Risk Factors for Adolescent Substance Abuse and Dependence

A national sample of 4,023 adolescents aged from 12 to 17 years of age were interviewed about substance abuse, experiences with victimization, substance abuse within their family, and posttraumatic reactions to identify risk factors of substance abuse. The survey was done by speaking to adolescents over the telephone and asking them questions. Of the 4,023 people who participated, 3,161 were a national probability household sample of adolescents. The other 862 people were considered a probability oversample of adolescents from households in designated central cities. This oversample was created to diversify the number of racial and ethnic minority respondents. Each participant was classified by three dimensions: age, race and gender. Prior to each adolescent interviewed on the phone, a guardian was briefly interviewed and was asked a few questions. There was a 75% participation and completion rate for the sample.

The telephone interview was based to collect information including demographic characteristics, adolescent and familial substance use patterns, witnessed violence, sexual assault and physical assault. The questions asked concerned age, gender, race, nonexperimental alcohol use, nonexperimental marijuana use, hard drug use, age of onset for nonexperimental substance use, substance dependence, sexual assault, physical assault, witnessed violence, familial alcohol problems, and familial drug use. Interviewers asked the adolescents if they were assured privacy so they could answer freely. The questions asked were more or less closed-ended questions. Each participant was protected by federal law governing research study.

Identifying and quantifying risk relationships between study variables and substance use was the primary data-analytic strategy. The data was presented in the terms of sample prevalence and odds ratios. For examination of the impact of each variable on problem use of each substance above the effects of other variables, a five-step hierarchical logistic regression was modeled. The odds ratios of variable were adjusted for the variables on that step and the preceding step. The data was examined carefully to assure that the findings were indeed valid.

Findings of this study included adolescents who had been physically assaulted, sexually assaulted, witnessed violence, or who had fimaily member with alcohol or drug use problems had an increased risk for substance abuse or dependence. Also, posttraumatic