Variation in menstrual cycle length by age, race/ethnicity, and body mass index in a large digital cohort of women in the US

Abstract

AbstractBackgroundMenstrual characteristics are important signs of women’s health. We examined the variation of menstrual cycle length by age, race and ethnicity, and body weight using data collected from mobile menstrual tracking apps. Understanding how menstrual characteristics vary by these factors can provide important information for further study of environmental and social determinants of menstrual health.MethodsWe collected self-tracked menstrual cycle data from participants of the Apple Women’s Health Study. Demographic and lifestyle characteristics were self-reported from surveys. Linear mixed effect (LME) models were used to estimate the differences in cycle length associated with age, race/ethnicity, and body mass index (BMI), adjusted for possible confounders or predictors of cycle length. Cycle variability was estimated by the change of within-individual standard deviations of cycle length.FindingsA total of 165,668 cycles from 12,608 participants from US were included. After adjusting for all covariates, mean menstrual cycle length was shorter with older age across all age groups until age 50 and then became longer for those age 50 and older. Menstrual cycles were on average 1·6 (95%CI: 1·2, 2·0) days longer for Asian and 0·7 (0·4, 1·0) days longer for Hispanic participants compared to White non-Hispanic participants. Participants with Class 3 obesity (BMI≥40 kg/m2) had 1·5 (1·2, 1·8) days longer cycles compared to those with healthy BMI (18·5≤BMI<25 kg/m2). Cycle variability was smaller among participants in older age groups but became considerably larger for those in the oldest age categories (45-49 and 50+). Asian and Hispanic participants and those who were obese had larger cycle variability.InterpretationThis study demonstrated differences in menstrual characteristics by age, race and ethnicity, and obesity using data collected from mobile health apps. Future studies should explore the underlying determinants of the racial and ethnic differences in menstrual characteristics.FundingApple Inc.Research in contextEvidence before this studyWe searched PubMed for studies on menstrual cycles with age, body weight, and race/ethnicity using the term “((“Menstrual Cycle”[Mesh:NoExp] OR menstrual cycle*[tiab]) AND (“Time Factors”[Mesh] OR cycle length*[tiab] OR variability[tiab] OR variation*[tiab])) OR (“Menstruation”[Mesh] AND (“Time Factors”[Mesh] OR length[tiab] OR variability[tiab] OR variation*[tiab]) AND 1950[pdat]:1986[pdat]) AND (“Age Factors”[Mesh:NoExp] OR “Race Factors”[Mesh] OR “Racial Groups”[Mesh] OR “Ethnicity”[Mesh] OR “ethnology” [Subheading] OR “Body Mass Index”[Mesh] OR “Body Weight”[Mesh:NoExp] OR “Overweight”[Mesh] OR age[tiab] OR ages[tiab] OR race[tiab] OR races[tiab] OR racial[tiab] OR ethnic*[tiab] OR body mass[tiab] OR bmi[tiab] OR weight[tiab] OR overweight[tiab] OR obes*[tiab]) NOT (“Animals”[Mesh] NOT “Humans”[Mesh])” from 1950 to August 3, 2022. This search yielded 2,064 sources and we identified 30 research articles comparing menstrual cycle length and/or variability by age, race/ethnicity, and/or body weight in population. Another 14 research articles were added to the search results by knowledge of existing literature. Of the 44 articles identified, studies on changes of menstrual cycle length and variability with age reported consistent results, although the age range of the participants differed across these studies. In general, menstrual cycles were longer and less regular in the first few years following menarche, but over the subsequent decades cycles tend to become shorter and more regular with older age until age 40-45, after which they can become increasingly longer and irregular until menopause. Obesity was associated with menstrual irregularity. However, the association with menstrual cycle length was less consistently reported. Some studies found obesity was associated with longer menstrual cycles while other studies reported null associations. Compared to age and body weight, fewer studies have considered racial and ethnic differences of menstrual characteristics, although results from separate studies in Japan, China, and India suggested that females in these countries had longer cycle lengths compared to those observed in White females in US. Many of the prior studies of menstrual cycle characteristics relied on self-reported typical menstrual cycle length and regularity in survey questions, without acquiring actual menstrual cycle data. Most of those that used menstrual diaries to obtain cycle characteristics with higher accuracy were limited by relatively small numbers of participants. A few recent studies using menstrual tracking app data from a large number of users in different countries (mainly in Europe and North America) reported similar associations of age with cycle length and variability. However, results for body weight were still inconsistent. In addition, menstrual cycle characteristics by race and ethnicity have seldom been characterized using such data.Added value of this studyIn this large digital cohort study, we collected menstrual cycle data from menstrual tracking apps and factors related to menstruation from surveys to comprehensively compare the distribution of menstrual cycle length by age, race and ethnicity, and body mass index in a diverse US population. Using this dataset, our study results confirmed the non-linear changes of menstrual pattern with age throughout the reproductive lifespan as characterized in previous studies. More importantly, after controlling for potential confounders, we observed racial and ethnic differences of menstrual cycle length, in which participants who were Asian and who were Hispanic had longer menstrual cycles and higher variability. Obesity was associated with longer menstrual cycle length and higher cycle variability, especially among Hispanic individuals.Implications of all the available evidenceMenstrual cycle characteristics such as cycle length and variability/regularity have been recognized as important vital signs associated with gynecological conditions, fertility, cardiometabolic diseases, and mortality risk. Previous studies provided different estimates on menstrual parameters and current recommendations on normal menstrual cycle length and variability were based on limited quantitative evidence that was mainly generated in the White population. With the popularity of menstrual cycle tracking apps, our study demonstrated that this app-based data is a promising and powerful resource for research of menstrual health today. This study also provided valuable evidence on possible racial and ethnic disparities in menstrual cycle characteristics in a large free-living population. Future studies should consider the underlying environmental, social, and behavioral factors that drive the observed racial and ethnic differences of menstrual characteristics.