Plasma Ceramides as Biomarkers for Microvascular Disease in Diabetes: Evaluating the Relationship Between Ceramide and Multiple Outcomes

Abstract

Abstract Background We aimed to evaluate ceramides as biomarkers for diabetic microvascular disease. Methods 309 patients were prospectively enrolled from 2018 to 2020: healthy controls (group 1/N = 51), diabetes mellitus (DM) without Acute Myocardial Infarction (AMI) (group 2/N = 150), and DM with AMI (group 3/N = 108). Outcomes measured were coronary microvascular disease (CMD) using stress perfusion cardiac magnetic resonance imaging (outcome 1), retinal microvascular disease (RMD) using retinography (outcome 2), both (CMD & RMD) (outcome 3), or absence of microvascular disease (AMD) (outcome 4). Biomarker classification performance were evaluated using receiving operator curve analysis (AUC) and multiple logistic regression. Reference three ceramide ratios associated with diabetes were analyzed and compared with 11 ceramides (panel) previously identified by our study group. Results HbA1c mean values were 5.1% (group 1), 8.3% (group 2) and 7.6% (group 3). CMD was observed in 59.5% of patients, RMD in 25.8%, CMD&RMD in 18.8%, and AMD in 38.5%. The AUC using reference ceramide ratios for each outcome were: 0.66 (p = 0.012) (CMD), 0.61 (p = 0.248) (RMD),0.64 (P = 0.282) (CMD&RMD) and 0.67 (P = 0.010) (AMD). However, AUC using 11 ceramides improved significantly: 0.81 (p = 0.001) (CMD), 0.73 (p = 0.010) (RMD),0.73 (P = 0.04) (CMD&RMD) and 0.83 (P = 0.010) (AMD). Furthermore, specific ceramides features were identified for each outcome. Only increased C14.0 was positively associated with AMD (p < 0.001). Conclusions Plasma ceramides predict health status and microvascular disease sub-phenotypes in DM.