Longevity & AgingResearch PaperOpen Access

Wearable heart rate variability tracks premenstrual symptoms in women with PMS and PMDD

In 193 young women wearing a Huawei tracker, lower nocturnal HRV tracked worse premenstrual symptoms only in women with PMDs, hinting at a digital biomarker.

Friday, November 20, 2026 0 views
Published in Arch Womens Ment Health
Young woman's wrist with a fitness tracker on a nightstand at dawn, a glowing heart-rate waveform and a faint monthly cycle arc overlay

Summary

Premenstrual disorders (PMS and PMDD) often go undiagnosed for years because diagnosis needs two months of daily symptom charting. This prospective study of 193 Chinese medical students (68 with prospectively confirmed PMDs, 293 cycles) used a wrist-worn Huawei tracker to measure nocturnal heart rate variability (HRV), a marker of autonomic nervous system activity. HRV measures reflecting parasympathetic activity (SDNN, rMSSD, HF) fell before menses and rose afterward in all women, with a larger rebound in women without PMDs. Lower HRV tracked higher symptom scores only in women with PMDs, not in those without. This is an observational association with a surrogate marker, not a validated diagnostic test, and the cohort is young and homogeneous.

Detailed Summary

Premenstrual disorders affect roughly 20-30% (PMS) and 2-8% (PMDD) of women of reproductive age, yet diagnosis takes an average of about 12 years. The only formal route is two cycles of daily symptom charting. An objective, passively collected marker could shorten that path. Heart rate variability, a readout of autonomic balance, is a candidate because it shifts with the menstrual cycle and relates to anxiety, depression and irritability.

The researchers nested a prospective study in the Care of Premenstrual Emotion cohort of medical students at Sichuan University. Participants (mean age about 20) wore a Huawei Fitness Tracker 6 Pro, which uses photoplethysmography. HRV was taken from the 3-5 a.m. window with four layers of quality control. They also completed the Daily Record of Severity of Problems daily for one to two cycles. Of 193 women, 68 met prospective criteria for PMDs (41 PMS, 27 PMDD). The analysis used 293 cycles, mixed-effects models adjusted for age, menarche age, BMI and alcohol use, and inverse-probability weights to correct for oversampling of screen-positive women.

In all women, SDNN, rMSSD and HF declined in the premenstrual period and rose after menses. The rebound was more pronounced in women without PMDs, and women with PMDs stayed lower across most post-menses days. In the week before and after menses, lower HRV was associated with higher symptom scores in women with PMDs (rMSSD postmenstrual week β = -0.036 per SD) but not in those without PMDs (β = -0.001; P-for-difference <0.001). The signal was stronger for affective than physical symptoms. Before menses it was stronger in PMDD than PMS, while after menses the two groups were similar. LF/HF showed no clear separation. Results held when controls were restricted to 43 'pure controls' and in women with regular 26-31 day cycles.

The authors propose that the pattern fits hormonal effects on autonomic tone, with estrogen favoring parasympathetic activity and progesterone dampening baroreflex sensitivity. This is interpretation, not something the study tested. The authors suggest that impaired restoration of parasympathetic activity after menses may characterize PMDs, but this requires replication.

Caveats are substantial. The effect sizes are small, and the study shows association rather than diagnostic accuracy, so no sensitivity, specificity or prediction was assessed. The sample is young medical students, mostly not on hormonal contraception (95.5%), and the controls were mostly lighter-symptom women. Wrist PPG HRV agrees only moderately with ECG, and the cycle days were self-reported.

Key Findings

  • Nocturnal SDNN, rMSSD and HF fell premenstrually and rose after menses in all 193 women; the rebound was larger without PMDs.
  • Lower HRV tracked higher symptoms only in women with PMDs (rMSSD postmenstrual β = -0.036 per SD); no association without PMDs (β = -0.001).
  • The HRV-symptom link was stronger for affective than physical symptoms, and stronger premenstrually in PMDD than PMS.
  • LF/HF did not show the same pattern; results were similar in 43 pure controls and in regular-cycle subgroups.
  • Diagnostic accuracy was not tested, so HRV is a candidate correlate, not a validated PMD screening tool.

Methodology

Prospective observational study nested in a cohort: 193 women (68 with PMDs), 293 cycles, nightly (3-5 a.m.) wrist PPG HRV from a Huawei tracker with four-layer artifact filtering. PMDs were confirmed by daily DRSP charting. Mixed-effects models were adjusted for covariates and weighted by inverse probability, with sensitivity analyses for pure controls and regular cycles.

Study Limitations

Observational, with small effects and no diagnostic accuracy testing or prediction. The sample is young, Chinese medical students, which limits generalizability, and wrist PPG HRV correlates only moderately with ECG. Cycle days and symptoms were self-reported, many controls had subclinical symptoms, hormones were not measured, and the autonomic and hormonal mechanism is inferred rather than tested.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: