University of Southern California Dissertations and Theses

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

(USC Thesis Other)

Concordant and discordant drug use in intimate relationships: 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. UMI films the text directly from the 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 afreet reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. 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 tins copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor MI 48106-1346 USA 313/761-4700 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. CONCORDANT AND DISCORDANT DRUG USE IN INTIMATE RELATIONSHIPS: A LONGITUDINAL STUDY by Michelle A. Rice A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFOPl'IIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Counseling Psychology) December 1997 Copyright 1997 Michelle A. Rice Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9835068 Copyright 1997 by Rice, Michelle A. All rights reserved. UMI Microform 9835068 Copyright 1998, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA TH E GRADUATE SCHOOL UNIVERSITY PARK LOS ANQELES. CALIFORNIA 90007 This dissertation, written by .................. under the direction of hast. Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Grach 'tudies Date . . P . ? . !? .? .£ . h.» . . . A . ? . ? . ? . . DISSERTATION COMMITTEE Chairperson Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dedication I have decided to dedicate my dissertation to many people, who in unique ways, helped me to complete it. The past two years have been filled with challenges and struggles, which were manageable only with the support of these people. To begin with, I would like to acknowledge my committee members, all of whom were willing to join my committee on short notice. I'd like to thank Dr. Michael Newcomb for his willingness to assume the role of my chairperson, for offering his database, and for working with me from 3000 miles away, in order for me to finish on schedule. Dr. Joan Rosenberg and Dr. Mitch Earleywine were supportive and flexible, helping me to meet deadlines. This dissertation is also dedicated to my family and friends, who collectively cheered me on, supported me when I wanted to give up, and never stopped believing in me. My parents, Bob and Norma, offered their love and support endlessly throughout the process. I thank my sister and best friend, Tricia, for her love, encouragement, the crisis phone calls, and of course, the editing. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In addition to my family, my friends have played an immeasurable part in allowing me to complete my dissertation. Diana and Andrea have been with me through every high and low of graduate school. Jessica and Linda made my internship year at Dartmouth amazing, filled with friendship, laughs, tears, and covert operations. Steve, I thank, for his support, encouragement, friendship, and above all, his ability to make me smile, even while writing this dissertation. Finally, I want to acknowledge the members of the research seminar at Dartmouth, who pushed me to finish, and the staff at the Counseling Center of Lebanon for their patience and support. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract The study explores concordant and discordant drug use within intimate relationships and the impact of such drug use on relationships variables, negative affect and self- medication. Previous research on drug use in the context of intimate relationships has focused on individual drug use within the relationship, while drug use and the dynamic aspect of the relationship has not been explored. Concordant and discordant drug use may be associated with changes in quality of relationship and level of dysphoria and/or agitation within the relationship. The study prospectively assessed concordant and discordant drug use in relationships as each predicted or was predicted by several other variables. The sample of 95 men and 268 women were participants in a longitudinal study, with the last two data gathering points (Year 13 and Year 17) used for this study. A structural or path model was analyzed for across time effects. The final structural model fit well and resulted in eleven significant cross-lagged effects and 20 significant nonstandard effects. Support for the presence of assortative mating for drug use was found, yet the contribution of socialization to homophily for drug use was not supported. Discordant self drug use decreased good intimate relations over time, while social support for drug problems was found Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V to decrease self drug use, over time. Dysphoria increased self drug use, over time, lending support to the idea of self-medication of negative affect. Partner discordant drug use increased dysphoria and agitation in subjects, over time, advancing the self-medication research into the area of relationship dynamics. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vi Table of Contents CHAPTER 1 Introduction p. 1 Overview of the Study p. 4 Drug Use in Couples p. 5 Impact of Drug Use on Relationships p. 10 Drug Use and Negative Affect p. 17 Hypotheses p. 24 CHAPTER 2 Sample p. 26 Measures p. 27 Analyses p. 32 CHAPTER 3 Results p. 3 5 CHAPTER 4 Discussion p. 44 References p. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vii List of Tables and Figures Figure 1.1 Hypothesized Path Model p. 72 Table 2.1 Description of Sample in Year 17 p. 73 Table 3.1 Summary of Fit Indices p. 74 Table 3.2 Correlation Matrix - Male vs. Female p. 7 5 Figure 3.1 Male/Female Factor Loadings p. 76 Table 3.3 CFA Model Intercorrelations for 1st Order Factors P-77 Figure 3.2 CFA Model for First Order Factors p.78 Table 3.4 CFA Model Intercorrelations for 2nd Order Factors p.79 Figure 3.3 CFA Model for Second Order Factors p. 80 Figure 3.4 Final Path Model p. 81 Table 3.5 Specific Effects p.82 Table 3.6 Partner-Self Drug Use Correlations Over Time p.83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 Chapter 1 Introduction Researchers have identified factors that potentially influence drug use, especially among the adolescent population (Newcomb, 1986; Newcomb & Bentler, 1987). Much of this research centered on individual drug use, including how much individuals use, how drug use influences other variables such as education, peer relations and the precocious entry into adult roles (Newcomb & Bentler, 1985), and why individuals use drugs, for example to self-medicate distressing feelings (Khantzian, 1985). A gap in this research occurs when the focus shifts from individual drug use to drug use within the context of an intimate relationship. Within this focus, two areas of research have not been developed: concordant drug use, defined as the drug use that the two people in the relationship have in common, and discordant drug use, defined as the drug use of one member of the relationship that exceeds the concordant drug use. Much of the research on drug use in adult intimate relationships relies upon three factors: assortative mating, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 socialization, and homophily. Assortative mating and socialization are processes which researchers think contribute to homophily on drug use in intimate relationships (Yamaguchi & Kandel, 1993). Homophily is defined as ‘ ‘the degree of similarity between two members of a pair on a particular attribute. . ." (Kandel, Davies, & Baydar, 1990, p. 222). In this context, homophily is the degree of similarity in drug use of two people in a relationship or the concordant drug use. Assortative mating occurs when individuals with prior similarity on certain attributes (e.g., values, attitudes, personality, drug use, etc.) select each other for intimate relationships (Merikangas, 1982; Kandel, 1984). The occurrence of assortative mating in drug use has been documented, specifically in the use of cigarettes, alcohol, marijuana, and heroin (Kandel, 1984, Hall, Hesselbrock, & Stabenau, 1983a and b; Price et al., 1981). Assortative mating for drug use has been found to be a contributing factor in concordant drug use in studies by Yamaguchi and Kandel (1993) and Kandel et al.(1990). The other potential factor in homophily is the process of socialization. Socialization is the process in which individuals in a relationship, regardless of any prior similarities, influence each other's behavior over time. While socialization has been hypothesized as a factor Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contributing to homophily, researchers have experienced difficulty thus far in teasing out the specific contribution of socialization to homophily on drug use. Until now, research on drug use in the context of intimate relationships has focused on individual drug use within a relationship and on the factors contributing to homophily on drug use. Many areas remain unexplored within the topic of drug use in relationships, especially when the focus is shifted from individual drug use to concordant drug use. Also, while it is known that drug use occurs in relationships and the processes of assortative mating and socialization are thought to be main contributors tc such drug use, there may be other variables which may either influence concordant drug use or which are influenced by concordant drug use (e.g., quality of the relationship, number of divorces, level of dysphoria and/or agitation, and level of social support) . Beyond concordant drug use, there may exist discordant drug use, in which one partner's drug use exceeds the drug use shared in common. In situations where this occurs, questions arise about the possible implications for each member of the relationship in areas such as quality of relationship, negative affect, and social support. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 Overview of the Current Study The purpose of this study is to evaluate the relationship among concordant and discordant drug use within relationships and the following variables: good intimate relations, number of divorces, dysphoria, agitation, and level of social support. (Due to the nature of the data base, complete data is available only for subjects, with only drug use measures available for the partners of subjects.) This research will be possible due to the use of an extensive longitudinal data base sponsored by the National Institute of Drug Abuse, in which data are collected at four year intervals. The last two gathering points (year 13 and year 17) will be used for the current study. Following conceptualization of the current study, the data base was reviewed for measured variables that represented, approximately, the constructs identified as important to the study. The study will prospectively assess concordant and discordant drug use in relationships as each predicts or is predicted by variables such as good intimate relations, number of divorces, level of dysphoria and/or agitation, and level of social support for both drug use problems and relationship problems. The study includes these variables because they have been identified cc be factors in individual drug use studies and are likely to be influential Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 in the study of drug use in the context of intimate relationships. The measured variables have been organized into latent constructs, and will then be tested with confirmatory factor analyses to confirm the validity. Based on the adequacy of the confirmatory factor analyses, structural models will be analyzed to test the hypothesized paths between year 13 and year 17 latent constructs and variables . Conceptual Framework The review of the literature will be divided into three general areas: drug use in couples; the impact of drug use on relationships; and drug use and negative affect. In each area, both theory and empirical studies will be reviewed, thereby developing a conceptual framework for the current s tudy. Drug Use in Couples The three main factors focused upon in much of the research on drug use in adult intimate relationships are assortative mating, socialization, and homophily. These factors are interrelated in that assortative mating and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 socialization are processes thought to contribute to homophily on drug use (Yamaguchi S c Kandel, 1993) . The current research on each factor will be explored, as well as the interrelationship among factors. Assortative mating. Assortative mating occurs when individuals with prior similarity on certain attributes select each other for intimate relationships (Merikangas, 1982) and assortative mating has been found to exist for numerous factors in intimate relationships. At first, the research supported assortative mating for limited factors such as race, age, and social class (Hollingshead, 1950). Strong correlations, ranging from .59 to .89, were reported between spouses on attitudes and values, suggesting that individuals endorsing similar attitudes and values tended to select each other for relationships (Merikangas, 1982). Moreover, the research also suggests assortative mating for physical characteristics, and more abstract factors such as cognitive abilities and personality traits, (Caspi and Herbener, 1990), and some psychological diagnoses. Merikangas and Spiker (1982) found a high degree of concordance between in-patients and their spouses in the diagnostic category of affective disorders, including 85% of subjects with bipolar illness having a spouse with an affective disorder. Assortative mating also occurs for substance use (Kandel, 1984) . Several studies have documented this process Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 for various types of drugs. Price, Chen, Cavalli-Sforza, and Feldman (1981) looked at smoking behavior among couples in the United States and Sweden. They found support for assortative mating for smoking, with individuals seeking out other individuals with similar smoking patterns. In two studies on alcohol, Hall, Hesselbrock, and Stabenau (1983a and b) found strong evidence for assortative mating for alcohol use. The first study analyzed the alcohol use of three generations of each subject (Hall et al., 1983a). Subjects were selected randomly from three inpatient alcoholism treatment centers and completed measures concerning personality, childhood adjustment, and both individual and family drinking history. In comparing parental drinking histories, the results suggested that when the subject's mother was a problem drinker, the father was also a problem drinker in 70% of the cases. This concordance rate for alcohol use gave support to the existence of assortative mating for alcohol use. The second study analyzed the subjects, themselves, along with the drinking history of their spouse and in-laws, and obtained similar results (Hall et al. , 1983b) . The levels of heavy drinking and alcoholism in the spouses of the alcoholic subjects were higher than what would be expected in the general population. Specifically, 31% of the husbands of female alcoholic subjects were themselves, alcoholic, thereby supporting the occurrence of assortative mating for alcohol use. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 Assortative mating also occurs with illegal drug use. One study focusing on heroin use, conducted by Clark, Capel, Goldsmith, and Stewart (1972) found in 3 0% of their sample of couples, both the husband and wife were addicted to heroin. Moreover, Anglin, Kao, Harlow, Peters and Booth (1987) studied the similarity of drug patterns within heroin- addicted couples. When measuring the level of heroin addiction, Anglin et al. found a significant positive correlation of .58, for daily heroin use within the couple during the relationship. This finding lends support to the existence of concordant drug use as drug patterns within the relationship were very similar . Socialization. In addition to assortative mating, the process of socialization is believed to contribute to homophily (Kandel, Davies, & Baydar, 1990). Socialization occurs when individuals in a relationship influence each other's behavior over time, regardless of any prior similarities. In the current research, the specific contribution of socialization to homophily on drug use has been difficult for researchers to tease out. While not specifically targeting socialization, Anglin et al. (1987), did find results which are suggestive of a socialization process in relationships. In looking at the level of heroin addiction in couples, they discovered that the daily use of heroin for both members of the couple was correlated throughout the span of the relationship. After the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 relationship ended, however, the strength of the correlation decreased. These results suggest that the influence of one partner on the other's behavior may have been occurring and adding to the maintenance of the drug use. Homophily on drug use. Assortative mating and socialization are not mutually exclusive and both are believed to contribute to homophily on drug use in intimate relationships (Yamaguchi & Kandel, 1993). Homophily is the level of similarity or concordance in drug use that two people have in a relationship (Kandel et al., 1990). To date, two studies have focused on homophily on drug use (Kandel et al., 1990; Yamaguchi & Kandel, 1993). Kandel et al. (1990) utilized data from two points in time for each subject: a best friend relationship in adolescence and an intimate relationship in adulthood. Homophily on drug use existed at both points and was higher for illicit drugs as compared to legal drugs. Kandel et al. found support for homophily on drug use and in particular, isolated the contribution of assortative mating to homophily. They could not, however, tease out the contribution of socialization to homophily. The study may have been limited in this area because couples were not followed over time and therefore, the researchers did not have the continuity of one relationship. Yamaguchi and Kandel (1993) also explored homophily on drug use, this time utilizing data from couples at two points Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 in time. Homophily on drug use existed and assortative mating clearly contributed to homophily. As in the Kandel et al. (1990) study, Yamaguchi and Kandel were not able to separate out the contribution of socialization. The inability to measure the contribution of socialization may be due to the study's methodological limitations. While the subjects in the researchers' original longitudinal database were interviewed at two points in time (1980 and 1984), data from their partners was only gathered in 1984. Without data from the partners at both points in time, the measurement of the socialization process would be limited. The proposed study would take this research a step further, by looking at the subject's drug use and partner's drug use at two points in time, thereby potentially teasing out the contribution of socialization to homophily on drug use. The Impact of Drug Use on Relationships The research has established that drug use not only occurs in some intimate relationships, but also impacts these relationships in numerous areas. Both theoretical and empirical work will be presented to illustrate the impact of drug use on relationships. Family systems theory. From a theoretical viewpoint, family systems theory will be used in exploring the impact of drug use on relationships. To explore the family systems Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 approach, two concepts will be defined: the system and homeostasis. A system is defined as any dyad or family, the members of which are believed to affect and be affected by each other (Haber, 1983). In a system there is homeostasis or equilibrium. Homeostasis in a system occurs due to the presence of rules and acceptable behaviors established by the members of the system. These behaviors produce predictability within the system. According to family systems theory, there is a need for homeostasis in a system and any attempts at changing the system are met by counter mechanisms . An example of such a mechanism would be one family member becoming symptomatic or distressed in response to an attempt to change the system (Stanton, 1980; Stanton et al. , 1982). This symptomatic behavior then influences the actions of other family members, thereby countering any attempt at systemic change. In families where drug use exists, efforts to maintain homeostasis also occur but the system incorporates modifications specific to drug use. In the Systems Model of Alcoholism, Steinglass (1985) suggested that the family system becomes organized around alcohol. In such a system, the built-in mechanisms for maintaining homeostasis involve alcohol-dependent behaviors, which are behaviors that occur only when alcohol is being used. These behaviors include problem solving and conflict resolution, which in systems organized around alcohol, are associated with drinking. While Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 Steinglass focused on alcohol use, the same approach could be applied to any type of substance use within a system. Substance use in this context seems to produce patterned and predictable behaviors within a system and hence the homeostasis is preserved (Steinglass, Davis, & Berenson, 1977 ) . Steinglass (1985) proposed that a system organized around alcohol would likely be confronted with chronic problems which threaten the system's equilibrium. The use of alcohol and the accompanying behaviors provide only a temporary solution to the problem, in that when drinking is not occurring in the system, the system is unable to resolve conflicts, thereby leaving the system dependent on alcohol- related behaviors to maintain homeostasis. The system, therefore, becomes involved in a continuing cycle. Within the cycle proposed by Steinglass et al. , (1977), the homeostatic system, when its members are sober, may be threatened by various problems such as individual pathology, intrafamilial conflict or family-environment conflict. The presence of such a threat leads to instability in the system. The instability prompts the need to maintain homeostasis. In a system organized around alcohol or drugs, this instability will prompt drug use and drug-related behaviors in at least one member. These behaviors are enacted to solve or at least stabilize the problem temporarily and the equilibrium of the system is thus temporarily maintained. While the homeostasis Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. f 13 of the system is preserved in this cycle, the drinking or substance using behavior of its members is also maintained, (Berenson, 1976) . Family systems theory provides a theoretical base for understanding the impact of drug use in relationships while empirical research in several areas offers further support for this approach. According to family systems theory, when a member of a couple or family uses drugs, the family system becomes organized around the drug use, thereby changing or impacting the system (Steinglass et al. , 1977) . The idea that drug use in relationships has a systemic impact is supported by research in the following areas: early entry into adult roles; relationship outcomes; and the influence of social support on continuing drug use. The proposed study will further the exploration of the systemic impact of drug use by focusing on the impact of both individual and concordant drug use on relationship variables. Early entry into adult roles. One of the ways that drug use has been thought to impact relationships is in the timing of the assumption of adult roles. Several researchers have explored this area and found varying support for this idea (Newcomb & Bentler, 1985; Yamaguchi & Kandel, 1985a & b). Newcomb and Bentler (1985) found a significant positive relationship between adolescent drug use (specifically cigarettes and illicit drugs) and marriage or cohabitation in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. f 14 young adulthood. They suggested, based on their findings, that adolescent drug use may actually accelerate the "precocious acquisition of adult roles" (Newcomb & Bentler, 1985, p. 259). This precocious entrance into adult roles, in particular cohabitation, was found to be related to greater use of cigarettes, alcohol, marijuana, and other illicit drugs (Newcomb, 1986; Newcomb & Bentler, 1987). While Newcomb and Bentler (1985) found support for the relationship between early entrance into marriage or cohabitation and drug use, other researchers have found slightly different results (Yamaguchi & Kandel, 1985b; Kandel, 1984). Kandel (1984) focused solely on marijuana use in young adults and discovered that any use of marijuana in the previous year for both men and women was significantly associated with a reduced propensity of being married. Kandel concluded that young adults using marijuana were less apt to be participating in conventional adult roles, such as marriage. Yamaguchi and Kandel (1985b) also studied the relationship between marijuana and adult roles but added the variables illicit drug use and cohabitation. In results similar to Kandel (1984), Yamaguchi and Kandel (1985b) found that men and women who were currently using marijuana were more likely to postpone marriage. The researchers also discovered that current and former use of marijuana for men and women were significantly related to an increased likelihood to cohabitate before marriage. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Methodological differences may explain why Newcomb and Bentler (1985) found support for the early entrance into marriage while Kandel (1984) found support for postponement of marriage among drug users. Newcomb and Bentler >1985) utilized prior drug use data from adolescence to predict current role assumption, while Kandel (1984) looked at drug use within the last year and its relationship to adult role assumption, thereby looking at a more limited amount of time. Despite differences in findings, the above research illustrates the systemic impact of drug use on adult relationships. Relationship outcomes. Drug use in intimate relationships has also been found to be associated with relationship outcomes (Yamaguchi & Kandel, 1935a). In general, drug use appears to be positively related to separation and/or divorce in both cohabitation and marriage (Yamaguchi & Kandel, 1985a). Specifically, Yamaguchi and Kandel (1985b) found that current marijuana use greatly increased the probability of separation and/or divorce. Along similar lines, Newcomb and Bentler (1986) discovered that both male and female cocaine users were more likely to be divorced than were nonusers. Social support. The final area associated with the systemic impact of drug use is social support. Certain kinds Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 of support have been hypothesized to both reinforce drug use, while other social support may actually mediate drug use. Newcomb and Bentler (1986) found that both male and female cocaine users perceived approval from their peers and community for their cocaine use, thereby reinforcing their drug use. In looking at the impact of marijuana use, Kandel (1984) found a similar social support influence on drug use. She found that the use of marijuana by a spouse/partner had a strong influence on one's own drug use for both men and women. Social support is often obtained, to some degree in all forms of interpersonal contact (Newcomb, 1990) and may also have a mediating effect on drug use. Newcomb and Bentler (1988b) investigated the impact of adolescent drug use and social support on seven areas of life for young adults. The areas of life considered were problems with drugs, psychosomatic complaints, relationship problems, emotional distress, work problems, health problems, and family problems. Newcomb and Bentler found that teenage social support reduced problems in each of the seven life areas for young adults, including problems with drugs, hence lending support for the possible mediating influence of social support. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 Drug Use and Negative Affect As the literature illustrates, drug use in the context of relationships impacts many areas, from the timing of the acquisition of adult roles to the role social support plays in ongoing drug use. While such variables have been studied in individual drug use, the proposed study shifts the focus to drug use in intimate relationships, adding the dynamic aspect of the dyad as a system. An individual's drug use within a system, has implications for the system itself, as well as for the other members of the system. One such implication is the possible presence of negative affect for one or both members of the dyad. The relationship between negative affect and drug use will be discussed, from a theoretical and empirical standpoint. Self-medication hypothesis. The self-medication hypothesis proposes that certain individuals are more apt to use substances to lessen or increase affect due to an inability to control such emotions on their own (Khantzian, 1990) . Murphy and Khantzian (1995) suggested that such individuals use drugs as a result of ego impairments and disturbances in their sense of self, especially in the areas of self-care and self-regulation. The idea of self-medication via drug use rests on the existence of developmental flaws and deficits in substance abusers which are proposed to have Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18 occurred in infancy and childhood (Khantzian, 1990). The amount of emotional distress an individual can tolerate and find relief from is based on how well he/she has been able to develop and internalize the ability to regulate affect and care for the self. Khantzian (1977) suggested that substance users attempt to use drugs as a way of compensating for their inability to cope with emotional distress. The substance an individual utilizes to self- medicate depends greatly on the dominant affective states of their inner emotional life, along with the individual's relationships with others (Khantzian, 1990). Cocaine and other stimulants are often chosen by individuals who are depressed and have low self-esteem. Khantzian (1985) suggested that stimulants are activating and energizing, thereby countering the fatigue of depression, while increasing self-esteem by increasing feelings of assertiveness. On the other end of the affective spectrum is rage and hostility, for which Khantzian (1990) proposed individuals utilize narcotics, especially opiates, as a means of medicating. Such individuals often experience aggressive and rageful feelings. The opiates are thought to have a containing and soothing effect on such feelings (Khantzian, 1985). Finally, there are a group of individuals whose feelings are constricted and overly contained and Khantzian (1990) suggested that this group self-medicates with alcohol. The disinhibiting effect of alcohol allows an individual to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 experience affect for brief periods of time - emotions that without alcohol are experientially out of reach. Khantzian's (1990) self-medication hypothesis suggests that certain individuals may be more apt to use drugs as a way to decrease or increase a wide array of affect. For the purpose of this study, dysphoria and agitation will represent the affective variables. The construct of dysphoria will consist of measures of depression, suicidal ideation, and purpose in life. The agitation construct will be comprised of measures of hostility and anxiety. There have been several empirical studies published in the literature, which attempt to study these affective variables and their association with relationship variables and drug use. Dysphoria. Several interesting findings exist in the literature, when considering the possible relationship between depression and substance use. Harlow, Newcomb, and Bentler (1986) looked at substance use, depression, self derogation, suicide ideation, and purpose in life and found that substance use in men was associated with increased depression and self-derogation. For women, substance use was related to decreased purpose in life and increased suicidal ideation. Several more recent studies provide further support for the self-medication of depression. Lerman et al. (1996) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 explored