BackgroundThis study aimed to evaluate whether social, familial and personal factors can predict incident and prevalent depressive symptoms in Chinese adults aged ≥ 45 years using data from the China Health and Retirement Longitudinal Study (CHARLS).MethodsStudy subjects without depressive symptoms from CHARLS at baseline were enrolled. Depressive symptoms were defined by the 10-item Center for Epidemiologic Studies Depression Scale. Statistical adjustment, subgroup exploration and unmeasured confounding assessment were undertaken to derive reliable estimates.Results1681 (27.04%) of 6215 subjects who had no depressive symptoms in 2011, suffered one or more depressive symptoms in 2018. Multivariate analyses showed that number of grandchildren (odds ratio [95% confidence interval]: 1.06 [1.02, 1.10]), social activity score (0.95 [0.91, 0.98]), instrumental activities of daily living (IADL) (1.35 [1.11, 1.65]) and number of comorbidities (1.16 [1.10, 1.22]) were independently and significantly associated with the presence of incident depressive symptoms. Further categorization revealed significance for social activity score (odds ratio [95% confidence interval]: 0.78 [0.69, 0.89] and 0.71 [0.53, 0.95] for 1–5 and > 5 vs. 0), IADL (1.35 [1.11, 1.65] for yes vs. no) and number of comorbidities (1.38 [1.20, 1.58], 1.44 [1.16, 1.81] and 2.42 [1.54, 3.80] for 1–2, 3–4 and > 4 vs. 0) associated with incident depressive symptoms. Restricting analysis to wave IV data in 2018 observed significant association of number of grandchildren, social activity score, IADL and number of comorbidities with prevalent depressive symptoms.ConclusionsThe present study findings support the marked contribution of social activity score, IADL and number of comorbidities to incident and prevalent depressive symptoms in Chinese middle-aged and older adults.