Analysis of menstrual diary data across the reproductive life span applicability of the bipartite model approach and the importance of within-woman variance.

The nature of variability in menstrual function has not been adequately described or quantified across the reproductive life span. This article evaluates the applicability of the bipartite model approach to the analysis of menstrual data and the relative importance of within-woman variability across the reproductive life span using data from the Tremin Trust data, a large prospective study in which women maintained menstrual diaries throughout their reproductive life. We first consider how the boundaries of the Gaussian portion of the distribution change with age, and reflect upon the implications of these distribution changes for definitions of normal cycling. We next estimate the change in mean cycle length, in between- and within-woman variance and in the probability of having a nonstandard cycle across the reproductive life span. Finally, we characterize the dynamics of menstrual cycling within women over time at various points in the reproductive life span.