Estimating the median age at menarche with a Logit model : application to African DHS surveys

Abstract

Background : The study investigates the median age at menarche in African countries. Data Source and Methods : DHS surveys provided information on the proportion ever-menstruating among women age 15-19. A linear-logistic regression model was used to derive the median age at menarche (statu quo method). Some 139 DHS surveys were analysed, for a total of 304,826 women. Results : Results show a wide range of variations of the median age at menarche among the 35 countries investigated, from 13.1 to 15.2 years. Conclusion : Country variations were correlated with various variables indicating the level of development (Income per capita; Caloric intake; Urbanization; Child mortality; Fertility; Age at marriage). Two variables were particularly important: food intake and urbanization.