Effect of HDL disk and LDL dimer presence on lipoprotein particle number determination and subclassification

Abstract High-density and low-density lipoproteins (HDL and LDL) are established analytical targets for diagnosis and risk stratification of numerous chronic diseases. This study investigates potential sources of bias in lipoprotein particle counting (HDL-P and LDL-P), focusing on the most atheroprotective small-HDL and most pro-atherogenic small-LDL. Plasma samples were fractionated using asymmetric-flow field-flow fractionation (AF4), coupled with hydrodynamic size measurement and comprehensive liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis of each fraction. Concentration-size profiles were deconvoluted into 10 HDL and 8 LDL Gaussian subspecies. Molecular volume ratios were used to evaluate proposed particle models, providing evidence for the presence of s-HDL disk and s-LDL dimers, as sources of bias in calculated HDL-P and LDL-P when spherical particle geometry is assumed. Matching apoA1/HDL-P and apoB/LDL-P to consensus values enabled correction of mass diameters (k*d m ), with k≈1.2 for s-HDL disks, and k≈0.8 for s-LDL dimers. These corrections resulted in closer agreements of the calculated HDL-P and LDL-P values with those reported by nuclear magnetic resonance spectroscopy (n = 666). After corrections, average AF4-LC-MS/MS versus NMR differences decreased from 20 to 4.9% for HDL-P and from −47% to −5.7% for LDL-P; x-y correlations improved from 0.83 to 0.85 and 0.51 to 1.03, respectively; and both platforms showed a dominant presence of s-HDL and s-LDL subclasses. Overall, the results support standardization of clinical methodologies for lipoprotein subclass measurement around apoA1 and apoB, and harmonization of HDL and LDL subclass definitions based on composition and structural characteristics rather than strictly enforced size cutoffs. Graphical abstract