Menstrual cycle length variation by demographic characteristics from the Apple Women's Health Study

Abstract

Menstrual characteristics are important signs of women's health. Here we examine the variation of menstrual cycle length by age, ethnicity, and body weight using 165,668 cycles from 12,608 participants in the US using mobile menstrual tracking apps. After adjusting for all covariates, mean menstrual cycle length is shorter with older age across all age groups until age 50 and then became longer for those age 50 and older. Menstrual cycles are on average 1.6 (95%CI: 1.2, 2.0) days longer for Asian and 0.7 (95%CI: 0.4, 1.0) days longer for Hispanic participants compared to white non-Hispanic participants. Participants with BMI ≥ 40 kg/m² have 1.5 (95%CI: 1.2, 1.8) days longer cycles compared to those with BMI between 18.5 and 25 kg/m². Cycle variability is the lowest among participants aged 35-39 but are considerably higher by 46% (95%CI: 43%, 48%) and 45% (95%CI: 41%, 49%) among those aged under 20 and between 45-49. Cycle variability increase by 200% (95%CI: 191%, 210%) among those aged above 50 compared to those in the 35-39 age group. Compared to white participants, those who are Asian and Hispanic have larger cycle variability. Participants with obesity also have higher cycle variability. Here we confirm previous observations of changes in menstrual cycle pattern with age across reproductive life span and report new evidence on the differences of menstrual variation by ethnicity and obesity status. Future studies should explore the underlying determinants of the variation in menstrual characteristics.

Fig. 1. Differences and 95%CIs of menstrual cycle length on the 25th and 75th percentiles by age, ethnicity, and BMI.

CIs, confidence intervals; BMI, body mass index, P25: 25th percentile; P75: 75th percentile. Exclusively adjusting for age, ethnicity, and BMI, and additionally for smoking, alcohol drinking, parity, physical activity, education, perceived stress scores, and MacArthur scale of subjective social status. Missing values in age, ethnicity, and BMI were excluded. Missing values in other covariates were treated with missing indicator. The other ethnicity group includes American Indian or Alaska Native, Middle Eastern or North African, Native Hawaiian or other Pacific Islander, or other unspecified ethnicity. Underweight: BMI < 18.5 kg/m² healthy: 18.5 ≤ BMI < 25 kg/m², overweight: 25 ≤ BMI < 30 kg/m², class 1 obese: 30 ≤ BMI < 35 kg/m², class 2 obese: 35 ≤ BMI < 40 kg/m², class 3 obese: BMI ≥ 40 kg/m². Error bars indicate 95%CIs.

Fig. 2. Differences and 95%CIs of the mean and median menstrual cycle length with BMI by ethnicity groups.

CIs, confidence intervals; BMI, body mass index Adjusting for age, smoking, alcohol drinking, parity, physical activity, education, perceived stress scores, and MacArthur scale of subjective social status. Missing values in age, ethnicity, and BMI were excluded. Missing values in other covariates were treated with missing indicator. Analysis was restricted to participants who were under age 50 years and had BMI < 40 kg/m², and to white, Black, Asian, and Hispanic participants to avoid having strata with few observations. Underweight: BMI < 18.5 kg/m²; healthy: 18.5 ≤ BMI < 25 kg/m², overweight: 25 ≤ BMI < 30 kg/m², obese: 30 ≤ BMI < 40 kg/m². Error bars indicate 95%CIs.