Prevalence, Comorbidity, and Sociodemographic Correlates of Psychiatric Disorders Reported in the All of Us Research Program.

IMPORTANCE

All of Us is a landmark initiative for population-scale research into a variety of health conditions, including psychiatric disorders.

OBJECTIVE

To analyze the prevalence, comorbidity, and sociodemographic covariates of psychiatric disorders in the All of Us biobank.

DESIGN, SETTING, AND PARTICIPANTS: We estimated prevalence, overlap, and sociodemographic correlates for psychiatric disorders as reported in electronic health records for All of Us release 5 (N = 331 380).

EXPOSURES

Social and demographic covariates.

MAIN OUTCOMES AND MEASURES

Psychiatric disorders derived from International Statistical Classification of Diseases, Tenth Revision, Clinical Modification, codes across 6 broad domains: mood disorders, anxiety disorders, substance use disorders, stress-related disorders, schizophrenia, and personality disorders.

RESULTS

The analytic sample (N = 329 038) was 60.7% female (mean [SD] age, 50.9 [16.8] years). The prevalence of disorders ranged from 11.00% (95% CI, 10.68% to 11.32%) for any mood disorder to less than 1% (eg, obsessive-compulsive disorder, 0.18%; 95% CI, -0.16% to 0.52%), with mood disorders being the most common and personality disorders being the least. There was substantial overlap among disorders, with the majority of participants with a disorder (30 113/58 806, approximately 51%) having 2 or more registered diagnoses and tetrachoric correlations ranging from 0.43 to 0.74. Comparisons of prevalence across demographic categories revealed that non-Hispanic White people, individuals with low socioeconomic status, women and individuals assigned female at birth, and sexual minority individuals are at greatest risk for most disorders.

CONCLUSIONS AND RELEVANCE

Although rates of disorders among the All of Us cohort are lower than in the general population, considerable variation, comorbidity, and disparities exist across social groups. To improve the practice of equitable precision medicine, researchers can use comprehensive health data from large-scale resources such as All of Us.