Smartwatch-derived exercise metrics as predictors of early hypertension: a prospective observational study.

BACKGROUND: Early identification of individuals at risk for hypertension is essential for effective cardiovascular disease. Physiological and activity metrics derived from consumer smartwatches may offer a practical, noninvasive approach to identify individuals at increased risk before the clinical onset of hypertension. METHODS: In this 12-month prospective observational study, 230 normotensive adults aged 30-60 years were followed using consumer smartwatches. Baseline wearable predictors were calculated as the mean of the first 30 days of valid data after enrollment and included heart rate variability, resting heart rate, and time spent in moderate-to-vigorous physical activity. Incident hypertension was defined according to current European guidelines using standardized office blood pressure measurements obtained at follow-up. RESULTS: During follow-up, 28 participants (12.2%) developed hypertension. Individuals who developed hypertension exhibited lower baseline heart rate variability and spent less time in moderate-to-vigorous physical activity compared with those who remained normotensive. In multivariable logistic regression analysis, lower heart rate variability, lower levels of physical activity, and higher body mass index were independently associated with incident hypertension. An interaction between autonomic variability and physical activity was observed, indicating that individuals with both reduced autonomic regulation and low physical activity had the highest predicted risk. Machine-learning models showed improved statistical discrimination compared with clinical variables alone and were used as complementary exploratory analyses. CONCLUSIONS: Smartwatch-derived autonomic and physical activity metrics were independently associated with the development of hypertension over a 12-month period. These findings from an observational study suggest a potential role for wearable-derived physiological parameters as digital biomarkers for early hypertension risk stratification, although further validation in larger and externally replicated cohorts is required.