Abstract
BACKGROUND: Cognitive models of post-traumatic stress disorder (PTSD) propose that trauma entails cognitive alterations of increased distrust and perceived threat from others. We tested whether these predictions also hold in individuals with varying levels of childhood maltreatment (CM), which is much more prevalent than traumatic events as required for a PTSD diagnosis. We hypothesized that higher levels of CM would entail greater distrust and perceived threat, and that distrust would be more change-resistant in participants with more CM. METHODS: The study was pre-registered; the pre-registration protocol, data, and code are available at https://osf.io/pufy2/ . We recruited 549 participants (M age = 29.2, 74.5% women) for an online study via websites related to CM, Borderline Personality Disorder, and via snowball method on social media. Participants self-reported their level of CM on the childhood trauma questionnaire (CTQ). Next, they played two rounds of a hypothetical distrust game, indicating the perceived trustworthiness of avatars by way of estimating expected monetary deductions from them (i.e. higher amounts indicating greater distrust). After the first round, we provided participants with the feedback that very little money was taken from them. We expected those with more CM to be less responsive to the positive feedback and to adapt their estimates less in the subsequent round. Following the distrust game, participants completed an emotion rating task in which they rated the emotional expressions of 60 faces on a scale from 'very negative' to 'very positive'. We included angry, fearful, and happy facial expressions, and expected individuals with higher CM levels to provide more negative ratings. We conducted linear mixed effects models with random intercepts for raters and stimuli (crossed), and modelled random slopes for all within-person predictors. RESULTS: As hypothesized, higher levels of CM were associated with higher levels of distrust and a weaker decrease in distrust following positive feedback. Further supporting our hypotheses, individuals with higher levels of CM showed more negatively shifted emotion ratings. CONCLUSIONS: Increased distrust and perceived interpersonal threat following trauma, as proposed in cognitive models of PTSD, likely also apply to individuals with CM, following a dose-response relationship. We discuss clinical implications of considering any level of CM as a potentially relevant treatment-factor, even when a trauma-related disorder is not the main diagnosis, and propose future research avenues.