Associations of accelerometer-derived daily step counts with future health risks: A phenome-wide association study...

Physical inactivity is a critical global public health concern. While guidelines often focus on moderate-to-vigorous physical activity, step-based targets are more feasible for contemporary lifestyles and provide a practical tool for clinicians and nurses to support patient education, lifestyle counseling, and long-term health management. However, evidence linking wearable device-measured daily steps to a broad spectrum of specific incident disease phenotypes remains limited.

To systematically examine the associations between accelerometer-derived daily step counts and the risk of more than 400 incident diseases.

A prospective, population-based cohort study using a Phenome-Wide Association Study approach.

The UK Biobank, a large-scale resource with participants from England, Wales, and Scotland.

We included 61,047 to 69,853 UK Biobank participants providing a full week of acceleration data between 2013 and 2015.

Daily steps were calculated using the validated Verisense algorithm from seven-day accelerometer data. Disease phenotypes were derived from electronic health records and categorized via Phecode Map v1.2 based on standard International Classification of Disease, 10th revision codes. Cox proportional hazards models were used to analyze 435 disease phenotypes (grouped by Phecodes) with ≥120 events. Models were adjusted for age, sex, ethnicity, education, smoking, alcohol, body mass index, deprivation, and self-reported health. Dose-response relationships were explored using restricted cubic splines.

During a median follow-up of 7.9 years, an increase of 5000 steps/day was associated with a lower risk of 64 incident diseases (HR range: 0.57 to 0.95). The strongest inverse associations were observed for type 2 diabetes (HR 0.80, 95% CI 0.76-0.84), acute renal failure (HR 0.82, 95% CI 0.78-0.86), Parkinson's disease (HR 0.57, 95% CI 0.49-0.67), essential hypertension (HR 0.93, 95% CI 0.91-0.95), and cholelithiasis (HR 0.77, 95% CI 0.71-0.83). For these conditions, the minimum dose for substantial benefit ranged from 8697 to 13,240 steps/day. Conversely, five conditions showed higher risk with increased steps, primarily musculoskeletal issues like internal derangement of the knee (HR 1.15, 95% CI 1.07-1.23) and osteoarthrosis (HR 1.15, 95% CI 1.06-1.24). The remaining 366 disease phenotypes showed no significant associations.

Higher daily step counts are associated with a 5% to 43% lower risk for a wide array of diseases across the human phenome. These findings support the use of step-based metrics as a simple, effective tool for clinical monitoring and health promotion, providing accessible alternatives to traditional intensity-based physical activity guidelines and offering potential applications for healthcare professionals in patient counseling and lifestyle management.