a more significant impact on subsequent adjustment than desirable stressful life events. Accordingly, it would be appropriate to create stressful life event item pools so 46 that level of desirability would be relatively unequivocal in each item, and in such a way that scores could be separately calculated for desirable and undesirable stressful life events. Subjective Judgments of Stress Another issue of concern has been how to weigh the severity of individual stressful life event items. Specifically. researchers dif— fer on whether each such item should be assigned a specific and universal weight as originally proposed by Holmes and Rahe (1967) or whether respondents should be asked to judge the impact the event had on them. We have already seen that this issue may be of greater theoretical than empirical importance. in view of the very high correlation between stressful life event scores calculated on the basis of independent weightings and those based on subjective judgments of stress. Dohrenwend (1979) takes an unequivocal position on the need to measure the impact of stressful life events independently of self— reports. He writes that it is incredible that some researchers have advocated scoring the magnitude of life events . . . in terms of subjective ratings by the individuals whose stress experiences in relation to their psychopathology are being studied. . . . This procedure is vir— tually guaranteed to confound the relationship between stress and psychopathology. Typically, for example. psychiatric pa— tients rate such events as more stressful than nonpatients . . . just as persons who have experienced heart attacks after a particular life event or series of life events are likely to rate such events as more stressful than a person who has survived the events without a heart attack (p. 8). Mueller et a1. (1978) contrasted 187 newly admitted mental health center patients and a randomly selected nonpatient sample of 321 persons drawn from the same geographic area in terms of the incidence of reported stressful life events during the preceding 30 days. The life events in the questionnaire were categorized ac— cording to their judged desirability and according to their judged independence of psychological status. Patients reported a significantly greater number of events at the time of entrance into treatment than did the nonpatient control group. This difference was primarily due to the reports of undesir— able events and was not found in the case of events judged to be independent of psychological status. The authors also administered a brief self—report measure of general well—being to both groups and found that, in both the patient and control groups, the fre- quency of reported stressful life events was significantly associ— ated with less positive well—being. The authors concluded that "the 47 difference between patients at admission and nonpatients appears to result from patients reporting more undesirable events of the kind that could be confounded with their current psychological state" (pp. 21—22). It is this very confounding that Dohrenwend is seeking to avoid. Schless et al. (1977) contrasted reported stressful life events during the previous year in the case of 56 newly admitted psychi— atric inpatients with those reported by a matched group of 56 medical and surgical patients and a larger community sample of persons interviewed as part of a general population epidemiological study. Psychiatric patients reported significantly more events than the medical and surgical patient group who, in turn, reported sig— nificantly more events than the community sample. Eight specific events were reported significantly more often among the psychi— atric patients than the medical and surgical group. Of these, the largest differences were in increased arguments with the spouse. increased arguments with the f amily, and a child leaving home. The authors are, however. cautious in proposing a cause—effect rela— tionship. since the two largest differences in reported specific life events may in fact have been symptoms of. rather than causes of. the psychiatric disorder—again a confounding of stressful life events and psychopathology. By way of contrast, Sarason et al. (1979) have developed the Life Experience Survey—a measure that asks respondents to rate a series of potentially stressful life events as either desirable or m— desirable and in terms of the degree of personal impact the event had. Horowitz et al. (1979) have developed the Life Event Ques- tionnaire. a scale designed to make a detailed study of the subjec— tive impact of stressful life possible. These authors proceeded from the belief that what was needed was an instrument that measures the current degree of subjective impact experi— enced as a result of a specific event. With such an instrument investigators can observe individuals over periods of , time following the occurrence of an event. compare subgroups for degree of subjective distress after a particular life event, or contrast life events in terms of their relative impact on different populations (p. 209). The scale these authors developed is used to report the impact of any specific stressful event. The items refer to the subject’s feel— ings during the past week regarding the event. Factor analysis of the 20 items in the scale revealed two clusters of 7 and 8 items, one measuring intrusion (“Images related to it popped into my mind") and the other measuring avoidance ("I stayed away from things or situations that might remind me of it"). The scale has been found useful in assessing the effectiveness of brief psycho— therapy undertaken to treat stress response syndromes, and has been found to differentiate such patients from medical students 48 reporting on their reactions to the first cadaver dissection, with medical students scoring substantially lower on both the intrusion and avoidance scales. Grant et a1. (1976) administered the same stressful life event scale to a sample of 171 psychiatric patients. 181 controls. and 165 relatives of the patients and controls. All participants were asked to assign weights to each event just as had been done in the origi— nal research by Holmes and Rahe (1967). While rank order correla— tions of judged severity of each of the 43 items across the three groups were very high, psychiatric patients consistently attached a greater magnitude to the events than did controls. While patients and controls listed "marital separation" as the fourth ranked event in terms of stressfulness, for example, patients weighted the event as nearly 35 percent more stressful than did the controls. Askenasy et al. (1977) have shown that there are substantial variations in subjective judgments of the severity of stressful life events as a function of educational level and cultural background of the judges. The authors suspect that the near universal consensus inferred from previous re— search . . . is in point of fact a middle—class consensus that is shared by social circles with values and interests in common in many different parts of the world (p. 438). In addition to the assessment of stressful life events by the use of standard weighted event items or by the reliance on self— reported measures of undesirability and impact, Brown (1974; see also Brown and Harris 1978) has proposed a third procedure involv— ing judgments of impact made by independent raters who follow a carefully designed interview format and who do not have informa— tion about outcome or perceived stress. Brown. agreeing with Dohrenwend. argues that it is crucial to be able to assess the importance of a stressful life event independently of knowing subsequent outcome if one is ever to develop a viable theory of how stressful life events may be causally related to illness. In the absence of that careful rating, three sources of invalidating contamination exist. In direct contamination . . . the respondent may report more disturbing life events to make sense of his illness; in indirect contamination. anxiety of the person or some other such trait may lead both to a greater reporting of life events and sub— sequent illness. And in spuriousness the account of life events may be accurate. but the association of events with illness may be due to another factor such as general anxiety. which leads both to a greater tendency to experience distress in respon)se to life events and also a greater illness rate (1974, p. 237 . 49 Brown suggests, however. that in assigning a standard index of importance to each event, an enormous amount of individual inf or— mation is sacrificed. Furthermore, unless each stressful life event is carefully described, each subject can interpret the event differ— ently. Brown believes that a stressful life event can only be under— stood in its social context and that careful interviewing is necessary to appreciate that context. Brown and Harris (1978, pp. 63—99) provide an excellent chronicle of their efforts to develop this strategy further in the study of the causal effects of stressful life events. In their more recent work, Brown and Harris have tried to preserve the objectivity of their methods for the identification of stressful life events while at the same time permitting the assessment by the interviewer of a. wide variety of other aspects of the event. including the meaning and significance the event appears to have had for the person. Tennant et al. (1979) have reported that raters can reliably judge the severity of stressful life events on an individual basis, as Brown has proposed. by determining the unique social context within which the event took place. These ratings are made on the basis of a wide ranging semistructured interview. and the severity of re— ported life events is judged independently of knowing the subject's response or the outcome of the event. This procedure of individu— ally assessing the contextual threat of stressful life events stands in contrast to the more traditional procedure of assigning fixed weights to each event in a finite list. In spite of the high correlations of stressful life event scores calculated on the basis of these two very different theoretical per— spectives. the importance of disentangling causes from prodromal signs of illness or from the understandable retrospective effort on the part of ill people to understand and explain their illnesses seems very persuasive. Accordingly. the effort to develop measures of stressful life event significance that are independent of their judged subsequent negative consequences seems critical if a persuasive theory of cause and effect is to be advanced. The Need for Prospective Studies Starting in 1974. researchers who have examined the state of stressful life event research have been manimous in their asser— tions that the field now needs to inaugurate a series of prospective cohort studies. Epidemiologists have developed two general re— search strategies in examining the possible connections between certain events and subsequent disorders. One research design, called the case—control method, contrasts a group of persons with a particular disorder (the cases) with a matched group of persons without the disorder (the controls) to determine whether a sus- pected antecedent event has occurred significantly more often in 50 the group of cases than in the control group. The other research design, called the cohort method, contrasts a group of persons who have undergone a particular event with a matched group of persons who have not Imdergone the event, to determine whether a sus— pected outcome has occurred significantly more often in the case of the group of persons who have undergone the event than in the matched group of persons who have not undergone the event. MacMahon and Pugh (1970) write: A case—control is usually less costly than a cohort study—in terms of both time and resources—and is therefore frequent— ly undertaken as a first step to determine whether or not an association exists between the suspected cause and effect. . . . Cohort studies may then be undertaken to gain added confi— dence in the existence of a relationship and to measure more accurately its strength (p. 43). Dohrenwend and Dohrenwend (1974) have suggested that one of the central questions yet to be answered is: What is the risk that some form of disability will follow the occurrence of stressful life events? With regard to this question. they stated. Only studies of cohorts who differ with respect to stressful life events provide any information about the magnitude of the risk that illness or disability will actually follow these events, information without which the practical implications of re— search on stressful life events is far from clear (p. 315). In a later publication, Dohrenwend and Dohrenwend (1978) reviewed the direct and indirect evidence relating life stress to illness. These authors concluded that the correlates of stressful life events are not limited to any particular types of disorder. On the contrary. life events have been shown to be related to many somatic disorders including heart disease . . . fractures. and childhood leukemia . . . . to performance deficits among teachers and college stu— dents . . . . and to psychological disorders including acute schizophrenia . . . depression . . . and suicide attempt (p. 8). 0n the other hand, however, they pointed out that current research findings "do not provide a clear picture of the nature and strength of this relationship" (p. 9; also see Rahe 1979, p. 3). Dohrenwend and Dohrenwend (1978) indicated that typical case—control studies do not provide the information that we need to estimate the magnitude of the risk that disorder will follow as a consequence of experiencing stressful life events. To get this information we need cohort 51 studies based on samples of the population of persons who have experienced whatever life events are of interest rather than case—control studies based on samples of persons who have become ill (p. 12). In considering what directions further research should take. Goldberg and Comstock (1976) suggest that further retrospective studies seem impractical both because of biased recall and because of possible interaction between events and diseases, particularly mental illnesses. More in order are well—designed prospective studies using random samples of the population, a range of outcome illnesses. out— come variables clearly distinguishable from independent variables, and an account of the effects of mediating factors (p. 156; see also Dean and Lin 1977). Finally, Paykel (1978) recently reviewed much of the research linking stressful life events with the onset of psychiatric illness and was able to calculate a measure of relative risk based upon an ana— lysis of a series of studies contrasting psychiatric patients with general population controls. Paykel estimated that the risk of psychiatric illness following major stressful life events was on the order of 2—7 times greater than when no significant life events were reported. Risks appeared greater following more stressful life events. and greater for subsequent depression and suicide attempts than for schizophrenia. Perhaps as important as this summary statement, Paykel noted that the vast majority of persons under— going stressful life events do not become ill. and that to understand how background factors play a role in determining who does he- come ill, it will be necessary to move away from retrospective studies and toward prospective studies in which persons undergoing some stressful life event are matched with persons not undergoing the event. in order to be able to identify the actual excess risk of illness associated with the event. Because there do not appear to be any stressful life-event— specific diseases or disorders, any case—control study in which per— sons with specific disorders are contrasted with suitably matched controls without the disorder runs the risk of significantly under— estimating the deleterious consequences of any particular stressful life event or group of such events. Only cohort studies can identify all such consequences. Very few prospective cohort studies have been reported in the literature. In addition to Warheit's study (1979) already described. a report on an additional study examining the relationship of changes in social support over time and changes in psychological adjustment will be useful. In this study. Holahan and Moos (1981) administered two surveys 1 year apart to the husbands and wives in a randomly selected 52 sample of 267 families in the San Francisco Bay area. The scales included a number of measures of social supports, negative life change events during the year between the two surveys, and as— sessed depression and psychosomatic symptoms. While the relation— ship between social support and psychological maladjustment was low in an absolute sense, the study did find that decreases in social support in the family and work environments was significantly related to increases in psychological maladjustment 1 year later. The need to inaugurate prospective studies seems to be advanced by virtually every investigator who has examined stressful life events. That is, the issue seems without controversy. Accordingly. it remains only to underline once more the limitations inevitably inherent in retrospective studies and the power of prospective studies to demonstrate how and Imder what circumstances stressful life events may serve to precipitate illness. Summary This chapter has examined a number of important methodologic— al issues that need to be considered when studying the effects of stressful life events. What seems clear is that the examination of stressful life events is a sufficiently new field so that studies that are especially sound from a methodological point of view have the prospect of having an unusually important impact. This review has already noted the importance of identifying unique sets of potentially stressful life events in the case of special populations. At the same time, however. virtually everyone is at risk of having to deal with certain stressful life events. According— ly. to increase the comparability of different data sets, it would be useful to include in uniquely developed stressful life event item sets those items that are ubiquitous in nature and that might apply to virtually all demographically defined population subsets. Also. if the field of stressful life event research is to thrive, every effort should be made to improve the reliability and validity of stressful life event scales. Four different stressful life event scales have been identified in this chapter—the original Social Readjustment Rating Scale. developed by Holmes and Rahe (1967); the Life Experiences Survey, developed by Sarason et al. (1979); the Psychiatric Epidemiology Research Interview Life Events Scale, developed by Dohrenwend et al. (1978); and the Life Event Questionnaire. developed by Horowitz et al. (1977). In a recent study, Kale and Stenmark (1983) contrasted these four stressful life event scales in terms of the correlations of scale scores with a dependent measure of psychological adjustment. They found that the Life Event Questionnaire was a ”significantly better predictor of adjustment than the other three scales” (p. 441). par— ticularly in the case of women. Other studies contrasting the vari— 53 ous stressful life event scales that are available need to be under— taken to arrive at rational conclusions regarding the circumstances in which a particular stressful life event scale should be used. Finally. the validity of reports of stressful life events in retro— spective studies has been thrown into question in a number of studies. In addition to the strategy of undertaking prospective studies, it might be useful to solicit several sets of stressful life event reports in the case of the same study participant. In some cases, as, for example. when the participant is a psychiatric or medical patient, it might be useful to solicit reports from friends or from family members to identify stressful life events that the study participant might have overlooked. In other cases, it might be appropriate to solicit reports from both husband and wife to collate them with the ultimate objective of increasing the validity of the reports. 54 Chapter 5 EXAMPLES OF PREVENTIVE INTERVENTION PROGRAMS A number of preventive intervention programs based on the stressful life event paradigm will be described in this chapter. As stated earlier, the programs eligible to be included in this chapter are limited by a fairly narrow definition of stressful life events as those events that are more or less contemporary in nature. There is a striking parallel between the published research dealing with preventive intervention programs related to stressful life events and the early research examining psychotherapy out— come, in that many of the preventive intervention studies currently being reported are primarily descriptive rather than empirical in character. A large number of publications deal with preinterven— tion analyses of specific stressful life events and with the con— ditions under which these events appear to lead to negative out— comes. These reports set the stage for preventive intervention. An equally large number of studies in the recent literature provide interesting clinical descriptions of preventive intervention pro— grams with little information of an empirical nature as to program effectiveness. Such studies do not always have control or com— parison groups, do not always include a quantitative approach to the assessment of program effects, and rarely have extended followup interviews to assess the duration of program impact. Finally, there are a growing number of published reports of pre— ventive intervention programs that seem both programmatically persuasive and methodologically sound. Were a very stringent set of criteria to be applied in deciding which programs to include. the chapter would fail to be repre— sentative of the efforts that constitute the bulk of the literature. On the other hand, including indiscriminate reference to too wide an array of studies runs the risk of failing to differentiate among program reports that vary in the quality of their methodology. In this chapter. we will take a midground position by diff erentiating, wherever indicated. between experimental preventive intervention programs that are not only of clinical importance but are also methodologically more sophisticated and rigorous and what will be called clinical studies that appear promising but relatively weaker from the point of view of evaluation research. 