University of Southern California Dissertations and Theses

Adolescent predictors and adult consequences of alcohol dysphoria: A longitudinal study

(USC Thesis Other)

Adolescent predictors and adult consequences of alcohol dysphoria: A longitudinal study

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. U M I films the text directly from tfie original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough. substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send U M I a complete manuscript and there are missing pages, these w ill be noted. Also, if unauthorized copyright material had to be removed, a note w ill indicate the deletion. Oversize materials (e.g.. maps, drawings, charts) are reproduced by sectioning the original, t>eginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact U M I directly to order. Bell & Howell Information and Learning 300 North Zeeb Road. Ann Arbor. M l 48106-1346 USA UIVQ 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ADOLESCENT PREDICTORS AND ADULT CONSEQUENCES OF ALCOHOL DYSPHORIA: A LONGITUDINAL STUDY by Thomas Fletcher Locke A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial FuLQliment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Education - Counseling Psychology) August, 1999 Copyright 1999 Thomas F. Locke Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. U M I Number. 9955501 UMI UM I Microform9955501 Copyright 2000 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor. Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation, w ritten by ................................ under the direction of h.x& Dissertation Committee, and approved b y all its members, has been presented to and accepted by The Graduate School, in partial fulfillm ent of re quirements for the degree of DOCTOR OF PHILOSOPHY D eano^^ggliliàte Studies D a te.... DISSERTATION COMMITTEE i«ee • • • •••••■ #1 ••••• ♦ e ••••••••••••••••• • Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table o f Contents List o f Tables...................................................................................................................................... iv List o f Figures...................................................................................................................................... v Abstract................................................................. -vi I. Introduction..................................................................................................................................... I Substance Abuse and Mental Disorders...................................................................... 3 Diagnostic Issues with Co-morbidity.............................................................. 3 Epidemiological Studies................................................................................ 5 Prevalence o f and Diagnostic Categories o f Alcoholism and Depression................ 7 Relationships Between Alcoholism and Depression.................................................. 8 Clinical Studies................................................................................................. 9 Concerns About Clinical Studies................................................................... 11 Community Studies....................................................................................... 12 Theoretical Perspectives.................................................................................................... 15 Depression and Alcoholism; Biological, Psychological, Social Theories 18 Biological Factors............................................................................................... 19 Psychopharmacology o f Alcohol................................................... 19 Genetic Factors in Alcoholism .........................................................19 Neuropsychology o f Depression......................................................21 Psychological Factors.................................................................................... 22 Cognitive Theory o f Depression......................................................22 Self-derogation Theory...................................................................... 23 Self-medication Hypothesis.............................................................. 23 Social Factors............................ 24 Alcohol Problems............................................................................... 24 Behavioral Genetic Perspective........................................................25 Gender Differences in Alcoholics....................................................26 General Risk Factors for the Female Alcoholic........................... 28 Stage-specific Risk Factors............................................................... 30 Depression................. 32 Constructs Analyzing the Consequences o f Alcoholic Depression....................... 33 Family Bonding.................................................................................................. 34 Social Conformity...................................... 35 Relationship Satisfaction...................................................................................36 Perceived Opportunity....................................................................................... 40 Employment............................... 42 Research Hypotheses......................................................................................... 45 II. Methodology.................................................................................................................................. 47 Participants.............................................................. 47 Measures...............................................................................................................................50 Adolescence.........................................................................................................50 Young Adulthood............................................................................................... 52 Adulthood............................................................................................................53 Statistical Procedures........................................................................................ 55 III. Results............................................................................................................................................ 60 Gender-related Differences...............................................................................................60 Sequence o f Analyses.............................. 66 Statistical Fit Indices..................................................................................................... 66 Method o f Developing Structural M odels.....................................................................68 Equality Constraints...........................................................................................................68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. iii Alcohol Problems, Dysphoria & Alcohol Dysphoria Over Time..............................69 Alcohol Problems, Dysphoria & Psychosocial Variables Over Time.......................70 Integrating Alcohol Dysphoria with Psychosocial Variables Over Time 70 Adolescent Predictors and Later Outcomes..................................................................71 Young Adulthood Predictors and Adult Outcomes..................................................... 72 IV. Discussion....................................................................................................................................81 Research Findings............................................................................................................. 82 Predictors o f Alcohol Use and Depression............................................... 82 Outcomes o f Alcohol Problems, Dysphoria and Alcohol Dysphoria 85 Predictors o f Alcohol Problems and Dysphoria............................................90 Theoretical Implications.............................................................................................. 92 Implications for Intervention and Clinical Practice..................................................... 95 Limitations and Research Implications....................................................................... 99 V. Reference....................................................................................................................................... 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV List o f Tables Table 2 .1 — Descripdoa o f Sample.............................................................................................. 48 Table 12 . - Summary o f Latent Constructs and Measured Variables....................................... 57 Table 3.1 — Multiple-group Analyses o f Factor Models.............................................................. 61 Table 3.2 - Table o f Factor Loadings by Gender......................................................................... 63 Table 3.3 — Factor Intercorrelations by Gender............................................................................ 65 Table 3.4 — First-order Factor Intercorrelations.......................................................................... 73 Table 3.5 — Second-order Factor Intercorrelations..................................................................... 74 Table 3.6 — Predictor and Outcomes Effects..................................................................................75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. L ist o f Figures Figure 2 .1 — Factor Model................................................................................................................. 59 Figure 3.1- Second- order Confirmatory Factor Model............................................................. 77 Figure 3.2 —Second-order Structural Model.............................................................................. 78 Figure 3 J — Full Second- order Confirmatory Factor Model..................................................... 79 Figure 3.4 — Second- order Factor Model.............................................................................— 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V I ABSTRACT The co-occurrence o f alcohol use and depression is a significant issue that has had a great deal o f clinical and research attention. Although numerous investigations have been conducted, differing results have been reported about the relationships between alcoholism, depression and psychosocial variables. Problems in the existing base o f knowledge have been highlighted by the National Enstitute on Alcohol Abuse and Alcoholism, indicating that most research in this area has been conducted on or included people within a clinical population and is cross-sectional in nature (NIAAA, 1991). Thus, generalizability to the community at large is restricted and an analysis o f temporal sequencing o f behavior has been lim ited This research addresses the concerns raised by NIAAA, as it uses subjects fi’ om the general population and is longitudinal in nature. Using prospective data from a community sample assessed when they were adolescents age 17-19, eight years later, when they were young adults, and, eight years after that when they were 33-35 years o f age, longitudinal and cross-sectional analysis using Structural Equation Modeling examined the complex relationships between alcohol use, depression, and psychosocial factors hypothesized to predict or be consequences. Alcohol use and depression were found to be related for adolescent females, but not for males, having a combined effect that I have called alcohol dysphoria; subsequent analyses were conducted on a female sample (N=305). The developmental pathway for alcohol dysphoria in women indicates that once established in young adulthood, it tends to persist. It was (bund that alcohol use and depression have additive effects that have stronger predictive capacity than either does alone. In young adulthood, it is a strong predictor o f decreased marital, relational and job satisfaction; and, an increase in alcohol Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. v il dysphoria &om adolescence to young adulthood predicts decreased perceived opportunity in adulthood. Several specific and direct effects o f the hypothesized predictor factors in adolescence, family bonding, parental divorce, social conformity, relationship satisfaction, and perceived opportunity^, were found to indirectly effect the development o f alcohol dysphoria by influencing both alcohol problems and depression independently. The results o f this investigation are discussed, as are theoretical, and clinical implications. The focus o f prevention efforts and screening for alcohol problems or depression in female adolescents should be expanded, as alcohol dysphoria appears to be a complex phenomenon, detectable at that time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CH A PTERl INTRODUCTION This work examines the co-occurrence o f alcohol problems and depression longitudinally within a sample o f people from the community at large, with a focus on determining the predictors and consequences o f such behavior. Substantial and numerous investigations have been conducted on alcoholism, depression, and their combined effects, yielding interesting results that have helped us, as a profession, understand some o f the mechanisms at work. However, there are inherent limitations that restrict the applicability and generalizability o f results, particularly within the general population. Namely, much o f it has used a clinical or treatm ent population and/or has been cross- sectional in design. This current study, using a community sample has greater generalizability and provides specific information about how the comorbid effects o f alcoholism and depression impact people who are not part o f a clinical or treatment population. This work differs firom and adds to the existing body o f knowledge in several ways. First, since participants were selected firom the general population results have greater generalizability and applicability. Data were obtained firom the UCLA Longitudinal Study o f Growth and Development database (Newcomb, 1997a), a study which began in 1976 and has followed subjects from adolescence through early to m iddle adult years. Second, participants have been followed since 1976 and have had repeated measurements conducted. This allows for an exam ination o f the stability and change ii^ the behaviors under investigation overtim e and, facilitates an understanding o f the temporal sequencing. Third, a powerful statistical methodology (structural equation Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 modeling) was used that allows causal pathways to be established for both the unique effects o f alcohol problems and depression as well as their combined effects. Overall, this work has the potential to inform theoretical understanding, research foci, intervention efforts and service utilization in a different way than much existing research. Factors in adolescence hypothesized to predict alcohol problems, depression, and their com orbid effects (labeled alcohol <fysphoria) in this investigation will be relationship satisfaction, attitudes o f social conformity, family bonding, parental divorce, and sense o f perceived opportunity. Outcomes in adulthood will be assessed by examining relationship satisfaction, m arital satisfaction, sense o f perceived opportunity, job satisfaction and job stability. This chapter highlights and describes relevant literature as a firamework for the current work, examining limitations in the existing knowledge base. A detailed account o f the methods used is presented followed by the results obtained The results are discussed and integrated within existing theory and research. First, I will begin this chapter by discussing general issues surrounding dual disorders and epidemiological research illustrating the prevalence o f co-occurring substance abuse and mental disorders. The focus will then narrow to an examination o f the relationships found to exist between alcohol use and depression in studies using both clinical and community samples. Theoretical perspectives on alcohol problems and depression from the biological, psychological and social domains will be investigated. Finally, psychosocial predictor and outcome variables under investigation will be briefly explained and their hypothesized relationships with alcoholism and depression explored Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Substance Abuse and Mental Disorders Understanding the complex co-existence o f psychological disturbance and disorders o f substance use and abuse is o f great import in contemporary society. In both research and practice, this status is commonly termed comorbidity and indicates the existence o f dual or multiple diagnoses, taking many forms and encompassing a wide range o f diagnostic categories. The term comorbidity can include virtually any medical or psychiatric disorder or diagnosis as well as any chemical dependency or abuse classification (Merikangas & Gelemter, 1990). People w ith comorbid disorders show a great deal o f measurable variability, including: “ ... number o f and types o f diagnoses, severity o f substance abuse and extent o f psychiatric impairment, the number and types o f psychosocial problems, availability o f social support systems, levels o f motivation, and personal strengths” (Daley, Moss & Campbell, 1993). This variability in presentation has, to some extent, clouded the understanding o f the antecedents, courses and outcomes o f co-occurring disorders. The extent o f the conditions is documented below as well as consequences affecting individuals, families, work and social structures. Diagnostic issues with co-morbiditv. In understanding the relationships between substance abuse or dependency and mental disorders, it is important to be aware o f the concepts o f primary and secondary disorder. Determining the order o f occurrence o f the mental disorder and the substance abuse or dependency disorder will establish which is primary and which is secondary. “The rational for using the primary-secondary concept involves improved prediction o f familial clustering o f the psychiatric disorder, implications for treatment, and improved outcome prediction” (Cook & Winokur, 1995). Various implications stem firom this primary-secondary distinction. Assessment and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 treatment o f dual-disordered individuals requires a encompassing approach in which the clinician needs to be aware o f three salient concepts: 1) the individual may be using a substance to self-medicate a psychiatric disorder; 2) the individual may be experiencing psychiatric symptoms as a result o f substance use; and, 3) the individual may be experiencing psychiatric symptoms and using substances independently (Spar, 1996). Further, Daley, Moss, and Campbell, (1993) suggest three conceptual categories or subgroups for patients with dual diagnosis in order to better understand and effectively address their treatment issues and needs: 1) Patients who may suffer from a primary mental disorder who also meet criteria for chemical dependency or whose periodic substance use causes problems which warrant treatm ent This group includes patients with both chronic mental illnesses as well as those with acute episodes. 2) Patients who may suffer from a primary chemical dependency disorder who also experience psychiatric problems. 3) Patients whose histories are so complex and intertwined that it is difficult to determine which diagnosis is the primary diagnosis. This third type o f patient often exhibits severe problems caused or exacerbated by either the mental disorder or the chemical dependency disorder or both. Examining people with dual disorders longitudinally allows for a determination o f the tem poral sequencing o f the disorders, to better establish the primary classificatiorL Some o f the diagnostic issues raised when assessing substance-abusing individuals who have psychiatric disorders are addressed by McKenna and Ross (1994). They conclude that the occurrence o r presence o f any number o f the following seven variables increases the risk o f an individual developing a dual disorder: 1) positive history o f sexual abuse, 2) emotionally m otivated substance use rationale, 3) symptoms Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 during sobriety, 4) anxiety-panic during sobriety, 5) length o f sobriety periods (less than one year associated with dual diagnosis), 6) age o f onset o f substance dependence (later age associated w ith dual diagnosis), 7) use o f four or more different substances. These seven factors can be used to help identify individuals who are at potential risk for developing dual disorders. Spar (1996) adds that childhood sexual abuse is found in 49 percent o f substance abusing women and 24 percent o f men; and that childhood physical abuse is found in 33 percent o f substance abusing women and 20 percent o f substance abusing men. It is very likely that childhood physical abuse is also a risk factor for dual disordered patients. Epidemiological studies. The existence o f comorbid diagnoses or co-occurring disorders is commonly believed to be wide-spread, with distinctions by type o f disorder, diagnostic nomenclature, method o f measurement, and sample characteristics. Within this review, the focus will be on the co-occurrence o f substance abuse and psychological disturbance. This relationship has been examined in various ways using both clinical populations and the general population. Regier, Farmer, Rae, Locke, Keith, Judd and Goodwin (1990) summarize the data obtained from the National Institute o f Mental Health Epidemiological Catchment Area Study (EGA). This study obtained data from interviews with over 20,000 people at five sites across the U.S. using the NIMH Diagnostic Interview Schedule; a highly structured interview schedule used to assess alcohol, drug, or mental disorders. The sample was obtained from the general US community and institutional populations (households, mental hospitals, nursing homes, and prisons). This sample is considered highly representative as it was standardized to US census data on the basis o f age, sex, and ethnicity. Reiger et al. (1990) have shown that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. within this combined sample o f the general population and an institutionalized population about 33 percent o f people will, at some tim e in them life, have an alcohol, drug or mental disorder. Further, they found that among people with a diagnosis o f any mental disorder at some time in their lives, 22.3 percent had a history o f alcohol abuse- dependence. Further, among those with alcohol disorders, they found about 44 percent had at some time in their lives been diagnosed w ith at least one other mental disorder. A second large-scale epidemiological stucfy, the National Comorbidity Survey (NCS), conducted by Kessler, Nelson, McGonagle, Ediund, Frank and L eaf (1996) yields results showing higher prevalence rates than the NIM H Epidemiological Catchment Area Study just described- They examined comorbidity w ithin a sample o f people from the general population and found that between 41 percent and 65 percent o f respondents with a lifetime history o f an addictive disorder also have had at least one mental disorder. Further, about 50 percent o f people with a lifetime history o f a mental disorder have also had a history o f at least one addictive disorder. It can be concluded that a significant portion o f people within both general and clinical population samples will, at some time in their lives, experience some type o f psychological disturbance or episode o f substance abuse, many of which are co occurring. It should be noted that the ECA study incorporated institutionalized populations and obtained lower rates, while the NCS stucfy excluded such populations. Nevertheless, the NCS study obtained higher rates. Methods o f assessment may, in part, influence the outcome rates and conclusions o f epidemiological studies. The studies described above both used interviews, but w ith differing schedules. This lack o f standardization in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 assessment procedures within differing studies may contribute to the disparate results obtained in these and other studies. The temporal relationship between co-occurring disorders is not clear firom the studies described. W hen answering the question about the lifetime history o f a given disorder, they do not adequately address whether the disorders, in fact, existed at the same time, which may be the primary or secondary disorder and the contributing fitctors o f one that may influence the other. The longitudinal research conducted in this investigation will further illuminate the temporal relationships o f the conditions under study, going beyond prevalence rate to examine behaviors that predictor and consequences o f alcoholism and depression. Prevalence o f and Diagnostic Categories o f Alcoholism and Depression Within contemporary research, the co-occurrence o f mood disorders with substance abuse is generally thought to be among the more prevalent disorders in the dual-disordered population (Brown & Schuckit, 1988; Cook & Winokur, 1995; Hesselbrock, Hesselbrock & Workman-Daniels, 1986). According to the American Psychiatric Association (1994) the majority o f adults (approaching 90 percent) at some time in their lives have consumed alcohol, and about 60 percent o f males and 30 percent o f females have had one or more adverse alcohol-related life events. People who excessively use alcohol can fall into two broad diagnostic categories within the D SM IV taxonomy, alcohol abuse and alcohol dependence. Alcohol abuse typically involves failure to fulfill major role obligations due to drinking while dependence, in addition, involves physiological tolerance and withdrawal, having a substantial effect on functioning. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Within the existing diagnostic nomenclature, people who are experiencing dysphoria can meet the criteria for somewhat different disorders based on the severity, length o f depressive experience, as well as associated and related features. The main diagnostic categories are M ajor Depressive Disorder, Dysthymic Disorder, Schizoaffective Disorder, M ood Disorder due to a General M edical Condition, Substance-Induced Mood Disorder, Dementia, and the depressive cycle o f a Bipolar Disorder. The common thread is incidence o f dysphoric mood. The prevalence rates o f the disorders vary. The lifetim e prevalence o f M ajor Depressive Disorder within a community population ranges between 10 to 25 percent for women and 5 to 12 percent for men (American Psychiatric Association, 1994). Lifetime prevalence for Dysthymia is about 6 percent. Prevalence data for Schizoaffective Disorder, Mood Disorder due to a General Medical Condition, Substance-Induced Mood Disorder, Dementia, and the depressive cycle o f a Bipolar D isorder are variable and/or lacking. Major problems exist when one adopts a dichotomous, either/or perspective when attempting to place people within arbitrary diagnostic categories. As will be discussed later, a great deal o f variabflity in the literature exists, in part, because o f varying definitions and arbitrarily assigned criterion points adopted by researchers; this study uses continuous measures to address this issue. Relationships Between Alcoholism and Depression Differing results have been reported about the relationship between depression and alcoholism. This may be, in part, due to the fact that it is difBcult to disentangle features that are unique to each disorder. Symptoms that are the result o f chronic alcohol use are often similar to those found in depressive individuals (Schuckit, 1986; Schuckit, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1979). Given this, some have argued that alcoholism and depression are two manifestations o f the same disorder. However, evidence to the contrary exists within family, twin and adoptive studies (NIAAA, 1991). A study by M aier and M erikangas (1996) comparing a clinical population w ith community controls highlights this. They found that these two disorders segregate independently in the families studied and argue that the strong levels o f comorbidity between alcoholism and depression are due to a causal, intra-individual relationship between the two disorders. In other words, the presence o f alcoholism may cause depressive symptomatology or the reverse, but the familial transmissions o f the two are independent. Schuckit (1986) also found that while both alcoholism and affective disorders have a degree o f genetic or familial transmission, they are two distinct disorders with different prognoses and treatments. He adds that depressive symptomatology is likely to develop in the course o f alcoholism and that drinking is likely to increase during the course o f an affective disorder, with about 5-10 percent meeting the criteria for secondary alcoholism. In a small subsample o f male college students with co-existing depression and alcoholism. Vaillant (1995,1993) asserts that alcoholism is rarely the result o f depression, but it is often a m ajor causative factor. Kasch and Klein (1996), also reviewing the ECA study data, found that an earlier age at onset is associated with increased comorbidity in m ajor depression and alcoholism. Clinical studies. In clinical samples, the frequency o f occurrence o f m ajor depression in hospitalized alcoholics has been reported to range from 8 to 53 percent (Merikangas & Gelem ter, 1990). This wide span and relative lack of consistency, in part. