Email updates

Keep up to date with the latest news and content from Cardiovascular Diabetology and BioMed Central.

Open Access Open Badges Original investigation

Non-HDL-cholesterol as valid surrogate to apolipoprotein B100 measurement in diabetes: Discriminant Ratio and unbiased equivalence

Michel P Hermans1*, Frank M Sacks2, Sylvie A Ahn3 and Michel F Rousseau3

Author Affiliations

1 Endocrinology and Nutrition Unit, Université catholique de Louvain, 10 Avenue Hippocrate, Brussels (1200), Belgium

2 Nutrition Department, Harvard School of Public Health, 665 Huntington Avenue, Boston (Ma 0215), USA

3 Division of Cardiology, Université catholique de Louvain, 10 Avenue Hippocrate, Brussels (1200), Belgium

For all author emails, please log on.

Cardiovascular Diabetology 2011, 10:20  doi:10.1186/1475-2840-10-20

Published: 28 February 2011



Apolipoprotein B100 (apoB) is a superior indicator of CV risk than total or LDL-C. Non-HDL-C represents a simple surrogate for apoB in hypertriglyceridemic and/or T2DM patients. ApoB and non-HDL-C show high correlation, although the degree of mutual concordance remains debated in CV risk evaluation.


We used the Discriminant Ratio (DR) methodology to compare the performance of non-HDL-C with that of apoB to rank diabetic patients according to dyslipidemia and to establish the underlying relationship between these variables taking measurement noise and intra-/intersubject variation into account, and to derive an unbiased equivalence equation.


Fasting total C, HDL-C, apoB and triglycerides were measured in 45 diabetic patients. The DR of the underlying between-subject standard deviation (SD) to the within-subject SD was calculated from duplicates. Correlation coefficients between pairs were adjusted to include an estimate of the underlying correlation.


Mean values [day 1 (1SD)] were 143 (36) mg/dl (non-HDL-C) and 98 (24) mg/dl (apoB). The DR's of both parameters were similar (1.76 and 1.83) (p = 0.83). Pearson's product-moment correlation coefficient between tests was very high (0.94), reaching unity (1.00) after attenuation adjustment. The unbiased equation of equivalence relating apoB to non-HDL-C had a slope of 0.65 and an intercept of 6.3 mg/dl.


The discrimination power of non-HDL-C is similar to that of apoB to rank diabetic patients according to atherogenic cholesterol and lipoprotein burden. Since true correlation between variables reached unity, non-HDL-C may provide not only a metabolic surrogate but also a candidate biometrical equivalent to apoB, as non-HDL-C calculation is readily available.