Sleep patterns linked to irregular heartbeat risk in Fitbit users
Each extra hour of deep sleep was linked to 13.5% lower relative risk; each extra hour of rapid eye movement sleep to 11.3% lower risk.
Our read
Biomarkers- Studied:
- Cohort study of 43,704 Fitbit-linked participants without atrial fibrillation, median 3.2 years
- What changed:
- Atrial fibrillation risk was 13.5% lower with each additional hour of deep sleep
- The catch:
- The findings concern wearable users without previously diagnosed atrial fibrillation
- Why it’s aging research:
- It bears on staying healthy in midlife and later life.
Editor of record: Boris Djordjevic · AgingResearch.News
The 43,704 Fitbit-linked participants in this observational cohort had no previous diagnosis of atrial fibrillation, an irregular heartbeat. They came from the All of Us Research Program and had at least 10 nights of usable wearable data. Over a median follow-up of 3.2 years, health records documented 335 new diagnoses.
More deep sleep and more rapid eye movement sleep, a stage often associated with dreaming, were each linked to lower risk after adjustment for clinical risk factors and daily steps. Participants who developed atrial fibrillation averaged 0.77 hours of deep sleep nightly, compared with 0.99 hours among those who did not. An indirect measure of heart-rate variation, based on day-to-day changes in estimated average daily heart rate, was not independently linked to risk.
Why it matters
For monitoring heart health as people age, the authors suggest that passively recorded sleep stages may offer a scalable signal of atrial fibrillation risk.
Caveats
Sleep stages came from Fitbit estimates and diagnoses from health-record codes; people with longer deep or rapid eye movement sleep may also have had unmeasured health differences. The findings concern wearable users without previously diagnosed atrial fibrillation and were reported in a conference abstract.
The paper
Marshfield Clinic Rsch Institute; MCHS
AHA Scientific Sessions 2026 Abstract, Nov 2, 2026
Presented at AHA Scientific Sessions 2026, abstract 11107



Comments
Be the first to comment.