Sleep breathing patterns linked to death risk in adults
Adults with high nighttime breathing rates and variability had a near five-fold higher three-year incidence of death than those low on both measures.

Our read
Biomarkers- Studied:
- Cohort study of 5,679 adults, three years, against the low-risk group (Dose-response)
- What changed:
- Death from any cause was higher in the high-risk group compared with the low-risk group (hazard ratio 2.15)
- The catch:
- The conference abstract did not report absolute death rates
Editor of record: Boris Djordjevic · AgingResearch.News
The 5,679 adults in the Sleep Heart Health Study averaged 63 years old; 52% were women. In an observational analysis, researchers used chest-belt recordings from an overnight sleep test to measure each person’s median, or middle, breathing rate and how much it varied.
Higher nighttime rates and greater variability were each linked to death from any cause and death from heart or blood vessel disease over three years. The group high on both measures included 412 people, compared with 3,828 low on both. After adjustment for age, sex, smoking and several health conditions, the high-risk group had a hazard ratio of 2.15 for death from any cause compared with the low-risk group. This measure describes relative risk over time, not the absolute chance of dying.
Why it matters
Tracking health risks in later life may benefit from measurements collected during sleep. The authors suggested that consumer and contactless devices may support wider use of nighttime breathing for long-term risk monitoring.
Caveats
Health differences not captured in the adjustments could have contributed to both altered sleep breathing and higher mortality. The conference abstract did not report absolute death rates.
- Combined nocturnal respiratory rate risk status
- Median nocturnal respiratory rate
- Nocturnal respiratory rate standard deviation
- Humans
The paper
University of California San Diego
AHA Scientific Sessions 2026 Abstract, Nov 2, 2026
Presented at AHA Scientific Sessions 2026, abstract 18322



Comments
Be the first to comment.