Impaired Rest–Activity Rhythm Characteristics Predict Higher Risk of Incident Type 2 Diabetes in UK Biobank Participants
Authors' conclusion
Rest-activity rhythm characteristics measured from actigraphy data may serve as digital biomarkers for predicting type 2 diabetes risk.
Study context
- Population
- UK Biobank participants with actigraphy data; 13 rest-activity rhythm metrics derived
- Participants
- 83,887 participants
- Exposure
- Rest-activity rhythm metrics, quintile 1 vs quintile 5
- Outcome
- Incident T2D (self-reported data and health records)
Methodological detailsactigraphy, Cox proportional hazards models, random forest models
- Actigraphy
- Cox proportional hazards models
- Random forest models
Main finding
As reported in the abstract
Incident T2D by rest-activity metric
Hazard ratio (95% CI)P value
Lower pseudo-F statistic (Q1)vs Q5
1.27(1.09–1.46)
P < 0.001for trendLower amplitude (Q1)vs Q5
2.56(2.21–2.97)
P < 0.001for trendLower midline estimating statistic of rhythm (Q1)vs Q5
2.59(2.24–3.00)
P < 0.001for trendLower relative amplitude (Q1)vs Q5
4.64(3.74–5.76)
P < 0.001for trendLower M10 (Q1)vs Q5
3.82(3.20–4.55)
P < 0.001for trendHigher L5 (Q5)vs Q1
1.88(1.62–2.19)
P < 0.001for trendLater L5 start time (Q5)vs Q1
1.20(1.04–1.38)
P = 0.004for trendObservational data: associations, not effects.
Further findings
Relative importance in risk prediction (random forest)
- Most rest-activity metrics ranked as top predictors of diabetes incidence vs traditional diabetes risk factors
Subgroups of age, sex, BMI and shift work status
- Findings consistent across subgroups
Original publication
Impaired Rest–Activity Rhythm Characteristics Predict Higher Risk of Incident Type 2 Diabetes in UK Biobank Participants. Diabetes Care. Published 26 June 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →