Epidemiology of DSM-5 Alcohol Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions III.
Level 4 - case-series / case-control
Nationally representative cross-sectional survey based on retrospective self-report interviews.
PubMed 26039070 · doi:10.1001/jamapsychiatry.2015.0584
What was done
Face-to-face structured interviews were conducted with 36,309 civilian, noninstitutionalized US adults aged 18 years and older as part of the 2012–2013 National Epidemiologic Survey on Alcohol and Related Conditions III (NESARC-III). Data were collected between April 2012 and June 2013 to assess 12-month and lifetime prevalence of DSM-5 alcohol use disorder (AUD) by severity (mild, moderate, severe), associated sociodemographic characteristics, psychiatric comorbidities, disability, and treatment utilization.
What was found
The 12-month and lifetime prevalences of DSM-5 AUD were 13.9% and 29.1%, respectively. Prevalences were highest among men (17.6% 12-month; 36.0% lifetime), Native Americans (19.2%; 43.4%), White respondents (14.0%; 32.6%), younger adults (26.7%; 37.0%), and never-married adults (25.0%; 35.5%). Severe AUD was highest in the lowest income tier (1.8% 12-month; 1.5% lifetime). Only 19.8% of individuals with lifetime AUD reported ever receiving treatment. Significant psychiatric comorbidities across all AUD severity levels included other substance use disorders, major depressive disorder, bipolar I, antisocial personality disorder, and borderline personality disorder, with odds ratios ranging from 1.2 (95% CI, 1.08–1.36) to 6.4 (95% CI, 5.76–7.22). Modest associations were observed for panic disorder, specific phobia, and generalized anxiety disorder (odds ratios 1.2 to 1.4).
Why it matters
This study provides the first comprehensive national baseline for DSM-5 alcohol use disorder in the United States, demonstrating that AUD affects nearly three in ten adults over their lifetime, is heavily comorbid with other psychiatric conditions, and remains overwhelmingly untreated.
Limits
The cross-sectional design cannot establish causality or determine the temporal sequence between AUD and comorbid conditions. Data relied entirely on retrospective self-report during face-to-face interviews, introducing potential recall and social desirability biases. The survey sampled only noninstitutionalized civilian adults, excluding homeless individuals not residing in shelters, incarcerated populations, and active-duty military personnel, which may underestimate true population prevalence.