Comparison of ultracentrifugation and a precipitation method for high-density lipoprotein cholesterol quantitation in insulin-dependent diabetic patients

We compared sodium phosphotungstic acid and magnesium chloride precipitation method for high-density lipoprotein (HDL) cholesterol quantitation with the ultracentrifugation method in 64 insulin-dependent diabetic patients with plasma triglyceride less than 3 mmol/l. The cholesterol content of HDL after precipitation of very-low-density lipoprotein (VLDL) and low-density lipoprotein (LDL) was 86% +/- 3% of the cholesterol content of HDL (q greater than 1.063) determined after ultracentrifugation at q = 1.063 (1.33 +/- 0.05 mmol/l vs 1.55 +/- 0.06 mmol/l; p less than 0.001). HDL cholesterol determined after precipitation closely correlated to HDL cholesterol determined after ultracentrifugation (r = 0.97; p less than 0.001). The absolute difference between the HDL cholesterol values obtained by the two methods was correlated to HDL cholesterol (ultracentrifugation) (r = 0.75; p less than 0.001), but it was not correlated to VLDL cholesterol, LDL cholesterol, triglyceride, HbA1c, blood glucose or serum albumin. LDL cholesterol calculated by use of Friedewald's formula was 108% +/- 4% of the cholesterol content of LDL (q = 1.019 to 1.063), determined after ultracentrifugation, but the calculated and the ultracentrifugally determined LDL cholesterol values were closely correlated (r = 0.98; p less than 0.001). These results suggest that during sodium phosphotungstic acid and magnesium chloride precipitation of plasma from diabetic patients, a constant fraction of HDL cholesterol is co-precipitated, resulting in a systematic difference in HDL cholesterol quantitation when compared with the ultracentrifugation method.