The concentration of small-diameter HDL particles (HDLp1) predicts the incidence of coronary calcium score in diabetic subjects without established cardiovascular disease

Abstract

Abstract Introduction Type 2 diabetes (DM2) decisively influences the evolution of coronary atherosclerosis. However, this is a heterogeneous and poorly understood process. While in 50% of diabetics the chance of incident Coronary Calcium Score (CAC) in 5 years exceeds 40%, in 10% of diabetics the chance of incident CAC is less than 10%. To better understand the progression of coronary atherosclerosis measured by CAC in diabetics without cardiovascular disease, we studied the impact of the concentration of the small-diameter subfraction of HDL (HDLp1, 7 to 8nm). HDLp1 has been implicated as a central mediator of the anti-inflammatory effect of the HDL pool and could play a specific role in CAC incidence among diabetics. Methods From 2370 individual with two CAC evaluations and hydrogen spectroscopy/magnetic resonance (H1-NMR) plasma profile available, we found 461 diabetic individuals, non-users of statins, without previous cardiovascular disease, with CAC=0 in the first measurement and a second CAC measurement after an average period of 5 years. [HDLp1] concentration was measured by H1-NMR and divided into tertiles. The incidence of CAC was defined as an initial CAC of 0 followed by CAC>0 in the second measurement. Stepwise forward logistic regressions were used to identify risk factors for CAC incidence. Discrimination analysis was performed by calculating the area under the ROC curve (AUROC). Results A total of 143 (31.0%) patients had incident CAC. In stepwise logistic regression, alongside factors such as age, gender, systolic blood pressure, smoking, LDL-C, C-reactive protein and HbA1c, the 3rd tertile of [HDLp1] was associated with 55.2% (95% CI 15-110%, p =0.0039) and 140% (95% CI 32-341%, p<0.001) higher risk of CAC incidence compared to the 2nd tertile and 1st tertile, respectively. When compared to a model composed of 10 confounding variables (age, gender, systolic blood pressure, BMI, smoking, Lp(a), LDL-C,triglycerides, C-reactive protein and HbA1c) for predicting the incidence of CAC, the diameter of HDLp1 substantially increased the AUROC from 0.651 (95% CI 0.597-0.705) to 0.684 (95% CI 0.631–0.736), p=0.027 for the DeLong test comparing AUROCs. A restricted cubic spline demonstrated a strong association between [HDLp1] and CAC incidence. Conclusion The concentration of HDLp1 improves the predictive capacity for the incidence of CAC in diabetic individuals, non-users of statins, without cardiovascular disease.Spline HDLp1 concentration vs CAC incideROC for incident CAC