Development and evaluation of cardiovascular disease-specific frailty index: a machine learning based analysis of the UK biobank
Authors' conclusion
Newly developed FICVD demonstrated better performance in prediction and risk stratification of CVD.
Study context
- Population
- UK Biobank participants
- Participants
- 366,622 participants
- Exposure
- Cardiovascular diseases (CVD)-specific frailty index vs traditional frailty index
- Outcome
- Incident CVD
Methodological detailsmachine learning (elastic net regression), development, temporal validation and spatial validation cohorts
- Machine learning (elastic net regression)
- Development, temporal validation and spatial validation cohorts
Main finding
As reported in the abstract
Incident CVD (development cohort)
| Group | HR per 1 SD increase | AUC, 10-year CVD |
|---|---|---|
| FICVD | 1.40 (95% CI: 1.37-1.43) | 0.649 |
| Frailty index (FI) | 1.26 (95% CI: 1.24-1.27) | 0.605 |
HR per 1 SD increase
FICVD1.40 (95% CI: 1.37-1.43)
Frailty index (FI)1.26 (95% CI: 1.24-1.27)
AUC, 10-year CVD
FICVD0.649
Frailty index (FI)0.605
Observational data: associations, not effects.
Further findings
| Outcome | Fold risk (95% CI) |
|---|---|
| CVD risk by FICVD and genetic risk (PRS)High FICVD and high genetic risk vs Low FICVD and low genetic risk | 2.84 (2.79–2.96) |
Temporal and spatial validation cohorts
- Similar patterns observed in both validation cohorts
Original publication
Development and evaluation of cardiovascular disease-specific frailty index: a machine learning based analysis of the UK biobank. Eur J Prev Cardiol. Published 22 August 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →