Objectives: Several pediatric advisory groups have recommended selective screening for dyslipidemia in children. Low-density lipoprotein cholesterol (LDL-C) is measured clinically with the Friedewald calculation in fasting samples. Nonfasting measurement of LDL-C would be clinically useful in children.
Design and methods: In the present study, we examine the performance of two surfactant-based direct LDL-C assays in paired samples, fasting and nonfasting, from 100 children.
Results: LDL-C in the fasting state was significantly lower with the Friedewald estimation: 2.43 +/- 0. 61 mmol/L, N-geneous (Genzyme Corp.) direct LDL-C: 2.30 +/- 0.59 mmol/L, and Roche (Roche Diagnostics) direct LDL-C: 2.32 +/- 0.57 mmol/L than with ultracentrifugation-dextran-sulfate-Mg(2+) precipitation (UC-DS): 2.47 +/- 0.64 mmol/L. Moreover, there was increased negative bias using nonfasting samples with N-geneous: 2. 25 +/- 0.56 mmol/L and Roche: 2.26 +/- 0.56 mmol/L compared with fasting UC-DS. Correlation with US-DS was highest for Friedewald (r = 0.974) and fasting N-geneous (r = 0.973), and lowest with nonfasting N-geneous (r = 0.849) and Roche in fasting (r = 0.891) and nonfasting samples (r = 0.747). The sensitivity at LDL-C concentration of 2.85 mmol/L for the two direct methods when either fasting or nonfasting samples were used, was lower than that obtained with Friedewald.
Conclusion: Overall, these direct LDL-C assays demonstrated limited utility in screening children but may be useful in the management of lipid disorders.