Large-scale plasma proteomics improves prediction of heart failure among MASLD individuals: A prospective cohort study
American Journal of Preventive Cardiology, 2026
Xiong Z., Chen S., Huang H., Lai S., Kuang L., Chen H., Zhang B., Wang Y., Li Y., Zhu H., Li Y., Huang E., Liu D., Mao C., Li Z.
| Disease area | Application area | Sample type | Products |
|---|---|---|---|
Metabolic Diseases CVD Hepatology | Patient Stratification | Plasma |  Olink Explore 3072/384 |
Abstract
Background
Metabolic dysfunction-associated steatotic liver disease (MASLD) substantially elevates the risk of heart failure (HF). While large-scale proteomics improves HF prediction in general populations, its incremental predictive value beyond standard clinical models in MASLD remains unexplored.
Objective
To identify plasma protein biomarkers for incident HF in MASLD and evaluate the predictive utility of integrating these signatures with the predicting risk of cardiovascular disease events (PREVENT) clinical model.
Methods
We prospectively analyzed 17,091 individuals with MASLD at baseline. Multivariable and LASSOsingle bondCox regressions were applied to 2911 plasma proteins to identify optimal predictors. Predictive discrimination and reclassification were assessed using Harrell’s C-index, time-dependent area under the curve (AUC), net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA).
Results
Over a median follow-up of 13.56 years, 953 incident HF events occurred. Integrating the PREVENT model with a 37-protein panel substantially improved predictive discrimination (C-index 0.805 vs. 0.723; ΔC-index 0.082, 95%CI: 0.064–0.100). Moreover, a parsimonious model containing only 5 proteins (NT-proBNP, WFDC2, LTBP2, BCAN, HAVCR1) delivered a meaningful incremental improvement over the PREVENT baseline (C-index 0.769 vs. 0.723; ΔC-index 0.046, 95%CI: 0.027–0.064). Pathway analyses indicated these proteins associating with systemic inflammation and extracellular matrix remodeling.
Conclusions
Large-scale proteomics significantly enhances HF risk prediction in MASLD, providing a robust tool for identifying high-risk individuals who may benefit from intensive clinical monitoring and preventive strategies.