关键词: 抑郁倾向, 奖赏正波, 晚期正电位, 受试者工作特征曲线

Abstract: Depression is one of the most prevalent mental diseases in the world, which is characterized by continuous downcast mood, slow thinking, and decreased energy. In recent years, predicting the risk of depression through physiological technology has become a hotspot in the field of neuroscience. In the current study, two event-related potential components, the reward positive (RewP) and late positive potential (LPP), were selected as predictors of depressive tendency. The amplitude differences of individuals with or without depressive tendencies were compared by artificial dichotomy, and the predictive effect of ERP components on individual depressive tendency after one year was investigated. Finally, the prediction effectiveness of the two indicators and their conjoint prediction effectiveness are compared through the analysis of the receiver operating characteristic (ROC) curve.
Study 1 referred to the classical "simple gambling task" paradigm and compared amplitude differences under the conditions of gain and loss between groups. The reward positive was indexed as the amplitudes of gain minus amplitudes of loss. Then, amplitudes of the reward positive was used as an indicator to predict the individual's depressive tendency after one year by logistic regression. The results indicated that individuals with depressive tendency had a reduced neural response to reward feedback. The amplitude of the reward positive were significantly negatively correlated with the score of depressive symptoms. The positive reward at baseline can negatively predict one's depressive tendency one year later.
In Study 2, the amplitude of late positive potential induced by positive affective pictures, neutral affective pictures and negative affective pictures was recorded and compared between groups. The amplitudes of late positive potential induced by these three valences of pictures were used as indicators to predict the individual's depressive tendency and depressive symptom score one year later by logistic regression. The results showed that the amplitude of late positive potential induced by positive affective pictures in individuals with depressive tendencies was significantly lower than that in individuals without depressive tendencies. The amplitude of late positive potential evoked by positive affective pictures was significantly negatively correlated with depression symptom scores. The late positive potential induced by positive affective pictures can significantly and negatively predict one's depressive tendency one year later. The late positive potential may be more suitable for the horizontal diagnosis of depressive tendency than the longitudinal prediction.
Study 3 examined the efficacy of the reward positive, late positive potential induced by positive affective pictures and the Beck depression inventory-II as independent indicators of depressive tendency. The analysis of receiver operating characteristic demonstrated that these three indicators can effectively predict individual depression tendency one year later, and there was no significant difference in the prediction effectiveness among them. When the reward positive and late positive potential induced by positive affective pictures were used as the conjoint indicators of depressive tendency, its Youden-index can exceed any single indicators.
The results of this study support that there are defects in the reward circuit of depressed individuals. This study verified the effectiveness and specificity of the reward positive and late positive potential as neural indicators of individual depression tendency. The current research provides a relatively novel perspective and reference for the diagnosis and prevention of depression.

Key words: depressive tendency, reward positive, late positive potential, receiver operating characteristic curve