Altered reward and effort processing in children with maltreatment experience: a potential indicator of mental health vulnerability

Abstract

In this longitudinal study of children and adolescents with a documented history of maltreatment, we investigated the impact of maltreatment on behavioral and neural indices of effort-based decision making for reward and examined their associations with future internalizing symptoms. Thirty-seven children with a documented history of maltreatment (MT group) and a carefully matched group of 33 non-maltreated children (NMT group) aged 10-16, completed an effort-based decision-making task during functional magnetic resonance imaging (fMRI). Internalizing symptoms were assessed at baseline and again 18 months later. Computational models were implemented to extract individual estimates of reward and effort sensitivity, and neural signals during decision-making about different levels of reward and effort were analyzed. These were used to predict internalizing symptoms at follow-up. We identified lower effort-related activation in the anterior cingulate cortex (ACC), a prespecified region-of-interest, in the MT relative to the NMT group. No group differences were observed in the striatum, or in behavioral indices of reward and effort processing. Lower effort-related ACC activation significantly predicted elevated internalizing symptoms at follow-up in the MT group. These findings suggest that disrupted effort-related activation may index latent vulnerability to mental illness in children who have experienced maltreatment.

Fig. 1. Effort-based decision making for reward task and behavioral descriptives.

A On each trial, participants were presented with an offer consisting of a given reward (4, 8, or 12 apples) for a given effort (20%, 50%, or 80% of their maximum voluntary contraction) presented as apples on a tree with a yellow bar on the trunk indicating the effort level. They indicated their decision by applying a mild force to either the left- or right-hand dynamometer. If they decided to accept the offer, the tree reappeared on either the left or the right side of the screen indicating which hand dynamometer the participant had to use to apply the required effort level, which they had to sustain for at least 2 s over a 4 s period. Depending on whether the participant was successful or not they received feedback; points were never deducted. B Mean acceptance rates by reward and effort for MT and NMT group.

Fig. 2. ACC ROI location and activation levels.

A ACC ROI showing decreased activation with increasing effort levels in the MT group compared to the NMT group (pre-specified ACC ROI mask applied to two-sample t-test map in SPM). B Group difference in average ACC ROI activation with increasing effort levels during the decision/choice phase (error bars +/− 1 SE). *p < 0.05.

Fig. 3. Partial regression plot.

Prediction of internalizing symptoms from baseline ACC effort-related activation (with increasing effort levels) in the MT group, controlling for baseline emotional symptoms, sex, age, pubertal status, and MT severity.