Improved prediction and risk stratification of major adverse cardiovascular events using an explainable machine learning approach combining plasma biomarkers and traditional risk factors
Authors' conclusion
Integrating traditional risk factors and biomarkers improves prediction and risk stratification of new-onset MACE. The BRS shows promise as a tool for identifying high-risk individuals, with the potential to support personalized CVD prevention and management strategies.
Study context
- Population
- UK Biobank participants free of baseline major adverse cardiovascular events (MACE); median age 58.0 years; 45.4% male
- Participants
- 229,352 participants
- Comparison
- Traditional risk factors combined with biochemical and metabolomic biomarkers vs traditional models (Age + Sex, ASCVD)
- Outcome
- Prediction of new-onset MACE:
- Discriminative performance
- C-index
Methodological detailsmachine learning (CatBoost, SHAP), Cox proportional hazards models, UK Biobank
- Machine learning (CatBoost, SHAP)
- Cox proportional hazards models
- UK Biobank
Performance
As reported in the abstract
Discriminative performance for new-onset endpoints
C-index, hemorrhagic stroke (suboptimal)
0.699
C-index, other outcomes
>0.75
C-index, cardiovascular diseases (CVD)-related mortality (highest)
0.822
Further findings
| Outcome | RR (95% CI) | |
|---|---|---|
| CVD death by BRS risk groupHigh-riskvs Low-riskStrongest effect observed for CVD death | 2.76 (2.48–3.07) | |
| 0.20.5125Relative risk (log scale) |
Key predictors of new-onset MACE
- Key predictors: cystatin C, HbA1c, GlycA, GGT; IGF-1 and DHA potentially protective
Original publication
Improved prediction and risk stratification of major adverse cardiovascular events using an explainable machine learning approach combining plasma biomarkers and traditional risk factors. Cardiovasc Diabetol. Published 2 April 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →