Key result
Combining wearable heart rate and symptoms identifies Long COVID with ~95% ROC-AUC versus symptoms alone.
Why the study?
Long COVID remains difficult to diagnose due to limited objective biomarkers, often leading to underdiagnosis or misdiagnosis.
Does combining wearable sensor heart rate data with symptom features improve the prediction of Long COVID status in individuals with acute SARS-CoV-2 infections?
Cohort (n=126)
No
Does combining wearable sensor heart rate data with symptom features improve the prediction of Long COVID status in individuals with acute SARS-CoV-2 infections?
Mean Difference: 0.044 (95% CI 0.04–0.048)
Absolute Event Rate: 0.951% vs 0.907%
Combining longitudinal wearable heart rate data with symptom features significantly improves the diagnostic precision of machine learning models for detecting Long COVID.
May support combined wearable-symptom models for Long COVID detection; extends symptom-only approaches but leaves open prospective validation.
Despite the millions of individuals struggling with persistent symptoms, Long COVID has remained difficult to diagnose due to limited objective biomarkers, often leading to underdiagnosis or even misdiagnosis. To bridge this gap, we investigated the potential of utilizing wearable sensor data to aid in the diagnosis of Long COVID. We analyzed longitudinal heart rate (HR) data from 126 individuals with acute SARS-CoV-2 infections to develop machine learning models capable of predicting Long COVID status using derived HR features, symptom features, or a combination of both feature sets. The HR features were derived across six analytical categories, including time-domain, Poincaré nonlinear, raw signal, Kullback-Leibler (KL) divergence, variational mode decomposition (VMD), and the Shannon energy envelope (SEE), enabling the capture of heart rate dynamics over various temporal scales and the quantification of day-to-day shifts in HR distributions. The symptom features used in the final models included chest pain, vomiting, excessive sweating, memory loss, brain fog, heart palpitations, and loss of smell. The combined HR- and symptom-feature model demonstrated robust predictive performance, achieving an area under the Receiver Operating Characteristic curve (ROC-AUC) of 95.1% and an area under the Precision-Recall curve (PR-AUC) of 85.9%. These values represent a significant improvement of approximately 5% in both the ROC-AUC and PR-AUC over the symptoms-only model. At the population level, this improvement in discrimination could lead to clinically meaningful reductions in misclassification and improved patient outcomes, achieved through a non-invasive diagnostic tool. These findings suggest that wearable HR data could be used to derive an objective biomarker for Long COVID, thereby enhancing diagnostic precision.
No takes yet. Share an insight, caveat, or question.
Uwakwe et al. (2025) conducted a cohort in Long COVID (n=126). Combined wearable heart rate and symptom data model vs. Symptom data alone model was evaluated on ROC-AUC for Long COVID detection (Mean difference 0.044, 95% CI 0.040-0.048). A machine learning model combining longitudinal wearable heart rate data and symptom features achieved an ROC-AUC of 95.1% for detecting Long COVID, a 0.044 improvement over a symptoms-only model.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: