The prevalence of subthreshold psychiatric symptoms and associations with alcohol and substance use disorders: from a nationally representative survey of 36,309 adults

Abstract

Abstract Background Our aim was to describe a broad number of subthreshold psychiatric symptoms (SPS) in a nationally representative population and evaluate associations with substance use. SPS describe groups of symptoms with significant pathology, but that do not quite meet full psychiatric diagnostic criteria. They have been associated with significant impairment and cost. Methods The National Epidemiologic Survey on Alcohol and Related Conditions-III was a multistage, weighted, cross-sectional survey completed in the United States in 2013 comprising 36,309 noninstitutionalized adults. We report lifetime prevalence rates of 14 SPS related to mood, anxiety, trauma, eating, and personality disorders. We then evaluate associations with lifetime alcohol use disorders (AUD) and all substance use disorders (SUD) using logistic regression and adjusted odds ratios. SPS and psychiatric diagnoses were mutually exclusive (could not co-occur). Results Lifetime prevalence of having at least one of 14 SPS was 57% compared with 37% for the related psychiatric disorders. This was similar for males and females, in contrast to psychiatric disorders in which prevalence was 42% in females and 31% in males. Otherwise, overall SPS and disorders had similar prevalence patterns across sociodemographic characteristics. Subthreshold personality symptoms had the highest prevalence rates (schizotypal 21.3%, antisocial 18.3%, and borderline 17.6%), followed by posttraumatic stress (13.1%). Subthreshold bipolar and depression had lifetime prevalence rates of 2.7 and 8.5%, respectively. Prevalence rates of subthreshold anxiety symptoms ranged from 2.2% (agoraphobia) to 9.8% (specific phobia). Subthreshold eating disorder related symptoms had the lowest prevalence rates (anorexia 1.5% and bulimia 1.7%). Half (seven) of the SPS had significantly increased odds of lifetime AUD. This number increased to 12 for all SUD. Subthreshold antisocial personality symptoms had the highest odds of AUD (2.2; 95% CI 2.00–2.37) and SUD (3.5; 95% CI 3.22–3.81). Conclusions We found high lifetime SPS prevalence rates and significant associations with AUD and SUD. To our knowledge, this is the first published study evaluating a broad number of SPS. This indicates possible opportunities for early intervention and prevention but requires additional research and development of infrastructure and guidelines to better understand and manage patients who experience SPS.