This study utilizes survey data from CHARLS 2011–2020, selecting menopausal women aged 40–54 at baseline (2011) and including only those who participated in at least two of the five follow-up surveys. The analysis of depressive symptoms first involves a univariate analysis using the baseline data. Subsequently, both generalized linear model and generalized linear mixed model are constructed with the longitudinal data to perform multivariate analysis, and the goodness-of-fit of these models is compared. The results show that: ① Baseline data reveals significant differences in the incidence of depressive symptoms among menopausal women based on urban-rural residence, education level, marital status, number of children, income, sleep quality, exercise habits, self-rated memory, and self-rated health. ② Longitudinal analysis indicates a U-shaped trend in the incidence of depressive symptoms among Chinese menopausal women, initially decreasing and then increasing. In addition to the variables mentioned above, birth cohort, social engagement, and menopausal status were also found to have significant effects on mental health. ③ The influencing factors on mental health varies across different birth cohorts. Based on these findings, the study recommends to strengthen attention to the mental health of menopausal women, improve their psychological well-being, and provide differentiated mental health support for different demographic groups.