Exploring the individual contributions of the 8 adverse childhood experiences (ACEs) domains to substance use.

Background: Traditionally, adverse childhood experiences (ACEs) are grouped together into a cumulative score to predict substance use.

Objective: To determine the effect of individual ACE domains (physical, emotional, sexual abuse, exposure to domestic violence, parental separation, household incarceration, household substance use and household mental illness) on later life substance use.

Participants and setting: BRFSS data from 2019 to 2022 were used to construct 3 cohorts analyzing 5 different substance use outcomes.

Methods: Logistic regression models were constructed to determine adjusted odd ratios (aORs) of substance used for each ACE domain. Random forest models served for sensitivity analyses.

Results: Effect sizes varied across ACE domains and substances. Household substance use was the strongest predictor, with the highest aORs for smoking (aOR = 1.336, 95 % CI [1.266-1.410]), e-cigarette use (aOR = 1.451, 95 % CI [1.317-1.597]), and binge drinking (aOR = 1.256, 95 % CI [1.181-1.335]). Emotional abuse was the second most influential domain, particularly associated with marijuana use (aOR = 1.475, 95 % CI [1.349-1.613]). In contrast, exposure to domestic violence showed no significant associations with any substance use outcome. Sensitivity analyses using random forest models validated these findings, emphasizing the differential contributions of individual ACE domains to substance use behaviors.

Conclusions: These findings emphasize that ACEs are not all entirely equal, and a cumulative score does not represent the individual effect each ACE has on substance use outcomes. Tailoring interventions and policies toward addressing specific ACE domains could significantly reduce substance use rates in later life.