时间：2026年10月10日（周六）中午 12:00
地点：清华大学吕大龙楼11层1100
报告主题：Longitudinal Pathways Linking Personality Traits and Depressive Symptoms in Older Adults: A Cross-Lagged Panel Network Analysis Across Two HRS Cohorts
报告人：博士研究生王子茂
内 容：
The longitudinal relationship between personality and depression is often framed by two competing hypotheses: the vulnerability model (traits predispose to symptoms) and the scar model (symptoms alter traits). However, evidence remains inconclusive, potentially because prior research has typically examined depression as a unitary construct, overlooking whether these models operate at the level of specific symptoms.
We applied a longitudinal network approach to data from two independent cohorts of older adults in the Health and Retirement Study (HRS). Cross-lagged panel network models were estimated separately for each cohort to examine directional associations between the Big Five personality traits and eight depressive symptoms across three waves spanning eight years.
Vulnerability-consistent pathways were more extensive and broadly consistent across cohorts and intervals. Neuroticism emerged as a broad vulnerability trait, predicting subsequent increases across multiple symptoms. Lower Extraversion predicted diminished positive affect and difficulty getting going; lower Conscientiousness predicted the latter. In contrast, scar-consistent pathways were more selective, with restless sleep predicting higher Neuroticism and feeling unmotivated predicting lower Conscientiousness.
By examining these models longitudinally at the symptom level, this study moves beyond the simple “both-occur” conclusion for the personality–depression relationship. Our findings suggest that, among older adults, this relationship is characterized by more extensive vulnerability-consistent pathways, accompanied by more selective scar-consistent pathways. These findings provide a more differentiated understanding of prospective personality–depression associations and highlight candidate pathways for future prevention and intervention research.
