Variability of menstrual cycles by age, polycystic ovary syndrome, and early-life cycle irregularity in the apple Women's Health Study.

backgroundPolycystic ovary syndrome is a common endocrine disorder, characterized by oligomenorrhea and androgen excess. Only a few studies have addressed the natural history of menstrual cycles among women with polycystic ovary syndrome and/or irregular cycles, most with limited sample size and homogenous populations.

objectiveWe describe age-related differences in menstrual cycle length and regularity for those with and without a diagnosis of polycystic ovary syndrome. STUDY

designWe included participants in the Apple Women's Health Study, a digital US cohort, who consented and enrolled between November 2019 and March 2024 and provided menstrual logging data (on days of menstrual bleeding) for at least 3 cycles without hormone use, pregnancy, or lactation. We evaluated menstrual cycle length among 3 mutually exclusive groups: those reporting a diagnosis of polycystic ovary syndrome (group 1: polycystic ovary syndrome), those without polycystic ovary syndrome but with early-life irregular (unpredictable) cycles (group 2: early-life irregular), and those without polycystic ovary syndrome and with early-life regular (predictable) cycles (group 3: early-life regular). Polycystic ovary syndrome status was based on self-reported physician diagnosis. "Early-life irregular" was defined as reporting not spontaneously establishing regular cycles within 4 years after menarche for those who did not report a polycystic ovary syndrome diagnosis. Menstrual cycles were categorized by participant's age into 8 groups (<20, 20-24, 25-29, 30-34, 35-39, 40-44, 45-49, and 50+ years). We used linear mixed effect models with random participant-specific intercepts to estimate a mean menstrual cycle length by age groups. Log-linear models for residual variance were used in the linear mixed effect models to estimate how within-individual standard deviations of menstrual cycle length (as a measure of cycle irregularity) vary across age groups. P values for interactions were calculated to compare across groups.

resultsData from 160,206 menstrual cycles across 15,586 participants were analyzed, including 18,875 cycles from 1842 participants with polycystic ovary syndrome. Mean age at enrollment was 33 years for all groups. Prevalence of obesity (body mass index ≥30.0) was 62% in the polycystic ovary syndrome group, 36% in the early-life irregular group, and 35% in those with early-life regular cycles. Around 22% of the cycles in the polycystic ovary syndrome group were 30 to 34 years of age and 24% were ages 35 to 39. Menstrual cycle length patterns across the reproductive lifespan differed among groups. Compared to group 3 with early-life regular cycles, group 1 (polycystic ovary syndrome) and group 2 (early-life irregular) consistently had longer mean menstrual cycle length in early reproductive years, with the difference between groups decreasing with age up to ≤40 years. Covariate-adjusted results showed similar trends. Decreases in cycle irregularity (within-individual standard deviations) for groups 1 and 2 were also seen across the 20s and 30s age groups.

conclusionAt younger ages, persons with polycystic ovary syndrome or early-life irregularity had longer and more irregular cycles than those with regular cycles, as expected. These differences among groups diminished with age, as cycle characteristics of those with polycystic ovary syndrome or early-life irregularity became more similar to those of persons with early regular cycles.