Dynamic emotional updating as a computational marker of well-being

Abstract

Mood shapes behavior, physiology, and well-being. While mood continuously fluctuates in response to external events, it can also remain surprisingly stable, raising a fundamental question: what shapes emotional well-being, mood stability or variability? Quantifying mood along these two dimensions is challenging. We combined a closed-loop mood modulating task with computational modeling to derive quantitative markers of emotional stability and variability and map them onto the continuum from well-being to depressive symptoms. Participants (n=209) experienced adaptive reward-based mood modulation and modeling revealed that higher well-being associates with mood variability and stronger weighting of recent events, whereas depressive symptoms were associated with greater mood stability and stronger anchoring to earlier events. Results replicated in an independent test-retest dataset, which also showed that these mood updating parameters are reliable across days. Our findings identify adaptive responsiveness to recent experiences as a quantitative marker of well-being and unravel the temporal mood dynamics in health and depression.