the relationship between depressed mood and smoking. A positive relationship was found between the presence of depressive symptoms and the level of nicotine, providing support for the self-medication hypothesis (Khantzian, 1985). Wallen (1992) found that women in substance abuse treatment had significantly higher levels of hopelessness than did men, with 73.8% of the female subjects reporting hopelessness as compared to 52.9% of the male subjects. Along the same lines, Weiss, Griffin, and Mirin (1992) tested the self-medication hypothesis by examining the motivation for drug use in hospitalized drug abusers. They discovered that women were significantly more apt to use their drug of choice to counter symptoms of depression than were men (76% of the female subjects versus 58% of the male subjects). Weiss et al. did find, however, that the presence of the diagnosis of major depression increased the likelihood that men would also self- medicate for depression. Depression has also been found to be related to level of relationship satisfaction. Beach and O'Leary (1993) studied new marital relationships and found that marital satisfaction predicted later levels of depression. In other words, decreased marital satisfaction was related to increased depression over time. Burns, Sayers, and Moras (1994) found similar results in their study of patients with mood disorders, finding a significant negative correlation of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 -.42, between severity of depression and relationship satisfaction. Agitation. While the self-medication hypothesis suggests that drug users may self-medicate feelings of anxiety and hostility, there have not been any empirical studies conducted to test this hypothesis. The literature does, however, include several studies which explored the relationship between hostility and relationship variables. A strong negative correlation between hostility and marital satisfaction for both men and women has been found in the research (Hawkins, 1968; Smith, Pope, Sanders, Allred, & O'Keeffe, 1988) . Hafner and Spence (1988) conducted a cross- sectional study on marital satisfaction, broken down based on length of marriage. In their findings, satisfaction in marriages of short duration (1-6 years) was predicted more by psychological symptoms such as depression, than by hostility, yet in marriages lasting 7-16 years, hostility was thought to be the main predictor of marital satisfaction. Moreover, Newton, Kiecolt-Glaser, Glaser, and Malarkey (1995) discovered that hostility and distressed marital interactions were positively correlated for men. Until now, the self-medication hypothesis has not been applied to drug use in intimate relationships. Although empirical studies do not exist in this area, some hypotheses can be derived from the existing literature. The research has Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22 shown a negative relationship between marital satisfaction and depression (Beach & O'Leary, 1993), along with a positive relationship between depression and drug use (Harlow et al., 1986; Wallen, 1992) . By combining these findings, it would seem that as marital satisfaction decreases and depression increases, drug use would also increase. In the area of hostility, the current research has found a negative relationship between marital satisfaction and hostility (Hawkins, 1968). Taking this finding a step further, based on the self-medication hypothesis, it would be expected that with increased hostility, drug use would also increase. Finally, Newcomb (1994) established that a significant negative correlation existed between relationship quality and drug use. Combining the above findings in the context of an intimate relationship creates several possible hypotheses. If one partner's drug use increases, perhaps due to an increase in negative affect of some kind, the discordant drug use would also be expected to increase. The increase in drug use in one partner may then lead to a decrease in marital satisfaction for the spouse. This reduction of marital satisfaction could, hypothetically, lead to an increase in depression and/or hostility and ultimately to a subsequent increase in drug use for the spouse. In reviewing the literature, a conceptual framework was established for the proposed study. The idea of homophily on drug use, with the contributing factors of assortative mating Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 and socialization, provided a foundation for the study of concordant drug use. Previous studies in this area have been unable to tease out the separate contribution of socialization, in part due to methodological limitations. The proposed study will attempt to take the research in the area of drug use in couples a step further by exploring the contribution of both assortative mating and socialization to homophily on drug use. In addition, this study will address discordant drug use in relationships, an idea which has not been focused on in the current literature. In addition to looking at drug use in relationships, the current study will explore how drug use also impacts intimate relationships in numerous areas. The review of family systems theory provided a theoretical framework, establishing that in couples or families in which one member uses drugs, the family system becomes organized around the drug use, thereby impacting the system (Steinglass et al. , 1977). The impact of drug use in relationships will be explored by focusing on the relationship between concordant and discordant drug use and several relationship variables, including the number of divorces and the quality of relationship, as well as social support for drug use and relationships. Finally, the current study will also further the research on self-medication and negative affect. Until now, the study of the relationship between negative affect and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24 self-medication has focused on the individual. The proposed study will shift the focus to intimate relationships, making possible a study of numerous variables which may influence or be influenced by the presence of negative affect and/or self- medication . The review of the literature has lead to several hypotheses, which the current study will attempt to address. Hypotheses 1. Partner polydrug use and self polydrug use are expected tc be positively correlated. 2. Concordant (couple) polydrug use is expected to increase from Year 13 to Year 17. 3 . Discordant polydrug use will generate decreased good intimate relations, over time. 4. Discordant polydrug use should increase number of divorces, over time. 5. Social support for drugs is expected to be negatively correlated with self polydrug use. 6. Social support for drugs is expected to decrease self polydrug use, over time. 7. Social support for relationships is expected to be positively correlated with good intimate relations. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 8. Social support for relationships should increase good intimate relations, over time. 9. a) Good intimate relations is expected to be negatively correlated with dysphoria, b) Good intimate relations is expected to be negatively correlated with agitation. 10. a) Self polydrug use is expected to be positively correlated with dysphoria, b) Self polydrug use is expected to be positively correlated with agitation. 11. a) Concordant (couple) polydrug use should decrease dysphoria, over time, b) Concordant (couple) polydrug use should decrease agitation, over time. 12. a) Discordant polydrug use in the partner should increase dysphoria, in self, over time, b) Discordant polydrug use in the partner should increase agitation, in self, over time. 13. a) Dysphoria will increase discordant polydrug use, in self, over time, b) Agitation will increase discordant polydrug use, in the self, over time. 14. A gender difference is expected in the relationship between dysphoria and self polydrug use. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. E5sJ |S- Ale. I E h lI S elf Polv D L ' Self Poiv D L ' S-llI Couple DL Couple , D L Partner Polv DL' Partner Polv D L v P-Alc. P-Alc. P-lll. Sexual Satis DAS Xjood Intimate .Relation DAS Dependence | pepend ence | Dating Competence Dating Comp. < ? of divorces past 4 years * of divorces current rel. [CES-Dl CES-D Suicidal Ideation tysphona Purpose in Life Agitation Agitation Familv Fam ily soaal support drugs soaal support drugs I Friends Fnends | Agency- Family Fam ilv social sup— rel Friends ipport import Fnends Agency Sex Figure 1.1. Hypothesized path model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26 Chapter 2 Method Sample Data were from a longitudinal study of personal development and drug use, which began in 1976. The original sample included 1634 junior high school students in the 7th, 8th, or 9th grade, in 11 Los Angeles County schools. The five school districts selected yielded a representative sample of Los Angeles County students in terms of socioeconomic status (SES) and ethnicity. Oversampling of lower SES schools was utilized in order to offset a predicted lower participation rate of lower SES students (Newcomb & Bentler, 1988a). For the current study, the two latest waves of data were used, in which the average ages were 25.5 (Year 13 follow-up) and 29.9 (Year 17 follow-up). Only those individuals who were in a relationship at the Year 13 follow- up were included in the sample. Approximately 3/4 of all subjects were in an intimate relationship at Year 13 and this proportion did not differ by gender (Newcomb, 1994). The inclusion criterion was utilized to ensure that all subjects Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27 were involved in an intimate relationship at at least one point in the study. A complete description of the sample is provided in Table 2.1. The ages of the subjects ranged from 25-32. The ethnicity of the sample was 67% Anglo with the remaining 33% distributed across African-American, Latino, and Asian- Pacific participants. The average education level was two years of college, while ranging from the ninth grade to doctoral degrees. More than 63% of the sample was employed full-time and 40% earned an income of over $50,000. Over 2/3 of the sample was living with a spouse and over 1/2 of the sample were parents. Measures At Year 13 and Year 17, identical scales were utilized for the polydrug use, good intimate relations, divorce, dysphoria, agitation, social support for drugs, and social support for relationships. The current relationship variable was not included in Year 13 due to the inclusion criteria for the sample, requiring that subjects be in an intimate relationship at Year 13. Drug Use. At both data gathering points, three scales of drug use were used as indicators of a self polydrug use latent variable and a partner polydrug use latent variable. These scales included frequency of cigarette use, alcohol, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 and illicit drugs (a sum of items that included cannabis, cocaine, heroin, hypnotics, stimulants, narcotics, PCP, inhalants, and hallucinogens) . The subject was asked to indicate how often he/she had used the drug in the past six months on a scale ranging from never (0) to more than once a day (5). The partner's drug use was measured on a shorter but similar questionnaire, completed by the subject. The drugs comprising the illicit drug scale differed slightly on the partner drug use questionnaire and included cannabis, cocaine, uppers, downers, psychedelics, narcotics, and PCP. The partner drug use scale also ranged from never (0) to more than once a day (5). Good intimate relations. Four measures are used to reflect the latent construct of Good Intimate Relations. Each scale is described below. Sexual satisfaction is reflected in three single items: happy with sex life, problems with sex, and intercourse satisfaction. The first item is rated on a 7- point scale, ranging from terrible (1) to delighted (7). The problems with sex item is rated on a 5-point scale, ranging from never (1) to always (5) . Finally, the intercourse satisfaction item is based on a 7-point scale, ranging from very dissatisfied (1) to very satisfied (7). Internal Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 consistency of this scale is adequate, given the small number of items (d = .53) (Newcomb & Rickards, 1995). Dyadic adjustment was measured using an abbreviated version of the original Dyadic Adjustment Scale (DAS, Spanier, 1976) . The DAS was designed to measure the quality of marriage. Subjects were asked to assess their current relationship or the most recent one, if they are not currently in a romantic relationship. The scale is broken into four subscales - consensus, affectional expression, cohesion, and satisfaction. Items for the abbreviated version were selected on the basis of the largest factor loadings for each scale (Newcomb, 1994). The reliability for the shortened version is good, with an alpha of .89 (Newcomb & Rickards, 1995) . The five-item dependence scale from the Depressive Experiences Questionnaire (DEQ; Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982) was utilized to measure dependence. The DEQ assesses feelings about the self within interpersonal relationships. The Dependence Subscale of the DEQ has good reliability with an alpha of .81 (Blatt et al. , 1982) . The Dating Competence scale was created from a subscale of the Dating and Assertion Questionnaire (Levenson & Gottman, 1978). The items are based on a four-point scale, ranging from never (1) to always (4). Sample items include "get a second date with someone you dated once" and "have an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 intimate relationship with a member of the opposite sex" . In this study, a shortened version of the dating scale was used, utilizing five of the original nine items. The shortened version had good reliability (d = .72) (Newcomb & Rickards, 1995). Number of divorces. The number of divorces in the eight years prior to the assessment in year 13, and in the four years prior to year 17, were noted for each subject. Dysphoria. Three separate measures were utilized to measure the latent construct of dysphoria. Each scale is described below. The Center for Epidemiologic Studies - Depression Scale (CES-D), constructed by Radloff (1977), is a 20 item self-report measure of depression. Subjects were asked to indicate how often in the past week they felt sad, depressed, fearful, lonely, happy, etc. on a 4-point scale ranging from rarely (0) to most of the time (3). The 20 items were grouped into 4 indicators based on factor analysis: positive affect, negative affect, impaired motivation, and impaired relationships (Harlow, Newcomb, & Bentler, 1986) . These subscales have reliability estimates ranging from .59 to .82 (Radloff, 1977) . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 The Suicide Ideation Scale is a five-item measure with each item on a 5-point Likert scale, ranging from never (1) to always (5). Three of the items were originally from Petrie and Chamberlain's (1983) Suicide Behavior Subscale (“I have been thinking about ways to kill myself", "I have had recent thoughts about dying", and “I have told someone I want to kill myself"). Two additional items were created for this longitudinal study: "I imagine my life will end with suicide", and "I have made attempts to kill myself" (Harlow et al., 1986). Higher scores are indicative of greater suicidal ideation. The Purpose in Life measure is a version of Crumbaugh and Maholick's (1964) Purpose in Life test (PIL). The measure consists of 20 items, each based on a 7-point scale, ranging from strongly disagree (1) to strongly agree (7). Higher total scores indicate greater purpose in life. Agitation. The latent construct of agitation consists of scales measuring hostility and anxiety. The hostility scale was taken from the Hopkin's symptom checklist (Uhlenhuth, Balter, Mellinger, Cisin, & Clinthorne, 1983) . The reliability for this scale is .80 (Scheier & Newcomb, 1993). The anxiety scale was also derived from same checklist. The reliability for this scale is .75. For both the hostility and anxiety measures, the three items with the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 largest factor loadings in each area were included in the scales. Social Support. Two areas of social support are assessed - support for drug problems and support for relationship problems. The quantity of support in each area was measured on a 9-point scale, ranging from no units of support (0) to nine or more units of support (9) . Subjects were asked how many sources of support they had in each area among a) family members; b) friends; and c) community or private agencies. Satisfaction with the support from the three sources noted above was measured on 5-point scales, ranging from very dissatisfied (1) to very satisfied (5) (Newcomb and Chou, 1989). Analyses Each model analysis was performed utilizing the EQS structural equations program (Bentler, 1989). After the study was conceptualized and the measured variables were organized into latent constructs, confirmatory factor analysis (CFA) models were constructed. The CFA models were used to determine the adequacy of the hypothesized factor loadings, the degree of model fit, and the latent construct intercorrelations. Multiple-group analyses were conducted to assess whether models for men and women were reliably different. If the models are not substantially different, one Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 overall model is then utilized as a basis for testing structural or path models. In the CFA models, all factor loadings were freed and the factor variances were fixed at 1.00. Each of the factors were allowed to correlate freely, as were the residuals. Models were accepted as adequate when the chi-square/degrees of freedom ratio was less than 2.00 or the Comparative Fit Index (Bentler, 1989) was greater than .90 (Newcomb, 1994). The factor structures from the CFA models were used as a basis for the structural or path models. All Year 13 constructs were initially allowed to predict all Year 17 outcomes. Specific or nonstandard effects were added to the model using modification indices and were restricted to the following across-time paths: a) from Year 13 measured variables (or residuals) to Year 17 measured variables; b) from Year 13 latent constructs to Year 17 latent constructs; and c) from Year 13 measured variables (or residuals) to Year 17 latent constructs. The multivariate Lagrangian Multiplier Test was used to add specific effects and then the Wald test was used to remove nonsignificant paths (Newcomb, 1994) . There were some limitations to this study, which will be introduced here and further explained in Chapter 4. For the most part, the limitations stem from data that was not available, due to the fact that this study was designed after Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34 the data was collected as part of a longitudinal study. As a result, assessing break-ups of relationships over time becomes difficult. While divorce is measured, break-ups of other intimate relationships are not measured and therefore present a limitation to the study. Another related shortcoming of the study, is the lack of information on changes of partners over time, therefore subjects may report being in a relationship at both points in time, yet may have changed partners. Despite these limitations, this study offered the means to assess concordant and discordant drug use within relationships. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1 Description of Sample in Year 17 Variable Male(N=95) Female (N=268) Total (N=363) Age (yeai?) Mean 29.9 29.9 29.9 Range 25-32 28-32 25-32 Ethnicity (%) African-American 3 13 11 Latino 15 12 13 Anglo 74 65 67 Asian-Pacific 8 9 9 Education (years) Mean 14.7 14.1 14.3 Range 10-18 9-18 9-18 Number of children (%) None O i l One 45 39 41 Two 42 34 36 Three or more 13 25 22 Income for past year (%) Under $10,000 3 9 8 $10,000 to $30,000 18 22 21 $30,000 to $50,000 35 30 31 Over $50,000 44 39 40 Living Situation C % 1 Alone 5 8 8 Parents 2 6 5 Spouse 79 66 69 Cohabitation 8 7 7 Dormitory 0 0 0 Roommates 6 4 5 Single Parent 0 9 6 No Regular Place 0 0 0 Current Life Activity (%) Unemployed, laid off, fired 2 6 5 Military 5 0 1 College 3 5 2 Child rearing/part time job 0 12 9 Part - time job 4 5 4 Full-tim e job 81 57 63 Child rearing/homemaker 0 18 13 None 4 1 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 Chapter 3 Results The results are organized into three segments, reflecting the order in which the analyses were completed. The first segment contains the multiple group analyses, an assessment of differences between the male and female models. Based on the multiple group analyses, confirmatory factor analysis models were completed and analyzed in the second section. Finally, the results of the final structural or path model are presented. A summary of all fit indices is presented in Table 3.1. Multiple Group Analyses Due to a smaller number of males in the sample, as compared to women, utilizing one overall model for the structural or path analyses would be more statistically powerful. In order to justify using such a model, differences between the male and female models needed to be minimal. If the differences were not substantial, then one overall model could be used, with sex included as a measured variable. The Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 following analyses were utilized to assess for differences between the male and female models . The Box M omnibus test was first utilized to compare the male and female covariance matrices. The Box M omnibus test is a Chi-square statistic, which if significant, allows for multiple group analyses to be conducted. The Box M test was statistically significant, indicating that differences existed between the male and female models. In order to delineate these differences, several multiple group analyses were conducted. A series of multiple group analyses were conducted, with the two groups being men and women. A summary of the model fit indices and nested difference tests is found in Table 3.1. The first confirmatory factor analysis (CFA) model(referred to as "all free") allowed all paths to be freely estimated and fixed all factor variances at 1.0. This model fit fairly well, with the CFI=.88, and x2:df=1.54. In the next CFA model, all factor loadings were constrained to equality between groups (referred to as 1 1 FL=“ ) . The following example illustrates what was meant by constrained by equality: the factor loading for subject cigarette use, year 13 - males is made to equal the factor loading for subject cigarette use, year 13 - females. This model also fit fairly Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 well, with a CFI=.88, and x2;df=1.54. A nested difference test was utilized to compare these two models. The resulting chi-square statistic was statistically significant, indicating that significant differences were present between the two models. Specifically, two paths had significant differences in factor loadings for males and females - the path between partner illicit drug use, Year 17 and Partner Polydrug Use, Year 17; and the path between suicidal ideation, Year 17, and Dysphoria, Year 17. Next, a CFA model was created in which both factor loadings and covariances were constrained to equality between the two groups (referred to as "FL and COV=") . This model was then compared to both the "all free" and "FL=" models. The nested difference test between the "FL and COV=" model and the "FL=" model, resulted in a chi-square statistic that was not statistically significant, indicating that there were not significant differences between these two models . A nested difference test was also conducted for the "FL and COV=" model and the "all free" model. In comparing these two models, a statistically significant chi-square statistic was found, indicating a significant difference between these two models. As a result, seven instances of significant gender differences were found, as summarized in Table 3.2. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 The multiple group analyses resulted in several significant differences between the male and female models. These differences, taken in the context of all potential factor loadings and correlations among constructs within the model, are not substantial in number. Due to the lack of substantial statistically significant differences, one overall model will be utilized for the structural or path analyses, with the measured variable of sex added to the model. CFA Models Prior to constructing the across-time path models, confirmatory factor analysis (CFA) models were formulated, in order to verify that the measured variables were statistically reliable indicators of the latent constructs. Separate CFA models were constructed for first order factors and for second order factors. In each model, all factor loadings were freed and factor variances were fixed at 1.00, in order to identify the constructs. All of the factors and independent measured variables (Divorce, Year 13 and Year 17, Sex, and Current Relationship) were allowed to correlate freely, as were the uniqueness of repeated measures (i.e. the residual of "CES-D" Year 13, was allowed to correlate with the residual of "CES-D" Year 17). The CFA models are a necessary step in constructing and analyzing structural path models. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 The results of the initial CFA models are presented below in figures 3.2 and 3.3, and in Tables 3.3 and 3.4. The figures include the latent constructs, depicted by large ellipses, and the measured variables, represented by rectangles. Each latent construct has arrows pointing towards the measured variables which are indicators of that construct. The smaller ellipses represent the residual variance of the measured variables. This depicts the unique variance that combined with the measurement error makes up the error term. The factor intercorrelations between all latent constructs and the measured variables of Divorce, Year 13 and Year 17, Sex, and Current Relationship, are presented in Table 3.3 and 3.4. The factor intercorrelations reflect both within-time and across-time correlations. The CFA model for the first order factors fit well, according to the criteria cited earlier: x2 = 13 84.46, pc.001, CFI=.91, and x2:df ratio= 1.72. All hypothesized factor loadings were significant in the expected directions. The results of the model are depicted in figure 3.2, and the factor intercorrelations are given in Table 3.3. The CFA model for the second order factors also fit well: x2=1482.44, pc.001, CFI=.90, and x2:df ratio= 1.76. As in the first model, all of the hypothesized factor loadings were statistically significant in the expected directions. The results of the model are given in figure 3.3, and the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 factor intercorrelations are included in Table 3.4. The results of the first and second order CFA models suggest that the measured variables are reliable indicators of the latent constructs in this study. Structural or Path Model The factor structures confirmed in the above mentioned CFA models were utilized as the basis for the following structural or path model. The model contains the same constructs and measured variables as used in the CFA models. All within-time constructs and their related disturbance terms were allowed to correlate freely. Initially, all Year 13 constructs and measured variables were allowed to predict all Year 17 outcomes. The significant paths between the latent constructs and measured variables in the model are presented in figure 3.4. There were also numerous "nonstandard" or “specific" paths added to the model, utilizing the multivariate Lagrangian Multiplier Test. Next, using the Wald test, all nonsignificant paths and correlations were deleted. The significant nonstandard effects are summarized in Table 3.5. In the nonstandard paths, either the predictor or the outcome reflected a measured variable that served as an indicator of a latent construct. Although the significant paths and the nonstandard effects are presented separately for the purpose of clarity, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41 it is important to note that they are derived from the same final model. The final structural model fit the data well, X2 = 1389.50, p< .001, CFI= .92, X2:df= 1.58. As in the figures for the CFA models, the latent constructs are represented by ellipses and the measured variables of divorce, current relationship, and sex, are depicted by rectangles. The smaller ellipses denote the residual variance. Numerous significant paths were found among the constructs (including the divorce variables, current relationship, and the sex variable). The paths are illustrated in figure 3.4. As shown in the figure, Couple Polydrug Use increased Couple Polydrug Use, while also indirectly increasing Self Polydrug Use and Partner Polydrug Use. Additionally, Divorce, past eight years, decreased Agitation. Good Intimate Relations decreased Couple Polydrug Use. Dysphoria increased both Self Polydrug Use and Agitation. Agitation, on the other hand, increased Divorce, past four years. Social Support for drug problems decreased Couple Polydrug Use, Self Polydrug Use, and Dysphoria, while increasing Social Support for relationship problems. Finally, Social Support for relationship problems decreased Agitation and increased Social Support for drug problems. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42 In addition to the significant paths noted, 20 nonstandard parameters were added and are summarized in Table 3.5. Seven of the nonstandard effects reflect drug use predictors and the remaining 13 nonstandard effects reflect drug use consequences. Among the 20 significant effects, three types of paths are present: from a latent construct to a measured variable; from a measured-variable residual to a latent construct; and from a measured-variable residual to a measured variable. As noted above, there were seven significant nonstandard effects that reflected drug use predictors in the final model. The effects are as follows: a) Good Intimate Relations decreased Self cigarette use and Partner cigarette use; b) Dependence increased Partner Polydrug Use; c) Suicidal ideation increased Self Polydrug Use; d) Hostility increased Partner Polydrug Use; e) Friend support for drug problems decreased Self Polydrug Use; and f) Family support for drug problems decreased Partner Polydrug Use. Numerous significant nonstandard effects were also found, reflecting drug use consequences. The effects are as follows: a) Couple Polydrug Use decreased dyadic adjustment; b) Self Polydrug Use decreased dyadic adjustment and purpose in life; c) Self cigarette use decreased Dysphoria and increased Agitation; d) Self illicit drug use increased Dysphoria; e) Self alcohol use decreased being in a current Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43 relationship; f) Partner Discordant Polydrug Use decreased both dating competence, purpose in life, and family support for drug problems (all in the subject) , while increasing suicidal ideation and hostility, in the subject; and g) Partner cigarette use increased Divorce, past four years. In summary, there were significant stability paths across time for all of the constructs found in both Year 13 and Year 17. Numerous cross-lagged effects existed, as well as many nonstandard effects. Additional Smaller Model In addition to the overall model, a smaller model was created to test the socialization hypothesis, which predicted that concordant drug use would increase over time. Only subjects indicating being in a relationship at Year 13 and Year 17, while also indicating no divorce in the intervening time, were included in the smaller model. These criteria were used in an attempt to control for changes in partners during the time between Year 13 and Year 17. Correlations between self and partner polydrug use, over time, did not increase significantly. Correlations between self and partner in the drug subcategories (cigarette use, alcohol use, and illicit drug use) were also not significantly different, over time. All self-partner drug use correlations are presented in Table 3.6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44 Chapter 4 Discussion Until