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 reflect the variatioa in the assessment criteria, tim e o f measurement and sample characteristics. This disparity is evident in the following review. A range o f fi-equencies has been obtained from studies that examined occurrence o f depressive symptomatology in inpatient alcoholics, hi a clinical sample o f male alcoholics. Brown and Schuckit (1988) found that 42 percent had clinically significant levels o f depression at admission to inpatient alcoholic treatm ent h i a sam ple o f fomale alcoholics, Nunes, Quitkin, and Berman (1988) found a frequency o f 36 percent Other frequencies o f depressive symptoms in alcoholics observed in inpatient settings are 16 percent (Herz, Volicer, D Angelo, & Gadish, 1990) and 32 percent (Dorus, Kennedy, Gibbons & Ravi, 1987). Roy (1996) investigated the clinical characteristics and presentations o f depressed alcoholics. He compared men with a diagnosis o f alcoholism and depression with male alcohohcs who never had been depressed and found those with both alcoholism and depression varied from the control group in a number of significant ways, suggesting a poorer clinical presentation in depressed alcoholic men. Also, depressed alcoholics had experienced more life events with negative impact, more negative events caused by alcohol and more suicidal behavior than control alcoholics; in addition, they had a family history o f depression and suicidal behavior. Hesselbrock et al. (1986) found that the onset o f alcoholism occurred at a similar age for patients with and without a history o f major depression, suggesting that it may not affect the developmental o f alcoholism. However, they did find that both male and female alcoholics with a history o f m ajor depression more frequently reported drinking to forget their worries or escape their problems (relieve depressive symptoms) than non-depressed male or female alcoholics. They reported that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. II although there were a large proportion o f male alcoholics in their sample, gender did not have a differential effect o f m ajor depression o f the course o f alcoholism for either sex. Concerns about clinical studies. A m ajor concern about the overall state o f existing research in this area is raised by the National Institute on Alcohol Abuse and Alcoholism (NIAAA, 1991). They indicate that most research on comorbidity has been conducted on or included people within a clinical population. Clinical samples may have unique characteristics because people with multiple ailments may tend to seek treatment at a higher rate than the general population. Consequently, this may increase the likelihood that estimates about the nature or prevalence o f co-occurring disorders in the community population based on clinical samples may not be accurate. Part o f the variability in studies o f the comorbidity o f alcoholism and depression is the result o f differing methods o f assessing both alcoholism and the depressive symptomatology (Cook & Winkur, 1995). Keeler (1979) has shown that rates o f depression co-occurring with alcoholism range from 8 to 44 percent depending on the type o f assessment used (clinical interview, structured interview, self-report inventory); Merikangas and Gelemter (1990) report ranges from 8 to 53 percent Cook and Winokur (1995) also suggest that variabihty in the timing o f assessments can produce differing results as alcohol intake and subsequent detoxification can induce psychiatric symptomatology. Based on this, it is reasonable to assume that a subject assessed for depressive symptomatology shortly after consuming alcohol or discontinuing use, will have results that are different from those o f a subject measured after a period o f sobriety. This point is illustrated in Brown and Schuckit's (1988) study Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 where depressive symptomatology was found to abate rather quickly over the course o f treatm ent Clinically significant levels o f depressive symptomatology decreased from 42 percent on admission to 12 percent in week two and to 6 percent four weeks later. The issue o f diagnostic confusion between alcoholism and affective disorder is addressed by Schuckit (1986), who concludes that the following factors contribute to this problem: 1) alcohol can cause depressive symptoms, 2) signs o f temporary serious depression can follow prolonged drinking, 3) drinking can escalate during primary affective episodes, 4) depressive symptoms and alcohol problems occur in other psychiatric disorders, and, 5) a small portion o f patients have indepencfent alcoholism and affective disorder. The confusion and lack o f consistency in defining the differences between sadness and depression and between drinking and alcoholism as well as the differences in the primary and secondary disorder distinctions, further clouds the issue. It is difficult to disentangle the contributions o f each disorder with respect to levels o f functioning within an individual. Communitv studies. The co-occurrence o f alcoholism and depression in community studies will be examined using studies involving adults and one study involving adolescents. In a sample from New Haven, Connecticut, Weissman and Meyers (1980) found that 71 percent o f people with a lifetime history o f alcoholism also met criteria for a lifetim e history o f major depression, while 15.4 percent o f people with a current diagnosis o f alcoholism also met the criteria for a current diagnosis o f major depressioiL Helzer et al. (1988) reported sim ilar findings utilizing the ECA data. They found that 78 percent o f the primary alcoholic males had a history o f major depression and 44 percent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13 o f the primary alcoholic females had a history o f major depression. Helzer and Pryzbeck (1987), again utilizing the ECA data, found a greater proportion o f people with a history o f alcohol dependence and/or abuse to have had a history o f major depression. They found that male alcoholics were slightly more likely than the overall ECA population to have had a diagnosis o f depression (5 and 3 percent respectively), where female alcoholics were much more likely to have had a diagnosis o f depression than the general population (19 and 7 percent respectively). Further, t h ^ indicate that, for men, alcoholism often precedes the onset o f depression (78 percent o f men studied), while in women, depression often precedes the alcoholism (66 percent o f women studied). For both men and women, the alcoholism is less severe when depression precedes alcoholism when compared to the opposite. Schuckit (1994) asserts that individuals with major depressive disorder are likely to develop alcohol dependence at a rate sim ilar to the general population but alcoholics are more likely than the general population to demonstrate severe depressive episodes. In a longitudinal study o f male college students. Vaillant (1995, 1993) found that those who abused alcohol were five tim es more likely to report being severely depressed than those who did n o t Vaillant (1995) states in about 28 percent o f this subgroup o f his sample (4 out o f 14) the reported depressive symptomatology could be entirely explained by the alcoholism. In about 43 percent (N=6) o f this subgroup, the first episode o f major depression occurred an average o f 12 years after the criteria for alcohol abuse had been m et Finally, in about 28 percent (N=4) o f this subgroup, the m ajor depressive disorder preceded the alcoholism. V aillants assertions support other views in the literature (e.g.. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 Schuckitt, 1986) and are supported by his longitudinal data, but it is somewhat disconcerting that they are based on such a small sample size. The association between (kpression and alcoholism in a com m unia sample o f adolescents, aged 18-19 years was studied by Deykin, Levy and W ells (1987). They found that o f those with alcohol abuse, 22.8 percent also had major depression, compared with those who did not abuse alcohol, only 6.9 percent had m ajor depression. Further, they found that it was more common for m ajor depression to be an antecedent o f alcoholism. Although a consistent and significant relationship between alcoholism and depression can be found in studies involving community samples, it tends to be lower in magnitude than the relationship found in clinical samples (Merikangas & Gelemter, 1990). This may, in part, be due to the fact that there is an increased likelihood for people suffering fiom these two conditions to seek treatment and thus have a higher level o f representation in clinical samples. In sum, it is clear that alcoholism and depression are related. For the purpose of this research, the relationship between alcohol use and depression will be conceptualized and operationalized using continuous measures. Continuous measurement will allow for a more complete understanding o f the relationships between these variables rather than examining people using an arbitrary cut-off point. Structural Equation Modeling will be used to examine the complex interrelationships among measured variables and the latent constructs. In short, rather than deciding that people are depressed or alcoholics when they reach X point, the continua o f alcohol use and dysphoric mood w ill be examined, using a statistical methodology that will allow for the complexities described. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15 Theoretical Perspectives For at least the past 150 years clinicians and researchers have developed theories in an effort to classify and understand alcoholics. Multiple theories have been advanced with difièrent perspectives, as people differ in their personal and drinking-related characteristics. In his review, Barbor (1996) asserts that typology theories can be divided into three distinct periods, the prescientifîc period o f clinical speculation (1850-1940), the Jellinek era o f review and synthesis (1941-1960) and the post-Jellinek period o f increasingly sophisticated empiricism (1960-present). The prescientifîc period includes work from William Carpenter, including his 1850 essay entitled “On the Use and Abuse o f Alcoholic Liquors in Health and Disease” where he describes three different types o f “oinomama” or wine manias, acute, periodic, and chronic (as cited in Barbor, 1996). Acute was described as sudden urges to drink with irregular patterns o f intoxication; periodic described as progressive binge drinking; and, chronic as a preoccupation and overwhelming desire for regular alcohol consumption. Jellinek (1960) identified five major subtypes o f alcoholics, based on etiologic determinants, alcohol process elements and damage elements: alpha, beta, gamma, delta and epsilon (as cited in Barbor, 1996). Two o f these types represent levels sufficient to be considered disease subtypes: the gamma alcoholic who ofien loses control and drinks because o f psychological factors; and, the delta alcoholic who can abstain if needed, but drink primarily because o f social and economic influences. Typological classification systems in the post-Jellinek period have been based more on research and clinical empiricism than the earlier period’s characteristic speculation and intuition. Criteria have included gender, family history, coexisting Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 psychopathology, age at onset, symptom developm ent and degree o f dependence. M ajor systems include those by Cloninger (1987), Zucker (1987), Barbor et al. (1992) and Del Boca (1994). Two types o f alcoholism were differentiated by Cloninger (1987), Type I o r “m ilieu limited” and Type H or “male limited”. Type I alcoholics typically have a later age at onset, develop psychological dependence rather than physiological, have a low need for novelty seeking and tend to avoid harmful events. In contrast. Type II alcoholics often have an age o f onset before age 25, are unable o r have difSculty in abstinence, have a high need for novelty seeking, low harm avoidance, and are more likely to be male. Zucker (1987,1994) developed a typological system with four different types o f alcoholism, antisocial, developmentally cumulative, negative affect and developmentally limited. Antisocial alcoholism is thought to have a genetic basis and poor outcome, it is characterized by an early onset o f alcohol problems and antisocial behavior. This type is typically seen in males. Developmentally cumulative alcoholism is initially limited and is culturally normative. For both males and females (m ore frequently male), over their life course, the consumption elevates to a dependent level. Negative affect alcoholism occurs primarily in women and is characterized by using alcohol for mood regulation. It has been linked to a fomily history o f unipolar depression and those with it have a higher suicide risk. Typically, there is a later onset o f symptoms o f alcoholism with comorbid depressive and anxious symptoms. Developmentally lim ited alcoholism occurs in both males and females and is characterized by heavy drinking in late adolescence, which decreases to socially acceptable levels as the individual matures and adopts adult roles. M uch o f Zucker’s recent work emphasizes the use o f tw o major types o f alcoholism. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 antisocial and non-antisocial (particularly in men). He asserts that the presence or absence o f antisocial symptoms is “a major distinguishing feature in virtually all o f the alcoholism subtyping schemes developed during the past generation” (Zucker, Ellis, Bingham & Fitzgerald, 1996, p.47). Offering another typology, Barbor et al. (1992) differentiated between Type A and Type B alcoholics. In comparison to Type A alcoholics. Type B alcoholics have an earlier onset, increased childhood risk factors such as conduct disorder and ADD, a higher degree o f familial alcoholism, more current psychopathology, greater severity o f dependence, more polydrug abuse and a more chronic treatm ent history. Reanalyzing the data from the Barbor (1992) study described above, Del Boca (1994) developed a four-group typology. This system o f classifying alcoholics was comprised o f a low risk-low severity group, an intem alizer group, an extemalizer group, and a high risk-high severity group. The low risk-low severity group is as its name describes. Comorbid drug use, conduct problems and antisocial personality characterized the high risk-high severity group. There was a significant gender effect noted between the intemalizer and extemalizer groups. The intemalizer group received this name because o f the way its members expressed their feelings and responded to their environments, and was comprised o f significantly more women than men. Members o f this group reported symptoms o f depression and anxiety, engaged in relief drinking, and were severely alcohol dependent with consequent physical problems. Members o f the fiaurth group, the extemalizer group, were mainly male. This group reported high levels o f alcohol use, experienced social consequences from alcohol use, and antisocial personality (Del Boca & Hesselbrock, 1996). As noted, there are significant gender differew es between the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18 intemalizer and extem alizer group. The group whose membership was mainly female was characterized by mood and anxious symptoms and highly dependent on alcohol. This intemalizing group may share sim ilar characteristics with Zucker’s (1987) negative affect alcoholics. Depression and Alcoholism: Biological. Psychological, and Social Theories Theories postulated to explain the etiology o f alcoholism and/or depression have been developed from the biological, psychological and social perspectives. As knowledge has developed, awareness o f complex and multiple-causation pathways has become apparent In this review, current research and theory support the concept that no one single causative factor wül be found to be a necessary and sufficient cause. A variety o f interrelated biopsychosocial factors may predispose or increase risks for an individual (Zucker & Gomberg, 1986). A concem is that in prior research and theory, there may have been excessive emphasis on an either/or binary approach borrowed from the natural sciences, while a broader multi-causal or systems perspective would be more appropriate. It appears that there are multiple etiological factors involved as well as various combinations o f factors, so that explanation may always ultimately be a matter o f the individual case. At the same time, various factors are proposed to contribute or account for the development o f alcoholism and/or depression. It is within this framework that this section proceeds. A selective review o f the literature follows investigating these issues and helping to support the idea that, while no single factor is responsible for the development o f alcoholism and/or depression, using multiple perspectives has m erit Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 Biological Factors Psvchophannacologv o f alcohol. The prim aiy pharmacological efifect o f alcohol is graded, reversible central nervous system depression (Julien, 1996). Salient to this investigation, the pharmacology o f alcohol may substantially contribute to the secondary depression commonly found in alcoholics. Genetic factors in alcoholism. Researchers have demonstrated a genetic or familial link to alcoholism. Cotton (1979) conducted a review o f 39 studies and concluded that alcoholics were approximately five tim es more likely to have an alcoholic relative than non-alcoholics. The DSM-IV confirms this but gives a slightly low er rate, suggesting that the risk for alcohol dependence is three to four times higher in close relatives o f people with alcohol dependence. This risk increases as the number o f afflicted close relatives increases, with closer genetic relationships and with m ore severe problems exhibited in the affected relative (American Psychological Association, 1994). The discovery o f a more specific genetic effect would lead to the identification o f some people at risk, while at the same time, bring a better understanding o f the genetic influences. Because not all persons with a genetic predisposition become alcoholic, a more thorough understanding o f genetic effects would assist in understanding the role o f environmental influences. It may be that genetic factors need to be viewed as vulnerability to alcoholism. A significant number o f large-scale tw in studies, comparing monozygotic twins o f alcoholics with dizygotic twins o f alcoholics and also with the general population have been conducted. Heath (1995) reviews adoption and twin studies and concludes that evidence indicates an important genetic influence on alcoholism risk, which is consistent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 0 across different studies despite using different research methodologies. Specifically, data based on U.S. samples suggest that genetics account for about 60% o f the variance in alcoholism risk in both m en and women, and that tw in studies suggest that there is some effect o f family environm ent Adoption studies based on U.S. samples suggest that an adoptee's family environm ent may account for about 33% o f the variance in alcoholism risk. He suggests that data fiom Scandinavian studies yield a slightly lower estimate o f explained variance by genetic influences o f 39% and a variance o f 15% for family environmental influences. Kaij conducted one o f the early studies during the 1950s in Sweden and found a higher percentage o f m on^gotic (61 percent) than dizygotic (39 percent) twins to have demonstrated problems with alcohol (Kaij cited in Heath, 1995). Clearly, there appears to be a strong genetic link. Risk may be greater in male offspring o f alcoholic fothers (Reich, Cloninger, Van Eerdewegh, Rice & Mullaney, 1988). Schuckit (1995) reports on data from a series o f studies between 1970 and 1995, investigating the idea that sons o f alcoholics would have a reduced response to a controlled dose o f alcohol, predisposing them to develop alcohol- related problems. In a ten-year follow up study, over h a lf (56%) o f the subjects with a positive family history for alcoholism , whose reactions to alcohol were low when tested at time one (demonstrated little impairment), had developed alcoholism, compared with 14% o f subjects whose reactions were high. Although there is some disagreement in the literature (O'Malley & Maisto, 1985), the author suggests that level o f response to alcohol may mediate the effects o f fam ilial transmission. Although it is clearly difficult to disentangle the unique contributions o f genetics and environmental influences, there appears to be a significant genetic relationship. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 Neuropsychology o f depression. Studies o f the neuropsychology o f depression baye yielded contradictory results. Much o f the research has focused on mature adults and/or elderly indiyiduals, as they are likely to present complex neurocognitiye differential diagnostic questions, requiring neuropsychological assessm ent There is contradictory eyidence in many areas o f cognitiye functioning for depressed indiyiduals. Some haye argued that global deterioration is associated with depression (Siegel, Gureyich & Oxenkrug, 1989; Siegffied, 1985), others argue against it (Williams, Little, Scales & Blockman, 1987; Popkin, Gallagher, Thompson & Moore, 1982). Howeyer, Boone, Lesser, Miller, W ohl, Berman, Lee, Palmer and Back (1995) concluded that there are relatiyely distinct cognitiye profiles, corresponding to the presence and seyerity o f depressioit They found the presence o f depression to be manifest by declines in nonyerbal skills (Performance IQ, yisual memory), while increasing seyerity o f depression was associated with declines in information processing speed and executiye skills. They indicate that these declines are typically subtle and generally not o f clinical releyance. Further, they also argue that this pattern is consistent with that found in younger depressiye patients. Boone et al. (1995) hypothesize that there are two conditions necessary for the emergence o f a depressiye episode. First, there needs to be an abnormal organization of the right hemisphere, which constitutes a predisposition for depression. The right hemisphere (especially temporal) contains the neural substrate for emotion (Gainotti, Caltagirone & Zoccolotti, 1993). Second, there needs to be an interruption in the left hemisphere mechanism found in the ftontal subcortical region that controls affect (Gainotti, Caltagirone & Zoccolotti, 1993). In sum, Boone et al. (1995) hypothesized that. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 2 in order for an individual to experience a depressive episode, there needs to be a pre existing cerebral abnormality in the right hemisphere that is normally inhibited by the left hemisphere. When something interrupts this control, a depressive episode is likely to develop. Overall, the state o f the literature in the neuropsychology o f depression is contradictory, due to a v aried o f reasons such as: 1) medication status was often not controlled, most studies have their patients on antidepressant medication; 2) concurrent medical illnesses were not controlled, many samples included patients who had a host o f different medical issues that could compromise the results; 3) history o f electro- convulsive therapy was not controlled for and ECT is known to be associated with neuropsychological deficits; 4) many studies fail to use adequate normative data, researchers have used the normative data published w ith the given tests as a control. Further, often when control groups have been used, dramatic age and education differences exist between patients and control groups (Boone, 1998). Nevertheless, a neuropsychological perspective offers one set o f possible explanations for depression. Psychological Factors Many psychological theories have explanations for why individuals develop depression and/or alcoholism. Several theories will selectively be reviewed. Cognitive theory o f depression. In this model, thought determines feelings and behavior. Psychological problems are the result o f faulty ways o f thinking and distorted attitudes toward oneself and others (schemas). According to Beck and Weishaar, (1989) systematic bias occurs in processing information in most psychological disorders. This bias applies to both external (views o f world/others) and internal (coming ftom within the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 body) information o r messages. Depressive individuals tend to have a negative bias and cognitive errors in their information processing (Beck, Rush, Shaw & Emery, 1979). They tend to have a negative view o f the self, the world, and the future (negative triad) and tend to perceive the self as inadequate, deserted and worthless (Beck & Weishaar, 1989). The negative view o f the self is manifost when the depressive person blames their setbacks and problem s on personal inadequacies without considering circumstantial explanations. They believe that they lack necessary qualities to bring them happiness. The second component, the negative view o f the world manifests in the depressive's tendency to interpret everyday experiences in a negative manner, selecting certain components o f experience that conform to their negative perspective. The third component, the negative view o f the future, reflects the depressive person's tendency to expect their present difficulties to continue, they tend to expect failure (Beck, 1987). Self-derogation theorv. The theory o f self-derogation, advanced by Kaplan, conceptualizes deviant behavior as resulting from a self-esteem motive (Kandel, 1980). According to this theory, people engage in deviant behavior in order to restore a sense of self that has been damaged by prior devaluing experiences. Thus, psychological distress resulting from a damaged sense o f self or poor self-esteem, may lead to deviant behavior. Alcoholism may occur in an environment that, although considered deviant, may provide the support the individual is lacking. Self-medication hvpothesis. It is postulated that substance abuse may be part o f an individual's response to emotional or psychological distress. In fact, some researchers argue that self-m edication is one o f the major reasons for the overuse o f and dependence Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 4 on substances (Khantzian, 1985). W ithm this framework, alcohol use may be hypothesized to ameliorate psychological distress or symptomatology. Social Factors Alcohol problems. Social and cultural contextual factors are associated with an increased risk o f developing alcoholism. Peele (1987) states that culturally imparted attitudes and beliefs are important determinants o f patterns o f consumption. He states that since the use o f alcohol is a norm for many groups of young people in the U.S., the norm will be followed. Vaillent and Hiller-Sturmhofel (1995) empirically validated the concept that alcohol use follows cultural or sub-cultural norms. In a longitudinal study begun between 1940 and 1944 they selected 456 men from Boston inner-city schools, the "core city sample", and have followed them since. They had a variety o f ethnic backgrounds, including Irish, Polish, Russian, Italian, Northern and Southern European; 61 percent o f their parents were immigrants. From this sample they were able to determine that the rate o f alcohol abuse and dependence were five times less common in m en o f Italian and other Southern European descent when compared with other ethnic groups. The authors suggest that part o f this difference may be due to cultural influences and beliefs about appropriate patterns o f usage. In Italian culture alcohol is regularly consumed with meals and introduced to children in a responsible manner; "moderation is encouraged, intoxication is proscribed" (p. 155). Hawkins et al. (1992) suggests the following social factors are among those that are associated with alcohol problems: 1) Extreme economic deprivation occurring concurrently with childhood behavior problems increases the risk for later alcoholism. 2) Neighborhood disorganization- specifically, neighborhoods with high population density. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 lack of natural surveillance o f public places, high residential mobility, physical deterioration, low levels o f attachment to neighborhood, high rates o f adult and juvenile crime. Neighborhood disorganization has been hypothesized to impair a parent's ability to foster prosocial values in their children including, appropriate usage o f alcohol. 3) Laws and norms favorable toward behavior affect usage patterns. Laws stating to whom alcohol may be sold and how it is to be sold help to determine who is legally able to obtain the substance. Taxation o f alcohol affects its price, in turn affecting consumption. Behavioral genetic perspective. Looking at the etiology o f alcoholism from the perspective that certain temperament characteristics act through an ongoing process o f reciprocal interaction with the social environment. Tarter and Vanyukov (1994) discuss the etiology o f alcoholism from a developmental behavior genetic perspective. From this view, alcoholism is a "multidimensional endpoint phenotype" preceded by a series o f "intermediary phenotypes" that are detectable early in life. They suggest that "temperament deviations in infants and young children negatively affect the quality o f the parent-child relationship so that the ensuing behavior disposition o f the child increases the risk for alcoholism" (p. 1096). Five temperament phenotypes are associated with the elevated risk for developing alcoholism. 