Large-Scale Plasma Proteomics Improves Prediction of Peripheral Artery Disease in Individuals With Type 2 Diabetes: A Prospective Cohort Study
Hancheng Yu, Jijuan Zhang, Frank Qian, Pang Yao, Kun Xu, Ping Wu, Rui Li, Zixin Qiu, Ruyi Li, Kai Zhu, Lin Li, Tingting Geng, Xuefeng Yu, Danpei Li, Yunfei Liao, An Pan, Gang LiuOBJECTIVE
Peripheral artery disease (PAD) is a significant complication of type 2 diabetes (T2D), yet the association between plasma proteomics and PAD in people with T2D remains unclear. We aimed to explore the relationship between plasma proteomics and PAD in individuals with T2D, and assess whether proteomics could refine PAD risk prediction.
RESEARCH DESIGN AND METHODS
This cohort study included 1,859 individuals with T2D from the UK Biobank. Multivariable-adjusted Cox regression models were used to explore associations between 2,920 plasma proteins and incident PAD. Proteins were further selected as predictors using least absolute shrinkage and selection operator (LASSO) penalty. Predictive performance was assessed using Harrell's C-index, time-dependent area under the receiver operating characteristic curve, continuous/categorical net reclassification improvement, and integrated discrimination improvement.
RESULTS
Over a median follow-up of 13.2 years, 157 incident PAD cases occurred. We observed 463 proteins associated with PAD risk, primarily involved in pathways related to signal transduction, inflammatory response, plasma membrane, protein binding, and cytokine-cytokine receptor interactions. Ranking by P values, the top five proteins associated with increased PAD risk included EDA2R, ADM, NPPB, CD302, and NPC2, while BCAN, UMOD, PLB1, CA6, and KLK3 were the top five proteins inversely associated with PAD risk. Incorporating 45 LASSO-selected proteins or a weighted protein risk score significantly enhanced PAD prediction beyond clinical variables alone, reaching a maximum C-index of 0.835.
CONCLUSIONS
This study identified plasma proteins associated with PAD risk in individuals with T2D. Adding proteomic data into the clinical model significantly improved PAD prediction.