Living environmental quality and incident depressive symptoms among middle-aged and older adults in China: a national longitudinal study

Abstract

Depressive symptoms are common in mid-to-late adulthood in China. Evidence linking living environments to depression has often focused on single exposures, and prospective evidence on the association between an integrated measure of living environmental quality (LEQ) and subsequent depressive symptoms remains limited. Using the China Health and Retirement Longitudinal Study (CHARLS), we followed adults aged ≥ 45 years without depressive symptoms in 2013 through 2020. Depressive symptoms were defined as CES-D-10 ≥ 10. LEQ was a composite score covering housing structure, indoor thermal comfort, drinking water source, cooking/heating fuels, and ambient PM2.5 exposure (higher scores indicate poorer conditions). Associations between LEQ and incident depressive symptoms were examined using logistic regression with sequential adjustment for demographic, socioeconomic, lifestyle, and clinical factors. Restricted cubic splines assessed dose–response shape, and subgroup analyses tested effect modification. Among 2,194 participants, 698 (31.8%) met the CES-D-10 threshold for depressive symptoms in 2020. Poorer baseline LEQ was associated with higher odds of depressive symptoms identified in 2020 after full adjustment, including education and household wage and bonus income (OR per 1 point increase = 1.096; 95% CI 1.010 to 1.190). Compared with the lowest LEQ tertile, the highest tertile had higher odds of incident depressive symptoms (OR = 1.327; 95% CI 1.033 to 1.706). The spline curves showed an approximately increasing pattern, and the statistical tests did not detect evidence of nonlinearity. Associations were broadly consistent across subgroups, with no significant interactions. Poorer baseline multidomain living environmental quality was associated with higher odds of depressive symptoms identified in 2020 among Chinese adults aged ≥ 45 years who were free of depressive symptoms at baseline.