Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Depression in wives of alcoholics : the role of perceived social support from family and friends Wilks-Reynolds, Janet H. 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1996-0601.pdf [ 7.99MB ]
JSON: 831-1.0054014.json
JSON-LD: 831-1.0054014-ld.json
RDF/XML (Pretty): 831-1.0054014-rdf.xml
RDF/JSON: 831-1.0054014-rdf.json
Turtle: 831-1.0054014-turtle.txt
N-Triples: 831-1.0054014-rdf-ntriples.txt
Original Record: 831-1.0054014-source.json
Full Text

Full Text

DEPRESSION IN WIVES OF ALCOHOLICS: THE ROLE OF PERCEIVED SOCIAL SUPPORT FROM FAMILY AND FRIENDS by  JANET H, WILKS-REYNOLDS B.S., Brigham Young University, 1970 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR T H E D E G R E E OF M A S T E R OF ARTS in THE F A C U L T Y OF G R A D U A T E STUDIES Department of Counselling Psychology  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH C O L U M B I A September 1996 © Janet H. Wilks-Reynolds  In  presenting  this  thesis  in partial  fulfilment  of the requirements  for an advanced  degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further copying  agree that permission for extensive  of this thesis for scholarly purposes may be granted  department  or  by  his or  her  representatives:  publication of this thesis for fjnancial  It  is  understood  of  (Vt< <^4-e^£-6<^u<^^ x  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  '/OffraS^  U),  that  copying or  gain shall not be allowed without my written  permission.  Department  by the head of my  -  Wives of Alcoholics  ii  Abstract Perceived social support has been shown to moderate the incidence of depression in the face of adversity in a variety of populations, although this relationship has not been previously verified in a sample of wives of alcoholics. Theory in relation to social support and depression is explored. About twice as many women as men experience depression, and women are more often the providers than the recipients of social support, thus both social support and depression are gender-biased phenomena. The gendered aspects of these constructs are discussed and related to the experience of wives of alcoholics. Work, marriage, and friendships are explored as potential contexts of social support, and a history of professional treatment of wives of alcoholics is reviewed. Professional support has been lacking, and these women have been unjustly pathologized. In a sample of 116 wives of alcoholics, stepwise regression analyses were conducted for two measures of depression as the dependent variables, with perceived social support from family, perceived social support from friends, and various demographic variables as the independent variables. Perceived social support from family was found to explain the majority of the variance for both measures of depression. Perceived social support from friends and unemployment explained additional variance only for the B D I . Number residing in the house explained additional variance only for the S C L - 9 0 - R depression subscale. Family income explained additional variance for both measures.  Wives of Alcoholics  TABLE OF CONTENTS Abstract  ii  Table of Contents  iii  List of Tables  vi  List of Figures  vii  Dedication  viii  Chapter I  Introduction The Problem Purpose of the Study  Chapter II  1 7 8  Background Information Depression in Women Depression and Cognitive Process Depression as a Gendered Phenomenon Issues of Unequal Power Socialization Shapes the Evolving Self Issues of Dependence Family Roles and Intimate Ties Loss of Control Loss of Self Summary of Depression Social support Aspects of Social Support Social Support as a Modifier of Stress and Depression Interaction of the Personal and the Social Perceived Social Support Social Support as a Gendered Phenomenon Summary of Social Support Gender Norms and the Alcoholic Family System Contexts For Self-Definition Employment and Depression Marriage as a Context For Social Support Women Are More Vulnerable to Marital Stress Loss of Support Within the Marriage Alcohostages Marital Distress and Depression in Women Friendships as a Context For Social Support Theories Past and Present About Wives of Alcoholics The Disturbed Personality Hypothesis  10 11 12 12 14 14 15 16 17 20 20 21 21 22 24 26 28 32 33 38 39 41 41 43 45 46 47 50 50  Wives of Alcoholics  The Sociological Stress Theory The Psychosocial Theory Family Systems Theory Codependency Theory The Language of Pathology Socialized to be Codependent Preserving the Hierarchy of Patriarchy  Chapter III  Literature Review Social Support as a Modifier of Stress Recent Qualitative Studies of Wives of Alcoholics Wives of Alcoholics and Employment Summary Hypotheses  Chapter IV  98 98 99 99 100 101 102 103 104 106 107 108 110 111  Results Simple Statistics For M a i n Variables Correlation Matrix Multiple Regression Analyses  Chapter VI  66 66 76 89 93 97  Method Sample Screening Inclusion Criteria Exclusion Criteria M a i n Variables Social Support Depression Demographic Predictor Variables Instrumentation Perceived Social Support-Family and Perceived Social Support-Friends Beck Depression Inventory Symptom Checklist 90-Revised-Depression Subscale Data Analysis  Chapter V  53 55 56 58 59 61 62  114 114 115 117  Discussion General Findings For the M a i n Variables Hypothesis I Perceived Social Support From Family Perceived Social Support From Friends Hypothesis II Employment Status Religious Participation  121 121 124 124 125 125 125 126  iv  Wives of Alcoholics  Number of Children and Number Residing in the House Years L i v e d with Partner Race Age, Education Level, and Annual Family Income Limitations Possibilities For Further Research Generalizability Conclusions  References  127 128 128 129 130 133 137 138  142  Wives of Alcoholics  List of Tables 1. 2. 3. 4.  Descriptive Statistics For M a i n Variables Pearson-r Correlations Between Independent and Dependent Variables Results For Stepwise Regression Analysis With B D I as the Dependent Variable Results For Stepwise Regression Analysis With SCL-90-R-Depression as the Dependent Variable 5. Indicators of Depression Utilized by the B D I and the S C L - 9 0 - R Depression Subscale  115 116 * 117 ) ^  118 122  vi  Wives of Alcoholics vii  List of Figures 1. Mean SCL-90-R-Depression Scores For Each Income Category  120  Wives of Alcoholics viii  DEDICATION To my late husband and companion of nearly 25 years, Justin L . Wilks, the father of my six children, with and for whom I began this degree so that he could be released from his second job and spend more time with his family. But instead, he was released from this life at 7:40 a.m. Saturday, July 2, 1994, when cancer took him from us with little warning, at the age of 47. Never have I felt such soul-wrenching pain and grief. Now he can witness this completion in spirit only. He leaves a legacy of excellence in education, work, and service to others that won't be forgotten. To my new husband, Daniel R. Reynolds, who brought me happiness again and patiently endured his bride writing this thesis. His kind and gentle nature has made the unbearable bearable, and I am so grateful that he came into my life. To my Maker, who heard my cries and stayed with me in my darkest hour, giving me strength to rise to the occasion and persevere in the face of life's trials, even when I didn't know if I could make it through.  Wives of Alcoholics  1  CHAPTER I Introduction This is a study of perceived social support and how it is related to risk of depression in wives of alcoholics. According to Kelly (1955), individuals develop constructs to explain the people, events, and environmental contexts in their lives. Emotional states arise out of this complex process of constructing meaning within the context of a dynamic, ongoing interaction with circumstances and significant others. A situation is stressful to the extent that it is perceived as a threat, and thus the meaning subjectively assigned to the situation is the critical determinant of the effect it will have on the individual's internal emotional state (Turpin & Lader, 1986). The discussion that follows will focus specifically on factors relevant to perceived social support and depression that are particularly salient for wives of alcoholics as they make sense of their circumstances, which include the problem of their husbands' alcoholism. Depression is associated with themes of helplessness (Seligman, 1974), defeat (Beck & Weishaar, 1989), and loss (Jack, 1991). It is a gender-biased phenomenon, at least twice as common in women as men (Nolen-Hoeksema, 1995), with the difference being that women are more emotionally impacted by stressors (Kessler & McLeod, 1984) occurring to close others, especially to spouse (Turner & Avison, 1989). Women are socialized to feel responsible for others (Wethington, McLeod & Kessler, 1987), and the gender difference in incidence of depression is partly due to the cost of caring about those around them (Belle, 1982; Kessler, McLeod & Wethington, 1985). Women's sense of connection to others, derived via closeness with mother for an extended time in early development, fosters a moral sense of responsibility for others (Gilligan, 1993) and an increased ability to recognize their needs.  Wives of Alcoholics  2  Self esteem, a construct negatively correlated with depression (Shaver & Brennan, 1991), is related to the quality of attachments for women (Jack, 1991), who are socialized to need love (Radloff & Monroe, 1978). Because of this, they are more likely to derive self-esteem outside themselves, which renders them more vulnerable to depression (Jack, 1991). Secure attachment in a love relationship is particularly important to a woman, and in order to ensure it, she may progressively disconnect from herself and her own personal needs, acquiescing to the desires of her partner in a self-defeating way and silencing her own voice in the process (Jack, 1991). This is not uncommon among women experiencing depression. Depressed women have been found to be less likely to share in decisions and more likely to carry the burden of housekeeping and child care (Whisman & Jacobson, 1989). Predisposing factors for depression in women include severely threatening life events and the lack of an intimate, confiding relationship (Brown & Harris, 1978). Social support has been found to moderate the effect of negative life events in other populations of women (Brown, Andrews, Harris, Adler & Bridge, 1986), and it may operate by fostering a sense of purpose and hope, a belief that one can influence outcome, and an increased ability to adapt to change (Lloyd, 1995). Perceived social support is more related to outcome than actual support received (Cohen & Wills, 1985), perhaps because it captures the ongoing dynamic interaction of the individual with the social environment. It is a cognitive adaptation made by the individual that evolves over the course of a lifetime of social experience (Sarason, et al., 1991). Internal working models of self and other are created through interaction with early attachment figures, modified through ongoing interaction with the social environment, and  Wives of Alcoholics  3  utilized to interpret other people and new situations as they arise (Bowlby, 1988). Secure attachment and learning to trust foster confidence in later social interaction (Champion, 1995). Believing that others care and feeling accepted and acceptable influence the process of cognitive appraisal and meaning-making (Lloyd, 1995). Knowing one is loved reduces anxiety and facilitates exploration, reasonable risk-taking, and seeking help when needed (Sarason, Sarason & Pierce, 1990). Cognitions about self, others, and situations affect many aspects of helpseeking and help-giving (Brewin, 1995). Social support does not occur in networks of balanced mutuality. W o m e n generally provide substantially more social support than they give, a condition known as the support gap (Belle, 1982). Women are the preferred providers of social support and are most often sought by members of both sexes when needs arise (Buhrke & Fuqua, 1987). Large social networks have been found to actually be detrimental to homemakers due to the demands they face from others for support (Veiel, 1995). A depressed person is less likely to perceive social support, and less likely to utilize it due to fear of negative evaluation after self-disclosure (Lloyd, 1995). Rook (1984) found negative social interactions to be particularly powerful because they are given more credibility than positive social interactions as meaning is constructed. Lack of emotional support and negative comments from a spouse or intimate partner are particularly devastating for women, fueling their natural tendency to self-blame (Andrews & Brewin, 1990) and contributing to depression (Hooley & Teasdale, 1989). Depression in women is associated with having a poor marriage and a lack of close confidants (Roy, 1978). Marital stressors (Ilfeld, 1982) such as nonfulfillment of role obligations, lack of reciprocity, and a sense of not being accepted evidence a lack of  Wives of Alcoholics  4  emotional support in marriage, and when a woman feels let down by her spouse, risk of depression is significantly increased (Brown, et al., 1986). Such experiences are common for wives of alcoholics. The perceived inability to influence outcome is a major factor in depression and is experienced as helplessness and loss of control (Sarason, Sarason & Pierce, 1990) by these women. Interviews with wives of alcoholics in previous works (Asher, 1992; Banister & Peavy, 1994; Wiseman, 1991) give a sense of the meaning they make of their lives, perceptions about self and the availability of social support, and the state of their emotions. It is important to recognize the multi-level context within which these wives attempt to deal with the problem of their husbands' alcoholism. A s these women interact in a reciprocal way with the problem of alcoholism as it evolves, the meaning they assign to events is affected by cultural mandates in concert with family dynamics (Asher, 1992; Banister & Peavy, 1994; Wiseman, 1991) and prior life experience (Bowlby, 1969, 1973, 1980). The way in which wives of alcoholics define themselves is heavily influenced by alcoholinduced changes in the marital relationship (Asher, 1992: Wiseman, 1991), set against the backdrop of cultural expectations about women's responsibility for relationships and socialized prescriptions about appropriate female behavior. Banister and Peavy (1994) define culture as "the acquired ideas, beliefs, and knowledge that a particular group of individuals use to make meaning of their experience and generate cultural behavior" (p. 209). It is beyond the scope of this study to focus on the experience of male alcoholics, although it is acknowledged that they are also affected by the cultural context in which they live, as they engage in their own subjective process of meaning-making in light of socialized expectations about masculine behavior. The  Wives of Alcoholics  5  alcoholic family is a rigidly gendered system (Nol, 1991) in which drinking functions to relieve the rigidity imposed by gender prescriptions by helping family members as well as the alcoholic "to express or suppress aspects of self or behavior that run counter to their ideas about appropriate male and female behavior" (Bepko, 1986, p. 68). Wives of alcoholics are groomed by their own sex-role training to internalize the alcoholic's projection of responsibility for the drinking as well as other family problems (Wiseman, 1991), which may lead to chronic depression (Burnett, 1984). A s women, they are taught to measure their worth in terms of success in relationships. A s alcohol becomes the "other woman" (Asher, 1992), and the marriage relationship deteriorates, the wife of an alcoholic feels a sense of failure in her primary relationship and redefines herself more negatively within that context (Asher, 1992; Bannister & Peavy, 1994; Wiseman, 1991). W h i l e feeling responsible to make her marriage relationship work and judging herself harshly for not succeeding, she is often lacking important sources of support due to the social isolation brought on by her sense of shame and embarrassment at not being able to get her husband to stop drinking (Asher, 1992; Wiseman, 1991). Enacting social roles provides a sense of personal value (Thoits, 1983), and depression occurs when a role central to identity is disrupted (Oatley & Bolton, 1985). The marital role is often centra!to a woman's identity, yet i f she has an alternative context for self-definition, she is likely to be less devastated by the loss of the marital relationship. Employment, for example, has been found to build confidence, self-esteem, independence, and a sense of accomplishment for women in general (Tebbets, 1982), and wives of alcoholics in particular (Casey, Griffin & Googins, 1993).  Wives of Alcoholics  6  It is not unusual for the availability of social support to be reduced during the time of greatest need (Parry, 1995), as friends are alienated by the husband's alcoholism (Wiseman, 1991). For the wife of an alcoholic, self-esteem is progressively eroded, and a sense of failure and powerlessness may well contribute to depressed affect (Jack, 1991). In broader populations, social support has been found to moderate the negative effects of stress while increasing coping resources (Burchfield, Hamilton & Banks, 1982; Gertsel, Riessman & Rosenfield, 1985; Hibbard, 1985).  Social experience is very much reflected in well-being and mental health (Argyle, 1987), and emotions are related to social interactions (Gilbert, 1995). People feel good when they feel cared for, valued, and respected, and belong in a desirable social group. They feel badly if they are alienated and not cared for. Wives of alcoholics learn to suppress their emotions in order to survive. They often suffer physically and emotionally abusive behavior at the hands of the alcoholic (Bepko, 1988), and this operates to inhibit their direct expression of anger. They may feel confused by conflicting messages of approach and avoidance from alcoholic husbands (Burnett, 1984), and they commonly experience loss of trust, powerlessness, guilt, self-blame, and isolation (Banister & Peavy, 1994). Stressful life experience is often associated with the development of psychiatric disorder, a complex interaction that is moderated by personality variables, type of stressor, and resources available (Lloyd, 1995). Pretorius (1994) found negative life events to be positively correlated with depression, and support to be negatively correlated with depression. Social support plays an important protective role against depression (Henderson, 1992; Veiel, 1995), with both the  Wives of Alcoholics  7  interpersonal and the intrapersonal being salient dimensions of the support variable (Sarason, et al., 1990). Social support measured subjectively as perceived social support has demonstrated a stronger relationship to onset, course, and outcome of depression (George, Blazer, Hughes & Fowler, 1989) and other psychiatric symptoms (Billings & Moos, 1985; Brugha, 1995; Henderson, 1981; Henderson & Moran, 1983; Holahan & Moos, 1981) in the face of adversity than have objective measures of social support. Henderson (1981) states that "neurotic symptoms emerge when individuals consider themselves deficient in care, concern, and interest from others" (p. 397). Thus the meaning made of events is more predictive of outcome than objective reality.  The Problem Wives of alcoholics have generally experienced a lack of understanding and support from society and professionals and a loss of support in their marital relationships, while also being isolated by shame and a perceived loss of self-efficacy. Historically, professionals have been inclined to collude with the alcoholic's denial and projection of responsibility for his drinking onto his wife. The one exception is A l - A n o n , which strives to help the wife disengage from the alcoholic, yet the wife's attendance seems to have little effect on the husband's sobriety (Wiseman, 1991) and may be a mixed blessing because the wife is taught to accept the lifelong designation of being codependent (Asher, 1992). She must struggle to recover from the disease of codependence, which may be synonymous with attempting to recover from female socialization (Krestan and Bepko, 1991), even though the validity of codependence as a construct is questionable, and empirical substantiation is sorely lacking (Troise, 1992).  Wives of Alcoholics  8  Wives of alcoholics seem to find the greatest emotional support from other wives of alcoholics, but they, too, are overburdened and limited in resources. A s her husband's alcoholism worsens, support networks are lost, and she becomes more isolated. Loss of the emotional intimacy of the marriage relationship is particularly devastating for the wife of an alcoholic. "Alcoholism can cause many aspects of married life to 'come apart' as deficiency in one area affects another" (Wiseman, 1975, p. 174). In light of her socialization to be responsible for the emotional well-being of the family (Bepko, 1986), she is particularly vulnerable to selfblame; and marital distress is likely to have a serious impact on her psychological well-being, as she finds herself powerless to bring about the desired outcome of sobriety (Kobasa, 1979; Sarason, et al., 1990). Her sense of herself as a capable and valuable person is severely diminished, and she redefines herself more negatively as the alcoholism evolves. She becomes progressively more disconnected from her own needs and feelings as she suppresses them in order to survive. Due to her sense of shame, she may be inhibited from seeking outside help or perceiving social support as available to her, and depression can be seen as a logical possibility.  Purpose of the Study Interaction in an alcohol-embedded marriage, set within the cultural context of socialized expectations for appropriate female behavior, plays an important part in the way the wife of an alcoholic defines herself and the problems she faces. The dimension of perceived social support taps the dynamic product of person in process with environment over the course of time, reflecting the meaning she makes of her social environment. The purpose of this study is to evaluate perception of social support in wives of alcoholics as a moderator of depression in the face of the unique set of adverse conditions experienced by this population, as well as to explore  Wives of Alcoholics  demographic variables that may bear upon the relationship between perceived social support and depression. The demographic predictor variables under study are employment status, religious participation, number of children, number of persons residing in the house, years lived with husband or partner, race, age of wife, education level of wife, and annual family income.  9  Wives of Alcoholics  10  CHAPTER II Background Information To fully understand how the variables in this study are related in a population of wives of alcoholics, it is relevant to look at what is known about depression in women in general and the cognitive process that is associated with depression. Depression is at least twice as common in women than men (Nolen-Hoeksema, 1995), particularly i f they are married (Gove, 1972; Radloff, 1975), and thus depression will be discussed as a gendered phenomenon. The impact of depression in women w i l l be explored in terms of unequal power, the constraints of socialization, and issues of dependence for women. Roles that women enact in families and the importance to them of having an intimate, confiding relationship will also be discussed, along with loss of control and loss of self as elements of depression for women. Theory about social support and how it operates to mediate stress and reduce risk of depression w i l l be reviewed in objective and subjective terms, with the personal and the social in dynamic reciprocal interaction spawning perceptions about self and the social environment as a supportive resource. The effects of culturally-based gender prescriptions will be considered because the present study involves a sample of women, and women are much more likely to be the providers of social support than the recipients (Belle, 1982), an imbalance that w i l l render women's experience of social support networks very different from men's. Gender training, deeply embedded in cultural mandates, is a fundamental aspect of growing up that profoundly influences the development of social life (Champion, 1995) and can not be ignored in any discussion of social support and its effects in a population of women. Gilbert (1995) states:  Wives of Alcoholics 11  The social construction of gender does much to encourage or dissuade males and females to confide in others and develop close emotional, supportive, and affectionate relationships, or avoid them. This cultural scripting of gender affects the styles of friendships (both between and across genders), the organization of networks, the degree of intimate confiding and sharing that takes place in supportive relationships, and the areas that are shameful and run counter to gender definitions (p. 124). Gender norms will also be discussed specifically in relation to the alcoholic family system, which tends to be rigidly gendered (Bepko, 1986, 1988, 1989; Nol, 1991). In addition, the contexts of employment, marriage, and friendships will be evaluated, as well as factors influencing the presence or absence of social support in these contexts for this population. Finally, past and present theories about wives of alcoholics will be reviewed, including the Disturbed Personality Hypothesis, the Sociological Stress Theory, Family Systems Theory, and Codependency Theory. Most theories have pathologized these women in the absence of empirical data. Wives of alcoholics are under significant stress as they attempt to cope with their husbands' alcoholism, while at the same time they experience loss of social support, and thus they are at risk for depression.  Depression in Women This discussion is limited to "neurotic" rather than "endogenous" depression (Fowles & Gersh, 1979). We are talking about the kind of depression that is experienced to some degree by everyone in response to the circumstances life presents (Pearlin & Johnson, 1977), also known as the "common cold of pyschiatric practice" (Silverman, 1968). Severe psychiatric disturbance was screened out of the sample used for this study because it is less likely to be affected by life  Wives of Alcoholics  12  events and interactive experience. Brugha, et al. (1990) found outcome of endogenous depression to be unrelated to most social network variables, while outcome of neurotic depression was significantly related to social support. Neurotic depression is the type more likely to be reactive in nature, influenced by stressful circumstances and factors embedded in the social environment, including social support. Depression and Cognitive Process According to B e c k ' s (1974) cognitive model, depression involves themes of defeat (Beck & Weishaar, 1989), perceiving the self as worthless, the world as meaningless, and the future as hopeless. This conclusion is said to derive from a cognitive process in which assumptions about self and others evolve within a context of early social interaction, and new experiences are perceived and interpreted on the basis of these assumptions (Fennell, 1989). When negative life events occur, negative thoughts based on negative assumptions about self are triggered, and causes are attributed to internal factors (Coyne, 1989), thereby causing a reduction in self-esteem and an increased risk of depression. Depressed persons often experience hostility and criticism from significant others (Hooley, Orley & Teasdale, 1986; Vaughn & Leff, 1985), which can insulate them from positive experiences that might challenge their negative view of themselves (Swann & Predrriore, 1985). Berne (1966) proposed that depressed affect is linked to inadequate recognition or "stroking," which is a supply born of the social environment that is in short supply for wives of alcoholics.  Depression as a Gendered Phenomenon In western society, depression is overrepresented in the female population by a ratio of two to one or higher across all age groups from 18 to over 65 (Nolen-Hoeksema, 1995; Paykel,  Wives of Alcoholics  13  1991; Shaw, Kennedy & Joffe, 1995; Weissman & Klerman, 1977). This difference is not genetically caused or biologically driven (Merikangas, Weissman & Pauls, 1985; NolenHoeksema, 1995), and it is not explained by differences in income, education, or occupation (Turner & Avison, 1989). It is not caused by a deficit in women's ability to cope (Kessler, McLeod & Wethington, 1985) or a difference in number of stressful life events (Dohrenwend, 1973; Turner & Avison, 1989). Instead, it is the differential emotional impact of stressors rather than different exposure to stress (Kessler, 1979; Kessler & McLeod, 1984). The difference between men and women in emotional impact of stressors appears to derive from gender differences in social experience (Gove & Tudor, 1973; Turner & Avison, 1989). For example, marital and parenting roles involve a different set of stresses, rewards, and resources for men and women (Gove, 1972; Gove & Tudor, 1973). Turner and Avison (1989) found men to be substantially more vulnerable to events occurring to self, and women to be substantially more vulnerable to events occurring to others, especially events occurring to spouse. This differential responsiveness to life events accounted for more than 22% of the difference in rate of depression between men and women. According to Jack (1991), "Women most often become depressed over disruption or conflict in close relationships, whereas men respond with depression to the loss of an ideal or an achievement-related goal" (p. 7). She further states: "Women perceive danger in interpersonal situations that pose a threat to connection, while men are threatened by situations that impinge on autonomy" (p. 13). Circumstances that contribute to depression are different for men and women, and they are related to gendered cultural prescriptions learned in the socialization process.  Wives of Alcoholics 14  Issues of unequal power. Power distribution in marriage and sex-role attitudes have been found to be related to depression in women. Whisman and Jacobson (1989) found depressed wives to be less likely to share in decision-making, which is a long accepted indicator of power distribution, and more likely to carry the burden of household tasks and childrearing than nondepressed wives. In addition, they found that wives who were depressed scored significantly lower in masculine personality traits than wives who were not depressed. Kaslow and Carter (1991) point out that many ethnic and religious cultures value dominant, powerful men and submissive, respectful women, creating a power imbalance that also contributes to a lack of intimacy—lack of power and intimacy being recurring themes in their work with depressed women. They further argue that "the suppression of anger, the desire to please others, the need to accommodate to others' expectations, and the propensity to accept blame are characteristics encouraged in women" and "associated with depression" (p. 168). Radloff (1975) found no difference between men and women in experiencing feelings of anger, but found women to internalize them while men were more likely to express them outwardly. Likely due to their gender training, women have been found to attribute failures to lack of ability, and accomplishments to luck or hard work, whereas the reverse is true for men, who tend to attribute failures to external factors and successes to internal factors (Abramson and Andrews, 1982; Frieze, 1975). Attributional processes are part of the meaning-making process that contributes to depressed affect (Beck, 1967).  Socialization shapes the evolving self. Sex-role socialization has trained women to be self-sacrificing and to feel guilty when they consider their own needs (Rubin, 1979). "Females learn that they must be the nurturers who never ask for nurturing" (Bepko & Krestan, 1985, p. 62), and they do not feel entitled to ask for what they need (Jack, 1991; Lederer & Brown, 1991).  Wives of Alcoholics 15  Their sense of self is embedded in relationships by their moral sense of caring for others (Belle, 1982). They are socialized to be more sensitive to the needs of others and to feel more responsible to meet those needs (Kessler & McLeod, 1984). According to Wethington, McLeod, and Kessler (1987), "Women's roles obligate them to respond to the needs of others" (p. 145). Those who feel responsible for others may also feel responsible for the effects of stressors on others (Brugha, 1995). Women are more emotionally affected than men by the negative events of those they care about (Wethington, et al., 1987). Gender differences in depression are partly due to the "cost" of caring about others (Belle, 1982; Kessler, et al., 1985; Turner & Avison, 1989). It is not just the event, "but the soil on which it falls" (Paykel, 1978, p. 250). An objective tally of life events is less related to depression than the subjective experience of stress, and even normal life-stage transitions place greater demands on women (Kaslow & Carter, 1991). Issues of dependence. Degree of emotional and economic dependence on others may be a factor in higher rates of depression in women. Dependency is defined as "an inordinate attachment to a person, cause, or organization, so that self-esteem is derived from the external object and not from an internalized source or from autonomous actions" (Jack, 1991, p. 16-17). Independence in women is not culturally valued, or women would not be fearful about becoming more independent (Rubin, 1979). Excessive dependence on others for one's sense of value increases risk of depression (Birtchnell, 1988; Overholser, 1990) and likelihood of self-blame for negative events (Overholser, 1990). Klein, Harding, Taylor, and Dickstein (1988) found depressives to score significantly higher in dependency and self-criticism than normal controls. High reliance on others for reinforcement makes a person more vulnerable to social loss, and stress can amplify this effect because it tends to increase affiliation needs in normal adults  Wives of Alcoholics 16  (Overholser, 1990). Women are taught to give family life high priority, and they may become emotionally dependent on family well-being for their sense of identity and personal adequacy. For women, "Self-esteem is tied to the quality of attachments; feelings of guilt, shame, and depression are associated with the failure of intimate ties" (Jack, 1991, p. 13). Self-esteem, which is highly negatively correlated with depression (Shaver & Brennan, 1991) , is fragile when one is dependent on external sources for validation and nurturance (Hirschfeld, Klerman, Chodoff, Korchin & Barrett, 1976). Depressed persons usually experience diminished feelings of self-worth (Becker, 1979) and may feel undeserving of help and support (Gilbert, 1995).  Depression negatively biases cognitive appraisal of self and others, including  the risks and benefits of social interaction (Pietromonaco & Rook, 1987); thus the depressed person is less likely to perceive and utilize available support (Brewin, 1995; Lloyd, 1995) and less likely to risk self-disclosure because of embarrassment and fear of negative evaluation (Gilbert, 1995; Lloyd, 1995). Interaction with a depressed person may not be pleasant and may actually alienate potential sources of support (Coyne, 1976; Brewin, 1995; Lane & Hobfoll, 1992) .  . Family roles and intimate ties. Family life can be very demanding for women. Mothers  who carry the majority of responsibility for child care without nurturance for themselves are at highest risk for depression (Belle, 1982), although being able to confide in their husbands substantially reduces this risk (Belle, 1982; Brown & Harris, 1978; Costello, 1982; Roy, 1978). The role of mother is low in control and prestige and high in demands and responsibility (Gove & Tudor, 1973), yet wives and mothers are held accountable for the well-being of their husbands  Wives of Alcoholics 17  and children and assumed to be at fault if their husbands are unhappy or their children are in distress (Barnett & Baruch, 1987). Brown and Harris (1978) found four factors that contributed to onset of depression in a sample of women, 83% of whom were married: 1) number of recent severely threatening life events; 2) lack of an intimate, confiding relationship with a husband or partner; 3) loss of her mother before age 11; 4) presence of three or more children at home under the age of 14. Roy (1978) confirmed the last three of these predisposing factors and found lack of full time or part time employment to also be a significant contributor to the onset of depression, a factor Brown and Harris (1978) found to be salient only in the presence of the other four factors. In a study of 449 women, Costello (1982) found social class, employment status, number of children at home, and loss of mother before age 11 not to be related to the onset of depression, but he did find that lack of closeness with an intimate partner and severe life events were related to risk of depression. Severely threatening life events and lack of an intimate, confiding relationship with their alcoholic husbands are highly salient predisposing factors for depression in wives of alcoholics. This could be compounded for those who are unemployed or have three or more children under the age of 14. Loss of control. According to Lloyd (1995), the salience of stressors is determined by their intensity, undesirability, and unexpectedness, along with the amount of adjustment required and perceived uncontrollability. Loss of a sense of control is often evident in depressed persons and is more often experienced by women than men because of their role demands (Radloff, 1975). Individuals who experience a sense of control over events in their lives are less distressed than those who feel like helpless victims (Johnson & Sarason, 1978). Feeling in control is  Wives of Alcoholics 18  achieved intrapersonally by believing one can bring about desired outcomes and creating a history of successfully influencing the course of events in one's life, thereby developing an internal locus of control (Sarason, et al., 1990). Powerlessness and an external locus of control are associated with depression (Benassi, Sweeney & Dufour, 1988). A sense of self-efficacy is fostered by exploring the environment, being willing to take reasonable risks, and attempting to solve problems when they arise, which requires a view that mistakes are not disastrous (Sarason, et al., 1990). Kobasa (1979) coined the term "hardiness" to describe a set of personality traits that protect against psychiatric symptoms in the face of stress: 1) a belief that one can control or influence life activities; 2) deep involvement and commitment to influencing life activities; 3) a view that change is an exciting challenge that fosters personal growth. Feeling in control is fostered interpersonally by the responsiveness of the social environment to one's needs and behaviors. Social investment of time, attention, and resources from valued others boosts self-esteem and provides reassurance that one is approved of and valued (Gilbert, 1995). Investment signals from others help the individual to feel attractive to others, free to make requests, and able to confide without risking attack, rejection, or involuntary submission (Gilbert, 1995). "The sense of acceptance is not just a matter of confidence that others w i l l be responsive in times of need. It also concerns the person's belief that he or she is an interesting, worthy person, an appropriate stimulus for the attention of others" (Sarason, et al., 1990, p. 141). Feeling accepted, valued, and supported contributes to the individual's sense of having control in significant areas of life and may help to foster a more positive view of self, others, and situations.  