关键词: 快感缺失, 亚临床高危群体, 心理干预, 强化学习模型, 控制感

Abstract: Anhedonia, defined as a marked reduction in the capacity to experience pleasure, is a transdiagnostic feature and a core symptom across multiple psychiatric disorders, including major depressive disorder, schizophrenia, and bipolar disorder. Despite its clinical significance, existing pharmacological and psychological interventions have shown variable and often modest efficacy in alleviating anhedonic symptoms, particularly during subclinical or prodromal stages. One possible contributing factor is that anhedonia has predominantly been approached as a relatively stable deficit in hedonic capacity, whereas accumulating evidence suggests that it may reflect dynamic alterations across multiple components of reward processing, with reward learning playing a particularly central role. Clarifying how reward learning deficits emerge and how they may be modulated prior to the onset of diagnosable psychiatric disorders is therefore critical for early prevention and intervention efforts. Recent theoretical and empirical advances indicate that reward learning serves as a crucial bridge between reward anticipation and consummatory experience, and that effective reward learning depends not merely on objective reward contingencies but also on individuals’ subjective perceptions of control over outcomes. Learning signals may be substantially attenuated when perceived control is low, even under identical objective contingencies. These observations point to perceived control as a potentially important factor shaping reward learning processes in anhedonia. However, the dynamic characteristics of perceived control’s modulation of reward learning in anhedonia, the underlying neurocomputational mechanisms, and the feasibility of targeting perceived control as an intervention have not yet been systematically characterized.
Grounded in meta-reinforcement learning theory and locus of control theory, the present study adopts a multimodal and multi-level research framework to investigate the dynamic modulation of reward learning by perceived control in individuals with anhedonia, with a particular focus on subclinical high-risk populations at elevated risk for mental disorders. The overarching aims are to characterize how reward learning deficits vary as a function of perceived control, to elucidate the mechanisms linking perceived control to alterations in reward learning, and to develop and empirically evaluate novel intervention paradigms targeting perceived control. Central to this research is the proposed Dual-Layer Reward Learning Hypothesis with Internal-External Loops. This hypothesis suggests that subjective perceptions of reward-behavior contingency strength, beyond objective contingencies alone, may dynamically influence immediate reward learning processes (Internal Loops) while also contributing to the updating of longer-term beliefs about reward-behavior associations (External Loops). Through iterative interactions between these two layers, perceived control may help attenuate maladaptive reward learning biases and, in turn, be associated with reductions in anhedonic experiences.
This study comprises three interrelated components. First, Study 1 aims to examine reward learning deficits as a function of perceived control through two experiments: 1) in healthy individuals, manipulating prior perceived control via a sensorimotor synchrony paradigm to assess how the external loop affects the reward learning process; 2) comparing computational parameters and adapted models across different groups of individuals, with a particular focus on the reward learning characteristics of subclinical populations, to identify shared/unique mechanisms and to clarify the modulatory role of perceived control. Second, Study 2 investigates the neurocomputational mechanisms underlying control-related modulation of reward learning in anhedonia. Using functional magnetic resonance imaging (fMRI) and probabilistic reward learning tasks, the study examines how perceived control affects neural representations, while also assessing modulation within and between core reward-related regions. Third, Study 3 aims to develop and evaluate intervention paradigms targeting perceived control through 1) laboratory-based task enhancing action-outcome contingencies and 2) psychological intervention based on positive autobiographical recall, and to assess the immediate and sustained effects on reward learning and anhedonia.
By integrating experimental manipulation, computational modeling, neuroimaging, and randomized controlled interventions, this study aims to advance a mechanistic understanding of anhedonia and evaluate the potential of perceived control as a theoretically informed intervention target. The findings are expected to contribute to early prevention strategies for mental disorders and to support a shift toward intervention approaches that place greater emphasis on enhancing positive psychological functioning, alongside traditional symptom-focused frameworks.

Key words: anhedonia, subclinical high-risk population, psychological intervention, reinforcement learning model, perceived control