How Predictors of Prescription Drug Abuse Vary by Age

Abstract

Previous studies that have identified the predictors of prescription drug abuse have either focused on a specific age group or pooled all age groups together into one sample. This approach constrains the predictors to have the same effect across age groups. In this study, we use the 2001 to 2003 National Survey on Drug Use and Health to estimate separate models across five age groups for the past year nonmedical use of prescription drugs. The results indicate that several factors (e.g., gender, race/ethnicity marital status, other substance use) have quite different correlations with prescription drug abuse across age groups. This suggests that more accurate profiles of prescription drug abusers can be obtained by estimating separate models for different age groups.