Abstract 14120: Proteomics of Diabetes-Related Heart Failure: The Atherosclerosis Risk in Communities (ARIC) Study

Abstract

Introduction: Data on the pathophysiology of diabetes-related heart failure (HF) mainly stems for animal studies . There is a paucity of large-scale human studies on proteomic alterations that characterize diabetes-related HF. Hypothesis: Diabetes-related HF has unique proteomic markers. Method: We analyzed 4,955 plasma proteins using SomaLogic v4.0 on 10,189 ARIC Study participants (Visit 2, 1990-92, mean age: 57±7 years, 56% women, 22% Black adults, 14% with diabetes [n=1477]). We used Cox regression to assess proteins associated with diabetes-related HF, in a 2/3 discovery sample and then a 1/3 validation subset, using Bonferroni correction of P -values. Results: Over a median follow-up of 24 years, there were 2417 HF events overall (605 among individuals with diabetes). After adjustment for demographics, clinical, and metabolic and biological factors (Figure), 18 proteins were associated with HF specifically among individuals with diabetes in the discovery sample ( p<10 -6 ), with 13 confirmed in the validation sample ( p<0.05/18 ). The validated proteins included cartilage intermediate layer protein 2 and N-terminal pro-BNP, and collagen alpha-1(XV) chain, which are not unique to diabetes-related HF. Of the validated proteins, 11 were uniquely associated with HF among diabetic individuals, and included : myocilin, lymphocyte activation gene 3 protein, and collectin-11. The latter proteins were not associated with HF among individuals without diabetes. Conclusions: A large-scale proteomics approach identified unique proteomic markers of diabetes-related HF. Pathway analysis will shed additional light on the exact role of the novel proteins in the pathogenesis of diabetes-related HF.