We compared six precipitation methods for high-density-lipoprotein cholesterol quantitation with an ultracentrifugation method, the accuracy of which was improved by correcting for 4% manipulative loss in the d greater than 1.063 fractions. For purposes of comparison, the apoprotein B-associated cholesterol (average 56 mg/L) measured by immunoassay in the d greater than 1.063 fractions was subtracted. In 65 plasma samples from men, women, and children, a heparin-Mn2+ procedure with Mn2+ at 46

more »

... ol/L produced results slightly higher (+16 mg/L), while results with Mn2+ at 92 mmol/L averaged slightly lower (-8 mg/L) than the comparison ultracentrifuge method. Results that were about 5% low were obtained by the dextran sulfate 500-Mg2+ (-25 mg/L) and phosphotungstate-Mg2+ (-31 mg/L) methods. A heparin-Ca2+ method produced results 10% high (+58 mg/L). Results by a polyethylene glycol-6000 precipitation method were 12% low (-64 mg/L). Precision was better with the two heparin-Mn2+ and the dextran sulfate 500-Mg2+ procedures, with CVs of 4%, intermediate with phosphotungstate-Mg2+ and polyethylene glycol-6000 (CV 6-7%), and poorest with heparin-Ca2+ (CV 10%). Precipitation by phosphotungate-Mg2+ appeared more sensitive to reagent concentration and temperature variations than either the heparin-Mn2+ or dextran sulfate 500-Mg2+ methods. We conclude that these precipitation methods are not equivalent, but give rise to significant systematic differences in high-density-lipoprotein cholesterol quantitation.