A transdiagnostic evaluation of adult temperamental profiles and their interactions with childhood adversities across major psychiatric disorders

Abstract Background: Temperamental profiles and adverse childhood experiences (ACEs) have been reported in most major psychiatric disorders. However, the nature of this relationship is complex and under-studied. Methods: A total of 1162 individuals (N=532 affected; N=457 unaffected first-degree relatives (FDRs) and N=173 healthy controls) were included. Affected individuals and their FDRs were from 393 multiplex families of schizophrenia, bipolar disorder, obsessive-compulsive disorder, or alcohol use disorder. They were assessed using the self-reported Adult Temperament Questionnaire (ATQ), and the ACEs International Questionnaire. We performed frequentist and Bayesian mixed-effects regression models, to look for associations between the ATQ domains, ACEs score and their interaction with each diagnosis/familial risk. External validity of our final model was tested to predict diagnosis in an independent cohort. Receiver operator characteristic plots and the area under the curve (AUC) were used to evaluate its prediction accuracy. Results: All temperamental domains: negative affectivity (NA), extraversion (EV), effortful control (EC) and orienting sensitivity (OS) had significant associations in a unique configuration within each of the four diagnoses and their familial risk. ACEs, and its interactions were significantly associated with alcohol use disorder only. In the external validation dataset, the final model was found to successfully predict AUD (AUC=0.82), BD (AUC=0.75) and OCD (AUC=0.74), but not SCZ (AUC=0.53). Conclusions: These findings provide unique insights into the complexities of temperament in major psychiatric disorders, and emphasize the importance of evaluating temperament in conjunction with childhood adversity to characterize risk.