Social Media Reward Sensitivity and Depressive Symptoms Over Time

Social media platforms are designed to sustain user engagement by leveraging human reward systems. Given established links between reward processing and mental health, modelling social media use through a reward learning framework may clarify mechanisms underlying mental health vulnerability. In this study, we analysed novel linked questionnaire and Twitter posting data from 301 young adults in the Avon Longitudinal Study of Parents and Children (ALSPAC). A hybrid reward learning-habit model was fitted to longitudinal posting behaviour to estimate individual differences in sensitivity to social media rewards (‘likes’), including learning rate, reward learning weight, and subjective posting cost (vigour cost). Cross-sectionally, lower vigour cost (i.e., lower subjective cost of posting) and higher learning rate (i.e., learning faster from new social media rewards) were associated with greater mental health symptoms. Longitudinally, higher reward learning weight, reflecting greater reliance on reward-driven versus habitual posting policies, predicted increased depressive symptoms one year later. These findings suggest that individual differences in computational markers of sensitivity to social media rewards are associated with mental health, and may represent a cognitive mechanism linking platform engagement to vulnerability.