Your Breathing Changes a Week Before Your Period

Summary

Can my wearable actually detect PMS before my period starts?

Partially. Across ~2.5 million sleep and symptom records from 2,941 users, nighttime breathing rate was the one physiological metric with a meaningful effect size (Cohen's d > 0.2, FDR-adjusted p < 0.05), rising roughly a week before menstruation and staying elevated until a week after. Other markers like heart rate, HRV, and sleep fragmentation shifted in the expected direction but the effects were small.

Why do I feel awful during PMS when my recovery score says I'm fine?

Because PMS is loud subjectively but quiet physiologically. Emotional and physical symptom tags showed large effects (|d| > 0.2) on PMS days, while objective metrics like nighttime heart rate, HRV, and sleep latency only shifted by small, statistically significant amounts. If your app says recovery is optimal but you feel exhausted, trust yourself — your intuition is more sensitive than the sensor.

How early before my period do physiological changes actually start?

About a week. Nighttime breathing rate begins climbing roughly 7 days before period start in cycles where PMS is reported, and remains elevated until about a week after menstruation begins. This makes breathing rate one of the earliest and most reliable wearable signals of the luteal phase shift.

What's happening to my body during PMS according to wearable data?

The pattern points to increased sympathetic activation: slightly higher nighttime heart rate, slightly lower HRV, small increases in sleep fragmentation and sleep latency, and a small decrease in daytime activity. None of these are dramatic, but together they describe a state of lower recovery and higher strain — which lines up with the common feeling of being 'just a bit off.'

Why do I report more cramps during PMS than during my actual period?

In this dataset, cramp tagging was actually higher on PMS days than on period days. This likely reflects tagging behavior — many women accept cramps as a given during menstruation and stop logging them — though some may genuinely experience more cramping premenstrually. Either way, the assumption that cramps peak during the period itself isn't supported by the user-reported data.

Are sugar cravings during PMS actually a real thing?

Yes, and it's the strongest tag correlation in the dataset. Among all symptom and behavior tags users logged, PMS correlated most highly with sugar cravings. So the chocolate craving stereotype shows up clearly in real wearable and self-report data.

Should I trust how I feel or what my wearable tells me during the luteal phase?

Trust how you feel. The analysis found that emotional symptoms produced larger effect sizes than physical ones, and both dwarfed the subtle physiological shifts visible on wearables. Around 90% of women experience some PMS, so if your energy dips in the luteal phase, treat it as a real biological pattern and adjust — even if your recovery score disagrees.

How was PMS isolated from normal cycle variation in this study?

The researchers compared luteal phase days with PMS reporting against luteal days without it, restricting the analysis to weeks before menstruation onset for a cleaner signal. They applied 30-day rolling Z-scores so each user was compared against her own physiological baseline, then used linear mixed-effects models to detect within-person shifts rather than population averages.