Wives of Alcoholics 19  Gender training actually predisposes women to experience loss of control and feelings of helplessness (Radloff, 1975), a central element in Seligman's (1974) model of depression yet also compatible with Beck's (1974) cognitive model. Helplessness is defined as "when the outcome occurs independently of all voluntary responses" (Seligman, 1975, p. 17), or in other words, when it is perceived that the self can not influence outcome, which ultimately leads to giving up. Concluding that outcomes are out of one's control impairs motivation to effect change as well as perception that choices may be effective, while leading to an increased state of emotionality that may include fear, anxiety, and depression (Garber, Miller & Seaman, 1979). Although wives of alcoholics often make valiant attempts to stop the drinking, they often experience themselves as powerless to bring about the desired outcome (Asher, 1992, Wiseman, 1991). According to Radloff and Monroe (1978), the cultural stereotype portrays women as being in need of help and protection and less able to take care of themselves out in the world, which carries with it an implicit message of inadequacy. As children, females experience fewer rewards and punishments (less attention) and less autonomy than their male counterparts, which means they have less opportunity to learn self-reliance and independence (Radloff & Monroe, 1978). As adults, their achievements are given less significance and value (Lavach & Lanier, 1975), being ignored, punished, or negatively labeled (Johnson, 1974). Not being allowed to compete freely may influence them to give up and become depressed because achievement and competent effort is often subtly undermined (Radloff & Monroe, 1978). Social forces historically have acted to keep women dependent, both emotionally and economically, on their husbands, with young children and limited support from community or extended family increasing their sense of isolation and powerlessness. "Early socialization toward dependency on  Wives of Alcoholics 20  people and 'needing love' increases the dependency" (Radloff & Monroe, 1978, p. 206) on a love relationship and reduces a woman's power therein (Epstein, 1974). Loss of self. According to Jack (1991), who reports the experiences of depressed women, depression is similar to grief because it involves a sense of loss and sadness, and she asks, "what is lost: the other or the self?" (p. 21). Because the female "self is interwoven with "other," women are vulnerable to a progressive disconnection from self and personal needs as they seek to connect and establish intimacy, which can be elusive as men seek to protect their autonomy. His more detached style may contribute to anxiety on her part, and to protect against the possibility of loss, she may offer compliance in the relationship, silencing her own voice in the process. Jack (1991) states that the cultural inequality of men and women complicates the attainment of true intimacy as it may cause the woman to alter herself in self-defeating and inauthentic ways in order to strive for secure attachment. "She protects her partner from her true reactions to support the illusion of intimacy" (p. 68), and she may even endure some hurt and injustice if she thinks it will help him to love her. Attachments are central to a woman's experience, and the very traits that facilitate her success at connection—being empathic, having flexible ego boundaries (Chodorow, 1974), and being willing to care for others—also render her vulnerable to selflessness and depression.  Summary of Depression Depression is a complex interaction of internal and external dynamics that is often precipitated in the face of adverse life events. It is at least twice as common in women as men, which is most likely due to women's socialized sense of responsibility for the well-being of others. Women often define themselves externally, based on success in close family relationships  Wives of Alcoholics 21  (Braverman, 1986), and they may get out of touch with their own personal needs and feelings in the process of striving for secure attachment (Banister & Peavy, 1994; Jack; 1991). When stress is high, the availability of a close, confiding relationship with an intimate partner can help protect against the onset of depression (Brown & Harris, 1978). Without access to a confidant and with their own well-being externally based, women may experience a sense of helplessness because they can not control the domain for which they feel responsible. This description may well apply to many wives of alcoholics.  Social Support Plutarch once said that the attention of friends "easeth grief and pain" (Brugha, 1995, p. 2). A social network has been defined as "those with whom the person has regular face-to-face interaction and some degree of commitment" (Barnett & Parker, 1995, p. 213). The most commonly used definition of social support was set forth by Cobb (1976): "Information leading the subject to believe that he is cared for and loved, esteemed and valued, and that he belongs to a network of communication and mutual obligation" (p. 300). Another description references the degree to which one's social needs are satisfied through interaction with others (Lloyd, 1995). Brown and Harris (1978) refer to the presence of an intimate other in whom to confide as a critical element for women.  Aspects of Social Support Social support may involve everyday social participation or practical help in times of crisis (Veiel, 1985) and includes both instrumental (tangible or material) and expressive or emotional (esteem enhancing) dimensions (Brugha, 1995). A person may be good at giving instrumental support, such as fixing things, but poor at giving emotional support (Gilbert, 1995).  Wives of Alcoholics 22  According to Champion (1995), emotional support is particularly important in protecting against psychiatric disorder. Other types of support include the social companionship of shared activities, informational support which provides knowledge and specialized guidance, and motivational support or encouragement to pursue a goal or sustain effort in the face of adversity (Champion 1995). These kinds of support can give new input to cognitive appraisal, thereby altering moods and attitudes (Lloyd, 1995). Social Support as a Modifier of Stress and Depression Social support has been explored at length in an effort to explain why some persons develop mental or physical illness in response to stress and others do not (Brugha, 1995). In a survey of research on the role of life events in the onset of depression, Champion (1990) found that the rate of severe events was greater for those having vulnerability factors such as a lack of social support, particularly the lack of a confiding relationship with an intimate partner for women. She also suggests the possibility that depression and lack of intimacy could be caused by the same event, a situation which may apply to wives of alcoholics. Variables moderating the stress-depression relationship have been generally classed into two categories, which are: characteristics of individual disposition, such as personality, interpersonal skills, and coping style, and characteristics of the environment, such as social support (Pretorius, 1994). Personality variables which may mediate the stress-distress relationship are hardiness or resilience, self-esteem, locus of control, and favorable disposition toward stimulation and change (Lloyd, 1995). A person with high self-esteem is less likely to personalize negative events, a person with a sense of control is less likely to give up, and a person with a change-as-stimulating mindset is more able to adapt. Social support and  Wives of Alcoholics 23  personality mediate the effect of negative life events to the extent that they: 1) affect the individual's ability to maintain meaning and purpose in life; 2) affect the individual's belief he or she can influence outcomes; 3) bolster self-esteem; 4) affect the individual's ability to adapt to change without feeling overwhelmed (Lloyd, 1995). Personal and social factors exert their effects on psychological state via a direct effect on health through satisfying roles and rewarding experiences in the social network, independent of the presence of adversity, or via an indirect effect on health by buffering stress through increasing host resistance, self-esteem, and coping, having no effect in the absence of stress (Barnett & Parker, 1995; Brugha, 1995; Henderson, 1988; L l o y d , 1995). These are the main or direct effects model and the buffering model, and there is evidence to support both (Cohen & W i l l s , 1985; Sandler & Barerra, 1984; Turner, 1983), although the concept of the buffering effect appears to be somewhat more widely supported (Cohen, M c G o w a n , Fooskas & Rose, 1984; W i l c o x , 1981). For example, Henderson (1981) found a lack of perceived social support to be related to the development of neurosis under conditions of high but not low adversity. The converse of the buffering hypothesis is that the lack, or perceived lack, of social support may contribute to stressful events and exacerbate symptoms (Champion, 1990; Rook, 1984).  Thoits (1986) conceptualized social support as coping assistance in response to environmental demands on the individual. Coping is defined as "constantly changing cognitive and behavioral efforts to manage specific external and/or internal demands that are appraised as taxing or exceeding the resources of the person" (Lloyd, 1995, p. 51). Coping may be problemfocused (instrumental) or emotion-focused (expressive) in an effort to deal with the problem or the emotions it evokes. It has been theorized that encouragement and coping assistance bolster  Wives of Alcoholics 24  self-esteem, thereby countering the deleterious effects of stress (buffering hypothesis). The ensuing success provides a sense of mastery and control, which counters a sense of hopelessness (Lloyd, 1995). Thoits (1983) has suggested that ongoing strains negatively impact self-esteem by decreasing a sense of control or by causing the loss of a valued role. This may be particularly salient for wives of alcoholics, who unsuccessfully attempt to control their husband's drinking and experience loss of the role of intimate partner (Asher, 1992, Wiseman, 1991). Interaction of the Personal and the Social Psychosocial theory is a particularly useful tool in the discussion of social support because both individual and environmental factors are relevant. Both psychosocial and systems theory suggest that the relationship between the interpersonal and the intrapersonal is reciprocal (Brugha 1995). According to Henderson (1984): The individual is largely the author of his social environment and not just its passive recipient. Individuals who lack social support lack social relationships and this must often be due to personality defects. These in turn may also confer vulnerability to psychiatric disorder . . . . social support and personality are inextricably linked (p. 51). The external social environment includes experiences with others within the context of life events which may be either normative, or typical of a particular life stage, or non-normative, in that they are developmentally unexpected when they happen (Champion, 1995). A person's inner resources include physical characteristics and biological predisposition as well as self-esteem, ego strength, intelligence, the ability to plan and adapt, and internal working models fashioned via his or her "cognitive capacity to acquire and organize a knowledge of self and others" (Champion, 1995, p. 62). Each individual constructs a self-schema of  Wives of Alcoholics 25  personal traits and capabilities that reflects his or her experience with close others and is used as a reference for interpreting new situations. According to B o w l b y ' s (1969, 1973, 1980, 1988) notion of secure attachment, internal working models of self are projected onto others as a means of understanding their behavior. Information that is inconsistent with the self-schema may be ignored or minimized (Brewin, 1995). Early attachment provides the prototype for the formation of social ties in later life and determines how social support will be utilized (Champion, 1995). From their own experience, children create internal working models of self and their main attachment figures. Trust in others to provide support evolves within the context of that early social interaction, as responses from early attachment figures are internalized (Bowlby, 1969, 1973, 1980; Brewin, 1995; Gilbert, 1995). Securely attached children are more confident in seeking help and comfort (Champion, 1995). According to L l o y d (1995), i f the world is perceived as supportive, exploration and reasonable risk-taking are more likely because anxiety and preoccupation with self are reduced. Self-preoccupation is a central aspect of anxiety because it focuses on worrying about potential failure and the dangers of exploring new options, thus inhibiting task performance. "Highly anxious people do not explore, persist, or embark on new ventures because their attention is directed at self rather than the task" (Lloyd, 1995, p. 142). Secure attachment in childhood enhances self-efficacy and one's ability to initiate relationships without anxiety or fear of rejection (Sarason, et al., 1990). Poor quality relationships in early life predispose to poor quality relationships later, modified to a greater or lesser degree by subsequent social interaction. According to Sarason, Levine, Basham and Sarason (1983), the most relevant aspects of social support are the perception that others are available when needed and the degree of  Wives of Alcoholics 26  satisfaction with the adequacy of the support that is perceived. A situation is stressful to the extent that it is perceived as a threat, thus the meaning the individual associates with the stressor is the critical element that determines the effect of the stressor (Turpin & Lader, 1986). A moderator of the stress-depression relationship would likely be subject to internal perception in much the same way as the stressor itself. "The evidence is that, as a causal factor in neurosis, the crucial property of social relationships is not their availability, but how adequate they are perceived to be when the individual is under adversity" (Henderson, 1981, p. 396). Perceived Social Support Social support has both objective and subjective dimensions. The presence of social networks does not guarantee that the individual will perceive them as useful (Lloyd, 1995). Barerra (1986) defines three categories of social support, including characteristics of the social network and the individual's embeddedness therein, actual enacted social support, and perceived social support, which encompasses cognitive appraisal of the availability and adequacy of support, as well as feelings of satisfaction with it. Perceived social support also involves a sense of "as is" acceptance, inferred about self from social interaction, which contributes to a positive self-image and a view of others as benign, as well as an expectation of positive outcomes (Sarason, et al., 1990). Perceived social support "is part of a constellation of cognitions that drives social behavior and accounts for differences in interpretations regarding the behavior and motivation of others" (Sarason, etal., 1991). H i g h perceived social support indicates a belief that specific others w i l l be accepting and available i f needed, at least most of the time. " K n o w i n g that one is loved by others who would willingly do what they can to help has a powerful effect on personal working models" (Sarason,  Wives of Alcoholics 27  et al., 1990, p. 141). B o w l b y (1969, 1973, 1980) theorized that internal working models become a stable aspect of personality that governs expectations about self and others and guides social interaction. Self-reports of perceived social support have been found to be stable over periods as long as three years, even when there were changes in social networks (Sarason, Sarason & Shearin, 1986). This may be explained in that prior experience in families, which affects perceptions of interpersonal risks and benefits, is constant (Brewin, 1995). Lakey and Heller (1988) found profound differences between perceived social support and objective acts of support and suggest that a sense of being cared for and valued is not • primarily anchored in the experience of receiving help or in specific relationships, but is integrated into personality. Lakey and Cassady (1990) found perceived social support to be highly correlated with measures of personality, inversely with trait anxiety and dysfunctional attitudes, and directly with self-esteem, with those scoring low in perceived social support viewing supportive behaviors more negatively due to perceptions of being unloved and unvalued. They concluded that feeling supported is a psychological phenomenon based on an internal representation of self in relation to others. Other individual variables, such as heredity, biological processes, and temperament, may also influence perceptions of social support (Sarason, et al., 1990). Persons who perceive themselves as lacking in support tend also to be less socially competent, more introverted, more anxious, intolerant of psychological problems in others, and more pessimistic about social relationships, as well as being more likely to have experienced unhappy childhood relationships with parents and to have a passive stance toward life events and an external locus of control (Parry, 1995).  Wives of Alcoholics 28  Perceived social support also appears to be strongly influenced by environmental factors (Brugha, 1995). T w o effects of supportive environments are low anxiety and a sense of acceptance (Sarason, et al., 1990), with social support and anxiety being negatively correlated (Sarason, et al., 1987; Sarason, et al., 1983). Sarason, et al. (1987) found through factor analysis that a perception of being loved and valued is central to perceived social support. Sarason, et al. (1986, 1991) found perceptions of social support to accurately reflect actual views of family and friends, which were independently obtained. They also found individuals' perceptions of specific relationships to be distinct. These researchers (1991) conclude: This construct represents an interaction between the needs and abilities embodied in an individual's personality and the distinct relationships developed in the course of social interactions. In a sense, perceived social support is a cognitive adaptation that individuals make given the constraints and opportunities, both real and imagined, that are placed upon them by a history of experiences that result in working models of self and others (p. 235). Perceived social support is best conceptualized as an individual difference variable as well as a provision of the social environment. It is more related to outcome than actual support received (Cohen & W i l l s , 1985; George, et al., 1989; Kessler & M c L e o d , 1984; Wethington & Kessler, 1986), which may affect outcome by its impact on perceived social support (Sarason, et al., 1990). A s a variable, perceived social support may best predict whether stress w i l l result in depression because it taps both the intrapersonal and the interpersonal domains. Social Support as a Gendered Phenomenon The care of mother to child is the prototype of social support (Cobb, 1976), and Belle (1982) has equated the expressive function identified by Parsons and Bales (1955) with the  Wives of Alcoholics 29  giving of social support. It has been assumed that social support occurs in networks of mutuality with balanced give and take (Caplan, 1976), but this view has ignored the fact that "women are overrepresented among providers of support" (Wethington, et al., 1987, p. 152). Bograd (1990) states: "In our society, mothers are defined as all-knowing, all-giving, all-responsible, and allsacrificing" (p. 73). G i v i n g all, sacrificing self, and taking the blame when something goes wrong in the family is the unspoken job description for mothers. Women in general usually give more support than they receive, which is referred to as the support gap (Belle, 1982, 1982b), a disparity that is unlikely to be without consequence. Lower levels of social support would theoretically translate into less of a stress-buffering effect, and increased rates of depression would be a logical result. Marriage has a protective function against suicide and mental health problems requiring treatment for men but not for women (Belle, 1980), which may be because more of the support in marriage is given by wives to their husbands. During times of stress, both men and women are more likely to seek out women for support because the male sex role discourages disclosure of feelings (Buhrke & Fuqua, 1987) and men lack training in the giving of emotional support (Belle, 1982). Gender identity may cause men to show their power and toughness while downplaying their capacity to empathize, nurture, and be affectionate (Gilbert, 1994). Caregiving activities have been seen as feminine and unmanly, which is an issue of rank in a society where overt power is highly valued and women generally have less of it than men do (Gilbert, 1995).  Women enact the majority of caregiving in our society in that they take primary responsibility for dependent members such as children and the elderly or handicapped (Guberman, 1990). Caregiving is "both the identity and the activity of women in western society" (Graham, 1983), with women carrying the major responsibility for the emotional and  Wives of Alcoholics 30  physical well-being of the family, both nuclear and extended. Social exchange theorists remind us that social ties are of a dual nature in that social interaction brings with it both rewards and costs (Homans, 1974). Rook (1984) found 38% of friendships and 36% of family ties to involve negative interactions, thus an extensive social network may be more associated with an overload of responsibility than a relief (Belle, 1982, 1982b), particularly for women. Veiel (1993; 1995) found a larger kin support network to be related to increased symptoms of depression and major depressive episodes in women who do not work outside the home ( r = .67 compared to .23 for employed men and .07 for employed women). In fact, for each additional person in the kin support network, the chance of a major depressive episode in nonemployed women doubled (Veiel, 1995). Guberman (1990) found that women assumed the major burden of caring for dependent adults, with sons involved only i f there was no female alternative. Women were found to provide the bulk of physical care, housework, and emotional support, and they experienced social isolation and lack of respite due to the monopolizing nature of the task, as well as a lack of financial autonomy resulting from spending their time doing unpaid work. They reported that they assumed the major burden of caregiving due to internalized sex-role expectations, a sense of obligation to family, pressures from other family members, a sexual division of labor where the male was the major breadwinner, and lack of viable alternatives. These factors are likely salient in women assuming the greater responsibility for parenting dependent children as well. Wethington, et al. (1987) found that wives were more frequently involved and spent more time helping during network crises than husbands. This was true whether or not the wife was employed.  Wives of Alcoholics 31  There are two general perspectives that seek to explain why women assume the major part in caregiving (Guberman, 1990). One looks to internal psychological factors stating that the qualities or characteristics required to give care are central to female identity, an expression of women's natural feelings of compassion. The other looks externally at the social organization of caregiving within the context of a sexual division of labor, with caregiving being an expression of women's position in society. The disparity between men and women as providers of support is most likely created by both internal and external forces at work in a reciprocal way, as social experience affects intrapsychic meaning-making, and different choices are made that affect social interaction. Women are described by theorists of female psychology to have "flexible ego boundaries" and less well-defined "self-other distinction" (Chodorow, 1974) than men. This likely derives from women's early sense of connection with mother, which endures as a developmental stage much longer than for men, who create some separateness from mother to find their identity as masculine beings via the physically and emotionally illusive role model of father (Pittman, 1990). With this in mind, it would not be surprising for women to naturally develop an orientation toward affiliation and men to become more inclined toward autonomous achievement. Hammen, Ellicott, Gitlin & Jamison (1989) found that the type of stressful life events to which unipolar depressives were vulnerable was specific to their desires for affiliative or autonomous achievement. The emotional development of men and women is very different, and in adulthood, women are more likely to derive satisfaction from affiliation, especially in intimate, confiding relationships, whereas men are more likely to find satisfaction in autonomous task accomplishment (Cohen & Wills, 1985).  Wives of Alcoholics 32  Due to developmental differences, the kind of social support men and women need and give is likely to be different. Billings and M o o s (1982) found that family support was more important to women than work support, whereas the reverse was true for men. East (1989) found father support to be inversely related to perceived interpersonal risks in adolescent boys and directly related to perceived interpersonal benefits in adolescent girls, although parental care has been found to be positively related to perceived social support in both sexes (Sarason, et al. 1986, 1991; Flaherty & Richman, 1986). It has been found that depressed men and women derive different benefits from support networks. Fondacaro and M o o s (1987) found that perceived family support was positively related to problem-focused coping in women and negatively related to emotional discharge in men. Brugha, et al. (1990) found that recovery from depression in women was positively related to number of primary support persons and satisfaction with social support, while in men recovery from depression was positively related to living as married and negatively related to amount of negative social interaction. They concluded that men and women are equally dependent on social support but utilize different aspects of it due to differences in their socialization. When it comes to providing support, men may be more likely to make instrumental, task-oriented contributions, whereas women may provide greater amounts of emotional support in addition to physical help (Gilbert, 1995).  Summary of Social Support Perceived social support has been found to better moderate the stress-depression relationship than objective measures of actual enacted support or network characteristics. It is a construct that is highly correlated with personality variables, yet has been shown to accurately reflect actual views of family and friends. It evolves out of a complex lifelong interaction of the  Wives of Alcoholics 33  individual with his or her social environment and results in internal working models of self and others by which the meaning of events and the motives of others are subjectively interpreted. Due to differences in sex-role socialization, women often provide more social support than they receive, which may be related to their higher incidence of depression (Belle, 1982). The sense of connection women experience in early development translates into a moral sense of responsibility for others (Gilligan, 1993) and enables women to better recognize the needs of others and believe they can be met (Miller, 1976). When carried to extreme, women's sensitivity to the needs and experiences of others can bring about a loss of self (Chodorow, 1974; Jack, 1991), particularly in the absence of a close, confiding relationship with an intimate partner, which may contribute to depression (Brown & Harris, 1978; Champion, 1990; Costello, 1982; Roy, 1979).  Gender Norms and the Alcoholic Family System The social context of wives of alcoholics can not be understood without first recognizing that life in families and in society is experienced differently for women than for men. Men are deemed "normal" and women are viewed as "other," with women's characteristics being more negatively valued (Downing, 1991). Women are seen as less mentally healthy than men (Jack, 1991), and standards for mental health for men and women have been based on sex-role stereotypes (Broverman, Broverman, Clarkson, Rosenkrantz & Vogel 1970). Broverman, et al. (1970) state: Thus, for a woman to be healthy, from an adjustment viewpoint, she must adjust to and accept the behavioral norms for her sex, even though these behaviors are generally less socially desirable and considered to be less healthy for the generalized competent, mature  Wives of Alcoholics 34  a d u l t . . . . Acceptance of an adjustment notion of health, then, places women in the conflictual position of having to decide whether to exhibit those positive characteristics considered desirable for men and adults, and thus have their femininity questioned, that is, be deviant in terms of being a woman; or to behave in the prescribed feminine manner [and] accept second-class adult status (p. 6). Therapists have traditionally been male, while patients have been largely female. Even today, problems are seen, causes are attributed, and solutions are implemented from a predominantly male perspective. In our male-dominated culture, women's needs and problems are considered to be less important but more pathological than men's (Anderson & Holder, 1989; Krestan & Bepko, 1991). "Women operate within a context that defines them as less adequate, less powerful, and less worthwhile than men, and men operate within a context that compels them never to identify with women lest they lose their power" (Bepko, 1986, p. 68). From the male perspective, to be controlled or to be weak is feminine and therefore frightening (Nol, 1991). Denial in the male alcoholic is extremely resilient, a concept that can be explained by gender training (Nol, 1991). Although to be able to "hold his liquor" is considered evidence of masculinity (Ettorre, 1986; Lammers, 1991; Lemle & Mishkind, 1989), the acknowledgement of alcohol dependency is acceptance of weakness, which he cannot reconcile with his identity as a male (Nol, 1991). Often he displays a "macho" facade, yet he drinks to quell his sense of loss of control and fears of inadequacy as a man (Bepko & Krestan, 1985; Rogalski, 1986; GriffinShelley, 1986). Bepko (1989) states: Being socialized in an inherently erroneous belief system sets the stage for addiction, since incongruity may arise between what one believes and how one actually feels. If a male must  Wives of Alcoholics 35  show dominance but actually feels vulnerable, he may employ some addictive agent to either disqualify his vulnerability or enhance his sense of dominance (p. 408). In addition, the alcoholic may accuse his wife of being deficient in her femininity, "a projection and displacement of his own gender concerns" (Burnett, 1984, p. 53). N o l (1991) states that many addicted men grew up in rigidly gendered families with fathers who were absent or emotionally unavailable, leaving a murky model of masculinity. Chodorow (1974) contends that due to inaccessability of the father, boys were left to use cultural stereotypes as a model for masculine behavior. According to Pittman (1990), cultural stereotypes depict and encourage hypermasculinity and the development of the macho man persona. In the process, many men distance from their softer emotions, relying on women to mediate their relationships and then blaming them i f something goes wrong. This is apparent in movies such as On Golden Pond, as women "automatically take the emotional responsibility for explaining and interpreting the men to the children, thus maintaining the men's immobility to deal with emotional relationships" (Krestan & Bepko, 1991, p. 57). Females living with alcoholism in the family frequently suffer very serious consequences resulting from men striving for the hypermasculine ideal, including rape, incest, physical abuse, depression, eating disorders, and sexual dysfunction (Bepko, 1988). The macho man searches for a powerful pseudoself to cover his dependency needs (Nol, 1991; Bepko, 1988; Bepko & Krestan, 1985), using alcohol excuse for his behavior. A common theme in alcoholic behavior is a style of approach and avoidance wherein the male alcoholic may "display hostility, violence or threats of violence, and/or distancing and emotional unavailability" alternated with "feelings of dependence, remorse, and pleas for closeness coupled with good-natured  Wives of Alcoholics 36  sentimentality" (Burnett, 1984, p. 53). M c C o r d & M c C o r d (1960) suggested, based on interviews with 225 men, that drinking satisfies men's need to feel taken care of while maintaining the independent masculine facade that drinking defines (Bepko & Krestan, 1985; Downing, 1991). According to Griffin-Shelley (1986), "Exaggerated sex-role behavior and attitudes—'macho' men and 'seductive' women—are common among alcoholics and drug addicts" (p. 1307). The male alcoholic projects blame for his behavior onto his wife, who is an easy scapegoat due to her own socialization. She w i l l either accept the blame and become chronically depressed, or she w i l l blame others, including the alcoholic, and be filled with chronic anger and resentment (Burnett, 1984). A n American wife, who participated in a study by Wiseman (1991), said this about the alcoholic's adeptness at passing on responsibility: " A n alcoholic is so good at playing with your mind that I was convinced for so long that even though I knew he had a problem, I was convinced that it was my fault, and I really thought I was crazy" (p. 29). She will feel guilty, alienated, and shamed, yet responsible to stay with him "for better or worse," rather than to desert him in time of need. She may also fear what he might do i f she leaves (Wiseman, 1991). If "she feels that her self-esteem and economic welfare are tied to him, she will be invested in the denial and rationalization process right along with him, even when this involves assuming blame for his illness" (Burnett, 1984, p. 53).  Working within the family system, the influence of gender differences as a determinant of adaptive responses must be recognized. Through the male lens of Systems Theory, the woman in an alcoholic system is said to be enmeshed, fused, or undifferentiated, yet to become differentiated, by definition she must assume the socialized traits of the masculine role, including  Wives of Alcoholics 37  independence, autonomy, and the aloofness of reason detached from emotion; in other words, for women, differentiation is nonnormative (Bepko, 1986). Gender identity formation is a crucial aspect of personality development, which takes place within the cultural context of females being considered "less than" males (Nol, 1991). This is further reinforced by the teaching in the alcoholic family system that all members are secondary to the alcoholic (O'Gorman, 1991). If the wife of an alcoholic also grew up in a family with an alcoholic father, which is frequently the case [Casey, Griffin, and Googins (1993) found an incidence of 50%], she has learned to discount her own perceptions and trust external cues from the family environment as to what is appropriate behavior in an ever changing scenario which has different rules during "wet" (drinking) and "dry" (not drinking) times in the alcoholic cycle—she learns dependency and an external locus of control in this way because the consequences are severe i f she does not (O'Gorman, 1991). She w i l l likely acquire dysfunctional roles and a distorted sense of identity. She w i l l derive her identity from gratifying the needs of others while neglecting the self, will have difficulty asking for help, and will carry a great deal of suppressed anger (Kokin & Walker, 1989). A woman from an alcoholic family of origin has an increased risk of depression (Downing, 1991; V a n Den Bergh, 1991), which is amplified if the alcoholic system is reproduced in her marriage. A s an adult, an external locus of control will continue to keep her looking outside herself toward others for approval to help her feel important and worthwhile, and can set her up for addictions of her own, such as eating disorders, relationship addiction, prescription drug abuse, and alcoholism (Van Den Bergh, 1991, Bepko, 1988). O f the one in four alcoholics who is female, 50% of them are married to alcoholics (Schaap, Schellekens & Schippers, 1991). If the wife of an alcoholic drinks, she is relieved from her sense of overresponsibility and  Wives of Alcoholics 38  perfectionism as she flips into the underresponsible role in the alcoholic family system (Bepko, 1988).  Contexts For Self-Definition According to symbolic interactionist theory, personality evolves and appropriate social conduct is learned through social interaction, which is an essential part of normal development (Thoits, 1983). The self is defined as a set of identities enacted in reciprocal role relationships. Enacting roles provides a sense of being valued and needed and provides purpose, meaning, and order to existence, thereby contributing to mental health (Thoits, 1983). More identities are seen as increasing meaning and reinforcing a sense of personal value in existence. This is known as the identity accumulation hypothesis (Thoits, 1983). Being a functioning person in more than one context provides additional opportunities for defining the self as valuable and capable while challenging messages of inadequacy, such as projections of blame from an alcoholic husband. Oatley and Bolton (1985) state: "Depression occurs when an event disrupts a role that had been primary in providing the basis for a person's sense of self and there are no alternatives that allow that sense of self to be maintained" (p. 372). Thus, wives of alcoholics who define themselves by their marital role become vulnerable to depression as the intimacy of the marriage is lost.  Self definitions are derived through an ongoing interactive process and can be extremely resistant to change because each new social experience is processed through a view of self that has been shaped by many past interactions (Parry, 1995). The sense of self is only eroded through a series of ongoing negative interactions. Brown, et al. (1986) found that a negative view of self was a vulnerability factor in the development of depression after the occurrence of a severely threatening life event. Self-blame and negative self-appraisal are associated with more  Wives of Alcoholics 39  social withdrawal (Brewin, McCarthy & Furnham, 1989) and are usually experienced along with a sense of failure, such as in a task or in a relationship (Dunkel-Schetter, Folkman & Lazarus, 1987)—not an unlikely scenario for the wife of an alcoholic (Asher, 1992; Banister & Peavy, 1994; Wiseman, 1991). Over time, repeated projections of blame from the alcoholic and the experience of being powerless to stop his drinking (influence outcome) can cause her to question her capability and personal worth and redefine herself more negatively (Banister & Peavy, 1994), thereby increasing her vulnerability to depression. Isolation is defined by Thoits (1983) as possessing few or no social role identities, or in other words, lacking involvement in role relationships. In testing her identity accumulation hypothesis, she found that more role identities (identity accumulation) was associated with less psychological distress. She also found that females showed significantly greater distress than males, and distress decreased with age, education level, and family income. Additional social identities were found to enhance well-being rather than result in role strain or role conflict. Employment and Depression W o r k outside the home can help to protect women under stress from depression (Brown & Harris, 1978; Warr & Parry, 1982). Benefits women have reported from being employed include confidence, self-esteem, a sense of accomplishment, dignity, and independence (Tebbets, 1982). Turner and A v i s o n (1989) studied the effect of employment in a chronically strained population and found that employed women were less depressed than unemployed women despite increased stress. They found that the work role acted as a buffer against the impact of events to significant others. In order to make predictions about the effect of employment, it is important to consider desire to be occupationally involved, quality of the nonoccupational  Wives of Alcoholics 40  environment, and quality of employment relationships (Warr & Parry, 1982). If the nonoccupational environment is adverse, a positive relationship between paid employment and well-being is more likely because the job may provide what is lacking elsewhere, such as opportunities for self-esteem and social support, while giving respite from domestic stress, such as a disturbed marital relationship (Warr & Parry, 1982). Paid employment w i l l be the most beneficial to a woman i f her desire to be employed is high, her nonoccupational environment adverse, and her employment relationships good. In a review of 38 studies, Warr & Parry (1982) found the relationship between paid employment and women's well-being to be stronger for working class than middle class subjects, indicating that employment had a greater effect on well-being under conditions where it reduced financial stress. They found that in no cases did employed women have significantly lower personal well-being than nonemployed women, although there were no significant differences between employed and nonemployed married women with children at home. Welch and Booth (1977) found a slight advantage for employed mothers over nonemployed mothers i f their children were above preschool age. Nye (1963) found part time employed mothers to be the most satisfied with their daily work. Mostow and Newberry (1975) found that women employed at least half time recovered from depression more quickly than homemakers. Veiel (1995) found homemakers to be more likely to face demands for social support, to have less opportunity to escape from the demands of family, and to be more likely to experience depression. H e states: . . . the detrimental effects of supportive social networks on both recovery and relapse involved women who were dependent on their families. Having a (male) partner and being a homemaker often implies that they are dependent, emotionally as well as economically, on  Wives of Alcoholics 41  their families. It is argued that this dependency on one's family alters the way social support is solicited, received, and given, both within the family and without (p. 159-160). According to Gove and Tudor (1973), the role of homemaker is low in status, economic reward, appreciation, and external structure, yet high in frustration. If life at home is unpleasant, she is left without an alternative context for self-esteem and personal fulfillment. Marriage as a Context For Social Support Marriage is "the most rewarding and yet the most demanding of all interactions within our culture . . . . a complex relationship involving emotional, interpersonal, familial, legal, vocational, physical, and spiritual dimensions that comprise the intricate fabric of human behavior" (Paolino & M c C r a d y , 1977, p. 5). Marriage can potentially bring great happiness or great pain. Self-esteem for women is closely related to the quality of intimate relationships, particularly marriage (Brown & Harris, 1978). Relationships with intimate partners are of great importance in providing the social support that mediates the stress-depression relationship (Brewin, 1995). Brown and Harris (1978) found that for the women they studied, the most important social relationship needed as a resource for social support was one with a sexual partner, involving closeness, confiding, trust, and free expression of feelings. Negative comments from such a confidant can lead to increased self-blame (Andrews & Brewin, 1990), and threatened or actual loss of this kind of an intimate relationship is extremely stressful (Lloyd, 1995). R o y (1978) found that women being referred for treatment as depressed was associated with having a poor marriage and the lack of close confidants. Women are more vulnerable to marital stress. Pearlin and Lieberman (1979) found that women are more vulnerable than men to the stress of a problematic marriage. Wethington, et al.  Wives of Alcoholics 42  (1987) state: "Women are more emotionally responsive than men to the quality of interpersonal ties" (p. 153), and they found the quality of the marital relationship to be more highly correlated with emotional well-being in women than men. They concluded that the absence of emotional support within marriage is potentially more devastating for women. According to Braverman (1986), "The need for attachments and relationships is so central to a woman's sense of self that she may experience the loss of a relationship as a significant diminution of s e l f (p. 94). Degree of marital impairment and severity of depression have been found to be directly related in women (Weissman & Paykel, 1974), and wives often attribute the cause of marital distress to self (Heim & Snyder, 1991). Brown, et al. (1986) found that lack of emotional support and a sense of being "let down" in the presence of adversity were associated with the onset of depression in women. Hooley and Teasdale (1989) found criticism from spouse to accurately predict relapse into depression. H e i m and Snyder (1991; Snyder and Heim, 1992) found marital disaffection to best predict depression, explaining one third of the variance in men and one half of the variance in women. It seems apparent that spousal support contributes strongly to well-being and acts as a buffer against stress (Barnett & Parker, 1995).  Ilfeld (1982) defined marital stressors as "circumstances or conditions of daily marital life that are generally considered problematic or undesirable, including non-fulfillment of role obligations, lack of reciprocity between the marital partners, and a feeling of nonacceptance by one's spouse" (p. 483), this being an ongoing circumstance rather than a discrete event. Companionable relating depends on reciprocality to create a sense of belonging, identity, and safety, and the loss of these elements can lead to feeling like an outsider (Gilbert, 1995). Oatley and Bolton (1985) theorized that "depression occurs when events disrupt roles by which people define their self-worth" (Lloyd, 1995, p. 50). Women, and especially homemakers, often define  Wives of Alcoholics 43  themselves by the success of their marital and family roles. Ilfeld (1982) found that self-esteem and self-efficacy showed a strong and significant inverse relationship to marital stressors, with frequency of marital arguments the strongest predictor of marital stress. Loss of support within the marriage. For the wife of an alcoholic, the alcohol-embedded marital context is filled with stress and is not a supportive resource. According to Paolino and McCrady (1977), "The demands on any marriage become especially severe when one or both partners have psychological weaknesses or are experiencing serious troubles. It is a tragic fact that alcohol misuse and a satisfying marriage are incompatible" (p. 6). Banister and Peavy (1994) report that wives of alcoholics struggle with the reality that marriage for them is far from their culturally schooled expectation, and they feel let down over this disparity and powerless in their marital relationship; As the male alcoholic's perception of self and others becomes distorted, and he is cut off from the outside world, he becomes more dependent on his wife and exerts more control over her in an effort to prevent her becoming independent from him (Denzin, 1987). Daily interaction with an alcoholic husband teaches these wives to suppress their emotions and get out of touch with their own needs, causing a gradual weakening of self (Jack, 1991). Themes of powerlessness, loss of trust, guilt, self-blame, and isolation are common (Banister & Peavy, 1994). Wilhelm and Parker (1988) refer to care (affiliation-hostility) and control (dominance-subordination) as two dimensions of intimate relationships. It could be said that wives of alcoholics lack both care and control in their primary relationship.  Intrapersonally, the wife of an alcoholic may become overresponsible for others partially to compensate for her lowered self-esteem, which deteriorates as she realizes that she is less important to the alcoholic than his bottle (Wiseman, 1991). In a complementary way she may  Wives of Alcoholics 44  become less able to meet her own personal needs and feels less entitled (Lederer & Brown, 1991) to ask for them to be met by others (Bepko, 1988). She is less able to express anger directly as she is afraid that it may give her husband an excuse to drink more and respond with anger or violence. Her voice is effectively silenced, she is unable to be heard, and she is at risk for depression (Jack, 1991). She may find it difficult to trust because she has been let down many times by an unreliable partner (Kokin & Walker, 1989; Wiseman, 1991). As the drinking behavior escalates, she is likely to become alienated from her husband and lose the companionship of the marriage relationship (Wiseman, 1991). Communication is the vehicle of empathy and understanding in a marriage, and at first she attempts to weave it around his drinking patterns, but as his alcoholism progresses, she may give up in despair (Wiseman, 1991). Physical attacks or threats, verbal insults (symbolic threats), neglect (ignoring, not listening), and removal of investment (withdrawal of love, attention, support, help) are putdown signals (Gilbert, 1995) that attack status and a sense of worth, and they are common in the experience of wives of alcoholics (Asher, 1992; Banister & Peavy, 1994, Wiseman, 1991). The ritual of meals is a basic source of structure in family life, but the excessive drinking kills his appetite, and he usually avoids mealtimes. She may view preparing a fine meal as "a creative act, an expression of love, intelligence, and imagination . . . . Food thrown out symbolizes all sorts of failures to her . . ." (Wiseman, 1991, p. 133-134). As the disheveled alcoholic stinks of sweat and stale beer and throws up on or wets the bed, sex is likely to become a turn off for her, and he increasingly struggles with impotence because of the alcohol (O'Farrell, 1990). Social gatherings often become an embarrassment to her as his behavior increasingly becomes a social disaster (Asher, 1992; Wiseman, 1991). Friends are thereby alienated and she is likely to become  Wives of Alcoholics 45  even more isolated and less confident as she is shamed by her husband's alcoholism, thinking that if she was somehow a better person, he wouldn't need to drink (Wiseman, 1991). Wiseman (1991) states from a symbolic interactionist perspective that people define the situations they are in and.construct their reactions based on the meaning they assign. A s their experiences change their definitions of reality, their responses also change. When we view the actions of the wife of an alcoholic through her eyes, her behavior makes sense. Once she defines his drinking as alcoholism, she may try to solve the problem and gain control by becoming perfectionistic—a better mother, wife, cook, and housekeeper—thinking that i f she can provide him a low stress, well-ordered environment, then he w i l l not need to drink. She tries many strategies to deal with his problem, but she ultimately fails. When he does not stop drinking, she is likely to feel, powerless and uncared for, losing confidence in her ability to influence outcome, and low self-esteem and depression may result. Alcohostages. About 90% of wives of alcoholics remain in their marriages (Kokin & Walker, 1989; Downing, 1991), even though as many as half of them have separated at one time or another (Wiseman, 1991). In contrast, only 10% of men married to alcoholic women remain in their marriages (Downing, 1991). Because the male way is the standard, wives of alcoholics have been pathologized as masochistic for their sense of loyalty and commitment (Kokin & Walker, 1989). It must also be remembered that they have fewer choices due to lower wages for women, lack of marketable skills, and responsibility for dependent children. K o k i n & Walker (1989) refer to wives of alcoholics as alcohostages, drawing a parallel between them and hostages in captivity. They are shamed by their lack of control, they are isolated by their feelings of shame, they often see no way out, and they may even bond to their captors in an effort to  Wives of Alcoholics 46  reduce anxiety. Unable to obtain freedom from the alcoholism, they try to survive and to adjust the best they can. Wives may be held hostage by economic circumstances, their own values, and even by love. Hostages are often affected long after their captivity is over, and lack of support after victimization can inflict a "second injury" according to Dr. Martin Symonds, a psychiatrist at the New York University School of Medicine (Kokin & Walker, 1989, p. 68). Even if her husband achieves sobriety, the road is not easy for the wife of an alcoholic. Practical problems, such as debt and property damage, often remain as reminders of the struggle with alcohol. While the drinker's memories have been cushioned by alcohol, hers have not—she can remember it all. She has a lot of stored feelings of hurt and anger that she felt necessary to suppress during the drinking, and it takes time before she is ready to trust again. When the crisis is over, she finally has the luxury of coming apart, as the dam may break and her feelings emerge. She has been deeply affected, possibly even scarred, by the experience, and it cannot be instantly erased. Is it safe to reinvest her feelings? The future is unpredictable and the past is unresolved. The process of healing will take time. According to Bepko (1986), the wife may experience "a deep sense of emptiness when she relinquishes her carefully constructed sense of s e l f . . . . old roles and behaviors have been lost or discarded, but no new ones have emerged to take their place" (p. 73). A period of adjustment is required to rebuild trust.  Marital distress and depression in women. The breakdown of affiliative and supportive bonds can be quite painful (Gilbert, 1995), and this effect is likely to be more pronounced for women, who often define their sense of self by their relationships with significant others (Gilligan, 1993; Jack, 1991). Ilfeld (1982) suggests that social stressors associated with everyday roles are as important in the stress-depression relationship as life events or changes, and he found  Wives of Alcoholics 47  in a study of men and women ( n = 2299) that marital stresses were more highly correlated with depression than parenting, job, financial, or neighborhood stresses. Parental stressors for mothers and job stressors for men also reached statistical significance. The present study utilizes a sample of wives of alcoholics who are all experiencing significant marital distress, which is not uncommon among marriages affected by alcohol (O'Farrell & Birchler, 1987). An alcohol-embedded marriage, often characterized by negative interaction (Jacob & Krahn, 1988), likely creates the kind of stress that is often associated with depression for wives of alcoholics. This may happen through negating the supportive potential of the marital context. The lack of a close marital relationship for these wives is likely to have a significant impact on their emotional well-being, as has been demonstrated in other populations of women (Brown & Harris, 1978; Costello, 1982; Roy, 1979). Friendships as a Context For Social Support Even though parents heavily influence the development of their children's internal working models even into adulthood, relationships with friends have the potential to modify perceptions of self and others (Sarason, et al., 1991). However, in a population of wives of alcoholics, there may be some reticence to fully tap this resource. The ability to mobilize support depends on the structure of the social network, willingness and ability of potential supporters to provide what is needed, and individual attitudes and behaviors that inhibit seeking or accepting help (Fisher, Goff, Nadler & Chinsky, 1988), such as low self-esteem and fear of dependency or judgment (Parry, 1995). Orner (1987) found that veterans from the war in the Falklands felt unable to talk to others because they felt marked by their experience, making them different from others in a way that was personally relevant to them. Stigmatized individuals feel different and  Wives of Alcoholics 48  tend to withdraw from social support systems and isolate themselves due to feelings of shame and embarrassment, fearing "social interactions in which their 'secret' might be unmasked" (Brewin, 1995, p. 104). This may be particularly relevant for wives of alcoholics who feel personally at fault for their husbands' drinking. Ironically, avoiding others causes them to have less access to information about the normalcy of their feelings and experience given their situation (Brewin, 1995). Even for mentally healthy persons, social support processes may present problems (Parry, 1995). "All help has the potential for negative messages about the self (Parry, 1995, p. 282). Seeking and receiving social support can have self-esteem costs (Fisher, et al., 1988) if the individual views offers of support as evidence of their own inferiority (Gilbert, 1995). Friends are more likely to expect reciprocity in the giving of support (Champion, 1995), and "caring debts" (Pearlin, 1985) can be a source of stress because of feeling obligated or unable to repay, with those who give more feeling taken advantage of or exploited (Gilbert, 1995). Accepting help may be viewed as a loss of autonomy and a reminder of what the individual is lacking, which may be interpreted as a loss of status (Gilbert, 1995). In an effort to maintain status in the eyes of others, the need for support is weighed against how others may perceive and respond to the need, and fear of negative evaluation may inhibit expression of the need for help or perceptions about its availability (Gilbert, 1995). Asking for help that is not forthcoming could be experienced as rejection, thereby attacking the individual's sense of being a person worthy of consideration (Buunk & Hoorens, 1992). Because a person does not seek social support, or avoids closeness and intimacy, does not mean that they do not want it. Rather they may be distrustful or too shame-prone to seek it  Wives of Alcoholics 49  . ...... Shame arises from the fear of being seen as weak, inferior, and the object of scorn, which motivates concealment and deception (Gilbert, 1995, p. 123, 128). The way a person self-presents is related to status and the avoidance of shame within the context of values assigned by the social environment, and loss of status and approval negatively impacts mood and behavior (Gilbert, 1995), thereby potentially contributing to depression. Her husband's alcoholism could conceivably represent a loss of status to the wife of an alcoholic and negatively alter her perception of social support. She may also compromise potentially supportive relationships in an effort to avoid her husband's anger, particularly if she has a history of physical or sexual abuse (Coyne & Downey, 1991). If the wife of an alcoholic decides to seek help, the result may still be inadequate to address her needs. According to social comparison and rank theories, support is most likely to be solicited and received from socially similar others, who may also have low self-esteem, few resources, and little ability to provide assistance (Gilbert, 1992, 1995; Thoits, 1986). Ironically, adverse life events may directly cause depletion of potential support (Parry, 1995), particularly when the stressors are chronic. Research has found that people tend to find expressions of negative affect unattractive and alienating (Buunk & Hoorens, 1992; Lane & Hobfall, 1992). It is clear that there are many forces at work that inhibit the perception and utilization of social support for wives of alcoholics. Affiliation with friends may actually be deleterious if the stressor is embarrassing or evokes social disapproval (Buunk & Hoorens, 1992), as does alcoholism. The wife of an alcoholic, like others, makes a cost-benefits analysis in subjectively determining whether social support is available to her (Fisher, et al., 1988; Parry, 1995).  Wives of Alcoholics 50  Theories Past and Present About Wives of Alcoholics The negative social context faced by wives of alcoholics has evolved over time as alcoholism has come to be viewed as a disease. In search of cause and maintenance factors, early work undertaken to determine the etiology of alcoholism focused on wives of alcoholics more than on the alcoholics themselves. As Bepko (1986) reminds us, "Women are assumed to be responsible ultimately for the emotional well-being of everyone" (p. 71). Historically, as wives of alcoholics have sought professional help, they have often been assigned blame instead of being given support. Though they may respond to carrying the burden of change because of their sense of relational responsibility, they may continue to struggle with a lowered sense of self as a result of living in an alcohol-embedded marriage, which is not a supportive context (Banister & Peavy, 1994). The Disturbed Personality Hypothesis The Disturbed Personality Hypothesis (DPH), proposed by Lewis (1937) and based on psychoanalytic theory, held that the woman who marries an alcoholic or pre-alcoholic is a deeply disturbed individual from a dysfunctional family of origin and that her behavior is primarily an expression of her intrapsychic state. Rooted in psychic determinism, this theory assumed that the spouse did not marry the alcoholic by accident and that her own mental process lead her to the alcoholic, contributes to his misuse of alcohol, and renders her unable to leave (Paolino & McCrady, 1977). According to this theory, she chooses a weak, needy, and dependent man that she can dominate in order to mask her own inadequacies and satisfy her deeply neurotic needs (Whalen, 1953), which she has repressed, thereby rendering them more powerful in determining marital dynamics and drinking behavior (Paolino & McCrady, 1977). She is characterized as an  Wives of Alcoholics 51  "excessively anxious, sexually inadequate, guilt ridden, and abnormally angry woman with pathogenic childhood experiences" (Paolino & McCrady, 1977, p. 13). Gaertner (1939) suggests that she is an aggressive woman chosen by the alcoholic to meet his masochistic needs. As Boggs (1944) put it, the wife of the alcoholic "knocks the props from under him at all turns, seemingly needing to keep him ineffectual so that she feels relatively strong and has external justification for hostile feelings" (p. 463). From this perspective, her husband's weakness allows her to exercise neurotic control over vacated masculine roles, and his absent or impaired sexuality operates as a convenient complement to her frigidity. It is further suggested that because of the wife's masochistic need to suffer and play the martyr, she stays with her alcoholic husband, who provides a convenient vent for stored hostility and aggression carried over from her pathogenic childhood (Gaertner, 1939). Both the alcoholic and his wife are portrayed as seeking to satisfy sadomasochistic needs in a reciprocal process of inflicting pain. Gaertner (1939), a female psychotherapist who studied 15 alcoholic marriages and was one of the first to report about wives of alcoholics, recommended this approach to treatment: If the caseworker can help the wife release some of the pent-up hostility aimed at the husband, he will be spared at least a little of the nagging and ridicule which emphasize his inadequacy as the head of the family and are part of the pattern driving him to drink (p. 43). Whalen (1953), also a female clinician, described four types of wives who marry alcoholics to meet their own neurotic needs: Suffering Susan who marries an alcoholic to punish herself (masochism); Controlling Catherine who marries a weak alcoholic man that she can control to defend against her distrust of men; Wavering Winifred who marries an inadequate man so she can feel needed as a defense against profoundly low self-esteem; and Punitive Polly who  Wives of Alcoholics 52  feels a need to punish men by competing with and defeating them. Lewis (1954), a female psychotherapist who had worked with 50 wives of alcoholics, described them as insecure, dependent women, confused about their sexual identity, who engage in rationalization to meet their aggressive need to be punitive. She cited obesity, vomiting, and preoccupation with food as evidence of unresolved oral and dependency needs and described them as "literally hungry for love and acceptance" (p. 12). Price (1945) interviewed 29 wives of alcoholics and subjectively described them as insecure, anxious, hostile, and dependent, but resentful and aggressive because their abnormal dependency needs go unmet. According to Fox and Lyon (1955), the alcoholic's wife unconsciously wants no part of his recovery and ensures that the drinking problem remains unsolved because his recovery would threaten her neurotic need for him to be helpless and inferior. Price (1945) stated: "Unconsciously and perhaps even consciously, she fought treatment of her spouse as one more way she could keep him inadequate" (p. 623).  The Decompensation Hypothesis (DH), an  extension of the Disturbed Personality Hypothesis (DPH), claimed that if her alcoholic husband attained sobriety, the wife would decompensate into more severe personality disturbance and deterioration (Edwards, Harvey & Whitehead, 1973). These theories were formulated on the basis of clinical impressions rather than empirical data, and were not supported by later research. Clinical cases often illustrate abnormal more than normal cases, and it has been argued that they do not represent the general population of alcoholics and their wives (Paolino & McCrady, 1977). For example, wives of alcoholics who do not decompensate upon their husband's sobriety will not have the same need to seek therapy.  Wives of Alcoholics 53  One of the problems in studying theories based on the Freudian model is that its principles cannot be precisely quantified and objectively tested (Paolino & McCrady, 1977). Following are several studies that are part of a much larger body of research challenging the validity of the DPH and the DH: In contradiction to the Decompensation Hypothesis, Burton and Kaplan (1968) found the marital relationship and mental health of both partners to improve when alcohol abuse was decreased or eliminated. Using the Minnesota MultiphasicPersonality Inventory (MMPI), Ballard (1959) found alcoholic husbands to be significantly more disturbed than controls, and their wives to be at least as well adjusted as controls and better adjusted in terms of their ability to adapt, thus showing wives of alcoholics to be psychologically normal. This was later corroborated by Kogan, Fordyce, and Jackson (1963), who found that wives of alcoholics did not display significant personality disturbance. The idea of one particular deviant personality typefor alcoholics or their spouses has been consistently shown to be false (Paolino & McCrady, 1977). Although the Disturbed Personality Hypothesis, which viewed wives as villains, has been formally abandoned for lack of empirical substantiation, some practitioners remain under its influence (Kokin & Walker, 1989), and some tenets of the DPH have been incorporated into a new theory that enjoys widespread acceptance today (Troise, 1992). The assumptions underlying these theories are culturally embedded, and therefore they persist. The Sociological Stress Theory Mowrer (1940), who compared 25 wives of alcoholics to wives of nonalcoholics, was actually the first researcher to consider the wife's behavior as a response to the situational stress of living with an alcoholic husband, but the idea was not pursued further at the time. Later the  Wives of Alcoholics 54  Sociological Stress Theory was developed by Jackson (1954) and based on three years of work with 55 women who were members of Al-Anon, a self-help organization for relatives and friends of alcoholics that is patterned after Alcoholics Anonymous. The main flaw in Jackson's (1954) work was that she only studied wives of alcoholics that were members of Al-Anon, who have been shown to be less symptomatic, more educated, and of higher socioeconomic status than wives of alcoholics that are not members of Al-Anon (Bailey, 1965, 1967). Utilizing interviews and recordings of Al-Anon meetings, Jackson (1954) concluded that families of alcoholic men go through seven critical stages of adjustment as they attempt to cope with the alcoholism. Wives of alcoholics were seen as victims, with the stress of living with an alcoholic viewed as causing personality disturbance in the wife. Her attempts to adjust and cope were postulated to result in enabling him to continue to drink, even though her intention was to achieve stability. The wife rationalized her mate's behavior as she lied and covered up for him, in an attempt to excuse his behavior, which alienated friends, increased social isolation, and eroded her self-esteem. When she could no longer deny the alcoholism, she endeavored to control it by begging him to stop drinking, by refusing sex, and by pouring out his alcohol, which only angered him and gave him reason to drink more. Usually some crisis occurred that created chaos and disorganization, and she felt angry and fearful for herself and her children, as well as powerless to change the situation herself. Finally, accepting the permanence of the situation, she would take over the responsibilities that he had neglected or abandoned (Kokin & Walker, 1989), seeking to escape the problem by reorganizing the family without the husband. If the husband achieved sobriety, the family was reorganized again to include him, which required the wife to risk becoming emotionally vulnerable again despite past hurts.  Wives of Alcoholics 55  This theory continued to view wives of alcoholics as suffering from personality disturbance, but viewed it as brought on by the cumulative crisis (Jackson, 1954) of living with an alcoholic spouse. The wife was still cast in an enabling role, and viewed as provoking the alcoholic to drink. Maxwell (1977) is quoted as stating that the wife of the alcoholic "actually helps her husband to drink" (Decker, Redhouse, Green & Starrett, 1983, p. 464). Although the existence of Jackson's seven stages has not been supported by quantitative studies (Paolino & McCrady, 1977), her concept that the psychological symptoms displayed by the wife of an alcoholic are induced by the stress of living with an alcoholic has been clearly demonstrated by research, of which a few examples will be included here. Bailey, Haberman, and Alksne (1962) compared four groups of married women and found that as the distance from the alcoholic behavior increased, symptoms in wives decreased. Kogan and Jackson (1965) found the greatest psychological disturbance in wives of drinking alcoholics, significantly less in wives of abstinent alcoholics, and the least in wives of nonalcoholics. Bailey (1967) found similar results in the same three groups, except that wives of abstinent alcoholics did not differ from wives of nonalcoholics. The Psychosocial Theory It became apparent that wives of alcoholics could not be classified into a homogeneous group (Edwards, et al., 1973), and thus the next emerging theory, the Psychosocial Theory, combined the elements of the previous two. Some wives were said to have disturbed personalities due to inadequate mothers and unhappy childhoods, thus blaming other women, while some were described as reacting to the stress of their husbands' alcoholism. A 1973 pamphlet from the U.S. National Council on Alcoholism referred to the wife's personality as  Wives of Alcoholics 56  "even more disturbed than her alcoholic husband's," and a 1981 article by Roth declared that "sometimes the non-alcoholic spouse needs the alcoholic to be sick" (Decker et al., 1983, p. 467468). The wife was accused of depriving her alcoholic husband of his masculinity and was seen as deviant if she took over masculine roles. In order for the alcoholic to recover, it was held that his wife would need to relinquish control of masculine roles (Decker, et al., 1983). Thus theorists continued to blame the wife of the alcoholic, who was really more a victim than a perpetrator, justifying the inequality of the system by looking for and constructing defects in the victims of that inequality. According to Kokin and Walker (1989) and Asher (1992), the role of the wife in the husband's alcoholism likely has been overstated, considering the fact that single men also manage to escalate into alcoholism without a wife to participate in the process. In this case attribution of responsibility is usually toward the mother, again blaming women, even though inacessability of the father may be more relevant to feelings of inadequacy as a man, with the distant or absent father providing only an elusive model of manhood (Holten, 1990; Nol, 1991; Pittman, 1990). A new focus on attribution of responsibility for alcoholism by theorists and treatment providers must include ways in which the alcoholic justifies himself by the meaning he assigns to any. given situation and the choices he makes—that he is a primary player in the drama and can be empowered to make different choices.  Family Systems Theory Interaction-based Family Systems Theory, introduced in the 1950's, brought forth the idea that all family members play a role in the functioning of each individual, and that to understand the behavior of one individual, it is necessary to understand the processes of the family, group in which he lives (Becvar & Becvar, 1988). Families are considered to be governed  Wives of Alcoholics 57  by rules, with different family members filling various roles, both practical and emotional. From this point of view, it is conceivable that the family structure can require a family member not to function in a capable way in order to serve a particular purpose and create balance in the system (Paolino & McCrady, 1977). The family member, bearing the symptom is referred to as the identified patient, yet the symptom is seen as being maintained by the family system as a whole. The family is seen as symptom-maintained as well as symptom maintaining, with this reciprocal influence creating an equilibrium that resists change as a means of minimizing disorder (Becvar &Becvar, 1988). This concept applied to alcoholic families assumes that alcoholism is a symptom that plays an adaptive function in the family system. In fact, according to Steinglass, Bennett, Wolin & Reiss (1987), "alcoholism is a condition that has the capacity to become a central organizing principle around which family life is structured" (p. 9). For example, drinking may provide a way for the alcoholic to feel powerful by expressing strong feelings, while at the same time avoiding responsibility by blaming the alcohol for his behavior (Schaap, et al., 1991). As the family interacts around the alcoholic behavior, they create new adaptive responses that may act to reinforce the drinking cycle, which in many cases oscillates between "wet" (drinking) and "dry" (not drinking), (Bepko, 1986; O'Gorman, 1991). Stability becomes organized around the alcohol as the alcoholic husband engages in avoiding responsibility and his wife necessarily assumes more responsibility (O'Farrell & Birchler, 1987). The systemic element is evident in a study of 227 couples (comprised of an alcoholic husband and a non-alcoholic wife) by Wright and Scott (1978), who found the wife's participation in two or more types of treatment (inpatient, outpatient, post-treatment counseling, Al-Anon) to be positively associated with the husband's abstinence.  Wives of Alcoholics 58  With the advent of Systems Theory, emphasis shifted from the wife of the alcoholic to the marriage as a family subsystem, characterized by role confusion, dependency conflicts, and psychiatric disturbances (Schaap, et al., 1991). The family systems approach continues to be in use today as a viable framework from which to conceptualize family functioning and deliver therapeutic interventions, though it has been criticized by feminists as detouring responsibility away from the alcoholic, blaming mothers and wives, and ignoring cultural factors such as gender training and power differences (Boss & Thorne, 1989). Codependency Theory Codependency Theory evolved in the early 1970's after alcoholism became conceptualized as a disease. This "pop" psychology concept, which has been utilized by treatment programs, therapists, and Al-Anon, defines all wives of alcoholics as inevitably codependent and has thus returned the clinical focus to wives, who are seen as coming from dysfunctional families of origin in which children lack close, intimate relationships with parents (Troise, 1992). Lacking the experience of intimacy in their primary relationships, these women are thought to seek out partners who will allow them to maintain emotional distance. Their husbands' alcoholism provides them the distance they need from the intimacy they fear (Beattie, 1987; Schaef, 1989). This theory bears remarkable similarity to the Disturbed Personality Hypothesis, except that with the more recent normalization of androgynous gender roles, wives of alcoholics are not pathologized for taking over traditionally masculine roles, which is evidence that "designations of deviance are culturally relative and change over time" (Asher, 1992, p. 189).  Wives of Alcoholics 59  Codependency theory views wives as "enabling" their husbands to drink when they directly engage the alcoholism and its consequences, yet "such behavior is often the only option available to a wife who must assume responsibility for her and her family's welfare given her alcoholic spouse's inability to do so" (Troise, 1992, p. 53). Interestingly enough, Orford, et al. (1975) found engagement or involvement-type strategies, which could be referred to as codependent behaviors, to be more associated with resolution of the drinking problem than disengagement strategies, such as might be encouraged by Al-Anon. Wiseman (1991) found A l Anon attendance by wives not to be related to husbands attaining sobriety. The language of pathology. Codependence has been declared to be a sickness, and therefore the disease label has been applied. Designation as a medical condition provides an aura of legitimacy yet becomes a directive that treatment is needed to aid in what is described as a lifelong process of recovery. Codependence is thereby a no-exit model (Asher, 1992), because recovery is ongoing without an endpoint other than death. Since the medicalization of alcoholism, recovery has become big business. Asher (1992) states: "Medicalization increases directly with its economic profitability" (p. 196). The label of codependent or co-alcoholic bears "the stigma of deviance and the medical trappings of sickness and recovery" (Asher, 1992, p. 1880). Codependency is "the outcome of a social process of constructing a medical reality, rather than of discovery or identification of an objective, pre-existent, distinguishable illness" (Asher, 1992, p. 190; Asher & Brissett, 1988). Language functions as a powerful agent in shaping our definitions of self and reality. Terms like "enabler" or "codependent" suggest the presence of an accomplice. "If alcoholism is a disease, how could the wife have given it to her husband?" (Kokin & Walker, 1989). Through  Wives of Alcoholics 60  the process of medicalization, wives of alcoholics are assigned the label of deviance to compensate for their husband's alcoholism and thereby maintain the existing power differential (Krestan & Bepko, 1991). These labels make women look negative, feel negative, and suffer negative treatment at the hands of professionals. Alcoholism has been classified as a disease because it causes a physiological interaction within the body. Diabetes is also a disease, affected by the ingestion of a substance called sugar, yet we do not talk about being co-diabetic (Decker, et al., 1983). Wegscheider-Cruse & Cruse (1990) actually construe codependency to be a brain disorder caused by an excessive rush of brain chemicals reinforcing certain behaviors that cause these chemicals to be released, thus creating an addiction to those behaviors. Examples are: dopamine released by running, causing excitement; serotonin released by overeating or relationship dependency, giving comfort; and norepinephrine released by workaholism or caretaking, providing a sense of power and control. This is an apparent effort to justify the disease concept of codependency, which seems to have been carried too far. Although the alcohol treatment industry widely embraces the construct of codependence, empirical substantiation for the existence of such a condition is nonexistent (Gierymski & Williams, 1986; Troise, 1992). Troise (1992) sought to empirically verify one measurable facet of the construct of codependency. He compared wives of alcoholics with wives of nonalcoholics in terms of their ability to experience high levels of intimacy with their best or closest friend outside of the marital dyad. He divided these two groups into four based on whether they came from an alcoholic family of origin. He found no significant effect on intimacy frequency, intimacy intensity, or intimacy total from husband alcoholism, parent alcoholism, or the interaction of the two. This is the first study to test a personality deficit attributed hypothetically to codependents, and it did not support the idea that wives of alcoholics, who have been declared  Wives of Alcoholics 61  to be codependent strictly on the basis of their husbands' alcoholism (Asher, 1992), could be differentiated on the basis of impaired capacity for intimacy.. Although in theory codependence can apply to men as well as women, in practice women are the primary bearers of this label (Krestan & Bepko, 1991; Kokin & Walker, 1989). According to Kokin and Walker (1989), the theory of codependency is grounded in the Freudian idea that women possess biologically-based masochistic tendencies which cause them to enjoy suffering. Troise (1992) declares Codependency Theory to be a "thinly disguised version of previously abandoned Freudian concepts which blame women for the difficulties of men" (p. 53). Socialized to be codependent. Although definitions vary, according to Walters (1990), "Codependents fit the archetypes usually drawn to describe women—overinvolved, depending upon others for approval, not taking care of herself, having poor boundaries, intoxicated with relationships, too willing to assume blame, putting the needs of others before herself (p. 54). As the wife of an alcoholic stated, "I came to discover that in some ways I was being raised to fit that codependency position" (Asher & Brissett, 1988, p. 338). Tavris (1992) observes that the symptoms ascribed to codependency are synonymous with the female role, while the prescription for recovery is to develop traits of the stereotypic male, evidence that women's way of being is more negatively valued on a cultural level. Individual men are occasionally assigned this label, but only when they act in traditionally feminine ways. Sometimes it seems as if anyone who subordinates his or her own needs to take care of others might be labeled codependent. Does this mean that altruism and unselfishness are diseases?.. . It would be a pretty cold world if everyone put their own needs first all the time (Treadway, 1990, p. 41).  Wives of Alcoholics 62  As stated by Walters (1990), "Mature behaviors are conceptualized as taking care of oneself, putting one's needs first, loving oneself. Whatever happened to the poor reputation of narcissism?" (p. 55). The tendency for women to be self-blaming makes them receptive to the label of codependency, which in turn promotes further self-blame (Asher, 1992). In reporting the experience of wives of alcoholics, Asher (1992) states that formal intervention programs "narrow her identity to 'codependent,' enhancing her readiness for self-blaming" (p. 207). Wives of alcoholics, who thought their actions were aimed at family survival and caretaking, have to be taught that they are codependent (Asher, 1992). Their socialization to accept responsibility for others and their history of self-blame make them easy targets for promoting such a label of deviance. Imposing the label of "codependent" on wives of alcoholics is like a re-enactment of the alcoholic's projection of blame. Preserving the hierarchy of patriarchy. In viewing the wife as co-alcoholic or codependent, responsibility is detoured away from the male alcoholic: Women are ascribed more pathology in this culture than men, so in any situation where the male clearly has the impairment, it must be the case that the woman as well as her children are sick, too. The codependency l a b e l . . . [fosters] denial of male accountability (Krestan & Bepko, 1991, p. 52). This is a result of hierarchical thinking, which dictates that men must be stronger than women in order to be masculine; if men display anything that may be construed as weakness, it seems logical to dilute accountability by redefining it as sickness and then declaring women to be  Wives of Alcoholics 63  equally sick. When interviewed by Tavris (1990), Harriet Lerner, a family systems therapist, said in relation to codependence, that society is more comfortable with women who feel inadequate, self-doubting, guilty, sick, and "diseased" than with women who are angry or confronting . . . . Women get more sympathy and support when they define their problems in medical rather than political terms (P- 43). But then, the medical model involves being sick and subordinating to a higher authority (Walters, 1990). Women are given the power to create dysfunctional family patterns that feed a problem which is overrepresented in the male population (75% of alcoholics are male, according to Schaap, et al. 1991), just as they are seeking power and equality of a different kind (Walters, 1990). "Until women collectively have acquired an equal share in the power to develop and impose labels, the controlling of women through an imputation of spoiled identity will persist" (Schur, 1984, p. 235). According to Bepko (1988), the family system mediated by alcohol becomes a precarious equilibrium between the extremes of overresponsibility and underresponsibility. Women are culturally conditioned to be overresponsible for the emotional and physical well-being of others, and when the husband takes an underresponsible role through excessive drinking, his wife does more of what she has already learned to do as she attempts to make up for the deficit (Krestan & Bepko, 1991). Her behavior is a normal response to a desperate situation and should not be thought of as sickness (Krestan & Bepko, 199.1; Kokin & Walker, 1989). Her actions are geared toward survival (Asher, 1992; Wiseman, 1991), as she attempts to reduce anxiety and create order in a very disordered environment. In an effort to preserve her family, she gradually takes  Wives of Alcoholics 64  on the roles her husband has vacated in an effort to minimize potentially disastrous consequences for her family. Of necessity, she plays a more autonomous role than she ordinarily would, while he plays a more dependent role, his self-awareness dulled by alcohol. The addictive system becomes a mirror for the fundamental paradox of gender issues in families: men are socialized to show autonomy and hide dependence while women are socialized to show dependence and hide autonomy. Alcohol permits either suppression or expression of impulses that run counter to these contradictory restraints (Bepko, 1988). In this way alcohol functions to maintain the sex-role status quo. The language of codependency locates the problem in individuals instead of in the larger structure of society (Krestan & Bepko, 1991), which is part of the system that supports alcoholism in families. Codependency is a social construction of assigned meaning and not an objective condition (Asher, 1992; Asher & Brissett, 1988). The concept of codependency began as a way to help wives of alcoholics understand how they can get drawn into a destructive system, reacting with well-intentioned but ineffective behaviors. To frame codependency as a "disease" which requires "recovery" is to pathologize women's socialized way of being in that overfunctioning in relationships has been the accepted cultural norm for women (Krestan & Bepko, 1990). In chastisement of traditional family therapists, who are inclined to blame the wife and mother in the family for its ills, Goldner (1987) states: "There is something fundamentally disrespectful and punitive about an approach that exaggerates women's domestic power, trivializes and pathologizes their domestic contributions, and then holds them ultimately responsible for family functioning" (p. 115). It is disturbing that attempts to adjust on the part of wives of alcoholics are viewed by many  Wives of Alcoholics' 65  professionals as symptoms of psychological illness. The wife's struggle to stabilize in spite of the reality of the alcoholism could as easily be interpreted as a sign of health rather than sickness, depending on what one is looking for (Kokin & Walker, 1989). She struggles to hold the family together through the crisis despite her own feelings of fear, hurt, and shame. She is thought to be in denial if she tries to maintain an appearance of well-being, and she is accused of covering up if she lies to the boss to protect his job and her family's economic well-being. Caplan (Introduction, Kokin & Walker, 1989) says that wives of alcoholics use all the strength and creativity they can muster to cope with the adversity they face: "Such people, when male, are usually called heros, not codependents" (p. 13). Codependency needs to be redefined as overresponsibility and overresponsibility needs to be understood as a positive impulse gone awry. Relational responsibility needs to coexist with responsibility to self, and the feminine emphasis on feeling needs to be acknowledged and celebrated while the feminine focus on relationship needs some redirection without its being pathologized" (Krestan & Bepko, 1991, p. 63). Whitfield (1989) describes codependence as "lost selfhood" (p. 19), a condition similar to Jack's (1991) description of depressed women. According to Mendenhall (1989), the family members of an alcoholic must suppress their self-awareness in order to survive, with their needs going unmet and their feelings unexpressed.  Wives of Alcoholics 66  CHAPTER III Literature Review Previous work pertaining to the impact of life stress and social support on depression will now be explored. Because this has not been specifically accomplished quantitatively with wives of alcoholics, studies will be reviewed that investigate the modifying effect of social support on the stress-depression relationship in samples other than wives of alcoholics. Due to the genderbiased nature of depression and social support, studies employing samples undifferentiated by sex or not including women will not be reviewed. Even though in the current study, onset factors associated with depression are of greater interest than recovery factors, studies that address either will be considered because social support appears to be relevant in both situations. Findings from three recent qualitative studies of wives of alcoholics will also be discussed, in terms of wives' experience of the stress of their husbands' alcoholism, availability of social support, and feelings of depression. The concluding focus will be on two studies examining the impact of employment status on wives of alcoholics, and the workplace as a potentially supportive context.  Social Support as a Modifier of Stress There is solid evidence of the protective role of social support against depression (Henderson, 1992; Monroe & Johnson, 1992), although comparing individual studies is complicated by differing operationalizations of the construct of social support. It may be measured objectively by evaluating social network structure, degree of social integration and participation, or quality and content of social intimacy in close relationships (Gottlieb, 1985). It may be measured subjectively utilizing individual perceptions of the availability and adequacy of  Wives of Alcoholics 67  support resources (Sarason, et al., 1990). Stronger relationships between social support and depression have been found using subjective measures (George, et al., 1989; Henderson, 1981). A general view from Gottlieb (1985) of how social support improves coping, thereby buffering the deleterious effect of stress, is that the individual first cognitively appraises the stressful situation to evaluate the degree of threat to personal well-being and then assesses personal and environmental coping resources. The social factor may operate covertly, as downward comparisons made with persons in worse condition inspire appreciation of what is good, and upward comparisons with those who have overcome, kindle hope. Overtly, the discussion of problems with supportive others can reduce self-blame and encourage the formulation of problem-solving strategies, as well as result in tangible forms of aid. Stress is defined by Burchfield, et al. (1982) as that which provokes intense reaction and is perceived as threatening by an individual. They hypothesized that stressors related to a highly valued need are more threatening and therefore more likely to induce illness. Affiliative needs were studied as a representation of highly valued needs in a sample of 44 female undergraduate psychology students. They were differentiated by testing from a group of 134 on the basis of strength and degree of fulfillment of affiliative needs and degree of interpersonal stress, and then randomly assigned into four groups with 11 in each. The groups were fulfilled and unfulfilled needs for affiliation and high and low interpersonal stress. Findings were that individuals with unfulfilled needs for affiliation and high stress experienced more symptoms of illness than those with fulfilled needs for affiliation and high stress or unfulfilled needs for affiliation and low stress. A larger sample would improve reliability and internal validity of the data.  Wives of Alcoholics 68  Wives of alcoholics, due to alienation from their husbands and attrition of social ties as alcoholism worsens, could be expected to have unfulfilled needs for affiliation in a context of high stress. Therefore, based on this study, more symptoms of illness could be expected to be manifest in wives of alcoholics, although generalizing from undergraduate university students to wives of alcoholics, even though both groups are female, would exceed the limits of external validity in making conclusive statements about wives of alcoholics. However, a direction for further study is indicated. Deterioration of the marital relationship and social isolation, as experienced by wives of alcoholics, could lead to unfulfilled needs for affiliation, and therefore could be hypothesized to lower self-esteem and to increase psychological symptoms of depression. Billings, Cronkite, and Moos (1983) studied negative events, life strain, and social resources in a sample of 424 depressed men and women seeking psychiatric treatment compared to a random sample of 424 nondepressed matched community controls. Those with symptoms of biologically based depression, manic symptoms, or alcohol abuse were screened out. The Health and Daily Living Form was used to assess social background, stressors, coping, social resources, and daily functioning, and the Family Environment Scale and the Work Environment Scale were used to evaluate family and work related stresses and resources. Demographic factors taken into account were age, ethnicity, education level, and occupational prestige as determined by Duncan's Socio-Economic Index. Psychosocial factors were found to be significantly related to the presence of depression. More acute negative events and higher levels of ongoing life strains differentiated depressed from nondepressed persons, as did coping responses and social resources. Depressed persons  Wives of Alcoholics 69  used less problem-focused coping and more emotional discharge coping, even though they sought information and guidance. In addition, their social resources were more limited in both quantity and perceived quality of support, and they were more likely to be unemployed and looking for work than employed either part time or full time. Gender differences in results included that physical home environment as a stressor did not differentiate depressed men from controls as it did women, and strains of work setting or having an ill spouse did not differentiate depressed women from controls as it did men. Although 44.5% of the patients and matched controls were married, comparisons of results were not given on this basis, which could tend to dilute the differences in depression.between men and women in that unmarried men and married women tend to experience more depression. This factor seriously threatens internal validity. Two of the three general areas found by Billings, et al. (1983) to differentiate depressed persons from controls seem to apply to wives of alcoholics. They experience the ongoing strain of their husbands' alcoholism as well as attrition of social resources, including the loss of emotional intimacy in the marital relationship, although their efforts to cope could be construed as problem-focused in that they make direct attempts to alleviate the problem of alcoholism and its effects on the family in their effort to survive.  :  Brugha, et al. (1990) interviewed 130 depressed outpatient men and women in equal numbers, each evaluated at baseline and approximately four months later to determine changes in adversity and social network in relation to recovery from depression. Personality and psychiatric status was also assessed. Social support was measured using the Interview Measure of Social Relationships (Brugha, et a l , 1987). • Recovery rates were similar in men and women, however the relationship between social support and recovery was much stronger in female patients than  Wives of Alcoholics 70  in the total sample, and living as married was associated with recovery for men but not for women. For women, predictors of recovery from depression were: 1) number of primary group members named and contacted; and 2) satisfaction with social support. For men, predictors of recovery were: 1) negative social interaction; 2) living as married; and 3) number of non-primary group social contacts. They proposed that negative social interaction not showing up as a factor for depressed women may be due to them being more likely to deny its existence or to avoid confrontation, and men naming non-primary group social contacts may be due to talking about close relationships being less socially acceptable among men in British culture. They suggested that differences in socially acceptable behavior for men and women may explain the differences between them in predictors of recovery from depression. The findings most relevant to the current study were that satisfaction with social support was particularly salient for women, and living as married did not predict improvement in depressive symptoms for women. These results come from a group who presented for treatment, and the present study draws from a group whose alcoholic husbands are presenting for treatment, which could indicate that potential selection bias may be similar in both groups. Generalizing should be done with caution, however, because subjects were not presenting for treatment for the same reason. Also, Brugha et al. (1990) studied recovery factors rather than onset factors in depression, which could be somewhat different. Nevertheless, the results are interesting and merit further investigation. George, et al. (1989) interviewed 150 clinically depressed middle-aged and elderly inpatients (90 females and 60 males) and did follow-up telephone interviews 6-32 months later, using exact time to follow up as a control variable. The hypothesis that impaired social support at baseline would be related to poor outcome, with subjective social support having a greater effect  Wives of Alcoholics 71  than objective support, was confirmed. The Center for Epidemiologic Studies Depression Scale (CES-D) was used as the outcome measure of depression, and the Duke Social Support Index provided information about network size, interaction frequency, instrumental support, and subjective social support. Also considered were age, sex, marital status, psychiatric history, life events, and psychiatric comorbidity of dysthymia, alcohol abuse, or anxiety. Findings were that life events were not related to outcome, and the married and those low on social interaction and subjective social support were less likely to have recovered. An interaction of the effects of sex and marital status on recovery was not reported, but would have given more validity to their finding about marital status. Generalizability is limited due to age restriction of the sample, and inpatient status, as well as the fact that the criterion variable is recovery from depression rather than onset. The predictive value of subjective social support on recovery from depression is worth considering, however. Goering, Lancee and Freeman (1992) studied marital support and recovery from major depression in an inpatient sample of 45 women aged 18 to 65 who were living as married. Serious psychiatric disturbance and alcoholism were screened out of the sample. Both the women and 42 of their partners were interviewed at baseline and six months following. Diagnosis of depression was made using the Schedule for Affective Disorders, and severity of depression was assessed using the Hamilton Depression Score and the Global Assessment Scale. The LIFE weekly psychiatric-status scale was used to track the course of the depression during the time between baseline and follow-up, and marital support was assessed using the Camberwell Family Interview to evaluate the marital climate in conjunction with A Five Minute Free Speech Sample evaluating critical comments made and tone of voice used by both partners.  Wives of Alcoholics 72  Of the 45 subjects, 51% recovered and an additional 24% partially recovered during the six-month period. The strongest predictor of recovery was the woman's perception of support she received from her husband, with her view of their relationship, his communication with her, and his relationship to the children all predicting outcome. The husband's rating of the marriage prior to onset was also predictive of outcome, and partners showed high agreement in their ratings of the marriage. Of particular interest, subjects' perceptions of support were found to accurately reflect actual marital support. This was an interesting and thorough piece of research, although generalizing should be done with caution in that these results are based on major depression requiring hospitalization and focus on factors associated with recovery rather than onset. However, severity of depression was not found to be predictive of outcome and may not be an issue. Paykel, Emms, Fletcher, and Rassaby (1980) assessed life events, social support, and depression in 120 women at six weeks pospartum and found a 20% incidence of mild clinical depression. Single private interviews were conducted, and depression was evaluated using the Raskin Three Area Depression scale, which focuses on feelings reported, physical appearance, and secondary symptoms of depression. Life events were evaluated from the start of pregnancy to six weeks postpartum using the Interview for Recent Life Events, and interview questions designed to elicit objective information about actual social circumstances were evaluate social support and marital climate. Negative life events that were moderate to severe were associated with onset of depression, with 75% of those experiencing depression also experiencing these negative events. Depressed subjects tended to be younger, and poor marital communication and lack of household  Wives of Alcoholics 73  help from husband was associated with onset of depression in those experiencing negative events, as was lack of a confidant other than the spouse. These results highlight the mediating influence of support from others, especially husbands, in a population of women experiencing the strain of a new baby in addition to other stressful life events. Similar results might be expected in a sample of wives experiencing the ongoing strain as well as specific negative events associated with husbands' alcoholism. These findings, relevant to a population of women experiencing the strain of a life stage transition, lend strength to a hypothesis about wives of alcoholics, who often lack support from their husbands, even though conclusive statements would be inappropriate. Brown, et al. (1986) conducted a prospective study of social support, self-esteem, and onset of depression using a random sample of 400 working class women between the ages of 18 and 50 and having at least one child at home. Data collected at baseline included measures of social network and self-esteem as well as history of any psychiatric disorder within the previous year. Follow-up data was collected approximately 12 months after baseline data and included measures of life-event stress and social support received during any crises during the year. Of the initial sample, 353 women completed the follow-up interview. Of these, 50 were excluded from the analysis of onset factors because they were already depressed at baseline, leaving a sample of 303. The Present State Examination (PSE), short version, was used to collect information about symptoms during the 12-month interval, and the Self Evaluation and Social Support (SESS) schedule was used to interview respondents with regard to aspects of self, marriage, motherhood, housework, job, and social ties outside the home. Life events were measured using the Life Events and Difficulties Schedule (LEDS).  Wives of Alcoholics 74  Findings were that 91% of women who had an onset of depression during the year experienced a severe life event within six months of onset. Those with a negative evaluation of self were nearly three times as likely to become depressed in the presence of a provoking event, but negative evaluation of self was unrelated to onset without a provoking event. Of the 32 onsets of depression found, only two occurred without a provoking agent, thus the findings that follow relate to the 150 women who experienced a provoking agent. Results were given separately for the 101 married mothers and the 49 single mothers, and we will focus here on the findings for the married women because living as married (legally or common law) was part of the inclusion criteria in the sample of wives of alcoholics in the current study. Although the presence of a very close relationship, with or without confiding, was not found to be related to onset of depression, emotional support during a crisis from a husband or very close tie was very related to risk of becoming depressed, as was negative interaction with a husband or close tie. However, crisis support from a non-core tie was not related to reduced risk of depression. In addition, women who experienced confiding with a husband or close tie at baseline yet did not receive crisis support later when needed, thereby contributing to a sense of being let down, experienced a twenty-fold greater risk of becoming depressed in the presence of a provoking agent. Married women were more likely to experience a sense of being letdown. Thus the loss of expected support from a husband or close tie over time showed an especially high risk of onset of depression. The worst of severe events were "husband events," such as infidelity, mental or physical illness, legal problems, or job loss, which were highly related to a history of lack of support from him. These events were experienced by 27 of 101 married mothers who suffered severe events, with only one husband of the 27 providing crisis support. Of the 27 having husband events, 57% of those without other crisis support from a close tie  Wives of Alcoholics 75  became depressed, and 8% of those with crisis support from another close tie became depressed, indicating that alternative crisis support was extremely valuable in moderating the relationship of stress to the onset of depression. Receiving crisis support predicted risk of depression, although perceived helpfulness and availability of that crisis support was not found to be a factor. Quality of the relationship with the husband was highly associated with self-esteem, and those women with negative self-evaluation showed no more vulnerability to depression than women without negative self-evaluation if crisis support was received. These findings seem particularly relevant to wives of alcoholics because they are likely to experience more husband events and a sense of being let down, without crisis support from their husbands. In addition, they are more likely to experience an erosion of self-esteem and less likely to have other close ties in whom to confide due to their sense of failure at getting the drinking to stop. Verification of these findings on social support and depression is needed using samples of wives of alcoholics. Generalization of these findings is somewhat limited due to homogeneity of socio-economic status. Andrews and Brown (1988) report further on this same subsample of 150 working class women who experienced severe life events. Dependency was rated for each core tie or resource relationship as to extent of reliance on that person, showing moderate or greater dependence on at least one relationship in 57% of the women. Dependency was unrelated to risk of onset.of depression, but dependency on at least one tie was associated with a greater likelihood of receiving crisis support. Attitudinal constraints on confiding and care-eliciting behaviors were also explored as inhibitors of supportive processes and were found unrelated to risk of onset of depression in the 41% of the sample that displayed them. Negative evaluation of self was found  Wives of Alcoholics 76  to be associated with social isolation and could possibly be related to failure to develop social ties, according to these researchers. They did find that early inadequate parenting was associated with lack of core crisis support and confiding that did not elicit core support (non-optimal confiding), all of which were associated with increased risk of onset of depression. It was felt that for those experiencing early inadequate parenting, personality factors had a bearing on the support-depression link. Of the women who experienced early inadequate parenting, 11 were low in confiding, low in support from a core tie, and low in depression, possibly explained by resilience in their personality structure. Women without inadequate parenting who were low in confiding and support from core ties were found to be at high risk of becoming depressed (35%) and 5 of these 7 onsets experienced severe events involving a husband. The study reported in these two articles points up the importance of husband events to married women and identifies significant influence from the social environment in moderating the onset of depression; Many of their findings seem applicable to the situations reported in samples of wives of alcoholics by recent ethnographic research!  Recent Qualitative Studies of Wives of Alcoholics Asher (1992) studied 52 wives of alcoholics whose husbands were in treatment using taped, semistructured, open-question interviews of 75-90 minutes at the beginning of family treatment and at seven and 18 months following program participation. This work was conceptualized from a symbolic interaction perspective, which proposes that objective social events are assigned subjective meaning by social participants. Actions are seen as emerging from a composite of desires, feelings, goals, expectations and demands of others, group norms, selfconceptions, memory, and conceptions of possibilities. The individual assigns meaning to every  Wives of Alcoholics 77  given situation, and ambivalence arises when there are incompatible or contradictory social realities or expectations. Asher (1992) refers to the ambivalent process of defining the alcoholic situation, as a moral career for the wife of an alcoholic, with an altered self-identity as the result. For the wives of alcoholics in this study, this definitional ambivalence tended to arise in the areas of husband, marriage, and self. For example, initially ambivalence arose in defining alcohol as a problem or a "not-problem." Drinking was often defined by the wife as a result of work or family-related stress instead of as a serious problem, with self-blame a common result. Initial ambivalence was dealt with by normalizing problems through the definitions assigned. Alcoholic husbands also worked to negate the process of designating drinking as a problem. The process of "acknowledging refers to recognizing the existence of disparate meanings of selves and situations" (Asher, 1992, p. 28). Vascillation occurred between these contradictory meanings. Social interactions gradually helped define that drinking was a problem, and new ambivalence arose about how to define the self, the marriage, and the husband. These wives of alcoholics experienced humiliation, turmoil, marital apathy, and low self-esteem as they interpreted to themselves their interactions with their alcoholic husbands. Self-esteem and selfconfidence were lost through the process of redefining the self within the context of excessive drinking by husbands. Many women chose not to talk to others about the ambivalence they faced because they felt ashamed and embarrassed.  As the realization was made that the problem was serious and not routine, the problem tended to amplify and the ambivalence did not subside. The next process was valuating, or assessing positive and negative aspects of selves and situations and dividing them into "good" and "bad." Husbands were viewed both positively and negatively in conflicting ways,  Wives of Alcoholics 78  contributing to definitional ambivalence, and distrust grew as wives became more aware of discrepancies, tipping the scale toward the negative. Wives also saw themselves changing into persons they did not like, usually as they gave vent to their anger, about which they felt justified yet uncomfortable in expressing: Women in these circumstances experience discomfort with their angry outbursts because they are incompatible with traditional definitions of ladylike behavior. Such changes are disruptive to self imagery. If self-control is a natural attitude taken toward self, then a violation of this has severe implications for self, producing feelings such as helplessness, betrayal, self-violation, and emotional vulnerability. (Asher, 1992, p. 59-60) As the self became more negatively valued, blaming self for the husband's drinking and the marital problems became easier. Ironically, even though these wives saw themselves as influential enough to cause a dramatic drinking problem, they did not see themselves as influential in other ways. These women felt there was something inherently wrong with themselves and what they were doing, and this was reinforced by their husbands' criticism, anger, avoidance, and lack of caring. Verbal conflict and physical violence were not unusual, and verbal comebacks on the part of the wives tended to give way to avoidance and withdrawal as time went by. Experiences within the alcoholic milieu tended to be taken very personally by these women, as they felt a sense of differentness they felt compelled to mask with secrecy and social isolation. Some tried to keep the drinking secret from their children and were surprised to find that their children knew all along. Embarrassment was a problem, along with anger, for essentially every woman in this study. Based on Goffman's (1967) work with social  Wives of Alcoholics 79  relationships, Asher (1992) explains that in marriage, partners come to share a face, such that if one partner acts inappropriately in front of third parties, it is acutely embarrassing to the other partner, thereby feeding social isolation, secrecy, and the sense of terrible uniqueness endured as the wife of an alcoholic. Some alcoholic husbands were expert at disavowing the feelings of their wives. As one wife said, "The more he violated my feelings, the more I started repressing them until I didn't know how I felt" (p. 94). A minority of wives fit the description of an "emotionally divided self—a self turned against itself, disembodied, characterized by self-loathing and resentment" (Denzin, 1984, p. 282). The majority of wives had a more intense awareness of selves and situations, sharpened by dramatic definitional shifts and contradictions. The lack of communication in marriage to the alcoholic rendered mutual discussions, shared meanings, and problem solving impossible. Lack of support from their husbands during crises resulted in "lived experiences of rejection" (Asher, 1992, p. 104) and fears for the future. One common guilttinged fear was related to the effect the husband's drinking would have on the children, whom these women tried to protect. Periods of sobriety and relapse created for some women cycles of building and tearing down of self-image and self-performance.  Women in this study tended to take a stance of either placating or confronting. Placating tended to be an early response in the course of the alcoholic drama, but tended to evoke either intimidation or victimization from the alcoholic husband. Home was not a refuge, but a place where wounds were inflicted. As upheaval became interpreted as disrespect, wives rebelled against their low status through confrontation of varying intensity, peaking as the problem amplified and lessening as avoidance and withdrawal set in. The stances taken by these women  Wives of Alcoholics 80  indicate a creative style of survival, according to Asher (1992). Nearly two-thirds of these women worked outside the home, and many found this to be a release or an escape, as was domestic work for some homemakers. A few turned to religion, exercise, and hobbies. Those who turned outward to engage, in social interaction had the benefit of the responses of others to provide a contrasting view of self, which contributed to their ability to survive. The two behaviors found in this study to maintain definitional ambivalence and keep these wives in their alcohol-complicated marriages were bargaining by the husbands and hoping by the wives. In order to limit ambivalence, competing definitions of self-doubt must be reduced. This may be aided by legal or health problems, groups like Al-Anon, of the husband's own admission that he has a problem. Also, the wife may gain a sense of intentionality as she makes choices that work for her, such as setting limits on useless interactions, like arguing, with the alcoholic. Entry into treatment became official and public acknowledgement of alcoholism as the problem. Paths leading to treatment for the husbands of the women in this study were: personal decision; family pressure or crisis; medical concerns or interventions; and legal-judicial mandate. Currently, designation of the problem as alcoholism defines the wife as codependent or coalcoholic and adds new self-definitions:  The designation of [caretaking actions] as codependent means that women's experience is being judged in terms of 'appropriate' male experience, which is based on standards of independence and separation : . . . This explains, in part, why the women in this study had to be taught (in the family program) that some of their actions had been codependent in nature instead of protective, caring, or a form of marital obligation, which were what they had often  Wives of Alcoholics 81  initially perceived them to be. The voice of these women was not heard as the valid account of their experience. (Asher, 1992, p. 198) This study is quite interpretive in nature and very much viewed through the lens of symbolic interaction, though its claims are backed up by numerous statements from subjects. It provides a compelling, first-hand account of the complex reality of marriage to an alcoholic man. Internal validity seems reasonably strong, although results are not presented quantitatively, and quantitative studies are needed to corroborate the findings and enhance generalizability. Exact incidence of phenomena is murky, as numerical proportions are almost never provided, but it does clearly elaborate the problem of declining self-concept and the struggle for survival faced by wives of alcoholics. Support is not forthcoming within the marriage, as the alcoholic husband becomes an adversary instead of a confidant, and embarrassment, leading to social isolation and secrecy, inhibits utilization of outside support. According to this study, lack of both success and support in the marital relationship seems especially related to disintegration of self-esteem, and based on other work reviewed here, receding marital support could be expected to fail to moderate depressive symptoms. Wiseman (1991) studied 54 wives of alcoholics in Finland and 76 wives of alcoholics in the U.S., compared to 63 U.S. wives with nonalcoholic spouses and 25 U.S. wives and 22 Finnish wives of recovering alcoholics with at least one year's sobriety. Subjects were obtained by placing ads in newspapers, and the subsamples were chosen to reflect income and education levels of the main samples. Only wives from intact marriages were included in the groups, as in the sample used for the present study. Alcoholic husbands of the wives in the two main samples were not in treatment, which provides a broader look at the population of wives of alcoholics.  Wives of Alcoholics 82  Data were collected using in-depth, semi-structured interviews, lasting approximately two hours, and a five-page structured questionnaire focusing on relevant background information. Although the drinking patterns in the U.S. and Finland differ markedly, the experiences of the wives from these two different cultures were remarkably similar. Wiseman (1991), like Asher (1992), works within a symbolic interactionist framework. In the early stages of the alcoholic behavior, wives were inclined to make mental compromises to relieve the dissonance of values coming in conflict with reality. Bizarre behavior was rationalized by them to be consistent with their definitions of their husbands as rational beings. Upsetting events often escalated in a progression such as this: husbands became undependable in ways they were formerly reliable; temporary Jekyll-and-Hyde personality and behavior changes occurred; husbands were no longer sensitive to the safety and feelings of others; temporary personality changes became long-term; husbands became unstable on the job and financial affairs became an ongoing crisis; and husbands committed infidelity. As this escalation process clarified the problem as alcoholism, wives would then attempt in-home treatment, first directly with logical persuasion, nagging, emotional pleading, and positive and negative sanctions, then indirectly through subtle manipulations such as controlling the alcohol or money supply, creating the perfect environment in an effort to eradicate the need for drink, and taking over roles to reduce husbands' stress. Wiseman (1991) highlights the gender power imbalance: "When a powerful person makes a request of a subordinate, there is little difference between the request and a command. On the other hand, if the request is by the less powerful of the two, it becomes a petition for a favor" (p. 46), and petitions by wives of  Wives of Alcoholics 83  alcoholics inevitably fell on deaf ears. When in-home attempts failed, wives worked toward professional treatment, which eventually ended in dashed hopes and a return to drinking. Wives' sense of self evolved within a context of abusive behavior, including verbal aggression, physical violence, and destruction of property, with little support from courts and police. Wiseman (1991) referred to a cross-cultural study by Mac Andrew and Edgerton (1969) which showed that drunken behavior does not exceed the particular culture's expectations for drunken behavior. This seems to show that the drinker retains control of his behavior, thus implicating drinking as more of an excuse for bad behavior than a direct cause. Selective attacks on people or property is further evidence of the drinker's control of his behavior (Wiseman, 1991). The wives experienced a great sense of loneliness and insecurity through the erosion of the marital relationship. Aspects of the self were lost because their personal identity was interwoven with the marriage. Lack of dependability in the alcoholic was a source of anger and worry; he could not be trusted to watch small children, and money and possessions were at risk if he wanted alcohol. These wives felt distress over the physical and mental deterioration of their husbands and turned fear and hostility inward in the form of physical illness and depression. Their husbands' criticism and blaming compounded downward spiraling self-esteem and operated as a control to keep wives from having enough confidence to leave. These women needed the time and resources of others, yet almost half of the wives in both cultures said they had no family emotionally close enough to ask for help. [Burnett (1984) states that the alcoholic's family may actually conspire with and support the alcoholic's projection of blame onto his wife]. Family networks tend to be situation-specific, and there would be little promise  Wives of Alcoholics 84  of closure in the task of giving support due to a chronically high level of neediness. Wiseman (1991) found that friends were lost through a quiet attrition long before the wife was ready to ask for help. Those who remained were not generally inclined to get directly involved. In both Finland and the U.S., two styles of living with an alcoholic husband emerged. About one quarter in each culture just drifted along in a state of inertia and depression, living an emotionally and socially deprived existence, while about three-quarters in each culture survived by creating an independent life of their own while remaining in the marriage. In looking for factors that could differentiate these two groups, age, education, income, Al-Anon attendance, and being employed outside the home were not clearly definitive. Women building their own lives were slightly younger with the difference being more pronounced for Finnish wives than American wives. Level of education made no difference for American wives and a small difference for Finnish wives. Income was nearly identical for both groups from both cultures. About 40% of the women who attended Al-Anon built an independent life for themselves, while about the same proportion of nonattenders do the same. Being employed outside of the home was nearly universal for the survivor group of Finnish wives compared to an employment rate of 61% for the inertia group. For American wives, 68% of the survivors were not employed outside the home, and 52% of the inertia group were employed. Building an independent life was found to be more associated with having no children in both countries and only one child in the U.S. Age of children was the most defining factor, with mothers of young children being more likely to build an independent life, possibly because young children provided an alternate involvement and a set of social activities outside of the alcoholism (Wiseman, 1991). Mothers of adolescents were more likely to be in the groups taking no action, this possibly being also related to more years in the relationship with the alcoholic, according to Wiseman (1991). A surprise finding  Wives of Alcoholics 85  was that few in either country looked to religion as a resource, except for the form of spirituality presented by Al-Anon. Race or ethnic group was not explored beyond comparison between the two cultures and bears further exploration. Motivation, inner resources, and attitude seemed to be critical factors. Those who remained in a state of inertia and depression suffered from muddled thinking, being unable to identify and evaluate alternatives. They were unable to create a plan of action, they lacked optimism, and they seemed unable to be self-centered: They were socialized to nurture and were trapped by a sick husband who needed them. They lacked the courage to proceed in the face of opposition or lack of support from others. "Encouragement from significant others is helpful to a person charting the unfamiliar territory of self-development..." (Wiseman, 1991, p. 217). Wives who remained mired in inertia were less able to see options, exhibited symptoms of depression, and were more likely to see themselves as powerless and without viable choices, a factor in self-concept. Homemaking was chosen by one-third of the total wives in both cultures as a focal point of life satisfaction, with home arts and crafts, recreation with children, reading, and housekeeping providing meaning and a sense of control. About 40% of the total sample chose time spent outside of the home as a way to independence. Renewed energy invested in employment settings enhanced a sense of accomplishment, autonomy, and optimism and provided an alternative setting where drinking did not intrude. Taking enrichment courses and involvement in clubs and organizations also provided social enrichment and enhanced self-esteem. For many women, learning skills to accomplish instrumental tasks, for which they formerly relied on a man,  Wives of Alcoholics 86  provided a sense of independence and self-esteem. As these women were able to formulate plans and execute them, they felt more capable and optimistic. Life for these wives of alcoholics tended to spiral either upward or downward. "Activities and attitudes are cumulative and circular in their effect on the motivation and behavior of these wives, whether the thrust is ultimately independence or dependence" (Wiseman, 1991, p. 226). Common threads in the move toward independence were: they chose areas of their lives they felt they could change; they modified their behavior and thinking to be more like a man than a traditionally-minded woman; they viewed employment as a source of autonomy; and they became aware of the benefits of controlling their own lives. The more wives of alcoholics in this study put their own lives on hold, the more they sank into a state of depression. Prolonged social isolation adversely affected self-concept, impaired initiative, and increased their sense of helplessness: The description of the state of mind of prisoners whose existence is filled with debility, dependency, and dread closely parallels the quotes from those wives of alcoholics, who, feeling like prisoners in their own homes and in their marriages, have lapsed into a depression that reinforces their inertia. (Wiseman, 1991, p. 231) Essentially, the women who became lost in inertia felt little ability to bring about a positive outcome for themselves and became depressed, a logical result of social isolation and erosion of self, which are antithetical to social support. In interviewing wives whose husbands had attained sobriety, none claimed that everything returned to normal upon sobriety. Wives remembered the suffering caused by their husbands' drinking, and sobriety put them in the position of needing to restructure lives  Wives of Alcoholics 87  constructed to function without their husbands. Cessation of drinking made it more safe to express feelings they were previously too afraid to express, for fear of upsetting the alcoholic. Sobriety necessitated marital adjustment as these feelings were processed, and that required time. Although subjects in Wiseman's (1991) study were not randomly selected, they were drawn from a more diverse group than wives whose husbands come for treatment, which worked to reduce the selection bias factor. The marital relationship was assessed in comparison to wives of nonalcoholics, and the effects of sobriety for wives were evaluated by studying wives whose husbands were at least one year into abstinence. Reliability of the data is indicated in that many different women tell of similar experiences. Comparison across cultures creates depth and strengthens external validity. Data was rechecked with respondents and was also corroborated by 33 U.S. and 15 Finnish male alcoholics as a check on internal validity. Overall, this work was well done, informative, and highly consistent with other work. Banister and Peavy (1994) interviewed five Caucasian women who were recruited from a drug and alcohol center and currently married to alcoholics with duration of marriage ranging from 19 to 32 years and age ranging from 47 to 60 years. The Developmental Research Sequence (DRS) Method was used, which consists of 12 tasks through which the study of language is used to discover dimensions of meaning in cultural experience. In this situation, the alcoholic's wife employs cultural meanings to make sense of her husband's alcoholism. Through this process, themes of common experience are developed and rechecked for accuracy with each informant.  These women lived in a state of constantly being on guard and tense because their husbands' behavior was unpredictable and inconsistent. They experienced a pervasive lack of  Wives of Alcoholics 88  trust and a feeling of powerlessness in their marital relationships and learned over time to suppress their emotions. One of them said she never shared her painful feelings with anyone, almost concealing them from herself as well. These women felt trapped by a gradual erosion of self that occurred through ongoing interactions with their alcoholic husbands in which perceptions of self became progressively more distorted, and they became more isolated and lonely as they accepted blame for the situation in which they lived. As their husbands also became more isolated from society by their drinking and became more dependent on their wives, they exerted more control over them in an effort to keep them from leaving or becoming more independent. All of these women believed they were not strong enough to leave. Having been socialized to be self-sacrificing, they felt guilty if they considered their own needs, and they . became out of touch with what they needed. These women feared being on their own at mid-life because of limited social options and the possibility of financial deprivation. They felt a strong sense of disillusionment and despair as their expectations about marriage were vastly discrepant from their reality. Interaction within the alcohol-embedded marriage brought about a weakening of self, to which they were made vulnerable by internalized cultural expectations to be dependent, self-sacrificing, and self-blaming. Though they felt trapped, they were immobilized by their own fear of change. This study was limited in size of sample and range of age, which limit generalizability, yet the findings were very similar to those of Asher (1992) and Wiseman (1991), and the cumulative evidence is very persuasive. Further study employing quantitative methods is indicated.  Wives of Alcoholics 89  Wives of Alcoholics and Employment Even though Wiseman (1991) did not find work outside of the home to clearly  • .  differentiate wives of alcoholics who lapsed into a state of inertia and depression from those who successfully coped by creating an independent life of their own, other researchers have found employment to be helpful to wives of alcoholics as they strive to cope with their situation (Casey, et al., 1993; Googins & Casey, 1987). The emergence of the two-breadwinner family has changed the structure of families in general, and the effect of the workplace in the lives of employed wives of alcoholics'was previously unstudied. The added responsibility of outside work for wives of alcoholics could conceivably lead to anxiety and depression or independence and fulfillment. Casey, et al. (1993) studied 60 wives of alcoholics who were employed at least 20 hours per week and whose husbands had been admitted to residential treatment. These wives ranged in age from 23 to 62 years of age with 95% having completed high school and 29% having completed college. Most (95%) were white, and a majority (63%) were Catholic, which limits generalizibility of results somewhat. None of these wives reported being dependent on drugs or alcohol themselves, but half reported being children of alcoholics, which is a high figure compared to the general American population, wherein recent studies report an incidence of 11.5 to 15.7% being children of alcoholics (Myers, et al., 1984). Data were collected one month following their alcoholic husbands' discharge from treatment using face-to-face interviews in conjunction with three standardized measures. The Health and Daily Living Indices was used to measure health and general social functioning on the basis of physical symptoms, medical conditions, depressed mood, smoking symptoms, and  Wives of Alcoholics 90  indicators of coping. The Family Environment Scale was used to assess the quality of family functioning, and the Job Satisfaction Scale was used to evaluate the worker's evaluation of specific aspects of the job as well as overall affective reaction to the job. Results were reported in terms of work performance, job satisfaction, and work-based social supports. The majority of these working wives of alcoholics reported little, if any, impact of their husband's drinking on five aspects of job performance, including relations with her boss (92%) and coworkers (76%), job performance (85%), general functioning (65%), and attitude at work (59%). A large minority (40%) reported beginning or quitting a job due to husband's drinking, and 37% missed work due to husband's drinking, with almost twice as many missed days annually as a normal population of community wives who are employed. As high as 35% felt significantly affected by husband's drinking in at least one aspect of job performance, attitude at work being the most commonly cited (24%). Of these working wives of alcoholics, 57% rated themselves as "very satisfied" with their jobs and 37% rated themselves as "somewhat satisfied," which is a very high level of job satisfaction compared to the normal population (Finney, Moos & Mewborn, 1980). For 73% of these wives, at least one person at work knew about their husband's drinking. The remaining 27%, who reported telling no one at work about their husband's drinking, were significantly less likely to report that the job helped them cope with their husband's drinking. In other words, wives who maintained complete privacy at work about their husband's drinking, thereby isolating themselves from potential sources of support, did indeed receive less support at work. In looking at how coworkers found out about their husbands' drinking, although 55% voluntarily told someone on their own, 48% demonstrated behavior indicating that home life was  Wives of Alcoholics 91  affecting their work, even though 59% to 92% reported their husbands' drinking to have little effect on the five aspects of job performance. Apparently, they felt somewhat less affected than they actually were, perhaps due to denial used as a coping mechanism, thereby distorting perception, according to these researchers. This explanation is supported by the fact that these wives scored significantly higher on avoidance coping in this study than is found in community populations (Moos & Moos, 1986). Although work has been viewed as a stressor, it can have supportive aspects as well, according to this study. In fact, "working wives of alcoholics reported levels of job satisfaction significantly higher (94% compared to 83%, p < .0001) than a community population [Finney, et al., 1980] composed of 488 working wives" (Casey, Griffin & Googins, 1993, p. 127), which seems to indicate that employment can be a very rewarding part of life for many wives of alcoholics. Nevertheless, this sample of employed wives of alcoholics resembled a depressed population in terms of physical symptoms, medical conditions, depressed mood, and smoking, thus reflecting the challenges they face. Overall, the majority of these wives saw their work more as a haven than as a source of stress. In addition to the value of work in providing income and bolstering self-esteem, establishing relationships in the workplace adds a dimension to one's social support system. Social support can buffer the stress of an alcoholic husband and, further, relationships can change the meaning of the situation and help control stress symptoms in many ways . . . . Whereas the work role can be a source of stress, it can also serve to reduce stress from other roles. (Casey, Griffin & Googins, 1993, p. 129)  Wives of Alcoholics 92  Thus it seems that the workplace is likely to add to the social support system of wives of alcoholics, at least for wives of alcoholics who are white, middle class women. This study adds important information to the body of knowledge about wives of alcoholics, although generalizability is limited in terms of ethnicity and socio-economic status. Googins & Casey (1987) did a pilot study, of ten working wives of alcoholics in which they evaluated the usefulness of support systems in three settings—at work, within the family, :  and outside of the family. These women ranged in age from 30 to 66 years. Half of therri : worked in clerical positions and half in management or professional positions. Of the ten women, 30% identified husband's drinking as their primary reason for working, and the other 70% chose to work for other reasons. Three women felt strongly that work increased the stress in their lives and that their husbands' drinking negatively impacted their job performance and ability to function at work. Yet more than two-thirds strongly agreed that work was a haven from a disordered home environment and felt their job performance was not impaired by their husbands' drinking. All of the women sought support from at least one person at work, and the response to their requests was overwhelmingly supportive, with 78% of the 18 identified support persons described as supportive and the other 22% described as supportive at least some of the time. No one was viewed as unsupportive. These wives of alcoholics were somewhat less likely to look for support within their families (including extended family), with 70% seeking support along familial lines. Of the 26 support persons identified, 38% were rated supportive, 15% unsupportive, and 46% giving a  Wives of Alcoholics 93  combination of supportive and unsupportive responses. It was suggested that family denial of the problem contributed to perceived lack of supportive behavior. Outside supportive persons or resources such as Al-Anon, female friends, and therapists were sought by 70% of these wives, while 30% sought no outside help. Of the 34 people or programs involved with these women, 76% were seen as supportive, 6% as unsupportive, and 18% as giving a mixed response. Therapists and Al-Anon were viewed as being the most supportive, while friends of either gender were found to be less able to provide useful help. Overall, work was viewed as being 78% supportive, outside the family following closely at 76% supportive, and family significantly less at 38% supportive. Googins and Casey (1987) concluded that working wives of alcoholics generally have a definite advantage over nonemployed wives of alcoholics because there is a greater likelihood that their environment will provide needed support. They suggest that the family environment may be the least supportive due to family history and the emotionality of the alcohol issue, whereas "the work environment does indeed offer a locus of meaningful relationships distant enough from the family baggage but close enough to provide understanding and support" (Googins & Casey, 1987, p. 63). They also suggest that the refuge provided by work may enable wives of alcoholics to postpone getting the help that could actually intervene and help to solve the problem, a pitfall wives need to be made aware of. This study is limited by the number of subjects, yet it does point a direction for further study in that it highlights a difference in supportiveness between family and work environments.  Summary Many studies have indicated a relationship between social support and depression, several of which have been cited here. Burchfield, et al. (1982) found that unfulfilled needs for  Wives of Alcoholics 94  affiliation in a context of high stress were associated with more symptoms of illness in a sample of women. In the etiology of depression, salience of stressors is different for men and women, and social support appears to be more vital to women in moderating stress and depression (Hibbard, 1985) because of their relational orientation. Although living as married is protective for men, it appears to be a mixed bag for women, with the quality of the marriage being strongly related to their sense of well-being. Perceptions of marital support by depressed women have been found to accurately reflect actual support and to be associated with recovery from depression (Goering, et al, 1992), and lack of communication and support from husband have been associated with depressive symptoms in other populations of women (Paykel, et al, 1980; Brown, et al., 1986),. particularly in the face of negative life events. Brown, et al. (1986) found husband events to be the worst of severe events related to risk of depression in a population of women, a dynamic magnified by a sense of being let down during a crisis and reduced by alternative support from a close other. It can be seen that social factors are strongly influential in moderating risk of depression, although intrapsychic factors such as resilience can affect how the social is processed.  Common threads emerge in all three qualitative studies of wives of alcoholics reviewed here. These wives interact within a progressively deteriorating family environment, without a supportive marital relationship and within a cultural context that holds women responsible for family well-being. In the process, their sense of self as being capable of influencing outcome (Seligman, 1974) is severely diminished, and they are at risk for depression. Asher (1992) outlined the process by which the definition of alcohol as a problem caused wives of alcoholics to redefine themselves more negatively. Shame and self-blame caused them  Wives of Alcoholics 95  not to seek support by talking to others and drove them into social isolation. Embarrassment and anger were universal feelings for these women, yet anger often felt too uncomfortable to express. Lack of support from their husbands became "lived experiences of rejection" (Asher, 1992, p. 104), and lack of success in the marital relationship was often interpreted to themselves as personal failure. Turning outside the family provided a contrasting view of self for many women. Wiseman's (1991) findings are very similar to Asher's (1992), but with the addition of a comparison group made up of wives of nonalcoholic husbands, which is used in studying the marital relationship. Wives experienced great loss of personal identity and attrition of friends as their marriages were deteriorated by alcoholism. Wiseman (1991) noted two distinct groups among wives of alcoholics. Those who were depressed seemed unable to be self-centered, were socialized to be nurturers, and lacked the courage to proceed without the support of others. Placing their own lives on hold mired them in inertia, which was associated with depression, and prolonged social isolation and interaction with the alcoholic eroded self-esteem. The more independent wives developed traditionally male traits, rounding themselves out into more androgynous, versatile beings—they became agents for themselves and chose a positive course of action. According to Wiseman (1991), the women socialized in a traditional way "believe that self-interest is a proper attitude for men, but not for women. Rather than making the most of themselves, their job is to make the most of their husband" (p. 226). Self-worth for these women was derived externally. Employment has been found to provide wives of alcoholics with a welcome rest from a chaotic home environment, an alternative context for self-definition, and a source of social  Wives of Alcoholics 96  support (Casey, Griffin & Googins, 1993; Googins & Casey, 1987). They experienced significantly higher job satisfaction than community controls, even though they experienced more depressed mood and physical symptoms. The workplace and other outside people and resources were found to be more effective resources of support than family and extended family, a finding that highlights the potential value of therapeutic intervention as a supportive context. The effect of husbands' alcoholism on nonalcoholic wives is very significant, and for the benefit of these women struggling to survive, it is time for those who would be helpers to discard the biases of the past: "To overcome their own myopia and the alcoholic's denial, both relative and clinician must learn to conceive of alcoholism as a disease that causes depression, marital breakup, and unemployment, not as a symptom that results from such distressing events" (Van Den Bergh, 1991, p. 53). Society is more sympathetic to children who grow up in alcoholic homes than to wives of alcoholics, yet children depend on their mothers' well-being for their own. When wives of alcoholics are able to maintain a positive identity and get their personal needs met, they are able to reduce the consequences for their children of growing up in an alcoholic home (Kokin & Walker, 1989). The mother, as a primary support resource for her children, may be able to moderate how they experience paternal alcoholism. For the sake of their children and themselves, wives of alcoholics need a more supportive cultural context in terms of attitudes and meaning-making, beginning in the therapist's office and extending to the families and friends darkened by the shadow of alcohol abuse. Does perceived social support moderate depression in the face of chronic stress for wives of alcoholics as it does in other populations? How does employment outside the home affect  Wives of Alcoholics 97  perceived social support as it relates to onset of depression as wives attempt to deal with their husbands' alcoholism?. What other variables have an effect on the stress-depression relationship in wives of alcoholics?  Hypotheses Hypothesis I Perceived social support from family and perceived social support from friends are significantly inversely related to depression in wives of alcoholics. Hypothesis II Selected demographic variables explain some of the variance in the relationship between perceived social support from family or friends and depression in wives of alcoholics. Variables to be explored as potential contributors to the variance in the relationship between the main variables listed in Hypothesis I are employment status, frequency of attendance at religious services, number of children, number of persons residing in the home, years lived with husband or partner, race, age of wife, education level of wife, and annual family income.  Wives of Alcoholics 98  CHAPTER IV Method This study utilized data collected through The Alcohol Recovery Project (TARP) (Friesen, Conry, Grigg, & Weir, 1995). Therefore, this research was designed within the parameters of data collected through the instruments chosen by the larger research and recovery project, which had the primary purpose of studying the outcome of three different therapeutic approaches, which were Supported Feedback Therapy-Individual Format, Experiential Systemic Therapy-Individual Format, and Experiential Systemic Therapy-Couples Format. The present study utilized self-report data collected from wives of alcoholics during the screening and pretreatment phases of the larger project (TARP), thereby depicting the experience of wives of alcoholics prior to therapeutic intervention, yet at a point when their husbands have consented to begin treatment for alcoholism. At this point, they are attempting to cope on their own without professional help, yet professional help for the problem of alcoholism is in sight.  