THE NATURE OF THE SUBSTANCES IN DIETARY FAT AFFECTING THE LEVEL OF PLASMA CHOLESTEROL IN HUMANS

Abstract
Three dietary experiments have been performed in which 52, 48, and 38 male students and a few members of staff participated as experimental subjects. In the first experiment 52 subjects ingested for 8 days a diet in which butter provided 60% of calories. During the next 8 days subgroups were given rations supplemented with a-tocopherol or [beta]-sitosterol, substances known to be present in corn oil in significant amounts. The [alpha]-tocopherol had no effect but the [beta]-sitosterol caused a highly significant decrease in plasma cholesterol. The same conditions were used for the first 8 days of the second experiment in which 48 subjects participated. The subgroups were given diets in which 30% of calories were supplied by butter and 30% by various corn oil fractions obtained by vacuum distillation. All groups showed highly significant decreases in plasma cholesterol and there did not appear to be any effective fractionation of the plasma cholesterol depressant factor in the corn oil, although the largest drop was obtained in the case of the most volatile fraction derived from the corn oil. This preparation contained most of the unsaponifiable material and had a slightly lower iodine number than the other fractions. In the third experiment, 38 subjects ingested for 8 days a diet providing 60% of calories from corn oil. Subgroups were transferred to diets supplying 40% of calories from butter fractions obtained by vacuum distillation. All groups showed highly significant increases, the greatest increase being noted in the case of the most volatile fraction, which contained most of the unsaponifiable material. It has been postulated that the potent plasma cholesterol elevating action of butterfat is at least partly dependent upon the presence of the unsaponifiable fraction. The identity of this factor and whether it acts alone or in conjunction with certain types of fatty acid residues are questions that remain to be answered.