Exploring the link between childhood maltreatment and substance use disorder using a prospective, genetically informative and experimental approach
Background: We study risk for substance use disorders (SUD) in a prospective cohort with objectively recorded childhood maltreatment (CM) using designs that allow controlling for familial confounding. We also investigate the impact of CM on emotional regulation.
Methods: We used medical records and national registers in 525 young adults (20-37 years) with prospectively and objectively documented severe CM, 1,979 clinical controls (unexposed former child and adolescent psychiatry patients), 1,388 matched healthy controls; and their parents and siblings. Using Cox models, we explored SUD risks in CM exposed and unexposed controls, and in siblings within the same family. In a subsample, we compared childhood trauma questionnaire (CTQ) scores in those exposed to CM with controls. Using functional magnetic resonance imaging (fMRI), we compared emotional regulation, using an emotional conflict task, in CM exposed subjects with no lifetime SUD (CM+/SUD-, N=23), with CM exposed with lifetime SUD (CM+/SUD+, N=25), and healthy controls (CM-/SUD-, N=24). Results CM significantly increased SUD risk (crude HR: 6.61, 95%CI: 5.81-7.53; HR adjusted for sex, birthyear, externalizing problems, parents’ SUD and socioeconomic factors: 3.50, 95%CI 2.95, 4.16). Although attenuated, SUD risk was 3-fold increased in CM exposed compared with their unexposed siblings (adjusted HR: 3.12, 95%CI 2.21, 4.42). ROC analysis showed a fair performance of the CTQ to identify individuals with CM (AUC = 0.78, sensitivity 0.65, specificity 0.75), but poorer in participants with lifetime SUD (AUC = 0.62, sensitivity 0.60, specificity 0.36). fMRI indicated increased activity in CM+/SUD- subjects in the right ventromedial prefrontal cortex, left anterior insula, midcingulate cortex (per-voxel p = 0.002, cluster corrected at alpha = 0.05, ps < 0.01), compared to CM+/SUD+ and CM-/SUD-.
Conclusion: CM significantly increased risk for SUD also when accounting for both recall bias and familial confounding. Differences in emotional regulation indicate underlying adaptive factors in those exposed to CM resilient to developing SUD.