Modeling Menstrual Cycle Length at the Approach of Menopause Using Bayesian Changepoint Models

• Activity Number: 257
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Xiaobi Huang and Michael R. Elliott*+
• Companies: University of Michigan and University of Michigan
• Keywords: Variance models ; Multiple imputation ; Women's health ; Final menstruation period ; TREMIN

Abstract

Menopausal transition is a critical period in women's lives, as physiologic changes and health practices during this time frequently define women's chronic disease risk profiles. We develop a Bayesian change point model to jointly model the mean and variability of menstrual cycle length together with the cessation of menstruation. The model incorporates separate mean and variance change points for each woman and a hierarchical model to link them together, along with regression components to include predictors of menopausal onset. Data are from TREMIN, a 70-year old longitudinal study of menstrual calendar data throughout women's life courses. We use multiple imputation to deal with different forms of missingness and develop innovative methods to predict unobserved final menstruation periods. We extend our model to incorporate growth mixture models to further define patterns of change.