Performance of LDL-C calculated with Martin's formula compared to the Friedewald equation in familial combined hyperlipidemia

Abstract

Background and aimsA novel method to estimate low density lipoprotein cholesterol (LDL-C) has been proposed by Martin et al. This may permit a more accurate estimation of cardiovascular risk, however, external validation is needed. Here, the performance of LDL-C using this new method (LDL-N) is compared with LDL-C estimated with Friedewald equation (LDL-F) in familial combined hyperlipidemia (FCHL), a common primary dyslipidemia in which apolipoprotein B containing particle composition is abnormal and interferes with LDL-C estimation. MethodsA total of 410 FCHL subjects were included. LDL-C was estimated with both the Friedewald equation (LDL-F) and the novel formula (LDL-N). Apolipoprotein B levels and non- HDL-C were recorded. The correlation and concordance between LDL-F and LDL-N and both Apolipoprotein B and non-HDL-C levels were calculated. Analysis stratifying for triglyceride tertiles and FCHL lipid phenotypes was also carried out. ResultsThe correlations between LDL-N and Apo B and non-HDL-C were ρ = 0.777 (95%CI 0.718–0.825) and ρ = 0.735 (95%CI 0.648–0.816), respectively. The corresponding correlations for LDL-F were ρ = 0.551(95%CI 0.454–0.637) and ρ = 0.394 (95%CI 0.253–0.537), respectively. In mixed dyslipidemia or isolated hypertriglyceridemia, these correlations were significantly better using LDL-N. With respect to concordance, LDL-N performed significantly better than LDL-F when considering apoB <90 mg/dL (κLDL-N = 0.495 vs. κLDL-F = 0.165) and non-HDL-C <130 (κLDL-N = 0.724 vs. κLDL-F = 0.253). ConclusionsIn FCHL, LDL-C estimation using Martin's formula showed greater correlation and concordance with non-HDL-C and Apo B compared with the Friedewald equation.