now, the research on drug use within intimate relationships, has focused on an individual's drug use, as well as on the factors contributing to homophily on drug use. When the focus is shifted from individual drug use to drug use within the context of the relationship, a gap occurs in the current literature. The current study has attempted to address several areas in this gap, which have not been developed, as yet, in the drug use literature. Past research on drug use in intimate relationships has relied on the concept of homophily on drug use, with assortative mating and socialization thought to be the contributing factors. Previous studies have found assortative mating to be a factor (Kandel, 1984), while teasing out the contribution of socialization has remained difficult (Kandel et al., 1990) . In addition to addressing homophily on drug use, the current study also focused on the idea of concordant and discordant drug use in couples. The implications of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45 concordant and discordant drug use for each member of the relationship were explored in areas such as intimate relations, social support, and negative affect. In this chapter, the results of the present study will be discussed in the context of the conceptual framework presented in Chapter 1. While some of the hypotheses suggested were not confirmed by the results, other findings that were not originally predicted, were found to be significant. Both the hypotheses presented earlier and the other findings will be addressed. In addition, implications for theory, as well as for intervention and prevention will be discussed. Finally, the limitations of the study will be focused upon, along with some directions for further research. Maior Findings The hypotheses suggested in Chapter 1, will be addressed in three sections: drug use in couples; the impact of drug use on relationships; and drug use and negative affect. In each section, the results will be integrated with relevant theoretical and empirical literature. Drug use in couples. The main processes thought to be involved in homophily on drug use in intimate relationships are assortative mating and socialization (Yamaguchi & Kandel, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46 1993). Self and Partner Polydrug Use were positively correlated at both Year 13 and Year 17, suggesting the presence of assortative mating. The correlations were statistically significant and strong in magnitude (Year 13: r=.72, p<.001; Year 17: r=.77, pc.001). Significant, positive correlations between self and partner, also occurred in drug subcategories (i.e. cigarette use, alcohol use, and use of illicit drugs), indicating that the subjects and their partners tended to have similar levels of drug use. The support for assortative mating in this study further confirms similar findings in the literature (Price et al., 1981; Hall et al., 1983a and b; and Anglin et al., 1987). The prediction that concordant (couple) drug use, should increase over time (i.e. the process of socialization) was not confirmed by the results. A smaller model was utilized to test this prediction, in an attempt to control for changes in partners during the time between the data gathering points. Only subjects indicating being in a relationship at both Year 13 and Year 17, while also indicating no divorce in the intervening four years, were included in the smaller model. It was hoped that these criteria would lower the possibility of a change in partners affecting the results. With the smaller model, no support was found for a statistically significant increase of the correlation between Self Polydrug Use and Partner Polydrug Use, over time. In addition, the comparisons of self and partner use of cigarettes, alcohol, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47 and illicit drugs, over time, were also not statistically significant. Limitations in the database may help to explain, in part, the lack of support for this hypothesis. While the current study improved on limitations in the current literature (Yamaguchi & Kandel, 1993; Kandel et al., 1990), by looking at the subject's and partner's drug use at two points in time, the study's database had limitations which may have hindered finding support for socialization as a factor in homophily on drug use. All of data on the partner's drug use was reported by the subject, thereby introducing some bias into the database. The subject may have given a subjective assessment of the partner's level of drug use. In addition, aside from the divorce variable, there was no way to monitor changes in partners between data gathering points. Subjects who were not married, may have changed partners, thereby affecting the ability to measure the true influence of one partner's drug use on the other partner. From a more conceptual viewpoint, the lack of support for the socialization process has several possible explanations. Assortative mating may actually be the predominant factor in homophily on drug use, as supported by the strong correlations found between self and partner polydrug use. In addition, the process of socialization may occur early in relationships and be a less powerful factor. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 The database did not account for the length of the relationship and therefore, couples in this study may have already experienced the socialization process, prior to the Year 13 of the longitudinal study. Finally, in some extreme cases of discordant drug use in couples, divorce may be a solution to the distress in the relationship and hence, socialization may not occur due to the ending of the relationship. The impact of drug use on relationships. Based on family systems theory, when drug use occurs in a couple or family, the system becomes organized around the drug use and hence, the system has been impacted by the drug use (Steinglass et al. , 1977) . In the current study, the concept of discordant drug use in an intimate relationship was introduced. Discordant drug use, meaning the individual drug use which exceeds the drug use shared in common by the couple, is part of the residual terms for the latent constructs of Self Polydrug Use and Partner Polydrug Use. Newcomb (1994) suggested that a portion of the residual term actually reflects some unique aspect of the variable or construct. The results support the idea that drug use, specifically discordant polydrug use, does impact intimate relationships, while also giving credence to the notion that social support may influence drug use in relationships and relationship outcomes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49 Discordant polydrug use was found to impact specific relationship outcomes. There was partial support for the prediction that discordant drug use would decrease Good Intimate Relations, over time. While there was not a significant cross-lagged effect, two significant nonstandard or specific effects were present, indicating a relationship between discordant polydrug use and specific indicators of Good Intimate Relations. Self Discordant Polydrug Use decreased dyadic adjustment over time and Partner Discordant Polydrug Use decreased dating competence over time. The variables of dyadic adjustment and dating competence were both indicators of the latent construct, Good Intimate Relations. It seems that the presence of discordant drug use in either member of the couple impacted aspects of relationship outcomes for the subject. A second prediction was made with regards to discordant drug use and was not confirmed by the results. It was hypothesized that discordant polydrug use would increase the number of divorces over time. This prediction was supported by previous research which had found a positive relationship between drug use and separation and/or divorce (Yamaguchi & Kandel, 1985a and b; Newcomb & Bentler, 1986). Two significant specific effects were found however, that hint at the proposed relationship. These results indicate that an unique effect of partner cigarette use is associated with increased divorces over time, while an unique effect of self Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 alcohol use is associated with a decrease in the probability of being in a relationship over time. It may be that the level of discordant drug use among the couples in the sample was not high enough to create a situation prompting divorce. Another possibility would be that some of the married couples could have separated, but not divorced during the intervening four years - a variable not included in the current study. Social support has been suggested as a mediating influence on drug use and in relationship problems (Newcomb & Bentler, 1988), and the results of the current study support this idea. As predicted, social support for drug problems was significantly correlated with decreased self polydrug use in both Year 13 and Year 17. Social support for drug problems was also found to decrease self polydrug use, over time, with a significant cross-lagged effect between these two latent constructs. In addition, a significant nonstandard effect was also present, indicating that the measured variable, social support for drug problems from friends (an indicator of the latent construct of social support for drug problems), decreased self polydrug use, over time. While the Newcomb and Bentler (1988) study focused on teens and social support, the current study moves the research into the context of adult intimate relationships and the role of social support as a mediating influence on drug use. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 Social support has also been found to reduce relationship problems for teens (Newcomb & Bentler, 1988) and support was found for this idea in adult relationships. Social Support for Relationship Problems was significantly correlated with increased Good Intimate Relations in both Year 13 and Year 17. The prediction that Social Support for Relationship Problems would increase Good Intimate Relations, over time, was not, however, confirmed by the results. Drug use and negative affect. An individual ' s drug use within a family system, has implications for the system itself, as has been shown in the results, as well as implications for the other members of the system. One such implication is the presence of negative affect for one or both members of a couple, as well as the possible impact on relationship satisfaction. The self-medication hypothesis suggests that some individuals may use substances to lessen painful affect (Khantzian, 1990). The results support the idea that negative affect, relationship satisfaction and drug use are interrelated, while also moving the research on the self-medication hypothesis into the context of intimate relationships. The level of good intimate relations, the presence of negative affect, and drug use were found to be closely related. Significant correlations existed between decreased Good Intimate Relations and increased Dysphoria and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 Agitation. This finding supported the results of previous studies (Beach & O'Leary, 1993; Burns et al. , 1994; Hawkins, 1968; and Smith et al., 1988) . Dysphoria was significantly correlated with increased Self Polydrug Use, confirming the work of Harlow et al.(1986), and Wallen (1992). In addition. Agitation and Self Polydrug Use were also positively related, providing empirical support for the self-medication of feelings of anxiety and hostility, which until this time, did not have any empirical studies. The results involving negative affect and drug use, across time, provided the strongest support for the self- medication hypothesis, in the context of intimate relationships. The prediction that Dysphoria would increase Self Polydrug Use over time, was supported by both a significant cross-lagged effect and a significant nonstandard effect. In the cross-lagged effect, Dysphoria increased Self Polydrug Use over time, supporting Khantzian's (1990) idea that individuals self-medicate painful affect. The nonstandard effect offered partial support in that suicidal ideation, an indicator of Dysphoria, increased Self Polydrug Use over time. The results did not, however, support a cross time effect for Agitation. The findings also combine the self-medication hypothesis with the dynamic aspect of a relationship, producing results which move the drug use literature a step ahead. Three Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 significant nonstandard effects partially supported the prediction that discordant polydrug use in the partner would increase Dysphoria and/or Agitation in the subject, over time. Discordant Partner Polydrug Use increased suicidal ideation and decreased purpose in life (both variables are indicators of Dysphoria) . Discordant Partner Polydrug Use also increased hostility (an indicator of Agitation) over time. These findings suggest that the discordant drug use of one member of the couple impacts the other member via an increase in negative affect. While many predictions in this area were confirmed, two hypotheses were not supported by the results. It was predicted that a gender difference would exist in the relationship between Dysphcria and drug use over time and this was not supported. Previous research had found a tendency for women to be more apt to self-medicate depression as compared to men (Wallen, 1992). The second unconfirmed hypothesis predicted that Couple Polydrug Use would decrease Dysphoria and/or Agitation over time. Drug use may instead be increasing within the relationship to counter any increases in negative affect, thereby maintaining a balance rather than actually decreasing the negative affect. While this remains purely speculation due to the findings, it would support the self-medication hypothesis in that negative affect is medicated with substances but not resolved. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 Additional Findings In addition to the findings presented above, several interesting results occurred, which were not predicted. Two significant cross-lagged effects occurred, involving Couple Polydrug Use. Both Good Intimate Relations and Social Support for Drug Problems decreased Couple Polydrug Use over time. These results suggest that when relationship satisfaction is high and/or when social support is present, the level of drug use shared in common in a relationship is likely to decrease. An additional finding involving social support was also present in the results. Social Support for Drug Problems increased Social Support for Relationship Problems over time, and vice versa. This finding suggests that the presence of one form of social support probably increases the presence of other forms of social support over time. Finally, several interesting yet unpredicted results occurred involving negative affect. In significant cross lagged effects, Divorce in the past eight years decreased Agitation over time, while Agitation increased Divorce in the intervening four years. Dysphoria was also found to increase Agitation over time. These results illustrate the impact of negative affect within the intimate relationship. Social support seemed to also have a mediating influence on negative affect. Social Support for Drug Problems decreased Dysphoria, suggesting that receiving support for drug problems, might Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 lessen attempts to self-medicate and lead to more resolution of dysphoric feelings. Social Support for Relationship Problems decreased Agitation over time, indicating that the presence of social support during difficulties in relationships may decrease the accompanying agitation. Theoretical Implications The findings of the current study support the theories discussed previously, and in some cases, strengthen the theory. Assortative mating and socialization have been hypothesized as contributing factors to homophily on drug use in intimate relationships (Yamaguchi & Kandel, 1993). There was strong support for the presence of assortative mating, suggesting that individuals tend to select partners with similar attributes, in this case, similar drug use. The process of socialization, in which individuals in a relationship influence each other's behavior over time, regardless of prior similarities, was not supported by the results. Socialization remains an important part of the theory of homophily, but has not been empirically supported by any study thus far. The results provided added support for family systems theory, as it relates to drug use in relationships. Steinglass(1985) introduced the Systems Model of Alcoholism, suggesting that family systems become organized around Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 alcohol (or other substances), when at least one member is using drugs. Such a system is often confronted with problems which threaten the equilibrium of the system, prompting substance-dependent behaviors to maintain the balance in the system. The results show that drug use in relationships has a systemic impact. The discordant drug use within a relationship, in particular, negatively impacted the quality of relationship. This finding was unique in that previous research has not focused upon the drug use existing beyond that which is shared in the relationship, thereby lending added support to systems theory. The results also revealed several instances in which the system impacted drug use. The presence of social support for drug problems lessened both individual drug use and the concordant drug use, suggesting a positive systemic impact on drug use. Good intimate relations and satisfaction in a relationship also decreased the level of concordant drug use. Perhaps positive social support and improved intimate relations help to change the system, in that substance- dependent behaviors used to maintain homeostasis are replaced by more healthy behaviors. The study supported and advanced Khantzian's body of work on the self-medication hypothesis (Khantzian, 1977; 1985; 1990). The self-medication hypothesis suggested that certain individuals are more apt to use substances to lessen Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 or increase affect due to an inability to control such emotions on their own (Khantzian, 1990) . Until now, the theory had been supported by several empirical studies, focusing on depression and substance use. Dysphoria and individual drug use were strongly related in the current study. In addition, a relationship between agitation and drug use was also documented - the first empirical support self- medication of such affect. Finally, the dynamic aspect of the relationship was considered, in the context of self- medication and negative affect. The drug use of one partner has a definite impact on the affect of the other person. In viewing these results, it is known that the drug use of one partner affects the affect of the other, and may therefore, increase the probability of more self-medicating of negative affect. Clinical Implications The study advanced the current research, on drug use by shifting the focus from the individual to drug use within an intimate relationship, and by doing so, provided clinicians with important information. Based on the results, there are clinical implications for the areas of assessment, treatment and prevention. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 In the area of assessment, two important suggestions arise from the results. First, due to the presence of assortative mating and concordant drug use in couples, clinicians need to assess for drug use within a relationship or system, rather than as an individual issue. Substance abuse assessment could be improved by focusing on both the level of drug use for the client and for the client's significant other. In addition, the impact of a partner's discordant drug use on the client should be assessed, based on the findings that discordant drug use of the partner is associated with increases in negative affect for the other person. Drug use and feelings of dysphoria and agitation have been found to be significantly associated. Clients presenting with affective symptoms should also be assessed for drug use, to appraise the level of self-medication occurring. In the realm of treatment, the study adds support for the idea of treating substance abuse within the context of a family system as suggested by Stanton and Shadish (1997).They conducted a meta-analysis of drug abuse treatment outcome studies and found that family therapy had better results than individual therapy, peer group therapy, and family psychoeducation. Drug use does not often occur in a vacuum, without the involvement and influence of others. Drug use impacts the family system and the system can also change, to impact the drug use. It seems that by involving the other members of the system in the treatment of a member's drug Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 abuse, more lasting changes can be made in the system. Family systems theory suggests that the system becomes organized around the drug use and then drug-dependent behaviors occur to maintain the equilibrium (Steinglass et al. , 1977). By addressing the entire system, an alternative organization and maintaining behaviors can be introduced. Edwards and Steinglass (1995) found that involving the entire system, via family therapy, was especially effective in the initial phase of treatment. The results of the study also point toward preventative measures in countering drug use. Social support for both drug problems and relationship problems was significantly related to decreases in both drug use and relationship problems . Establishing programs within the community which offer support for relationship issues could serve as a preventative measure. Already established programs such as Alcoholics Anonymous, provide such social support for drug related problems. Clinicians could provide preventative treatment by utilizing such programs as referral sources for their clients. Limitations There were some limitations to this study, as introduced in Chapter 2, which will now be addressed further. The data for this study were acquired from an extensive longitudinal Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 research project, funded by the National Institute of Drug Abuse. While these data offered a means of prospectively assessing concordant and discordant drug use in relationships, there were also limitations inherent in the data set. The limitations in the study stem, for the most part from data that were not available, due to the fact that the study was designed after the data were collected. The data limitations are all associated with relationship measurement issues. With the particular data available, the assessment of break-ups in relationships over time was difficult. While divorce was a measured variable to assess the endings of marriages, many relationships were not assessed. For instance, individuals who were dating or cohabiting, could end a relationship during the intervening four years between data gathering points, and this would not appear in the data. Along the same lines, a separation could occur in a married couple during the intervening time and this again would not be measured. In addition, individuals could have been in a relationship at both points in time, yet have different partners. This lack of measurement of changes in relationship status could have affected the results, especially in the attempt to show the presence of socialization in a couple over time. In the results, the hypothesis predicting that concordant drug use would increase over time, due to socialization, was not confirmed. The Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 inability to monitor changes in relationship status likely affected these analyses. An additional limitation occurs in the drug use data. While information was available on both the subject and partner drug use, the subject completed the questions concerning partner drug use. The accuracy of the partner's level of drug use comes into question, when someone else is assessing use levels. More information would also have been available if the partner had filled out a complete questionnaire, similar to that of the subject. While the results of the current study are abundant in number, a broader data set would produce richer outcomes. Directions for Future Research This study has prompted ideas for future research, both to improve upon the limitations addressed above and to extend the research begun here. The study of homophily on drug use is an important one and could be continued. The contribution of socialization to homophily remains difficult to tease out, empirically. Further research focusing on this portion of the homophily research would add to the greater understanding of drug use in intimate relationships. One possibility to tease out socialization would be to utilize a sample of couples, collecting the same data from each member of the couple and following the couples over time. By controlling for changes Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 in partners and in relationship changes, the contribution of socialization might be isolated. Another intriguing expansion of the current study would be to extend the self-medication research into the area of abuse in intimate relationships. Battered women are fifteen times more likely to abuse alcohol and nine times more likely to abuse drugs than women who are not in abusive relationships (Stark & Flitcraft, 1996). At the same time, women in abusive relationships are three times more likely to be diagnosed with depression as compared to other women. It seems that a relationship may exist among the variables of abuse, negative affect (in this case depression), and drug use. On the other side, the physical abuse of the batterer has been viewed in relation to drug use in the literature, not as a cause of the violence, but as a concurrent factor (Van Hasselt, Morrison, & Bellack, 1985; Leonard, Bromet, Parkinson, Day, & Ryan, 1985) . The possibility that self- medicating of hostile, agitated affect is occurring in batterers exists and is a future research topic. Conclusions This study shifted the focus from the individual's drug use in an intimate relationship to the idea of concordant drug use in relationships. In addition, the concept of discordant drug use in relationships was also introduced. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Concordant drug use does occur in some relationships, supporting the idea of assortative mating for drug use. Individuals tend to seek similar individuals for intimate relationships and the similarities include level of drug use. The study also showed that discordant drug use occurs in relationships and influences the quality of the relationship and the presence of negative affect. All of the findings, however, should be interpreted with some caution until research with less limitations replicates the results. The results also furthered the research on self- medication and negative affect. Feelings of dysphoria and agitation are self-medicated by some individuals, using various substances to lessen the painful affect. The discordant drug use of a partner evoked increased feelings of dysphoria and agitation in the other member of the relationship. Putting these findings together produces a possible cycle, in which the drug use of one member produces negative affect in the other, which may in turn lead to self- medication of those painful feelings. Based on the study, social support was found to be a possible mediating variable for both drug and relationship problems. Social support for drug problems lead to decreased drug use, individually and as a couple. Social support for relationship problems was related to both improved intimate relations and decreased agitation. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64 In summary, this study found numerous results which added to the understanding of drug use in the context of intimate relationships. The dynamic aspect of an intimate relationship was taken into account by viewing a couple as a system. Drug use impacts a system via decreased quality of relationship and at times, complete deterioration of the relationship. It impacts the individual within the system through negative affect and the accompanying need to self- medicate that affect. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 References Anglin, M.D., Kao, C., Harlow, L.L., Peters, K. & Booth, M.W. (1987). Similarity of behavior within addict couples.Part I. Methodology and narcotics patterns . International J o u r n a l o f the A d d i c t i o n s , 22, 497-524. Beach, S.R. & O'Leary, K.D. (1993). Marital discord and dysphoria: For whom does the marital relationship predict depressive symptomatology? J o u r n a l o f Social a n d P e r s o n a l R e l a t i o n s h i p , 10, 405-420. Bentler, P.M. (1989). E Q S : S t r u c t u r a l E q u a t i o n s P r o g r a m M anual. Encinco, CA: Multivariate Software, Inc. Berenson, D. (1976) . Alcohol and the family system.In P. J. Guerin (Ed.) F a m i l y T h e r a p y : T h e o r y a n d P r a c tice. New York:Gardner Press, 284-297. Blatt, S.J., Quinlan, D.M., Chevron, E.S., McDonald, C., & Zuroff, D. (1982). Dependency and self-criticism: Psychological dimensions of depression. J o u r n a l of C o n s u l t i n g a n d C l inical Psychology , 50, 113-124. Burns, D.D., Sayers, S.L., & Moras, K.(1994). Intimate relationships and depression: Is there a causal connection? J o u r n a l o f C o n s u l t i n g a n d C l i n i c a l Psychology, 62, 1033-1043. Caspi, A. & Herbener, E.S.(1990). Continuity and change: Assortative marriage and the consistency of personality in adulthood. J o u r n a l o f P e r s o n a l i t y and S o c i a l Psychology, 58, 250-258. Clark, J.S., Capel, W.C., Goldsmith, B.M., & Stewart, G.T. (1972). Marriage and methadone:Spouse behavior patterns in heroin addicts maintained on methadone. J o u r n a l of M a r r i a g e a n d the Family, 34, 496-502 . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 Crumbaugh, J.C. & Maholick, L.T.(1964). An experimental study in existentialism: The psychometric approach to Frankl's concept of noogenic neurosis. J o u r n a l o f C l i n i c a l Psychology , 20, 200-207. Edwards, M.E. & Steinglass, P. (1995). Family therapy treatment outcomes for alcoholism. J o u r n a l o f M a r i t a l a n d F a m i l y Therapy, 21, 475-509. Haber, R.A.(1983). The family dance around drug abuse. The P e r s o n n e l a n d G u i d a n c e Journal, 61, 428-431. Hafner, R.J. & Spence, N.S. (1988). Marriage duration, marital adjustment and psychological symptoms: A cross- sectional study. J o u r n a l o f C l i n i c a l P s y c h o l o g y , 44, 309-316. Hall, R.L., Hesselbrock, V.M., Stabenau, J.R.(1983a). Familial distribution of alcohol use:I.Assortative mating in the parents of alcoholics. B e h a y i o r Genetics, 13, 361-372. Hall, R.L., Hesselbrock, V.M. , Stabenau, J.R.(1983b). Familial distribution of alcohol use: II.Assortative mating of alcoholic probands. B e h a y i o r Genetics, 13, 373-382. Harlow, L.L., Newcomb, M.D., Sc Bentler, P.M. (1986). Depression, self-derogation, substance use, and suicide ideation: Lack of purpose in life as a mediational factor. J o u r n a l o f C l i n i c a l Psychology, 42, 5-21. Hawkins, J.L. (1968). Associations between companionship, hostility, and marital satisfaction. J o u r n a l o f M a r r i a g e a n d the Family, 30, 647-650. Hollingshead, A.B. (1950) . Cultural factors in the selection of marriage mates. A m e r i c a n S o c i o l o g i c a l Review, 15, 619-627. Kandel, D.B. (1984). Marijuana users in young adulthood. A r c h i v e s o f Ge n e r a l Psychi a t r y , 41, 200-209. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 Kandel, D., Davies, M., & Baydar, N. (1990). The creation of interpersonal contexts:Homophily in dyadic relationships in adolescence and young adulthood. In L.N. Robins Sc M. Rutter (Eds.).S t r aight and Devious Pathways f r o m C h i l d h o o d to Adulthood. Cambridge:Cambridge University Press, 221-241. Khantzian, E.J. (1977). The ego, the self, and opiate addiction: Theoretical and treatment consideration. In J.D. Blaine & D.A. Julius (Eds) . P s y c h o d y n a m i c s o f D r u g D e p e n d e n c e :NIDA R e s e a r c h M o n o g r a p h 12. Washington, D .C .:U.S.Department of Health, Education, and Welfare, 101-117. Khantzian, E.J. (1985). The self-medication hypothesis of addictive disorders: Focus on heroin and cocaine dependence. The A m e r i c a n J ournal o f P s y c h i a t r v , 142, 1259-1264. Khantzian, E.J.