Assessment of the validity of the frequently used lipid indices for predicting LDL peak particle diameter in a large cohort of 1955 normal and dyslipidemic subjects

Objectives: To assess the relationship between commonly used lipid indices and LDL peak particle diameter (LDL-PPD) in a large cohort of 1955 subjects. Design and methods: Four statistical methods were used for comparison: correlation, concordance analysis, kappa statistics and receiver operating characteristic curve (ROC) analysis. Results: Plasma triglyceride (TG) levels, TG/HDL-C, LDL-C/apoB, total cholesterol (TC)/TG, LDL-C/TG, and TG/apoB ratios were best correlated with LDL-PPD but none of these ratios accounted for more than 45% of the variation in LDL-PPD. Moreover, across the range of the lipid indices and LDL-PPD quintiles, just under 40% of the values were concordant, with kappas varying between 0.20 and 0.25. Finally, plasma TG levels and the lipid ratios yielded areas under the ROC curve between 0.78 and 0.80. Conclusions: The present study does not Support the concept that plasma TG levels and the commonly used lipid indices may be considered as adequate surrogates for the small dense LDL phenotype. Our data also indicate that various lipid indices are not superior to plasma TG levels alone as a clinical tool for prediction of the small dense LDL phenotype. (C) 2009 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.