Determinants of age at natural menopause in the Isparta Menopause and Health Study

Abstract

The influence of body weight and weight change on age at natural menopause (ANM) is controversial. We aimed to explore these associations by overcoming two possible reasons for inconsistent findings: (1) a lack of information on weight change from a fixed time point up until menopause and (2) a possibility of capturing a spurious association between later ANM and greater weight gain because of a longer time span. Cross-sectionally, we evaluated static (body mass index [BMI] at age 25 y and height) and dynamic (history of premenopausal episodic weight loss and BMI gain rate) anthropometric indicators as predictors of ANM in Cox proportional hazards models. BMI gain rate was calculated based on weight change from age 25 y to menopause (or to current age in premenopausal women). Of 1,106 women, 46% were postmenopausal. Dynamic anthropometric indicators were stronger determinants of ANM compared with static indicators. Being in the higher quartile of BMI gain rate and having a history of episodic weight loss of more than 5 kg were independently associated with later ANM (hazard ratio [95% CI], 0.832 [0.765-0.905] and 0.433 [0.344-0.546], respectively). After controlling for history of episodic weight loss of more than 5 kg, smoking and premenopausal type 2 diabetes emerged as time-dependent predictors of ANM (hazard ratio [95% CI], 1.371 [1.093-1.720] and 1.640 [1.141-2.356], respectively). Premenopausal BMI gain rate and premenopausal history of episodic weight loss of more than 5 kg are independently associated with later ANM. Smoking predicts an earlier menopause after the age of initiation and type 2 diabetes predicts an earlier menopause after the age of diagnosis. The latter two associations are significant only after adjusting for a history of episodic weight loss of more than 5 kg.