(1986). A contemporary psychcdynamic approach to drug abuse treatment.American Journal o f D r u g and A l c o h o l Abuse, 12, 213-222. Khantzian, E.J. (1990). Self-regulation and self-medication factors in alcoholism and the addictions: Similarities and differences. In M.Galanter (Ed). Recent D e v e l o p m e n t s in A l c o h o l i s m , V o l . 8. New York:Plenum Press, 255-271. Leonard, K.E., Bromet, E.J., Parkinson, D.K., Day, N.L., S c Ryan, C.M. (1985). Patterns of alcohol use and physically aggressive behavior in men. J o u r n a l o f S t u d i e s on Alcohol, 46, 279-282. Lerman, C., Audrain, J., Orleans, C.T., Boyd, R., Gold, K . , Main, D., Sc Caporaso, N. (1996) . Investigation of mechanisms linking depressed mood to nicotine dependence. A d d i c t i v e Behaviors, 21, 9-19. Levenson, R.W. S c Gottman, J.M. (1978) . Toward the assessment of social competence. J o u r n a l o f C o n s u l t i n g a n d C l i nical Psychology , 46, 453-462. Merikangas, K.R.(1982). Assortative mating for psychiatric disorders and psychological traits. A r c h i v e s o f General Psychiatry , 39, 1173-1180. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 Merikangas, K.R. & Spiker, D.G.(1982). Assortative mating among in-patients with primary affective disorder. P s y c h o l o g i c a l M e d i c i n e , 12, 753-764. Murphy, S.L. & Khantzian, E.J. (1995). Addiction as a "self- medication" disorder:Application of ego psychology to the treatment of substance abuse. In A. M. Washton (Ed). P s y c h o t h e r a p y a n d S u b s t a n c e Abuse: A P r a c t i t i o n e r ' s Ha n d book. N e w York: The Guilford Press, 161-175. Newcomb, M.D. (1986). Cohabitation, marriage, and divorce among adolescents and young adults. J o u r n a l o f S o c i a l a n d P e rsonal R e l a t i o n s h i p s , 3, 473-494. Newcomb, M.D.(1990). Social support by many other names: Towards a unified conceptualization. J o u r n a l o f S o c i a l a n d P e r s o n a l R e l a t i o n s h i p s , 7, 479-494 . 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 of C o n s u l t i n g a n d C l i n i c a l Psvchologv, 62, 463-476. Nev/comb, M.D. & Bentler, P.M. (1985) . The impact of high school substance use on choice of young adult living environment and career direction. J o u r n a l of D r u g Education, 15, 253-261. Newcomb, M.D. & Bentler, P.M.(1986). Cocaine use among young adults .Advances in A l c o h o l a n d S u b s t a n c e Abuse, 6, 73- 98 . Newcomb, M.D. & Bentler, P.M.(1987). Changes in drug use from high school to young adulthood: Effects of living arrangement and current life pursuit. J o u r n a l o f A p p l i e d D e v e l o p m e n t a l Psycho l o g y , 8, 221-246. Newcomb, M.D. & Bentler, P.M. (1988a).C o n s e q u e n c e s of A d o l e s c e n t D r u g Use. Newbury Park, CA: Sage. Newcomb, M.D. & Bentler, P.M. (1988b). Impact of adolescent drug use and social support on problems of young adults: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 A longitudinal study. J o u r n a l o f A b n o r m a l P sychology , 97, 64-75. Newcomb, M.D. & Chou, C.P. (1989) . Social support among young adults: Latent-variable models of quantity and satisfaction within six life areas. M u l t i v a r i a t e B e h a v i o r a l R e s e a r c h , 24, 233-256. Newcomb, M.D. & Rickards, S.(1995). Parent drug-use problems and adult intimate relations: Associations among community samples of young adult women and men. J o u r n a l o f C o u n s e l i n g P s y c h o l o g y , 42, 141-154. Newton, T.L., Kiecolt-Glaser, J.K., Glaser, R., & Malarkey, W.B. (1995). Conflict and withdrawal during marital interaction: The roles of hostility and defensiveness.P e r s o n a l i t y a n d Social P s y c h o l o g y B u l l e t i n , 21, 512-524. Petrie, K. & Chamberlain, K. (1983). Hopelessness and social desirability as moderating variables in predicting suicide behavior. J o u r n a l o f C o n s u l t i n g a n d C l i n i c a l P s y c h o l o g y , 51, 485-487. Price, R.A., Chen, K.H., Cavalli-Sforza, L.L., & Feldman, M.W. (1981). Models of spouse influence and their application to smoking behavior. Social B i ology, 28, 14- 29 . Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the general population. A p p l i e d P s y c h o l o g i c a l M e a s u r e m e n t , 1, 385-401. Scheier, L.M. & Newcomb, M.D. (1993) . Multiple dimensions of affective and cognitive disturbance: Latent-variable models in a community sample. P s y c h o l o g i c a l A s s e s s m e n t , 5, 230-234. Smith, T.W., Pope, M.K., Sanders, J.D., Allred, K.D., & O'Keeffe, J.L.(1988). Cynical hostility at home and work: Psychological vulnerability across domains. J o u r n a l o f R e s e a r c h in P e r s o n a l i t y , 22, 525-548. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 Spanier, G.B.(1976). Measuring dyadic adjustment: Mew scales for assessing the quality of marriage and similar dyads. Journal of M a r r i a g e a n d the Family, 38, 15-28. Stanton, M.D. (1980) . Drug abuse and the family. In M. Andolfi & F. Zwerling (Eds) . D i m e n s i o n s o f F a m i l y Therapy. New York: The Guilford Press, 29-46. Stanton, M.D. & Shadish, W.R. (1997) . Outcome, attrition, and family-couples treatment for drug abuse: A meta-analysis and review of the controlled, comparative studies. P s y c h o l o g i c a l Bulletin, 122, 17 0-191. Stanton, M.D., Todd, T.C., Heard, D.B., Kirschner, S., Kleiman, J.I., Mowatt, D.T., Riley, P., Scott, S.M., Sc VanDeusen, J.M. (1982) . A conceptual model. In M.D. Stanton & T.C. Todd (Eds) . T h e F a m i l y T h e r a p y o f D r u g A b u s e a n d A ddiction. New York: The Guilford Press, 7-30. Stark,E. & Flitcraft, A. (1996). W o m e n at Risk: D o m e s t i c V i o l e n c e a n d Women's Health. Thousand Oaks, CA: Sage Publications, 157-200. Steinglass, P.(1985). Family systems approaches to alcoholism. Journal o f S u b s t a n c e A b u s e T r e a t m e n t , 2, 161-167. Steinglass, P., Davis, D.I., & Berenson, D. (1977) . Observations of conjointly hospitalized "alcoholic couples" during sobriety and intoxication for theory and therapy. F a m i l y Process, 16, 1-16. Uhlenhuth, E.H., Balter, M.B., Mellinger, G.D., Cisin, I.H., Sc Clinthorne, J. (1983) . Symptom checklist syndromes in a general population. A r c h i v e s in General Psychiatry , 40, 1167-1173. Van Hasselt, V.B., Morrison, R.L., Bellack, A.S. (1985). Alcohol use in wife abusers and their spouses. A d d i c t i v e Behaviors, 10, 127-135. Wallen, «J.(1992) . A comparison of male and female clients in substance abuse treatment. J o u r n a l o f S u b s t a n c e A b u s e Treatment, 9, 243-248. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71 Weiss, R.D., Griffin, M.L., & Mirin, S.M.(1992). Drug abuse as self-medication for depression: An empirical study. A m e r i c a n J o u r n a l o f D r u g a n d A l c o h o l A b u s e , 18, 121-129. Yamaguchi, K. & Kandel, D.B.(198 5a). Dynamic relationships between premarital cohabitation and illicit drug use: An event-history analysis of role selection and role socialization. A m e r i c a n S o c i o l o g i c a l Review, 50, 530- 546. Yamaguchi, K. & Kandel, D.B. (1985b). On the resolution of role incompatibility: A life event history analysis of family roles and marijuana use. A m e r i c a n J o u r n a l of Sociology, 90, 1284-1325. Yamaguchi, K. & Kandel, D.(1993). Marital homophily on illicit drug use among young adults: Assortative mating or marital influence? S oci a l Forces, 72, 505-528. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 Year 13 Year 17 [s^Aic Partner Poly Partner Sexual Satis | Ua sI Sexual Satis Good Intimate Relation ood Intimate Relation DAS I ^Dependence | Dating Comp # of divorces past 4 years #of divorces current rel Suicidal Ideation Suicidal Idejtion sphoria Dysphona Purpose in Life I'urpose in Life rV I Agitation J Hostihtv Agitation Anxiety xiety social Farrulv social support dnjgs support | Friends | Agency Agency Family Family social support social Friends support . rel. I Friends V — Agency | Agency | Sex Figure 1.1. Hypothesized path model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 Table 2.1 Description of Sample in Year 17 Variable M ale(N =95) Female (N=268) Total (N=363) Age (years) Mean Range Ethnicity (%) African-American Latino Anglo Asian-Pacific 29.9 25-32 3 15 74 8 29.9 28-32 13 12 65 9 29.9 25-32 11 13 67 9 Education (years) Mean 14.7 Range 10-18 Number of children (%) None 0 One 45 Two 42 Three or more 13 Income for past year (%) Under $10,000 3 $10,000 to $30,000 18 $30,000 to $50,000 35 Over $50,000 44 14.1 9-18 1 39 34 25 9 22 30 39 14.3 9-18 1 41 36 it 8 21 31 40 Living Situation (%) Alone 5 Parents 2 Spouse 79 Cohabitation 8 Dormitory 0 Roommates 6 Single Parent 0 No Regular Place 0 Current Life Activity Unemployed, laid off, fired 2 Military 5 College 3 Child rearing/part time job 0 Part - time job 4 Full -time job 81 Child rearing/homemaker 0 None 4 8 6 66 7 0 4 9 0 6 0 5 1 2 5 57 18 1 8 5 69 7 0 5 6 0 0 1 2 9 4 63 13 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74 Table 3.1 Summary of Model Fit Indices and Nested Tests Model X2 df x2 :df p CFI NN FI Multiple G roup Analyses: A. All free 2318.77 1502 1.54 <.001 .88 .85 B. FL= 2382.30 1544 1.54 <.001 .88 .85 Difference test (B-A) 63.53 42 — <.025 — -- C. FL and COV= 2516.38 1664 1.51 <.001 .88 .86 Difference test (C-B) 134.08 120 — >.05 — — Difference test (C-A) 197.61 162 — <.05 — — D. FL, COV and E's= 2591.05 1697 1.53 <.001 .87 .86 Difference test (D-C) 74.67 33 — <.001 — — Difference test (D-B) 208.75 153 — <.005 — — Difference test (D-A) 272.28 195 — <.001 — — Overall Model: CFA, 1st O rder Factors 1384.46 807 1.72 <.001 .91 .89 CFA, 2nd O rder Factors 1482.44 840 1.76 <.001 .90 .88 Final STR 1389.50 866 1.58 <.001 .92 .91 Note. Dashes indicate that data were not applicable. CFA= confirmatory factor analysis. CFI= comparative fit index (Bentler, 1990). NNFI= non-norm ed fit index (Bentler, 1989). FL= factor loading. An equal sign indicates that these param eters were constructed to equality between sex groups. STR= Structural or Path Model Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I ( ) ( ) > d 2 ( ) > d : q K ) > d : . i b '() > d :u 75 3s yi !u N — V g V > a ^ X < - . 2 s 3« v. ■o. g C * •i r* -n 2 - 2 . X a - s a t f l £ . < r Si ^ > 7 5 n . r — X ^ » o S ' & l ? c - ^ c; -- ' - j “ = - ^ ^ ; ! j 5 ' j i - ^ 'i 2 w 2 s = - ? * - j ! - 1 j s t_— f • ^ • - 1 * " f c m V I ? k t p . \ m ^ ■ * * • — T ? w w ~ vj f. ^ v; “ f'*' r > J v + - - — * V I ^ J £ N * * - - O N > V iro ^ vj vj ji vj * J ' - J '*) '*) ^ X ^ ^ c C Se ~ j- ; -n “ — 'f f* A. iu j. 2 ' ■ ' T - ' V I X * X L - wV,J'7**C'Ci,5v'£X3'3i:Wstf3e -5S55 — bww'Siic'i- — c : — w ^ r ° > " — 9 b — r u r - o v i m “ T _ ~ 1 H C i ~ £ . - J r 5 , Z - • „ • , — ! u 5 * “*^#3w c^s 55 i . w i- * . j » i ? — ^ vn -n ^ -i * 2 * f - w * . w j - j ~ O vi '-n X 5 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le .12 Currddtion Matrix - M ale v s . Fem.ile # of divorces past S years .74/ 69 Suicidal Ideation .7 8 /-.SO Purpose m Life Hostility Agitation Anxiety I Family | ^, 63 - 60 | PnenJs| Agency social support drugs 38/.69 Family social support SS/.74 Fnends 39 / 38 * of divorces past 4 years Current Rel. SO SO ^uiadal Ideation phona I Hostility Agitatio | Anxiety Family rpport Family social support ^ drugs Sex Figure 3.1. M ale/fem ale factor loadings. * p<.02 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 8 Sox / / Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78 Year 13 Year 17 © - ► [ S ] CD— <g>— ► C T CZ2)-^I p^ r. I P M c . | © - ► [ s = Satis Partner Polv DC Intimate Relation Suicidal Ideation Purpose in L i f e Hostility | Anxiety"| Agitation I CM)— ► [ Family 7y y social support drugs Agency CD— ►[? social Fnends ( ^ 8 6 ) ^.[ Ag e n c y Sex S-Alc | Partner Polv DL' | P-Alc Good Intimate Relation I Sexual Satis r[DAr]^ (io) | Dependence Dating Comp. ] • * - © # of divorces # of divorces past 8 years past 4 years current r e l . Suicidal Ideatinn Oygphnnn | 66 ^82 Purpose in Life Hostility Agitation L^64 Anxiety social support drugs h*— CE> Family D * — © Fnends — C D Agency Familv social support Fnends Figure 3,2- CFA model for first order factors. All paths significant at p<.001. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. u x i> J :p zo>d 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.4 C F A Modelinteirorrelations fu r 2 n d Order Factors Year 13 Year 17 80 © — ►CHoi] O —►[ZHZl (2)— ► G i n © - ► r^ar . ( § ) - ► [ P-Alc. | ® > [ p¥] (fc j)— ^► |Sexuol Satis O ^ fDAS ^97^ ^ [Dependence © -► (Silence Partner Partner I Sexual Satis I DAS 1 ^ ( - t v ) | Dependence Cood Intimate Relation Good Intimate Relation Dating Comp. # of divorces # of divorces past 4 years past 8 years current rel. | Hostility © ) -------► I Anxiety | Family ( © ------ ^ ► [Friends [■ Agitation I 76 T social support 39— drugs Agency ® ------- ► [ Family (3 9 ^ — -^ ► [“Friends Agency CES-D I-. - ( E ) Suicidal Ideation - .JO*} 3vsph on a] ( Dysphoria L _ E l^ . \ y ^ 8 3 " " Suicidal ^ — ( E ) © — ► Purpose in — Life. Purpose in Life ( E ) Hostility Agitation L.65 Anxiety I social support drugs Family ~ T ^ — ( E ) Friends Agency Family social support Fnends Sex Figure 3-3. CFA model for second order factors. All paths significant at p<.001 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 Year 13 Year 17 .02 Self Self .89' .81 Couple DL f Couple ^ V , D L .83' Partner Poly D L Partner .30 -.16' - .22' -.05' Cood Intimate Relation Coud Intimate Relation .54' .23' # of divorces past 4 years .93 # of divorces .12' current rel. .68*' Dysphoria Dysphoria ) ■ * — © - . 12 ' .3 4 ' Agitation I Agitation -.13' social support drugs social support drugs .71 .4 1 .16' social ST' .30' import Sex Figure 3.4. Final path model. *p<.05, **p<.01, ***p<.001 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.5 Specific Effects Not Depicted in Figure 82 -Predictors Consequences Latent Variable Manifest Variable Latent Variable Manifest Variable Standardized Path Coefficient --------------- Drug Use Predictors---------- Good Intimate Relations Good Intimate Relations self cigarette use partner cigarette use dependence (R) Partner Polvdrug Use suicidal ideation(R) Self Polydrug Use hostility (R) Partner Polydrug Use friend social support Self Polydrug -drugs (R) Use family social support Partner Polydrug -drugs (R) Use Drug Use Consequences — Couple Polydrug Use Self Polydrug Use (R) Self Polydrug Use (R) dyadic adjustment dyadic adjustment purpose in life self cigarette use (R) Dysphoria self cigarette use (R) Agitation self alcohol use (R) self illicit drug use (R) Dysphoria Partner Polydrug Use (R) Partner Polydrug Use (R) Partner Polydrug Use (R) Partner Polvdrug Use (R) Partner Polydrug Use (R) current relationship dating competence suicidal ideation purpose in life hostility family social support -drugs partner cigarette use (R) divorces past 4 years -. 10* * .09* .07* .18*** .13** .14** .10** .14** .15** .11* .08* .12* * .12** .07* .27** .17*** .20*** .13** .12* * .21*** Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.6 Partner-Self Drug Use Correlations Over Time 83 Drug Year 13 Year 17 Polydrug Use .79 .91 Cigarette Use .41 .47 Alcohol Use .41 .40 Illicit Drug Use .48 .43 Note: N(total) = 298; N(women) = 218; N(men) = 80. Subjects included were in a relationship at Year 13 and Year 17 and had no divorces in the intervening time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IMAGE EVALUATION TEST TARGET (Q A -3 ) 1. 0 l.l 1 .2 5 1 2.8 * y* ! ” UZ 1 . 4 lii 12 . 2 2.0 1 . 8 150mm 6" V J* * /> V o / /4 P R L IE D A IIW 1G E . Inc . = 1653 East Main Street - ■ Rochester. NY 14609 USA — Phone: 716/482-0300 - = ~ - = Fax: 716/288-5989 O 1993. Applied Image. Inc.. All Rights Reserved Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A cross-cultural comparison of marital power and dyadic adjustment among American, Indo-American, and East Indian dual-career and single-career couples

