The dose-response association between wearables-quantified incidental physical activity and type 2 diabetes: a prospective cohort study
Brief vigorous daily activity was linked to lower type 2 diabetes risk.
*Prospective cohort study; Level 2 (OCEBM).
Citation
Sabag A, Biswas RK, Koemel NA, et al. The dose-response association between wearables-quantified incidental physical activity and type 2 diabetes. Journal of Clinical Endocrinology & Metabolism. 2026. doi:10.1210/clinem/dgag316
Background
Many adults struggle to do planned exercise, but everyday activities such as stair climbing, brisk walking, and active chores may improve metabolic health. This study examined whether wearable-measured incidental activity predicts new type 2 diabetes.
Patients
22,857 UK Biobank adults who reported no regular exercise. Exclusions included baseline diabetes, insulin use, inability to walk, missing data, inadequate wearable data, and diabetes diagnosed within 12 months.
Intervention
Higher daily incidental activity measured by wrist wearable: light, moderate, vigorous, and total activity energy use.
Control
Lower or no moderate-to-vigorous incidental activity; models used minimum or low activity reference levels.
Outcome
New type 2 diabetes, mainly from primary care records, with hospital and death-record linkage.
Follow-up Period
Mean 7.8 years.
Results
| Exposure | Daily amount | Type 2 diabetes risk |
|---|---|---|
| Total incidental activity energy use | Median 24.96 kJ/kg/day | Hazard ratio 0.46 (95% CI 0.36 to 0.59) |
| Vigorous incidental activity | Median 4.75 minutes/day | Hazard ratio 0.62 (95% CI 0.53 to 0.73) |
| Moderate incidental activity | Median 23.70 minutes/day | Hazard ratio 0.67 (95% CI 0.52 to 0.86) |
| Light incidental activity | Median 89.89 minutes/day | Hazard ratio 0.79 (95% CI 0.63 to 0.98) |
Vigorous and moderate activity showed clearer dose-response patterns than light activity. One minute of vigorous incidental activity corresponded to about 5.9 minutes of moderate or 46.6 minutes of light activity for similar estimated diabetes risk reduction.
Limitations
Observational design prevents proving causation. UK Biobank participants may be healthier than average. Exercise status was self-reported years before wearable measurement. Wearables may misclassify some activities, and unmeasured lifestyle factors may remain.
Funding
Australian public/charity grants; no industry funding. One author advised Longevity League Ltd.
Clinical Application
Encourage nonexercising adults to add brief vigorous daily activities; evidence supports counseling, not replacing formal diabetes-prevention programs.
Journal Club
Discussion questions and an EBM quiz for this paper. How to run a journal club →
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.