Comparison of Sampson and extended Martin/Hopkins methods of low-density lipoprotein cholesterol calculations with direct measurement in pediatric patients with hypertriglyceridemia.

Abstract

The Friedewald equation is the commonly used method of low-density lipoprotein cholesterol (LDL-C) calculation, requiring reflex to direct LDL-C measurement when triglycerides (TG) ≥ 400mg/dL. Recently formulated Sampson and extended Martin/Hopkins methods have been validated with TG up to 800mg/dL and thus have the potential to replace direct LDL-C measurement. Given the growing prevalence of childhood dyslipidemia, the objective of this study was to compare Sampson and extended Martin/Hopkins methods of LDL-C calculation with the direct measurement in a pediatric cohort with 400≤ TG ≤ 799mg/dL. This study retrieved standard lipid panels and corresponding direct LDL-C measurements of 131 patients with 400≤ TG ≤ 799mg/dL from a pediatric population. Following the application of Sampson and extended Martin/Hopkins calculations, calculated values were compared with direct LDL-C measurements using ordinary least squares linear regression analysis and bias plotting. Both Sampson and extended Martin/Hopkins LDL-C calculations exhibited a strong correlation with the direct measurements (Pearson r=0.89) in patients with 400≤ TG ≤ 800mg/dL. Average percentages of bias of 45% and 21% were found between the direct LDL-C measurements and Sampson or extended Martin/Hopkins calculations, respectively. Both Sampson and extended Martin/Hopkins calculations are applicable as clinical alternatives of direct LDL-C measurement in pediatric patients given 400≤ TG ≤ 799mg/dL.