55 Stresses Associated With Parenthood Experimental Preventive Intervention Programs Stressful life events Sin-rounding childbirth and early parenthood and childhood have attracted the attention of mental health pro— f essionals for the past two decades or more. Early preventive in— tervention studies tended to be primarily descriptive in nature. More recently. reports of experimental preventive intervention programs have begtm to appear. Minde et al. (1982) examined the birth of a premature infant as a stressful life event. They noted existing documentation of the greater difficulties experienced by parents of such babies (Kaplan and Mason 1960) and briefly re— viewed attempts to investigate this phenomenon. Previous studies have apparently failed to interpret their results as suggesting a connection between changes in maternal caretaking and improved maternal self—esteem or autonomy. Minde et al. also critiqued past experiments for their provision of either emotional or practical support, whereas both seemed necessary. They, therefore, designed a study which would offer both types of assistance within a self—help format. Minde et al. chose their sample from among the parents of rel— atively healthy but premature infants admitted to the Neonatal Intensive Care Unit of The Hospital for Sick Children in Toronto. The experimental groups consisted of 4 to 5 families. In addition, a veteran mother who had recently given birth to a premature baby acted as group facilitator. and an experienced neonatal nurse served as coordinator. The sessions initially focused on the sharing of fear, guilt, and depression and then moved into more didactic agendas. These agendas included the identification of the devel— opmental and medical needs of premature infants. Time was also devoted to practical tasks (e.g., f'mding babysitters) as well as to the familiarization of group members with community resources. Attention was directed to such issues as working mothers. the role of the father, and parents' emotional availability to their infants. The group coordinator answered specific questions and offered individual support between meetings. an opportunity for assistance used by approximately 60 percent of the parents. Control group parents received routine ward care. Data earlier reported by the authors (Minde et a1. 1980) sug— gested that the experimental mothers were more socially stimulating with their infants and shared their feelings with them more easily. Their children reciprocated by showing more social and inde— pendent behaviors such as general playing, food sharing, and self—feeding (Minde et al. 1982. p. 245). One year later, experimental group mothers showed "a higher 56 degree of personal autonomy and positive changes in their rela— tionship with other people" (p. 254). The authors speculated that "the gain in autonomy. . . may allow a woman to utilize her interpersonal abilities and sensitivities with more ease and less conflict and in this sense make her into a more caring mother" (p. 254). This interpretation was viewed as consistent with the increased social behavior. as well as the ability to demonstrate affection and concern. observed among these mothers. Minde et a1. (1982) concluded that the self—help groups constituted one practical application of Engel's biopsychosocial model [1980] . . . the groups allowed care to go on at various systems levels and converted a poten— tially very disorganizing and disintegrating experience into one which taught these parents competence and self—reliance (pp. 255—256). These levels included not only the medical care that was accorded to the infants during hospitalization but also the emotional support represented by the opportunity for mourning. information sharing, and general ventilation offered by the groups. The authors also suggested that the combination of emotional and educational sup— port gave parents a sense of mastery, one modeled by the veteran mothers who bridged the gap between professional and patient. They hypothesized that "the increase in the group mothers' auton— omy may be a direct result of our systems—oriented treatment program, although a replication of our study in another setting is clearly warranted" (p. 256). In Israel. Weingarten and Feuchtwanger (1978) examined the effect of weekly discussion groups on women's experiences of the birth of a new baby. Led by the family doctor and a social worker. these groups met for 2 months and included 10 patients with in— fants up to 4 months old. Ten mothers who declined an invitation to attend these meetings served as a comparison group. The authors underscored the stresses presented by the changes in family dy— namics associated with childbirth and connected these changes with the frequency with which new mothers visit doctors. They hypothesized that "if group discussions help mothers to cope with some of their doubts, the need for individual consultation should lessen" (p. 232). Members generated the topics for group discussion. These topics included the role of the husband in child care. the mother's will— ingness and ability to meet her infant's needs, the presence of ambivalent feelings toward the baby. separation issues, relation— ships with the mother's mother and mother—in—law. and the place of children in the family. The authors found that during the tenure of the group. partici— pating mothers brought their babies to the doctor less frequently than did nonparticipating mothers. Yet. Weingarten and Feucht— 57 479-088 0 — 85 — 3 wanger cautioned against overgeneralizations drawn from these data as the sample was small, the comparison group was not strictly comparable to the intervention group, and mothers may have used medical services outside the clinic involved in the study. The authors also noted, however, that consultation rates among the experimental and comparison group members differed only during the study period, suggesting "that some need for individual con— sultation may have been fulfilled by the group meetings" (p. 241). Weingarten and Feuchtwanger concluded that "whereas the baby is the overt focus of individual clinical consultations. some of the real needs of the mother-baby pair were satisfied in the mothers' group" (p. 241). The authors then suggested that family doctors adopt an active role in health education, perhaps via organizing group meetings among patients with common concerns. Schaeffer et a1. (1981) assessed the impact of two types of in— tervention on children‘s placement in foster care. After conducting a pilot project, the authors mdertook a major research effort designed to: 1. Ascertain the differential efficacy of very brief crisis in— tervention compared with more extensive short—term crisis treatment 2. Measure the effect in terms of duration of stay in first foster—care placement. number and duration of subsequent placements, duration until return to biological parents when appropriate, and duration of stay with biological parents 3. Ascertain cognitive, social. and emotional changes in the subjects 4. Develop an epidemiological profile of the conditions that lead to foster placement 5. Ascertain the effects of parental pathology and the interval between foster—home placement and the initiation of inter— vention on foster—care experiences The study participants, drawn from New York State, were im- mediately referred for the brief form of intervention. and both biological and foster parents were included whenever possible. This treatment approach consisted of a maximum of 10 sessions with the project's clinical staff. These sessions involved initial psycho— logical examinations and recommendations to biological and foster parents. Two—thirds of these families were then randomly assigned to the group receiving the more extensive treatment. These re— spondents received a maximum of 20 individual crisis-oriented therapy sessions for the children and 10 guidance sessions for the biological and/or foster parents. Meetings with the children gener— 58 ally focused on mourning and adaptation. At the 6— and 12—month points, children were given new psychological examinations, and interviews were conducted with biological and/or foster families and with the mental health personnel involved in the cases. Median IQ for the subjects was 100; no child was discovered to have a primary diagnosis of mental retardation. Moreover. these scores seemed to be lowered somewhat by restricted experiences, emotional deprivations, and developmental deficits. Evidence of conflicts and symptoms deriving from stressful situations was uncovered, but the authors also asserted that they were "impressed with most of the children's resiliency and drive to grow, which was usually evident simultaneously with reactions of narcissistic hu— miliatio " (p. 53). Although followup data were not sufficient to test the hypotheses under consideration, Schaeff er et al. noted that 29 children from 15 families left the foster—care system to live with biological relatives; 6 of these received the brief treatment, and 20 of these received the more extensive version. In addition. 10 children moved from their original foster home to another. Five of these received the brief intervention, while 5 received the more extensive treatment. Despite recent trends limiting permanency planning in foster care to return to biological family or adoption, Schaeffer et al. (1981) suggested that long—term foster care might be desirable in some instances. These included cases in which the biological tie can be neither severed. due to strong emotional connections or older age, nor consummated. due to the immaturity or problems of the mother. The authors planned to extend the followup period by 18 months owing to the placement changes that tended to occur at the 12— to 18—month point and to then engage in extensive sub— sample analysis. They ended this interim report with the conclusion that data from the current study will serve to help build a body of knowledge about the value of preventive intervention strate— gies. particularly brief situational crisis interventions, in meeting the needs of foster children and their families (p. 56). Clinical Studies Friedman and Cohen (1980) reported on a peer support group focused on the emotional effects of miscarriage. Led by the authors (a psychiatrist and a health educator), the group consisted of five womenand met for eight sessions. The project was designed to serve both therapeutic and educational functions for the par— ticipants. All members were married and suffered miscarriages of consciously desired first pregnancies during the previous year. This intervention approach was viewed as filling an important gap in services centering on stressful life events. Friedman and Cohen noted that, despite the mourning and grief following a miscarriage, S9 support from medical and mental health professionals is often absent. In this way. a contradiction between the woman's subjec- tive experience of loss and the environment's dismissal of the miscarriage as unimportant was seen as established. Although no statistical analyses could be undertaken. the authors concluded that "the group successfully carried out its task of helping members cope with the loss of a pregnancy through mis— carriage" (p. 45). Indeed, all participants had been either severely or mildly depressed when the group began but evidenced a marked reduction in symptomatology at termination. At the 6—month followup, all members were pregnant again. The participants per— ceived the group as an effective way to deal with grief and dis— appointment. obtain and off er comfort and encouragement, correct distortions. develop insight, and cultivate active coping strategies. The transition to parenthood can be viewed as a stressful life event, particularly when it involves a cesarean procedure. Lipson (1982) employed both participant—observer and individual interview techniques to investigate the functions and effects of two support groups comprised of women who had undergone cesarean sections during the past 3 years or who were currently pregnant and ex— pected to deliver via this route. Her rationale for this project can be found in studies exploring women's experience of cesarean childbirth. Indeed. women perceive cesarean childbirth as a less positive event than vaginal delivery. Cultural factors and current trends in childbirth have come to place value on natural methods and on a woman's responsibility for her birth performance. More— over, the physiological strain of pregnancy and delivery can com— bine with the experience of surgery to precipitate a major situ— ational crisis. Lipson noted that almost 50 percent of cesarean mothers suffer from one or more surgical complications, some quite severe. These factors increase the risk of psychological trauma in a cesarean mother. The negative impact can linger for months or years. indicating a lengthy and difficult assimilation process. The cesarean procedure has been linked to delayed mother—infant bonding, problems in the spousal relationship. and limitation of family size. It has also been statistically associated with higher rates of child abuse. Lipson identified three main potential trouble areas for cesarean mothers: 1. Problems in self (e.g., unexpected memories of the proce— dure, strong negative feelings about the birth, post partum depression, unmet expectations for childbirth. and self— esteem and body image alterations) 2. Relationships with children (e.g., delayed bonding, negative feelings toward the infant, concerns about the effect of the cesarean procedure on infant's health and on the likelihood of child abuse) 60 3. Relationship with the spouse (e.g.. sexual problems) Lipson (1982) noted improvements in all three areas among members. She traced this progress to the functions characterizing cesarean support groups and concluded that such groups provide a number of valuable services. This combination of services encourages women to work through feelings and fears asso— ciated with a past cesarean birth. In many cases. a good second cesarean provides a corrective experience for a woman who may have spent months or years struggling with Impleas— ant feelings or memories about the first birth. and concerns about its effects on her infant and spouse. And finally. groups allow women to prepare for and maximize their chances for more positive experiences in the future (p. 28). In this connection, the author underscored the groups' contri— bution to the emotional preparation for cesarean deliveries (e.g.. cleaning up "old business") as well as to the information—gathering procedure (e.g.. awareness of such options as having the father present in the delivery room and engaging in self—care activities during the recovery period). Lipson also pointed out that group members tended to become "more confident and assertive health care consumers" (p. 27). The transition to motherhood was viewed as a stressful life event by Silverman and Murrow (1976) in their investigation of the La Leche League (LLL), a mutual—help organization made up of nurs— ing mothers. LLL consists of local chapters that provide a forum to share experience. information, advice. and support concerning breastfeeding and mothering. It grew from the difficulties involved in obtaining such services from pediatricians and family members. Each chapter generally includes a leader and 10 to 30 mothers. Silverman and Murrow examined the effects of LLL by inter“ viewing four leaders and five members of urban and suburban groups. They identified 11 main areas in which LLL assisted mothers: Provision of an accepting, personal approach Preparation for the birth of the child Handling problems in the hospital Understanding the period of exhaustion following the birth of the baby 5. Preparation and support in facing the difficulty of the sudden, full—time task of motherhood Physical preparation for nursing Dealing with the physical problems of the mother 5‘5”!"2" 61 8. Dealing with the physical problems of the baby 9. Addressing the social—emotional problems of the mother 10. Handling the behavior of the baby 11. Dealing with generalized feelings of guilt and anxiety Kagey et a1. (1981) investigated the impact of participation in mutual support groups for parents of newborns. They examined the efforts of the Perinatal Task Force organized by the Greater Lynn, Massachusetts, Community Mental Health Center‘s Primary Pre— vention Team. This task force developed 28 support groups led by trained volunteer facilitators who were also parents. Each group comprised five to eight members and focused on such areas as child development, women's issues, and intrafamilial relationships. The authors conducted their study 2 years into the operation of the program by mailing questionnaires to members. More than 9 out of 10 endorsed the overall helpfulness of the groups and said they would recommend them to a friend. The groups were also viewed as helpful in facilitating adult social contact. increasing positive valuation of the parenting role, decreasing social isolation in the parenting role, enhancing understanding of child develop— ment, and increasing child—caring skills. On the other hand. less than one—fourth of the respondents reported improvement in their spousal relationships. The authors suggested that this last finding may reflect either the relative lack of attention paid to marital relations or the possibility that, in some cases, the groups had a negative effect on such relationships. Crissey (1977) reviewed data collection by the Board of Control of State Institutions and the Child Welfare Research Station of the State University of Iowa to examine the effects of adoption on the development of adoptees. Spanning 40 years, the studies under examination focused on four groups of children who were adopted by nonrelatives. The first group (13 children) was drawn from low socioeconomic families deemed inadequate; natural mothers had 10s of 70 through 79 and were described as dull and incapable of child—rearing. These children were placed in higher socioeconomic status homes before they were 6 months old. The family back— grounds of the second group (11 children) were judged to be even more inferior than that of the first; their natural mothers were identified as retarded. with 103 in the 53 to 67 range. Placement into adoptive homes was made before they were 6 months old, but although these placements generally represented improvements over the original settings, some variation in quality was noted. The third group (13 children) came from problematic family backgrounds, including some incidence of psychosis and low 10s. The children were themselves identified as retarded in infancy. They were placed in planned environments during infancy and pre— 62 school and then returned to an orphanage when the maximum benefit of the special program was achieved. Most were later sent to adoptive homes. The fourth group (12 children) was also born into disadvantaged families, with maternal IQs varying from 36 to 85. These children were identified as normal in infancy, but they remained in an orphanage due to minor problems preventing early adoption. By the time these were solved. the children no longer scored as normal in mental development. Followup information indicated an absence of major maladjust— ment among adults who had been placed into Group 1 as children. Most Group 2 members were "at least reasonably well adjusted" as adults (p. 195). As adults. none of the Group 3 subjects had "eX— perienced discernible mental health problems and none . . . had social or emotional problems beyond those encountered in everyday living" (p. 198). Group 4 children, on the other hand, grew up to"fit the classical stereotype of the mentally retarded, minimally skilled, unemployed or unemployable individ " (p. 199). With more than three million births taking place each year in the United States, continued interest in identifying stressful life events associated with parenthood and early childhood will surely con— tinue. The clinical base for the development of methodologically sound preventive intervention programs seems strong enough so that one might hope for an increasing number of sound experimen— tal studies to be reported in the next few years. Meanwhile. pre— experimental studies need to be continued in areas that do not yet have an adequate enough analytic foundation to justify the devel— opment of experimental preventive intervention programs. Physical Illness in Children Experimental Preventive Intervention Programs Another area that has recently attracted a good deal of interest is the stressful character of physical illness in the case of children. Associated with the interest in physical illness has been the grow- ing interest in the medical care system as a source of stress. Among experimental preventive intervention programs that have been reported in the literature, the evaluation of programs de- signed to reduce the stresses associated with illness and medical care are unusually sound and compelling. Noting that the social environment of the hospital may produce stress for child surgical patients and their mothers, Skipper and Leonard (1968) developed a program whereby nurses systematically interacted with hospitalized children's mothers in the hopes that reduced stress on mothers would result in a subsequent reduction in the stress on the children. A total of 80 children between ages 3 and 9, hospitalized for tonsillectomy. were randomly divided into 63 two groups. In the case of the control group, routine hospital pro— cedures were in effect. In the case of the experimental group, from the moment of admission. while routine attention was paid to the child, the focus of interaction on the part of the nurse was with the mother. The nurse attempted to create an atmosphere which would facilitate the commmica— tion of information to the mother, maximize freedom to verbalize her fear, anxiety. and special problems. and to ask any and all questions which were on her mind. The information given to the mother tried to paint an accurate picture of the reality of the situation. Mothers were told what routine events to expect and when they were likely to occur——including the actual time schedule for the operation (p. 277). The interaction was characterized as "expressive, yet affectively neutral. person—oriented rather than task—oriented. nonauthori— tarian. specific (not diffuse) and intimate" (p. 277). Two experi— mental subgroups were identified—one in which the additional interaction was limited just to the time of admission, and the other in which the additional interaction occurred at the time of admis— sion and at other critical times during the hospitalization as well. Preoperative and postoperative differences between the children in the experimental and control groups were dramatic in terms of blood pressure. pulse rate, temperature. frequency of postoperative vomiting, time from surgery Imtil first voiding. postoperative fluid intake, regular nursing staff evaluation of adaptation to the hOS— pital. and postoperative home experiences during the first week after discharge. In every case where a significant difference was found, it favored the children in the experimental group. Where differences were found between the children as a function of the level of interaction between nurses and mothers, the more intense level of interaction yielded better results in the case of the children. . In a similar study, Wolfer and Visintainer (1975) examined the effects of psychologic preparation and supportive care by nurses on mothers' and children‘s reactions to pediatric hospitalizations. Their sample comprised 80 children aged 3 to 14 who had not ex- perienced a previous hospitalization during the past year, did not suffer from any chronic disease. and had no other medical or psy— chological conditions. Five categories of real or imagined dangers were identified, including: The fear of discomfort, pain, or death Anxiety associated with separation from parents and friends Fear of the unknown Uncertainty about limits and expectations Relative loss of control, autonomy. and competence .U‘PP’PE“ 64 Relevant information about procedures was provided in stage— appropriate form to the children and parents during stressful points. Information, instruction, and support designed to minimize stress and facilitate coping on the part of the child were provided by one nurse who was present at all critical times during the hos— pitalization. Regular nursing care was provided to the control group. Both behavioral and physiological dependent measures were used, including ratings of the children‘s emotional state. cooperative— ness, ease with which fluids were taken postoperatively, and of recovery room upset as reflected in medication decisions. Pulse rates and time to first postoperative voiding were also measured. Adjustment after discharge was measured via a Posthospital Be— havior questiormaire developed by Vernon et al. (1966). Using this instrument, the parent compared the child's pre— and posthospital behavior in terms of six factors: 0 General anxiety and regression 0 Separation anxiety 0 Anxiety about sleep Eating disturbance Aggression toward authority Apathy—withdrawal Using a questionnaire administered prior to discharge, the par— ent's experience was assessed in five ways: Manifest upset Coping/cooperation rated upon admission Anxiety Satisfaction with care Adequacy of information received The authors found that the experimental group suffered from less upset and fewer cooperation difficulties during hospitalization than did the control group. They also reported fewer posthospital problems. This analysis rested on data showing that experimental group children demonstrated greater ease of fluid intake. less time to first voiding, lower heart rates, lower incidence of resistance to anesthesia induction, and better postdischarge adjustment scores. Moreover. experimental group parents emerged as less anxious and more satisfied. Wolfer and Visintainer (1975) suggested that "these results seemed to provide strong support for the beneficial effect of systematic preparation and support for hospitalized children and parents" (p. 254). Clinical Studies Benjamin (1978) examined children's sudden life—threatening illness and subsequent recovery as a stressful life event for both 65 children and parents. The author described the effects of psycho- logical counseling on the families of stricken latency—age children. Such counseling centered on the problems commonly encountered by parents of acutely ill children who subsequently recover. Ac— cording to Benjamin, "the most pervasive difficulty observed has been a sense of helplessness and lack of control experienced by both the afflicted children and their parents" (p. 288). This problem was said to derive from the rapidity of both illness onset and re— covery, which did not allow for the mobilization of appropriate defenses. Escalation of parental and child anxiety often resulted. frequently manifested by overprotectiveness. Meetings with parents were designed to help them attain a sense of continuity with regard to their children. Meeting topics included preparing the families for the behavioral changes probable in the children and the consequent need for setting limits. The author also stressed the development of a sense of parental competence. With children. Benjamin attempted to reconcile fantasy with reality, especially in connection with death imagery. A followup visit was arranged at discharge. Although Benjamin did not report actual statistics (only eight children were involved), she did conclude that our experience with this brief work during the recovery phase of acute illness has been encouraging and suggests that the time invested during the early recovery phase represents a wise investment of clinical time and service (p. 290). With the persuasive evidence that some aspects of being ill and receiving medical treatment are unduly stressful, and that remark— ably little effort is needed to reduce some of those stresses signifi— cantly. it seems likely that research studies examining intervention programs designed to facilitate coping with the stress associated with illness are likely to continue to be reported. If the results of preventive intervention studies continue to be positive. significant permanent changes in certain aspects of the medical care system may well be undertaken as a consequence of these research re— ports. Thus. it can be predicted that one of the most fruitful areas of preventive intervention research based on the stressful lif e event paradigm may be located within the medical care system. Adolescent Role Transitions Experimental Preventive Intervention Programs Early theorizing regarding crisis and crisis intervention identi— fied developmental crises as particularly amenable to preventive intervention. Role transitions take place at predictable moments in lifespan development, and there has been a high level of interest in facilitating these role transitions wherever possible. The past 66 several years have witnessed the beginning of experimental pre— ventive intervention studies targeted at individuals undergoing such developmental crises. Felner et al. (1982) examined the effectiveness of a preventive intervention program that was designed to facilitate the transition from junior to senior high school—a common and predictable stressful life event. Moves from one school to another have often been found to result in decreases in academic achievement, in— creased classroom behavior problems. and heightened anxiety, particularly in interpersonal situations. The Felner et al. project took place during the summer prior to entrance into high school and involved a total of 65 students. A control group of 120 matched nonproject students was also identified. The project was designed to increase levels of teacher and stu— dent support during the transition to high school and to reduce the difficulties of mastering the tasks associated with the transition. All project students were assigned to one of four designated home— room teachers. During the summer, these teachers contacted the parents of their students in order to increase the amount of per— ceived social support available from teachers, to decrease the students' sense of anonymity. and to help students gain access to important information about school expectations and regulations. In addition. the school schedule and organization were revised so that students in the experimental project would take their four primary academic subjects only with other project students. Thus, it was hoped. a stable peer support system could be established. Both project and control students were evaluated during the midyear and again at the end of the school year. Students were evaluated in terms of their self—concepts, their perceptions of the social climate of the school, and their grades and attendance records during both the previous and the current school years. Control students were found to be absent from school signifi— cantly more often in the first high school year than students in the intervention project and were found to have earned significantly poorer grades. While the self-concepts of students in the experi— mental program generally remained stable, control group students showed marked declines in those concepts during their first year in high school. Finally, students in the experimental program reported signif— icantly more positive feelings about the social climate in the school than did students in the control group. In particular. exper— imental students reported higher levels of teacher support. af— filiation. and involvement; significantly more favorable impressions of the personal growth—enhancing aspects of the school environ— ment; and significantly more favorable judgments of how well organized the school was. In virtually all cases, the differences between experimental and control groups were due to declines in the ratings by control students rather than to increases in ratings by experimental students. 67 The authors conclude their report by noting that overall, these findings support the arguments that attempts to understand and modify school environments . . . can be fruit— fully adapted to preventive programs designed to increase people's ability to cope with the adaptive tasks of life transi- tions (Felner et a1. 1982. p. 288). A similarly conceptualized program took place at a college cam— pus ‘and involved assisting a group of 200 freshman students in mastering the stresses associated with the transition from being a high school senior living at home to being a college student living away from home (Bloom 1971). The unusual vulnerability of college freshmen to stresses during the early months of college is well known. While about 5 percent of the student body in the average university seeks psychiatric help each year, the incidence of help—seeking is unusually high among freshmen. The dropout rate is twice as high for freshmen as for seniors. This project developed, carried out. and evaluated a strategy designed to provide mean— ingful interventions to assist freshmen in this role transition. The theoretical constructs that were most useful in concep— tualizing the project and in planning its specific objectives and activities were related to the developmental tasks of adolescence. Review of the literature suggested that students could be helped toward the completion of five major tasks: 1. The development of independence as well as appropriate interdependence 2. The ability to recognize and deal with uncertainty 3. The development of a personal set of values and standards that may or may not reflect the values and standards of peers or parents 4. The development of a sense of sexual identity and of satis— faction with one's own masculinity or femininity 5. The development of mature interpersonal relationships and social skills The project strategy was to develop an ongoing process with a defined group of college students so that these research findings could be fed back to them in a manner they would find growth inducing and stress reducing. It was hoped that this feedback would help the student achieve emotional maturity, adapt successfully to the college community, avoid psychological disability. and stay in school. The process was designed to: 68 1. Provide a sense of membership in a group and thus reduce feelings of isolation 2. Give group members some reference facts with which to compare themselves, thus reducing feelings of uniqueness 3. Provide group members with the opportunity to express their reactions to the university 4. Give group members some intellectual tools by which to better understand the stresses acting on them and their re— actions to these stresses 5. Provide formalized opportunities (in the form of question— naires) for group members to think through their own beliefs) 6. Provide a resource person for the students to talk with in the event of some crisis Information was collected by means of questionnaires that were administered during orientation week and at key times during the year—~after 1 month in college. after Christmas vacation. shortly after the start of the second semester, and just before final exam— inations in the second semester. Progress reports based on analyses of these questionnaire responses were prepared by the author at irregular times and distributed to all of the participating students. Evaluations of the pilot project were generally favorable. In contrast to a comparison group of freshmen. the experimental group had a somewhat higher survival rate. and. of those no longer on the campus, a significantly larger portion of experimental group students continued their academic involvement in other settings. In addition, of students who left the university, a majority of the members of the experimental group continued to live away from home, while a majority of the comparison group returned to their parental home. Clinical Studies Heller (1975) focused on geographic mobility as a stressful life event (see also Asher and Bloom 1983). She investigated the effect on child adjustment of preventive services given to parents who had recently relocated. These services grew out of a 1973 Federal staffing grant'establishing the Staten Island Children's Community Mental Health Center. Group sessions were organized to "provide a practical method of intervention and prevention of problems" (p. 493). The meetings were designed to enhance parents' sensi— tivity to their children's needs and emotions. to increase their abil— ity to deal with general concerns regarding their children, and to diminish everyday anxieties. "The goal was to prevent the need for 69 costly. long—term help that might become necessary if problems were not dealt with immediately" (p. 493). Recruitment occurred via local schools and PTA chapters. Nine groups, each consisting of a leader and approximately 10 members. met for 10 weeks during the study period. Sessions were devoted to child—centered discussions, and an interactive format was employed to facilitate the expression of such individual concerns as sibling rivalry, problematic peer relations, discipline versus punishment, and building trust. Parents requiring more extensive assistance were referred to appropriate agencies. Evaluation was accomplished by questionnaires administered during the final session of each group. Responses indicated that the expectations of most parents were met. The discovery that their problems were shared emerged as the most satisfying aspect of the group experience. Additional benefits centered around the honesty. openness, and support provided by other members, gaining new insights in child development, increased security in the parenting role, and an opportunity for structured ventilation. The major problem cited was that the time passed too rapidly. As a result of this perceived success, the program was solidified and expanded. Maturational crises that occur with role transitions in the normal life trajectory represent usually promising targets for the devel— opment of experimental preventive intervention programs. The potential for such programs has not been reached by any means, and it is hoped that sound preventive intervention programs will be increasingly reported in the literature in the next few years. Specific Adult Stressors Experimental Preventive Intervention Programs A variety of specific adult stressors have been identified as potential targets for the development of preventive intervention programs. A number of these programs will be described in the next two sections. This section will deal with preventive inter— vention programs designed to reduce the stress associated with such events as entering into psychotherapy, physical illness, and marital disruption. Preventive intervention programs that have been designed to be of help to persons undergoing bereavement will be discussed in the next section. Viewing entrance into psychotherapy as a stressful life event, Jacobs et a1. (1972) focused on the problems specific to the treatment of lower-socioeconomic—class patients. They noted both the attendance problems and the slower progress reported for members of this group as well as the fact that these clients may represent a particularly needy population. They also described data showing that lower—class patients were unlikely to be referred for 70 insight therapy and inferred that doctors have "little confidence in the treatment situation" (p. 667). Jacobs et al. suggested that the "culturally determined expectations" (p. 667) of such patients may conflict with the expectations carried by mental health profes— sionals, particularly in the areas of reconstructive work and open— ness. They concluded that "the problem may be viewed as a func— tion of the social distance between middle—class doctors and lower—class patients" (p. 667). The authors identified the first interview as the critical point at which social distance is either established, leading to premature termination, or reduced. Jacobs et al. assessed the impact on subsequent treatment his-- tory of preparing lower—class patients and psychiatric residents by, first, orienting some patients and some psychiatric residents. Then, 120 low—socioeconomic—status psychiatric clinic patients were randomly assigned to one of four conditions: oriented patient to be seen by oriented psychiatric resident. oriented patient to be seen by nonoriented resident. nonoriented patient to be seen by non— oriented resident, and nonoriented patient to be seen by oriented resident. Orientation of patients and of residents required about 15 min- utes and was undertaken by the chief resident. Psychiatric resi— dents were told about the difficulties lower—class patients might have exploring their feelings, accepting the concept of psycho— logical motivation, and tolerating delay in receiving immediate help. Patients were told about how discussing problems with a psychiatrist was different from discussing problems with a family doctor or surgeon. particularly when talking about personal feelings or emotions. Nonoriented patients seen by nonoriented residents were judged to be significantly less likely to profit from insight therapy. were seen significantly more often only for evaluation (without offers for treatment). and reported significantly less improvement than all other groups. Jacobs et a1. (1972) placed their study in the context of the current need for mental health professionals to reach lower socioeconomic groups previously deprived} of psychiatric care. The authors suggested that the results obtained thus far seem to support the idea that even a limited educational procedure may be of considerable value in increasing the motivation and ability to both patient and doctor to work with each other (p. 673). They concluded that these findings tend to support the hypothesis that a procedure that dimin» ishes social distances will increase the frequency of contact of lower—class patients with their doctors, and. concomitantly. that doctors will then tend to view such patients as more ae— 71 cessible to some form of psychotherapeutic intervention (p. 673). Paralleling the interest in physical illness as a stressful life event among children. a number of studies have examined the effectiveness of preventive intervention programs designed to reduce the stressful character of physical illness in adults. Again. just as in the case of children. many of these studies have been concerned with the medical care system as a source of stress. Johnson and Leventhal (1974) investigated the effects of prepa- ration on reactions during a gastrointestinal endoscopic examina— tion-a noxious medical procedure. They began with the assump— tion that threatening situations evoke two types of responses: emotional reactions that include subjective and autonomic signs of emotion and behavior to reduce emotion and... danger—oriented reactions that include awareness of the objec— tive features of the threat agent and behaviors designed to control the degree of potential danger (p. 710). The authors noted that prior research in this area had indicated that these factors are largely independent of each other. Johnson and Leventhal, however. operationalized these variables in relation to two types of preparatory messages: one involving a sensory description of the experience and the other involving a description of appropriate examination behavior. Four groups were thus formed: Members receiving the sensory description Members receiving the behavioral instruction Members receiving both sensory and behavioral messages Members of the control group. who received no preparatory information PPN!" Patients undergoing a gastrointestinal endoscopic examination were assigned to one of these four groups; two observers who were unaware of group assignment charted the relevant information during the patients’ hospitalization. Several dependent measures were used to gauge reactions during the procedure, including amount of tranquilizer administered. changes in heart rate, avoidance movements. gagging, and time required to insert the endoscope. These were seen as reflecting the degree to which danger—controlling behaviors were implemented. Johnson and Leventhal (1974) found a slight reduction in tran- quilizer dosage, in difficulties of stabilization of heart rate. and in prevalence of gagging among those receiving the behavioral in— structions. The group receiving sensory description was given less tranquilizer and showed more stable heart rates than did those in the control group; these differences were. however. restricted to 72 subjects under 50 years of age. Patients in the sensory description group also suffered less gagging but, again. no effect was obtained in relation to time required for insertion of the endoscope. Subjects receiving both behavioral and sensory information exhibited more stable heart rates (among patients under 50). reduced gagging. and an increase in the time measure. This increase was attributed to the attention these subjects paid to an otherwise automatic be— havior. Tranquilizer dosage was slightly reduced among patients under 50 years of age but was increased among those over 50 years old. The authors concluded that preparatory communication can reduce distress and increase compliance with recommended action in a difficult, real—life setting. . . . But there was clearly less separation or inde— pendence of effects than was anticipated (p. 717). Similar studies have been reported by Egbert et a1. (1964) and by Langer et a1. (1975). In the case of the Egbert et a1. study, a sam— ple of 97 patients undergoing abdominal surgery was divided into two randomly selected groups. Members of the control group were treated in a routine preoperative manner. In particular. as was routinely the case. they were not told about postoperative pain or its management. Members of the experimental group were given a good deal of information about postoperative pain. its character— istics, and its management. These conversations were in the hands of the anesthetists. Surgeons did not know which patients were receiving the special information and continued their practices as usual. In comparison with the control group patients. members of the experimental group used significantly less narcotic medication postoperatively. gave fewer indications of pain on the first post— operative day, and were ultimately discharged nearly 3 days earlier. Langer et a1. (1975) examined the effectiveness of two alter— native coping strategies in dealing with postoperative surgical recovery. They created four groups of 15 patients each who were undergoing elective surgery for which the prognosis was generally favorable. Groups were created by a stratified randomization pro— cedure that equated them for type of operation, severity of oper— ation. sex, age. and religious affiliation. One group was given pre— operative instruction in alternate styles of coping with stress; the second group was given some information preparatory to surgical procedures; a third group received both sets of information; and the fourth group received time—matched neutral interviews that were related to their illnesses but not to coping with surgery. While differences among the four groups were not dramatic, preoperative evaluations by nurses tended to favor all three in— tervention conditions when contrasted with the control group. The members of the control group requested more medication following surgery than either of the experimental groups and tended to stay in the hospital longer. Of the two intervention strategies, the pro— 73 cedure designed to increase coping skills tended to be more ef— f ective than the one that concentrated on providing information. Another stressful life event to which adults are vulnerable is marital disruption. Such disruption, particularly caused by separa— tion and divorce, has long been regarded as among the most stressful life events. Accordingly, it is not surprising that a number of studies have been undertaken to develop and evaluate preventive intervention programs designed to reduce the stress associated with the event. One such 6—month—long program was developed in Boulder. Colorado (Bloom et a1. 1982). Newly separated persons were located by means of the mass media (newspapers and radio stations. as well as posters in super— markets. laundromats, etc.) and direct mailing to human service agencies and to appropriate practitioners (attorneys, physicians. clergymen, mental health professionals). The program was de— scribed in terms that were deliberately ambiguous, with no mention of the availability of services to participants. Assignments to the experimental intervention and to a no— treatment control group were systematically alternated at the ratio of two persons in the intervention program for each one per— son in the control condition. All persons assigned to the interven— tion group were contacted by an intervention program staff mem— ber who administered the initial interview before indicating the availability of intervention services. Persons assigned to the con— trol condition were interviewed by advanced clinical psychology graduate student research assistants. Referrals to community agencies were made in this group as seemed appropriate. A total of 101 persons were assigned to the intervention program while 52 persons were assigned to the no-treatment condition. Interviews were conducted at the time of entrance into the study, and again 6 months later (by which time participation in the preventive in» tervention program was completed). Additional interviews were conducted 18, 30, and 48 months after entrance into the program. Each program participant was assigned to one of three full—time supervised program representatives, who were paraprofessionals with extensive volunteer counseling experience. Twenty program participants constituted a full case load, and never more than 60 participants were in the program at any one time. The program representatives provided emotional support and crisis intervention and served as the link between the participants and the rest of the program. Program representatives played an active outreach role. contacting the participants regularly. developing opportunities for social interaction both individually and in groups, making referrals to other parts of the program and to appropriate community agen— cies, and following up throughout the period of participation in the intervention program. In addition to the program representatives, five specialized study groups were available. each led by a subject—matter expert who was associated with the project a maximum of 4 hours per week. 74 Both group and individual contacts were made available, depending on the demand for services as well as the nature of the issues under consideration. One study group focused on the career planning and employment problems common to the newly separated-r finding employment. changing jobs. long—range occupational and career planning, and the development of marketable skills. The leader of this group was a psychologist experienced in career and educational counseling. The second study group focused on legal and financial issues—the establishment of credit, eligibility for loans, child custody. visitation rights, maintenance and child support, and di— vorce litigation. The leader of this group was an attorney experi— enced with clients who were seeking divorce. The third study group dealt with child rearing and single par- enting problems——children's reactions to the separation of their parents, visitation issues, behavior problems, and enhancing per— sonal adjustment. The study group leader was a psychiatric social worker who specialized in working with children. The fourth study group dealt with housing and homemaking issues—finding a place to live. home repairs. money management, consumerism. food pur— chase and preparation. and time management. The leader of this group was a home economist. The fifth study group dealt with is— sues associated with socialization and personal self—esteem-—en— hancing the self—concept, loneliness, resocialization among single people, and feelings of social and personal inadequacy. The leader was a psychiatric social worker with extensive clinical experience. Program participants were given the opportunity to evaluate their contacts with intervention program components monthly. The program was judged very favorably. Study group experiences were thought to be good; program representatives 'were judged to be skillful and easily available when needed. In addition, more than half of the 6—month analyses demonstrated either significant dif— f erences by group for the entire study population. or for only one gender. In all cases where significant differences were found. in— tervention group participants reported fewer problems and better psychological adjustment than members of the control group. Members of the intervention group also showed significant de— crease in general psychological problem scores across time. The intervention group significantly reduced its psychological distress and maladjustment score from pre— to posttest. In the case of the Anxiety and Neurasthenia Subscales of the Symptom Checklist, equally significant improvements from pre— to posttest were found. No equivalent improvement was found in the control group. The nature of the general improvement lends considerable support to the primary prevention goals of the program. Reductions in anxiety and in general fatigue. along with improved coping skills and physical well—being, are exactly the results one would hope to obtain in a program that attempts to reduce the incidence of di- agnosable psychopathology in a vulnerable high—risk group such as the newly separated. Preliminary analysis of 18— and 30—month 75 data indicates that the differences between intervention and con— trol group members continues to exist, and in some cases increases. Wertlieb et al. (1982) also studied the impact of a preventive intervention program on reaction to marital separation, but did not obtain equally favorable results. They selected 237 subjects from among members of a health maintenance organization (HMO) and formed three groups. The experimental group consisted of 65 sep— arated individuals who participated in the intervention program. The second group consisted of 62 separated individuals who served as a control. The third group consisted of 110 nonseparated indi— viduals who served as a control for marital disruption. Members of the two separated groups had been separated within the year prior to intervention. The experimental group attended eight weekly sessions involving lectures and discussions on such topics as the emotional effect of separation, the continuing relationship with the former spouse, and dating. Each meeting comprised 8 to 15 par— ticipants and was led by specially trained health care providers, including physicians, nurses. social workers, and psychologists. Psychosocial functioning was measured via self—reports on scales of anxiety, self-esteem. energy level, and organization. The mem— bers of the two separated groups completed these instruments 2 months prior to intervention and then 2, 6. and 12 months after the intervention. Medical utilization data were available for all subjects. The authors found that separated individuals increased their use of mental health services dming the 2 years surrounding the sep— aration. Wertlieb et al. attributed this phenomenon to their method of subject selection. Separated subjects were recruited through the offer of participation in a psychoeducational group. Comparisons between the two groups of separated subjects, on the one hand. and other separated HMO members, on the other. indicated that those who did not participate also did not utilize regular mental health services as frequently as did volunteers. Initial analyses revealed no differences among the three groups in terms of either medical utilization or psychOsocial functioning that could be attributed to the intervention program. Only when baseline medical utilization rates were controlled did evidence of a differential effect emerge. But even after this reanalysis, dif— f erences were marginal. with separated controls manifesting slight increases in medical utilization during the 6—month postinterven- tion period when compared with experimental group participants. Clinical Studies Paralleling these experimental preventive intervention studies are a number of clinical investigations that serve to set the stage for well—evaluated and well—conceptualized intervention projects that hopefully will be undertaken in the future. Gore (1978) studied the effect of social support on the physical 76 and mental health consequences of involtmtary unemployment. Using a longitudinal design. the author selected the sample from two communities. one urban and the other rural. and from among married men who had been stably employed at blue collar jobs prior to plant shutdowns. Preliminary findings indicating that the rural group suffered from fewer abnormalities despite a longer period of unemployment prompted a focus on the role of social context. Stress was measured in terms of stages in the process of in— voluntary unemployment, with the earlier ones presumed to be more stressful than the later ones. Number of weeks unemployed and economic deprivation provided other gauges of stress. The dependent variables included measures of depression. self—blame, reported illness symptoms, and level of serum cholesterol. Social support. a hypothesized intervening variable, was operationalized to include interpersonal relationships, activity levels. and oppor- tunities to engage in satisfying interactions which permitted dis— cussion about problems. Men scoring in the lowest third were classified as 'fimsupported," whereas those scoring in the upper two—thirds were classified as "supported." Perceived economic deprivation served as a second intervening variable. Gore found that "consistent differences in the dependent vari— able by level of support indicate that social support modifies se— verity of psychological and health—related responses to unemploy— ment” (p. 168). Yet, the support groups did not differ with regard to such tangible phenomena as weeks unemployed and economic deprivation. This finding was seen as consistent with the emphasis on emotional factors. As expected. rural respondents scored higher than urban subjects on the support measure. It was hypothesized that this finding occurred because of the strength of ethnic ties and the pervasive concern generated by plant closings in rural communities. Wolkon and Moriwaki (1977) examined the primary prevention functions of ombudsman programs. Originally conceived as pro— tectors of political rights, these programs also moderate the psy— chological impact of exogenous reality stresses by providing rele— vant resources and enhancing coping abilities. More specifically. the authors identified vocational and marital problems as the two main sources of such stressful life events. Yet, they cited data indicating that material difficulties were often unattended by mental health professionals. Wolkon and Moriwalci then noted that "the need for such help seems clear if psychological disturbance is to be prevented" (p. 232). The authors. therefore. postulated the importance of assistance in alleviating these problems in particular. Wolkon and Moriwaki focused on the ombudsman program or— ganized by KABC, a Los Angeles radio station. They selected a random sample of 200 letters received by the KABC ombudsman during 1970. These letters were analyzed in terms of the writer's presenting problem, the source of his/her complaint, the perception of the ombudsman's function, and the route employed for problem 77 resolution. Most problems (85 percent) pertained to financial and vocational matters, subjects not generally brought to the attention of mental health professionals. Other difficulties centered on bu— reaucratic red tape, individual rights. and privacy. as well as on interpersonal conflicts. Contacts with mental health professionals were established for the few cases involving psychiatric disturbances. The authors found a widespread perception of the ombudsman "as an effective problem—solving resource" and, to a lesser extent, "as a humanizing factor in an impersonal environment" (p. 234). Indeed, most respondents have tried and failed to obtain solutions to their problems prior to contacting this program. As a result, these people felt "alienated and frustrated in their efforts" (pp. 234—235). Wolkon and Moriwald thus underscored the ombudsman's ability to give "the individual emotional support through working with him to resolve his immediate problem and hopefully developing a more adaptive approach to crises" (p. 235). Roskin (1982) developed a preventive intervention program for persons who had sustained two or more significant stressful life events within the previous year. Intervention consisted of six weekly meetings involving didactic presentations and group discussions in the context of a supportive orientation. A total of 55 persons were involved in the program. Roskin found that the program appeared to have positive conse— quences for depression, anxiety, and interpersonal oversensitivity, and that the program was most helpful for those participants who had experienced the greater number of stressful life events. In particular. those participants who had experienced the death of a family member or close friend appeared to improve most. Wood et a1. (1978) evaluated a group comseling pilot project initiated to help cancer patients at a California community hos— pital. Originally co—led by a psychiatrist and a social worker, the group consisted of 15 members referred by a physician and was limited to 8 weeks. It focused on living and coping with cancer, including issues of diagnosis, treatment techniques, and role clar— ification within the group process. Evaluation of this experience revealed that most participants felt the group was generally help— ful, helped them face anxieties and f ears, served as a better forum for the discussion of their illness than their home environment. delivered useful factual information, helped them feel less iso— lated, was led by warm and supportive leaders, and was worthy of recommendation to other cancer patients. In addition. contact among members extended beyond the regular meetings and gener— ated interest among hospital staff. In a pilot study. Witkin (1978) tested two related hypotheses regarding mastectomy as a stressful life event. The first posited that the sexual relationship had a central impact on women's psy— chological response to this trauma, and the second, therefore. sug— gested the potentially beneficial influence of sexual counseling on 78 postmastectomy recovery. The experimental group was interviewed after the mastectomy while still hospitalized and again upon their return for postoperative checkup or chemotherapy. It was during the second session that the opportunity for counseling was pre— sented. In contrast to the reluctance often generated by the offer of counseling regarding sexual problems, all women approached in this instance agreed to participate. Witkin speculated that this response might reflect the anxiety and need accompanying the mastectomy procedure. The interview itself focused on the sexual history of the woman and her partner both before and after the surgery. Subjective evaluations of the psychological state of the woman were also noted. The counseling centered on concerns related to sexual intercourse, general interactions with the male partner. and mas— turbation. Women frequently expressed concerns about their post— mastectomy attractiveness to men and received statistically sup— ported reassurance that most men do not leave the woman after such a procedure. Masturbation was recommended for those women who had experienced sexual dissatisfaction prior to the mas— tectomy or who did not have a partner. Although no actual numbers were reported. Witkin found that "in the 41 women of the study sexual concerns were, indeed, at the core of their emotional and psychological status and that emotional support was very much in order to help them resolve those con— cerns" (p. 23). The author interpreted the data as confirming the importance of sexual concerns and the helpfulness of counseling. She further concluded that psychosexual functioning supplanted mortality issues as the central concern of recovering mastectomy patients and the partner's acceptance was primary in this connec- tion. Witkin also suggested that these patients regressed to a pe— riod of greater dependency and that, therefore, expressions of support were beneficial. She concluded that counseling should begin soon after the operation to facilitate the adoption of a realistically optimistic image of the future. These sessions, it was recom— mended. should involve comselors who are comfortable with issues of sexuality and who refrain from performing sex therapy per se. Witkin proposed that the "efforts of trained and qualified psycho— therapists to help patients deal with the emotional traumas of a mastectomy should be viewed as part of a team approach" (p. 27). Finlayson (1976) examined social support networks as coping resources utilized by wives following their husband's myocardial inf arctions. The sample consisted of Scottish women whose spouses had been either manual (blue collar) or nonmanual (white collar) workers, who were less than 60 years old and who had survived 12 months following the infarction. Favorable outcomes were those in which the husband returned to work and the wif e judged the illness as over; intermediate outcomes were those in which the husband worked but the wife defined the illness as not over; and unfavor— able outcomes were those in which the husband did not return to 79 work. Support was classified into various types of lay helpers and consultants: children. members of the husband‘s family of origin, members of the wife‘s family of origin. other, and the husbands themselves. The author found that wives whose husbands had favorable outcomes tended to be those who acknowledged support (in terms of lay help and con— sultation) from a wider range of sources. among whom hus- bands and non—kin appeared important for nonmanual [white collar] families and adult children appeared important for manual [blue collar] families. Conversely. wives whose hus— bands had less favorable outcomes tended to be those who ac— knowledged support from a narrower range of sources, often restricted to members of the families of origin of one or both spouses (p. 102). In addition, blue collar families seemed to have access to a nar— rower range of resources. The author hypothesized that the posi— tive impact of receiving support from the adult children of these families might derive from the broadening effect of upward mo— bility by this younger generation. In Australia, Porritt and Bordow (1976) examined the impact of an experimental crisis intervention program on 70 patients hos— pitalized for road trauma in Canberra. Four conclusions emerged from this study. First. the experience of injmy and treatment was rated as an intensely unpleasant crisis. Second, resulting disruption was evident 3 to 4 months after the event for all 30 patients re— ceiving no intervention. Third, followup reports revealed that those who had received the crisis intervention treatment (for an average of 5 hours) scored higher on followup measures of well—being. Fourth. those in the experimental group as well as those with bet— ter outcomes were given more support from significant others. The authors underscored the need for a road accident social service that would help victims throughout the period of crisis. Such a program would attend to the drinking problems that pre— cipitated many of these injuries and would rest on an active out— reach and appropriately confrontational approach designed to cap— italize on the opportunity to use the crisis for constructive pur— poses. The potential benefit of self—help groups consisting of road trauma victims was also mentioned. Kelly and Wallerstein (1977) have reported on their preventive intervention programs undertaken in California with children of divorcing families. According to the authors. "the overall goal was to minimize or prevent the consolidation of psychopathological response. and to facilitate the child's transition to new family relationships following divorce" (p. 23). They excluded children with histories of psychological problems and developed an asseSS— ment that. along with counseling experience. enabled the program 80 staff "to delineate intervention strategies and agendas. .cspecifi to the varying needs of children of different ages" (p. 24). This assessment focused on: 1. The developmental achievements of each child 2. The child's response to and experience of his/her parents‘ separation and divorce 3. The support network surrounding and available to each child These data were used to construct a divorce—specific diagnostic profile. Central to this profile was the particular interaction of divorce—induced stresses with the developmental and personality achievements of each child, and the extent to which these particular configurations had potential for pathological or adaptive developments (p. 29). One key strategic question centered around the degree to which intervention should be effected via child or via parents and child. The authors identified age—related responses to separation and divorce that shaped which intervention was adopted. Their work with preschool children, for example, was accomplished primarily through the parents. whereas their work with older children usually occurred with the children themselves. They described two models of treatment. The first was designed for children who were too young or who were unable to disclose feelings and conflicts to the therapist. This modality involved three to four 1—hour individual sessions. and 4—year followup reports suggested that "some of these very brief interventions were quite helpful and. perhaps more surprising, long—remembered by the children" (p. 30). The second approach served as a more extended crisis intervention and was generally appropriate for older children. (Intervention with later latency age and preadolescent children was carried out directly. in either brief or extended form.) Four goals were described for both models: 1. To reduce suffering, including anxiety, fear. depression. anger. longing 2. To reduce confusion in regard to the divorce and its aftermath 3. To increase the psychological distance between the divorce situation (or the divorcing parents) and the child in instances in which the child was involved in parental conflict 4. To resolve the idiosyncratic issues (e.g.. choosing between parents) 81 479-088 0 - as — 4 Kelly and Wallerstein (1977) found that therapeutic intervention with young children was not effective and that "the intervention of choice with preschool children who do not have a history of psycho— logical difficulty is that made on their behalf with their parents” (p. 31). Latency age children. it was suggested. "were unable to utilize denial as effectively as the younger children and unable to mobilize themselves, like the older children, into various coping activities which facilitated the mastery of psychic pain" (p. 33). In a majority of cases. intervention allowed for the assessment of latency age children's responses to their parents' divorce but failed to decrease anguish. Therefore. therapists were careful to respect the defenses of children in this age group. A divorce monologue. in which the therapist spoke of the experiences of other children. was a recommended technique. Latency age children were especially eager to see their fathers. and flexible. frequent visiting times were urged. Suarez et al. (1978) described the UCLA Section on Legal Psychiatry program designed to work with divorcing families referred by courts, by the litigants themselves, or by the attorneys involved in the divorce proceedings. Assuming a consultant role. the project staff acted as active collaborators "working to enhance the ftmction of the court and thus protect the interests of the children involved in litigation" (p. 273). Clinic participation oc— curred at the time of the divorce or during subsequent litigation and always included an evaluation of the child's entire domestic situation. Evaluation generally involved individual interviews, joint interviews with the parents. meetings with the attorneys. and psychological testing. as well as indirect techniques such as observation of play between parent and child. The authors also described a postdivorce clinic which can inter— vene after the court has ruled. Suarez et a1. noted that counseling undertaken subsequent to the decision—making process involves working directly with the parties in a thera- peutic way to create a constructive experience out of the postdissolution chaos and pain. Although the welfare of the child continues to be of primary concern, counseling focuses on the parents and the child may not be seen (p. 277). After an initial evaluation, conjoint counseling with two thera— pists was usually arranged for the ex—spouses. "In a case that is going smoothly, the parties will begin to negotiate with each other and to work through their current anger" (p. 279). New spouses were included in these sessions when appropriate. When the parents refused to work directly with each other, group therapy was often recommended. These groups were composed of divorced people who could not be seen with their spouses. The authors found that healthier people generally had better prognoses and smnmarized their work by stating that 82 the population. then, is divorced people. 90 percent court re— ferred (the others are self—referred). who cannot finish with each other and their divorces. and find themselves continuing the fight through the children. These people have unfinished business that they do not know how to resolve. It is our job to teach them how to do that and get on with their separate lives (p. 278). As has been seen. experimental and clinical studies of a very wide array of specific adult stressors and of ways to reduce their personal impact are now being reported in the literature. The existing knowledge base would seem to allow for a very significant increase in such studies. At the same time. however. analytic studies can be expected to increase our understanding of the array of stressful life events that can occur to adults and how preventive intervention programs might be designed, implemented, and evaluated in the future. Bereavement Experimental Preventive Intervention Programs Another adult stressful life event that has been studied is be— reavement. Parkes (1981) evaluated the impact of bereavement counseling on family members of 181 patients who died at St. Christopher's Hospice in London during 1970 through 1974. The selection of family members for followup was made on the basis of a questionnaire completed by nurses at the time of bereavement. All family members judged by the nurse as likely to cope badly or very badly were followed. All other family members (judged as likely to cope well, fairly. or doubtfully) were divided into a high— risk or low—risk group on the basis of their replies to a brief ques— tionnaire. Members of the high—risk group were randomly assigned to either an experimental or control condition. Volunteer counselors visited experimental subjects. Contact was established with a key person within families. most frequently the spouse. in general. these subjects were seen as requiring help. but a quarter of the participants seemed to be coping quite well. Two— thirds of the key persons needed to talk about their bereavement, and more than half were evaluated as having benefited from such expression. Specific problems were also evident during visits by the counselor. including interpersonal problems. housing. financial problems, the disposal of possessions. need for friendship, insomnia. actual physical illness. keeping the existing home. problems with crying, eating too little, hatmting memories. and difficulty main— taining relationships. In 50 cases. an extended version of the origi— nal assessment form was completed by volunteer counselors after 83 the first meeting; it revealed relatively high incidence of crying. anxiety. anger or bitterness. talking about the loss. and suicidal risk (which. in no case. was actualized). A health questiomaire was administered 20 months after the death. The study results reflected poorer outcomes among nonsupported high—risk subjects than among nonsupported low—risk subjects. Members of the control group were more likely to have four or more autonomic symptoms and to have increased their use of drugs, alcohol. and tobacco. In the entire sample, women, those who exhibited clinging behavior. older subjects, those likely to do poorly according to mn-ses‘ predictions. and subjects engaged in self—reproach tended toward the worst outcomes. Moreover, ex— perimental subjects emerged as better adjusted than control sub— jects in terms of overall outcome as well as in terms of autonomic symptoms. depressive scores. amdety level. physical symptoms. habit changes (i.e.. drug. alcohol, and tobacco usage). and use of health care. This trend was particularly pronounced among men. While the low-risk group obtained lower scores than high—risk controls. they were not appreciably different from high—risk experimental subjects. Parkes asserted that ”it seems. therefore, that the effect of the service has been to reduce the health risk of the 'High Risk' group to about the same level as the 'Low Risk' group" (p. 186). He con— cluded that the service reduces the need for drugs. alcohol, and tobacco among the bereaved and reduces the number of symptoms at— tributable to anxiety and tension from which they suffer. It may also reduce the use which they make of doctors and hos— pitals and improve their overall level of contentment (par— ticularly if they are men) (p. 186). Raphael (1978) has also examined the effect of preventive inter— vention on the bereavement process of recently widowed men. Be— ginning with the psychoanalytic view of mourning and melancholia, the author hypothesized that women with pathological relationships with their husbands would experience a more difficult bereavement period than those with less problematic relationships. Raphael de— veloped a set of criteria to identify a group of recently bereaved widows who were at high risk of becoming ill. These criteria in— cluded the widow's perception of her social network as nonsupport— ive during the bereavement crisis. and a preexisting pathological. ambivalent, or dependent marital relationship. In one study, 12 widows whose marriages qualified as pathologi— cal were seen for one or more sessions within the 3 months follow— ing their husbands' deaths. These meetings were designed to en- courage the expression of grief and promote the mourning process, especially with regard to its ambivalent components. A control youp receiving no intervention was also established. A followup 84 questionnaire was distributed 13 months after the death. The results indicated that 11 of the 12 experimental subjects evidenced a good health outcome, whereas 1 evidenced a poor out— come. Only 4 control group subjects exhibited a positive outcome. while 6 exhibited a negative outcome. Moreover, 9 of the widows who participated in the intervention program showed no depressive signs, and none of this group were hospitalized for depression. In contrast, only 4 control group widows were free of depressive symptomatology. and one required hospitalization. Raphael com— pared these data with those obtained from a group of widows with pathological marriages but with more positive perceptions of their social networks. These 6 additional subjects expressed no depres— sive complaints. They saw their social networks as actively helpful during the bereavement process. and the average number of per— ceived unmet needs was smaller for this than for the other groups. The author identified several specific types of unmet needs characterizing the particularly complicated mourning process deriving from ambivalent marriages. These involve needs to deal with guilt, to be encouraged to cry. to talk of the husband, to talk about the marital problems. to talk of the husband's bad points. to be encouraged to release f eelings, to receive optimistic support for the future, and to receive reassurance of their own good qualities and coping capacities. Such unmet needs, it was suggested. related to the resolution of the ambivalence complicating the mourning process. The author hypothesized a connection between pathological levels of ambivalence characterizing marital relationships and the development of melancholia. She also suggested that widows with an effective social network were able "to work through their loss. perhaps in the way Freud suggested, helping them with their strug— gles to release the bonds of love and hate from the lost object” (p. 309). Raphael concluded that professional support designed to ac— complish this release substantially reduced the risk of depression. She summarized by noting that this small study suggests that preventive support at the time of the crisis may be helpful in lessening the pathological ef— fects. In some bereavements ambivalence may predispose the mourning to melancholia. Crisis support, which facilitates the working through of this ambivalence, may prevent or lessen the melancholia (p. 310). In another study, Raphael (1977) investigated the effect of inter— vention on the reactions to spousal death of 200 recently widoWed women identified as facing a high risk of morbidity. Subjects were selected from among women less than 61 years old applying for pensions at the Social Security Department in Sydney, Australia. They were then classified according to risk status on the basis of four predictive indices: 85 1. A high level of perceived nonsupportiveness in the social network 2. A moderate level of perceived nonsupportiveness coupled with traumatic circumstances of death (e.g.. mtimely, unexpected, anger— or guilt—provoking) 3. A highly ambivalent marital relationship 4. The presence of a concurrent life crisis The 64 high—risk subjects were randomly assigned to either an experimental or a control group. Intervention consisted of counsel— ing sessions during the first 3 months of bereavement. The author used a model of selective ego support for the expression of such grieving affects as sadness, anger, anxiety. hopelessness, helpless— ness, and despair as well as a general facilitation of the mourning process. Control group members received no intervention. Outcome was measured with a health questionnaire that assessed changes in symptomatology. The questionnaire was mailed to all subjects 13 months after the death. Of the experimental group members providing followup inf orma— tion, 21 achieved a good outcome and 6 a bad outcome. Among the control group, 12 were classified as good and 17 as bad outcome. Indeed, the data were seen as reflecting the consistent superiority of the intervention group in terms of lessened health impairment. The more marked health deterioration of the high—risk control group included greater frequency of numerous symptoms (e.g.. sleeplessness. back pains, aches, greater number of doctor visits. more weight loss. increased alcohol intake, increased smoking. increased drug usage and accompanying concern, stoppage of men— struation. more severe depression, and diminished work capacity). Moreover. intervention group widows came to resemble the residual subjects who had not scored as high—risk candidates. whereas differences between control and residual groups emerged as highly significant. The author suggested that "it is possible that intervention has contributed to making this predicted high—risk group more like the residual group" (p. 1453) of the social network. In addition. perceived nonsupportiveness of the social network was found to be the most accurate predictor of those likely to benefit from intervention of the sort tested in this study. Raphael con— cluded that intervention to modify the process of bereavement toward the promotion of normal grief and the accomplishment of a sig— nificant degree of mourning appeared to be associated with a significant lowering of morbidity compared to findings in con— trol group subjects (p. 1454). 86 In a 2—year study of postbereavement adaptation of a group of 162 widows, Vachon et a1. (1980) contrasted adaptation in two groups. one of whom consisted of widows who were paired with other widows for purposes of providing emotional support and prac— tical assistance. These helper—widows had previously participated in a training seminar that examined issues related to widowhood, and the widow-to—widow program consisted of supportive telephone calls and both individual and small group meetings. These authors found that the intervention and control (those who were not part of the widow—to—widow program) subjects were equally disturbed 1 month after bereavement, but that widows in the intervention program were less depressed and less preoccupied with the past. One year after widowhood, the control group had caught up with the intervention group in personal adjustment. but the intervention group was superior in terms of their resocializa— tion. At 2 years, level of health was substantially better in the intervention group than in the control group. The authors concluded that those widows who received the intervention services followed the same general course of adaptation as the control subjects, but their rate of achieving a satisfactory level of adaptation was accelerated (see also Walker et al. 1977). Rogers et a1. (1980) described the transformation of the experi— mental program described above (Vachon et al. 1980) into an on- going program offered to widows. Although use of a self—help re— ferral system limited the membership of lower—class and immi— grant women. more than 1.500 widows had already participated. The program employed paid widow contacts similar to those in— volved in the original study. These widows were trained to rein— force existing skills. imderstand the reactions and needs of the newly widowed, share bereavement experiences, and utilize helping techniques and resources. The program. Community Contacts for the Widowed (CCW), offered four kinds of service: 0 One—to—one support 0 Group meetings 0 Volunteer work 0 Education The authors found that newly widowed women tended to prefer the one—to-one modality, whereas the group format appealed more to those who had been widowed for at least 6 months. CCW con— centrated on women who were widowed for 3 years or less. The average length of participation was approximately 1.5 years. Data indicated that a normal grief reaction requires 2 years and that contact with a self—help organization often occurred after 6 months had elapsed since the spouse's death. The authors also not— ed that middle—class women were most likely to seek help from 87 CCW and that those referred by professionals experienced yeater distress than did those who learned of CCW via other means. In addition. younger women. particularly those with young children. generally needed a greater number of hours of help and generally contacted the organization earlier in their bereavement. The authors concluded that "CCW appears to be an effective and efficient program for delivering services to a high—risk group" (p. 847) Barrett (1978) investigated the effect of three types of inter— vention strategies on women's experience of widowhood. Widows responding to a newspaper armouncement describing the program first attended an orientation session and completed a pretest ques— tionnaire; 83 women participated. They were then assigned to one of three experimental groups or to a control waiting list. Self—help groups were designed to encourage members to assist each other in finding solutions to the problems of widowhood. Confidant groups sought to develop close friendships between pairs of widows. Con— sciousness—raising groups focused on enhancing awareness of how members' experiences as widows were connected to their status as women. Each group was led by a (non—widowed) female doctoral student in clinical psychology. After a 7—month treatment period. subjects completed a posttest and subsequently attended a followup session. Dependent variables included frequency of physical complaints. predictions of personal health, intensity of grief . attitudes toward widowhood and remarriage. social role involvement. quality of life changes since program involvement. open—ended comments, extent of social activity. and interaction among members. Measm-es of self—esteem. locus of control. life satisfaction. and attitudes to— ward women were also obtained. At posttest and f ollowup, subjects evaluated the extent to which the groups provided education and help. The author found greater change among experimental than among control group members. Participants in the experimental group more strongly endorsed the notion that women need autono- my and independence beyond that provided by traditional family roles. In this regard, Barrett suggested that the experience pro— vided a reference group with new norms. Also noteworthy were the more positive health predictions among experimental group mem— bers. a possible indicator of morale. as well as improvements in self—esteem. Plans for continuing contact among group members were interpreted as confirmation of the program's effectiveness. Gains in the intensity of grief were seen as reflecting acceptance of the negative emotional reactions in widowhood and a lessening of denial. Locus of control, life satisfaction. frequency of physical complaints. attitude toward widowhood. and attitudes toward women did not emerge as descriptors of changes accompanying participation. The consciousness—raising modality was identified as the most consistently effective approach. Barrett tentatively attributed this success to its content, which provided an outlet for anger. and to its highly structured format. The self—help approach was fomd to be the least effective. This result was hypothetically connected to the presence of a professional leader. Changes recorded among control group members were explained in terms of the therapeutic effect of the hope of a group experience. Clinical Studies Silverman (1967, 1972; Silverman et a1. 1975) designed a program to facilitate the transition from role of wife to role of widow in newly widowed persons less than 65 years old. Other widows. who had themselves already made the transition successfully, were used as helping agents. In her review of the literature and in her own studies. Silverman fomd that newly widowed persons rarely considered using tradi— tional mental—health—related agencies unless they had used them prior to widowhood. In f act. Silverman suggests that use of a psy— chiatric clinic may not be appropriate for widows. particularly during the early stages of mourning. Rather. what is required is a service that is available immediately after the death occurs. that can reach every new widow, that has legitimate access to the new— ly widowed population. and that can provide a range of services including all existing community resources. On the basis of this analysis. Silverman has concluded that "the most effective care— giver . . . is another widowed person who has recovered" (p. 44). While this project has not been quantitatively evaluated. quali— tative observations suggest that large numbers of newly widowed individuals find the widow caregivers enormously helpful to them in the crisis that follows the death of the spouse. Another approach to reduce the negative consequences of widowhood as a major life stressor is the self-help group. In a recent evaluation of such a group. THEOS. Lieberman and Barman (1981) have analyzed data from a large followup study of a group of widows and widowers who were either current members. former members, or those who had been invited to join the self—help pro— gram but chose not to do so. Several hmdred persons were involved in the followup study that included two questionnaires administered 1 year apart. Respondents were asked to describe their most pressing widow— related problems. Role problems. that is. problems concerned with establishing a role as a single individual, were mentioned by 61 percent of the respondents. Problems of depression were mentioned by nearly one—third of the sample. In evaluating the effectiveness of the self—help group, the authors divided the sample according to how actively they had been involved in the program. Their analysis 89 suggested that active participation in the self—help group positively affected the mental health status of the participants. particularly in reducing depression and in increasing self—esteem. Videka~Sherman (1982) investigated the relationship between degree of self—help group participation in Compassionate Friends. an American and Canadian organization aimed at helping the be— reaved. and levels of depression and personal change following the death of a child. Personal change, a central concept of Compassion— ate Friends, was monitored by a measure based on the work of Hoffman (1972). The concept of personal change derives from existential and ego psychology and was viewed as reflecting move— ment toward ego ideals, possibly accompanied by some deeper understanding of the self and of life. The author selected recently bereaved parents, that is, those who had lost a child within the past 18 months. The parents were drawn from 18 geographic areas in which chapters of Compassionate Friends were located; nonmem— bers and members with various levels of involvement were included in the sample. Data were collected in two waves, with all first— wave respondents receiving a second questionnaire several months to 1 year later. The sample was primarily female, married at the time of the first survey, and middle and upper—middle class. More than half of the parents had lost their children suddenly. Videka—Sherman found a linear relationship between psychologi— cal adjustment as reflected in personal growth and level of involve— ment in Compassionate Friends. The author noted that the nature of the personal change detected by this study corroborated the dimensions identified by researchers investigating existential change as a function of life crisis (Yalom 1980; Hoffman 1972). Such dimensions of change centered aromd the interpersonal arena and included empathy for others as well as shifts in intimate rela— tionships, belief systems, and self—representation. She concluded that these findings demonstrate that this self—help group for be— reaved parents does affect the psychological adjustment of parents to the death of a child. More specifically, involvement in Compassionate Friends sustains a positive assessment of the impact of the loss on the parents' self. while resulting in no differential effects in terms of levels of depression across involvement groups (p. 75). As can be noted. preventive intervention programs that are de— signed to be of assistance to people going through the process of widowhood appear to be sufficiently effective to encourage their continuation and more detailed evaluation. A number of different strategies have been developed to be of help to newly widowed persons. and it will be important to determine whether certain pro— gram strategies are more effective in specific settings or under specific circumstances. 90 Community-Wide Stressful Life Events Mental health professionals have involved themselves in efforts to be of help during community—wide stressful life events since the time of the Coconut Grove fire in Boston in 1942 (Lindemann 1944). As can be well understood, no such studies have been experi— mental in nature. That is, efforts were made to be of assistance to everyone in need, and no untreated control groups were created. Because a number of years must necessarily elapse between the time of a disaster and the publication of scholarly reports regard— ing crisis intervention programs that are mounted to be of assist— ance to its victims, relatively little is known about recent disas— ters. Initial publications have appeared. however. that are con— cerned with preventive intervention services that were developed subsequent to the 1977 Hanafi Muslim takeover in Washington. the 1980 Cuban sealift, and the 1981 Kansas City hotel disaster. But disasters that have an impact on the entire community are not limited to the United States. Singh and Raphael (1981) used a naturalistic approach to study the impact of bereavement counseling on the relatives of victims of a rail disaster in a suburb of Sydney. Australia. The disaster occurred in 1977 and involved the derailing of a commuter train. Many Victims were trapped mder debris for hours. thus heightening the trauma to relatives. A total of 83 people were ldlled and ap— proximately the same number injured. Bereaved families were contacted and classified according to risk of morbidity on the basis of such criteria as inadequate support. additional crises, and prob— lematic relationships with the deceased. A variety of counseling services were offered. Indeed. the authors noted that the emergen— cy nature of the situation prevented the provision of uniform coun— seling services or the systematic recording of relevant information. Approximately 1 year later, followup data were collected and analyzed to: 1. Assess subjects' physical and mental health and their current level of functioning 2. Assess the degree to which the loss had been resolved and to understand the subjects' perceptions of the helpfulness of both their social networks and the intervention 3. Compare these data with notes on the bereavement counseling The authors acknowledged that the absence of a control group other than those who refused the intervention and the lack of pre— bereavement assessments represented methodological problems inherent in their approach. Dependent variables were measured with a health questionnaire focused on changes since the accident; an author—constructed loss questionnaire tapping continuing affec— 91 tive distress. inability to give up the lost object, and inability to return to normal functioning; a questionnaire probing the social network at the time of bereavement; and a nondirective interview. These measures enabled the placement of the 44 subjects, most of whom were spouses of victims, into one of three categories: 1. Clearly good outcome. in which the person had returned to normal functioning and was resolving the grief process 2. Clearly bad outcome. in which grief was poorly resolved and level of functioning was not comparable to prebereavement levels 3. An intermediate category, in which people appeared to be coping well but did not seem to have grieved According to responses obtained from the health questionnaire, parents of victims were more 'poorly‘ adjusted than spouses. This finding was consistent with that emerging from the clinical judg— ments made during the interview. Although those who received intervention differed from those who did not only in regard to health scores. those who had found the counseling helpful emerged as better adjusted than others in terms of health. loss. and clinical assessments. In addition, more experienced counselors seemed to obtain better results. Moreover. those relatives with inadequate support networks and those who had not seen the body tended to— ward worse outcomes than those with adequate networks and those who had seen the body. Subjects identified as high—risk seemed to have done better than the low—risk group. Yet. the authors specu— lated that such a result might have derived from the use of risk criteria developed for a different population or from the greater effort expended by counselors aware of the classifications of their clients. Singh and Raphael concluded that their study "suggested that be— reavement counseling following disaster may play a role in lessen- ing postbereavement morbidity” (p. 212). They underscored both the status of parents as a particularly high—risk group and the need for well-trained counselors in this area. Sank (1979) described how the mental health staff of the George Washington University Health Plan dealt with the aftermath of the 39—hour takeover of 154 hostages by the Hanafi Muslim sect in three Washington. D.C., locations on March 9-11, 1977. The staff of the health plan was involved in this takeover because many persons held captive were members of this particular health main— tenance organization. The treatment format was short—term, crisis—oriented, broad— spectrum youp behavior therapy. Treatment began within a week after the hostages‘ release and was available to all hostages re— gardless of insurance coverage. About half of the hostages were 92 treated. Sank reported that "the victims gained a seme of comfort from the assurance that they were being followed professionally from the outset” and that ”there was a pleasant surprise that the Health Plan was seeking them out, regardless of their insurance affiliation" (p. 337). The treatment program, which appeared to be remarkably eff ective. had several distinctive features: First, a strateg of primary prevention was employed. Initial- ly, only a few of the hostage victims experienced any unpleas— ant effects from the ordeal, except for exhaustion. 'me Prac— tice felt confident. however (and their prediction was borne out), that there would be an emotional fallout from the expe— rience. They set about employing a treatment program focused on minimizing the effects of the tramna. They attempted to teach coping and support skills to the victims to assist them— selves, their colleagues, and their families. The Practice instituted a community—based approach to treatment. They acted on the belief that the most powerful tool available to them was the natural support system provided by past and future associations at the work place. Consistent with this belief in mutual support. they set about fostering it through group treatment, work4ite—based treatment. and the retention of an already existent supportive atmosphere. They reinforced openness. the solicitation of caring. and an eager positive response to these requests (p. 337). Perez (1982) has described and analyzed the organizational, social, and clinical issues that had to be dealt with in providing mental health services to the Cuban immigrants who arrived in Florida during the 1980 sealift. In particular. Perez described the work of the hastily formed interdisciplinary team who attempted to meet the needs of several hundred minors who fetmd themselves placed among several thousand adults in one of the refugee centers. Perez concluded: Ultimately. like most therapeutic interventions. the provision of ventilation. catharsis. and an honest. concerned and open relationship proved to be critical. A need was observed for a setting in which experiences and feelings could be freely and Openly explored. Many minors shared feelings of abandonment by families or guilt at having left a loved one behind. Most experienced considerable ambivalence over their decision to leave Cuba and come to the U.S. However, the discussion of feelings among the population of minors had previously not been acceptable. The simple provision of an open setting in which these feelings could be discussed proved quite helpful to many (p. 43). In the case of the 1981 Kansas City hotel disaster. Gist and Stolz 93 (1982) have described the response of the mental health commmiity following the sudden collapse of two aerial walkways that spanned the lobby of the Hyatt Regency Hotel. resulting in the deaths of 111 persons and injuries to more than 200 others. It was estimated that. counting individuals involved in the rescue effort as well as family and friends of those killed and iniured, more than 5,000 persons were at risk of developing negative reactions to that crisis. It was difficult to develop a prompt and effective professional response. The seven mental health centers that served Kansas City had no coordinated crisis response plan, and there were no guide— lines to help the mental health centers. Because nearly everyone who was ldlled or injured was a local resident, the impact of the crisis extended to the entire comrmmity. Finally, it was not possi— ble to identify all of the 5,000 persons who were judged to be at risk. A community—wide response was developed by a collaborative effort of the community mental health centers. the local mental health association. the Red Cross, and the local regional office of the US. Department of Health and Human Services. That response. developed within 72 hours of the tragedy, included three compo— nents—the creation of support groups at each of the mental health centers, the development of a training program for mental health professionals and for primary caregivers who would be a major source of support in the community. and a mental health information and education campaign that was distributed by the mass media. The messages were simple—reactions of shock and disbelief , of exhaustion. and of extended assessment and integra— tion were to be expected and were normal; sharing these feelings with others and accepting these feelings in oneself were important; and help was available. It was estimated that more than 10 percent of the at—risk popu— lation contacted a community mental health center for assistance. But later requests for direct clinical services for delayed reactions were far fewer than would have been anticipated on the basis of the literature. Thus, the authors concluded that longer term nega— tive consequences of this tragedy were diminished by ”the efforts of the Kansas City mental health community and media to provide coordinated. aggressive prevention programming immediately after the disaster" (p. 1139). Summary The vigor of the field of preventive intervention that is based on the stressful life event paradigm can be seen in the more than 40 experimental and clinical studies that have been described in this chapter. The programs vary enormously in terms of their objec— tives, the population being served, the nature of the services that 94 are provided. and the discipline that has gone into plarmins and evaluating the effort. Basic stressful life event research and preventive intervention program development have a synergistic relationship—each has the potential to strengthen the other. Just as basic research in the field of stressful life events is becoming sounder and more sophis— ticated, so are preventive intervention programs becoming more disciplined and persuasive. Together, there is every reason to believe that the next decade will witness significant breakthroughs in the development and evaluation of new preventive interventions based upon this important paradigm. 