and little, if any, research on the transition from use to abuse and dependence.
Several investigators have noted that protective factors can moderate the effects of risk conditions, thereby reducing vulnerability and enhancing resiliency (Garmezy, 1985; Werner, 1989; Brook et al., 1990; Rutter et al., 1990). Protective factors that have been suggested based on analyses of cross-sectional data include a positive mutual attachment between parent and child (Brook et al., 1990), nondeviant siblings, academic achievement, positive group norms (Hawkins et al., 1992), and dimensions of conventionality such as low rebelliousness and adherence to broad social norms. Moreover, the effect of one parent's drug use can be offset by the nonuse of the other parent. Additional protective factors that have been identified in young adulthood include employment, marriage, and childrearing responsibilities. It has been noted that several protective factors can ameliorate the negative effects of exposure to extreme stress (Garmezy, 1985). These include the child's temperament, a supportive family, and an external support system that reinforces the child's efforts at coping (Brook et al., 1986a; Labouvie and McGee, 1986). Further research is needed, however, to determine which protective factors are relevant at different developmental stages, and more attention also needs to be given to mechanisms by which protective factors influence the onset and progression of drug abuse.
Additionally, research on drug use has documented substantial racial and ethnic differences in drug use among adolescents, such as lower use of tobacco products among African American adolescents than white adolescents (IOM, 1994a; Johnston et al., 1995). The reason for this particular difference is unclear; influences may include the church, cultural consensus against youth tobacco use, or lack of attention from advertisers (IOM, 1994a). Research should be conducted to enhance understanding of racial and ethnic differences in the acceptability of tobacco use and how these differences may be used in the design of prevention interventions related to other drugs of abuse.
Numerous longitudinal studies have identified the childhood antecedents of adolescent drug use (e.g., Kellam et al., 1983; Pulkkinen, 1983; Baumrind and Moselle, 1985; Block et al., 1986; Brook et al., 1986a). However, far less research has been done on identifying childhood risk factors associated with drug abuse and dependence that are not associated with behavior problems but with individual vulnerability factors (such as genetic predisposition and emotional disorders). Additionally, there has