The abstract is read at the publisher; the summary is JClub's.
In MESA, adding proteomic and metabolomic scores to blood pressure polygenic risk scores improved incident hypertension prediction, raising AUC to 0.72 from 0.63.
- Why it matters: Hypertension risk prediction is limited by incomplete understanding of how proteomic, metabolomic, and genetic signals contribute beyond clinical measures.
- What they did: Elastic Net built ProtS and MetS for SBP/DBP and hypertension in MESA, then combined them with BP PRSs in incremental XGBoost models, externally validated in JHS and adapted via transfer learning.
- The result: In JHS, ProtS had the highest AUC (0.68) and smallest cross-cohort decline, while measured SBP/DBP improved discrimination (PRS+BP AUC 0.80 in MESA; baseline+BP AUC 0.74 in JHS).
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.