What does this research mean for the field?

Cardiac patients exhibited higher variability in R-wave amplitude and longer R-wave duration compared to healthy individuals using smartwatch-derived ECGs. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The study aims to investigate the accuracy and feasibility of smartwatch-derived ECG parameters in distinguishing between healthy individuals and those with cardiac disease.

February 8, 2026

Differences in smartwatch-derived single-lead ECGs in cardiac disease

Key Result

Smartwatch ECGs showed cardiac patients had higher R-wave variability, longer R-wave duration, and lower T-wave amplitude vs healthy individuals, detectable in 77.5% cases.

Key Points

The study aims to investigate the accuracy and feasibility of smartwatch-derived ECG parameters in distinguishing between healthy individuals and those with cardiac disease.
Recorded 30-second ECGs using an Apple Watch (Series 9) in healthy individuals and cardiac patients.
Analyzed P-wave, R-wave, and T-wave parameters with the NeuroKit algorithm for amplitude, variability, and duration.
Applied quality control to ensure data integrity by deleting poor cardiac cycles.
Analyzed data from 160 patients with at least 10 viable cardiac complexes in 124 individuals (77.5%).
Found increased R-wave variability and longer R-wave duration in cardiac patients compared to healthy individuals.
Noted reduced T-wave amplitude in cardiac patients, with no differences in T-wave variability or duration.

Structured PICO

Do automatically derived parameters from smartwatch single-lead ECGs differ between healthy individuals and patients with cardiac disease?

Population

160 individuals (124 with viable ECGs analyzed: 53 healthy and 79 with cardiac disease including coronary artery disease, severe valvular disease, heart failure, arrhythmias, and cardiomyopathy).

Intervention

30-second single-lead ECG recording using Apple Watch (Series 9) with automated parameter extraction via NeuroKit algorithm

Comparator

Healthy individuals (n=53)

Outcome

Differences in automatically derived ECG parameters (P-wave, R-wave, and T-wave amplitude, variability, and duration)surrogate

Smartwatch-derived single-lead ECGs can automatically extract wave parameters that significantly differ between healthy individuals and cardiac patients, suggesting potential utility for remote monitoring.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

Abstract Background Smart-watch derived single-lead electrocardiograms (ECG) are commonly available and increasingly used to detect arrhythmias. However, differences in those ECGs between healthy persons and cardiac patients have not yet been explored. Purpose Investigate the feasibility and accuracy of automatically derived ECG parameters from smartwatch ECGs. Methods We investigated the P-wave, R-wave, and T-wave in healthy people and patients with cardiac disease. A 30 second ECG was recorded with an Apple Watch (Series 9). The raw ECG data was exported, and the specifics of the single waves were automatically extracted: The NeuroKit algorithm calculated the amplitude, the variability, and the duration of each valid wave; then a template-matching quality control algorithm deleted poor cardiac cycles. Results We analysed a total of 160 patients. The algorithm detected at least 10 viable cardiac complexes in 124 patients (77.5%). Of those 124 patients, 53 were healthy and 79 had cardiac disease (coronary artery disease: 37 46.8%; severe valvular disease: 9 11.7%; heart failure: 39 49.4%; arrhythmias: 13 16.5%; cardiomyopathy: 19 24.1%). Cardiac patients were older (65.7±12.6 vs 53.8±15.8 years, p0.001), less frequently female (17.7% vs 68.2%, p0.001), and had a higher body-mass index (29.1±4.9 vs 26.9±5.9 kg/m², p=0.039). We found no difference for the P-wave (Fig 1). Cardiac patients had no difference in R-wave amplitude, however, they exhibited a higher variability in R-wave amplitude and longer R-wave duration than healthy individuals (Fig 1). T-wave amplitude was reduced in cardiac patients with no difference in T-wave amplitude variability or T-wave duration (Fig 1). As age and body-mass index were different between healthy individuals and cardiac patients, we performed linear regression analyses of these variables with ECG parameters (Fig 2). Age was correlated with lower P-wave amplitude (r²=0.051, p=0.027) and had no influence on R-wave (r²=0.029, p=0.100) or T-wave amplitude (r²=0.007, p=0.430) (Fig 2 A-C). Body-mass index was associated with higher P-wave amplitude (r²=0.129, p=0.001), but had no effect on R-wave (r²=0.009, p=0.395) or T-wave amplitude (r²=0.006, p=0.501) (Fig 2 D-E). Conclusion Our results show that smartwatch-derived single-lead ECGs provide measurements that are correlated with cardiac disease, and which could be used to monitor healthy individuals and changes in cardiac patients.

Bookmark