Variability of plasma HDL subclass concentrations in men and women over time

Plasma HDL subclasses were examined by gradient gel electrophoresis in repeated samples to assess variability over time. Absorbance of the protein stain was used as an index of mass concentrations at 0.01-nm intervals within five HDL subclasses: HDL3c (7.2 to 7.8 nm), HDL3b (7.8 to 8.2 nm), HDL3a (8.2 to 8.8 nm), HDL2a (8.8 to 9.7 nm), and HDL2b (9.7 to 12 nm). Three separate longitudinal studies of men showed that repeated samples of HDL over time were correlated most strongly within HDL2b, somewhat less within HDL2a, and more weakly within HDL3a, HDL3b, and HDL3c. As in men, repeated samples in women from two studies were significantly correlated within the HDL2b, HDL2a, and HDL3b intervals. Plasma HDL2b levels were significantly more stable in men than in women. Although the variability of HDL subclass measurements includes both methodological and physiological sources, differences in laboratory measurement error do not appear to explain the differences in correlations among subclasses. Specifically, analysis of 288 replications from frozen aliquots suggested that laboratory error had the least effect on correlations involving HDL3 subclasses and only slightly greater effect on correlations involving HDL2 subclasses. Our results suggest that for plasma sampled over time, the stability of HDL subclass levels increases with particle size. Prior reports of subclass-specific correlations between HDL and other variables (eg, diet, exercise, and other lipids) are unlikely to be artifacts of laboratory precision but could arise from subclass differences in variability that are physiological.