Investigating Sensorimotor Decision-Making in Depressed Individuals Using a Bayesian Approach

Depression is among the most prevalent and disabling mental health conditions worldwide. It is associated with low mood and motivation, cognitive inefficiencies, altered reward and punishment learning, and broader executive dysfunction that can disrupt daily functioning. Depression has also been linked to differences in sensory perception, sensitivity, and processing speed. However, how depression affects sensorimotor decision-making under uncertainty remains comparatively understudied. In this study, we examined how individuals with depression process and weigh prior information relative to current sensory evidence during sensorimotor decisions. Participants completed a visuomotor coin-catching task, and behavior was analyzed within a Bayesian Decision Theory framework. We tested whether reliance on current information differed between depressed and non-depressed individuals, and whether any differences depended on uncertainty in the prior or the sensory signal. Across behavioral and computational analyses, both groups combined prior knowledge and sensory input in a qualitatively Bayesian manner, placing greater weight on current information when it was less uncertain and/or when prior information was more uncertain. The groups did not differ significantly in overall relative reliance on current information. Despite these broad similarities, more subtle differences were apparent. Model-based analyses indicated that Bayesian models outperformed heuristic alternatives in both groups, but model fit was poorer in the depressed group. Both groups adjusted sensory weighting as prior uncertainty changed, but non-depressed participants showed about twice the shift observed in depressed individuals, although this difference was not statistically significant. Together, these findings suggest that depression does not eliminate Bayesian-like integration of priors and sensory evidence, but may potentially reduce sensitivity to changes in prior reliability. Future work should test whether individual differences in symptom profiles, task demands, and cognitive flexibility moderate these effects.