1) Behavior activity level- a rapid tempo or high activity level is associated with an elevated risk for alcoholism (see also Zucker & Gomberg, 1986). W hile substantial portions o f male alcoholics have a childhood history o f hyperactivity, children with "pure" ADHD are not at a substantially increased risk o f developing alcoholism. Rather, when children who are hyperactive are also aggressive, the risk increases (Phil & Peterson, 1991). 2) Emotionality-the propensity to be easily and intensely aroused. Children o f alcoholics report greater emotional reactivity and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 6 display somatic and behavioral symptoms o f anxiety. High psychological and ptycho- physiological reactivity or high emotionality is a phenotype for alcoholism risk. 3) Soothability- people at risk for alcoholism have a decreased capacity to rapidly return to a baseline level following emotional arousal. 4) Attention Span-Persistence- attentional difficulties and goal impersistence are associated with an elevated risk for alcoholism; and, 5) Sociability- behavioral disinhibition, manifested by aggression or antisocial behavior (Zucker & Gomberg, 1986), sensation seeking, impulsivity and social nonconformity are related to an increased risk for alcoholism (Tarter & Vanyukov, 1994). Gender Differences in Alcoholics Much o f the theory and treatment o f alcoholism in women has been derived foom the study o f drinking in men (Patterson, 1995) and most research on alcoholism has traditionally focussed on the male alcoholic. This may be due to a variety o f factors. Men’s behavior may call more attention to their alcohol problems, especially in the past. This may have been partially because a large proportion o f women were historically not in the workforce, and therefore their problems with alcohol were more likely to be concealed within home and family. In addition, cultural expectations o f behavior for women did not include the excessive use o f alcohol; such attitudes further supported women's use o f alcohol in a hidden context. Following is an expanded, topical outline Grom Lex (1991) o f some o f the differences between male and female alcoholics commonly cited in the literature. I . Familial/genetic factors: Women alcoholics are more likely to have an alcoholic role- model in their nuclear families and to engage in assortive mating practices, choosing alcoholic spouses (Lex, 1991). There appears to be disagreement on genetic factors in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 7 the literature, with some sources seeing genetic family influences as m ost likely similar in male and female alcoholics; with others seeing differences in the heritability o f alcoholism (Svüds, Velez & Pickens, 1994). 2. Onset: Alcoholic women usually have drinking problems at later ages than men do (Jung, 1994; Lex, 1991). According to Lisanslqr-Gomberg (1994), mean age a t onset o f problem drinking for males is 27.0 years and for females it is 46.2 years. 3. Consumption patterns: Women typically consume less alcohol than men and are less likely to drink daily, to drink continuously, to drink heavily or to engage in binges (National Institute on Drug Abuse, 1997; Jung, 1994; Lex, 1991). 4. Course o f disease progression: Women tend to have a later onset o f alcoholism, but then move more rapidly through the progressive stages o f alcoholism than men (telescoping) (Orford & Keddie, 1985 as cited in Lex, 1991). 5. Attribution o f the etiology o f alcoholism: Women are more likely than men to be able to pinpoint a specific recent life event that they believe served as a trigger for their alcoholism (Jung, 1994; Lex, 1991). 6. Dual diagnoses: Alcoholic men tend to have comorbid diagnoses o f antisocial personality disorder, and alcoholic women tend to have comorbid diagnoses o f affective disorders (Regier et al, 1990). 7. Physiological differences: Women have lower levels o f gastric alcohol dehydrogenase than men (Julien, 1996). This enzyme is lower in women and is responsible for metabolizing alcohol. This means that if women and men consume the same amount o f alcohol, women will have higher blood alcohol concentrations. In fact, first-pass metabolism o f alcohol in women is about fif^ percent less than in men Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 (Julien, 1996). This may increase women’s vulnerability^ to acute alcohol intoxication and chronic, long-term effects o f alcohol consumption. Also, women have different body composition and body weight than men do (51% lean body water volume in women, 65% in men). Because women weigh less and have a higher proportion o f body fat to water, the same dose o f alcohol results in greater impairment for women than for m en (Lex, 1991). W omen’s absorption rates and blood alcohol levels are affected by hormonal changes. Absorption is influenced by progesterone levels, which have been found to fluctuate during the menstrual cycle, and throughout pregnancy (Jung 1994; Lex, 1991). 8. Social consequences o f alcoholism: Typically, the consequences o f alcohol use or abuse are different for women and men. Men are less stigmatized for alcohol use than women, and often face few er social consequences (Jung, 1994). Traditionally, men typically experience problems in jobs or career paths, where women typically experience disruptions in fam ily life (Beckman & Amaro, 1986 as cited in Lex, 1991). 9. Personal responses to illness; The typical alcoholic woman is characterized as feeling more guilty, anxious or depressed (Robbins, 1989 as cited in Lex, 1991) General risk factors for the female alcoholic. Certain specific risk factors occur throughout a woman’s life; others have stronger effects at specific ages. This section will describe risk factors that have been associated or correlated with female problem drinking behavior over the life span. The following section will describe risk factors associated with specific stages in the life span. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 Demographic characteristics such as age, ethnicity, employment and marital status can help predict drinking behavior in women. Younger women tend to drink heavier and experience more severe behavioral and social consequences than older women do (Jung, 1994). W hite women (or accuiturated minority group members) are more likely to consume more alcohol and experience more severe consequences from alcohol consumption than Hispanic and African-American women (Wilsnack, W ilsnack & Hiller-Strumhofel, 1994). Women who have multiple roles, such as family, marriage and paid employment may have reduced risks o f alcoholism. However, women who work in “nontraditional”, male-dominated occupations report increased drinking and/or adverse drinking consequences (LaRosa, 1990). Women who are never-married, divorced, and separated tend to have relatively heavy rates o f alcohol consumption. Widowed women tend to have lowest rates, and married women intermediate rates (Jung, 1994; W ilsnack etal., 1994). A history o f childhood sexual abuse or sexual dysfunctions or problems is a risk factor for alcoholism in women. For female drinkers, a positive history o f incest and/or other childhood sexual abuses has been shown to be correlated with alcoholism. (Wilsnack et al., 1994). Also, women experiencing sexual problems or dysfunctions are at increased risk for alcoholism. However, it is diflBcult to differentiate whether sexual dysfunction may be the cause o f drinking or conversely, whether the excessive drinking may have resulted in sexual dysfunction. Further, alcohol use by women can lead to engagement in high-risk sexual behavior and possible sexual victimization (Norris, 1994). Additionally, in the case o f dually diagnosed patients. Spar (1996) suggests that childhood sexual abuse is found in 49% o f dually diagnosed, substance abusing women Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 and 24% o f men; and that childhood physical abuse is found in 33% o f dually diagnosed substance abusing women and 20% o f men. Partner alcohol consumption/pressure to use has been found to be positively associated with alcohol consumption in women. Women may imitate the drinking behavior o f a “higher-status” male whether in the fomily or in the workplace (Lisansky- Gomberg, 1994; W ilsnack e ta l., 1994). Depression has been found to be related to problem drinking behaviors in women (Lisansky-Gomberg, 1994). Also, negative personality foctors (high alienation/low self esteem, negative affect) are stronger predictors o f female drinking patterns than o f male drinking patterns (W ouldt & Bradley, 1996). Styles o f coping with stressors which have been characterized as wishful thinking such as escape-avoidance and denial, may appear more frequently among female alcoholics than male alcoholics (Lisansky-Gomberg, 1994). Stage specific risk factors. In order to conceptualize more fully the unique problems, prevention and treatment needs o f the female alcoholic, it is necessary to understand which risk factors determine or are associated with alcoholism at various stages in the life course. In adolescence, various risk factors have been associated with female problem drinking behavior. Lisansky-Gomberg, (1994) found associations between peer alcohol use or pressure to use and alcohol use in females. Further, associations have been found with problems such as antisocial, aggressive or deviant behaviors including vandalism, temper tantrums, rejection o f authority, and heightened impulsivity. Frequent school absences, school problems and low educational aspirations have been found to be Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 associated with female problem drmking behavior (Newcomb, Maddahian, Skager & Bentler, 1987 cited in Lisanslqr-Gomberg, 1994). Early alcohol intoxication and early use o f marijuana have been found to be associated with problem drinking among young women. Also, individuals who have expectancy effects o f alcohol to be used as removmg inhibitions or obtaining a high have higher incidence o f problem drinking behavior (Lisansky-Gomberg, 1994). Among young-adult women a different set o f potential risk factors emerge. Young- adult women who have role-related issues, health-related issues, lifestyle patterns with other drug use, and those who are in their thirties are more susceptible to alcoholism. Role-related issues such as being in college and being in a “non-traditional”, male-dominated workplace have been found to be associated with increased alcohol consumption in young-adult women. Women with reproductive disorders who cannot become mothers may be at increased risk for problem drinking behavior. Women who use other drugs, including nicotine, are at higher risk for problem drinking behavior (Lisansky-Gomberg, 1994). Finally, during the decade o f the thirties, the drinking rates o f women and men tend to converge, and some research has found associations between women and problem drinking behavior and/or alcohol-related consequences during this decade o f life (Fillmore, 1987). For middle-aged women (40 to 59) certain fectors that are also risk factors for other ages may make more o f an impact during this time period. Lisansky-Gomberg (1994) suggest that the minimal likelihood o f acquiring new roles, new jobs or new interests, “empty nest” status, heavy spousal drinking and marital disruption, abuse o f Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 Other psychoactive drugs, and, comorbid psychiatric diagnosis are risk factors during this period. After age 59, the typical female drinks less, and more abstain from drinking (Jung, 1994; Lex, 1991; Lisansky-Gomberg, 1994). However, Lisansky-Gomberg (1994) suggests several risk factors for alcoholism in this female, older-adult age-group. These risk factors include: problems or stress associated with retirement; use o f psychoactive drugs; and, for some moving into retirement communities. Depression A variety o f social factors contribute to or mitigate the development o f depression. These fectors and/or influences have produced a number o f differing social theories o f depression and much of existing research on the social context o f depression have not been guided by one particular cohesive theory (Gotlib, 1992). The development o f depression has been linked to interpersonal rejection (Marcus & Nardone, 1992; Segrin & Dillard, 1992), reduced social support (Billings, Cronkite, & Moos, 1983), social comparison (Sheeran, Abrams & Orbell, 1995), stressful life events (Frank & Spanier, 1995) and self-verification (Joiner, Katz & Lew, 1997). Depression may be manifest differently in different cultural groups. While the literature on culturally relevant counseling seldom focuses specifically on depression, certain themes emerge. One such theme includes the issues that occur because o f immigration. Many iimnigrants experience intense feelings o f loss for the families and social networks left behind, as well feeling isolated and pressured to change to conform to their new surroundings (Garcia-Preto, 1996). Another theme surrounds issues o f culturally appropriate expression of self, especially the expression o f feelings. For Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 example, in the Asian population, restraint o f feelings is associated with wisdom and maturity (Sue & Sue, 1990), while "obligation and shame are the mechanisms that traditionally help to reinforce societal expectations and proper behavior" (Lee, 1996, p.230). A belief that Afiican-Americans do not become depressed has existed in the past both within this group and also, to some extent, within the scientific population (Paniagua, 1994), It is clear that the conditions and norms o f behavior held by particular cultural groups may influence the perception and manifestation o f depression; nevertheless, depression exists in all groups, although it may have other labels. Constructs Analyzing the Predictors and Consequences o f Alcohol Dysphoria Following is a description o f the constructs that will be utilized in this research. Predictors o f alcohol use and (fysphoria will be examined with the following latent constructs; social conformity, relationship satisfaction, perceived opportunity and family bonding. Outcome measures are the latent constructs o f marital and relationship satisfaction, perceived opportunity, job satisfaction and job stability; they will be conceptualized as consequences o f alcohol dysphoria. The measured variables that make up the latent constructs will be briefly described here, and specified in Chapter Two. Framing the proposed research within a developmental perspective is pertinent because o f the longitudinal nature o f the present investigation spanning several important developmental stages, using data obtained fiom individuals when they are adolescents, young adults and middle adults. Forrest, (1983) using a developmental perspective, suggests that certain childhood and adolescent characteristics may, in part, be predictive of subsequent alcoholism and perhaps o f a combination of alcoholic and depressive Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 4 symptomatology proposing that "most alcoholics have been unhappy during childhood. Flat or blunted affect, social withdrawal, self-preoccupation and the other clinical symptoms o f depression are central to the childhood adjustments o f many alcoholics" (p. 104). During adolescence, he states that although some experience direct depressive symptoms, alcoholics are ofien characterized as having been verbally and physically aggressive with an angry interpersonal demeanor in adolescence. This demeanor can lead to interpersonal rejection in friendship, romantic and 6m ilial relationships and subsequent depressive episodes. Also, as adults, alcoholics experience events and circumstances that may induce o r exacerbate depressive symptomatology. Forrest argues that marital dissatisfaction is higher in an alcoholic population than the general population and that alcoholics experience more vocational instability and dissatisfaction, difficulty in spiritual or religious matters and, with their own children, problematic or inadequate parenting practices. Given this perspective, the need to more fully explore underlying or latent concepts and how they may relate to the lives o f adults who are experiencing alcohol dysphoria is im portant The present investigation will utilize Structural Equation Modeling techniques to examine longitudinal data fiom early adolescence through young and middle adulthood. This procedure will allow for complex analyses o f the relationships between alcohol dysphoria and the latent constructs over time. Familv Bonding Family bonding is conceptualized as a predictor variable for alcohol use and dysphoria in this study. The latent construct consists o f two items, “Good relationship with parents”, and “Good relationship with family”. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 Aspects o f the environment within the family have been shown to affect alcohol use and depression in different ways. Parenting practices, levels o f conflict and bonding impact the use o f alcohol in offspring. M odeling usage patterns and im parting beliefs about consumption are other important contributions (Hawkins, Catalano & Nffller, 1992). Low bonding to family as characterized by parent-child interactions with a lack of closeness and a lack o f maternal involvement in activities impact substance initiation (Brook, Lukoff & Whiteman, 1980; Kandel et al., 1979 as cited in Hawkins et al., 1992). Adolescents from families characterized by feelings o f parental trust, warmth and involvement show decreased alcohol use (Hundleby & Mercer 1987, as cited in Hawkins et al., 1992). One specific risk factor fr>r alcohol use is poor relationship w ith parents (Newcomb, Maddahian & Bentler, 1986). Depression has been linked to decreased social frmctioning and fewer close relationships (Gotlib & Lee, 1989). Social Conformity Culturally and socially imparted attitudes and beliefs are im portant determinants o f patterns o f alcohol consumption (Peele 1987). Socially conforming attitudes are used as a predictor variable for alcohol use and dysphoria in this study. Conceptualized as “Social conformity”, the latent construct consists o f three multi-item scales reflecting law abidance, liberalism and religiosity. Conforming to societal norms has been shown to impact alcohol use and is seen as a risk factor for alcohol and drug problems (Peele, 1987; Newcomb, M addahian & Bentler, 1986). Low social conformity has been found to be correlated w ith deviant behavior, drug use (McGee & Newcomb, 1992), and alcoholism (Smith & Newman, 1990; Sovani, 1987). It has predicted hostility, decreased sense o f purpose in life. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 increased self-derogation and negative self-image (Newcomb, Schefer & Bentler, 1993). Low social conformity has been associated with increased depression and suicide attempts (Burch, 1994; Newcomb et al., 1993); and, has been shown to moderate intoxication and difGculty at work (Newcomb, 1995). Social conforming attitudes and behavior are important to consider in this investigation as t h ^ im pact alcohol use behaviors, psychological factors associated with depression and psychosocial function. Relationship Satisfaction A comprehensive examination o f how people function in relationships, involves several domains. Those addressed here are satisfaction with marital or intimate relationships, divorce, satisfaction with same-sex relationships, satisfaction with relationships with opposite-sex friends, and relationships with peers. Latent constructs o f marital and relationship satisfaction. M arital satisfaction will be examined using the Dyadic Adjustment Scale in adulthood. Relationship satisfaction will be examined with the following measured variables: "Satisfaction with close friends", the "Satisfaction with opposite-sex friends" item and the "Good relationship with peers", also incorporating incidence o f divorce. In adolescence, relationship satisfaction will serve as a predictor variable, in adulthood, an outcome variable. Alcohol problems and relationship satisfaction. The literature on marital satisfaction shows that it has been operationalized by examining levels o f marital satisfaction, family dysfunction, rates o f divorce, sexual relations and other relationship aspects. Among alcoholics, the relationship between alcohol use and relational quality has been well documented. High levels o f family dysfunction have been found to be related to alcohol use (McKay, Longabaugh, Beattie, Maisto & Noel, 1992). Alcoholic Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 families have been characterized by poor communication patterns» lack o f mutual warmth and support» spousal abuse and poor role functioning. N ot surprisingly, spouses o f alcoholics expressed greater dissatisfaction in all the areas o f fam ily functioning, than alcoholics (Suman & Nagalakshmi, 1995). Couples with an alcoholic husband have been found to demonstrate increased relationship distress on measures o f marital stability, change desired, and positive communication behaviors. (O'FarreU & Birchler, 1987). Marital dissatisfaction and divorce rates have been shown to be as much as seven times greater in alcoholics than the general population (Schafi, Lavely Sc Jafife, 1975). On sexual measures, alcoholic couples report lower frequency o f intercourse, more change desired in intercourse frequency, greater misperception about the amount o f change in sex frequency desired by their mate, and a higher frequency o f disagreement about sex. (OTarrell, Choquette & Birchler, 1991). Noting a gender effect, N oel, McCrady, Stout and Fisher-Nelson (1991) found that couples with an alcoholic husband were more dissatisfied with each other and themselves in their role functioning than couples with an alcoholic wife. In addition, they found that alcoholic wives engaged in more positive communication with their husbands, and alcoholic husbands were m ore negative toward their wives. On the other hand, differing qualities o f alcohol consumption may differentially affect relational satisfaction. Under some circumstances, alcohol use has been associated with increased marital satisfaction; suggesting that the relationship between these two variables may not be entirely linear. Jacob, Dunn and Leonard (1983) found that high alcohol consumption was associated with high marital satisfaction and reduced symptomatology in the spouses o f steady drinkers, but not binge drinkers. In some cases Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 alcohol use may have adaptive consequences for the marriage and family life o f some alcoholics, serving a protective role (see also Steinglass, 1981). Social support and personal relationships have been shown to be related to alcohol use (Newcomb & Bentler, 1986). Medora and W oodward (1991) studied loneliness in alcoholics, finding significant negative relationships between the number o f years alcohol was consumed, self-esteem, self-rated m arital satisfaction, self-rated job satisfaction and loneliness. Also, they found differences between the loneliness scores o f subjects who had indicated varying degrees o f happiness during the previous year, and betw een men and women (women were significantly lonelier than men were). Depression and relationship satisfaction. Depression has been shown to influence the quality and satisfaction o f relationships. Research has shown that people involved in committed relationships who are experiencing depressive symptoms are at risk for dissatisfaction in their relationships. About 50 percent o f depressed, married individuals have been shown to experience distress their marriages. (Rounsaville, W eissman, Prusoff & Herceg-Barton, 1979). Degree o f m arital distress has been shown to be related to the level of depression (Schmaling & Becker, 1991). Couples with depressed members perceive marital interactions to be more hostile, less friendly, and more dom inated by their partners than did those who were not depressed (McCabe & Gotlib, 1993). People who are prone to depression may be likely to marry partners whose personalities are not conducive to marital satisfaction or positive marital relations (Brown, Bifulco, Harris & Bridge, 1986). In a sample o f both male and female university students, those who exhibited dysphoric symptoms responded that both they and their pariners were behaving Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 more coldly and that they perceived their relationships to be less complementary (Thompson, W hiffen, & Blain, 1995). D ifferential gender effects have been noted. Married women as a group have been shown to be at high risk for depression when they are dissatisfied with their marriages and unable to confide in their husbands (Weissman, 1987). Dysphoric mood is related to low rates o f pleasing and high rates o f displeasing marital exchanges and to marital dissatisfaction in women (Assh & Byers, 1996). McCabe and Gotlib (1993) found couples, but particularly wives who are depressed, perceived their family life to be more negative than those who were not depressed. Further, depressed wives displayed increased negative verbal behavior over the course o f interactions. Across relational domains, Barnet and Gotlib (1988) suggest that dependency and introversion are stable trait characteristics o f dejx'essives. However, in interpersonal relationships people tend to display behaviors and traits in varying styles. Depressed people do not consistently present themselves as depressed or negative, nor do they consistently m aintain negative cognitions (Gotlib, 1983); presentation o f self tends to be more situational in nature (Meyer & Hokanson, 1985). Overall, depressed individuals report having sm aller and less supportive social networks and are less skillful in interpersonal interactions (Gotlib, 1992). This is supported by findings that suggest the frequency o f social contacts and dissatisfaction with one's friends is related to depression (Rosenbaiun, Lewinsohn & Gotlib, 1996; Gotlib & Lee, 1989). Having little or no social support system leads to a much higher risk for depression as social support is known to mitigate the effects o f negative life stressors (Maxmen & Ward, 1995). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 It is apparent that both alcoholism and depression are related to satisfaction with relationships. The measured variables and constructed âcto rs (Dyadic Adjustment Scale, Divorce, "Satisfaction with close ftiends", "Satisfaction with opposite-sex friends", and "Good relationships with peers") allow for a considerable exploration o f the relationships involved. Overall, across the studies reviewed, alcoholics and depressives have lower levels o f satisfaction than those without the given sym ptom atolo^. Satisfaction, social support and personal relationships are related to alcohol use and depression. The predictor variable o f relationship satisfaction should allow for an exploration of these effects longitudinally. In terms o f outcome variables, it is reasonable to assume that combining alcoholism and depression into a latent construct o f alcohol dysphoria may potentially show an exacerbation o f the relationships described above and alcohol dysphoria would be significantly related to decreased levels o f satisfaction in differing relational domains. Perceived Opportunity Latent construct o f perceived opportunity Perceived opportunity is conceptualized as reflecting an individual's future orientation. In this research, the construct o f perceived opportunity will consist o f the following measured variables: "Satisfaction with future opportunities" item, "Satisfaction w ith school and work" item, and "Satisfaction with being what you want to be" item. In adolescence, it will be used as a predictor variable, in adulthood, an outcome variable. Alcohol problems, depression and perceived opportunity. Although conceptualized differently by different researchers, perceived opportunity, agency, future orientation and outlook on life have been found to be associated with alcohol problems Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41 and depression. This construct has also been examined as it relates to variables that may mitigate alcohol use and/or depression such as selfesteem (Stein, Newcomb & Bentler, 1992), quality o f life (Russo, Roy-Byme, Reeder & Alexander et al., 1997) and anhedonia (Schrader, 1997). Negative future orientation and hopelessness are associated with suicidal ideation (Beck, Rush, Shaw & Emery, 1979) and suicidal behavior (Petrie & Chamberlain, 1983). People suffering from both depression and alcoholism suffer an additive or synergistic effect o f two separate disorders, resulting in a disproportionately high level o f suicidality (Cornelius, Salloum, Mezzich, Cornelius, Fabrega, Ehler, Ulrich, Thase & Mann, 1995). Stein, Newcomb and Bentler, (1992) found perceived opportunity to be related to levels o f self-esteem with a differential gender effect For men, they found the opportunity for personal achievement and self-fulfillm ent is important in their future sense o f self-esteem and self-evaluation. In women, future selfesteem and self- evaluation was related to the perception o f good interpersonal relationships. Having a future temporal orientation was found by Lennings (1996) to be a significant predictor o f long-term abstinence in a study o f severely dependent alcoholics. Amodeo, Kurtz and Cutter, (1992) found purpose in life and life satisfaction to be important measures o f treatment progress in alcoholics. Consistent with the tenets o f Beck's cognitive theory o f depression (Beck, Rush, Shaw & Emery, 1979), depressed individuals tend to make long-range projections o f their current difficulties. Expecting that the currently experienced feelings will persist, depressed individuals tend to have a negative view o f the future. Perceived opportunity is related to depression and alcohol use. W hen measured in adolescence, perceived opportunity is predictive o f later alcohol use and dysphoria. In Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 2 adulthood it should be related to prior alcohol use and depression. It is reasonable to assume that when alcohol use and depression are combined into the construct o f alcohol dysphoria, their effects will have a stronger outcome. Employment Latent construct o f job satisfaction. In order to assess Job Satis&ction, the latent construct to be used will consist o f the following measured variables: "Happiness with work" item and the "Troubles with work" item. Latent construct o f iob stability. Job Stability will consist o f the following measured variables: the "Current job” item, the “Collecting unemployment” item and the "Number o f times fired" item. Alcohol problems and employment. The impact o f drinking on employment is shown in data from the 1988 National Health Interview Survey, consisting o f interviews with over 43,000 people in all 50 states and the District o f Columbia. Among men in the workforce, across occupations, 68 percent were current drinkers with an average daily intake o f 16 grams o f ethanol (roughly one drink). Further, 10.3 percent o f men were alcohol dependent (based on DSM-IDR criteria) with 3.9 percent being severely dependent. Among women in the workforce, across occupations 49.2 percent were current drinkers with an average daily intake o f 8.1 grams o f ethanol; 4.1 percent o f women were alcohol dependent (based on DSM-IHR criteria) and 1 percent were severely dependent (Parker & Harford, 1992). Alcoholism has been shown to affect employment in several ways. Moderate levels o f consumption are associated with increased income, while heavy drinking may be harmful to income (Mullahy & Sindelar, 1992; Parker & Harford, 1992). It is often Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 3 hypothesized that alcoholics would have a less stable pattern o f employment due to the adverse consequences o f heavy drinking, h i fact, failure to fulfill major role obligations (work, school, and home) is a criterion (DSM-IV) for alcohol abuse. The direct association between alcohol use and the hours/weeks worked, length o f employment and/or job loss specifically due to drinking problems has not been widely studied. (Mullahy & Sindelar, 1992). However, one study indicates heavy drinking has been associated with as much as a 40 percent increase in absenteeism firom work (Manning, Keeler, Newhouse, Sloss & Wasserman, 1991). Low job satisfaction has been shown to be associated with alcoholism (Hingson, Mangione & Barrett,1981; Mangione & Quinn, 1975). Greenberg and Grunberg, (1995) found more drinking when job satisfaction is low and people experience low job autonomy, low skill and lack o f participation in the decision making processes in the workplace. Increases in drinking are not found when jo b satisfaction is high, although the other dimensions remain low. They suggest that job satisfaction acts to mediate the effects o f low job autonomy, low skill and lack o f participation in decision making in the workplace on drinking to cope and subsequent heavy drinking. Depression and emplovment. Literature on the relationship between depression, job satisfaction or stability is scant. It is reasonable to hypothesize that the negative thought processes and schemata associated with the depressive individual (Beck, Rush, Shaw & Emery, 1979) would apply to the environment o f the workplace. According to the tenets o f cognitive theory, people's thoughts have a large role in determining how they feel and behave. People who are depressed tend to have a negative view o f the self, the world, and the future (negative triad); they tend to perceive the self as inadequate. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44 deserted and worthless (Beck & Weishaar, 1989). Individuals with increasingly severe levels o f depression tend to have thought processes that are increasingly dominated by negative ideas and thoughts. Pervasive negative ideas and thoughts are associated with low levels o f satisfaction in life and, one m ight hypothesize, probably in work. Alcohol use is prevalent among those who work and alcohol problems have been related to both job satisfaction and/or job stabili^. Job satisfaction has been shown to be related to alcoholism and to mediate the effect o f workplace experiences or events on drinking to cope with problems and subsequent heavy drinking. Depressives are believed to have pervasive negative thoughts with consequent diminished pleasure or satisfaction. The measured variables ("Happiness with work" and "Troubles with work") should allow for an exploration o f the relationships between alcohol problems, depression and job satisfaction. Based on Beck's theory outlined above, the contribution o f depression should at least maintain if not exacerbate the effect o f alcohol on job satisfaction and/or stability. The measured variables ( "Current job " , “Collecting unemploym ent, “Number o f times fired", “Trouble with work”, and “Happiness with work”) will address these issues and examine the relationship between alcohol problems, depression and job stability. In sum, there have been significant investigations of some o f the relationships discussed. M uch o f the literature has major limitations because it is cross-sectional in design. This is problematic because it is difficult to determine whether the relationships found are antecedents or consequences o f alcohol problems or depressiotL This longitudinal investigation will facilitate this determinatioiL Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 5 Research Hypotheses Based on the literature reviewed, it appears that both alcoholism and depression individually may predict o r be consequences o f family bonding, social conformity, relationship satisfaction, m arital satisfaction, job satisfaction and job stability. The latent construct o f alcohol (fysphoria to be used in this research w ill combine both alcohol use and depression, and allow fo r a realistic investigation o f how these two processes present in the general population. From the literature reviewed, it is apparent that it is difiScult to disentangle the unique contributions o f alcoholism and depression to the functioning o f an individual within the domains described. The latent construct o f alcohol dysphoria should establish the strongest relationships. The statistical method employed (SEM) will allow for a more complete understanding o f the complexities o f the relationships examined over time. The following hypotheses are proposed: Predictive hypotheses: 1) High levels o f social conforming attitude will decrease alcohol use and depressioiL 2) Increased levels o f perceived opportunity will decrease alcohol use and depression. 3) Increased levels o f relationship satisfaction w ill decrease alcohol use and depression. 4) Increased levels o f family bonding will decrease alcohol use and depression. Outcome hypotheses: 1) Increased alcohol dysphoria will decrease m arital satisfaction. 2) Increased alcohol dysphoria will decrease relationship satisfaction. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 6 3) Dicreased alcohol ^sp h o ria will decrease perceived opportunity. 4) Increased alcohol (fysphoria will decrease job satis&ctioiL 5) Increased alcohol dysphoria will decrease job stability. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 7 CHAPTERn METHODOLOGY Participants The participants in this study are from the UCLA Longitudinal Study o f Growth and Development. Begun in 1976 this study was originally designed to address the etiology o f drug-taking behaviors and has developed to include and emphasize the long term consequences o f such behavior. A community sample o f 1634 7*, 8* and 9* grade students from 11 Los Angeles County schools constituted the original participants. Sites were selected to oversample minority and lower socioeconomic areas. Subsequently, 8 waves o f data have been collected in study years 2,4,5,9,13,17 and 21. As the participants have aged, the focus has shifred toward examination o f the consequences o f alcohol use, drug use and poly-drug use as well as an appraisal o f m ajor adult roles (Newcomb, 1997a). In this work, subjects were examined at study year 5, when they were late adolescents, aged 17-19, at study year 13, when they were young adults, aged 25-27 and, outcomes were assessed at study year 21, when they were in adulthood, 33-35 years o f age. Data were available to study differences among and within subjects across several developmental stages, allowing for an understanding o f the antecedents and consequences o f alcohol dysphoria. A cross-sectional summary o f the sample selected for use in this research appears in Table 2.1. It consists o f individuals from whom data were collected at the three time points used. The sample is ethnically mixed, has a mean age o f about 35 years and contains m ore women than men. Most have graduated high school and have about 2 years o f college, most are living with a spouse and are employed. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 Table 2.1: Description o f Sample Variable M ale (N=121) Female (N=305) Total (N=426) Age fvears) Mean 34.95 34.90 34.92 Range 34-37 33-37 33-37 Ethnicity (Yo) African-American 8.3 13.1 11.7 Latino 13.2 8.9 10.1 White 71.9 68.2 69.2 Asian-Pacific 6.6 9.5 8.7 American fiidian 0 0.3 0.2 Education fvears) Mean 14.5 13.7 14 Range 11-18 9-18 9-18 Income for past vear (%) Under 10,000 4.9 3.6 4.0 10,000-30,000 2.5 21.3 16.4 30,000-50,000 90.1 73.1 77.5 Over 50,000 2.5 2.0 2.1 Current living situation (%) Alone 7.4 9.8 9.2 With parents 3.3 5.9 5.2 Roommates/relatives 8.3 6.2 6.8 Single parent with children 0.8 10.5 7.7 No regular place to stay 0.8 0.3 0.5 Married- spouse 19.8 12.5 14.6 Married- spouse/children 53.7 48.5 50.0 Cohabit w ith partner 5.8 6.2 6.1 Number o f children (%) None 0.8 1.3 1.2 One 14.0 21.3 19.2 Two 67.8 58.7 61.3 Three 14.0 11-8 12.4 Four 2.5 5.2 4.5 Five 0.8 1.6 1.4 Current emplovment situation Unemployed, laid off, fired 0.8 3.6 2.8 Military 4.1 0 1.2 H alf tim e work/Child rearing 1.7 12.5 9.4 College/University 0.8 2.6 2.1 Part-time work 1.7 6.2 4.9 Full-time work 88.4 52.1 62.4 Child rearing/Homemaker 2.5 21.0 15.0 None 0 2.0 2.1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 9 Attrition Analysis As noted, participants in this study are part o f a longitudinal investigation that began in 1976. Because o f the significant loss o f participants since the study’s conception, extensive attrition analyses have been conducted to determine whether attrition could reflect any systematic influence or bias. Newcomb (1992) performed a series o f analyses on data fiom m ultiple measures o f personality, emotional-distress, and social support to address the 60% loss in sample size fiom 1980 to 1988. Procedures included: correlations, Bonferoni procedure, average (absolute) point-biserial correlation, and stepwise multiple regression analysis. Based on these analyses, it has been determined that that “very little o f the attrition rate between 1980 and 1988 was due to self-selection on the basis o f drug use o r personality traits” (p.266). Two variables were found to significantly differentiate the groups, but accounting for only 1 % o f the variance between groups. These results indicated that “those who continued in the study reported a better relationship with their family and greater use o f nonprescription cold medicine in 1980 than did those who did not continue in the study. Because participant drop-out was ostensibly not due to system atic self-selection or other influences on the basis o f personality factors, emotional distress social support or drug use, Newcomb (1992; 1993) concluded that the data are o f sufficiently high quality to enable structural modeling. As the study continues, additional analyses have been conducted on later years o f data collection to determine whether differences in sample size are due to bias or systematic differences between those who continue in the study and those who drop o u t Several analyses concluded that only a very small percentage o f attrition could be Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 accounted for by personality and drug use variables (Newcomb, 1997a; 1997b). No significant bias exists in the data between these variables, gender and relationship satisfaction variables (Newcomb, 1994) or intoxication at work and employment variables (Newcomb, 1995). Therefore, it can be concluded that the data are o f sufScient high quality to enable stnictural modeling and, afier multiple analyses, no systematic bias has been reported between those who continue in the study and those who drop-out Measures On the basis o f the theoretical and research literature reviewed, latent constructs were developed and variables specified to measure them. For a summary o f latent constructs and measured variables please see Table 2.2 and Figure 2.1. Adolescence. In order to examine the second-order latent variable o f alcohol dysphoria during the period o f adolescence, first-order latent constructs measuring both alcohol problems and dysphoria were utilized. Alcohol problems during adolescence were assessed by using one single-item and three multi-item scales. These scales are: “Quantity o f alcohol used”, “Use o f alcohol at school/work”, “Frequency o f alcohol consumption”, and “Times ‘super-high’ on alcohol”. 1) Quantity o f alcohol used is a single-item scale asking respondents to indicate "In the last six months, on those days when you had a drink o f beer, wine or liquor, about how many bottles o f beer, glasses o f wine, or mixed drinks did you have on a typical day?" Ranges o f response were from no drinks to 6 or more. 2) Use o f alcohol at school was assessed by asking respondents to indicate "hi the last six months, how many tim es have you been "high", "drunk", or "stoned" at school o r work?" Beer, wine and liquor are separately coded with a range o f responses between none and more than 40 times. 3) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 Frequency o f consumption was assessed by asking respondents, "About how often did you use beer, wine and liquor without a doctor's order in the last six month?” Responses range from never to more than once per day. 4) The times "super-high" category was assessed by asking respondents "in the past 6 months, how many times have you been super "high", "drunk", "stoned" on beer, wine and liquor (coded separately). Responses range from none to 40 or more times. Dysphoria during adolescence was assessed with three multi-item scales measuring depression, suicidal ideation and self-acceptance (Scheier & Newcomb, 1993). 1) A four-item depression scale asking people to respond on a 5-point scale to questions such as "Future often seems hopeless" was used to assess depression. 2) Suicidal ideation was assessed by four items asking people to respond on to questions such as "I've been thinking about ways to kill self' and "Imagined life would end with suicide" with responses ranging from never to always on a five-point scale. 3) S e lf acceptance was assessed by asking people to respond on a 5-point scale to questions such as "Like myself for what I am"; low se lf acceptance indicates dysphoria. Perceived opportunity in adolescence was assessed with three single-item scales measuring satisfaction with future opportunities, satisfaction with school and work, and satisfaction with being what you want to be (Newcomb & Jack, 1995). Each question asked people to respond on a 5-point Likert-type scale with ranges from l(very unhappy) to 5 (very happy). Social confbtm i^ in adolescence was assessed with three, four-item scales: law abidance, liberalism, and religiosity (Newcomb & Bentler, 1988). These scales reflect the degree to which an individual conforms to societal expectations and norms, using a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 semantic dififerentfal form at An example o f an item on the law abidance scale is “might shoplift certain items” and “wouldn’t know how o r want to”. An example from the liberalism scale is “approve o f many protests” and “approve o f a few protests”. An example from the religiosity scale is “am not religious” and “am a religious person”. Family bonding during adolescence was assessed by two multi-item scales measuring good relationship with parents and good relationship with fomily (Newcomb & Jack, 1995). The “Good relationship with parents” scale consists o f four bipolar items on a 5-point scale. An example o f an item on the relationships w ith parents scale is "Parents include me when they do things" and "Usually left out when they do things". The “Good relationship with family” scale consists o f four bipolar items on a 5-point scale. An example o f an item on the good relationship with fam ily scale is "Impossible to talk to my family about anything" and "Can talk to my family about anything". Relationship satisfaction during adolescence was assessed by two single-items and one multi-item scale measuring satisfaction with opposite-sex friends, satisfaction with close friends and good relationships w ith peers. The "Opposite-sex friends" scale is a single item on a five-point anchored rating scale, with responses ranging from "very happy" to "very unhappy". "Satisfaction with close friends" is a single item on five-point anchored rating scales with responses ranging from "very happy" to "very unhappy". The "Good relationship with peers" measure is a four-item measure on a five-point anchored rating scale with responses ranging from "very happy" to "very unhappy". Young adulthood. In order to examine the second-order latent construct o f alcohol dysphoria during the period o f young adulthood, first-order latent constructs measuring both alcoholism and dysphoria are utilized. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 Alcohol problems during the young adult period were assessed by a single-item scale assessing quantity o f alcohol used, a multi-item scale assessing frequency o f use, a multi-item scale assessing use o f alcohol at school or work and, a multi-item scale measuring alcohol-related problems. The quantity, use o f alcohol at school/work, and frequency scales are identical to those described above in the adolescent time period. The “Alcohol-related-problems” scale, designed to meet the DSM-UIR criteria for alcohol abuse, consists o f 29 items related to social, work and physiological frmctioning. Subjects were asked how ofren in the past year a given problem occurred from drinking alcohol, with responses ranging from never to more than once (Newcomb, 1992; Stacy, Newcomb & Bentler, 1991). Dysphoria during the young adult period was assessed with three scales measuring depression, self-acceptance, suicidal ideation and the CES-D (Scheier & Newcomb, 1993). The scales measuring depression, self-acceptance and suicidal ideation are identical to those described above in the adolescent period. The Center for Epidemiological Studies Depression Scale (Radlofif, 1977) asks respondents to rate the occurrence of various feelings such as "I felt depressed" or "I was happy" on a four-point scale with responses ranging from rarely to most o f the time. Adulthood. Outcome measures during this time period examined six areas and consisted o f one second-order latent construct (alcohol dysphoria) and frve first-order latent constructs: m arital satisfaction, relationship satisfaction, perceived opportunity, jo b satisfaction and job stability. The second-order latent construct o f alcohol dysphoria used the same scales described in the young-adult tim e period. Alcohol problems were assessed by a single- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 4 item scale assessing quantity o f alcohol used, a multi-item scale assessing use o f alcohol at school or work, a multi-item scale assessing frequency o f use and, a multi-item scale measuring alcohol-related problems. Dysphoria was assessed by with three scales measuring depression, self-acceptance, suicidal ideation and the CES-D. Marital satisfaction was assessed with a modified, 16-item version o f the Dyadic Adjustment Scale (Spanier, 1976). This scale contains four subscales designed to measure consensus, satisfaction, cohesion and affectional expression (Newcomb & Rickards, 1995). Also, an assessment o f family instability was be made by using an item that asked "I got divorced" in which the participant responds true or fidse for the past four years o f measurement. Relationship satisfaction was assessed by using the same two items described above for the adolescent period, a single-item measuring satisfaction with "Close friends" and a multi-item scale measuring "Good relationships with peers". A fifteen-item measure, the UCLA loneliness scale (Russell, Peplau & Cutrona, 1980), was added during this period. An example question from this scale is “I lack companionship” with responses being indicated on likert-type scale ranging from 1 for “never”, to 4 meaning “always”. Job satisfaction was assessed by two items addressing happiness or trouble with work. One item asks respondents to indicate on a 7-point scale the degree to which they were happy at work w ithin the past six months. Responses range from "terrible" to "delighted". The other item asks respondents to indicate on a 5-point scale the degree to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 which they were experiencing troubles at work within the last six months. Responses range from "no difiBculty" to "great difBculty" (Newcomb, 1995; Newcomb & Bentler, 1988). Job stability was assessed by three items which ask respondents to identify I) whether or not they have a current job; 2) if they are currently collecting unemployment; and, 3) the number o f times they were fired over each o f the past fi)ur years. For each year, the item responses were either yes or no (Newcomb, 1995). Statistical Procedures This research employs a structural equation modeling (SEM) technique to examine to complex relationships and reciprocal influences between the constructs described above. SEM allows for a simultaneous investigation o f multiple independent and dependent variables. Associations between latent constructs, measured variables and measured variable residuals are investigated in a confirmatory approach to multivariate analysis (Byrne, 1994). SEM has two phases, a measurement model, which is the confirmatory factor analysis model (CFA) and a structural model. The CFA evaluates the measurement models in the study that are specified in advance o f the analyses. This theory-driven approach determines whether the observed variables do represent the latent constructs in a statistically significant way. As observed variables are measured directly, their variance is broken down into two components: I) variance associated with the latent construct or factor; and, 2) the residual or disturbance term, signifying the variance not accounted for by the factor. There is some theoretical disagreement about the meaning o f the residual or disturbance terms (Gerbing & Anderson, 1984). This study will adopt the guidelines set forth by Newcomb (1994), indicating that residual terms indicate the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 6 variance not associated with the construct, instead o f only representing measurement error. In the CFA, measured variables w ith high factor loadings are assumed to m easure an unobserved factor, called a latent construct. The latent constructs capture the common associations among the measured or observed variables. At this point, the fit o f the CFA model is tested, determining whether or not the model accurately represents the data. A com bination o f three criteria are used fo r this determination: the chi-square/degrees o f rieedom ratio to be less than 2; the Comparative Fit Index (CFI) greater than .90; and the B entler Non-Normed Fit hidex (NNFI) greater than .90 (Bentler, 1992; Byrne, 1994). The next step is to develop and test the structural model to examine the m ultiple relationships between latent constructs, observed variables, residual and disturbance terms. The initial structural model uses the final CFA as a foimdation and replaces across-time correlations with directional paths. The model is then refined and developed as the relationships are tested; fit is determined to be adequate by using the same guidelines described above. Relationships between constructs are reported as are the specific effects, reflecting associations between variables other than latent constructs. The EQS program version 5.7a (Bentler, 1998) was used for all analyses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ______________________________________________________________________________________________57 Table 2.2; Summary o f Latent Constructs & M easured Variables Adolescence: Alcohol Dysphoria Alcohol Problems Quantity o f alcohol used (I ) Use o f alcohol at school/work (3) Frequency o f alcohol consumption (3) Times "super high” on alcohol (3) Dysphoria Depression Scale (4) Suicidal Ideation (4) S e lf acceptance (4) Perceived Opportunity Sati^action with future (I) Satisfaction with school/work (I) Satisfaction to be what you want (I) Social Conformity Law abidance (4) Liberalism (4) Religiosity (4) Parental Divorce Family Bonding Good relationship with parents (4) Good relationship with family (4) Relationship Satisfaction Satisfaction with opposite-sex fiiends (I) Satisfaction with close friends (1+1) Good relationship with peers (4) Young Adulthood: Alcohol Dysphoria Alcohol Problems Quantity o f alcohol used (1) Use o f alcohol at school/work (3) Frequency o f alcohol consumption (3) Alcohol-related problems scale (3) Dysphoria CES-D Depression (4) Suicidal ideation (4) Self-acceptance (4) Adulthood: Alcohol Dysphoria Alcohol Problems Quantity o f alcohol used (I) Use o f alcohol at school/work (3) Frequency o f alcohol consumption (3) Alcohol-related problems scale Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 8 Table 2.2 (Continued) Dysphoria CES-D Depression (4) Suicidal ideation (4) Self-acceptance (4) Perceived Opportunity Satîàlâction with future (1) Satisfaction with school/work (1) Satisfaction to be what you want (I) Dyadic Adjustment Scale Cohesion (4) Satis&ction (4) Consensus (4) Affect expression (4) Relationship Satis^ction Satisfaction with opposite-sex fiiends (I) Good relationship with peers (4) UCLA loneliness scale (15) Family Instability Job Satisfaction Happiness with work (1) Troubles with work (I) Job Stability Current job (1) Collecting unemployment (1) Number o f times fired (1) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 .Study year ): 17-19 years old Study year 13 25-27 years old Dysphona ton Seal ïuiadal Idean Dysphona Self-acceptance}^-------- Parental Divorce Sat opposite sex' |Sat close mem Gooa relations —ufith pfpTS__ uooa reianons . ■ with parmTT Xjooa reianons — fW T T lît-»- Dysphoria )uicidal Idean Law abidance Religiosity Sat with future Mtwithsch/ battoDewnar you wanr elaoonshi Sanstactio Family Bonding Study year 21: 33-35 years old Quantity Quantity Jseschooi/wori r Alcohol^ Jseschool/wori T^AlcohoI^ Frequency A-ProbletnsJ Frequency 1-ProbletnsJ Times Super Hi Alcprob scale f\lc prob scale Alcohol Consensus DAS Cohesion Sat with future bat to De whZ vmi want Family JCLA lonelinesj#^^ eianonshi Satisfacti Sat close men uooo reiauons with pf n a Happy w/ work .roubles w/w Job Insabiiity # limes nred Louecnng rrnwnplnymeat Figure 2.1 F actor M odel Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 0 CHAPTER m RESULTS Results o f the data analyses will be presented in two m ajor sections. The initial section o f this chapter examines gender differences between m easured variables and factors under investigation. The second section consists o f two sequences o f analyses. The first examines alcohol use and <fysphoria only, and adds the second-order construct o f alcohol dysphoria. The second phase incorporates the psychosocial variables under investigation with alcohol use, dysphoria and alcohol (fysphoria. For each phase o f data analysis, a confirmatory factor analysis was run followed by a structural equation model. For each phase, the first step involved examining first-order constructs to determine whether the model adequately fit the data. Then second-order constructs o f alcohol dysphoria were added to ascertain th e existing interrelationships Gender-related Differences Because o f the comparatively small number o f male subjects in this sample, models were run to determine whether differences exist within the sample based on gender. Four sets o f multiple group analyses were run to estim ate these differences. The first set o f analyses allowed factor loadings to be estimated fireely. The second set o f analysis constrained factor loadings to be equal between m en and women. The third set o f analyses constrained factor loadings and all correlations between factors, to determine differences based on gender. The fourth set o f analyses constrained 6 c to r loadings, correlations between factors, correlations between factors and variables as well as error variances, to determine differences based on gender. If few o r no differences emerged in these analyses, differences betw een males and females in this sample could be viewed as Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 minimal. This would justify the use o f a single, combined sample o f men and women. Tables 3.1 and 3.2 summarize these results. Table 3.1: M ultiple group analyses o f factor models. Model 1) Multiple Group CFA A) All Free Chi Square 3495.6 d f 2128 X 2/df prob. 1.64 Nnfi .83 CFI .858 B) Factor Loadings Constrained 3542.96 2177 1.63 .834 .859 C) Factor Loadings & Corr. Constrained 3707.97 2282 1.62 .835 .852 D) Factor Loadings & 3782.18 Factor Corr. & Factor-Variable Corr. & Error Constrained 2332 1.62 .836 .85 2) Difiference Tests A-B 83 49 <01 C-B 165 105 <01 D-C 74 50 <01 D-B 239 155 <01 The four multiple-group CFA analyses yielded non-significant differences, indicating no significant differences on the parameters constrained by gender. When looking at differences between models where factors are constrained and where factors and their correlations w ith either control variables and/or error are constrained, the models are statistically the same. This indicates that the differences in correlations, factor estimates and error between men and women are non-significant. Factor loadings by gender are presented in Table 3.2. As can be seen, among the 51 variables in this study, only one (use o f alcohol at school/work- time two) varied significantly by gender. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 2 Factor intercorrelations by gender are presented in Table 33. As can be seen, several relationships vary by gender. The relationships between alcohol problems and dysphoria as well as between perceived opportunity and relationship satis&ction are different for boys and girls at tim e 1. There are sex differences in the relationship between social conformity at tim e 1 and alcohol problems at tim e 2. The relationship between parental divorce and depression at time I are different for boys and girls. Sex differences exist between parental divorce at time 1 and perceived opportunity at time 3. The relationship between alcohol problems at time 2 and relationship satisfaction at time 3 differ for men and women. Sex differences exist between perceived opportunity and parental divorce are different for men and women at time 3 and, between jo b satisfaction and job stability at time 3. The relationship between family instability at tim e 3 and DAS differ by sex. In sum, based on the analyses conducted, it is determined that the use o f a single model combining women and men is not appropriate for these data. Sex differences exist on the primarv constructs under investigation, alcoholism and dvsphoria. This indicates that for men in this sample alcoholism and depression are unrelated during adolescence, but they are for women. To further explore this relationship, subsequent analyses will focus only on the female portion o f the sample. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Table 3.2: Table o f factor loadings by gender Variable Male Female P-value Adolescence: Alcohol Problems Quantity o f alcohol used .688 Use o f alcohol at school/work .611 Frequency o f alcohol consumption .804 Times "super high" on alcohol .823 Dysphoria Depression Scale .934 Suicidal Ideation .442 Self-acceptance -.774 Perceived Opportunity Satisfaction with future .769 Satisfaction with school/work .521 Satisfaction to be what you want .776 Family Bonding Good relationship with parents .876 Good relationship with âm ily .747 Relationship Satisfaction Satisfaction with opposite-sex fiiends -.341 Satisfaction with same-sex + close fiiends -.799 Good relationship with peers -.699 Young Adulthood: Alcohol Problems Quantity o f alcohol used .702 Use o f alcohol at school/work .589 Frequency o f alcohol consumption .649 Alcohol-related problems scale .661 Dysphoria CES-D .828 Depression .869 Suicidal ideation .744 Self-acceptance -.887 Middle Adulthood: Alcohol Problems Quantity of alcohol used .643 Use o f alcohol at school/work .336 Frequency o f alcohol consumption .741 Alcohol-related problems scale .577 .760 .423 .884 .744 .871 .250 -.795 .744 .489 .688 .857 .876 -.340 -.713 -.744 .732 .255 .777 .693 .678 .868 .673 -.846 .802 .256 .758 .652 ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns < 0 5 ns ns ns ns ns ns ns ns ns ns Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 4 Table 3.2 (Continued) Dysphoria CES-D .819 .736 ns Depression .895 .905 ns Suicidal ideation .580 .602 ns Self-acceptance -.916 -.850 ns Perceived Opportunity Satisfaction with future .859 .804 ns Satisfaction with, school/woric .820 .677 ns Satisfaction to be what you want .784 .791 ns Dyadic Adjustment Scale Consensus .887 .809 ns Satisfaction .799 .849 ns Cohesion .627 .679 ns Affect expression .467 .408 ns Relationship Satisfaction UCLA Loneliness scale -.909 -.819 ns Satisfaction with same-sex + close fiiends .578 .685 ns Good relationship with peers .670 .801 ns Job Satisfaction Happiness with work .754 .709 ns Troubles with work .394 .545 ns Job Stability Current job -.166 -.130 ns Number o f times fired .260 .645 ns Collecting unemployment .804 .658 ns ________Ns = non-significant________________________________________________ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.3: Factor Intercorrelations bv Gender 6 5 Ale ProbS DvsphS PercOpS FamBonS Rei SatS SocCbnfS AlcProb 7 Dysph 7 Ale Prob9 Ale ProbS .31 -.30 -.24 .06 -.77 .55 .11 .53 DysphS .03* -.72 -.55 .56 -.42 .17 .46 .23 PercOpS -.08 -7 5 .38 .32 .40 -.19 -.33 -20 FamBonS -3 0 -.60 .38 — -2 6 .43 -.14 -.34 -20 Rei SatS -1 0 .76 -.70* -.45 -.19 .08 .32 .12 SocConfS -.80 -.27 .19 .66 -.15 61 -26 -.51 Ale Prob7 .55 .00 -.07 -.13 -.08 -.40* .32 .65 DysphZ .02 .37 -.41 -.26 .45 -2 2 2 7 21 Ale Prob9 .40 .11 .00 -.37 .08 -.45 .70 .30 Dysph9 .09 -42 -.43 -.41 .44 -.38 .08 .72 .14 Pere Op9 -1 2 -.36 .38 .30 -.39 .36 -.16 -.60 -.21 DAS9 -2 4 -3 2 .27 .40 -16 .39 -.16 -.17 -.25 Rei Sat9 -10 -.38 .39 .44 -.44 .35 .05* -.61 -.09 Job Sat9 -.01 -.36 .43 .17 -.50 .13 -.08 -.51 -.24 Jobinsta9 -.32 .13 -.04 .38 .18 38 -.31 .21 -2 5 Par Div -.14 .o r -.15 -.12 .06 -.11 .13 .04 .19 Divcree -.05 -16 .12 .16 -.06 .15 -.00 -.15 -.08 Dysph9 Pere Op9 □AS9 Rei Sat9 Job Sat9 Jobirtsta9 Par Div Divorce .04 -.03 .07 -.04 -.10 -.04 .36 -.15 Ale Prob5 .54 -.42 -.34 -.44 -.29 -.02 .26" -.25 Dysph5 -.28 .28 22 .19 .27 .01 -.10 .08 PercOp5 -.33 .27 .24 .38 .25 00 -.22 .26 FamBonS .32 -.18 -.31 -.31 -.11 -.04 .05 -.11 RelSatS -21 .15 .15 .19 31 .15 -.35 .07 SocConfô .15 -.24 -05 -.16 -3 2 -.18 .26 .04 AlcProb7 .63 -.55 -31 -57 -35 -01 .09 -06 Dysph7 .12 -13 -12 -.06 -14 -.01 .29 -16 Ale Prob9 -.78 -.45 -.77 -.60 -01 .14 -25 Dysph9 -.86 .42 .64 .70 .07 .09' .19 Perc0p9 -.32 .32 .42 .35 .03 -.06 .39" DAS9 -.87 .69 40 — * .41 .04 -.04 .10 RetSat9 -.72 .84 -.01* .55 — .09 -.09 .12 JobSat9 -.07 .13 .00 .04 34* — -.16 .03 Jobinsta9 .10 - . i r -.10 .02 -.14 -.02 .10 Par Div -.15 .10 .16* .11 .09 .17 .10 Divorce Data from adolescence is signified with a S , from young adulthood with a 7, and adulthood with a 9 Males are presented in lower mangle, females in upper triangle. Scores that differed significantly are in bold. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 Sequence o f Analyses The analyses conducted consisted o f two phases, utilizing the same sequence o f procedures. First, a confirmatory factor analysis was conducted on the data to determine whether the observed variables do, in fact, represent the latent constructs in a statistically significant way. W hen this fit was judged to be adequate, a structural model was then developed and tested. Initially, the relationships between alcohol and depression were examined alone to determine the stabili^ and relationships over time. A confirmatory factor analysis was conducted with only the first-order latent constructs, to determine whether the model fit the data befi)re imposing second-order constructs. A structural model was then developed and tested on these first-order constructs. Then, second-order constructs o f alcohol dysphoria were introduced and a confirmatory factor analysis was conducted on the associated factors and variables. A structural model was then developed on these second-order constructs, and judged to be adequate. Following this, the second- order constructs were reintroduced into the entire data set with the psychosocial variables. A first-order confirmatory factor model was developed to determine whether the model fit the data in a statistically significant way. Once judged to be adequate, a structural model was developed on this data. Then, second-order constructs were introduced and a confirmatory factor model was developed and judged to be adequate. Finally, a second-order structural model was developed, utilizing all the data imder investigation. Statistical Fit Indices Statistical fit o f the models to the data was determined by a combination o f three criteria: the chi-square/degrees o f fi’ eedom ratio to be less than 2; the Comparative Fit Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 7 Index (CFI) greater than .90; and the Bentler Non-Nonned Fit Index (NNFI) greater than .90 (Bentler, 1992; Byrne, 1994). Alcohol problems and dvsohoria over tim e The first-order CFA model involving the constructs of alcohol problems and dysphoria fit the data well: CFI = .95, Chi square/df = 2.28, NNFI = .93. A second-order CFA was developed adding the construct o f alcohol dysphoria over tim e to the model. This was found to have an adequate fit: CFI = .92, Chi square/df = 2.36, NNFI = .90. Even though the Chi square/df ratio is slightly above 2.0, the magnitude o f the CFI ratio allowed this to be deemed sufficient The confirmatory factor analysis is depicted in Figure 3.1. The second-order structural model had a good f it CFI = .94, Chi square/df = 2.05, NNFI = .92. Figure 3.2 depicts this second-order structural model. Alcohol problems, dvsphoria and psvchosocial variables over time. The psychosocial variables under investigation were then added to the constructs o f alcohol problems and dysphoria. The fit o f this initial CFA was judged to be adequate based on the following: CFI = .89; Chi square/df = 1.64, NNFI = .87; this model is depicted in Figure 3.3. A first-order structural model was then developed and judged to have an adequate fit: CFI = .89, Chi square/df = 1.64, NNFI =.87. The second-order constructs o f alcohol dysphoria were then added to the model, a CFA and a structural model were developed. This CFA was judged to have an adequate fit: CFI = .87, Chi Square/df ratio = 1.81, NNFI = .85; this model is depicted in Figure 3.3. The second-order structural model fit the data well based on the following: CFI = .89, Chi Square/df ratio = 1.64, NNFI = .87; this model is depicted in Figure 3.4. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 In sum, ail m odels tested were found to fît the data well. Although some variability in the criteria was experienced, allowances were made based on the size o f the model tested. Method o f developing structural models Once the CFA m odel fit was adequate and the constructs were correlated in the hypothesized directions, structural models were developed for the latent constructs, h i these models, the residual or disturbance terms o f the constructs were allowed to correlate freely. Structural models were initially based on final CFA models. Structural models were then further developed by adding paths based on the fîndings o f the Lagrange Multiplier test for adding parameters. This test identified significant paths and correlations to add to the model. Final models were judged to be adequate by the same criteria used for the CFA models. The following across-time relationships were evaluated: latent constructs at tim e one to latent constructs at tim es two and three; latent constructs at time one to measured variables at times two and three; measured variables, residual and disturbance term s at time one to latent constructs and measured variables and disturbance terms at time two and three; measured variables, residual and disturbance terms at time two to latent constructs and measured variables and disturbance term s at tim e three. This analysis allowed for an exam ination o f the predictors and consequences o f the constructs and variables under investigation. Eoualitv Constraints There were equality constraints placed upon certain variables and factors to bolster the stability o f the model and ensure that the same constructs over time were Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 equal. The measured variables constrained were “frequency o f alcohol consumed” and “depression scale” scores. This constitutes a partial invariance (Byrne, 1994). Factors constrained were alcohol problems and dysphoria across time. This process has the effect o f mathematically equating the alcohol and dysphoria factors at the three time periods, ensuring equality o f meaning over time. This was necessary to control for the one different measurement variable across time. It may be recalled that the “Times super high” at time one is a different variable than the “Alcohol problems scale” used in the alcohol problems construct at times two and three. Also, the CES-D was added to the Dysphoria construct at time two and three. Alcohol Problems. Dvsphoria & Alcohol Dvsphoria Over Time To determine whether the measured variables were significantly representative o f the latent constructs under investigation, a CFA was conducted on the data before attempting to test structural models among the latent constructs. Figure 3.1 depicts the initial factor structure, the standardized factor loadings and residual variances o f the measured variables in the CFA model. Overall, the standardized factor loadings were large and significant, indicating that the measured variables were reliable indicators o f the latent factors. Following this, second-order constructs o f alcohol dysphoria were developed and integrated into the model. Alcohol dysphoria was found to be a stable construct In adulthood, alcohol dysphoria is positively predicted by alcohol dysphoria in young adulthood and is positively predicted by alcohol dysphoria in adolescence. Alcohol dysphoria in young adulthood is positively predicted by alcohol dysphoria in adolescence. Alcohol problems in young adulthood positively predicts scores on the depression scale in adulthood. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 Dysphoria in young adulthood positively predicts use o f alcohol at school and/or work and ratings on the alcohol problems scale in adulthood. Both quantity o f alcohol used and ratings on the alcohol problems scale in young adulthood positively predict CES-D scores in adulthood. These relationships are depicted in Figure 3.2. Alcohol Problems. Dvsphoria and Psvchosocial Variables Over Time When integrating the psychosocial variables into the existing model, it was necessary to determine whether the measured variables were significantly representative o f the latent constructs under investigation. A CFA was conducted on the integrated model before attempting to test structural models among the latent constructs. In this model, the standardized factor loadings were large and significant, indicating that the measured variables were reliable indicators o f the latent factors. The intercorrelations among the factors were correlated in expected directions with many significant relationships. The first-order structural equation model was then developed, with paths replacing the across-time correlations from the CFA. Table 3.4 presents the factor intercorrelations based on this first-order analysis. Overall, as can be seen, many statistically significant relationships exist within the data. Integrating Alcohol Dvsphoria with Psvchosocial Variables Over Time Alcohol dysphoria in adolescence positively predicts alcohol dysphoria in young adulthood, which, in turn, positively predicts alcohol dysphoria in adulthood. Dysphoria in adolescence negatively predicts marital satisfaction as measured by the DAS and negatively predicts perceived opportunity and relationship satisfaction in adulthood. Perceived opportunity in adolescence positively predicts perceived opportunity in adulthood. Parental divorce negatively predicts relationship satisfaction in adulthood Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71 while relationship satisfaction in adolescence does not predict i t Fam ily bonding in adolescence positively predicts fam ily stability and marital satisfaction as measured by the DAS in adulthood. Alcohol dysphoria in young adulthood is a strong negative predictor o f a number o f factors in adulthood: m arital satisfaction as measured by the DAS, relationship satisfaction and Job satisfaction. A n increase in alcoholic dysphoria from adolescence to young adulthood predicts perceived opportunity in adulthood. Dysphoria in young adulthood predicts relationship satisfaction in adulthood. Alcohol problems in young adulthood negatively predicts job satisüiction in adulthood. These relationships are depicted in Figure 3.4. Additional specific effects o f variables in adolescence and young adulthood are presented in Table 3.6. As can be seen, a number o f effects exist for both adolescence and adulthood. Adolescent predictors and later outcomes. A number o f factors and variables in adolescence predict later effects. Fam ily Bonding in adolescence negatively predicts alcohol quantity and scores on the depression scale in adulthood. Good relationship with parents in adolescence negatively predicts alcohol problems and use o f alcohol at work or school in young adulthood, while good relationship with family in adolescence positively predicts relationship satisfaction in adulthood. Satisfaction to be what you want to be in adolescence negatively predicts dysphoria in young adulthood. Satisfaction with school or work in adolescence negatively predicts alcohol problems and suicidal ideation in adulthood. Satisfaction with the future in adolescence positively predicts Job Satis&ction in adulthood. Relationship Satisfaction in adolescence negatively predicts CES-D Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 depression and suicidal ideation in adulthood. Satis&ction w ith relationships with the opposite sex in adolescence negatively predicts Dysphoria in adulthood. Good relationship with peers negatively predicts use o f alcohol at work or school in adulthood. Social Conform i^ negatively predicts alcohol problems scale scores in adulthood, while religiosity in adolescence negatively predicts depression scale scores in adulthood. Parental Divorce in adolescence negatively predicts satisfaction w ith the future in adulthood. Young adulthood predictors and adult outcomes. Specific variables and factors in young adulthood have a significant effect on variables and/or factors in adulthood. Use o f alcohol at work or school in young adulthood negatively predicts happiness at work in adulthood, while quantity o f alcohol consumed positively predicts job instability. Dysphoria in young adulthood positively predicts frequency o f alcohol consumed in adulthood, while it negatively predicts satisfaction with the future and number o f times fired from a Job in adulthood. CES-D scores in young adulthood negatively predict marital satisfaction as measured by the DAS in adulthood, while they positively predict loneliness and trouble with work in adulthood Suicidal ideation in young adulthood positively predicts Dysphoria in adulthood and negatively predicts perceived opportunity in adulthood. The second-order structural model was refined by use o f the lagrange m ultiplier te st Final fit was good. Comparative Fit Index (CFI) = .89, Chi square/df ratio = 1.64. Figure 3.4 depicts this second-order structural model. As can be seen a number o f significant relationships ex ist Also, many significant correlational relationships exist. / Table 3.5 presents the factor intercorrelations based on this second-order analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 Ale ProbS 1.00 DysphS PercOpS FamBonS Rei SatS SocConK AlcProb? Dysph? Ale ProbS DysphS .31“ * 1.00 Pere OpS -.30“ * -?1*“ i.OO FamBonS -.25*“ -5S*“ .38*“ 1.00 Rei SatS .06 -.56*“ -.31*“ -.26*“ 1.00 SocConfS -?9*“ -.42*“ .40*“ 43*“ -18* 1.00 AleProb? .55*“ .1?* -.18* -.15* .08 -.61*“ 1.00 Dysph? .11(10 .45*“ -.32*“ -.33*“ -.33*“ -.25*“ .32*“ 1.00 Ale ProbS .53*“ .24*“ -.20“ -.20“ .12 -.51*“ .65*“ .21*“ 1.00 DysphS .04 .53*“ -.2?*“ -.33*“ -.32*“ -.21“ .15* .62*“ .1 1 (1 * ) PercOpS -.03 -.41*“ .2?*“ .2?*“ -.18* .15(1*) -.23*“ -.54*“ -.13(10 DASS 0? -.33*“ .21“ .24*“ -.31*“ .15(1*) -OS -.30*“ -.12(10 Rei SatS -.05 -.43*“ IS* .38*“ -.31*“ .IS* -.16* -.5?*“ -.05 Job SatS -11 -.28*“ .2?*“ .25*“ -.10 .32*“ -.32*“ -.33*“ -.13 JobinstaS -OS -OS .08 .04 -.08 .20* .22“ -.10 -.02 Popmom .13 .05 -.15* -.12* .05 -.10 .12(10 .04 .12(10 Divcorce -.05 -.17“ ,13(io .15 -.05 .14(10 -.01 -.14* -.07 DysphS Pere OpS DASS Rei SatS Job SatS JobinstaS Popmom Divorce 1.00 -78“ * 1.00 -.44"“ .42*“ 1.00 -.76*“ .62*“ .40*“ 1.00 -59“ * 6S*“ .36*“ 3S*“ 1.00 -.O S .01 .04 -0? IS* 1.00 09 -.10 -.10 .04 -.15 -.01 -.14* .O S 16“ .10 .12 17* 1.00 - .10(1 *18 ) 1.00 Table 3.4: First-order Factor Intercorrelations Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 4 AlcOepS Pere Op5 FamBonS Rei SatS SocConfS AlcOep? AfcOepS Pere Op9 DAS9 RelSat9 Job Sat9 Jobinsta9 1.00 PercOpS FamBonS Rei SatS SocConfS AlcOep7 AicOep9 PercOp9 DAS9 .92" '.71— .66— .81— .44— .58— .49— -.35— -.52— -.38— -.15 1.00 .37— -.31 — .42— .43— -.40— .27— .21- .19" .27— .08 1.00 -24— .44— -.43— -.47— .27— .24— .39— 2 5 — .05 1.00 - 20" .40— .4 3 - -.18" -.3 1 - -.3 1 - -.10 -.08 1.00 -.56— -.43— .15(11) .14(11) .20" .32— 21" 1.00 .78— -.75— -.38— -.75— -.57— -24" 1.00 - 1.00— -.58— -.95— -.77— -.14 1.00 .42 .62 .68 .10 1.00 .40 .35 .04 Rei Sat9 Job Sat9 Jobinsta9 1.00 .3 9 - .07 1.00 .19" 1.00 Tables .5: Second-order Factor Intercorrelatioas Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75 Table 3.6: Predictor and outcome effects Predictor Constructs Later Outcomes Standardized Estimate at Time 1 Family Bonding Construct: FAMILY BONDING alcohol quantity T3 -.11* FAMILY BONDING depression scale T3 -.07* + Relationship Satisfaction Construct: RELATIONSHIP SATISFACTION cesd T3 -.12* RELATIONSHIP SATISFACTION suicide T3 -.10* + Social Conformity Construct: SOCIAL CONFORMITY alcohol problems scale T3 -. 16*** Predictor Variables Later Outcomes Standardized Estimate at Time 1 Variables From Family Bonding Construct: good relations w / parents ALCOHOL PROBLEMS T2 -.16* good relations w / parents use o f alcohol at work/school T2 -. 16* good relationship with fam RELATIONSHIP SAT T3 .17* Variables From Perceiyed O p p o rtu n ity Construct: satisfaction with school/work ALCOHOL PROBLEMS T3 -. 10* -i- satisfaction with school/work suicide T3 -.11* satisfaction with future JOB SATISFACTION T3 .17* satisfaction w/ what you want to be DYSPHORIA T2 -. 10* + Variables From Relationship Satisfaction Construct: sat with rei opposite sex DYSPHORIA T3 -.07* + good relationship with peers use o f alcohol at work/school T3 -. 12* Varibles From Social Conformity Construct: religiosity depression scale T3 -.11 *** Parental Diyorce: parental diyorce satisfaction with future T3 -.07* + Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 Table 3.6 (Continued) P redictor C onstructs a t Tim e 2 Dvsnhoria Construct: DYSPHORIA DYSPHORIA DYSPHORIA Later Outcomes Standardized Estimate alcohol firequency T3 satisfaction w ith future T3 number o f tim es fired T3 .14** -.15** -.17* P redictor V ariables a t Tim e 2 Later Outcomes Standardized Estimate Variables From Alcohol Problems Construct: use o f alcohol at work/school happy at work T3 -.15** alcohol quantity JOB INSTABILITY .23* + = 1 tailed test o f significance In all cases, CONSTRUCTS are in ALL CAPS, variables in lower case Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C D ■ D O Q. C g Q. ■ D C D C / ) C / ) 8 ■ D 3. 3 " C D C D ■ D O Q. C a O 3 " O o C D Q. Quantity ( § ) * ( ^ s ) ^ Quantity Jse school/wort (^22) ^ Frequency Alcohol ,88*i)t Problems (!94)'^ s e school/wori (30) ^ Frequency Alcohol 80**1. Problems (35) ^ Quantity Jse school/worli (ao) ^ Frequency ^ 6 ) ^ Times Super Hi Ale prob scale Ale prob scale Alcohol J)ysphoriay )epression Scab .31** Alcohol Dysphoria. CES-D ( 94) ^ Suicidal Ideatioi Dysphoria j ^ 3 ^ ^ Depression Seal Self-acceptance .78***. Study year S; 17-19 years old ([56) ^ Suici Suicidal Ideatioi ^ 0) ^ Self-acceptance CES-D ( j s ) ^ Depression Seal Suicidal Ideatioi ( S s ) ^ Self-acceptance Alcohol .77**!. Problems Alcohol Dysphoria Study year 13; 25-27 years old Study year 21:33-35 years old ■ D C D C / ) C / ) Figure 3.1: Second-Order Confirmatory Factor Model 7 8 ( 4 ^ * Quantity se school/wori Frequency ^ W * Times Super ffi suicidal Ideatiof 2^ CO Seal i^nSelf-ecceptanceliir^***’ uiadalld \lcohol Aie prob ^ w )* p e p resiio a S c ai modal Ideancr ( 3 ) * SelFacceptance Study year 5: 17-19 years old Study year 13:25-27 years old Study year 21:33-35 years old Figure 3 2 : Second-Order Structural Model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 Study yearS: 17-19 yeusold Quantity Times Super Hi SUu^year 13:25-27yeats old Study year 21:33-35 years old Quandly ( 3 ^ seschooiAv Alcohol Alcohol * ^ P ro b Ie m s Alcohol Frequency Frecfuency Frequency ^ y ^ Alcprobscale ^ ? y ^ Ale prob scale Alcohol Alcohol Dysphorie Depression Seal ^iy^pepressioii Seal uicidal Idean uiddal Idean uicidal (soy^ Law abidance ^ sîy ^ Religiosity ^4?y^ Sat with future 7 4 * # * Sat with Anure (5y^P «w ithscfa^ 3 7 bat to oe wnat uou-acaas bat to oe wnat i.S O ramuy Insnhiliiy Parental Divorce ( S y ^ Sat opposite JCLA ionelin t close then uooareianons i U t f a withnrers ( ^ y ^ Happy w/ work uooQ retaoons with family roubles w/ I2*fltaü) ^% y^ #dtnesAred 70**f Job Instabiliw ^^^y^ ^jood reianons /CqV^I uoou relations .H f l t f a p a r m t s jnfmplnyme Figure 3.3 Second-order Confirmatory Factor Model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Study year 5:17-19 years old Stuffy year 13:25-27 years old 8 0 Stuffy year 21:33-35 years f)ld .8 2 ,.75 .86 Alcohol Problems Alcfshol \ Problems X Alcohfd Problems 49 Alcohol Dysphoria Alcohol Dysphoria Dysphoria Dysphoria .45, ' Social ^ iCoofbrmit% .