Characterization and Validation of Adverse Childhood Experiences Data in the All of Us Research Program
TL;DR
The findings establish the All of Us ACE dataset as a reliable resource for epidemiologic, clinical, genomic, and precision medicine research on childhood adversity.
Abstract
Adverse childhood experiences (ACEs) are major determinants of lifelong health, yet few large precision medicine cohorts integrate standardized ACE measures with longitudinal clinical, genomic, and participant-reported data. In this cross-sectional study, we characterized the newly released 11-item ACE questionnaire using data from 137,946 All of Us Emotional Health History and Well-Being survey respondents. 90,540 completed all 11 items and 91,871 could be classified across all eight Centers for Disease Control and Prevention (CDC) ACE domains. The questionnaire demonstrated good internal consistency (Kuder-Richardson Formula 20 = 0.79), and the derived eight-domain score showed good reliability (Kuder-Richardson Formula 20 = 0.75). Compared with participants eligible to complete the survey, respondents were disproportionately White and non-Hispanic, whereas Black or African American and Hispanic participants were underrepresented. Increasing ACE burden was independently associated with higher odds of clinical and social determinant outcomes, with the strongest associations observed for post-traumatic stress disorder, food insecurity, bipolar disorder, suicidal ideation and self-harm, and substance use disorder. Outcome prevalence generally increased with ACE burden, supporting dose-response relationships. These findings establish the All of Us ACE dataset as a reliable resource for epidemiologic, clinical, genomic, and precision medicine research on childhood adversity.