1E-04 - Comparative risk of sedative-related use disorders following benzodiazepine treatment versus alternative treatments in adolescents and young adults

Background: The risk for misuse and dependency with benzodiazepine (BZD) treatment is the subject of a boxed-warning. However, it is unclear how often a sedative, hypnotic and anxiolytic disorder (SHA-UD) develops following prescription BZD treatment, and whether this occurs more frequently than with alternative medications in adolescents and young adults, a high-risk population. Understanding the risk of SHA-UD in adolescents and young adults prescribed BZD treatment can inform clinical management decisions.

Objectives: To estimate the incidence of a SHA-UD diagnosis in adolescents and young adults prescribed BZD treatment compared to initiators of an alternative medication.

Methods: We utilized MarketScan (2016-2022) administrative claims data covering privately insured enrollees in the US. We included adolescents (12-17 years) and young adults (18-24, 25-29 years) newly initiating BZD or trazodone treatment (comparator medication) with a 1+ year washout period with no prior BZD or trazodone use and without a prior substance use disorder diagnosis. Incident SHA-UD was defined as an inpatient/outpatient ICD-10-CM SHA-UD diagnosis (F13.x, excluding ‘in-remission’) within two years of the first prescription dispensing. We estimated the cumulative incidence of a SHA-UD diagnosis with Kaplan Meier analysis and hazard ratios (HR) with Cox regression. For adjusted results, we applied propensity score inverse-probability of treatment weighting within each age group.

Results: We identified 217,158 BZD new users, 68.3% were female and 13.6% aged 12-17y, 48.2% aged 18-24y, and 38.3% aged 25-29y. The majority (72.6%) had a prior mental health diagnosis and prior antidepressant prescription (53.8%). Two years after the initial BZD prescription, the cumulative incidence of a SHA-UD diagnosis was 0.45% for adolescents, 1.07% in young adults 18-24y and 0.47% in young adults 25-29y. The comparator cohort included 103,221 trazodone new users. In the propensity-score weighted, trimmed cohorts, the risk of a SHA-UD diagnosis was elevated for BZD treatment compared to trazodone treatment in those aged 18-24y (HR:1.55, 95%CI:1.26-1.90) and 25-29y (HR:2.13, 95%CI:1.19-3.80). In adolescents, the risk of a SHA-UD diagnosis was similar in BZD and trazodone initiators (HR:0.84, 95%CI:0.55-1.28).

Conclusions: In young adults, new BZD treatment was associated with increased risk of incident SHA-UD diagnosis compared to an alternative medication (trazodone); young adults aged 18-24 years appear more vulnerable to risks of SHA-UD. In adolescents, the comparable risk of SHA-UD in BZD and trazodone initiators may signify more cautious BZD prescribing and monitoring in this younger age group as recommended.