The Relationship of Family Structure to Adolescent Drug Use, Peer Affiliation and Perception of Peer Acceptance of Drug Use
Jenkins, Jeanne E., Zunguze, Sabina T., Adolescence
Alcohol and other drug use among adolescents has been the focus of numerous empirical investigations. To unravel the complex etiology of this phenomenon, several factors have been studied. Among these, family and peers have been identified as prominent proximal influences.
Associations have been found between familial variables and children's drug involvement. Among others, parents' drug use (Ellis & Stone, 1979; Kandel, 1973; Newcomb, Huba, & Bentler, 1983), parental drug-use attitudes (Jessor & Jessor, 1977; Korsnick & Judd, 1982), child-rearing practices (Brook, Whiteman, & Gordon, 1983; Brook, Whiteman, Gordon, & Brook, 1985), and family disruption or lack of cohesion (Babst, Miran, & Koval, 1976; Jenkins & Guidubaldi, 1992) have all been found to relate to children's use of illegal substances.
Family structure appears to be an important variable. Several studies have shown that youth from disrupted families evidence more frequent substance abuse (Needle, Su, & Doherty, 1990; Doherty & Needle, 1991). In a study of psychoactive substance use disorder in youth suicide, findings revealed that parental divorce was more frequent among adolescents and young adults with the disorder (Runeson, 1990). Stein, Newcomb, and Bentler (1987) found that their family-disruption factor was significantly correlated with adolescent drug use, albeit largely mediated through lack of social conformity. In their review of the literature, Ganong and Coleman (1993) noted that children in stepfamilies evidenced more internalizing behavior problems (e.g., depression, anxiety) and more antisocial behavior than did children residing in nuclear families. Interestingly, stepchildren were found not to differ from children in single-parent families on either internalizing or externalizing behaviors. Other studies have found that socioemotional and behavioral problems among children and adolescents frequently reemerge with the introduction of a stepparent (Bray, 1988; Hetherington, Cox, & Cox, 1985).
Unfortunately, small and socioeconomically constricted samples and utilization of combined drug-use variables have limited the generalizability of some of the findings. In addition, some studies have failed to identify children's living arrangements or used confounded family-status variables (i.e., neglected to distinguish between different types of single-parent and intact households). For example, it is not uncommon to find intact-family categories inclusive of stepfamilies in studies on the effects of alcohol and other drug use. Similarly, in those studies that consider family structure, typically only the predominant mother-headed (single-parent) household is included.
Hirschi's (1969) social control theory and Jessor and Jessor's (1977) problem-behavior theory offer an explanation for the relationship between drug use and family structure. These theories propose that family problems may create an environment that is not positively grounded in traditional social institutions, which can lead to deviant behaviors, such as drug involvement. Inasmuch as divorce and family reconstitution have been found to be related to some negative outcomes for children, these theories, in part, explain the hypothesized association.
In addition to family influences, affiliation with drug-using peers has been found to represent a strong correlate of drug experimentation, and thus is a critical risk factor for adolescents (Hawkins, Lishner, & Catalana, 1985; Needle et al., 1986). Peer environments accepting of drug use may be more inviting to adolescents from nontraditional family structures, particularly younger adolescents, who are not only struggling with developmental challenges, but also coping with the stresses associated with single-parent and stepparent families.
The present study sought to contribute to our understanding of adolescent drug involvement and improve upon previous designs by using several drug indicators and a large sample representative of different family types. …