-.18» DAS Perceived Perceived Parental Divoi ■larionshi] itisfactiot Jarionshi] itisfactioii Job isfactii Family Bonding Vlistabili^/ Figure 3.4: Second-order Factor Model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 CHAPTER IV DISCUSSION: This study was designed to explore the complex relationships between and the impact o f alcoholism, depression and psychosocial factors in adolescence and young adulthood, and the consequences o f these factors on psychosocial functioning, alcohol use and depression in adulthood. Several components o f this stucfy^ make it somewhat unique when compared to much o f existing literature: the sample, the design, and the statistical methodology used. A community sample was used, consequently, the results and implications have greater generalizability than those from clinical or treatment samples. The constructs were investigated longitudinally, temporal sequencing was established, and the predictors, consequences and developmental implications o f alcoholism and depression within the comprehensive domains addressed were determined. Further, this work employs Structural Equation Modeling, a powerful statistical tool, allowing causal relationships between measured variables and constructed factors to be established. SEM can statistically separate the unique effects and combined effects o f the constructs and variables under investigation, allowing for an in-depth analysis. These results have the potential to inform etiological theory, contemporary research and intervention efforts. When originally designing this study and reviewing the literature, a combined sample, using both m en and women, was conceptually appropriate. However, the analysis of gender differences yielded a significant gender effect, that alcohol problems and dysphoria were not significantly related for the men in our sample during the adolescent period. Therefore, for this report analyses were conducted on a female-only sample. The Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 results o f this investigation with women supported many o f the hypotheses proposed for the original combined sample. Alcohol dysphoria is a stronger predictor than either alcoholism o r depression alone or other psychosocial variables under investigation fo r the domains o f marital satisfaction, perceived opportuni^, relationship satisfoction and job satisfaction. There is a gender effect when the relationship between alcoholism and depression is examined in adolescence. For girls, they are significantly correlated, but not for the boys in this sample. Further, once alcohol dysphoria is established in young adulthood, it is very likely to persist into adulthood. This chapter provides a discussion o f the results for each o f the hypothesized constructs. The significant across-tim e paths and intercorrelations between variables, variable residuals, factors, and factor disturbances are discussed. Following this, an integration o f the results is presented; implications for theory and intervention are discussed. The methodological issues and limitations are presented as well as research implications. Research Findings Predictors o f Alcohol Use and Depression Several factors as well as specific variables in adolescence predicted alcohol use and depression in young adulthood and adulthood, hi addition, numerous correlational relationships exist between factors within and across time periods. Social conformitv. The hypothesis that attitudes o f social conformity will decrease alcohol use and depression was supported Increased social conformity, namely religiosity, law abidance, and reduced liberalism, decreases alcohol problems in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. «3 adulthood; religiosity alone decreases later depression. These relationships are consistent with existing research findings (Burch, 1994; Newcomb, M addahian & Bentler, 1986; McGee & Newcomb, 1992; Smith & Newman, 1990; Sovani, 1987). Attitudes o f social conform ity in adolescence have 6 r reaching effects, suggesting that attention to this variable may be significant in prevention efforts. Social conformity predicts future alcohol use and depression, and, in addition, correlates with many psychosocial domains Social conformity in adolescence was correlated with many o f the psychosocial factors across time. Increased social conformity^ was correlated with decreased alcohol problems, dysphoria and alcohol dysphoria in both young adulthood and adulthood. Further, it was correlated with increased satisfaction in employment, marriage and relationships, an increased sense o f perceived opportunity and more stability in employment. Perceived opportunity. The hypothesis that increased levels o f perceived opportunity will decrease later problems with alcohol and depression was supported. An increase in the sense o f agency or perceived opportunity in adolescence, specifically satisfaction with school, decreases both alcohol problems and suicidal ideation in adulthood. Satisfaction with “what you want to be” decreases dysphoria in young adulthood. Perceived opportunity is a risk and/or protective factor to consider in adolescence, as it is strongly related to later problems. Correlational relationships exist between increased perceived opportunity and decreased alcohol problems, dysphoria and alcohol dysphoria Also, increased perceived opportunity was related to increased levels o f satisfaction in marriage, relationships and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 4 employment in adulthood. These fîndings highlight the importance o f future outlook or sense of agency in adolescence, continuing into adulthood. Relationship satisfaction. The hypothesis that relationship satisfaction will reduce later alcohol problems and dysphoria was supported. It was found th at being satisfîed in relationships during adolescence reduces depression and suicidal ideation in adulthood. Also, the specifîc effect o f satisfaction in relationships with the opposite sex leads to decreased dysphoria in adulthood, while being satisfîed with peer relationships decreases the likelihood o f using alcohol at school or work in adulthood. Learning in adolescence to develop and maintain satisfying relationships is another important aspect to consider in prevention. Relationship satisfaction was correlated with decreased levels o f alcohol problems, dysphoria and alcohol dysphoria in young adulthood and adulthood. It was also correlated with increased perceived opportunity and increased satisfaction in marriage and relationships. Familv bonding. The hypothesis that increased bonding to the femily during adolescence decreases later alcohol problems and dysphoria was supported. Having a good bond with one’s family decreases quantity o f alcohol consumed and depression in adulthood. Further, specifîcally, having a good relationship with parents is important, as it decreases alcohol problems and use o f alcohol at school or work in young adulthood. A good relationship with one’s fam ily during adolescence increases relationship satisfaction in adulthood. Family bonding in adolescence was correlated with decreased alcohol problems, dysphoria and alcohol dysphoria during young adulthood and adulthood. It was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 also correlated with increased perceived opportunity^ and satisfaction with relationships, marriage and employment in adulthood. This is consistent with previous findings (Brook, Lukof^ & W hiteman, 1980; Hawkins et al., 1992; Hundleby& M ercer 1987; Newcomb, Maddahian & Bentler, 1986) and clearly demonstrates that the nature o f family relationships in adolescence significantly impacts many areas o f functioning in young adulthood and adulthood. Outcomes o f Alcohol Problems. Dvsphoria and Alcohol Dvsphoria In this investigation, focused on the women in the sample, alcohol (fysphoria was found to be a stable construct from adolescence through adulthood. Alcohol dysphoria in adolescence leads to alcohol (fysphoria in young adulthcxxi, which leads to alcohol dysphoria in adulthcxxL The relationship between alcohol dysphoria in young adulthood and adulthood is extremely strong, indicating that once established, it tends to persist. In addition, alcohol dysphoria was found to be a strong predictor construct, leading to later problems in marital adjustment, perceived opportunity, relationship and job satisfaction. These relationships are discussed below. Marital satisfaction. The hypothesis that alcohol dysphoria will negatively impact marital satisfaction was supportecL Alcohol dysphoria during young adulthrxxi decreased marital satisfaction to a greater extent than psychosocial factors. Other effects that were o f significantly impact were dysphoria in adolescence and, depression as measured by the CES-D in young adulthood. Both depression and later alcohol dysphoria lead to decreased marital satisfaction, with alcohol (fysphoria being more influential. This is consistent with literature on both alcohol and depression, and their relationship to marital satisfaction. Alcoholism has been found to influence aspects of marital satisfaction Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 6 (McKay et al., 1992; OTarreU & B irchler, 1987; O ’Farreli & Birchler, 1991 ; Suman & Nagalakshmi, 1995; Schafî, Lavely & JafiFe, 1975), as has depression (Rounsaville et al., 1979; McCabe & Gotlib, 1993; Thompson, WhifFen, & Blain, 1995), with a gender efiFect noted with women (Assh & Byers, 1996). The combined construct o f alcohol (fysphoria goes beyond existing literature in that it is a strong predictor over tim e and has a greater impact than alcoholism or depression alone. In addition, marital satisfaction was positively correlated with m any o f the psychosocial factors under investigation: perceived opportunity, fam ily bonding, relationship satisfaction and social conformify in adolescence; perceived opportunify, relationship satisfaction and job satisfaction in adulthcxxL It was negatively correlated with alcohol dysphoria at all time pericxls. The combined effects o f alcohol and dysphoria are additive in this sample, and are stronger than their unique effects. Therefore, while it is clear that m<x)d-state during adolescence and young adultho(xi negatively impacts later marital satisfaction, when both alcohol problems and depression are present, they experience a synergistic effect that more strongly diminishes later marital satisfaction. Relationship satisfaction. The hypothesis that relationship satisfaction would be negatively impacted by alcohol dysphoria was supportecL Alcohol dysphoria in young adulthood had a greater negative im pact on relationship satisfaction in adulthcxxi than other variables. In magnitude, the im pact o f this factor was followed by dysphoria in young adulthood ancL dysphoria and parental divorce in adolescence. This goes beyond existing research, which dcx;uments a relationship between alcoholism (Newcomb & Bentler, 1986; Medora & WocxiwarcL 1991), depression (McCabe & Gotlib, 1993) and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 7 aspects o f relationship satisfaction. Consistent with the hypotheses set forth here, evidence is provided o f a synergistic or additive strength o f alcohol (fysphoria on relationship satisfactioiL Additionally, having a good relationship with one’s family in adolescence increases relationship satisfaction in adulthood. In adulthood, relationship satisfaction was positively correlated with perceived opportunity, fam ily bonding, relationship satisfaction and scxnal conformify in adolescence; and, m arital satisfaction, perceived opportunify, and job satisfaction in adulthood. It was negatively correlated with alcohol (fysphoria across time, with an almost perfect correlation at time three. People in this sample who are depressed in adolescence or young adulthood, but particularly those who experience alcohol dysphoria during young adulthocxi, are likely to have de(nreased relationship satisfaction in adulthood. Further, during adolescence it appears that the mcxxi-state and/or whether one’s parents obtained a divorce is a better predictor o f adult relationship satisfaction than relationship satisfaction during adolescence. Relationship satisfaction is potentially a fluid construct, influenced over time by alcohol- and mood-related factors. Perceived opportunitv. Alcohol dysphoria negatively impacted perceived opportunify consistent with the hypothesis set forth in this investigation. However, the effect o f alcohol dysphoria followed a different course on this factor, as an increase in the level o f alcohol dysphoria from adolescence to young adultho<xi was required to decrease adult perceived opportunify. Another effect was that dysphoria or mo(xi-state in adolescence decreased perceived opportunify in adulthood, as did suicidal ideation in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 young adulthood. Having a sense o f perceived opportunity in adolescence is maintained over time, although it is strongly impacted by other factors throughout the life course. Perceived opportunity in adulthood was negatively correlated with alcohol dysphoria across time, with a perfect relationship existing in adulthood. Perceived opportunity in adulthood was positively correlated w ith perceived opportunity, family bonding, relationship satisfaction and social conformity in adolescence; and, perceived opportunity, marital satisfaction and job satisfaction in adulthood. Often conceptualized as “future orientation” in the literature, both alcoholism (Lennings, 1996; Amodeo et al., 1992) and depression (Stein, Newcomb & Bentler, 1992; Beck et al., 1979; Petrie & Chamberlain, 1983) have been found to be negatively related to perceived opportunity. The results obtained are consistent with the hypothesis set forth in the beginning o f this investigation, namely, that the combined effects o f alcohol and depression will have an additive effect and be a stronger negative predictor of perceived opportunity than alcoholism or depression alone. However, a change in the level overtime is required for the impact to occur. Job satisfaction. Job satisfaction in adulthood was decreased by alcohol dysphoria as hypothesized. Alcohol dysphoria in young adulthood decreases job satisfaction more than other constructs under investigation. Also, alcohol problems in young adulthood were found to decrease job satisfaction. In addition, several specific effects were found to predict job satisfaction. A specific effect o f the perceived opportunity construct, namely, “satisfaction with future” in adolescence, was found to increase job satisfection in adulthood, hi addition, use o f alcohol at work or school during Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 young adulthood negatively impacts “happiness at work”, while depression as measured by the CES-D in young adulthood increases “trouble at work”. Job satisfaction was negatively correlated with alcohol dysphoria a t all time periods. It was positively correlated with perceived opportunity, fam ily bonding, and social conformity in adolescence and perceived opportunity, m arital satisfaction, relationship satisfaction and job instability in adulthood. If in young adulthood, one experiences comorbid problems w ith alcohol and depression, it is likely to lead to a decreased sense o f job satisfaction in adulthood. Although alcohol problems alone are related, alcohol dysphoria it is more likely to decrease job satisfaction. Further, if in adolescence one experiences an increased sense o f satisfaction with the future, it can have the positive effect o f having an increased sense o f happiness with work in adulthood. Job instabilitv. The hypothesis that alcohol dysphoria would predict decreased job stability was not supported; however, certain effects were noted. Quantity o f alcohol used during young adulthood increases job instability; and, dysphoria in young adulthood increases the “number o f times fired”. Therefore, increased quantity o f alcohol consumed in young adulthood will increase number o f times fired, decrease the chance o f having a current job, and increase the probability o f collecting unemployment Job Instability was positively correlated with social conformity in adolescence, alcohol dysphoria in young adulthood, and divorce in adulthood. In conclusion, as can be seen a number o f significant relationships exist between the variables and factors under study. Alcohol dysphoria in young adulthood has substantial negative effects as the individual matures into adulthood. For the women in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 0 this sample, it appears that alcohol (fysphoria will lead to decreased functioning in many domains o f life. Predictors o f Alcohol Problems and Dysphoria Alcohol problems. M ultiple pathways influence the development o f alcohol problems later in life. Alcohol (fysphoria in adolescence increased alcohol problems in adulthood. Having a positive bond to one’s family in adolescence, decreases the quantity o f alcohol used in adulthood, and having good relationship with one’s parents decreases alcohol problems and the use o f alcohol at school or work in young adulthood. Satisfaction with school or w ork in adolescence reduced alcohol problems in adulthood. Peer relations are also important, as having good relationship with peers in adolescence decreases use o f alcohol at school or work in adulthood. Further, having socially conforming attitudes in adolescence reduces alcohol-related problems in adulthood In young adulthood depression not associated with alcohol problems significantly increased the use o f alcohol at school and work, the fi~equency o f alcohol consumed and alcohol- related problems in adulthood. Alcohol problems in adolescence, young adulthood and adulthood were correlated with many o f the psychosocial factors under investigation. In adolescence, alcohol problems were positively correlated with dysphoria and later alcohol problems, and negatively correlated with perceived opportunity, family bonding and social conformity. Alcohol problems in young adulthood were positively correlated with dysphoria, and later problems with alcohol, and negatively correlated with perceived opportunity, fam ily bonding and social conformity in adolescence, and relationship satisfaction, job satisfaction and job instability in adulthood. Adult alcohol problems were positively Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 correlated with alcohol problems and dysphona earlier and later in life, and negatively correlated with perceived opportunity, fam ily bonding, parental divorce and social conformity in adolescence, and marital satisfaction and perceived opportunity in adulthood. There is no single path that leads to the development o f alcohol problems. Rather, numerous factors are interrelated including mood-state, family environment, attitudes o f social conformity, future outlook and relationship satisfaction. This is consistent with the view in the literature that alcohol problems are multifaceted (Hawkins et al., 1992; Zucker & Gomberg, 1986; Tarter & Vanyukov, 1994). The current investigation shows that a female is prone to develop alcohol problems if, earlier in life, they have a depressed mood-state, poor family environment, beliefs that do not conform with social norms, as well as a poor sense o f satisfaction with relationships and a negative outlook on the future. Dysphoria. As with alcohol problems, several pathways lead to the development o f dysphoria. Family bonding during adolescence decreases depression scale scores in adulthood. In adolescence, several specific effects exist. Being satisfied with what you want to be decreases dysphoria in young adulthood. Also, being satisfied in relationships during adolescence decreases later suicidal ideation and depression as measured by the CES-D in adulthood. Further, satisfaction with relationships with the opposite sex in adolescence decreases dysphoria in adulthood. Religiosity in adolescence decreases depression scale scores in adulthood. Alcohol problems during young adulthood increased depression scale scores in adulthood. Quantity o f alcohol consumed and alcohol-related problems in young Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92 adulthood increased depression as measured by the CES-D in adulthood. Suicidal ideation during young adulthood increases dysphoria during adulthood. Dysphoria in adolescence, young adulthood and adulthood correlated with many o f the psychosocial variables under investigation. Dysphoria across tim e positively correlated with both alcohol problems and dysphoria across time. Dysphoria at all time points negatively correlated with, perceived opportunity, relationship satisfaction, family bonding and social conformity in adolescence. In adulthood it correlated with perceived opportunity as well as marital, relational and job satisfaction. Dysphoria in young adulthood was also correlated with divorce in adulthood. As with alcohol problems, there is no single path leading to depressive symptoms in later life. This finding is consistent with the literature reviewed (Gotlib, 1992; Marcus & Nardone, 1992; Segrin & Dillard, 1992; Billings, Cronkite, & Moos, 1983). Depression was shown to result from earlier alcohol use, poor family bonding, diminished satisfaction with relationships, attitudes o f social conformity and poor future outlook during adolescence/' Theoretical Implications The results o f this investigation have important theoretical implications for understanding the interrelationships between alcoholism, affect and psychosocial aspects o f adolescent and adult functioning. Theories to be discussed in this section include Zucker’s negative affect alcoholism (Zucker 1987; 1994), Del Boca’s intemalizer vs. extem alizer distinction (Del Boca, 1994), the self-medication hypothesis, the self derogation theory, and Beck’s (1987) cognitive theory o f depression. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 Zucker^s theory o f negative affect alcoholism received support in this investigation. He hypothesized four different alcoholisms (Zucker, 1987; 1994), antisocial alcoholism, developmentally cumulative alcoholism , developmentally lim ited alcoholism and negative affect alcoholism. He characterized negative affect alcoholism as more frequently occurring in women of middle class status with the typical sequela including dissatisfaction in social relationships w ith peers, in marriage and at work. This pattern was seen in the women in this stu(fy who had alcohol dysphoria. Alcohol dysphoria in young adulthood lead to decreased satisfection in relationships, marriage and employment, as well as a diminished sense o f perceived opportunity. W hile this concept is clearly supported by the data, it should be noted that alcohol dysphoria does not account for all o f the total alcohol problems experienced by these women. Zuckeris developmentally cumulative and developmentally lim ited alcoholisms, also thought to occur in women (but more frequently men) deserve attention and should be more fully explored. The antisocial type o f alcoholism, thought to occur primarily in men, may explain the gender effect noted in this study for the adolescent period. Future research might include an antisocial behavior latent construct, and utilize a combined or a male only sample. The distinction between those with an internalized style o f coping and those with an externalized style has been made by Del Boca (1994). Thought to be more prevalent among female alcoholics, those characterized as intem alizers displayed symptoms o f depression and anxiety, engaged in relief drinking, and were severely alcohol dependent with consequent physical problems. In this study, the intem alizer style o f coping, as manifest by the symptoms o f dysphoria, is supported. W omen with alcohol dysphoria Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94 have, by definition, comorbid symptomatology o f alcohol use and depression. This can be contrasted with more external behavior o f acting out and antisocial behavior, thought to be more prevalent in men. Depression or dysphoric symptoms are related to a portion, but not all o f the alcohol use in this sample o f women. I f anxiety^ symptoms were included, perhaps more o f the alcohol use would have been accounted for by this affectively- driven style o f alcoholism. A more comprehensive theory may need to be developed to explain coping style in women alcoholics that goes beyond being an “intemalizer^’. People who experience psychological, relational, social and occupational difficulties may abuse substances, in part, as self-medication to help ameliorate distress (Khantzian, 1985). The findings o f this study, showing a temporal constancy and mutually-predictive relationship between alcohol use and depression, lend support to this hypothesis. In addition, the high levels o f correlation o f alcohol dysphoria, alcohol use, and depression with the adult and adolescent constructs under investigation suggest that women in this study may self-medicate with alcohol to cope w ith distress in various psychosocial domains across time. The theory o f self-derogation, advanced by Kaplan (Kandel, 1980), received support in this investigation. According to this theory, people engage in deviant behavior in order to restore a sense o f self that has been damaged by prior devaluing experiences. In this investigation it was shown that poor bonding with one’s fam ily and little satisfaction in adolescence predicted alcohol use and depression. Beck’s cognitive theory o f depression received partial support in this study, indicating a potential systematic bias in information processing. According to Beck Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 (1987), depressed people tend to have a negative view o f the self^ the world, and the future (negative triad) and tend to perceive the self as inadequate, deserted and worthless (Beck & Weishaar, 1989). In this study, depression was negatively correlated with many o f the psychosocial processes under investigation. Since the measures were based on the subjects perception, responses may have been influenced by some o f the bias Beck proposes. Also, those with depression in adolescence may have experienced a self- fulfilling prophecy. Depression not associated with alcohol use in adolescence predicted decreased satisfaction in marriage, relationships and perceived opportunity in adulthood. As these depressed individuals saw the world in adolescence, so it became. Implications for Intervention and Clinical Practice The results o f this research yield a number o f significant implications. This analysis of the psychosocial pathways leading to alcohol use, depression and alcohol dysphoria and, the long-term outcomes, raises various issues in diagnosis, prevention and treatment that are important for clinicians to consider. First, it was determined that alcohol problems and depression have a strong, synergistic effect, which I have labeled alcohol dysphoria. Factors demonstrated to be predictors o f alcohol use and depression include poor family environment, non- conforming social attitudes, decreased sense o f perceived opportuni^ and decreased satisfaction in relationships. Alcohol dysphoria was found to be a strong predictor o f later problems in marital adjustment, a decreased sense o f perceived opportunity, and decreased satisfaction in interpersonal relationships and employm ent Consistent with existing research (& w kins et al., 1992), significant risk foctors or predictors were uncovered in this investigation. Because o f the significant impact o f Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 6 alcohol dysphoria, broadening the scope o f alcohol or depression screening and/or assessment in adolescence is important to consider. Alcohol use and depression should not be considered to occur in isolation or in a vacuum. Rather, they should be considered as part o f a complex set o f relationships that occur over the life-course. In adolescence, diagnostic efforts should not ju st look at whether or not the child is depressed o r using alcohol. A more thorough clinical investigation, looking at family environment, peer relationships, attitudes o f social conformity and sense o f perceived opportunity, would better establish the level o f risk for the development o f problems (Hawkins et al., 1992). It may be particularly salient when working with female adolescents, given the gender effect shown. Del Boca and Hesselbrock (1996) determined that female adolescents are often referred for alcohol treatment at a later time than male adolescents, and often their problems are more severe. They suggest that this delay may be due, in part, to expectations o f clinicians (W ierzbicki, 1993) that females are less likely to become involved in alcohol use; and, in part, to the tendency for women to internalize problems thus making them less visible. Alcoholism in women is socially perceived much more negatively than alcoholism in men. Clinical work and treatment availability are influenced by societal attitudes, encouraging some females to internalize the belief o f “feminine purity”, serving as both a destructive and a protective force in women alcoholics (Blume, 1990). One o f the more destructive forces which stem from these beliefs is that it may discourage alcoholic women from seeking treatment and, in fact, it encourages keeping the problem hidden (Jung, 1994; Lex, 1991). The ratio o f males to females in treatment is at least 4 or 5 to 1. Additionally, women alcoholics in the general health care delivery system often do Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 not have their alcoholism problems addressed (Blume, 1990). Medical and other professional sta£f may be less likely to diagnose or treat women alcoholics, because alcoholism may be seen as more congruent with male patients (Beckman & Amaro, 1984). Lesser representation in treatment does not necessarily mean that women benefit less firom treatm ent than men do. In fact, Nathan and Skinstad (1987) suggest that although women seek treatment at lower rates, men and women benefit firom treatm ent at comparable rates. Traditional treatment approaches may need to be tailored to fit the need to the female alcoholic because the etiology, risk factors, manifestation and course o f alcoholism is different for men and womerL Jung (1994) suggests that fam ily therapy, group counseling, separate all-female groups, and the use o f female therapists might be more appropriate for female alcoholics. Beckman (1994) raises some specific considerations as she describes what makes alcoholism treatment approaches appropriately female-oriented. It is delivered in a setting that is compatible with women's interactional styles and personal orientations. . . [such as] the need for and responsiveness to social relationships . . . It takes into account gender roles, fem ale socialization and women's place in society. . . It does not exploit w om en...nor does it support passive, dependent roles for women . . . It addresses women-specific treatm ent issues, (pp.206-7) She goes on to describe three types o f women-specific treatment components: 1) those addressing issues unique to women regarding female physiology, sexuality and reproductive functions, as well as gendered aspects o f psychological development and growth (see also Spampneto & Wadsworth, 1996); 2) those addressing issues o f the unique etiological characteristics o f female alcoholics such as increased incidence o f sexual assault and physical abuse; 3) those addressing issues that are common to both Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 8 women and men such as dual-diagnosis issues and polysubstance abuse. Expanding on these three categories, Beckman (1994) describes some specific components o f women- oriented alcoholism treatm ent She calls for broad and comprehensive services including treatment for other problems such as incest, sexual assault, and other mental health problems, health services, fam ily services, and services for children. She emphasizes the importance o f the development o f parenting skills, the development o f social roles, positive relationships, and social support the development o f self-esteem and adaptive coping mechanisms. Employment or vocational counseling and legal assistance should be available when needed and women-specific support groups are needed for afiercare. Similarly, clinicians may need to expand their view o f the dual-disordered patient In adolescents, treatment may be expanded to include family counseling, and focus on fostering the development o f social-support networks and positive peer relationships. Psychoeducational efforts to raise awareness o f long-term consequences o f alcohol dysphoria may be helpful. Once alcohol dysphoria is established, it is likely to persist, especially between young adulthood and adulthood. As this investigation has found, alcohol dysphoria has significant consequences, as it impairs or impacts numerous psychosocial domains in adulthood. Treatment efforts geared toward adolescents and young adults may successfully interrupt the development o f alcohol dysphoria and decrease the long-term consequences. As noted above, treatm ent efforts need to be multi dimensional, incorporating biological, psychological and social components (Zucker & Gromberg, 1986). This may be at odds with much o f contemporary psychopharmacological and short-term, narrowly focused treatm ent efforts, but a broader definition o f treatment practices is indicated. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 Limitations and Research Implications While this study goes beyond much o f the existing research and has a high level o f generalizabili^ to the com m unia population, several limitations exist including: general problems with self-report data, a lack o f identical measures across time, a lack o f biological measurements, and the exclusion o f m ales in the final models. Each o f these issues will be briefly discussed below. First, the use o f self-report measures without collateral validation always raises reliability and validity concerns. Measurement error may be inflated because o f self- presentation bias, hindsight bias, variations in the testing environment, and the potential o f multiple interpretations o f the questions (Isaac & Michael, 1995). Although statistically controlled for in this study, a second limitation involves the lack o f identical measures across time. For the adolescent period, the latent construct o f alcohol problems used measured variables (quantity o f alcohol used, use o f alcohol at school/work, firequency o f alcohol consumption and times "super-high" on alcohol) that were different from those in young adulthood and adulthood (quantity o f alcohol used, use o f alcohol at school or work, frequency o f use, and alcohol-related problems scale). The alcohol-related problems scale was not part o f the survey for the first wave o f data used, but was in later tim e periods. Further, relationship satisfaction was assessed in adolescence by using measures of satisfaction w ith opposite-sex friends, satisfaction with close fiiends and good relationships with peers; in adulthood, it was assessed by the same “satisfaction with close friends” and “good relationships with peers” as in adolescence, the UCLA loneliness scale was added in adulthood. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 Alcoholism is a biopsychosocial phenomena (Zucker & Gomberg, 1986), and, as such, relevant biological factors should be mcluded in future research. The inclusion o f neuropsychological/neurophysiological data would allow for a m ore thorough evaluation o f the processes and constructs described. Cortical substrates that may be impacted by alcohol use include processes associated w ith frontal or executive frmctioning and memory (M iller, 1991). Deficits in these processes also impact the person’s ability to plan, motivate, organize and sequence activities. It is logical to assum e that these cognitive impairments would influence how the individual functions in their relational, marital, social and work environments. Because alcohol use and depression in adolescence were not related for the males in this sample, final analyses were run on the females only. W hile this analysis provided valuable insight into the roles, predictors and consequences o f alcohol and dysphoria in women, it appears that the developmental course for men may be different Future research is needed to more fully examine this relationship and increase the generalizability o f results. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 REFERENCES American Psychiatrie Association (1994). Diagnostic and Statistical M anual o f Mental Disorders. Fourth Edition. Washington, DC: American Psychiatric Association. Amodeo, M ., Kurtz, N., & Cutter, H. S. (1992). Abstinence, reasons for not drinking, and life satisfaction. International Journal o f the Addictions. 27(6). 707-716. Assh, S. D., & Byers, E. S. (1996). Understanding the co-occurrence o f marital distress and depression in women. Journal o f Social & Personal Relationships. 13141 537-552. Barbor, T. F., Hofinann, M ., Del Boca, F., Hesselbrock, V., M eyer, R_, Dolinsky, Z., & Rounsaville, B. (1992). Types o f alcoholics. I: Evidence for an em pirically derived typology based on indicators o f vulnerability and severity. Archives o f General Psychiatry. 49. 599-608. Barbor, T. F. (1996). The classification o f alcoholics typology theories from the 19^ century to the present Alcohol Health & Research World. 20(1). 6-14. B arnet P A., & Gotlib, I. H (1988). Psychosocial functioning and depression: Distinguishing among antecedents, concomitants, and consequences. Psychological Bulletin. 104. 97-126. Beck, A. T. (1987). Cognitive therapy. In J. K. Zeig (Ed.), The Evolution o f Psychotherapy (pp. 149-178). New York: Bruner/Mazel. Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive Therapy o f Depression. New York: Guilford Press. Beck, A. T., & W eishaar, M (1989). Cognitive therapy. In R. J. Corsini & D. Wedding (Eds.) Current Psvchotherapies. (4"^ ed.), (pp.285-320). Itasca, Illinois: Peacock Publishers. Beckman, L. J. (1994). Treatment needs o f women with alcohol problems. Alcohol Health & Research World. 118)3.206-211. Beckman, L. J., & Amaro, H. (1984). Patterns o f women’s use o f alcoholism treatment agencies. In S. C. W ilsnack & L. J. Beckman (Eds.), Alcohol problems in women: Antecedents, consequences and intervention (pp.319-348). New York: Academic Press. Bentler, P. M. (1992). On the fit o f models to covariances and methodology to the Bulletin. Psychological Bulletin. 112.400-404. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 Billings, A. G., Cronkite, R. C., & Moos, £L H. (1983). Social-environmental factors in unipolar depression: Comparisons o f depressed patients and nondepressed controls. Journal o f Abnormal Psychology. 92(11.119-133. Blume, S. B. (1990). Chemical dependency in women: Im portant issues. American Journal o f Drug and Alcohol Abuse. 16(3& 4\ 297-307. Boone, K. B. (1998). Nemroosycholoev o f Depression. Paper presented at the Neuropsychology Seminar, Harbor-UCLA Medical Center, Torrance, CA. Boone, K. B., Lesser, I. M , M iller, B. L., Wohl, M , Berman, N ., Lee, A., Palmer, B., & Back, C. (1995). Cognitive functioning in older depressed outpatients: Relationship o f presence and severity o f depression to neuropsychological test scores. Neuropsychology. 9(31.390-398 Brook, J. S., Lukoff I. P., & Whiteman, M. (1980). Initiation into adolescent marijuana use. Journal o f Genetic Psychology. 137(1). 133-142. Brown, G. W., Bifulco, A., Harris, T., & Bridge, L. (1986). Life stress, chronic subclinical symptoms and vulnerability to clinical depression. Journal o f Affective Disorders. 11. 1-19. Brown, S. A., & Schuckit, M A. (1988). Changes in depression among abstinent alcoholics. Journal o f Studies on Alcohol. 49(5). 412-417. Burch, E. A. (1994). Suicide attem pt histories in alcohol-dependent men: Differences in psychological profiles, hitemational Journal o f the Addictions. 29(11), 1477-1484. Byrne, B. M (1994). Structural Equation Modeling with EOS and EQS/Windows Basic Concepts. Applications, and Programming. Newbury Park, CA: Sage Publishing. Cloninger, C. R. (1987). Neurogenetic adaptive mechanisms in alcoholism. Science. 236.410-416. Cook, B. L., & Winokur, G. (1995). Complications: Mental disorders. In J.H. Jaffe (Ed.) Encyclopedia o f Drugs and Alcohol: Vol. 1. (pp.295-300). New York: M acmillan Library Reference. Cornelius, J. R., Salloum, I. M , Mezzich, J., Cornelius, M D., Fabrega, H., Ehler, J. G., Ulrich, R. P., Thase M. E., & Mann, J. J. (1995). Disproportionate suicidality in patients with comorbid major depression and alcoholism. American Journal o f Psychiatry. 152131.358-364. Cotton, N.S. (1979). The fam ilial transmission o f alcoholism: A review. Journal o f Studies on Alcohol. 40. 89-116. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103 Daley, D. C., Moss, H., & Campbell, F. (1993). Dual disorders: Counseling clients with chemical dependency and mental illness (2nd ed.). Center City, NGnnesota: l^ e ld e n Educational M aterials. Del Boca, F. K. (1994). Sex, gender, and alcoholic typologies. In T.F. Barbor, V. Hesselbrock, R.E. Meyer & W.J. Shoemaker (Eds). Types o f Alcoholics: Evidence From Clinical. Experimental, and Genetic Research. (Vol. 708, pp. 34-48) New York: Annals o f the New York Academy o f Sciences. Del Boca, F. K., & Hesselbrock, M. N. (1996). Gender and alcoholic subtypes. Alcohol Health & Research World. 20(1). 56-62. Deykin, E. Y , Levy, J. C., & Wells, V. (1987). Adolescent depression, alcohol and drug abuse. American Journal o f Public Health. 77.178-182. Dorus, W., Ketmedy, J., Gibbons, R., & Ravi, S. D. (1987). Symptoms and diagnosis o f depression in alcoholics. Alcoholism: Clinical and Experimental Research. I L 150-154. Fillmore, K. M (1987). Women’s drinking across the adult life course as compared to men’s. British Journal o f Addiction. 82. 801-811. Forrest, G. G. (1983). Alcoholism. Narcisism and Psychopathology. Springfield, Illinois: Charles C. Thomas. Frank, E., & Spanier, C. (1995). Interpersonal psychotherapy for depression: Overview, clinical efficacy, and future directions. Clinical Psychology: Science & Practice. 2(4). 349-369. Gainotti, G , Caltagirone, C., & Zoccolotti, P. (1993). Left/right and cortical/subcortical dichotomies in the neuropsychological study o f human emotions. Cognition and Emotion. 7 .71-93. Garcia-Preto, N. (1996). Latino families: An overview. In M. McGoldrick, J. Giordano & J. K. Pearce (Eds.) Ethnicity and Family Therapvf2°^ ed.) (ppl41-154). New York: The Guilford Press. Gerbing, D. W., & Anderson, J. C. (1984). On the meaning o f within-factor correlated measurement errors. Journal o f Consumer Research. 1111). 572-580. Gotlib, I. H. (1983). Perception and recall o f interpersonal feedback: Negative bias in depression. Cognitive Therapy and Research. 7 .399-412. Gotlib, I. H. (1992). Interpersonal and cognitive aspects o f depression. Current Directions in Psychological Science. 1(5). 149-154. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104 Gotlib, L H., & Lee, C. M L (1989) The social flinctiomng o f depressed patients; A longitudinal assessm ent Journal o f Social & Clinical Psychology. 8(3’ ). 223-237. Greenberg, E. S., & Grunberg, L. (1995). Work alienation and problem alcohol behavior. Journal o f Health and Social Behavior. 36. 83-102. Hawkins, J. D., Catalano, R. F., & M iller, J. Y. (1992). Risk and protective Actors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psvcholoeical Bulletin. 1121 I t 64-105. Heath, A C. (1995). Genetic influences on alcoholism risk: A review o f adoption and twin studies. Alcohol Health & Research World. 20(31166-171. Helzer, J. E., Canimo, G. J., Hwu, H. G., etal. (1988). A cross-national comparison o f population survey w ith the diagnostic interview schedule, hi R. M. Rose & I. E. Barret (Eds.) Alcoholism: Origins and Outcomes.tpp.31-48^ New York: Raven Press. Helzer, J. E., & Pryzbeck, T. R. (1987). The co-occurrence o f alcoholism with other psychiatric disorders in the general population and its impact on treatm ent. Journal o f Studies on Alcohol. 49(31219-224. Herz, L. R., Volicer, L., D Angelo, N., & Gadish, D. (1990). Additional psychiatric illness by Diagnostic Interview Schedule in male alcoholics. Comprehensive Psvchiatrv. 3 0 .72-79. Hesselbrock, V. M., Hesselbrock, M. N., & Workman-Daniels, K. I. (1986). Effect o f major depression and antisocial personality on alcoholism: Course and motivational patterns. Journal o f Studies on Alcohol. 47131.207-212. Hingson, R. T. Mangione, T., & Barrett, J. (1981). Job characteristics and drinking practices in the Boston m etropolitan area. Journal o f Studies on Alcohol. 42. 725-738. Hundleby, J. D. & Mercer, G. W. (1987). Family and friends as social environments and their relationship to young adolescents' use o f alcohol, tobacco, and marijuana. Journal o f Marriage & the Family. 49(11151- 164. Isaac, S. & Michael, W. B. (1995). Handbook in Research and Evaluation for Education and the Behavioral Sciences (3”* ed ). California: Educational and Industrial Testing Services. Jacob, T., Dunn, N. J., & Leonard, K. (1983). Patterns o f alcohol abuse and family stability. Alcoholism: Clinical & Experimental Research. 7(4). 382-385. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105 Joiner, T. E., Katz, J., & Lew, A. S. (1997). Self-verification and depression among youth psychiatric inpatients. Journal o f Abnormal Psychology. 106f4L 608-618. Julien, R. M. (1996). A Prim er o f Drug Action: A Concise. Nontechnical Guide to the Actions. Uses, and Side Effects o f Psychoactive Drues. New York: Freeman. Jung, J. (1994). Under the Influence- Alcohol and Human Behavior. California: Brooks/Cole Publishing. Kaij, L. (1960) Alcoholism in Twins: Studies on the Etiology and Sequels o f Abuse o f Alcohol. Stockholm: Ahnqvist & W iksell International. Kandel, D. B. (1980). Drug and drinking behavior among youth. Annual Review o f Sociology. 6 .235-285. Kasch, K. L., & Klein, D. N. (1996). The relationship between age at onset and comorbidity in psychiatric disorders. Journal o f Nervous and Mental Disease. 1841 111. 703-707. Keeler, M. t t , Taylor, C. L., & M iller, W. C. (1979). Are all recently detoxified alcoholics depressed? American Journal o f Psychiatry. 136. 586-588. Kessler, R. C., Nelson, C. B., McGonagle, K. A., Edlund, M. J., Frank, R. G., & Leaf, P. J. (1996). The epidemiology of co-occurring addictive and mental disorders: Implications for prevention and service utilization. American Journal o f Orthopsychiatry. 66{l), 17-31. Khantzan, E. J. (1985). The self-medication hypothesis o f addictive disorders: Focus on heroin and cocaine dependence. The American Journal o f Psychiatry. 142(11), 1259-1264. LaRosa, J.H. (1990) Executive women and health: Perceptions and practices. American Journal o f Public Health. 8 0 .1450-1454. Lee, E. (1996). Asian American families: An overview. In M. McGoldrick, J. no & J. K. Pearce (Eds.) j New York: The Guilford Press. Giordano & J. K. Pearce (Eds.) Ethnicity and Fam ily Therapy (2“^ ed.) (pp 227-248), Lennings, C. J. (1996). Self-efBcacy and tem poral orientation as predictors o f treatment outcome. Alcoholism Treatment Quarterly. 1414). 71-79. Lex, B. W. (1991). Some gender difTerences in alcohol and polysubstance users. Health Psychology. 10(21121-132. Lisansky-Gomberg, E. S. (1994). Risk factors for drinking over a woman’s life span. Alcohol Health & Research World. (18)3.220-227. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 Maier, W., & Merikangas, K- (1996). Co-occurrence o f cotransmission o f affective disorders and alcoholism in èunilies. British Journal o f Psvchiatrv. 168. (suppl.- 30), 93-100. Mangione, T. W., & Quinn, R. P. (1975). Job satisfaction, counterproductive behavior, and drug use at work. Journal o f Applied Psvcholoev. 60.114-116. Manning, W. G., Keeler, E. B., Newhouse, J. P., Sloss E. M., & Wasserman, J. (1991). The Costs o f Poor Health Habits. Cambridge, MA: Harvard University Press. Marcus, D . K., & Nardone, M. E. (1992). Depression and interpersonal rejection. Clinical Psvcholoev Review. 12f4 1 433-449. Maxmen, J. S. & Ward, N. G. (1995). Essential Psvchopatholoev and its Treatment (2“* ^ ed). New York: Norton & Company. McCabe, S. B., & Gotlib, I. H. (1993). Interactions o f couples w ith and without a depressed spouse: Self-report and observations o f problem-solving situations. Journal o f Social & Personal Relationships. 10(41.589-599. McGee, L. & Newcomb, M. D. (1992). General deviance syndrome: Expanded hierarchical evaluations at four ages from early adolescence to adulthood. Journal of Consulting & Clinical Psvcholoev. 60(51.766-776. McKay, J. R., Longabaugh, R. Beattie, M. C. Maisto, S. A., & Noel, N. E. (1992). The relationship o f pretreatment family functioning to drinking behavior during follow- up by alcoholic patients. American Journal o f D rue and Alcohol Abuse. 18(4). 445-460. McKenna, C., & Ross, C. (1994). Diagnostic conundrums in substance abusers with psychiatric symptoms: Variables suggestive o f dual diagnosis. American Journal o f Drug and Alcohol Abuse. 20f41 397-412. Medora, N. P., & Woodward, J. C. (1991). Factors associated w ith loneliness among alcoholics in rehabilitation centers. Journal o f Social Psvcholoov. 13 If6). 769-779 Merikangas, K R., & Gelemter, C. S. (1990). Comorbidity for alcoholism and depression. Psvchiatric Clinics o f North America. 13(41613-632. Meyer, B. E., & Hokanson, J. E. (1985). Situational influences on social behaviors o f depression-prone individuals. Journal o f Clinical Psvcholoov. 41. 29-35. M iller, L. (1991). Predicting relapse and recovery in alcoholism and addiction: neuropsychology, personality, and cognitive style. Journal o f Substance Abuse Treatm ent 8(4). 277-91. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 0 7 Mullahy, J., & Sindelar, J. (1992). Effects o f alcohol on labor market success; Income, earnings, labor supply and occupation. Alcohol Health & Research World. 16(2). 134-139. Nathan, P. E., & Skinstad, A. H. (1987). Outcomes o f treatm ent for alcohol problems: Current methods, problems and results. Journal o f Consulting and Clinical Psvcholoev. 55(3). 332-340. National Institute on Alcohol Abuse and Alcoholism (1991). Alcohol alert: Alcoholism and co-occurring disorders. (No. 14 PH 302). National Institute on Drug Abuse (1997). Capsule Report: Women and Drug Abuse rWeb Page). httpy/wwwnidanih.gov/. Newcomb, M. D. (1990). What structural modeling techniques can tell us about social support. In LG. Sarason, B. R. Sarason & G.R. Pierce (Eds.)Social Support: An Interactional View. (pp26-63). New York: Wiley. Newcomb, M. D. (1992). Understanding the multidimensional nature o f drug use and abuse: The role o f consumption, risk factors, and protective factors. In M. Glantz & R. Pickens (Eds.) Vulnerabilitv to Drue Abuse. Washfogton DC: American Psychological Association. Newcomb, M. D. (1993). Effects o f adolescent drug use on adult mental health: A prospective study o f a community sample. Experimental and Clinical Psvchopharmacologv. 1(1-4). 215-241. Newcomb, M. D. (1994). Drug use and intimate relationships among women and men: Separating specific from general effects in prospective data using structural equation models. Journal o f Consulting and Clinical Psvcholoev. 62(3). 463-476. Newcomb, M. D. (1995). Prospective dynamics o f intoxication in the workplace: Personal and job-related predictors and consequences. Experimental and Clinical Psvchopharmacologv. 3(1), 56-74. Newcomb, M. D. (1997a). Psychosocial predictors and consequences o f drug use: A developmental perspective within a prospective study. Journal o f Addictive Diseases. 16(1), 51-89. Newcomb, M. D. (1997b), General deviance and psychological distress: Impact o f family support/bonding over 12 years from adolescence to adulthood. Criminal Behavior and Mental Health. 7 .368-400. Newcomb, M. D., & Bentler, P. (1986). Loneliness and social support: A confirmatory hierarchical analysis. Personalitv and Social Psvchologv Bulletin. 12114). 520-535. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 Newcomb, M. D., & Bentler, P. (1988). Consequences o f adolescent drug use: Impact on the lives o f voung adults. N e \^ u ry Park, CA: Sage Publication. Newcomb, M. D., Huba, G. J., & Bentler, P. M. (1981). A multidimensional assessment o f stressful life events among adolescents: Derivation and correlates. Journal o f Health and Social Behavior. 2 2 .400-415. Newcomb, M. D., & Jack, L. (1995). Drug use, agency, and communality: Causes and consequences among adults. Psvcholoev o f Addictive Behaviors. 9 (1). 67-82. Newcomb, M. D., Maddahian, E., & Bentler, P. M. (1986). Risk factors for drug use among adolescents: Concurrent and longitudmal analyses. American Journal o f Public Health. 76151 525-531. Newcomb, M D., & Rickards, S. (1995). Parent drug-use problems and adult intimate relations: Associations among community samples o f young adult women and men. Journal o f Cnunseline Psvcholoev. 42(2). 141-154. Newcomb, M. D., Scheier, L. M., & Bentler, P. M. (1993). Effects o f adolescent drug use on adult mental health: A prospective study o f a community sample. Experimental and Clinical Psvchopatholoev. 111-41215-241. Noel, N. E ., McCrady, B. S., Stout, R. L., & Fisher-Nelson, H. (1991). Gender differences in m arital functioning o f male and female alcoholics. Family Dvnanucs o f Addiction Ouarterlv- H4Y 31-38. Norris, J. (1994). Alcohol and fem ale sexuality. Alcohol Health & Research World. 18(31197-201. Nunes, E. N ., Quitkin, F., & Berman, C. (1988). Panic disorder and depression in female alcoholics. Journal of Clinical Psvchiatrv. 4 9 .441-443. OTarrell, T. J., & Birchler, G R. (1987). M arital relationships o f alcoholic, conflicted, and nonconflicted couples. Journal o f Marital & Family Theraov. 13(3). 259- 274. OTarrell, T. J., Choquette, K. A., & Birchler, G. R. (1991). Sexual satisfaction and dissatisfaction in the marital relationships o f male alcoholics seeking marital therapy. Journal o f Studies on Alcohol. 52(51.441-447. O'Malley, S. S., & Maisto, S. A. (1985). Effects o f family drinking history and expectancies on responses to alcohol in men. Journal o f Studies on Alcohol. 46141289- 297. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109 Paniagua, F. A. (1994). Assessing and Treating C ulturally Diverse Clients- A Practical Guide. Newbury Park, CA: Sage Publications. Parker, D. A., & Harford, T. C. (1992). The epidemiology o f alcohol consumption and dependence across occupations in the United States. Alcohol Health & Research World. 16(2). 97-105. Patterson, D.G. (1995). Alcoholism: Treating the "second sex.". Irish Journal o f Psychological Medicine. 12(2). 46-47. Peele, S. (1987). W hat can we expect from treatm ent o f adolescent drug and alcohol abuse? Pediatrician: D rug and Alcohol Abuse in Children and Adolescence. 14. 62-69. Petrie, K., & Chamberlain, K. (1983). Hopelessness and social desirability as moderating variables in predicting suicide behavior. Journal o f Consulting and Clinical Psvchologv. 5 1 .485-487. Phil, R.O., & Peterson, J.B. (1991). Attention-deficit hyperactivity disorder, childhood conduct disorder, and alcoholism: Is there an association? Alcohol Health & Research World. 15(11.25-31. Popkin, S. J., Gallagher, D ., Thompson, L. W., & M oore, M. (1982). Memory complaint and performance in normal and depressed older adults. Experimental Aging Research. 8 .141-145. Radloff, L. S. (1977). The CES-D scales: A self-report depression scale for research in the general population. Applied Psychological M easurem ent 1 .385-401. Regier, D. A., Farmer, M. E., Rae, D. S., Locke, B. Z., Keith, S. J., Judd, L. L., & Goodwin, F. K. (1990). Comorbidity o f mental disorders w ith alcohol and other drug abuse: Results from the epidemiologic catchment area (ECA) study. Journal o f the American Medical Association. 264(19). 2511-2518. Reich, T., Cloninger, C.R., Van Eerdewegh, P., R ice, J.P., & Mullaney, J. (1988). Secular trends in the familial transm ission o f alcoholism. Alcoholism: Clinical and Experimental Research. 12.458-464. Rosenbaum, M., Lewinsohn, P. M , & Gotlib, I. H. (1996). Distinguishing between state-dependent and non-state-dependent depression-related psychosocial variables. British Journal o f Clinical Psychology. 35(31.341-358. Rounsaville, B. J., W eissman, M. M , PrusofiF, B. A ., & Herceg-Barton, R. L. (1979). Marital disputes and treatm ent outcome in depressed women. Comprehensive Psvchiatrv. 2 0 .483-490. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. no Roy, A. (1996). Aetiology o f secondary depression in male alcoholics. British Journal o f Psvclu'atrv. 169.753-757. RusseU, D., Peplau, L., & Cutrona, C. (1980). The revised UCLA Loneliness Scale: Concurrent and descriminant validity evidence. Journal o f Personalitv and Social Psvchologv. 3 9 .472-480. Russo, J., Roy-Byme, P., Reeder D., & Alexander, M., et al. (1997). Longitudinal assessment o f quality o f life in acute psychiatric mpatients: Reliability and validity. Journal o f Nervous & Mental Disease. 185(3). 166-175, Schafi, M ., Lavely, R., & Jaffe, R. (1975). M edicationand the prevention o f alcohol abuse. American Journal o f Psvchiatrv. 132191 942-945. Scheier, L. M., & Newcomb, M D. (1993). Multiple dimensions o f affective and cognitive disturbance: Latent-variable models in a community sample. Psvcholoeical Assessment. 5, 230-234. Schmaling, K., & Becker, J. (1991). Empirical studies o f the interpersonal relations o f adult depressives. In J. Becker & A Kleinman (Eds.) Psvchosocial aspects o f depression, (pp. 169-185). New Jersey: Lawrence Erlbaum. Schuckit, M. A (1979). Alcoholism and affective disorder Diagnostic confusion. In D. W. Goodwin & C. K. Erickson (Eds.) Alcoholism and Affective Disorders: Clinical. Genetic and Biochemical Studies. New York: SP Medical & Scientific Books. Schuckit, M. A (1986). Genetic and clinical implications o f alcoholism and affective disorder. American Journal o f Psvchiatrv. 14312). 140-147. Schuckit, M. A. (1994). Alcohol and depression: A clinical perspective. Ygo Jahnsson Foundation VQI Medical Symposium: Depression: Preventive and risk factors. Acta Psvchiatrica Scandinavica. 8 9 0 77. Suppl), 28-32. Schuckit, M. A. (1995). A long-term study of sons o f alcoholics. Alcohol Health & Research World. 19(31172-175. Schrader, G. D. (1997). Does anhedonia correlate with depression severity in chronic depression? Comprehensive Psvchiatrv. 38151260-263. Segrin, C., & Dillard, J. P. (1991). The interactional theory o f depression: A meta analysis o f the research literature. Journal o f Social & Clinical Psvcholoev. 1 111). 43-70. Sheeran, P., Abrams, D., & Orbell, S. (1995). Unemployment, self-esteem and depression: A social comparison theory approach. Basic & Applied Social Psychology. 17(1-2) 65-82. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ill Siegel, B., Gurevich, D., & Oxenkrug, G. F. (1989). Cognitive impairment and cortisol resistance to dexamethasone suppression in elderly depression. Biological Psvchiatrv. 2 5 .229-234. Siegfried, K. (1985). Cognitive symptoms in late-life depression and their treatment. Journal o f Affective Disorders. 9f Suppl.). 533-540. Svikis, D. S., Velez, M L., & Pickens, R. L. (1994). Genetic aspects o f alcohol use and alcoholism in women. Alcohol Health & Research World. 18(31.192-196. Smith, S. S., & Newman, J. P. (1990). Alcohol and drug abuse-dependence disorders in psychopathic and nonpsychopatiiic crim inal offenders. Journal o f Abnormal Psvchologv. 99(41430-439. Sovani, A. (1989). A comparative study o f alcoholics and non-alcoholics along the parameters o f conformity and self-confidence. Indian Journal o f Applied Psvchologv. 26(2), 23-29. Spampneto, A. M., & Wadsworth, R. O. (1996). Addressing sexual issues in the treatment o f alcoholic women. American Treatment Ouarterlv. 1411 1. 1-9. Spanier, G. (1976). M easuring dyadic adjustment: N ew scales for assessing the quality o f marriage and sim ilar dyads. Journal o f M arriage and the Familv. 3 8 .15-28. Spar, J. (1996, September). Dual Disorders. Paper presented at the Substance Abuse Update- Training workshop. University o f California, Los Angeles School o f Medicine, Los Angeles, CA. Stacy, A. Newcomb, M D., & Bentler, P. (1991). Personality, problem drinkû% and drunk driving: Mediating, moderating and direct-effect models. Journal o f Personalitv and Social Psvchologv. 60151.795-811. Stein, J. A., Newcomb, M D., & Bentler, P. M. (1992). The effect o f agency and communality on self-esteem: Gender differences in longitudinal data. Sex Roles. 26(11/12), 465-483. Steinglass, P. (1981). The alcoholic femily at hom e: Patterns o f interaction in dry, wet, and transitional stages o f alcoholism. Archives o f General Psvchiatrv. 38151578- 584. Sue, D. W., & Sue, D. (1990). Counseling the Culturallv Different Theory & Practice (2nd ed.). New York: John Wiley & Sons. Suman, L. N., & Nagalakshmi, S. V (1995). Fam ily interaction patterns in alcoholic families. NIMHANS Journal. 13111.47-52. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 Tarter, R. E ., & Vanyukov, M. (1994). Alcoholism: A developmental disorder. Journal o f Counseling and Clinical Psvchologv. 62(61 1096-1107. Thompson, J. M., WhiflFen, V. E., & Blain, M . D. (1995). Depressive symptoms, sex and perceptions o f intimate relationships. Journal o f Social & Personal Relationships. 12(1), 49-66. Vaillant, G. E. (1993). Is alcoholism more often the cause or the result o f depression? Harvard Review o f Psvchiatrv.1(2). 94-99. VaiUant, G. E. (1995). The Natural Historv o f Alcoholism Revisited. Cambridge, Massachusetts: Harvard University Press. Vaillant, G. E., & Hiller-Sturmhofel, S. (1995). The natural history o f alcoholism. Alcohol Health & Research World. 20(3). 152-161. Weissman, M. M. (1987). Advances in psychiatric epidemiology: Rates and risks for depression. American Journal o f Public Health. 7 7 .445-451. Weissman, M. M., & Meyers, J. K. (1980). Clinical depression in alcoholism. American Journal o f Psvchiatrv. 137.372-373. W ierzbicki, M. (1993). Issues in Clinical Psvcholoev: Subiective Verses Objective Approaches. Massachusetts: Allyn & Bacon. W ilsnack, S.C., Wilsnack, R.W. & Hiller-Strumhofei, S. (1994). H ow women drink: Epidemiology o f women’s drinking and problem drinking. Alcohol Health and Research World. 18(3). 173-181. Wouldt B.D. & Bradley, J.R.(1996). Precursors, mediators and problem drinking: Path analytic models for men and women. Journal o f Drug Education. 26(1). 1-12. Zucker, R. A., Ellis, D. A., B in g h ^ , C. R., & Fitzgerald, H E. (1996). The development o f alcohol subtypes risk variation am ong alcoholic families during the early childhood years. Alcohol Health & Research W orld. 20(1). 46-54. Zucker, R. A., Fitzgterald, H. E., & M oses, H. D. (1994). Emergence o f alcohol problems and the several alcoholisms: A developmental perspective on etiologic theory and life course trajectory. In D. Cicchetti & D. J. Cohen (Eds.) Developmental Psvchopatholoev. Vol. 2: Risk. Disorder, and Adaptation, (pp. 677-709) N ew York: John Wiley & Sons. Zucker, R.A., & Gomberg, E.S. (1986). Etiology o f alcoholism reconsidered: The case for a biopsychosocial process. American Psvcholocist 4117). 783-793. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 113 Zucker, R. A. (1987). The four alcoholisms: A developmental account o f the etiologic process. In P.C. Rivers and N.B. Lincoln (Eds.) Alcohol and Addictive Behavior. University o f Nebraska Press pp. 27-83. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Biosocial antecedents of schizophrenia-spectrum personality disorders: A longitudinal study

PDF

A mechanistic study of the diels-alder reaction

PDF

Absorption and circular dichroism studies of the ethylene and butadiene chromophores

PDF

Child sexual abuse in a sample of male and female Hispanic and White nonclinical adolescents: Extending the reliability and validity of the Trauma Symptom Inventory (TSI)

PDF

Concordant and discordant drug use in intimate relationships: A longitudinal study

PDF

A study of the reaction of 2,4-dinitrobenzene-sulfenyl halides with ketones

PDF

A study of the reaction of tetramethyldiborane with sodium in liquid ammonia and the products derived from this reaction

PDF

Childhood precursors of adult reality distortion: A prospective study.

PDF

Competitive and sequential absorption and dissociation mechanisms in the infrared fragmentation of polyatomics

PDF

A longitudinal cross-ethnic study of sociocentricity in schizophrenia

PDF

A study of the double sulphates of some rare-earth elements with sodium, potassium, ammonium and thallium

PDF

13C NMR spectroscopic study of substituted diamondoid hydrocarbons and their carbocations (dications)

PDF

Adsorption of arenes and reactivity of pyrene and diphenylanthracene with singlet oxygen on different silicas

PDF

An ab initio effective two electron Hamiltonian and an effective one-electron operator in "many-body" and "Green's function" theories

PDF

A generic training model and a study of an in-service application: an alternative for professional education

PDF

A.The anions of C(60) and its pyrrolidine derivatives B.The search for hydridocobaloximes

PDF

Crystallographic studies of metal hydrides and other transition metal complexes

PDF

A photophysical study of pyrene adsorbed onto silicas of variable surface area and porosity

PDF

A study of the technology of color motion picture processes developed in the United States

PDF

A study of the kinetics of the reaction of sulfenyl halides with olefins

Asset Metadata

Creator Locke, Thomas Fletcher (author)

Core Title Adolescent predictors and adult consequences of alcohol dysphoria: A longitudinal study

Degree Doctor of Philosophy

Degree Program Chemistry

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical,statistics

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member), Crimmins, Eileen (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-559417

Unique identifier UC11354846

Identifier 9955501.pdf (filename),usctheses-c17-559417 (legacy record id)

Legacy Identifier 9955501.pdf

Dmrecord 559417

Document Type Dissertation

Rights Locke, Thomas Fletcher

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA