Predicting the timeline to the final menstrual period: The study of women's health across the nation

Abstract

Context: Predicting the final menstrual period (FMP) would help women know when their menopause transition will becompleted. Additionally, biological changes, suchasaccelerated bone loss, precede the FMP by at least 1 year. Objective: Our objective was to assess whether FSH, estradiol, or urinary N-telopeptide predict where an individual is on her timeline to FMP. Methods: The sample was 554 women from the Study of Women's Health Across the Nation. We modeled the probability of having crossed specified landmarks: 2 years before, 1 year before, and the FMP. We also modeled the probability of being in narrower intervals: 2 to 1 year before FMP, 2 years before FMP and FMP, or 1 year before FMP and FMP. Wedetermined the candidate markers that best predicted having crossed each landmark, with the optimum defined as the greatest area under the receiver-operator curve; created formulas for the probability of having crossed each landmark; and calculated sensitivity and specificity. Results: Final models included current estradiol and FSH (each as a fraction of 1 previous reference measure), age, menopause transition stage, race/ethnicity, and whether serum was collected during the early follicular phase. Areas under the receiver-operator curves of final models that predicted the probability of a woman having crossed 2 years before, 1 year before, and the FMP itself were 0.902, 0.926, and 0.945, respectively. If we classified women as having crossed the 2 years before the FMP landmark when predicted probability exceeded 0.3, sensitivity was 85% and specificity 77%. Conclusion: This model could help patients and researchers estimate the time to FMP. Copyright © 2013 by The Endocrine Society.