Endocrine Abstracts

JOINT3542

Introduction: Most clinical laboratories use the Friedewald equation to account for VLDL-cholesterol and subsequently calculate LDL-cholesterol. This equation has known limitations at high triglyceride levels or low LDL-cholesterol levels. Alternative equations, such as the Martin-Hopkins and Sampson-NIH equations, have been proposed. This study aims to compare LDL-cholesterol estimates derived from these three equations with directly measured LDL-cholesterol and to assess whether the choice of equation affects patients’ clinical management.

Methods: In this retrospective, monocentric study, all LDL-cholesterol samples analyzed at Amiens University Hospital between august 1^st, 2022 and July 31^st, 2023 were included (n = 7895). LDL-cholesterol levels calculated by the three equations were compared with measured LDL values using Pearson correlation coefficient. Comparisons were stratified by total triglyceride (TG) (0-400 mg/dL, 400-800 mg/dL and >800 mg/dL) and LDL-cholesterol levels (below or above 70 mg/dL). Subsequently, random samples were selected from each category (TG 0-400 mg/dL, TG 400-800 mg/dL and LDL-cholesterol <70 mg/dL), cardiovascular risk was estimated for each individual and the misclassification rates attributable to each equation were determined.

Results: In the TG 0-400 mg/dL category, all three equations yielded equivalent correlation coefficients (’r’) around 0.95 but the Sampson equation showed the highest concordance rate (24.85% vs 22,01% for Martin-Hopkins 20.45% for Friedewald). In the TG 400-800 mg/dL category, the correlation coefficients (’r’) was around 0.83 and both the Martin-Hopkins and Sampson-NIH equations had the highest concordance rates (14.51% and 13.70% vs 3.22% for Friedewald). In the LDL-cholesterol <70 mg/dL category, The Martin-Hopkins and Sampson-NIH equations had correlation coefficients (’r’) of 0.70 and 0.71, respectively, compared to 0.61 for Friedewald; concordance rates were 27.55% for the Martin-Hopkins, 26.78% for the Sampson-NIH and vs 24.25% for the Friedewald equation. Among people with TG 0-400 mg/dL, the proportion of misclassified individuals was 4,4% for the Sampson-NIH, 4,9% for the Martin-Hopkins and 6,4% for the Friedewald equation. Among people with TG 400-800 mg/dL, the proportion of misclassified individuals was 7,4% for the Sampson-NIH, 3,7% for the Martin-Hopkins and 14,8% for the Friedewald equation. The respective proportions were 14,7%, 14,7% and 17,6% among individuals with LDL-cholesterol <70mg/dL.

Conclusion: This study demonstrates that the three methods perform very well in people with TG 0-400 mg/dl, relatively good in people with TG 400-800 mg/dL and less well in people with LDL-cholesterol < 70 mg/dL. However, the Sampson and Martin-Hopkins equations are less prone to therapeutic misclassification errors than the Friedewald equation.