Time to generalized anxiety disorder onset after substance use disorder: a survival analysis from a national sample

Substance use disorders (SUD) frequently co-occur with Generalized Anxiety Disorder (GAD), yet the timing and likelihood of GAD emergence following SUD onset remain poorly characterized. This study estimated substance-specific cumulative incidence and hazard of incident GAD after alcohol use disorder (AUD; n = 6685), cannabis use disorder (CUD; n = 1500), and nicotine use disorder (NUD; n = 5673), benchmarked against adults with no lifetime SUD (n = 16,117). Data were drawn from the National Epidemiologic Survey on Alcohol and Related Conditions-III (NESARC-III), a U.S. nationally representative survey of 36,309 adults conducted in 2012-2013. Time-to-event was measured from age at SUD onset (or age 30 in controls) to age at first GAD episode. We applied survey-weighted Kaplan-Meier estimation of cumulative incidence and Cox proportional hazards regression with Taylor-series variance estimation and examined clinical and sociodemographic predictors of GAD onset within each cohort. All SUD groups showed a higher estimated likelihood of subsequent GAD onset relative to the no-lifetime-SUD control group. Notably, CUD conferred the highest 10-year cumulative incidence of GAD onset (7.06%) and the steepest early hazard trajectory, exceeding both NUD (5.45%) and AUD (5.22%), and far exceeding controls (1.34%). In a pooled any-SUD cohort combining the three groups (n = 9812; 862 incident GAD episodes), the 10-year cumulative incidence was 4.78%, and risk increased stepwise with the number of co-occurring substance use disorders. Across SUD groups, younger age, comorbid non-GAD anxiety, mood, and personality disorders were consistently associated with increased risk, whereas remission was protective. These results may inform targeted screening, monitoring, and prevention strategies to interrupt development of GAD.