Menopausal transition: predicting time to menopause for women 44 years or older from simple questions on menstrual variability.

Objective: To assess whether menstrual variability predicts time to menopause. Design: Analyses drew on 326 menstruating women, aged 44 to 56, who were followed until they reached menopause or the study ended. The women provided data on their menstrual characteristics at intake. We evaluated the utility of six definitions of menstrual variability for predicting time to ascertained menopause (final menstrual period + 12 months): (1) more than 90 days since the most recent menstrual period (n = 20); (2) 60 or more days of amenorrhea during the previous year (n = 71); (3) cycle lengths that varied by 19 or more days (n = 106); (4) cycle lengths too variable to report a usual length (n = 29); (5) cycles less regular than they had been at age 40 (n = 107); and (6) change in the duration or heaviness of menstrual flow compared with age 40 (n = 255). In addition, we evaluated hot flashes or night sweats during the previous week (n = 50) and age 50 or more years (n = 60) as predictors. Results: Definitions 1 to 5 predicted time to menopause; definition 6 did not. Definition 1 had the highest positive predictive value for ascertained menopause within 2 years and within 4 years; definitions 2 and 4 had low to moderate positive predictive values for ascertained menopause within 2 years but good positive predictive values for ascertained menopause within 4 years. For ascertained menopause within 2 years, definition 2 showed the best balance of sensitivity (94%) and specificity (91%). Conclusion: Simple questions about menstrual variability elicit information that is informative about proximity to menopause.