PDF

Childhood precursors of adult reality distortion: A prospective study.

PDF

Perceptions of parental behaviors by a sample of gifted and no gifted women: an exploratory study

PDF

The paradox of loss: Toward a theory of grief

PDF

A correlational study of feminist/womanist identity development and depression in women

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

Counselor Trainees' Perceptions Of And Attitudes Toward Sexual Harassment In Relation To Feminist Beliefs

PDF

An adolescent -parent conflict resolution skills training program for ethnically diverse families: A program evaluation study

PDF

Impact of language and culture on a neuropsychological screening battery for Hispanics

PDF

A Comparison Of Female Inmates With And Without Histories Of Prostitution On Selected Psychosocial Variables

PDF

Gender differences in motivation for sexual intercourse: Implications for risky sexual behavior and substance use in a university and community sample

PDF

A comparative study of the family processes of youth of Mexican descent who have dropped out of high school and those who have remained in school

PDF

Cognitive predictors of intention -behavior consistency in safer sex practices

PDF

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

PDF

Self-Concept In Adult Women: A Multidimensional Approach

PDF

A cluster analysis of the differences between expert and novice counselors based on real time training

PDF

A longitudinal cross-ethnic study of sociocentricity in schizophrenia

PDF

Childhood Sexual Abuse and Depressive Symptoms In A Lesbian Population: An Exploratory Study.

PDF

Behavioral and demographic predictors of breast cancer stage at diagnosis

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

Asset Metadata

Creator Rice, Michelle A (author)

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

Degree Doctor of Philosophy

Degree Program Counseling Psychology

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

Tag OAI-PMH Harvest,psychology, clinical,psychology, social,sociology, individual and family studies

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Newcomb, Michael (committee chair), Earleywine, Mitch (committee member), Rosenberg, Joan (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-314992

Unique identifier UC11350871

Identifier 9835068.pdf (filename),usctheses-c17-314992 (legacy record id)

Legacy Identifier 9835068.pdf

Dmrecord 314992

Document Type Dissertation

Rights Rice, Michelle A.

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