The association between earlier spousal retirement and depressive symptoms in China: a longitudinal study

Introduction Retirement is a major life transition that may influence mental health. In urban China, gender-specific statutory retirement ages often result in spouses retiring at different times. This study examined whether earlier spousal retirement was associated with depressive symptoms among middle-aged and older adults and explored the potential role of household income. Methods We used five waves of data from the China Family Panel Studies (2012–2022), comprising 5,652 person-wave observations from 3,059 individuals. Pooled ordinary least squares regression models with province and survey-wave fixed effects were used to examine the association between earlier spousal retirement and depressive symptoms (CES-D), with gender differences tested using interaction terms and stratified models. Robustness checks included an alternative outcome measure, inverse probability weighting, a correlated random effects model, alternative samples, and propensity score matching. The mediating role of household income was assessed using bootstrap mediation analysis. Results Earlier spousal retirement was associated with fewer depressive symptoms ( β = −0.323, p < 0.05), and this association remained consistent across robustness checks. The association was observed among husbands whose wives retired earlier ( β = −0.739, p < 0.01) but not among wives whose husbands retired earlier ( β = 0.111, p > 0.05), and the interaction between earlier spousal retirement and gender was significant in a fully interacted model ( β = 0.850, p < 0.05). Earlier spousal retirement was associated with higher household per capita income, which in turn was associated with fewer depressive symptoms. The indirect association through household income was significant in the full sample (indirect effect = −0.051, 95% CI : −0.083 to −0.022), accounting for approximately 15% of the total association. It was also significant among husbands (indirect effect = −0.086, 95% CI : −0.154 to −0.022) but not among wives. Conclusion Earlier spousal retirement was associated with fewer depressive symptoms, particularly among husbands. Household income may be one pathway underlying this association. These findings highlight the importance of considering retirement as a couple-level transition and suggest that the relative timing of spouses’ retirement may be relevant to mental health in later life.