BACKGROUND AND AIMS

In epidemiological trials and in clinical practices, it is relevant to have affordable and reliable methods to measure the main lipid cardiovascular risk factors, and in particular low-density lipoprotein cholesterol (LDL-C) plasma level. In this context, we aimed to compare the reliability of the Friedewald's (LDL-Cf) and Sampson's (LDL-Cs) equations with the LDL-value dosed by a validated dosage method (LDL-Cd) in a large cohort of children.

METHODS AND RESULTS

We considered the lipid values of 145 infants, 278 preschoolers, 810 scholar children, and 1372 adolescents (Total N. 2605, 1291 males, 1314 females), with mean total cholesterol (TC) = 169.8 ± 39.7 mg/dL, HDL-Cholesterol = 50.8 ± 12.7 mg/dL, non HDL-Cholesterol = 118.9 ± 35.9 mg/dL, Triglycerides (TG) = 90.3 ± 77.9 mg/dL, LDL-Cd = 106.2 ± 29.9 mg/dL, LDL-Cf = 100.9 ± 33.8 mg/dL, and LDL-Cs = 102.2 ± 33.4 mg/dL. Comparing the distance to the LDL-Cd, Friedewald's equation mildly but significantly underestimated in infants (3.4 ± 5.3 mg/dL), preschoolers (1.5 ± 7.1 mg/dL). Children (1.2 ± 2.2 mg/dL) and adolescents (1.1 ± 5.9 mg/dL) compared to Sampson's equation (all comparisons, p < 0.001).

CONCLUSIONS

Our analysis, being carried out on a large population sample, shows that Sampson's equation is more reliable than Friedewald's one at each considered age class and even for extreme TG values.