Ratio of dense to buoyant LDL subclass is associated with LDL density phenotype (VLDL-5)

Background: Dense LDL phenotypes are associated with increased atherogenicity, and are commonly evaluated for the purposes of atherosclerosis research and cardiovascular risk discrimination. Objective: To examine the ability of LDL subclasses, expressed as a ratio of dense-to-buoyant subclass, to predict LDL density phenotype. Methods: LDL subclasses and density phenotypes were measured with vertical auto profile ultracentrifugation in 1,339,898 consecutive lipid profiles between 2009 and 2011 from a clinical sample of US adults. Logarithmic LDL density ratio (LLDR) was calculated as ratio of dense-to-buoyant LDL subclasses, ln[(LDL₃-C + LDL₄-C)/(LDL₁-C +LDL₂-C)]; normally distributed after log-transformation. LLDR was compared to density phenotype using ROC C-statistic with optimum sensitivity and specificity cutpoints determined. Results: There was a strong, highly significant, monotonic increase in LLDR across progressively higher density phenotypes (p<0.001). Mean LLDR for Phenotype A was 0.122 (95% CI 0.121-0.123), Phenotype A/B was 0.751 (95% CI 0.750-0.752), and Phenotype B was 1.336 (95% CI 1.335-1.338). ROC analysis showed a strong association of LLDR with phenotype A, C=0.915 (0.914-0.915), p<10^-4, optimum cutpoint <0.398, sensitivity 72%, specificity 95%; and phenotype B, C= 0.923 (0.923-0.924), p<10^-4 optimum cutpoint >0.905, sensitivity 81%, specificity 86%. There was also a positive correlation between LLDR and LDL Max Time (R²=0.802). Conclusion: LLDR is a convenient, easily calculated, and continuous variable that is strongly associated with LDL density phenotype and LDL Max Time. Further research is needed to investigate the relationship between lipoprotein density and size, and whether LLDR provides more cardiovascular risk discrimination than LDL density phenotype.