Menstrual Cycle Variations in Wearable-Detected Finger Temperature and Heart Rate, But Not in Sleep Metrics, in Young and Midlife Individuals
Elisabet Alzueta,Marie Gombert-Labedens,Harold Javitz,Dilara Yuksel,Evelyn Perez-Amparan,Leticia Camacho,Orsolya Kiss,Massimiliano de Zambotti,Negin Sattari,Andres Alejandro-Pena,Jing Zhang,Alessandra Shuster,Allison Morehouse,Katharine Simon,Sara Mednick,Fiona C Baker
DOI: https://doi.org/10.1177/07487304241265018
Abstract:Most studies about the menstrual cycle are laboratory-based, in small samples, with infrequent sampling, and limited to young individuals. Here, we use wearable and diary-based data to investigate menstrual phase and age effects on finger temperature, sleep, heart rate (HR), physical activity, physical symptoms, and mood. A total of 116 healthy females, without menstrual disorders, were enrolled: 67 young (18-35 years, reproductive stage) and 53 midlife (42-55 years, late reproductive to menopause transition). Over one menstrual cycle, participants wore Oura ring Gen2 to detect finger temperature, HR, heart rate variability (root mean square of successive differences between normal heartbeats [RMSSD]), steps, and sleep. They used luteinizing hormone (LH) kits and daily rated sleep, mood, and physical symptoms. A cosinor rhythm analysis was applied to detect menstrual oscillations in temperature. The effect of menstrual cycle phase and group on all other variables was assessed using hierarchical linear models. Finger temperature followed an oscillatory trend indicative of ovulatory cycles in 96 participants. In the midlife group, the temperature rhythm's mesor was higher, but period, amplitude, and number of days between menses and acrophase were similar in both groups. In those with oscillatory temperatures, HR was lowest during menses in both groups. In the young group only, RMSSD was lower in the late-luteal phase than during menses. Overall, RMSSD was lower, and number of daily steps was higher, in the midlife group. No significant menstrual cycle changes were detected in wearable-derived or self-reported measures of sleep efficiency, duration, wake-after-sleep onset, sleep onset latency, or sleep quality. Mood positivity was higher around ovulation, and physical symptoms manifested during menses. Temperature and HR changed across the menstrual cycle; however, sleep measures remained stable in these healthy young and midlife individuals. Further work should investigate over longer periods whether individual- or cluster-specific sleep changes exist, and if a buffering mechanism protects sleep from physiological changes across the menstrual cycle.