HDL Particle Measurement: Comparison of 5 Methods.

BACKGROUND: HDL cell cholesterol efflux capacity has been documented as superior to HDL cholesterol (HDL-C) in predicting cardiovascular disease risk. HDL functions relate to its composition. Compositional assays are easier to perform and standardize than functional tests and are more practical for routine testing. Our goal was to compare measurements of HDL particles by 5 different separation methods. METHODS: HDL subfractions were measured in 98 samples using vertical auto profiling (VAP), ion mobility (IM), nuclear magnetic resonance (NMR), native 2-dimensional gel electrophoresis (2D-PAGE), and pre-ß1-ELISA. VAP measured cholesterol in large HDL2 and small HDL3; IM measured particle number directly in large, intermediate, and small HDL particles; NMR measured lipid signals in large, medium, and small HDL; 2D-PAGE measured apolipoprotein (apo) A-I in large (α1), medium (α2), small (α3-4), and pre-ß1 HDL particles; and ELISA measured apoA-I in pre-ß1-HDL. The data were normalized and compared using Passing-Bablok, Lin concordance, and Bland-Altman plot analyses. RESULTS: With decreasing HDL-C concentration, NMR measured a gradually lower percentage of large HDL, compared with IM, VAP, and 2D-PAGE. In the lowest HDL-C tertile, NMR measured 8% of large HDL, compared with IM, 22%; VAP, 20%; and 2D-PAGE, 18%. There was strong discordance between 2D-PAGE and NMR in measuring medium HDL (R2 = 0.356; rc = 0.042) and small HDL (R2 = 0.376; rc = 0.040). The 2D-PAGE assay measured a significantly higher apoA-I concentration in pre-ß1-HDL than the pre-ß1-ELISA (9.8 vs 1.6 mg/dL; R2 = 0.246; rc = 0.130). CONCLUSIONS: NMR agreed poorly with the other methods in measuring large HDL, particularly in low HDL-C individuals. Similarly, there was strong discordance in pre-ß1-HDL measurements between the ELISA and 2D-PAGE assays.