Abstract
OBJECTIVES: We identified peak annual incidence rates for medical and nonmedical use of prescription opioid analgesics, stimulants, sedatives, and anxiolytics (controlled medication), and explored cohort effects on age of initiation. METHODS: Data were gathered retrospectively between 2009 and 2012 from Detroit area students (n = 5185). Modal age at the last assessment was 17 years. A meta-analytic approach produced age-, year-, and cohort-specific risk estimates of first-time use of controlled medication. Cox regression models examined cohort patterns in age of initiation for medical and nonmedical use with any of 4 classes of controlled medication (opioid analgesics, stimulants, sedatives, or anxiolytics). RESULTS: Peak annual incidence rates were observed at age 16 years, when 11.3% started medical use, and 3.4% started using another person's prescription for a controlled medication (ie, engaged in nonmedical use). In the more recent birth cohort group (1996-2000), 82% of medical users and 76% of nonmedical users reported initiating such use by age 12 years. In contrast, in the less recent birth cohort group (1991-1995), 42% of medical users and 35% of nonmedical users initiated such use by age 12 years. Time to initiation was 2.6 times less in the more recent birth cohort group (medical use: adjusted hazard ratio [aHR] = 2.57 [95% confidence interval {CI} = 2.32-2.85]; nonmedical use: aHR = 2.57 [95% CI = 2.17-3.03]). CONCLUSIONS: Peak annual incidence rates were observed at age 16 years for medical and nonmedical use. More recent cohorts reported initiating both types of use at younger ages. Earlier interventions may be needed to prevent adolescent nonmedical use of controlled medication.