95 Chapter 6 CONCLUDING SUMMARY Examination of the conceptual and programmatic base of pre— ventive intervention that derives from the stressful life event paradigm should leave the reader with the sense of being on the threshold of a new era in the control of mental disorders. As was noted in the first three chapters of this monograph. the basic re— search documenting the complex role of stressful life events in the precipitation of both psychological and physical disorders is well under way, and, as has been seen in chapter 5. a number of preven— tive intervention programs have already been implemented on the basis of this paradigm. The stressful life event paradigm has intro— duced a new sense of hope that a proportion of psychiatric and physical disorders not currently preventable can be prevented. In this concluding chapter. some implications of the existing litera- ture will be examined. with an aim toward improving the quality of preventive intervention program plarming. implementation. and evaluation. ‘ The research paradigm for evaluating preventive intervention programs that are based on the stressful life event model differs from the more traditional evaluation of a specific disease preven— tion program. In the case of a specific disease. once a theory has been developed that links a presumed causative agent with the risk of developing the illness, a preventive program can be assessed by demonstrating that the risk of that disease in an exposed untreated group is significantly greater than in an exposed treated group— when the treatment is defined as the preventive intervention pro— gram. The essential dependent measure of success is the incidence of a single disease or disorder. In the case of the stressful life event paradigm, a model that is by nature nonspecific in character. evidence of preventive program effectiveness needs to be sought in a far larger arena. The existing basic research may suggest. for example. that a specific stressful life event is significantly associated with increased risk of coro— nary artery disease, neurotic depression, alcoholism, hypertension. suicide, schizophrenia, and automobile accidents. Under these con— ditions. evidence of preventive program effectiveness needs to be sought by assessing all of these dependent measures. This task is thus tmusually demanding. 96 Such evaluations. if they are to be persuasive in their results. need to be able to contrast outcomes in comparably selected groups who differ only in whether the preventive program is made available to them. That is, experimental and control groups need to be identified who have undergone the same stressful life event or set of events and who do not differ significantly on any demo— graphic characteristics that are thought to be associated with outcome. As was noted in the previous chapter, many preventive intervention programs have yet to be soundly evaluated. In addi— tion, prudence would suggest that where programs have been found to be effective. program replications would be most desirable— preferably with demographically different and larger samples of participants and with the application of the most powerful statistical tests of effectiveness consistent with the nature of the program evaluation design. A related problem with many of the studies that have been re— ported is that the preventive intervention program has not been described in enough detail to permit a true replication. Finally. our experience has been that preventive intervention program descrip— tions (whether or not they have been appropriately evaluated) exist in so many different source documents that locating them in a comprehensive manner is very difficult. Until such time that a clearinghouse of preventive intervention programs is established. agencies will have difficulty building on each other‘s experiences. Two facts are encouraging. however. First, if one is to judge by the published literature, a number of well—evaluated preventive intervention programs appear to be effective, that is, local agen— cies have choices about how -to proceed with implementing such programs. For them. the issue is to identify preventive intervention progams that are most effective or that are most appropriate to the specific commm'iity situation. Second. trial preventive inter— vention programs are not only highly desirable but are also rela— tively inexpensive. thhermore, within a short time it should be possible to obtain at least preliminary information about the program's effectiveness. One additional observation about the preventive intervention programs that have been identified and described should be made. With regard to the stressful life event paradigm. preventive inter— vention can be undertaken either by reducing the incidence of the stressful life event or by helping people cope with the event once it has occurred. Virtually all existing preventive program research has been based on the second of these two approaches. Very little research has been reported in which an effort was Imdertaken to reduce the likelihood that the stressful life event would occur—— such as. for example. by working in a community in which a fac— tory is planning to close in order to keep the factory open. or by working with intact families to lower the risk of marital disruption. Efforts to reduce the incidence of certain stressful life events can take mental health professionals out of their traditional re— 97 medial arenas. and there may be considerable uncertainty as to how to proceed in a setting where there is no traditional mandate to do so. It is imperative, however. to examine the possibilities of prevention by including studies that are designed to reduce the frequency with which the stressful life event itself takes place. One must be both elated and sobered by the current state of stressful life event research. On the side of elation is the fact that there is so much active research underway pertinent to some as— pect of the stressful life event paradigm. On the sobering side is the fact that the research thus far is only suggestive. Much more and much better quality preventive intervention research will have to be undertaken. reported, replicated. and critically analyzed be— fore it will have clear and unequivocal implications for comrmmity mental health program planning. It is the rare community that ought not be concerned with the stressful life events that impinge upon its citizens. Yet, few com— munities attend to these events on any systematic basis. With the growing evidence of the harmful consequences to physical health and psychological well—being of many forms of stress. and with the growing nationwide interest in stress management. it would seem timely to urge health and welfare agencies, both public and pri— vate. to begin identifying the life stressors that occur in their cormmmity. Such a process. especially when aided by the mass media, can help create an alert and informed citizenry and can help set the stage for responsible resource allocation for preven— tive intervention. No single list of stressful life events can be complete. Each com— munity has its own unique distribution of stressful life events and its own unique stressful life event history. Furthermore, since life stressors have complex consequences, it would be important to document the apparent effects of these stressful events in order to be able to argue persuasively that a preventive intervention pro— gram is indeed desirable. With regard to any particular stressful life event. two groups of people need to be identified and studied—those who cope well with the event and those who cope poorly. Locating people to study is not inordinately difficult. The list of people who undergo certain stressful life events is public, for example. persons who divorce or who are involved in automobile accidents of sufficient severity to require a police report. Victims of other stressful life events can be identified but in a roundabout marmer. For example, siblings or parents of children who have died can be located by examining death notices. Siblings and parents of children with life— threatening diseases can be identified by using hospital admission records. When human service agencies are sensitized to the need to learn more about the consequences of certain stressful life events, ad— mitting staff can make inquiry regarding precipitating events and can identify people who are helpseekers who might be thought of 98 as doing poorly and who are attempting to cope with a specific stressful life event. Finally. it is possible to locate persons who are undergoing stressful life events by the use of certain quasi—public records. In a community that is characterized by rapid immigration, it would be possible to work with the local telephone company. public utility company. or Welcome Wagon to identify newcomers to the com~ munity. If there is reason to believe that persons who live alone are unusually vulnerable to a number of particular stressful life events. the local community planning agency can help identify the section of the community that has a concentration of single room dwell- ings. Interviewing people who appear to be coping poorly. as well as those who appear to be coping successfully, would be important in the process of developing an appropriate local strategy for pre- ventive intervention. 99 REFERENCES Aahta. C.W. Psycho—eodal stress and health disturbances. Social Sci— ence and Medicine 8:77—90. 1974. Abraham. 11.8.. and Patterson. R.D. Psychological distress among the continuity elderly: Prevalence, chmcteriaties and implications for service. International Journal of Aging and Human Development 9:1—18. 1978-79. Albee. G.W. Primary prevention. Canada's Mental Health 27:5—9. 1979. Andrews. 6.; Tenmnt, C.; Hewson. D.M.; and Vaillant. G.E. Life event stress. social support. coping stifle. and risk of psyctwlosical impair— ment. Journal of Nervous mdMental Disorders 166:307-316. 1978. Antonovsh. A. Health. Stress and Coping. San Francisco: Josey-Baas. 1979. Asher. 5.1.. and Bloom. 8.1.. Geographic mobility as a factor in adhstment to divorce. Journal of Divorce 1983. Askenasy, A.R.; Dohrenweui. B.P.; and Dohrenwend, 8.5. Some effects of aocialclassandetlmicmtpmemberstdponfidsmmtsofthemagxfl— tude of stressful life events: A research note. Journal of Health and Social Behavior 18:432-439. 1977. Bane. MJ. Marital disruption and the lives of child-m. 1n: Levinger. G., and Moles. 0.. eds. Divorce and Separation: Context. Causes and Con— sequences. New York: Basic Books. 1979. Barrett. CJ. The effectiveness of widow‘s maps in facilitating change. Journal of Consulting Md Clinical Psychology 46:20—31. 1978. Barrett. J .E. The relationship of life events to the onset of neurotic dis— orders. In: Barrett. J.E.: Rose. R.M.; and met-man. 6.1... eds. Stress md Mental Disorder. New York: Raven Pres. 1979. Baum. A.; Singer. 1.15.; and Baum. C.S. Stress and the envirotment. Journal of Social [mes 37:4-35. 1981. Belfer. M.1..; Mulliken. 1.8.; and Cochran. T.C. Cosmetic surgery as an antecedent of life change. America: Journal of Psychiatry 136:199—201. 1979. ' Benjamin. P.Y. Psychological problems following recovery from acute life—threatening illness. American Journal of Orthopsychiatry 48:284- 290. 1978. Bloom. 3.1.. A miversity fresln'aan preventive intervention prom: Report of a pilot project. Journal of Conwlting and Clinical Psychology 37:235—242. 1971. Bloom. 8.1... Changing Patterns of Psychiatric Care. New York: Human Sciences Press. 1975. Bloom. 8.1.. Community Mental Health: A General l‘ntrofitction. Zd ed. Monterey. Calif.: Brooks/Cole. 1983. Bloom. B.1..: Aaher. 5.1.; and White. S.W. Marital disruption as a strenor: A review an! analysis. Psychological Balleth 85:867—894. 1978. Bloom. B.l..; Hodges. W.F.; and Caldwell. R.A. A preventive intervention mfortheneflbaeparated: initial evahnflmAmerlcanJoumal of Community Psychology 10:251-264. 1982. Blurnemhal. MD. Mental health among the divorced. Archives of General Pvehiatry 16:603-608. 1967. Boryamko. M.. and Borysenko. 1. Stress. behavior and kmmmity: Animal models and mediating mechanisms. General Hospital Psychiatry 4:59— 67. 1982. 100 kaoley. J. P.: Daniels. 1...;F and Jones. '1‘. C. The International Dictionary of noughts. Cl'iicaao: Form 1969. Bremen M. H. Mental Illness and the Economy. Cambridge. Mass; Harvard University Pres, 1973. Bremer. M..H influence of the social environment on peyohopetholog: ‘l'hehistm-icperspective. In: Barrett. J..;E Rose. R.M.: “Herman. 6.1... eds. Streasanallantal DisoMer. New York: Raven Preas. 1979. Briacoe. CH9]; Smith.J.B.;Rohins. 8.; Martens; endGaakin.F. Diva'ce mgapsyohietric disease. Archives of General Pochiatry 29:119—125. 19 Emmet. E..;J Parkinson. D..;K Seminal-LC; DimLO.;endGmdek. P.C. Mentalhealthofresidentsnearthe'l‘ln-ee Milelslendreactor:A comparative mm of selected m. Journal of Preventive Psychiatry 1:225—276. 1982. MG.W.Meuflng.meannement.endstressoflifeevents.h:Doln-en— mid. 8.5.. and Dohrenwend, B.P., eds. Streslul Life Events: Their Nature and Effects. New York: Wiley 5. Sons. 1974. Brown. 6.91.. and Birley. J.L.'l‘. Crises and life changes and the onset of schizophrenia. Journal of Health andSocial Behavior 9:203—214. 1968. Brown. 6.91.. and Harris. '1‘. Social Origins of Depresdon: A Study of Psychiatric Disorder in Women. New-York: Free Pres. 1978. Brown. G.W.; Harris. T.O.: and Pete. J. Life events and psychiatric dis— orders” . Part 2. Nature of the causal link. Psychological Medicine 3:159— 6. 1973. Bum, A.R. Ischaernic heart disease mortality and the business cycle in Australia. American Journal of Mic Health 69:772-781. 1979. Burke. R.J.. and Weir, T. Marital helping relatiommips: The moderators between stress and well-being. Journal of thology 95:121—130. 1977. Cannon, W. B. Bodily Changes in Pain. Hunger, Fear and Rage. New York: D. Appleton and Company. 1929. Caplan. G. Principles of Preventive Psychiatry. New York: Basic Books. 1964. Caplan. G. Mastery of stress: Psychological aspects. American'Journal of Psychiatry 138:413—420. 1981. Cassel. J. The relation of the urban environment to health: implications for prevention. Mount Shai Journal of Medicine 40:539—550. 1973. Cassel, J. Psychosocial processes and “stress": 'lheoretical formulation. International Journal of Health Services 4:471—482. 1974a. Cassel, J. An epidemiological perspective of psychosocial factors in disease etiology. American Journal of Public Health 64:1040—1043. 1974b. Cassel. J. The contributim of the social environment to host resistance. American Journal of Epidemiology 104:107—123. 1976. Catalano. R.. and Dooley. C.D. Economic predictors of depressed mood and stressful life events in a metropolitan community. Journal of Health and Social Behavior 18. 292—307. 1977. Chiribosa. D..;A Coho. A.; Stein.J..;A andRoberts, J. Divorce.su'essand Astudy in helpseekin; behavior. Journal of Divorce 3. 121—135.1979. Chirlhoga. D.A.. and Order. 1.. Stress responses among divorcing men and men. Journal of Divorce 195-106. 1977. Chiriboga. D.A.; Roberts. J.; and Stein. J.A. Psychological well-being during marital separation. Journal of Divorce 2:21—36. 1978. 101 Clo-1. PJ. A checklist for life event research. Journal of We Research 24:199—207. 1980. Cobb. S. A model for life events and their consequences. in: Dohrenwmd. 3.5.. and Dohrenwend. B.P.. eds. Stressful Life Events: Their Nature and Effects. New York: John Wiley 8: Sam. 1974. Cobb. 5. Social support as a moderator of life stress. Psychosomatic Medicine 38:300-314. 1976. Coddington. 11.0. The significance of life events as etiologic factors in the diseases ofchildren. l—Ameyafprofeasionalworkers. Journal of Psychosomatic Research 16:7—18. 1972a. Coddington. R.D. The significance of life events as etiologic factors in thediseasesofchildren. il-Astudyofanormalpopulation. Journal of Psychosomatic Research 16:205—213. l972h. Cohen. F. Personality. stress. and the development of plwsical illness. in: Stone. G.C.; Cohen. F.; and Adler. N.E.. eds. Health Psychology—A Hmdboak. San Francisco: Jossey—Bass. 1979. Cohen. F.. and Lazarus. R.S. Coping with the stresses of illness. in: Stone, G.C.; Cohen. F.; and Adler. N.E.. eds. Health Psychology—A Hand— book. San Francisco: Jossey—Bass. 1979. Cooke. 0.3.. and Greene. .l.G. Mes of life events in relation to symp— toms at the climacterium. Journal of Psychosomatic Research 25:5—11. 1981. Cooper. 8.. and Sylph. J. Life events and the onset of neurotic illness: An investigation in general practice. Psychological Medicine 3:421—435. 1973. Craso. M.A. Psychopathology in married couples. Psychological Bulletin 77:114—128. 1972. Crandall. J.E.. and Lehman. R.E. Relationship of stressful life events to social interest, locus of control. and psychological adjustment. Journal of Conndting and Clinical Psychology 45:1208. 1977. Crissey. M.S. Prevention in retrospect: Adoption follow—up. in: Albee. G.W.. and Joffe. J.M.. eds. Primary Prevention of Momthology. Hanover, N.H.: University Press of New England. 1977. Crook. T.. and Eliot. J. Parental death during childhood and adult depres- sion: A critical review of the literature. Psychological Bulletin 87:252— 259. 1980. Dean. A., and Lin. N. The stress—buffering role of social support: Pro— blems and prospects for systematic investigation. Journal of Nervous mdMental Disease 165:403—417. 1977. - Dohrenwend. B.P. Stressful life events and psychopathology: Some issues of theory and method. 111: Barrett. J.E.; Rose, R.M.; and Klerman, 6.1... eds. Stress and Mental'Disorder. New York: Raven Press. 1979. Dohrenwend. B.P.; Shrout. 9.8.: Egri, 0.: and Mendelsohn. F.S. Nonspecif- ic psychological distress and other dimensions of psychopatholog: Measures for use in the general population. Archives of General Psychi— atry 37:1229—1236, 1980. Dohremvend. 8.5. Social mess and community psychology. American Journal of Community Psychology 6:1-14. 1978. Dolmmwend. 8.5.. and Dohrenwend. B.P. Stressful Life Evans Their Nature and Effects. New York: John Wiley 5. Sons. 1974. Dol'n'enwmd. 8.5.. and Dohreuwend. B.P. Some issues in research on stressful life events. Journal of Nervous and Mental Disease 166:7—15. 1978. 102 Mum 8.8.. and Dohrenwend. B.P. Life stress and psychopatholog. in: Regier. D.A.. and Allen. 6., eds. Ride Factor Research in the Maior Mental Disorders. DHHS Pub. No. (ADM) 81-1068. Washington. D.C.: Supt. of Does.. U.S. Govt. Print. om. 1981a. Dohreuwend. 8.5.. and Dolreawend. B.P.. eds. Stressful Life Events and Their Contexts. New York: Prodist. 1981b. Dolreuwend. 8.5.; Krasnoff. L.; AW. A.R.: and Dohrenwend. B.P. Exemplification of a method for acalim life events: The PERI life events scale. Journal of Health and Social Bdnavior 19:205-229. 1978. Dooley. D.. and Catalano. R. Scanmic. life. and disorder changes: Time— series anabnes. America: Journal of Community Psychology 7:381-396. 1979. Dooley. D.. and Catalano. R. Bomomic change as a cause of behavioral disorder. Psychological Mletin 87:450—468. 1980. Dinner. D.L.; Patrick. V.; and F‘ieve. R.R. Life events at the onset of bipolar affective illness. American Journal of Psychiatry 136:508—511. 1979. Eaton. W.W. Life events. eccial supports. and psychiatric symptoms: A re— analysis of the New Haven data. Jownal of Health and Social Behavior 19:230—234. 1978. Egbert. L.D.; Battit. G.E.; Welch. C.E.; and Bartlett. MK. Reduction of postoperative pain by encouragement and instruction of patients. New England Journal of Medicine 270:825—827. 1964. Eisenberg. L. A research framework for evaluating the promotion of mental health and prevention of mental illness. Public Health Reports 96:31—19. 1981. Reel. 6.1.. The clinical application of the biopsychosocial mode. American Journal of Psychiatry 137:535—544. 1980. Erikson. K.T. Everything in Its Path: Destruction of Community in the Mfalo Creek Flood. New York: Simon and Schmter. 1976. Felner. R.D.; Ginter. M.; and Primavera. .l. Primary prevention during school transitions: Social support and enviromnental structure. Ameri— can Journal of Community Psychology 10:277—290. 1982. Figley. C.R.. ed. Stress Disorders Among Vietnam Veterans Theory, Research and Treatment. New York: Brunner/Mazel. 1978. Finlayson, A. Social networks as coping resources: Lay help and consul— tation patterns med by women in husbands' post-infarction career. Social Science and Medicine 10:97—103. 1976. Frank. J.D. Persuasion and Healing. Baltimore. Md.: Johns Hopkins Press. 1973. Friedman. M.. and Roseman. R.H. Type A Behavior and Your Heart. New York: Knopf. 1974. Friedman. R.R.. and Cohen. K. The peer support poup: A model for deal— ing with the emotional aspects of miscarriage. Group. the Journal of the Eastern Group Psychotherapy Society 4:42—48. 1980. Gallin. R.S. Life difficulties. coping. and the use of medical services. Culture. Medicine. aid Psychiatry 4:249-269. 1980. Gersten. J.C.; Lananer. T.S.; Elisabet-g. J.G.; and Crack. 1.. Child behavi— orandlifeevents:0ndesirablechangem-changeperae.h:Doln-en— wand. 3.5.. and Dohreuwend. B.P.. eds. Stressful Life Events: Their Nature and Effects. New York: John Wiley 8: Sum. 1974. Gersten. J.C.; Lanmer. T.S.; Eisenberg. J.C.; and Simcha~Fa3an. 0. An evaluation of the etiologic role of stressful life—change events in psy— chological disorder. Journal of Health and Social Behavior 18:228—244. 1977. 103 mean; one-e University Press. 1972. cm 3.1... and Connect. 6!. Life events andmhsequentillness. America Journal of Epidemiology 104:146—158. 1976. Gottlieb. B. Mutual-ban pm: Members' viewsoftheirheneflts and of roles for mm. Prevention in Hm Services 1:55—67. 1982. Grad. J.. andSaimbuI-y. P. Musdngthe counma'dtypsychiatricservice hChichester: Results. llilbank Memorial Find Quarterly 44: 246—278. 1966. Grant..;l Gust. M.: andYager. J. Scalinsoflife eventshypsychiatric patients and normals. Journal of Pochosomatic Research 20: 141—149. 1976. Granvold. D.K.: Pedler. L.M.; and Schellie. 5.6. A study of sex-role espectancy and female post-divorce aluminum. Journal of Divorce 2:383—393. 