Risks for early substance involvement associated with parental alcoholism and parental separation in an adolescent female cohort

Background: We examined timing of substance involvement as a joint function of parental history of alcoholism and parental separation during childhood. Method: Data were drawn from a large cohort of female like-sex twins [n= 613 African Ancestry (AA), n= 3550 European or other ancestry (EA)]. Cox proportional hazards regression was conducted predicting age at first use of alcohol, first alcohol intoxication, first use and regular use of cigarettes, and first use of cannabis and other illicit drugs from dummy variables coding for parental alcoholism and parental separation. Propensity score analysis was also conducted comparing intact and separated families by predicted probability of parental separation. Results: In EA families, increased risk of substance involvement was found in both alcoholic and separated families, particularly through ages 10 or 14 years, with risk to offspring from alcoholic separated families further increased. In AA families, associations with parental alcoholism and parental separation were weak and with few exceptions statistically nonsignificant. While propensity score findings confirmed unique risks observed in EA families, intact and separated AA families were poorly matched on risk-factors presumed to predate parental separation, especially parental alcoholism, requiring cautious interpretation of AA survival-analytic findings. Conclusion: For offspring of European ancestry, parental separation predicts early substance involvement that is not explained by parental alcoholism nor associated family background characteristics. Additional research is needed to better characterize risks associated with parental separation in African American families.