Diabetes Obes Metab. 2026 Sep 30. doi: 10.1111/dom.71399. Online ahead of print.
ABSTRACT
AIMS: Cardiovascular-kidney-metabolic (CKM) stages 0-2 represent a potentially modifiable period before overt cardiovascular complications, yet molecular markers that identify individuals at higher downstream risk remain insufficiently defined. We investigated whether plasma proteomic signatures could characterize early CKM-related molecular risk and improve prediction of incident cardiovascular, kidney and mortality outcomes.
MATERIALS AND METHODS: We analysed 43 709 UK Biobank participants with CKM stages 0-2 and measurements of 2920 plasma proteins. Multivariable Cox models evaluated protein associations with 11 incident outcomes after excluding participants with the corresponding condition at baseline. We further assessed modelled prediagnostic protein patterns, stage-related proteomic patterns across baseline CKM stage groups, and the incremental value of outcome-specific and recurrent protein panels beyond clinical models using Predicting Risk of Cardiovascular Disease Events (PREVENT)-aligned covariates.
RESULTS: A total of 157 proteins showed directionally concordant associations across all 11 outcomes, including 155 positive and 2 inverse associations. Population-level models across time-to-outcome intervals identified differences in GDF15, TNFRSF10B, WFDC2 and NT-proBNP between incident cases and matched reference participants. Eleven of the 155 proteins with positive associations across all 11 outcomes were consistently elevated across baseline CKM stage groups 0-2. In repeated internal validation, outcome-specific proteomic signatures improved discrimination beyond the clinical model, with ΔC-index values of 0.024-0.070. An exploratory recurrent 12-protein panel retained incremental value in internal evaluation, with ΔC-index gains of 0.018-0.052.
CONCLUSIONS: In early-stage CKM, plasma proteomic signatures captured preclinical molecular risk and improved prognostic discrimination beyond conventional clinical factors. These findings support further evaluation of targeted proteomic panels for earlier CKM risk stratification and preventive cardiometabolic management.
PMID:42816446 | DOI:10.1111/dom.71399