Abstract
We estimated HDL-cholesterol (HDL-C) in control human sera (n=17), pooled human sera with different HDL-C levels (n=5) and fresh human sera (n=12) by three precipitation methods based on different principles: phosphotungstate MgCl2 method (4 kits), polyethylene glycol method (3 kits) and isoelectric point method (1 kit) and the values were compared by HPLC method. All HDL-C values measured by precipitation methods were significantly lower than those by the HPLC method. Especially, the values measured by the isoelectric point method and phosphotungstate MgC12 method were remarkably lower than those by HPLC method. The polyethylene glycol method, however, showed values closest to those by the HPLC method. The most remarkable differences between values measured by precipitation method and HPLC method were noted in the control sera group and the largest inter-kit deviation was also noted in this group. The recovery rate of HDL-C in the supernatant fractions was the lowest and the deviation was the largest in this group. Analysis by the HPLC method of the supernatants revealed that there was no contamination from lipoproteins other than HDL in any case, although the distribution in sizes of HDL were often different from those of original sera. In conclusion, we speculated that HDL-C values lower and more variable than those demonstrated by HPLC were shown by precipitation methods due to the possible of precipitation of a part of the HDL by the reagents as well as the differences in reactivity of reagents to aggregated lipoprotein particles and apo-E rich HDL in the samples.
