Estimation of the distribution of age at natural menopause from prevalence data

Abstract

Nearly 30% of US women reach menopause (defined as cessation of menstrual periods) as a consequence of an operation. This biases the observable distribution of age at natural menopause. Another problem with estimating this distribution from a cross-sectional study is the clustering of reported age at natural menopause around ages ending in zero and five (Mac-Mahon B, Worcester J. Age at menopause, United States 1960-1962. Washington DC: National Center for Health Statistics, 1966. Vital and health statistics, Series 11: Data from the National Health Survey, no. 19. (DHEW publication no. (HSM) 66-1000)). This paper discusses the approach of Mac-Mahon and Worcester to this problem and compares it with a competing risks approach.