WeChat use and Moment posting on depressive symptoms in middle-aged and older adults: A national cohort study using quantile regression

Background Prior studies have suggested an association between WeChat use and depressive symptoms among middle-aged and older Chinese adults. However, most have relied on traditional linear or logistic regression models that estimate average effects, potentially overlooking heterogeneous associations across different levels of depressive symptom severity. Objective This study examined the differential associations of WeChat use and Moment posting with depressive symptoms across symptom severity levels in middle-aged and older Chinese adults. Methods A prospective national cohort study was designed using data from Wave 4 and Wave 5 of the China Health and Retirement Longitudinal Study (CHARLS). Participants aged 45–83 years with complete 10-item Center for Epidemiologic Studies Depression (CESD-10) Scale data were included. Quantile regression models were used to assess associations between WeChat/Moment use and depressive symptoms at the 25th, 50th, and 75th percentiles of CESD-10 scores. Subgroup and sensitivity analyses were conducted. Results Among 9,573 participants (WeChat usage rate: 15.75%; median CESD-10 score: 6.00 [IQR: 3.00–9.00]), WeChat use and Moment posting were inversely associated with depressive symptoms at the 25th, 50th, and 75th percentiles in the overall cohort and among middle-aged adults (P < 0.001). Among older adults, WeChat use was associated with lower CESD-10 scores at the 25th and 75th percentiles, while Moment posting showed a significant association only at the 25th percentile. Moment posting demonstrated a stronger inverse association with depressive symptoms than WeChat use alone (P < 0.05). Conclusions WeChat use was associated with lower depressive symptoms in middle-aged and older adults, with Moment posting particularly linked to such reductions among middle-aged individuals. The findings highlight a potential inverse association that warrants further investigation. However, causal inference is limited by the observational design.