Effects of acute stress on reward processing: A comprehensive meta-analysis of rodent and human studies

Abstract

Stressors can initiate a cascade of central and peripheral changes that modulate mesocorticolimbic dopaminergic circuits and, ultimately, behavioral response to rewards. Driven by the absence of conclusive evidence on this topic and the Research Domain Criteria framework, random-effects meta-analyses were adopted to quantify the effects of acute stressors on reward responsiveness, valuation, and learning in rodent and human subjects.

In rodents, acute stress reduced reward responsiveness (g = −1.43) and valuation (g = −0.32), while amplifying reward learning (g = 1.17). In humans, acute stress had marginal effects on valuation (g = 0.25), without affecting responsiveness and learning. Moderation analyses suggest that acute stress neither has unitary effects on reward processing in rodents nor in humans and that the duration of the stressor and specificity of reward experience (i.e., food vs drugs) may produce qualitatively and quantitatively different behavioral endpoints.

Subgroup analyses failed to reduce heterogeneity, which, together with the presence of publication bias, pose caution on the conclusions that can be drawn and point to the need of guidelines for the conduction of future studies in the field.