1979. Helios-1n. EA. Life—cycle stressors: An overview of research. Family and CWO Mental Health 2:71-101. 1980. m. 1.5.; Ream. 5.: Raitz. R.; Rampart. M.; an Coddinstm. R.D. magnificanceeflife eventsasoentributinzfsctorsmthediseasesof children. Journal of Pediatrics 83:119—123. 1973. Heller. M. Preventive mental health services for families new to the cormmmity. Hospital and Comrmnity Psydiiatry 26:493-494. 1975. Henderson. 5. Social relationships, advcmity and nemesis: Analysis of motive observations. Britten Journal of Psychiatry 138:391-398. 1981. Herman. S.J. Women. divorce and suicide. Joumai of Divorce 12107—117. 1977. Hetherinzton. E.M.; Cox. M.: and Coat. R. Divorced fathers. Family Coordinator 25:417—428. 1976. Hetherinston. E.M.: Cour. M.; and Cox. R. The aftermath of divorce. in: Stevens. J.H.. Jr.. and Matthews, M.. eds. Mother—Child. Father-Child Relations. Waddngton. D.C.: National Association for the Bducatim of Young Children. 1977. Hinkle. LE... Jr. nae effect of mute to cultm-al change. social change. and changes in interpersonalselationships on health. In: Dohrenwend. 8.5.. and Dohremwend. B.P.. eds. Stressful Life Events Their Nature and Effects. New York: John Wiley 8- Sons. 1974. Hoffman. 1. “Parental Adaptation to Fatal Illness in a Child." Unpublished doctoral dissertation. University of Chicago. 1972. Holahan. C.J.. and Moos. R.H. Social wt and psychological distress: A longimdinal analysis. Journal of Almormal Psychology 90:365—370. 1931. Holmes. T.H. Development and application of a quantitative measure of life change magnitude. in: Barrett. J.B.: Rose. R.M.; and Klerman. 6.1... eds. Streu mdllerttal Disorder. New York.- Rsven Press. 1979. Holmes. T.H.. and Manda. M. Life change and illness mentibility. in: Dohrenwend. 3.8.. and Dohrenweufl. B.P.. eds. Stressful Life Events: TheirNature and Effects. New York: John Wiley 3. Sons. 1974. Holmes. T.H.. and Rahe, R.H. The social readjastment rating scale. Journal of Psychosomatic Research 11:213—218. 1967. 104 Horowitz. M.; Wilner. N.: and Alvarez. VI. Impact of evmt scale: A measure of mbjective stress. Psychosomatic Hediche 51:209—218. 1979. Horowitz. M.J.; Schaefer. 5.; Hiroto, D.; Wilner, N.; and Levin. B. Life event questionnaires for measuring wenmtive mm. Psychosomatic Medicine 39:413-431. 1977. Horowitz. 18.1.. and Solomon. GJ’. Delayed Itreu respome syndromes in Vietnam veteran. In: May, C.R.. ed. Stress Disorder: Am Vietnam Veterans. New York: Banner/Mam. 1978. Budgets. R.W.; Mal-rim 1.1L; and Barchha. R.G. Life events and «net of primary affective disorders. Archives of Gm Poehiatry 16:134—145. 1967. lacoe. 1. Community psychology and the competent ooxmmmity. American Psychologist 29:607-613. 1974. Jacobs. D.: Charles. 13.; Jacobs. 1:; Weinstein. H.: and Ham D. Prepara— tion for treatment of the disadvantaged patient: Effects on disposition and outcome. American Journal of Orthopwchiatry 42:666—674. 1972. Jacobs. S.C.; Pmsofl'. B.A.; and Paykel. 8.5. Recent life events in echizoflrenia and depression. Psychological Medicine 4:444—453. 1974. Jenklm. C.D.; 'mthill, R.W.: Tamenbaum 5.1.; and Kirby. C. Social etressors and excess mortality from hypertensive diseases. Journal of Human Stress 5:29-40. 1979. Johmon. 1.8.. an! Leventhal. H. The effects of accurate expectations and behaviwalimtmctiommreactiomthrhganaimmedical examin- tion. Journal of Personality and Social Psychology 29:710—718. 1974. Johmon. 1.11.. and Samson. 1.6. Moderator variables in life stress re— search. In: Sarason. 1.6.. and Spielberser. C.D.. eds. Stress “Anxiety. Vol. 6. Wilmington. D.C.: Hemisphere, 1979. Kasey. J.R.; Vivace. 1.; and Lutz. W. Mental health primary prevention: The role of parent mutual mpport m. American Journal of Mic Health 71:166—167. 1981. Kale. “LL. and Stenmark. D.E. A comparison of four life event scales. American Journal of Community Psychology 11:451—458. 1983. Kaplan. D.M.. and Mason. EM. Maternal motions to premature birth viewed as an acute emotional disorder. American Journal of OrthopsyL chiatry 30:539—552. 1960. Kasl. S.V. Mortality and the laziness cycle: Some questions about research strategies when utilizing macro—social and ecological data. American Journal of Mic Health 69:784—788. 1979. Kelly, 1. Divorce: The adult perspective. In: Wolman. 3.. and Stricken G.. eds. Handbook of Developmental Psychology. New York: Prentice—Hall. 1982. Kelly. 1.8., and Walla-stein, LS. Brief interventims with children in divorcing families. American Journal of Orthopsychiatry 47:23—39. 1977. Keaeler. R.C. Stu-ea. nodal status. and psychological dim. Journal of Health and Social Behavior 20:259-272. 1979. Kimtrax W..andRoaser. R. Disaster: Effectsonrnentalandplwrlcal atate. Journd of Pochosomatlc Research 18:437-456. 1974. Kit-m. G.C.; inpata. H.Z.: Hohns. “ML; and Meyer-ins. S.Vl. Divorces andwidovrl:5imi1ar1tieaanddiflm.AmericalJanalofOrtho— psychiatrySO:291-301. 1980. Kitson. G.C., and m 11.8. The impact of divorce on adults. Calculation Court: RM” 15:20-24. 1977. Klanen. D.; Roth. A.: and Hormtra. R.K. Perception of life events as aim or losses in a community survey. Journal of Community Psychol— ogy 2:330—336. 1974. 105 Kobasa. S.C. Penonality and resistance to illness. American Journal of Community Psychology 7:413—423. 1979. Langer. 5.1.; Janis. LL; and Welter, LA. Reduction of psycholodcal stress in surgical patients. Journal of Emerirnental Social Psychology 11:155—165. 1975. Lamar. ’1‘.S. A twenty—two item screening score of psychiatric symptoms indicating impairment. Journal of Health and Hurnm Behavior 3:269— 276. 1962. Lazarus. R.S. Psychological Stress and the Coping Process. New York: McGraw—Hili. 1866. Lef‘f. Ml; Roatch. J.F.; and Bunney. W.E. Envircmnental factors pre— ceding the onset of severe depression. American Journal of Psychiatry 33:293—311. 1970. Lieberman. M.A.. and Barman. L.D. The impact of self-help grows on Mdow‘s mental health. National Reporter 4:2—6. 1981. Litton. R.J.. and Olson. E. The human meaning of total disaster: The Buffalo Creek experience. Psychiatry 39:1—18. 1976. Lin. N.; Simeone. R.S.; Easel. W.M.; and K110, W. Social support, stressful life events. and illness: A model and an empirical test. Journal of Health and Social Behavior 20:108—119. 1979. Undexnann. E. Symptomatoloy and management of acute grief. American Journal 0! Psychiatry 101:141-148. 1944. Lipson. J. Effects of a support group on the emotional impact of cesarean childbirth. Prevention in Hianm Services 1:17—29. 1982. Locke. S.E. Stress. adaptation. and immunity: Studies in humans. General Hospital Psychiatry 4:49—58. 1982. Lowenthai. M.F.. and Haven. C. interaction and adaptation: intimacy as a critical variable. American Sociological Review 33:20—30. 1968. Lynch. JJ. The Broken Heart. New York: Basic Books. 1977. MacMahon. 8.. and Pugh. ’1‘.F. Epidemiology: Principles and Methods. Boston: Little, Brown. 1970. Marin. P.A. Community Psychology: Concepts and Applications. New York: 'lhe Free Press. 1978. McGrath. J.E.. ed. Social and Psychological Factors in Stress. New York: Holt. Rinehart 8: Winstai. 1970. Medalie. J.l-i.. and Goldbourt. U. Angina pectoris among 10.000 men. American Journal of Medicine 60:910—921. 1976. Meyer. A. The life chart and the obligation of specifying positive data in psychopatholodcal diagnosis. In: Winters. E.E.. ed. The Collected Papers of Adolf Meyer. Vol. III. Medical Teaching. Baltimore. Md.: Johns Hopkins Prm 1951. Miller. P.M.. and lnghun. J.G. Reflections on the life—events—to—illness link with some preliminary ftxiings. in: Sarason. 1.6.. and Spielberger. C.D.. eds. Stress and Anxiety. Vol. 6. Washington. D.C.: Hemisphere. 1979. Minde. K.K.; Shosenberg. N.E.; and Marten. P.L. The effects of self-help minamattremrseuymmaternalautonomyandcaretaflng style 1 year later. in: Bend. L.A.. and Joffe. J.M.. eds. Facilitating Infant and Early Childhood Development. Hanover. N.H.: University Press of New England 1982. Minde. K.; Shooenberg. N.; Marten. P.; 'lhompson. 1.; Riplw. 1.; and Burns. S. Self-help gmups in a premature nursery—a controlled evalu— ation. Journal of Pediatrics 96:933-940. 1980. 106 Mueller. D.P.; Edwards. 0.91.; and Yarvis. R.M. Stressful life events and tric symptomatology: Change or undesirahility. Journal of Health and Social Behavior 18:307—317. 1977. Mueller. D.P.; Edwards, D.W.; and Yarvis, RM. Stressful life events and community mental health center patients. Journal of Nervous and Men— tal Disease 166:16-24. 1978. Myers, J.K.; Lindenthal. 1.1.; and Pepper. MP. Life events. social inte— gration and psychiatric symptomatoloy. Journal of Health and Social Behavior 16:421—427. 1975. Newman. C.J. Children of disaster: Clinical observations at Buffalo Creek. American Journal of Psychiatry 133:306—312. 1976. Oliver. J .M.. and Pomicter. C. Depression in automotive assembly—line workers as a function of unemployment variables. America! Journal of Community Psychology 9:507-512. 1981. Parad. H.J.. ed. Crisis Intervention: Selected Readings. New York: Family Service Associatim of America. 1965. Parad. i-i.J.; Resrzik. H.L.P.; and ‘Parad. LG. Emergency and Disaster Management. Bowie. Md.: Charles Press. 1976. Parkes. C.M. Evaluation of a bereavement service. Journal of Preventive Psychiatry 1:179-188. 1981. Panel. E.S. Contribution of life events to causation of psychiatric illness. Psychological Medicine 8:245—253. 1978. Paykel. 8.5. Causal relationships between clinical depression and life events. in: Barrett. 1.8.; Rose. R.M.; and Klerman. 6.1... eds. Stress and Mental Disorder. New York: Raven Press, 1979. Paykel, E.S.; Myers. J.K.; Dienelt. M.N.; Kierman. G.L.; Lindenthal. 1.1.; and Pepper. MP. Life events and depression. Archives of General Psychiatry 21:753—760. 1969. Pearlin. L.I., and Johnson. .l.S. Marital status. life—strain; and depressim. American Sociological Review 42:704—715. 1977. Pearlin, LL. and Schooler. C. The structure of coping. Journal of Health Md Social Behavior 19:2—21, 1978. Perez. R. Provision of mental health services during a disaster: The Cuban irmnigration of 1980. Journal of Community Psychology 10:40—47. 1982. Porritt. D.. and Bordow. S. Some implications on an experimental trial of crisis intervention with road tram iii—patients. Australian Journal of Alcoholism and Drug Dependence 3:136—139. 1976. Rahkin. LG. Stressful life events and schizophrenia: A review of the research literature. Psychological Bulletin 87:408—425. 1980. Rabkin, J .G.. and Struening. E.L. Life events. stress. and illness Science 194:1013—1020. 1976. Runoff. L. Sex differences in depression: The effects of occupation and marital status. Sex Roles 1:249—265. 1975. Rahe. R.H. The pathway between subjects' recent life changes and their near-future illness reports: Representative results and methodological issues. in: Dots-mend. 8.5.. and Dohrenwend. B.P., eds. Stressful Life Events: Their Nature and Effects. New York: John Wiley & Sons. 1974. Rahe. RJ-i. Life change evemsand mental illness: An overview. Journal of ”term Stress 5:2—10. 1979. Rafinei. B. Preventive intervention with the recently bereaved. Archives of Gmral Psychiatry 34:1450—1454. 1977. Raphael. B. Mom-hing and the preventim of melancholia. British Journal of Medical Psychology 51:303-310. 1978. 107 Raschke. HJ. The role of social participation ’in post—separation and post-divorce adjustment. Journal of Divorce 1:129—140. 1977. Redick. R.W.. and Johnna. C. Marital Status. Living Arrangements aud Family Characteristics of Admissions to State and County Mental Homitals cud Outpatient We Clinics. United States 1970. Statistical Note 100. National Bustimte of Mental Health. Washingm D.C.: Supt. d Dec" U.S. Govt. Print. Off" 1914. Reid. D.D. Predfltatiog mm factors in the occurrence ofunental disorders: Modal evidence. In: More. 5.14.. and Huxley. 14.. eds. Course: of Mental Disorders: A Review of Epidemiological Knowledge. 1959. New York: mm Memorial Fund. 1961. Robertson. N.C.'1here1atiomhipbetweenmarita1statmatdtheriskof paychiatflc referral. Brititu Journal of Podiatry 124:191—202. 1974. Rogers. 1.: Vachm. M.L.S.; man. W.A.; Sheldon. A; and Freeman. S.J.J. Aaelf—helppuwamforwldawsasanimiepeufient ootxmmityservice. Hospital & Commlnity Wry 31:844-847. 1980. Rankin. M. Coping with life changes—A preventive social work approach. American Journal of Community Psychology 10:331—340. 1982. Rowland. K.F. Environmental events predicting death for the elderly. Psychological Bulletin 84:349—372. 1977. Ryan. W. Preventive services in the aocial context: Power. pathology. and prevention. In: Bloom, 8.1... and Bock. D.P.. eds. Preventive Services in Mental Health Programs. Boulder. Colo.: Western Interstate Commis— sion for Higher Education. 1967. Sandler. I.N.. and Block. M. Life was and maladaptation of children. American Journal of Community Psychology 7:425—440. 1979. Sandler. l.N.. and Ramsay. T.B. Dimersional analysis of children's stress— ful life events. American Journal of Community Pathology 8:285—302. 1980. Sank. L. Community disasters: Primary prevention and treatment in a health maintenance organization. American Maxis-t 34:334—338. 1979. Saracen. LG. Life stress. self—preoccupation and social mpports. In: Sara— aon. 1.6.. and Splelberger. C.D.. eds. Stress aud Anxiety. Vol. 7. WW D.C.: Hemisfliere. 1980. Sarason. I.G.; Johann, J.H.; and Siegel, J.M. Assessing the impact of life changes: Dwdmmt of the life experiences survey. In: Sarason. 1.6.. and Spielberger. C.D.. eds. Stress and Anxiety. Vol. 6. Washington. D.C.: Hemisflmre, 1979. Schaeffer. M.H.; Kliman, G.W.; Friedman. M.J.; and Pasquariella. B.G. Children in foster care: A preventive service and research program for a hiqu risk population. Journal of Preventive Psychiatry 1:47-56. 1981. Schleal. A.P.; Teichman. A.; Mendels. 1.; and Di Giacomo. J.N. The role ofstreas namedpitatingtactor ofpaycluiauicillnes.8ritishloumal of Psychiatry 130:19—22. 1917. ' Setye. H. The general adaptation syndrome and the diseases ofadaptation. Journd of Clinical Endocrinology 6:117—230. 1946. 59170. ii. The Siren ofLUe. New York: McGraw—Hm. 1956. Selzer. ML. and Vlneknr. A. Life events. subjective stress and traffic accidents. America Journd of Pryciuiatry 131:903—906. 1974. Silverunan. ER. Servicummewidommnepsinamunof preventive intervention. Community Malta! Health Journal 3:37-44. 1967. Silver-man. P.R. W and preventive intervention. The Family Coordinator 21:95-102. 1972. 108 Silverman. P.R.; MacKenzie. D.; Pettipas. M.; and Wilson. [-2. Helping Each Other in Widowhood. New York: Health Sciences. 1975. Silver-man. P.R.. and Murrow. H.G. Mutual help during critical role transitions. Journal of Applied Behavioral Science 12:410—418. 1976. Singh. 8.. and Raphael, B. Postdisaster morbidity of the bereaved: A pos— sible role for preventive psychiatry? Journal of Nervous and Mental Disease 169:203—212. 1981. Skipper. J.K.. and Leonard. R.C. Children. mass. and hospitalization: A field capes-ham. Journal of Health and Social Behavior 9:275—287. 1968. Smith. W.G. Critical life-events and reventim strategies in mental health. Archives of General Psychiatry 25:103-109. 1971. Spanier. 6.8.. and Cam. Rf. Adjxtment to separation and divorce: An analysis of 50 case studies. Journal of Divorce 2: 241—253. 1979. Stole. L.; banner. ‘1‘.S.; Michael. S.T.; Opler. M.K.; and Rennie. T.A.C. Mental Health in the Metropolis. New York.- McGraw—Hill. 1962. Steinberg. H.R.. and Durell. J. A stressful social situation as a precipitant of schizoplrenic symptoms: An epidemiological study. British Journal of Pachiatry 114:1097—1105. 1968. Stem. G.M. The Mfalo Creek Disaster: The Story of the survivors' Unprecedented Lawmit. New York: Random House. 1976. Suarez. J.M.; Weston, N.L.; and Hartstein. N.B. Mental health interven— tiom in divorce proceedings. American Journal of Orthopsychiatry 48:273—283. 1978. Syme. 5.1... and Torts, C.P. Epidemiologic research in hypertension: A critical appraisal. Journal of Human Stres 4:43-48, 1978. Tauss. W. A note on the prevalence of mental distm-bance. Australian Journal of Psychology 19:121-123. 1967. Tennant. C.. and Andrews. G. A scale to measure the stress or! life events. Australian mid New Zealmd Journal of Psychiatry 10:27-32, 1976. Tennant. C.. and Andrews. G. The pathogenic quality of life event stress in neurotic impairment. Archives of General Psychiatry 35:859—863. 1978. Tennant. C.; Smith. A.: Bebbington. P.; and Hurry, J. The contextual threat of life events: The concept and its reliability. Psychological Medicine 9:525-528. 1979. Thomas. D.S.. and Locke. B.Z. Marital status. education and occupational differentials in mental disease. Milbank Memorial Fund Quarterly 41:145-160. 1963. Titchener. 1.1.... and Kapp. F.’l‘. Family and character change at Buffalo Creek. American Journal of Psychiatry 133:295—299. 1976. Totman. R. What makes life events streafiul? A retrospective study of patients Who have suffered a first myocardial infarction. Journal of Psychosomatic Research 23:193—201. 1979. Totman. R.; Kiff. 1.; Reed, 5.8.; and Craig. J.W. Predicting experimental colds in volunteers from different measures of recent life stres. Journal of Psychosomatic Research 24:155—163. 1980. Udell. B.. and Hormtra. R.K. Hospitalization and presenting problems. Comprehensive Psychiatry 16:573-580. 1975. Ursano. RJ. The Viet Nam era wisoner of war: Precaptlvity personality and the development of psychiatric illness. American Journal of Psychiatry 138:315—318. 1981. Ursano. R.J.; Boydstun. J.A.; and Wheatley. R.D. Psychiatric illness in us. Air Force Viet Nam prisoners of war: A five—year follow-up. American Journal of Psychiatry 138:310-314. 1981. 109 Vachon. M.L.S.; Lyell. W.A.L.; Rogers, J.; Freedman—Letofsy. K.; and Freeman. 5.1.1. A controlled study of self—help intervention for widows. American Journal of Psychiatry 137:1380—1384. 1980. Vernon. D.T.A.; Schulman. .l.l...; and Foley. LM. Changes in children's behavior after hospitalization. American Journal of Diseases of Chil— dren 111:581-593. 1966. Videka$herman. L. Effects of participation in a self—help group for be— reaved parents: Compassionate Friends. Prevention in Human Services 1:69—77. 1982. Vinokur. A.. and Selzer. M.L. Desirable vet-ms madesirable life events: Their relationship to mess and mental distress. Journal of Personality and Social Psychology 32:329—337. 1975. Walker. K.N.; MacBride. A.; and Vachon. M.L.S. Social support networks and the crisis of bereavement. Social Science and Medicine 11:35—41. 1977. Waller-stein, 1.5.. and Kelly. 1.8. Effects of divorce on the visiting father— child relationship. American Journal of Pachiatry 137:1534—1539. 1980. Warheit. GJ. Life events. coping. stress, and depressive symptomatolog. American Journal of Psychiatry 136: 502—507. 1979. Weingarten. M.A.. and Feuchtwanger. D. Group meetings with new mothers in a family practice: Report of a pilot project. Imeli Annals of Psychiatry and Related Disciplines 16:232—242. 1978. Weiss. R.S. The emotional impact of marital separation. Journal of Social Issues 32:135—145. 1976. Weiss. R.S. Going It Alone. New York: Basic Books, 1979. Wertlieb. D.: Budman. 5.; Demby. A.; and Randall. M. The stress of mari— tal separation: Intervention in a health maintenance organization. Psychosomatic Medicine 44:437-448. 1982. Wildman, R.C.. and Johnson, D.R. Life change and Langner's 22—item mental health hides: A study and partial replication. Journal of Health and Social Behavior 18:179—188. 1977. Williams, T. Post—Traumatic Stress Disorders of the Vietnam Veteran: Observations and Recommendations for the Psychological Treatment of the Veteran and His Family. Cincinnati. Ohio: Disabled American Veterans. 1980. . Witldn. M.H. Psychosexual counseling of the mastectomy patient. Journal of Sex 4. Marital Therapy 4:20—28, 1978. Wolfer. J .A.. and Visintainer. M.A. Pediatric surgical patients' and par— ents' stress responses and adjustment as a function of psychologic prep— aration and stress-point nursing. Nursing Research 24:244—255. 1975. Wolff. H.G. Life stress and bodily disease—a formulation. In: Wolff. H.G.; Wolf, S.G.. Jr.; and Hare. C.C.. eds. Life Stress and Bodily Disease. Baltimore: Williams and Wilkins Compaw. 1950. Wolkon. G.H.. and Moriwald. S. The ombudsman: A serendipitous mental health intervention. Community Mental Health Journal 13:229-238. 1977. Wood. P.E.: Milligan. M.; Christ. 0.; and Liff, D. Group oomneling for cancer patients in a community hospital. Psychosomatics 19:555-561. 1978. Yalom. LD. Existential Parchotherapy. New York: Basic Books. 1980. Zautra. A.. and Beier. E. The effects of life crisis on psychological adjustment. American Journal of Community Psychology 6:125—135. 1978. Zubin. 1.. and Spring, 8. Vulnerability—a new View of schizophrenia. Journal of Abnormal Psychology 86:103—126. 1977. 110 U.S. GOVERNMENT PRINTING OFFICE : 1985 O - 479-088 DEPARTMENT OF sP‘E'clA L FOURTH c LA-ss M'Ai L HEALTH & HUMAN SERVICES POSTAGE AND FEES PAID PHSlADAMHA Public Health Service PERMIT NO. 6-283 Alcohol, Drug Abuse, and Mental Health Administration RockVille MD 20857 Official Business Penalty for Private Usa. $3100 .LL NOTICE OF MAILING CHANGE El Check here, if‘you wish to discontinue re,Ceiving thi’SZt-‘ypezof publication; El Cheek here if- yo‘ur address has changed and you Wish to colnt‘inUe teoeiv‘ing this type of publication. (B'esure-to furnish yo’urcomplet'e, address including zip code.) Tearitbff-covér‘with address label 'and publication numberstill affixed and send to: National institute: of Mental Health Science Communication Branch 560.0'Fi'she'rs Lane (Room 1-5-99.) Flockvjlle’, Maryland 29857 DH HS Publication No. (ADM) 85-1385. Printed 13985 GENERAL UBBAflY-ILC.BEHKELEY BBUDSLBL?”