Longitudinal study of hormone levels and depression among women transitioning through menopause

Objective To address the association between depression and absolute levels of estradiol, and change in estradiol, among women during the menopausal transition. Methods This was a prospective, observational study conducted in Massachusetts, USA. Analyses reported here are based on the first three interviews following baseline (T1–T3). The participants were 309 women aged initially 43–53 years, who contributed a total of 728 observations (mean number of observations per subject 2.36). The main outcome measure was depression as measured by the Center for Epidemiologic Studies-Depression (CES-D) scale. Results CES-D score was not significantly associated with menopause status categories, nor was it associated with annual change in estradiol level (E2) (p = 0.19). The unadjusted association between log E2 and CES-D was negative (odds ratio 0.69) and statistically significant (p = 0.03). Upon adjustment for symptoms, however, the association remained negative but was no longer statistically significant (p = 0.26). Hot flushes/night sweats were positively associated with CES-D (p = 0.04), and trouble sleeping was strongly positively related to CES-D (p < 0.001). Conclusions Results provide strong support for the domino or symptom hypothesis, which posits that depressed mood is caused by vasomotor symptoms associated with changing estrogen levels. Estradiol did not have a direct effect, independent of symptoms. This study adds to the body of literature suggesting that any association found between menopause and depression is most likely to be explained by other factors, such as symptoms and sleep problems. Findings also highlight the importance of studying the complex relationship between hormone levels, sleep problems and vasomotor symptoms during the menopausal transition.