What Large-Scale NIH Wearables Cohort Data Show About Early Detection in Men

NIH-Funded Wearable Device Studies Link Continuous Metrics to Cardiovascular Risk Models for Men

CDC data show heart disease is the leading cause of death for men in the United States. It accounts for roughly one in four male deaths and registers higher prevalence and mortality than in women. The Apple Heart Study tested a smartwatch algorithm for atrial fibrillation detection across more than 419,000 adults. Irregular pulse notifications reached 0.52 percent of participants, and ECG monitoring confirmed atrial fibrillation in 34 percent of those who received notifications and completed follow-up.

What this means

NIH-supported research links continuous wearable device metrics to enhanced cardiovascular risk models in men. Heart rate variability, step counts, sleep patterns and activity levels can flag subclinical signals tied to atrial fibrillation and heart failure. Models that fold in these data have shown tighter risk stratification than tools limited to age, sex and single clinic readings. Gains stay investigational. Randomized trials have not yet proven that wearable-guided approaches cut hard cardiovascular events.

Key takeaways

• The Apple Heart Study sent irregular pulse notifications to 0.52 percent of its more than 419,000 participants. [1]
• ECG patch monitoring confirmed atrial fibrillation in 34 percent of those who followed up after a notification. [1][2]
• Heart disease accounts for about one in four male deaths in the United States and shows higher mortality than in women. [3]
• Wearable devices generate continuous data on heart rate, activity and sleep that complement single-point clinical measurements.
• Some analyses suggest sex-specific models may improve performance because male patterns of physical activity, heart-rate variability and sleep differ from female patterns.

Wearable metrics and cardiovascular signals

The Apple Heart Study evaluated a smartwatch algorithm at large scale for detection of atrial fibrillation. [1] Machine learning models trained on wearable heart rate, activity and sleep data associate with incident events such as atrial fibrillation and heart failure. These continuous streams differ from the limited snapshots gathered in routine office visits. Patterns observed in men have prompted calls for tailored model calibration.

Comparison with traditional risk scores

Standard tools such as the Framingham and ASCVD scores draw on a narrow set of static factors. Models that add wearable-derived variables track changes across days and weeks. Early NIH-supported findings associate the combined approach with identification of men placed in higher-risk groups that conventional scores might miss. Accuracy gains stay under active investigation. Outcome benefits have not been established in randomized trials.

Barriers to clinical integration

Most wearable studies draw participants who are younger, healthier and less racially diverse than the broader male population at highest risk. Adherence fluctuates. Motion artifacts and differences across device brands complicate the records. No device holds clearance specifically for male cardiovascular risk stratification.

Limitations

Participant populations in wearable studies are often healthier, younger, and less racially diverse than the general male population at risk. Device adherence, motion artifact, and inter-device variability reduce data quality and generalizability. Most studies detect associations but lack long-term randomized evidence that wearable-guided interventions reduce hard cardiovascular events. Regulatory pathways for AI-driven wearable risk scores are still evolving. No device is currently cleared specifically for male CV risk stratification.

Sources / References

[1] Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation. The New England Journal of Medicine. https://www.nejm.org/doi/full/10.1056/NEJMoa1901183

[2] Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation: The Apple Heart Study. PubMed. https://pubmed.ncbi.nlm.nih.gov/31722151/

[3] Heart Disease Facts. Centers for Disease Control and Prevention. https://www.cdc.gov/heartdisease/facts.htm