A 17-protein risk score improved heart failure prediction beyond clinical variables, polygenic risk, and NT-proBNP in individuals with type 2 diabetes, yielding a maximum C-index of 0.833 with an increment of 0.091.
During a median follow-up of 13.1 years, 298 out of 2198 participants with type 2 diabetes developed incident heart failure.
A total of 455 plasma proteins were significantly associated with incident heart failure in individuals with type 2 diabetes.
WAP 4-disulfide core domain protein 2 (WFDC2) was the top protein positively associated with increased heart failure risk.
Apolipoprotein C-I was the top protein inversely associated with heart failure risk.
A 17-protein risk score significantly enhanced heart failure prediction beyond clinical variables, polygenic risk, and NT-proBNP.
The proteomic predictors were selected from clinical and protein predictors using a LASSO-based variable selection approach.
This research suggests that measuring specific proteins in the blood can meaningfully improve the ability to predict which people with type 2 diabetes will go on to develop heart failure. The researchers analyzed blood samples from nearly 2,200 people with type 2 diabetes who were followed for over 13 years, during which almost 300 developed heart failure. Out of nearly 3,000 proteins measured, 455 were found to be linked to heart failure risk — with one protein called WFDC2 showing the strongest association with increased risk, and a protein called apolipoprotein C-I showing a protective association. The researchers then identified a set of 17 key proteins that, when combined into a risk score, substantially improved heart failure prediction compared to using standard clinical information, genetic risk scores, and an established heart failure blood marker called NT-proBNP alone. The combined model achieved a C-index (a measure of predictive accuracy) of 0.833, representing a meaningful improvement over existing approaches. The proteins identified are linked to biological processes such as cell adhesion, immune signaling, and inflammation, potentially offering clues about the mechanisms through which heart failure develops in people with diabetes. This research suggests that plasma protein profiling could one day be used as part of clinical risk assessment to identify people with type 2 diabetes who are at highest risk of developing heart failure, potentially enabling earlier preventive interventions. However, further validation in diverse populations would be needed before such approaches could be translated into routine clinical practice.
Yu H, Zhang J, Qian F, Liu J, Zhu K, Qiu Z, et al.. (2026). Plasma proteomics enhances heart failure risk prediction among individuals with type 2 diabetes: a prospective cohort study.. The American journal of clinical nutrition. https://doi.org/10.1016/j.ajcnut.2026.101216