Sample Subjects were obtained by referral from drug and alcohol treatment agencies and inpatient treatment programs after being processed through their normal intake procedures. Due to coverage of the project by television, radio, and newspapers, subjects also referred themselves. After being screened for participation using the inclusion and exclusion criteria listed below, a sample of 150 families was obtained from these sources. Of the 150 wives of alcoholics from whom data was collected at the conclusion of the screening phase, 133 responded at the pretreatment phase of data collection, and of these, complete data was obtained from 116, which comprise the sample for the present study. The attrition group compared to the study sample had  Wives of Alcoholics 99  a somewhat higher incidence of common law unions (24.2% vs. 16.5% ), was more likely to be employed full time (58.1% vs. 48.3%) and less likely to be employed part time (19.4% vs. 29.3%), averaged slightly fewer years together (10.6 vs. 12.4) and were slightly younger (34.8 vs. 38.5), but were quite similar in terms of incidence of unemployment as well as mean income, education level, number in the house, and number of children. Overall, the two groups seemed comparable, and attrition did not effectively bias the sample that remained. These women ranged in age from 22 to 66 years, with a mean age of 38.47 and a mean annual family income of approximately $47,900 (Canadian dollars), while ranging from under $10,000 to over $100,000. In years of formal education, they ranged from 8 to 18 years, with a mean of 12.9 years of schooling. The majority of them (83.5%) were legally married, they had a mean of 12.4 years together, and they ranged from one to 43 years together with their partners. They had from one to six children, with a mean of 2.4, and the number of people currently living in the house ranged from one to seven with a mean of 3.8. In terms of employment status, 17.3% were unemployed, 29.3% were employed part time, 48.3% were employed full time, and 5.2% were in school. Screening Inclusion Criteria. Families involved in this study met the following inclusion criteria at the point of screening: the father was alcohol dependent and had consumed alcohol within the past three months; the mother had not had a problem with alcohol dependency at any time during the previous five years; the couple had been sharing residence for at least a year were and were desirous of preserving the relationship; they were experiencing marital distress and were ready and willing to be involved in couples therapy if assigned to that treatment group; included in  Wives of Alcoholics 100  each family was at least one child living in the home, or out of the home but in frequent contact; blended and remarried families were included, and the children could be of either parent. Exclusion Criteria. The following exclusion criteria were employed at the point of screening with those families that met the criteria listed above: the father's alcohol dependency was not severe enough for him to exceed a score of five on the Michigan Alcoholism Screening Test (MAST, Selzer, 1971); use of alcohol by the mother was severe enough for her to exceed a score of four on the M A S T ; negligible marital distress was indicated by both husband and wife having a score of over 99 on the Dyadic Adjustment Scale (Spanier, 1976, 1990); severe psychiatric disturbance was indicated by either mother or father scoring exceptionally high on the psychiatric or depression subscales of the Symptom Checklist-90 Revised (Derogatis, 1983). The Edmonds Marriage Conventionality Scale was used to screen out those who would respond with a socially desirable bias rather than accurately reflect functioning. Adults were aged 21 to 65 years old, and none of the participants used English as a second language, which could be related to not fully understanding questions in the self-report test instruments used. As compensation for time required to fill out questionnaires, each family received $200.00. The sample for this study represents the population of wives of alcoholics who are not personally alcohol dependent and whose husbands have expressed willingness to begin treatment. These wives have one or more children at home or in close contact and are maritally distressed but not psychiatrically disturbed. Including those with endogenous depression or other psychiatric disturbance would have.introduced a confounding influence, potentially unaffected by social and family circumstances. The intention of this inquiry was to examine the relationship between perceived social support and depression in wives of alcoholics within the high-stress  Wives of Alcoholics 101  context of living with an alcoholic husband. The focus was on situational rather than endogenous depression.  Main Variables The overall flow theorized in the current study is that involvement in contexts outside the home provide information about self that competes with and potentially contradicts messages of projected blame from the alcoholic, thereby providing potential contexts for social support that contribute to greater perceived social support, which then operates to increase self-esteem and decrease depression. The main independent variables were perceived social support from family and perceived social support from friends, with employment status and other demographic variables derived from Wiseman's (1991) work, included as additional predictor variables. The dependent variable was depression, which was evaluated by two separate instruments, depression being essentially the core experience of the wives of alcoholics described by Wiseman as being mired in inertia. Self-esteem as measured by the Structural Analysis of Social Behavior-Introject Scale- , Affiliation Subscale (SASB-Intro-Affil; Benjamin, 1974, 1984) was considered as an additional dependent variable but was dropped due to its very high negative correlation, in this sample, of -.711 with depression as measured by the Beck Depression Inventory (BDI), which showed depression (as measured by the BDI) and low self-esteem (as measured by the SASB-Intro-Affil) to be constructs with considerable overlap. This finding is sufficiently significant to merit further analysis of these variables in a separate report utilizing the data from this sample. Elsewhere self-esteem has generally been found to be inversely related to depression, with a correlation of around -.60 (Shaver & Brennan, 1991)-  The two depression measures utilized here had a  Wives of Alcoholics 102  correlation in this sample of .596, indicating that they measured related but slightly different aspects of depression. Social Support Social support is conceptualized generally within the category identified by Gottlieb (1985) relating to the quality and content of close relationships. Social support was measured subjectively as perceived social support, in that the subjective view of the person involved is all important in determining the real impact of the support. According to Murray (1938), how a person sees the world is more relevant to predicting behavior than objective reality, which may be difficult to assess untainted by someone's subjective evaluation. Perceived social support is defined as "the extent to which an individual believes that his/her needs for support, information, and feedback are fulfilled" (Procidano & Heller, 1983, p. 2). According to symbolic interaction theory, situations defined as real are real in their consequences (Wiseman, 1991). In other words, if a person feels supported, the intrapsychic process of meaning-making is altered in positive ways, and new possibilities emerge. On the other hand, if supportive social networks are available but unperceived, they can not have an effect. Perceived social support is an indicator of the impact social networks have on the individual (Procidano & Heller, 1983). According to Gottlieb (1985), assessment of environmental coping resources is part of the cognitive process when facing stress.  The construct of perceived social support was measured separately for family and friends using two instruments: Perceived Social Support from Family (PSS-Fa) and Perceived Social Support from Friends (PSS-Fr). Subjects were asked to evaluate their support resources during the previous four weeks. Perceived social support from within the family reflects the degree to which family is viewed by its members to be a supportive resource in terms of support,  Wives of Alcoholics 103  information, and feedback. Perceived social support from friends refers to the degree to which friendship networks are viewed as a supportive context in terms of support, information, and feedback. These instruments are an indication of support received, particularly emotional support, although enactment and availability of support are included. Family and friends contexts for support differ by effect based on the situation and the population being studied, and therefore they should be differentiated (Procidano & Heller, 1983). Relationships with friends are often shorter in duration than are family relationships, due to life changes such as school and work-related moves. Also, social competence plays a greater role in the maintenance of friend relationships than family relationships, which come into being more by assignment than choice and are changed by the less frequent events of birth and death (Procidano & Heller, 1983). Depression The dependent variable is depression, which was measured using the Beck Depression Inventory (BDI) and the depression subscale of the Symptom Checklist 90-Revised (SCL-90-R). Depression has been conceptualized by Beck (1974) as a disturbance in cognition involving a negative view of self, the world, and the future. Errors in thinking, such as overgeneralization, magnification and minimization, personalization, selective abstraction, and arbitrary inference, can be activated under stress (McNamara, 1992). Individuals are likely to become depressed when they experience failure or when reality fails to meet their expectations (Shaver & Brennan, 1991), a situation not unknown to wives of alcoholics. According to Shaver, Schwartz, Kirson, and O'Connor (1987), all emotions fit within five basic categories, which are love, happiness, anger, sadness, and fear. Depression, which is  Wives of Alcoholics 104  also closely related to loneliness, alienation, external locus of control or helplessness (Benassi, et al., 1988), and low self-esteem, is a form of sadness (Shaver & Brennan, 1991). In fact, in recent years, depression has been the most prevalent mental health problem in the United States (Dean, 1985). Loneliness is a closely related concept that is difficult to separate from depression empirically, though it is not as broad in scope because it is limited to the interpersonal realm. Depression arises from the interpersonal as well as the circumstantial or situational, and thus loneliness can be seen as a subset of depression (Shaver & Brennan, 1991). Wives of alcoholics typically experience isolation as well as challenging circumstances (Asher, 1992; Wiseman, 1991) and thus are at risk for depression.  Demographic Predictor Variables Because all wives of alcoholics are not alike, the influence of demographic variables was explored, including employment status, frequency of formal religious participation, number of children, number of people living in the house, marital status, number of years lived with partner, race, age of wife, education level of wife, and annual family income. This information was obtained from the T A R P database, where it had been entered from the T A R P Family Demographics Form. Age of children was one of the variables studied by Wiseman (1991), but it was unavailable in the TARP database, thus number of people living in the house was utilized because it would include children still living at home and at least partially reflect age of children. In the study of wives of alcoholics by Wiseman (1991), there were two styles of coping that were evident: One group was lost in the inertia of depression, while the other group carved out independent lives despite remaining married to alcohol-dependent men. Satisfaction with work and work-related accomplishment were observed in the more independent group in  Wives of Alcoholics 105  Wiseman's (1991) study. Employment provided release or escape for many of the women Asher (1992) studied as well, and employment status was the additional predictor variable of greatest interest in the present study, second only to perceived social support. Number of religious services attended during the previous year was part of the data collected by TARP, and was also of interest because religious involvement would provide an alternative context for wives of alcoholics that could be experienced as supportive. Both Asher (1992) and Wiseman (1991) found regular religious involvement to be rarely pursued by wives of alcoholics, a finding which was verified by this inquiry. In the TARP sample, the distribution was badly skewed, with a range of zero to 60 religious services attended during the previous year and the majority reporting zero services attended. This fact rendered the data on religious participation unusable for regression analysis, which assumes normality of data, and therefore it was dropped. A similar situation existed in the TARP sample with respect to that 96% of the subjects were white, even though non-whites were not excluded from the study. Employment status was categorized as nonemployed, part time, full time, and in school. Each category was considered separately and coded as a dummy variable with "1" indicating the presence of that particular employment status and "0" indicating the absence of it. Frequency of attendance at religious services was reported as the number of times a church service was attended during the previous year, though the number of regular church attenders was too small to provide a basis of comparison. Number of children and years lived with partner were recorded numerically as continuous variables. Number in the house was recorded as the actual number of persons residing in the house, including the subject. Race was recorded as Native Indian, Oriental, Black, Hispanic, or White, however, almost all of the subjects were white, making any  Wives of Alcoholics 106  real study of the influence of race impossible, and therefore race was dropped as a potential variable in the regression analyses. Age of wife and education level were recorded in years as continuous variables. Income was grouped in increments of $10,000 up to $69,999, with $70,000 through $99,999 and $100,000 and over as the highest two categories out of nine.  Instrumentation Instruments for assessing the experience of members of families in which alcohol is a factor were selected by the larger research project (TARP) to be in harmony with Brofenbrenner's (1977, 1979) Ecological Assessment Model, which conceptualized a dynamic interaction between individuals and their environment. Human development was seen as evolving within a series of hierarchical contexts in which influence and change were bidirectional, with individuals having an impact on their environments and environments affecting individuals. This "progressive accommodation" (Brofenbrenner, 1979) between the evolving person and his or her environment was to be studied by contrasting the structural components of environmental systems, controlling for other sources of influence by using matched samples or random assignment. Individual assessment, as recommended by Conger (1981), was a complement to ecological assessment in the evaluation of the dynamic interaction between individuals and their environments. The present study utilizes instruments on the levels of the intrapersonal system, influenced indirectly by social network experience. Instruments were chosen based on their ability to measure specific qualities or characteristics, while at the same time meeting adequate standards of psychometric utility.  Wives of Alcoholics 107  Perceived Social Support-Family and Perceived Social Support-Friends Participants responded to two self-report questionnaires which were designed to quantify the degree of perceived social support from the two separate sources of family (PSS-Fa) and friends (PSS-Fr). Both measures have 20 items arrived upon through factor analysis, and both were implemented by TARP using a seven-point Likert-type answer format ranging from 1 (Strongly Agree) to 7 (Strongly Disagree). Seven points were scored for each item, with higher scores indicating greater perceived social support, giving a scoring range of 0-140 on each of the two scales, which were scored separately because they measure psychometrically distinct aspects of perceived support. According to the authors of these instruments, Procidano and Heller (1983), both the PSS-Fa and the PSS-Fr are internally consistent and measure constructs that can be differentiated empirically, with a correlation of .24, demonstrating that they are related but separate constructs. The PSS-Fa and the PSS-Fr are single-factor, homogeneous measures, and they were found to predict symptoms more accurately than life events or social network characteristics. Procidano and Heller (1983) found Cronbach's alpha to be .90 for Pss-Fa and .88 for PSS-Fr, and later work by Kurdek and Sinclair (1985-1986) found a Cronbach's alpha of .87 for PSS-Fa. The PSS-Fr. was shown to correlate positively with other measures of social assets (CPI, DAQ) and negatively with psychopathology as measured by the MMPI short form (Procidano & Heller, 1983). The PSS-Fa was shown to correlate positively with social competence (DAQ) and negatively with depression on the MMPI short form.  The authors found perceived social support as measured by these instruments to be relatively stable across situations in their validation study, with PSS-Fa unaffected by exposure to  Wives of Alcoholics 108  prior positive or negative self-statements and PSS-Fr subject only to negative induction. This may be explained in that friend networks are more subject to change while family relationships are of longer duration and less vulnerable to temporary changes in attitude. Beck Depression Inventory The Beck Depression Inventory (BDI) was one of the two instruments used to quantify the construct of depression by measuring its existence and severity. This scale, written at a sixthgrade reading level (Beck, Steer & Garbin, 1988), consists of 21 groups of items, originally drawn from clinical experience, which correspond to the cognitive, affective, motivational, and vegetative aspects of depression in terms of negative attitudes, performance difficulties, and somatic complaints. Each item provides a group of four statements that represent a range of responses relevant to a particular aspect of depression. Subjects in T A R P were instructed to base their responses on how they had been feeling during the previous week, including the day of the test administration. Score range is from 0-63, with 0-9 representing the normal range, 10-18 indicating mild to moderate depression, 19-29 showing moderate to severe depression, and 30-63 demonstrating severe depression (Beck & Beamesderfer, 1974). The BDI is used for detecting and assessing depression in both psychiatric (Piotrowski, Sherry & Keller, 1985) and normal populations (Steer, Beck & Garrison, 1986).  According to Fremouw, Perczel, and Ellis (1990), the BDI has high internal consistency, test-retest reliability demonstrated across populations, and correlates highly with other measures of depression. Nine studies of internal consistency of the BDI for psychiatric populations found coefficient alphas ranging from .76 through .95 with a mean of .86, and 15 nonpsychiatric samples had a mean coefficient alpha of .81 with a range of .73 to .92 (Beck, Steer & Garbin,  Wives of Alcoholics.109  1988), which indicates strong internal consistency across a wide range of subjects. Stability over time using test-retest was shown in five studies of nonpsychiatric populations ( n = 204 to 498) with Pearson product-moment correlation coefficients of .60 to .90 at time intervals varying from hours to weeks and up to four months (Beck, Steer & Garbin, 1988). Content validity is shown in that the BDI reflects the majority of criteria listed in the DSM-III, with three criteria purposely omitted because they fail to differentiate between depressed and nondepressed populations or are inappropriate for self-evaluation (Beck, Steer & Garbin, 1988). Concurrent validity is shown in at least 35 studies (Beck, Steer & Garbin, 1988) correlating the BDI with a wide variety of other measures of depression for psychiatric and nonpsychiatric populations respectively (mean correlation coefficients are given), including clinical ratings (.72 and .60), MMPI-Depression Subscale (:76 and .60), and the SCL-90-R (.55 and .60), which was also used by TARP and the current study. Correlations are generally somewhat stronger for psychiatric populations, but correlations are also significant in nonpsychiatric populations. The strength of the correlation of the BDI with the depression subscale of the Symptom Checklist 90 Revised in the nonpsychiatric subjects of the present study was essentially the same at .596 (p < .0001) as those just listed. Discriminant validity is shown by the BDI in differentiating normals from psychiatric, depressed, and alcoholic patients (Beck, Steer & Garbin, 1988). Construct validity is demonstrated as the BDI has detected relationships between depression and other behaviors, such as suicide, alcoholism, school adjustment, and anxiety (Beck, Steer & Garbin, 1988).  Wives of Alcoholics 110  Symptom Checklist 90 Revised-Depression Subscale The Symptom Checklist 90 Revised (SCL-90-R) is a 90-item self-report inventory of psychological symptoms with 9 primary symptom dimensions, including the depression subscale, which was used in the current study to provide an additional measure of depression. A total of 13 items comprise the depression subscale, tapping symptoms of clinical depression, such as suicidal thoughts, crying, and loss of vital energy and sexual interest, as well as a variety of the feelings associated with a depressed state, such as feeling worthless, trapped, lonely, and without hope. Cognitive elements are also included, such as the tendency to self-blame and worry, as well as losing interest in life and viewing it as an effort. The SCL-90-R utilizes a 5-point Likert scale ranging from 0 (Not at All) to 4 (Extremely). It has been used for assessment purposes across a wide variety of conditions involving mental and physical health. It is a "measure of current, point-in-time psychological symptom status" and "is not a measure of personality except indirectly, in that certain personality 'types' and 'disorders' may manifest a characteristic profile on the primary symptom dimensions" (Derogatis, 1983, p. 4).  Concurrent validity has been established through correlation with related subscales of the MMPI (Derogatis, Rickels & Rock, 1976) and the Hamilton Depression Scale (Weissman, Sholmskas, Pottenger, Prusoff & Locke, 1977), among others. Internal consistency has been demonstrated with alpha coefficients ranging from .77 to .90, and stability is shown by test-retest correlations ranging from .78 to .90. Derogatis and Cleary (1977) demonstrated convincing construct validity through factor analysis of the 9 dimensions based on data from 1,002 psychiatric outpatients, although results from others have been mixed (Cyr, Doxey & Vigna, 1988; Brophy, Norell & Kiluk, 1988). In the sample of wives of alcoholics studied here,  Wives of Alcoholics 111  correlations between the depression subscale of the SCL-90-R and the other subscales of the SCL-90-R ranged from .49 to .79.  Data Analysis Hypothesis I, which states that perceived social support from family and perceived social support from friends will be significantly inversely related to depression in wives of alcoholics, was tested first by examining Pearson-r correlations. Next, multiple regression analyses were done using the predictor variables that showed some relationship (P < .15) to the dependent variable of depression, as measured by at least one of the instruments. A correlation matrix was calculated to determine correlations, and their level of significance, between all of the variables individually. This provided a check that the two depression measures were not redundant and that the independent variables were not highly correlated and thus collinear. It also provided a sense of which variables were of greatest interest for further exploration. The demographic predictor variables having a significant relationship (P < .15) with at least one of the measures of depression were employment status (unemployed, employed part time), number of children, number of persons residing in the house, age of wife, education level of wife, and annual family income. Years together and being in school or employed full time were not significantly correlated with either measure of depression. A stepwise multiple regression analysis (Chatterjee & Price, 1991; Cohen & Cohen, 1983; Tabachnick & Fidell, 1989; Stevens, 1986), was employed to examine the independent variables of perceived social support from family, perceived social support from friends, and the demographic predictor variables, in relation to each of the two measures of depression in this sample of wives of alcoholics ( n = 116). According to Tabachnick and Fidell (1989), the  Wives of Alcoholics 112  optimal set of IV's is the smallest set uncorrected with each other that covers the dependent variable, and stepwise regression is a good method to achieve this end. Stepwise regression selects the independent variable having the largest correlation with the dependent variable as the first entry into the regression equation, followed by the independent variable remaining that will make the largest contribution to R . As additional independent 2  variables are added in decreasing order of their contribution to R , the previously added 2  independent variables are rechecked to verify their continued importance in the regression equation and dropped if they do not continue to explain a significant portion of the variance (Chatterjee & Price, 1991; Tabachnick & Fidell; 1989). This procedure is complete when there are no more independent variables that will make a statistically significant contribution to explaining the variance. In this study, variables not significant at the .15 level were not added into the regression equation. However, all but one of the variables selected in the regression analyses were significant at the .05 level. Two additional regression methods were employed as a check of results from the stepwise method, which is an improved version of the forward selection method (Chatterjee & Price, 1991; Norusis, 1988; Stevens, 1986)). The backward elimination process (Chatterjee & Price, 1991) considers all possible variables and then drops the least relevant variable at each step until only those that contribute significantly to the variance remain. This method yielded the same results as the stepwise method, thus strengthening the validity of the regression models obtained (Norusis, 1988). A setwise regression, which looks for the highest value of R based on 2  all possible combinations of independent variables, again verified the results of the other two methods. It is not uncommon to obtain different results when different methods are employed, and it is very encouraging when results from these different methods concur (Norusis, 1988).  Wives of Alcoholics 113  Plots of residuals, which represent the differences between obtained and predicted dependent variable scores (Tabachnick & Fidell, 1989), were examined to evaluate for normality of distribution,: linearity of relationship, homoscedasticity, and the existence of outliers. No problems were found in these areas, although it was appropriate to combine the lowest and highest income categories with the next closest category because there was only one observation in each of those categories. Because stepwise regression enters variables strictly on the basis of statistical without an eye to theory or causation (Cohen & Cohen, 1983), the regression models were checked for order of possible causality. According to Cohen and Cohen (1983), "no independent variable entering later should be a presumptive cause of an independent variable that has been entered earlier" (p. 120). The regression models for the two measures of depression as the dependent variable do not appear to violate this restriction.  Wives of Alcoholics 114  CHAPTER V Results Simple Statistics For Main Variables With a mean score of 14.7 ( n = 116) on the BDI, these wives were assessed to be moderately depressed, based on cut-off scores of 10-18 for mild to moderate depression as determined by Beck and Beamesderfer (1974), and they ranged overall from normal to severely depressed. For the SCL-90-R depression subscale (Dep), expressed in T-scores, the mean for these women was 78.6, which is more than two standard deviations above a T-score mean of 50. In other words, these women experienced significant depression overall, particularly on the SCL90-R-Dep, which tends to emphasize mood and depressed affect related to the stressful circumstances life presents more than clinical symptoms of depression, whereas the BDI may place a stronger emphasis on clinical symptoms, particularly somatic complaints. The mean PSS-Fa for these women ( n = 116) was 93.3 out of 140, which was 0.1 SD below that found by the authors (Procidano and Heller, 1983) in a group of undergraduate university students ( n = 205). In relation to the same group, the mean for PSS-Fr at 100.9 was about .2 SD lower. If these two groups are comparable, and they may not be, mean PSS-Fa and mean PSS-Fr in the study sample were only slightly less than what we might expect from a sample not generally under the stress of coping with an alcoholic spouse. In light of recent qualitative study of wives of alcoholics (Asher, 1992; Banister & Peavy, 1994; Wiseman, 1991), a greater difference might have been expected. To provide a family perspective on PSS, means were somewhat lower (89.5/Fa; 86.0/Fr) for alcoholic husbands and much, much lower (24.0/Fa; 17.5/Fr) for children (Friesen, et al., 1995), seemingly an indication of the detrimental effect of the alcoholic family on developing internal working models. Refer to Table 1 for additional  Wives of Alcoholics 115  information on descriptive statistics. Results are shown for the study sample ( n = 116) as well as for the original sample ( n = 150). There appear to be no significant differences due to attrition or incomplete data sets.  Table 1 Descriptive Statistics For Main Variables  n = 116  n = 150  Instrument  Mean  SD  Range  Mean  SD  BDI  14.71  8.27  1-35  14.79  8.49  SCL-90-R-Dep  78.62  19.40  78.11  19.04  PSS-Fa  93.30  20.50  30-140  93.86  20.81  PSS-Fr  100.92  23.15  42-139  100.85  23.07  (All distributions were approximately normal).  Correlation Matrix All main variables and demographic predictor variables were entered into the correlation matrix preliminary to conducting the multiple regression analyses. The predictor variables that showed a significant relationship (P < . 15) with either measure of depression are included in Table 2. Full time employment, years together, and being in school showed no significant relationship with either depression measure. Number of children and number residing in the house showed no significant relationship to the BDI. The two depression measures showed a correlation in this sample of .596 with P < .0001, thus indicating that they may evaluate slightly  Wives of Alcoholics 116  different aspects of the same construct. Modest correlations (.16 to .31) between independent variables were evident in areas that are logically related. Age showed a negative relationship with unemployment and number in the house and a positive relationship with number of children and years together. Education level was positively related to income and negatively related to unemployment. Number of children was positively related to number in the house. These correlations between independent variables were not high enough to indicate multicollinearity.  Table 2 Pearson-r Correlations Between Independent and Dependent Variables  Variable  1  1. B D I  1.0  2.  SCLdep  .596**  3.  PSS-Fa  -.458** -.411** 1.0  4.  PSS-Fr  -.187*  5. age  2  3  4  5  6  7  8  9  10  1.0  1.0 -.202*  .145 .144  6.  income  -.167  -.249*  7.  education  -.141  -.153  8.  unempl  .261*  .249*  9.  PTempl  -.227*  . 1.0  .155 -.148  .161  1,0  .158  .199*  1.0  -.256*  -.279*  -.205*  1.0 -.289** 1.0  10. # kids  .141  .296**  11. # house  .188*  -.303**  Only variables that showed a relationship with one of the depression variables at P < .15 are included. * P<.Q5 ** p<m\  1-0  Wives of Alcoholics 117  Multiple Regression Analyses  ,  Results were obtained separately for each measure of depression as the dependent variable using a stepwise regression and were corroborated using both a backward elimination procedure and a setwise regression analysis. All three methods gave the same results. An Rsquare of .30 was achieved for the dependent variable BDI in which PSS-Fa, PSS-Fr, unemployment, and annual family income combined in declining order to explain a total of approximately 30% of the variance. All variables were significant at the P < .05 level except family income, which was significant at the P < . 15 level. The most influential variable in the regression equation, which accounts for the majority (21% out of 30%) of the explained variance, was PSS-Fa. Results are shown in Table 3. The regression equation for the BDI as the dependent variable in this sample of wives of alcoholics whose husbands have agreed to begin treatment is: Y = 38.77 - 0.16X!-0.07X +3.33X3-0.51X .' 2  4  Table 3 Results For Stepwise Regression Analysis With BDI as the Dependent Variable  Step 1 2 3 4  Variable • Number Partial E n t e r e d Removed • In'R**2 PSSFA PSSFR UNEMP FAMINCOM  .1 0.2090 2 . 0.0457 3 • 0.0306 4 0.0139  Model R * * ?, 0.2090 " 0.2547 0.2853 0.2992 '  .  •'' - • C(p)  F  Prob>F  9.5160 . •• 30.1274 4.5005, • 6.9236 1.80.34 4.7910. 1.6664 2.2032  0.0001 0.0097 0.0307 .0.1406  A l l - v a r i a b l e s l e f t i n . the model a r e s i g n i f i c a n t a t t h e 0 .1500 l e v e l . . No o t h e r v a r i a b l e met the 0.1500 s i g n i f i c a n c e l e v e l f o r e n t r y • i n t o the model. •• :  An R-square of .27 was achieved for the dependent variable SCL-90-R-Dcp jh which PSS-Fa, number residing in the house, and annual family income combined in declining order to  Wives of Alcoholics 118  explain a total of approximately 27% of the variance. All variables were significant at the P < .05 level. The most influential variable in the regression equation, which accounts for the majority (20% out of 27%) of the explained variance, is PSS-Fa. Results are shown in Table 4. The regression equation for the SCL-90-R-Dep as the dependent variable in this sample of wives of alcoholics whose husbands have agreed to begin treatment is: Y = 110.35 - 0.40X, + 3.58X - 1.73X . 2  3  Table 4 Stepwise Regression Analysis For SCL-90-R-Depression as the Dependent Variable  Step 1 2 3  Variable Number E n t e r e d Removed In PSSFA NUMHOUSE FAMINCOM  1 2 3  Partial R**2  Model ,R**2  C(p)  F  Prob>F  0.1973 0.0439 0.0301  0.1973 0.2412 0.2713  7 0213 2 5116 0 0492  28.0272 6.5379 • 4.6256  0.0001 0.0119 0.0336  A l l v a r i a b l e s l e f t i n the model a r e s i g n i f i c a n t a t the 0.0500 l e v e l . No o t h e r v a r i a b l e met the 0.1500 s i g n i f i c a n c e l e v e l f o r e n t r y i n t o the model.  Plots of residuals were examined for both regressions to determine whether distributions were normal and relationships linear as well as to check for homoscedasticity and the existence of outliers. No problems were found, although an interesting possibility emerged in the plot of mean SCL-90-R-Dep scores for each level of family income when categories 1 and 2 were combined and categories 8 and 9 were combined due to only one observation in categories 1 and 9. Means for categories 2 through 5 were remarkably similar, ranging approximately between Tscores of 80 and 84 on the SCL-90-R-Dep, while categories 6 through 8 averaged T-scores of approximately 72 to 73, thus creating a step function or cut-off point where income may make a significant difference in depression scores, from Category 6 on up at incomes of over $50,000  Wives of Alcoholics  (Canadian dollars) per year. The difference in depression in this sample as measured by the SCL-90-R-Dep was nearly one standard deviation between incomes below and above this cutoff, with little difference before or after the cut-off point. See Figure 1.  Wives of Alcoholics 120  Figure 1 Mean SCL-90-R-Depression Scores For Each Income Category  MEAN_SCL-90-R-Dep 85 + A  A A A  80 +  75 +  70 + 2  3  4  5  6  7  8  Family Income Category (in Canadian dollars)  (Category 1 added to Category 2; Category 9 added to Category 8) Category 1,2 Category 3 = Category 4 = Category 5 = Category 6 = Category 7 = Category 8, 9  = up to $19,999 $20,000 - $29,999 $30,000 - $39,999 $40,000 - $49,999 $50,000 - $59,999 $60,000 - $69,999 = $70,000 & over  Wives of Alcoholics 121  CHAPTER VI Discussion It is at this point that the findings that have been presented will be evaluated and interpreted in comparison and contrast to previous research on social support as a moderator of depression, informed by ah understanding of the experience of wives of alcoholics. General findings from the descriptive statistics for each of the main variables will be discussed, and then findings will be evaluated in terms of the hypotheses that have been set forth. In addition, limitations of the present work will be addressed, along with generalizability to the population of wives of alcoholics and directions for further research, followed by conclusions based on the present work.  :  General Findings For the Main Variables Depression was not unusual for wives of alcoholics, who were moderately depressed with a mean of 14.7 on the BDI, and a T-score mean of 78.6 on the SCL-90-R-Dep, which is nearly two standard deviations above the normal range of 40-60. The circumstantial stress of living with an alcoholic husband was particularly apparent in the results of the SCL-90-R-Dep. The difference in severity of depression for the same sample on these two instruments supports the notion that they measure aspects of depression that are somewhat different. In comparing items from the two measures to evaluate similarities and differences in symptoms used as indicators of depression, a more clinical profile emerged for the BDI and a stronger emphasis on feelings that may arise from situations was apparent in the SCL-90-R-Dep. See Table 5 for symptoms tapped by both instruments or unique to each one. It appears that although these wives experienced only mild to moderate symptoms of clinical depression, they experienced substantial depressed affect in relation to the stresses of their everyday lives.  Wives of Alcoholics 122  Table 5 Indicators of Depression Utilized by the BDI and the SCL-90-R Depression Subscale  BDI  Both  SCL-90-R-Dep  Sense of failure Dissatisfaction Feel guilty Feel punished Irritable Difficulty with decisions Feel unattractive Sleep disturbance Appetite disturbance Weight loss Health concerns  Suicidal thoughts Sad (or) Blue Discouraged (or) Hopeless Dislike self (or) Worthlessness Blame self Crying Lost interest in things Everything is an effort Fatigue (or) Low energy Loss of sexual interest  Feeling trapped Feeling lonely Worrying too much  (or) indicates different terms used for similar symptoms (BDI listed first, SCL-90-R-Dep listed second)  The SCL-90-R-Dep seems to clearly articulate experiences of wives of alcoholics (Asher, 1992, Banister & Peavy, 1994; Wiseman, 1991) as it taps feelings of being trapped, lonely, worthless, and without hope, which are feelings that arise out of the circumstances in which they live their daily lives. In two cases, the SCL-90-R-Dep uses terminology that is more intense than the BDI—hopeless rather than discouraged and worthlessness rather than self-dislike. Selfblame, utilized by both instruments, is a core experience for wives of alcoholics as they interact with the problem of alcoholism. Norms are unavailable for both PSS instruments, and thus comparison can only be accomplished in relation to other research in which these instruments were used. It was a surprise to find that mean perceived social support from family was only 0.1 SD below that of undergraduate university students in three validation studies by Procidano and Heller (1983). Comparing current results to their work can only be done in a very general way due to basic  Wives of Alcoholics 123  differences between the populations under study. University students often have recently left home and are near the end of the developmental stage of separation from their families, which may operate to alter their needs for and perceptions of family support. A more ideal comparison group would be a population of women of a much wider age range. In view of the fact that social experience and expectations are very different for men and women, needs for and perceptions of social support are likely to differ by sex, and comparison information should be differentiated on that basis, Hibbard (1985) found men to perceive more inner control than women, which rendered support networks less critical to their well-being. The mean for our sample could be at or below normal for average community women, but suitable norms are not available for reliable comparison. Studies of community samples, differentiated by sex, for the purpose of establishing norms for the PSS-Fa and the Pss-Fr would be helpful. Compared to research by the authors (Procidano & Heller, 1983) in which the PSS-Fr was employed, the mean for perceived social support from friends for wives of alcoholics was 0.2 SD below that for undergraduate university students. With a mean age of 19, these university students were at an age where relationships with friends become relatively more important, yet more transitory, and may not really be comparable to the women in the present study, who are generally older, married, and involved in a family of procreation. They may have less opportunity to invest in friendships due to the demands of their immediate families and less desire due to their sense of shame about the alcoholic behavior. If we assumed that the findings of Procidano & Heller (1983) are a valid comparison, which we cannot, we might explain the reduction in perceived social support from friends being modest for these wives by the fact that 96 out of the study sample of 116 were employed or in school, and only 20 were unemployed. Previous research (Casey, Griffin & Googins, 1993; Googins & Casey, 1987) has shown  Wives of Alcoholics 124  employment to be a particularly positive context for wives of alcoholics, and positive involvement in alternative contexts could work to reduce their sense of isolation and low selfesteem.  Hypothesis I Perceived social support from family and perceived social support from friends is significantly inversely related to depression in wives of alcoholics. Perceived Social Support From Family That perceived social support from family is significantly inversely related to depression was clearly supported in this sample of wives of alcoholics, as has already been demonstrated in other populations (Brugha, et al, 1990; George, et al., 1989; Paykel, et al., 1980). PSS-Fa had a fairly strong inverse relationship to both measures of depression in the correlation matrix, and it was the variable that explained the majority of the variance in both regression analyses. Henderson (1984) has suggested the possibility that subjective measures of support may be contaminated by depressed affect, which seems unlikely in view of the fact that mean PSS-Fa and mean PSS-Fr in the study sample were relatively similar to those of other groups (Procidano & Heller, 1983). Modest mean differences in.PSS across groups may be evidence for conceptualizing it as an individual difference variable that is stable across time, situations, and changes in social networks (Sarason, et al., 1986). Perceived Social Support From Friends In the correlation matrix, PSS-Fr had a small but significant correlation with depression as measured by the BDI (-.187), but it showed no relationship to depression as measured by the SCL-90-R-Dep. Similarly, in the regression analysis with the BDI as the dependent variable, PSS-Fr was the variable explaining the most variance after PSS-Fa, adding 0.0457 to R in the 2  , '•'  Wives of Alcoholics 125  regression model, while in the regression analysis with the SCL-90-R-Dep as the dependent variable, PSS-Fr could not make a significant contribution to explaining the variance. This seems to be evidence for the notion that the BDI and the SCL-90-R-Dep capture different aspects of depression, even though they correlated with each Other at .596 in this sample. Considering that one's expectations of friends and family are likely to differ, and family ties usually are more enduring, it is logical that perceived support from family might bear more strongly on incidence of depression. A disparity between an individual's perception of what should be and what is may be greater arid carry more positive or negative emotional impact in relation to family than friend relationships because of a greater sense of Obligation and expected loyalty. In other words, there may be more of a potential to feel let down, which has been found to carry a particularly high risk of onset of depression in women (Brown, et al., 1986).  Hypothesis II Selected demographic variables will explain some of the variance in the relationship between perceived social support from family or friends and depression in wives of alcoholics. Variables to be explored as potential contributors to the variance in the relationship between the main variables given in Hypothesis I are employment status, frequency of attendance at religious services, number of children, number of persons residing in the: home, years lived with husband orpartner,race, age ofwife, educationleyel'of wife, and annual family income: Employment Status  ,  '  Full time employment was found to be unrelated to both measures of depression in the correlation matrix and the regression analyses.. This was somewhat of a surprise in light of the findings of previous research on employment in wives of alcoholics (Casey, Griffin & Googins, 1993; Googins & Casey, 1987) highlighting the positive benefits of employment for this group.  Wives of Alcoholics 126  It is possible that the stress of working full time may have cancelled out some of the benefits of having an alternative context for self-definition and social support for the women in our sample. Quality of the work environment as a supportive context was not specifically evaluated, nor was the women's desire to be employed, which could have clarified the dynamics that were operating. The correlation matrix did show a negative relationship between part time employment and depression as measured by the BDI, indicating a possibility that part time employment might provide the benefit of an alternative context without the time pressure that full time work involvement may impose. Part time employment was not a significant contributor to variance in the regression analysis for the BDI, however. Again, additional information about the work context and how these women perceived their work experience would have been helpful. Unemployment bore a significant relationship with both measures of depression in the correlation matrix, but retained significance only in the regression analysis for the BDI, where it added 0.0306 to R . Perhaps unemployment is more related to clinical symptoms of depression 2  than to the stress of situations. This relationship may be reciprocal in that depressive symptoms may inhibit job-seeking, and being unemployed may contributed to depressed affect and a sense of helplessness. Unemployment may be stressful for wives who want to work (Moss & Plewis, 1977), who are under financial stress (Keith & Schafer, 1980), or who have an emotionally adverse home environment (Warr & Parry, 1982), all of which may apply to wives of alcoholics. Religious Participation It is curious that, similar to the findings of Asher (1992) and Wiseman (1991), the present findings reflect a lack of formal religious participation. This does not preclude the possibility of personal spiritual involvement, but it does indicate that these wives generally did not seek support from others in a formal religious setting, perhaps due to practical concerns, such as time  Wives of Alcoholics 127  constraints and family needs, or perhaps due to a sense of shame and embarrassment. Further inquiry would be needed to explain this phenomenon. About 92.6% attended religious services four times or less during the previous year, with the majority not attending religious services at all. The data was badly skewed and therefore unusable in the regression analysis. However, no correlation with either measure of depression was found when the correlation matrix was run. Number of Children and Number Residing in the House Age of children was not available in the TARP database, but number of children showed a very small relationship to depression as measured by the SCL-90-R-Dep in the correlation matrix. However, any significance of number of children was not apparent in the regression analysis. Wiseman (1991) found having more children and older children to be more associated with inertia and depression in the wives of alcoholics she studied, yet Brown & Harris (1978) found three or more children under age 14 to be a vulnerability factor for depression in women. Even though these works did not agree on age of children, in both, number and age of children combined to impact vulnerability. Number in the house was the closest variable in the TARP database that would reflect both of these factors, though it would not simulate an age of 14 as a cut-off point. With the many problems faced by adolescents today, a child older than 14 might continue to contribute to family stress, and number in the house would provide an indicator of the household burden a responsible, caretaking adult in the home might experience. Thus number in the house was used in the present study to substitute for number and age of children in past work. Number in the house was found to be unrelated to the BDI and mildly related to the SCL90-R-Dep in the correlation matrix. The same situation was found in the regression analyses, where number in the house added 0.0439 to R , with SCL-90-R-Dep as the dependent variable, 2  Wives of Alcoholics 128  second only to PSS-Fa, and it did not make any significant contribution to the variance of the BDI. Number in the house likely impacts depressed affect relevant to the daily, stresses of life more than symptoms of clinical depression, and in that light, these results seem logical. Years Lived With Partner  .  ,.  Years together and whether legally married or common law were unrelated to either measure of depression in the correlation matrix or the regression analyses. It may be that once the wife of an alcoholic has defined the problem as alcoholism and herself as inadequate (Asher, 1992), the passage of additional time in the alcohol-embedded marriage may not be critical to whether or not depression is present. Race As mentioned previously, the sample of wives of alcoholics studied here was 96% white, and thus it was not possible to study the effect of race on perceptions of social support as related to depression. Of the study sample, only four wives were non-white, and only four of their husbands were non-white. Data on race was collected for the categories of Native Indian, Oriental, Black, Hispanic, and White, and though drinking is found in populations from these other categories, they were poorly represented in the study sample despite the fact that ethnic groups reside in areas from which the sample was obtained. A selection factor may have operated here. It may be that non-white families that are impacted by alcoholism are less likely to present for treatment or volunteer for research due to cultural beliefs about males and about alcoholic behavior. In addition, media publicity may have differentially impacted these groups. Whatever the reason, inferences can only apply to a predominantly white population of wives of alcoholics.  Wives of Alcoholics 129  Age, Education Level, and Annual Family Income  *  Age,.education level, and income were not found to be clearly definitive in differentiating depressed women from survivors in Wiseman's (1991) work, although survivors were slightly younger. In the present study, age was not related to the B D I in the correlation matrix but was significantly negatively related to the SCL-90-R-Dep, which could reflect a reduction in depressed affect from the daily stresses of life as age increases. This could just be a natural result of.developmental factors, such as children growing up and becoming less dependent. Age was not a significant factor in either regression analysis of depression as the dependent variable. I  Education level 'showed a small negative relationship to both depression measures in the  correlation matrix but was unrelated in both regression analyses, and it appears from the data available not to be a defining factor in incidence of depression among wives of alcoholics. Further detail about their subjective experience of education may be enlightening, however. For example; did she feel a sense of capability or inadequacy in educational settings? Does the knowledge obtained have relevance to her life in the present?  ;  Brown, et al. (1986) chose to study working class women because, they were more at risk for onset of depression due to economic factors (Brown & Harris, 1978), and have been found to experience enhanced well-being when economic stress was reduced. In the present study, annual family income was negatively related to both measures of depression in the correlation matrix, although more strongly to the SCL-90-R-Dep. A similar pattern was apparent in the regression analyses, with income adding 0.0139 to R (P < .15) with the BDI as the dependent variable and 2  0.0301 to R (P< .05) with the SCL-90-R-Dep as the dependent variable. Though less critical 2  than other variables, income remains a significant factor, particularly, it seems, at incomes below $50,000 (Canadian dollars), as seen when mean SCL-90-R-Dep scores were plotted against  Wives of Alcoholics 130  income in the study sample, as explained previously and shown in Figure 1. Increased income appears to reduce depressed affect relevant to the everyday stresses of life after basic financial needs are met, but after a threshold is achieved that addresses basic financial needs, additional income does not appear to further decrease depressed affect. This phenomenon needs further study to determine its validity.  Limitations Selection factors may cause the sample to be less than fully representative of the population of all wives of alcoholics. Many families dealing with the specter of alcohol abuse seek to deny or conceal the problem of alcoholism due to shame, embarrassment, and feeling marked by their experience, and they may not respond to the opportunity to participate in treatment and research. Cultural meanings attached to alcoholic behavior in different ethnic groups may contribute to a reticence to present for treatment. Newspaper and radio coverage may also differentially impact these groups, thereby influencing selection. However, within the limits of practicality, these limitations are unavoidable. Using intact families will eliminate many of the wives who divorce because of the alcoholism, and they are not likely to have the same kind of characteristics as those who stay. Wives who remain in the alcohol-embedded marriage may be more likely to be traditionallygendered women who feel responsible to be caretakers (Belle, 1982; Bepko, 1986), looking for approval and self-definition outside themselves and thus being more vulnerable to a loss of self (Jack, 1991). As previously noted, those women who divorce may only represent 10% of all wives of alcoholics (Downing, 1991; Kokin & Walker, 1989), so the sample seems not to be seriously restricted by this exclusion, even though it is not fully representative of all wives of alcoholics and will therefore limit generalizability somewhat. These limitations may be  Wives of Alcoholics 131  substantially offset by the fact that remarkably similar experiences among wives of alcoholics are found even across diverse groups (Wiseman, 1991). Perhaps the largest source of selection bias is that all of these couples involved in The Alcohol Recovery Project have come for the treatment which is offered. This was necessary because the overriding purpose of TARP was to study treatment and recovery, and therefore subjects needed to be willing to be involved in treatment. O'Farrell, Kleinke, Thompson, and Cutter (1986) studied 35 couples who accepted therapy compared to 28 couples who did not. They found that couples who accepted had sought more help during the previous year and had more marital separations, with husbands being more educated, happier in their marriage than their wives, employed full-time, and having a larger number of alcohol-related arrests. In couples who did not accept treatment, the husband was older and had more alcohol-related hospitalizations. They lived farther from the clinic, and the wives showed better marital adjustment. The conclusion was that the unhappiness of the wife was a motivating force to get them to treatment. Based on this work (O'Farrell, et al., 1986), it might be a concern that SES may be higher in the group that accepted therapy due to the husband's greater education and likelihood of full time employment. The husbands being older and having more health-related problems from alcohol in couples rejecting therapy may mean that the alcoholism is further advanced, problems for wives may be more severe, and wives may also be older. Their reported greater marital satisfaction may indicate denial or resignation. In any case, the groups showed significant differences, and selection bias may pose a threat to internal validity. An important difference between O'Farrell, et al.'s (1986) study and the one proposed here is cost of therapy. At The  Wives of Alcoholics 132  Alcohol Recovery Project, treatment was free for participation in research, which may help to counter selection bias, especially with regard to SES. Also, as noted previously, the attrition group in the present work, half of whom dropped out and may have been reticent to accept therapy, were of similar income and education level as the study sample. In terms of the process of defining the alcoholic behavior as alcoholism (Asher, 1992), the alcoholic's wife may experience a senseof validation through her husband being accepted for therapy because objective testing has confirmed that drinking is indeed a problem, and the ambivalence (Asher, 1992) she may have been feeling is thereby resolved. She may perceive an increase in available support merely by coming to a treatment center, and she may feel a glimmer of increased hope that change is possible. This could operate to reduce mean differences i n perceptions of social support as compared to other groups, and it may explain the low mean differences that were found here between the study sample and the sample used for validation studies by Procidano and Heller (1983). As mentioned previously, there are no norm groups for PSS-Fa and PSS-Fr. Comparison to appropriate norm groups would enhance interpretation of information gathered through the use of these instruments. Random community samples used for . norm groups with separate statistics for men and women would be very helpful.  The use of self-report instruments can be a source of response bias in terms of inaccuracy of recall or misrepresentation. Warr and Parry (1982) reviewed various studies that failed to find differences in well-being in women based on employment status and concluded that self-report instruments failed to identify real differences which they were able to find using clinical measures. Screening procedures were used in the present study to help reduce any bias from selfreport by eliminating those who would tend to give socially desirable responses rather than  Wives of Alcoholics 133  accurate responses. Self-report instruments that were employed were utilized appropriately according to their design, and they were designed to be as effective as possible within the  ,  constraints of self-report as a vehicle for data collection. It has been suggested that depression scores on self-report scales may reflect moodrelated features of depression more than symptoms of clinical depression (Mitchell, Cronkite & Moos, 1983). The instruments used here for/evaluation, however, are well accepted and well respected,as valid measures of depression. Overall, despite some limitations, this study corroborates earlier qualitative work with wives of alcoholics; Wives of alcoholics have traditionally been misunderstood, and more research is heeded in an effort to understand the complex interaction of internal and external factors that combine to determine how they will meet the multiple challenges they face.  Possibilities For Further Research The present study is preliminary in nature in that it seeks to determine whether perceptions of social support moderate the incidence 6f depression in wives who are under the stress of their husbands' alcoholism, as it does in other populations. The influence of selected demographic variables was also explored, although further detail is needed about these wives' subjective experience of employment and parenting, for example, before variables not contributing to variance in the present work can be.confidently ruled out as irrelevant. Prospective studies would be helpful in more clearly identifying onset and recovery factors for depression in wives of alcoholics specifically, and pre-existing factors should be explored, such as family history of depression. Each individual's inner resources, such as personality, intelligence, ability to adapt, and beliefs about self and other, need to be explored along with  Wives of Alcoholics 134  objective and subjective evaluations of the external environment, including chronic and acute stress and the demands of life stage transitions. A wide variety of available information was informally included in the correlation matrix to get a larger picture of the data and how it may fit together in light of related theory. As mentioned previously, study of the SASB-Intro-Affil in relation to the BDI could be explored in a separate study of the present data. Though not included in the present study, all subscales of the SCL-90-R were also checked for relationships with variables under study, and they provided some interesting data for later exploration. PSS-Fa was very significantly related to all of the SCL-90-R subscales with Pearson-r correlations ranging from .29 to .44, which may lend evidence to the view of perceived social support from family as an individual difference variable (Sarason, et al., 1990) related to intrapsychic state. PSS-Fr was significantly related to only four of the nine SCL-90-R subscales, which were Interpersonal Sensitivity, Phobic Anxiety, Paranoia, and Psychotic, with correlations from .20 to .36 in the study sample. Influence of the intrapsychic domain is also apparent here, though perhaps less strongly than for the PSS-Fa. Exploring perceptions of social support in relation to other symptom dimensions is one possible avenue of further exploration in this population. How do these symptom dimensions impact perceptions of support, and how do perceptions of support impact other symptom dimensions? There could be a reciprocal relationship between these variables. In light of work by Warr and Parry (1982) as well as current results, it seems apparent that in order to more fully understand the influence of employment for wives of alcoholics, further information is needed about what kind of experience the work environment provides for each subject. Quality of work relationships, type and prestige of job, and amount of physical and psychological stress inherent in the work could be important factors in whether an employee  Wives of Alcoholics 135  thrives in the work setting. Work settings are likely to be as variable as home settings. Evaluation of the work environment would enhance our understanding of how employment influences perceptions of social support and risk of depression. Failure to control work environment variables could mask important relationships between depression and these variables. The indication of a step function relevant to income and mean SCL-90-R-Dep scores was a significant finding, and it needs to be explored further to verify if there is indeed a threshold of income below which income is more related to depressed affect and above which additional income does not further reduce mean depression scores, as shown in Figure 1. It could be possible that inadequate income contributes to depressed affect when basic financial needs are not fully provided for, but that above a level of basic need fulfillment, added income does not further reduce depressed affect. An enticing direction for new research is to determine the effect of gender role orientation on wives of alcoholics in terms of self-esteem, depression, and coping styles. It could be hypothesized that more wives of alcoholics will show many traditionally feminine traits and few traditionally masculine traits, especially those who are mired in inertia and depression, as found by Wiseman (1991). In previous work, wives of alcoholics saw themselves as more passive, submissive, feminine, and wifely than wives of nonalcoholics (Kogan & Jackson, 1965). This work needs to be updated. Androgynous wives would be less likely to conform to the codependency label, could be expected to show more self-esteem with less tolerance to a status quo built around alcoholism, and may be more likely to leave the alcoholic. Bepko (1986) suggests that rigid gender roles are what keep the alcoholic family system dysfunctional.  Wives of Alcoholics 136  Alcoholic women have been studied for gender-role orientation, but further study of the influence of gender is needed for wives of alcoholics. It would also be interesting to study the complementary gender issues faced by the male alcoholic, how they impact his drinking, aggressiveness, and self-esteem and to offer treatment to couples with an alcoholic member based on heightened awareness of their own socialization and the limitations it may impose on the way they view their lives. Comparisons before and after treatment that promotes gender awareness might be quite enlightening. Van Wormer (1989) described a male-specific alcoholism treatment group designed to help male alcoholics address their issues of masculinity. Empirical results of this type of treatment are needed, as gender issues may play a part in the alcoholic system (Bepko, 19986). Further study could be directed at including the impact of marital distress on PSS-Fa and the two measures of depression. It might be expected that degree of marital distress as measured by the DAS would make a significant contribution to explained variance relevant to risk of depression. An inverse relationship between marital distress and PSS-Fa is a possibility, in which case the two variables could be run in separate regression analyses with depression as the dependent variable, as well as with PSS-Fa as the dependent variable and marital distress and the demographic variables as predictors. If marital distress and PSS-Fa are not collinear, then marital distress could be added to the predictors used in the regression analyses of the present study. In light of information from work done by Casey, et al. (1993), who found that half of their sample of 60 wives of alcoholics were from alcoholic families of origin, it could be useful to study the influence of this variable on the relationships we have studied. Because this  Wives of Alcoholics 137  information was not gathered for wives in the alcohol-complicated families studied by TARP, it was not included in the present study. In view of extremely low PSS (Friesen, et al., 1995) among the children in the alcoholic families studied by TARP, the relationship between being from an alcoholic family of origin and perceived social support from family could be a strong one and might be explored in future work. The larger research project from which this one derives breaks the alcoholic husbands . into three groups of 50 each and compares them using different treatments: Supported Individual Therapy for the husband only, Experiential-Systemic Therapy for the husband only, and Experiential-Systemic Therapy for the couple. The effect of the couple therapy for the wife in terms of available and perceived support, the quality of the marital relationship, and self-esteem would be another interesting avenue to pursue.  Generalizability Findings from the present study are generalizable to the population of wives of alcoholics with intact marriages and at least one child, who are white, whose husbands are white, and whose husbands have agreed to begin treatment, with a full spectrum of age, employment status, education level, and income included within those parameters. Applying these results to wives who opt out of the alcohol-embedded marriage should be done with caution, as the study sample excluded many of the women who leave by requiring an intact marriage and a desire to remain intact to participate. A few of these women later separated, though it may have been fewer than in a full-spectrum sample of wives of alcoholics.  Wives of Alcoholics 138  Conclusions It is quite apparent from the findings of the present study that perceptions of social support from family are significantly inversely related to depression in the subset of wives of alcoholics described above. Annual family income also makes a contribution to explaining the variance in this relationship, although it may impact depressed affect associated with the stress of everyday living somewhat more than the clinical features of depression. Perceptions of social support from friends and being unemployed contribute to the variance in the BDI, which may tap more heavily the clinical symptoms of depression, but not to the SCL-90-R-Dep, which may touch more strongly on depressed affect associated with the stress of everyday living. Number in the house significantly contributes to the variance in the SCL-90-R-Dep but not in the BDI. Recognizing the strong psychometric properties of the BDI and the SCL-90-R-Dep, this seems to show that depression is a multidimensional phenomenon in which each instrument focuses more strongly on specific aspects of depression. In addition, perceived social support from family and perceived social support from friends relate to depression differently, with perceived support from family having the more profound influence, thereby supporting the idea that they are separate constructs and that perceived family support can have a profound effect on well-being in wives of alcoholics, as has been found in other populations.  Age and years together do not appear to impact the relationship between social support and depression in these wives of alcoholics, but it may be premature to make conclusive statements about number of children, employment status, and education level. Number and age of children living at home could be relevant, as indicated by number in the house being a significant factor in the regression analysis for the SCL-90-R-Dep. The subjective experience of  Wives of Alcoholics 139  employment needs to be explored in some detail to make explicit the conditions under which employment contributes to well-being for wives of alcoholics. Education may also have a differential impact based on the woman's subjective experience of academia, whether or not it contributed to her sense of herself as a capable person, and the personal relevance of knowledge obtained. More information about the experience of parenting may also be useful. Some children are more cooperative than others, and some have special needs that add to the responsibility carried by the mother who has been conditioned by culture and circumstance to be overresponsible. This study is only a beginning in the effort to define and quantify the role of perceived social support as it relates to depression in wives of alcoholics. There is much more to be done in identifying the factors at work in the complex interaction of the intrapsychic with the interpersonal, as the woman makes sense of her environment under conditions of ongoing stress, which include interaction with her alcoholic husband. Onset and recovery from depression are a product of the biological, the psychological, and the social in dynamic interaction, and interventions must be chosen based on awareness of the influence of all three dimensions. As stated by Coyne and Delongis (1986), "There are undoubtedly profound connections between having good relationships and well-being, but they are likely to be complex, reciprocal and contingent" (Lloyd, 1995, p. 46). Wives of alcoholics experience ongoing adverse conditions related to their husbands' alcoholic behavior, and adversity is a known precipitating influence in psychiatric disorder (Brugha, 1995). "Under adversity, it is those who construe their social relationships as inadequate who are more likely to develop symptoms" (Henderson, Byrne, & Duncan-Jones,  Wives of Alcoholics 140  1981; Barnett & Parker, 1995, p. 214). The essence of a supportive relationship is the communication of caring and love. The perception that one's well-being is important to significant others is protective in that it provides a feeling of being worthwhile, capable, and valued, with the availability of resources if needed (Sarason, et al., 1990). That social support is generally protective against depression is fairly well established (Henderson, 1992; Monroe & Johnson, 1992; Veiel, 1995), and women may have a particularly strong need for social support because of their relational orientation (Hibbard, 1985; Belle, 1982). When a married woman's marriage is not a source of support, she is particularly vulnerable to depression (Brown & Harris, 1978; Roy, 1978). According to Barnett and Parker (1995), women under stress need to be able to talk to others and be heard, they need practical assistance, and they need contact with others who have had similar experience but have achieved some resolution to the problem. Wives of alcoholics often experience chronic multiple stresses associated with their husbands' drinking. According to Henderson (1984), "Persons with high life-event exposure will tend also to have recent losses of social support" (p. 51). As a group, wives of alcoholics may experience high adversity, increasing isolation, and decreasing selfesteem and social support, placing them at significant risk for depression.  Marital and family relationships seem to be of particular importance in the moderating influence of social support. It would seem that family therapy based on a biopsychosocial model and informed by an understanding of the cultural experience of women, has considerable potential to meet the personal needs of wives of alcoholics to be heard and understood, while at the same time strengthening the family as a mutually supportive context. Changes on a personal, family, or societal level that give these women more active control in their own lives (Radloff,  Wives of Alcoholics 141  1975), a sense of being personally valued (Parry, 1988), and a sense of personal power to make choices in their own behalf (Lederer & Brown, 1991) could be expected to positively impact their mental health. Professionals who would be helpers must work toward this end.  Wives of Alcoholics 142  References Abramson, L . Y . & Andrews, D. E . (1982). Cognitive models of depression: Implications for sex differences in vulnerability to depression. International Journal of Mental Health, 11, 77-94. Anderson, C. M . & Holder, D. P. (1989). Women and serious mental disorders. In M . McGoldrick, C. M . Anderson & F. Walsh (Eds.), Women in families: A framework for family therapy (pp. 381-405). New York: W.W.Norton. Andrews, B. & Brewin, C. R. (1990). Attributions of blame for marital violence: A study of antecedents and consequences. Journal of Marriage and the Family, 52, 757-767. Andrews, B. & Brown, G. W. (1988). Social support, onset of depression and personality. Social Psychiatry and Psychiatric Epidemiology, 23, 99-108. Argyle, M . (1987). The psychology of happiness. London: Methuen. Asher, R. M . (1992). Women with alcoholic husbands: Ambivalence and the trap of codependency. Chapel Hill: The University of North Carolina Press. Asher, R. & Brissett, D. (1988). Codependency: A view from women married to alcoholics. The International Journal of the Addictions, 23(4), 331-350. Bailey, M . B. (1965). Al-Anon family groups as an aid to wives of alcoholics. Social Work, 10, 68-74. Bailey, M . B. (1967). Psychophysiological impairment in wives of alcoholics as related to their husbands' drinking and sobriety. In R. Fox (Ed.), Alcoholism: Behavioral research, therapeutic approaches (pp. 134-142). New York: Springer Publishing Co.  Wives of Alcoholics 143  Bailey, M . B., Haberman, P. & Alksne, H. (1962). Outcomes of alcoholic marriages: Endurance, termination or recovery. Quarterly Journal of Studies on Alcohol, 23, 610-623. Ballard, R. (1959). The interaction between marital conflict and alcoholism as seen through MMPI's of marriage partners. American Journal of Orthopsychiatry, 29(3), 528-546. Banister, E. M . & Peavy, R. V . (1994). The erosion of self: An ethnographic study of women's experience of marriage to alcoholic husbands. Canadian Journal of Counselling, 28(3), 206-221. Barnett, R. C. & Baruch, G. K. (1987). Social roles, gender, and psychological distress. In R.C. Barnett, L. Biener, & G.K. Baruch (Eds.), Gender and stress (pp. 122-143). New York: The Free Press. Barnett, B. & Parker, G. (1995). A test of the social support hypothesis. In T.S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 213-238). Cambridge, UK: Cambridge University Press. Barerra, M . Jr. (1986). Distinctions between social support concepts, measures, and models. American Journal of Community Psychology, 14, 413-45. Beattie, M . (1987). Co-dependent no more. New York: Harper & Row. Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. New York: Harper & Row. Beck, A . T. (1974). The development of depression: A cognitive model. In R. J. Friedman & M . M . Katz (Eds.), The psychology of depression: Contemporary theory and research (pp. 3-27). Washington, DC: V . H . Winston.  Wives of Alcoholics 144  Beck, A. T. & Beamesderfer, A. (1974). Assess of depression. The Depression Inventory. In P. Pichot (Ed.), Psychological measurements in psychopharmacology, modern problems in pharmacopsychiatry (Vol. 7) (pp. 151-169). Basel, Switzerland: Karger. Beck, A. T., Steer, R. A. & Garbin, M . G . (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77-100. Beck, A. T. & Weishaar, M . (1989). Cognitive therapy. In A. Freeman, K. M . Simon, L. E. Beutler & H. Arkowitz (Eds.), Comprehensive handbook of cognitive therapy (pp. 21-36). New York: Plenum Press. Becker, J. (1979). Vulnerable self-esteem as a predisposing factor in depressive disorders. In R. A. Depue (Ed.), The psychobiology of the depressive disorders: Implications for the effects of stress (pp. 317-334). New York: Academic Press. Becvar, D. S. & Becvar, R. J. (1988). Family therapy: A systemic integration. Neadham Heights, M A : Allyn and Bacon. Belle, D. (1980). Who uses mental health facilities? In M . Guttentag, S. Salasin, & D. Belle (Eds.), The mental health of women. New York: Academic Press. Belle, D. (1982). The stress of caring: Women as providers of social support. In L . Goldberger & S. Breznitz (Eds.), Handbook of stress: Theoretical and clinical aspects (pp. 496505). New York: The Free Press. Belle, D. (1982b). Social ties and social support. In D. Belle (Ed.), Lives in stress: Women and depression (pp. 133-144). Beverly Hills, CA: Sage Publications.  Wives of Alcoholics 145  Benassi, V. A., Sweeney, P. D. & Dufour, C. L. (1988). Is there a relation between locus of control orientation and depression? Journal of Abnormal Psychology, 97(3), 357-367. Benjamin, L. S. (1974). Structural Analysis of Social Behavior. Psychological Review, 81. 392-425. Benjamin, L . S. (1984). Principles of predicition using Structural Analysis of Social Behavior. In R. A. Zucker, J. Aronoff & A. I. Rabin (Eds.), Personality and the prediction of behavior (pp. 121-174). New York: Academic Press. Bepko, C. (1986). Alcoholism as oppression: The dilemma of the woman in the alcoholic system. In J. C. Hansen & M . A. Riche (Eds.), The family therapy collections: Women and family therapy (pp. 64-77). Rockville, M D : Aspen Publications. Bepko, C. (1988). Female legacies: Intergenerational themes and their treatment for women in alcoholic families. In L . Braverman (Ed.), A guide to feminist family therapy (pp. 97-111). New York: Harrington Park Press. Bepko, C. (1989). Disorders of power: Women and addiction in the family. In M . McGoldrick, C. M . Anderson & F. Walsh (Eds)., Women in families: A framework for family therapy (pp. 406-426). New York: W. W. Norton. Bepko, C. & Krestan, J. A. (1985). The responsibility trap: A blueprint for treating the alcoholic family. New York: The Free Press. Berne, E. (1966). Principles of group treatment. New York: Grove Press. Billings, A. G., Cronkite, R. C. & Moos, R. H . (1983). Social-environmental factors in unipolar depression: Comparisons of depressed patients and nondepressed controls. Journal of Abnormal Psychology, 92(2), 119-133.  Wives of Alcoholics 146  Billings, A . G. & Moos, R. H. (1982). Social support and functioning among community and clinical groups: A panel model. Journal of Behavioral Medicine, 5, 295-311. Billings, A. G . & Moos, R. H. (1985). Life stressors and social resources affect posttreatment outcomes among depressed patients. Journal of Abnormal Psychology, 94(2), 140153. Birtchnell, J. (1988). Defining dependence. British Jourmal of Medical Psychology, 61, 111-123. Boggs, M . H . (1944). The role of social work in the treatment of inebriates. Quarterly Journal of Studies on Alcohol, 4, 557-567. Bograd, M . (1990). Scapegoating mothers: Conceptual errors in systems formulations. In M . P. Mirkin (Ed.), The social and political contexts of family therapy (pp. 69-87). Boston, M A : Allyn and Bacon. Boss, P. & Thorne, B. Family sociology and family therapy: A feminist linkage. In M . McGoldrick, C. M . Anderson & F. Walsh (Eds.), Women in families: A Framework for family therapy (pp. 78-96). New York: W.W. Norton & Co. Bowlby, J. (1969). Attachment and loss: Attachment (Vol. 1). London: Hogarth Press. Bowlby, J. (1973). Attachment and loss: Separation, anxiety, and anger (Vol.2). London: Hogarth Press. Bowlby, J. (1980). Attachment and loss: Sadness and depression (Vol. 3). London: Hogarth Press. Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.  Wives of Alcoholics 147  Braverman, L. (1986). The depressed woman in context: A feminist family therapist's analysis. In J. C. Hansen & M . Ault-Riche (Eds.), The family therapy collections: Women and family therapy (pp. 90-101). Rockville, M D : Aspen Publications. Brewin, C. R. (1995). Cognitive aspects of social support processes. In T. S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 96-116). Cambridge, U K : Cambridge University Press. Brewin, C. R., MacCarthy, B. & Furnham, A. F. (1989). Social support in the face of adversity: The role of cognitive appraisal. Journal of Research in Personality, 23, 354-72. Brofenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32, 513-531. Brofenbrenner, U . (1979). The ecology of human development: Experiments by nature and design. Cambridge, M A : Harvard University Press. Brophy, C. J., Norell, N. K. & Kiluk, D. J. (1988). An examination of the factor structure and convergent and discriminant validity of the SCL-90-R in an outpatient clinic population. Journal of Personality Assessment, 52, 334-340. Broverman, I. K., Broverman, D. M . , Clarkson, F. E., Rosenkrantz, P. S. & Vogel, S. R. (1970). Sex-role stereotypes and clinical judgments of mental health. Journal of consulting and clinical psychology, 34(1), 1-7. Brown, G. W., Andrews, B., Harris, T., Adler, Z. & Bridge, L. (1986). Social support, self-esteem and depression. Psychological Medicine, 16, 813-831. Brown, G. W. & Harris, T. (1978). Social origins of depression: A study of psychiatric disorder in women. New York: The Free Press.  Wives of Alcoholics 148  Brugha, T. S. (1995). Social support and psychiatric disorder: Overview of evidence. In T. S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 1-38). Cambridge, UK: Cambridge University Press. Brugha, T. S., Bebbington, P. E . , MacCarthy, B., Stmt, E., Wykes, T. & Potter, J. (1990). Gender, social support and recovery from depressive disorders: a prospective clinical study. Psychological Medicine, 20, 147-156. Brugha, T. S., Sturt, E . , MacCarthy, B., Potter, J., Wykes, T. & Bebbington, P.E. (1987). The Interview Measure of Social Relationships: The description and evaluation of a survey instrument for assessing personal social resources. Social Psychiatry, 22, 123-128. Buhrke, R. A. & Fuqua, D. R. (1987). Sex differences in same- and cross-sex supportive relationships. Sex Roles, 17(5/6), 339-352. Burchfield, S. R., Hamilton, K. L . & Banks, K. L . (1982). Affiliative needs, interpersonal stress, and symptomatology. Journal of Human Stress, 8, 5-9. Burnett, M . M . (1984). Toward a model for counseling the wives of alcoholics: A feminist approach. Alcoholism Treatment Quarterly, 1(2), 51-60. Burton, G & Kaplan, H. M . (1968). Marriage counseling with alcoholics and their spouses—The correlation of excessive drinking behavior with family pathology and social deterioration. British Journal of Addiction, 63, 161-170. Buunk, B. P. & Hoorens, V . (1992). Social support and stress: the role of social comparison and social exchange processes. British Journal of Clinical Psychology, 31, 445-457. Caplan, G. (1976). The family as a support system. In G. Caplan & M . Killilea (Eds.), Support systems and mutual help: Multidisciplinary explorations. New York: Grune & Stratton.  Wives of Alcoholics 149  Casey, J. C , Griffin, M . L. & Googins, B. K. (1993). The role of work for wives of alcoholics. American Journal of Drug and Alcohol Abuse, 19(1), 119-131. Champion, L. (1990). The relationship between social vulnerability and the occurrence of severely threatening life events. Psychological Medicine, 20, 157-161. Champion, L. (1995). A developmental perspective on social support networks. In T. S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 61-95). Cambridge, UK: Cambridge University Press. Chatterjee, S. & Price, B. (1991). Regression analysis by example. New York: John Wiley & Sons, Inc. Chodorow, N. (1974). Family structure and feminine personality. In M . Z. Rosaldo & L. Lamphere (Eds.), Woman, culture, and society. Stanford, CA: Stanford University Press. Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine, 38(5), 300-314. Cohen, J. & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum Associates Publishers. Cohen, L. H., McGowan, J., Fooskas, S. & Rose, S. (1984). Positive life events and social support and the relationship between life stress and psychological disorder. American Journal of Community Psychology, 12(5), 567-587. Cohen, S. & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310-357.  Wives of Alcoholics 150  Conger, R. D. (1981). The assessment of dysfunctional family systems. I n B . B . Lakey & A . E. Kazdin (Eds.), Advances in clinical child psychology (Vol. 4) (pp. 199-242). New York: Plennum Press. Costello, C. G. (1982). Social factors associated with depression: A retrospective community study. Psychological Medicine, 12, 329-339. Coyne, J. C. (1976). Depression and the response of others. Journal of Abnormal Psychology, 85, 186-93. Coyne, J. C. (1989). Thinking postcognitively about depression. In A. Freeman, K. M . Simon, L. E. Beutler & H. Arkowitz (Eds.), Comprehensive handbook of cognitive therapy (pp. 227-244). New York: Plenum Press. Coyne, J. C. & Delongis, A. (1986). Going beyond social support: The role of social relationships in adaptation. Journal of Consulting and Clinical Psychology, 54, 457-460. Coyne, J. C. & Downey, G. (1991). Social factors and psychopathology: Stress, social support, and coping processes. Annual Review of Psychology, 42, 401-425. Cyr, J. J., Doxey, N. C. & Vigna, C. M . (1988). Factorial composition of the SCL-90-R. Journal of Social Behavior and Personality, 3, 245-252. Dean, A. (1985). Introduction. In A. Dean (Ed.), Depression in multidisciplinary perspective (pp. xi-xix). New York: Brunner/Mazel. Decker, J. T., Redhourse, J., Green, R. D. & Starrett, R. (1983). The wife of the alcoholic: Sexist stereotypes in the alcoholism literature. Journal of Sociology and Social Welfare, 10(3), 462-471.  Wives of Alcoholics 151  Denzin, N. K. (1984). On understanding emotion. San Francisco, C A : Jossey-Bass. Denzin, N. K. (1987). The alcoholic self. Newbury Park, CA: Sage Publications. Derogatis, L . R. (1983). SCL-90-R: Administrative, scoring and procedures manual-II for the revised verson and other instruments of psychopathology rating scale series. Towson, MD: Clinical Psychometric Research. Derogatis, L . R. & Cleary, P. (1977). Confirmation of the dimensional structure of the SCL-90-R: A study in construct validation. Journal of Clinical Psychology, 33, 981-989. Derogatis, L . R., Rickels, K. & Rock, A . (1976). The SCL-90 and the MMPI: A step in the validation of a new self-report scale. British Journal of Psychiatry, 128, 280-289. Dohrenwend, B. S. (1973). Social status and stressful life events. Journal of Personality and social Psychology. 28. 225-235. Downing, C. (1991). Sex role setups and alcoholism. In N. Van Den Bergh (Ed.), Feminist perspectives on addictions (pp. 47-60). New York: Springer Publishing Co. Dunkel-Schetter, C , Folkman, S. & Lazarus, R. (1987). Correlates of social support receipt. Journal of Personality and Social Psychology, 53, 71-80. East, P. L . (1989). Early adolescents' perceived interpersonal risks and benefits: Relations to social support and psychological functioning. Journal of Early Adolescence, 9, 374395. Edwards, P., Harvey, C. & Whitehead, P. C. (1973). Wives of alcoholics: A critical review and analysis. Quarterly Journal of Studies on Alcohol, 34, 112-132. Epstein, C. F. (1974). Changing sex roles: A review of sociological research developments and needs. Current Psychology, 283-315.  Wives of Alcoholics 152  Ettorre, B. (1986). Women and drunken sociology: Developing a feminist analysis. Women's Studies International Forum, 9(6), 515-520. Fennell, M . J. V . (1989). Depression. In K. Hawton, P.M. Salkovskis, J. Kirk & D . M . Clark (Eds.), Current behavior therapy for psychiatric problems: A practical guide (pp. 169234). New York: Oxford University Press. Finney, J., Moos, R. H. & Mewborn, C. R. (1980). Post-treatment experiences and treatment outcomes of alcoholic patients six months and two years after hospitalization. Journal of Consulting and Clinical Psychology, 43, 17-29. Fisher, J. D., Goff, B. A., Nadler, A . & Chinsky, J. M . (1988). Social psychological influences on help seeking and support from peers. In B.H.Gottlieb (Ed.), Marshalling social support: Formats, processes, and effects (pp. 267-304). Newbury Park, CA: Sage Publications. Flaherty, J. A. & Richman, J. A. (1986). Effects of childhood relationships on the adult's capacity to form social supports. American Journal of Psychiatry, 143, 851-855. Fondacaro, M . R. & Moos, R. H. (1987). Social support and coping: A longitudinal analysis. American Journal of Community Psychology, 15(5), 653-673. Fowles, D. C. & Gersh, F. (1979). Neurotic depression: The endogenous—neurotic distinction. In R. A. Depue (Ed.), The psychobiology of the depressive disorders: Implications for the effects of stress (pp. 55-80). New York: Academic Press. Fox, R. & Lyon, P. (1955). Alcoholism: Its scope, cause, and treatment. New York: Random House.  Wives of Alcoholics 153  Friesen, J., Conry, R., Grigg, D. & Weir, W. (1995). The Alcohol Recovery Project final report. Vancouver, B C . Frieze, I. H . (1975). Women's expectations for and causal attributions of success and failure. In M . T. S. Mednick, S. S. Tangri, & L. W. Hoffman (Eds.), Women and achievement: Social and motivational analyses (pp. 158-171). Washington, DC: Hemisphere. Fremouw, W. J., Perczel, M . & Ellis, T. E . (1990). Suicide risk: Assessment and response guidelines. Toronto: Pergamon Press. Gaertner, M . L. (1939). The Alcoholic Marriage. Thesis, New York School of Social Work. Garber, J., Miller, W. R. & Seaman, S. F. (1979). Learned helplessness, stress, and the depressive disorders. In R. A. Depue (Ed.), The psychobiology of the depressive disorders: Implications for the effects of. stress (pp. 335-363). New York: Academic Press. George, L. K., Blazer, D. G., Hughes, D. C. & Fowler, N. (1989). Social support and the outcome of major depression. British Journal of Psychiatry, 154, 478-485. Gerstel, N., Riessman, C. K. & Rosenfield, S. (1985). Explaining the symptomatology of separated and divorced women and men: The role of material conditions and social networks. Social Forces, 64(1), 85-101. Gierymski T. & Williams, T. (1986). Codependency. Journal of Psychoactive Drugs, 18(1). 7-13. Gilbert, P. (1992). Depression: The evolution of powerlessness. Hove, UK: Lawrence Erlbaum Associates.  Wives of Alcoholics 1  Gilbert, P. (1994). Male violence: Towards an integration. In J.Archer (Ed.), Male violence (pp. 352-389). London: Routledge. Gilbert, P. (1995). Attachment, cooperation and rank: The evolution of the need for status and social support. In T .S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 117-141). Cambridge, UK: Cambridge University Press. Gilligan, C. (1993). In a different voice: Psychological theory and women's development. Cambridge, M A : Harvard University Press. Goffman, E. (1967). Interaction ritual. Garden City, NY: Anchor Books. Goering, P. N., Lancee, W. J. & Freeman, J. J. (1992). Marital support and recovery from depression. British Journal of Psychiatry, 160, 76-82. Goffman, E. (1967). Interaction ritual. Garden City, NY: Anchor Books. Goldner, V . (1987). Instrumentalism, feminism, and the limits of family therapy. Journal of family psychology, 1(1), 109-116. Googins, B. & Casey, J. C. (1987). Dynamics of alcoholism in working families. Alcoholism Treatment Quarterly, 4(3), 47-66. Gottlieb, B. H. (1985). Social networks and social support: An overview of research, practice, and policy implications. Health Education Quarterly, 12(1), 5-22. Gove, W. R. (1972). The relationship between sex roles, marital status, and mental illness. Social Forces, 51, 34-44. Gove, W. R. & Tudor, J. F. (1973). Adult sex roles and mental illness. American Journal of Sociology, 78(4), 812-835.  Wives of Alcoholics 155  Graham, H. (1983). Caring: A labor of love. In J. Finch & D. Groves (Eds.), A labor of love. London, UK: Routledge and Kegan Paul. Griffin-Shelley, E . (1986). Sex roles in addiction: Defense or deficit? The International Journal of the Addictions, 21(12), 1307-1312. Guberman, N. (1990). The family, women, and caregiving: Who cares for the caregivers? In V. Dhruvarajan (Ed.), Women and well-being. London, UK: McGill-Queens University Press. Hammen, C , Ellicott, A., Gitlin, M . & Jamison, K.R. (1989). Sociotropy/autonomy and vulnerability to specific life events in patients with unipolar depression and bipolar disorders. Journal of Abnormal Psychology, 98(2), 154-160. Heim, S. C. & Snyder, D. K. (1991). Predicting depression from marital distress and attributional processes. Journal of Marital and Family Therapy, 17(1), 67-72. Henderson, A. S. (1981). Social relationships, adversity, and neurosis: An analysis of prospective observations. British Journal of Psychiatry, 138, 391-398. Henderson, A . S. (1984). Interpreting the evidence on social support. Social Psychiatry, 19, 49-52. Henderson, A. S. (1988). An introduction to social psychiatry. New York: Oxford University Press. Henderson, A. S. (1992). Social support and depression. In H. O. F. Veiel & U. Baumann (Eds.), The meaning and measurement of social support (pp. 85-92). New York: Hemisphere. Henderson, A. S., Byrne, D. G. & Duncan-Jones, P. (1981). Neurosis and the social environment. Sydney: Academic Press.  Wives of Alcoholics 156  Henderson, A . S. & Moran, P. A. P. (1983). Social relationships during the onset and remission of neurotic symptoms: A prospective community study. British Journal of Psychiatry, 143, 467-472. Hibbard, J. H. (1985). Social ties and health status: An examination of moderating factors. Health Education Quarterly, 12(1), 23-34. Hirschfeld, R., Klerman, G . Chodoff, P., Korchin, S. & Barrett, J. (1976). Dependency—self-esteem—clinical depression. Journal of the American Academy of Psychoanalysis, 4, 373-388. Holahan, C. J. & Moos, R. H. (1981). Social support and psychological distress: A longitudinal analysis. Journal of Abnormal Psychology, 90, 365-370. Holten, J. D. (1990). When do we stop mother-blaming? Journal of Feminist Family Therapy, 2(1), 53-60: Homans, G. L . (1974). Social behavior (2 ed.). New York: Harcourt Brace nd  Jovanovich. Hooley, J. M . , Orley, J. & Teasdale, J. D. (1986). Levels of expressed emotion and relapse in depressed patients. British Journal of Psychiatry, 148, 642-647. Hooley, J. M . & Teasdale, J. D. (1989). Predictors of relapse in unipolar depressives: Expressed emotion, marital distress, and perceived criticism. Journal of Abnormal Psychology, 98, 229-235. Ilfeld, F. W. (1982). Marital stressors, coping styles, and symptoms of depression. In L . Goldberger & S. Breznitz (Eds.), Handbook of stress: Theoretical and clinical aspects (pp. 482495). New York: The Free Press.  Wives of Alcoholics 157  Jack, D. C. (1991). Silencing the self: Women and depression. Cambridge, M A : Harvard University Press. Jackson, J. K. (1954). The adjustment of the family to the crisis of alcoholism. Quarterly Journal of Studies on Alcohol, 15, 562-586. Jacob, T. & Krahn, G. L. (1988). Marital interactions of alcoholic couples: Comparison with depressed and nondistressed couples., Journal of Consulting and Clinical Psychology 56(1), 73-79. Johnson, P. (1974). Social power and sex role stereotypes. San Francisco, CA: Western Psychological Association. Johnson, J. H . & Sarason, I. G. (1978). Life stress, depression, and anxiety: Internalexternal control as a moderator variable. Journal of Psychosomatic Research, 22, 205-208. Kaslow, N. J. & Carter, A. S. (1991). Depressed women in families: The search for power and intimacy. In T.J.Goodrich (Ed.), Women and power (pp. 166-182). New York: W.W. Norton & Co. Keith, P. M . & Schafer, R. B. (1980). Depression in one- and two-job families. Psychological Reports. 47. 669-670. Kelly, G . (1955). The psychology of personal constructs (Vols. I and II). New York: W.W. Norton & Co. Kessler, R. C. (1979). Stress, social status, and psychological distress. Journal of Health and Social Behavior 20. 259-272.  Wives of Alcoholics 158  Kessler, R. C. & McLeod, J. D. (1984). Sex differences in vulnerability to life events. American Sociological Review, 49, 620-631. Kessler, R. C , McLeod, J. D. & Wethington, E. (1985). Cost of caring: A perspective on the relationship between sex and psychological distress. In I.G. Sarason & B.R. Sarason (Eds.), Social support: Theory, research, and application. Boston, M A : Martinus Nijhoff Publishers. Klein, D. N., Harding, K., Taylor E. B. & Dickstein, S. (1988). Dependency and selfcriticism in depression: Evaluation in a clinical population. Journal of Abnormal Psychology, 97(4), 399-404. Kobasa, S. C. (1979). Stressful life events, personality, and health: An inquiry into hardiness. Journal of Personality and Social Psychology, 37, 1-11. Kogan, K. L., Fordyce, W. E. & Jackson, J. K. (1963). Personality disturbance in wives of alcoholics. Quarterly Journal of Studies on Alcohol, 24(2), 227. Kogan, K. L. & Jackson, J. L. (1965). Stress, personality, and emotional disturbance in wives of alcoholics. Quarterly Journal of Studies on Alcohol, 26, 486-495. Kokin, M . & Walker, I. (1989). Women married to alcoholics: Help and hope for nonalcoholic partners. Toronto: Macmillan of Canada. Krestan, J. & Bepko, C. (1991). Codependence: The social reconstruction of female experience. In C. Bepko (Ed.), Feminism and addictions (pp. 49-66). New York: The Haworth Press, Inc.  Wives of Alcoholics 159  Kurdek, L . A. & Sinclair, R. J. (1985-1986). Adolescents' view on issues related to divorce. Journal of Adolescent Research, 1, 373-387. Lakey, B. & Cassady, P. B. (1990). Cognitive processes in perceived social support. Journal of Personality and Social Psychology, 59(2), 337-343. Lakey, B. & Heller, K. (1988). Social support from a friend, perceived support, and social problem solving. American Journal of Community Psychology, 16, 811-824. Lammers, S. M . M . (1991). Problem drinking women: Sex role conflict and gender identity. In G . M . Schippers, S. M . M . Lammers & C. P. D. R. Schaap (Eds.), Contributions to the psychology of addiction (pp. 15-31). Berwyn, PA: Swets & Zeitlinger. Lane, C. & Hobfoll, S. E . (1992). How loss affects anger and alienates potential supporters. Journal of Consulting and Clinical Psychology, 60(6), 935-942. Lavach, J. F. & Lanier, H. B. (1975). The motive to avoid success in 7 , 8 , 9 , and 10 th  th  th  th  grade high achieving girls. Journal of Educational Research, 68, 216-218. Lederer, G. S. & Brown, F. H . (1991). The interlocking issues of gender and entitlement in alcoholic marriages. Family Dynamics of Addiction Quarterly, 1(3), 12-20. Lemle, R. & Mishkind, M . E. (1989). Alcohol and masculinity. Journal of Substance Abuse Treatment, 6, 213-222. Lewis, M . S. (1937). Alcoholism and family casework. Family, 18, 39-44. Lewis, M . S. (1954). The initial contact with wives of alcoholics. Social Casework, 35, 8-14. Lloyd, C. (1995). Understanding social support within the context of theory and research on the relationship of life stress and mental health. In T. S. Brugha (Ed.), Social support and  Wives of Alcoholics 160  psychiatric disorder: Research findings and guidelines for clinical practice (pp. 41-60). Cambridge, UK: Cambridge University Press. MacAndrew, C. & Edgerton, R. B. (1969). Drunken comportment. Chicago: Aldine. McCord, J. & McCord, W. (1960). Origins of alcoholism. Stanford, CA: Stanford University Press. McNamara, K. (1992). Depression assessment and intervention: Current status and future directions. In S. D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology (2 ed.), pp. 691-718. New York: Wiley & Sons. nd  Mendenhall, W. (1989). Co-dependency definitions and dynamics. In B. Carruth & W. Mendenhall (Eds.), Co-dependency: Issues in treatment and recovery (pp. 3-17). New York: The Haworth Press. Merikangas, K. R., Weissman, M . M . & Pauls, D. L. (1985). Genetic factors in the sex ratio of major depression. Psychological Medicine, 15, 63-69. Miller, J. B. (1976). Toward a new psychology of women. Boston, M A : Beacon. Mitchell, R. E., Cronkite, R. C. & Moos, R. H . (1983). Stress, coping, and depression among married couples. Journal of Abnormal Psychology, 92(4), 433-448. Monroe, S. M . & Johnson, S. L. (1992). Social support, depression and other mental disorders: In retrospect and toward future prospects. In H . O. F. Veiel & U . Baumann (Eds.), The meaning and measurement of social support. New York: Hemisphere. Moos, R. H . & Moos, B. (1986). Family Environment Scale manual (2 ed.). Palo nd  Alto, CA: Consulting Psychologists Press. Moss, P. & Plewis, I. (1977). Mental distress in mothers of preschool children in inner London. Psychological Medicine, 7, 641-652.  Wives of Alcoholics 161  Mostow E. & Newberry, P. (1975). Work role and depression in women: A comparison of workers and housewives in treatment. American Journal of Orthopsychiatry, 45, 538-548. Mowrer, H . R. (1940). A psychocultural analysis of the alcoholic. American Sociological Review, 5, 546-557. Murray, H. A. (1938). Explorations in personality. New York: Oxford University Press. Myers, J. K. et. al., (1984). Six-month prevalence of psychiatric disorders in three communities: 1980 to 1982. Archives of General Psychiatry, 41, 959-967. Nol, J. (1991). Self-object search: The role of addictions in a patriarchal culture. In N. Van Den Bergh (Ed.), Feminist perspectives on addictions (pp. 31-44). New York: Springer Publishing Co. Nolan-Hoeksema, S. (1995). Epidemiology and theories of gender differences in unipolar depression. In M . V . Seeman (Ed.), Gender and psychopathology (pp. 63-87). Washington, DC: American Psychiatric Press. Norusis, M . J. (1988). SPSS-X advanced statistics guide (2 Ed.). Chicago: SPSS, nd  Inc. Nye, F.I. (1963). Personal satisfactions. In F. I. Nye & L . W. Hoffman (Eds.), The employed mother in America. Chicago: Rand McNally. Oatley, K. & Bolton, W. (1985). A social-cognitive theory of depression in reaction to life events. Psychological Review, 92(3), 372-388.  Wives of Alcoholics 162  O'Farrell, T. J. (1990). Sexual functioning of male alcoholics. In R. L. Collins, K. E . Leonard & J. S. Searles (Eds.), Alcohol and the family: Research and clinical perspectives (pp. 244-271). New York: The Guilford Press. O'Farrell, T. J. & Birchler, G.R. (1987). Marital relationships of alcoholic, conflicted, and nonconflicted couples. Journal of Marital and Family Therapy, 13(3), 259-274. O'Farrell, T. J., Kleinke, C. L., Thompson, D. L . & Cutter, H. S. G. (1986). Differences between alcoholic couples accepting and rejecting an offer of outpatient marital therapy. American Journal of Drug and Alcohol Abuse, 12(3), 285-294. O'Gorman, P. (1991). Codependence and women: Unraveling the power behind learned helplessness. In N. Van Den Bergh (Ed.), Feminist perspectives on addictions. New York: Springer Publishing Co. Orford, J., Guthrie, S., Nicholls, P., Oppenheimer, E., Egert, S. & Hensman, C. (1975). Self-reported coping behavior of wives of alcoholics and its association with drinking outcome. Journal of Studies on Alcohol. 36(9). 1254-1267. Orner, R. J. (1987). Post-traumatic stress disorders in victims of the Falklands War: Syndrome and treatment. Paper delivered at the British Psychological Society Annual Conference, Brighton, UK. Overholser, J. C. (1990). Emotional reliance and social loss: Effects on depressive symptomatology. Journal of Personality Assessment, 55(3/4), 618-629. Paolino, T. J. & McCrady, B. S. (1977). The alcoholic marriage: Alternative perspectives. New York: Grune & Stratton.  Wives of Alcoholics 163  Parry, G. (1988). Mobilizing social support. In F.N. Watts (Ed.), New developments in clinical psychology (Vol. 2) (pp. 83-104). New York: John Wiley & Sons. Parry, G. (1995). Social support processes and cognitive therapy. InT. S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 279-294). Cambridge, UK: Cambridge University Press. Parsons, T. & Bales, R. (1955). Family, socialization, and interaction process. New York: The Free Press. Paykel, E. S. (1978). Contribution of life events to causation of psychiatric illness. Psychological Medicine. 8. 245-253. Paykel, E. S. (1991). Depression in women. British Journal of Psychiatry. 158, 22-29. Paykel, E. S., Emms, E. M . , Fletcher, J. & Rassaby, E. S. (1980). Life events and social support in puerperal depression. British Journal of Psychiatry. 136, 339-346. Pearlin, L . I. (1985). Social structure and process of social support. In S. Cohen & S. L . Syme (Eds.), Social Support and Health (pp. 43-59). New York: Academic Press. Pearlin, L. I. & Johnson, J. S. (1977). Marital status, life-strains and depression. American Sociological Review. 42. 704-715. Pearlin, L. I. & Lieberman, M . A. (1979). Social sources of emotional distress. In R. Simmons (Ed.), Research in community mental health. Greenwich, CT: J.A.I. Pietromonaco, P. R. & Rook, K. S. (1987). Decision style in depression: The contribution of perceived risks versus benefits. Journal of Personality and Social Psychology, 52(2), 399-408.  Wives of Alcoholics 164  Piotrowski, C , Sherry, D. & Keller, J. W. (1985). Psychodiagnostic test usage: A survey of the Society for Personality Assessment. Journal of Personality Assessment, 49, 115119. Pittman, F. (1990). The masculine mystique. Family Therapy Networker, 14(3), 41-52. Pretorious, T. B. (1994). Network orientation and social support as moderators of the stress-depression relationship: A South African study. Journal of Social Behavior and Personality, 9(1). 1-12. Price, G. M . (1945). A study of the wives of 20 alcoholics. Quarterly Journal of Studies on Alcohol, 5. 620-627. Procidano, M . E . & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal of Community Psychology, 11(1), 1-24. Radloff, L . S. (1975). Sex differences in depression: The effects of occupation and marital status. Sex Roles, 1(3), 249-265. Radloff, L. S. & Monroe, M . K. (1978). In L. S. Hansen & R. S. Rapoza (Eds), Career development and counseling of women (pp. 199-221). Springfield, IL: Charles C. Thomas Publishers. Rogalski, C.J. (1986). The male addict's consciousness of women in his life. In D. Howard (Ed.), The dynamics of feminist family therapy, pp. 259-274. New York: The Haworth Press. Rook, K. S. (1984). The negative side of social interaction: Impact on psychological well-being. Journal of Personality and Social Psychology, 46, 1097-1108.  Wives of Alcoholics 165  Roy, A . (1978). Vulnerability factors and depression in women. British Journal of Psychiatry, 133, 106-110. Rubin, L . B. (1979). Women of a certain age. New York: Harper & Row. Sandler, I. N. & Barerra, M . Jr. (1984). Toward a multimethod approach to assessing the effects of social support. American Journal of Community Psychology, 12(1), 37-52. Sarason, B. R., Pierce, G. R., Shearin, E. N., Sarason, I. G., Waltz, J. A. & Poppe, L . (1991). Perceived social support and working models of self and actual others. Journal of Personality and Social Psychology, 60(2), 273-287. Sarason, B. R., Shearin, E. N., Pierce, G. R. & Sarason, I. G. (1987). Interrelations of social support measures: Theoretical and practical implications. Journal of Personality and Social Psychology, 52, 813-832. Sarason, I. G., Levine, H. M . , Basham, R. B. & Sarason, B. R. (1983). Assessing social support: The social support questionnaire. Journal of Personality and Social Psychology, 44, 127-139. Sarason, I. G., Sarason, B. R. & Pierce, G. R. (1990). Social support: The search for theory. Journal of Social and Clinical Psychology, 9(1), 133-147. Sarason, I. G., Sarason, B. R. & Shearin, E. N. (1986). Social support as an individual difference variable: Its stability, origins, and relational aspects. Journal of Personality and Social Psychology, 50(4), 845-855. Schaap, C. P. D. R., Schellekens, I. & Schippers, G. M . (1991). Alcohol and marital interaction: The relationship between male alcoholism, interaction characteristics, and marital therapy. In G. M . Schippers, S. M . M . Lammers & C. P. D. R. Schaap (Eds.), Contributions to the psychology of addiction (pp. 65-86). Berwyn, PA: Swets and Zeitlinger.  Wives of Alcoholics 166  Schur, E. M . (1984). Labeling women deviant: Gender, stigma, and social control. New York: Random House. Seligman, M . E. (1974). Depression and learned helplessness. In R. J. Friedman & M . M . Katz (Eds.), The psychology of depression: Contemporary theory and research (pp. 83-113). Washington, DC: V.H.Winston. Seligman, M . E. P. (1975). Helplessness: On depression, development and death. San Francisco: W. H. Freeman. Selzer, M . L. (1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127, 1653-1658. Shaef, A. W. (1989). Escape from intimacy: The psuedo-relationship addictions. San Francisco, CA: Harper & Row. Shaver, P. R. & Brennan, K. A . (1991). Measures of depression and loneliness. In J. P. Robinson, P. R. Shaver & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (pp. 195-204). New York: Academic Press. Shaver, P., Schwartz, J., Kirson, D. & O'Connor, C. (1987). Emotion knowledge: Further exploration of a prototype approach. Journal of Personality and Social Psychology, 52, 1061-1086. Shaw, J., Kennedy, S. H. & Joffe, R. T. (1995). In M . V . Seeman (Ed.), Gender and psychopathology (pp. 89-111). Washington, DC: American Psychiatric Press, Inc. Silverman, C. (1968). The epidemiology of depression. Baltimore, M D : John Hopkins Press.  Wives of Alcoholics 167  Snyder, D. K. & Heim, S. C. (1992). Marriage, depression, and cognition: Unraveling the gordian Knot—Reply to Ettinger, et al. Journal of Marital and Family Therapy, 18(3), 303307. Spanier, G . B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and other similar dyads. Journal of Marriage and the Family, 38, 15-28. Spanier, G. B. (1990). Dyadic Adjustment Scale (DAS). In J. Touliatos, B. F. Perlmutter & M . A . Straus (Eds.), Family measurement techniques. Newbury Park: Sage Publications. Steer, R. A., Beck, A. T. & Garrison, B. (1986). Applications of the Beck Depression Inventory. In M . Sartorium & T. A . Ban (Eds.). Assessment of depression (pp. 121-142). Geneva, Switzerland: World Health Organization. Steinglass. P.. Bennett. L. A.. Wolin. S. J. & Reiss. D. (1987). The alcoholic family. New York: Basic Books. Stevens, J. (1986). Applied multivariate statistics for the social sciences. Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. Swann, W. B. & Predmore, S. C. (1985). Intimates as agents of social support: Sources of consolation or despair? Journal of Personality and Social Psychology, 49, 1609-1617. Tabachnick, B. G. & Fidell, L. S. (1989). Using multivariate statistics.  New York:  Harper Collins Publishers. Tavris, C. (1990). The politics of codependency. The Family Therapy Networker, 14(1), 43.  Wives of Alcoholics 168  Tavris, C. (1992). The mismeasure of woman. New York: Simon & Schuster. Tebbets. R. (1982). Work: Its meaning for women's lives. In D. Belle (Ed.), Lives in stress: Women and depression (pp. 83-95). Beverly Hills, CA: Sage Publications. Thoits, P. A . (1983). Multiple identities and psychological well-being: A reformulation and test of the social isolation hypothesis. American Sociological Review, 48, 174-187. Thoits, P. A . (1986). Social support as coping assistance. Journal of Consulting and Clinical Psychology, 54(4). 416-423. Treadway, D. (1990). Codependency: Disease, metaphor, or fad? The Family Therapy Networker. 14(1). 39-42. Troise, F. P. (1992). The capacity for experiencing intimacy in wives of alcoholics or codependents. Alcoholism Treatment Quarterly, 9(3/4), 39-55. Turner, J. (1983). Direct, indirect, and moderating effects of social support on psychological distress and associated conditions. In H . B. Kaplan (Ed.), Psychosocial stress: Trends in theory and research (pp. 105-155). New York: Academic Press. Turner, R. J. & Avison, W. R. (1989). Gender and depression: Assessing exposure and vulnerability to life events in a chronically strained population. The Journal of Nervous and Mental Disease, 177(8). 443-455. Turpin, G. & Lader, M . (1986). Life events and mental disorder: Biological theories of their mode of action. In H . Katschnig (Ed.), Life events and psychiatric disorders: Controversial issues. Cambridge, UK: Cambridge University Press.  Wives of Alcoholics 169  Van Den Bergh, N. (1991). Having bitten the apple: A feminist perspective on addictions. In N. Van Den Bergh (Ed.), Feminist perspectives on addictions, pp. 3-30. New York: Springer Publishing Co. Van Wormer, K. (1989). The male-specific group in alcoholism treatment. Small Group Behavior, 20(2), 228-242. Vaughn, C. E . & Leff, J. (1985). Expressed emotion in families: Its significance for mental illness. New York: Guilford Press. Veiel, H. O. F. (1985). Dimensions of social support: A conceptual framework for research. Social Psychiatry, 20, 156-62. Veiel, H . O. F. (1993). Detrimental effects of kin support networks on the course of depression. Journal of Abnormal Psychology, 102(3), 419-429. Veiel, H . O. F. (1995). Social support as a high risk condition for depression in women. In T. S. Brugha (Ed.), Social support and psychiatric disorder: Research findings and guidelines for clinical practice (pp. 145-162). Cambridge, UK: Cambridge University Press. Walters, M . (1990). The codependent Cinderella who loves too much . . . fights back. Family Therapy Networker, 14(4), 53-57. Warr, P. & Parry, B. (1982). Paid employment and women's psychological well-being. Psychological Bulletin. 91(3). 498-516. Wegscheider-Cruse, S. & Cruse, J. (1990). Understanding co-dependency.  Peerfield  Beach, FL: Health Communications. Weissman, M . M . & Klerman, G. L . (1977). Sex differences and the epidemiology of depression. Archives of General Psychiatry, 34, 98-111.  Wives of Alcoholics 170  Weissman, M . M . & Paykel, E. S. (1974). The depressed woman: A study of social relationships. Chicago: The University of Chicago Press. Weissman, M . M . , Sholmskas, D., Pottenger, M . , Prusoff, B. A. & Locke, B. Z. (1977). Assessing depressive symptoms in five psychiatric populations: A validation study. American Journal of Epidemiology, 106, 203-214. Welch, S. & Booth, A. (1977). Employment and health among married women with children. Sex Roles, 3, 385-397. Wethington, E. & Kessler, R. C. (1986). Perceived support, received support, and adjustment to stressful life events. Journal of Health and Social Behavior, 27, 78-89. Wethington, E., McLeod, J. D. & Kessler, R. C. (1987). The importance of life events for explaining sex differences in psychological distress. In R. C. Barnett, L . Biener, & G. K. Baruch (Eds.), Gender and stress (pp. 144-156). New York: The Free Press. Whalen, T. (1953). Wives of alcoholics: Four types observed in a family service agency. Quarterly Journal of Studies on Alcohol, 14, 632-641. Whisman, M . A. & Jacobson, N. S. (1989). Depression, Marital satisfaction, and marital and personality measures of sex roles. Journal of Marital and Family Therapy, 15(2), 177-186. Whitfield, C. L. (1989). Co-dependence: Our most common addiction—Some physical, mental, emotional, and spiritual perspectives. In B. Carruth & W. Mendenhall (Eds.), Codependency: Issues in treatment and recovery (pp. 19-36). New York: The Haworth Press. Wilcox, B. L. (1981). Social support, life stress, and psychological adjustment: A test of the buffering hypothesis. American Journal of Community Psychology, 9(4), 371-386. Wilhelm, K. & Parker, G. (1988). The development of a measure of intimate bonds. Psychological Medicine, 18, 225-234.  Wives of Alcoholics 171  Wiseman, J.P. (1975). An alternative role for the wife of an alcoholic in Finland. Journal of Marriage and the Family, 37(1), 172-179. Wiseman, J. P. (1991). The other half: Wives of alcoholics and their socialpsychological situation. New York: Aldine de Gruyter. Wright, K. D. & Scott, T. B. (1978). The relationship of wives' treatment to the drinking status of alcoholics. Journal of Studies on Alcohol, 39, 1577-1581.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items