Innovative Insights into Menstrual Health Monitoring Through Wearable Technology

An analysis of 1.2 million days of data from 2,596 women across 42,759 menstrual cycles demonstrates that cycle length strongly influences the magnitude of fluctuations in cardiorespiratory biometrics, while shorter sleep durations correlate with greater cycle length deviation. Resting heart rate reaches its minimum near menstruation and peaks before the next onset, with heart rate variability exhibiting the inverse pattern; respiratory rate and skin temperature follow similar cyclic rises. These patterns were quantified at daily resolution and shown to scale with both participant age and cycle length.

Sleep’s Grip on Cycle Stability

Participants sleeping six hours per night exhibited a 1.3 times greater odds ratio of cycle length deviation of three or more days compared with those sleeping eight hours, after adjustment for age, body mass index, sleep onset, and workout patterns. Greater variability in sleep duration likewise increased the odds of cycle deviation by a factor of 1.2. A within-participant comparison of 813 individuals who experienced both high and low sleep variability periods confirmed an odds ratio of 1.3 for more variable cycle lengths during high-variability weeks. Decreases of 10 percent or more in sleep duration occurred more frequently in the premenstrual week than in menstrual or postmenstrual weeks.

Wearable Data Methodology

Daily biometric and sleep measures were obtained from wrist-worn devices that recorded resting heart rate, heart rate variability, respiratory rate, skin temperature, and blood oxygen saturation, with menstruation status logged via a connected smartphone application. Generalized estimating equation models related sleep metrics to cycle outcomes while controlling for age, body mass index, and activity; generalized additive models isolated the independent effects of age and cycle length on biometric trajectories across the cycle.

Benchmarks for Next-Gen Tools

The resulting normative profiles of daily biometric changes across ages 18–50 and cycle lengths 21–35 days supply reference trajectories that future digital health applications could use to interpret individual wearable readings. Because the observed relationships between sleep behavior and cycle variability were derived from passively collected longitudinal data, they offer an ecologically valid basis for designing interventions aimed at stabilizing sleep to reduce cycle irregularity. The detailed characterization of physiological signals may eventually support health economics and outcomes research evaluations of wearable-based menstrual health monitoring by providing the empirical benchmarks needed for such assessments.