Large-scale plasma proteomics in the UK Biobank modestly improves prediction of major cardiovascular events in a population without previous cardiovascular disease
Authors' conclusion
Large-scale plasma proteomics with data-driven and literature-based protein selection modestly improves prediction of future MACE beyond TRFs.
Study context
- Population
- UK Biobank participants without previous CVD
- Participants
- 38,380 participants
- Comparison
- Protein panels added to SCORE2 or 10 traditional risk factors vs SCORE2 or traditional risk factors alone
- Outcome
- MACE:
- Fatal and non-fatal MI
- Stroke
- Coronary artery revascularization
- Follow-up
- 10-year follow-up
Methodological detailsmachine learning (extreme gradient boosting), data- and literature-based feature selection
- Machine learning (extreme gradient boosting)
- Data- and literature-based feature selection
Performance
As reported in the abstract
MACE prediction
AUC
0.771
P < 001
Proteins + SCORE2 · net reclassification index (NRI) = 0.140
AUC
0.767
P = 0.03
Proteins + 10 traditional risk factors (TRFs) · NRI 0.053
SCORE2
0.740
AUC
SCORE2 refitted
0.749
AUC
114 proteins
0.758
AUC
Proteins + SCORE2
- AUC
- 0.771
- NRI
- 0.140
Proteins + 10 TRFs
- AUC
- 0.767
- NRI
- 0.053
Further findings
Literature-based (113 proteins from 5 studies) and random (114 proteins) selection
- Literature-based selection improved risk prediction beyond TRFs; random selection did not
Original publication
Large-scale plasma proteomics in the UK Biobank modestly improves prediction of major cardiovascular events in a population without previous cardiovascular disease. Eur J Prev Cardiol. Published 28 March 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →