Heart failure & cardiomyopathies
Predicting Acute Decompensated Heart Failure Using Circadian Markers From Heart Rate Time Series
Authors' conclusion
Circadian metrics from a wrist-worn device showed progressive alterations over the 3 weeks preceding ADHF, offering potential early detection of HF decompensation with moderate prediction performance.
Study context
- Population
- Patients discharged after acute decompensated heart failure (ADHF) hospitalisation, monitored with a wrist-worn HR device for 6 months after reaching stable heart failure (HF)
- Participants
- 65 patients
- Comparison
- Circadian HR parameters in 21 days before HF events vs other periods
- Outcome
- Prediction of HF events:
- Unplanned outpatient visits for congestion
- Increased diuretics
- ADHF hospitalisation
- SCD
- Follow-up
- 21-day periods before an HF event
Methodological detailswrist-worn device, cosinor analysis, long short-term memory neural network
- Wrist-worn device
- Cosinor analysis
- Long short-term memory neural network
Performance
As reported in the abstract
ADHF prediction by circadian-marker algorithm
Sensitivity
74%
Specificity
73%
AUC
74%
P < 0.001
Amplitude contribution to feature importance (strongest predictor)
62%
Further findings
Circadian changes in 21 days preceding HF events · 21 days before HF events
- Mesor elevated: 70.6 vs 73.6 b.p.m.
- P < 0.001
- Amplitude reduced: 8.3 vs 4.9 b.p.m.
- P = 0.046
- Acrophase shift: 11.3 vs 12.2 h
- P = 0.706
Original publication
Predicting Acute Decompensated Heart Failure Using Circadian Markers From Heart Rate Time Series. ESC Heart Fail. Published 28 August 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →