As far as the interrelations of the lipids were concerned a definite correlation was found between the development of atherosclerosis and an increase of serum cholesterol that was out of all proportion to the increase of serum lipid phosphorus and neutral fat. When these last two lipid constituents rose almost parallel with the serum cholesterol (as they did in some alloxan-diabetic rabbits), then the development of atherosclerosis was inhibited. This correlation was independent of the diabetic state, per se. It appeared likely that the marked elevation of serum neutral fat and lipid phosphorus in the diabetic animals was due to mobilization of body fat because of the disturbed carbohydrate metabolism. Because of their hydrophilic and emulsifying properties, it was thought probable that the elevation of the phospholipids was the important factor responsible for the stability of serum cholesterol. That neutral fat played a role, however, could not be denied.

In normal rabbit sera, as we have previously shown, only small proportions of the lipid phosphorus and cholesterol are "readily extractable" (i.e., unattached or only loosely attached to protein). On the other hand, in every case in which the serum lipids were elevated, the greater proportion of the lipid phosphorus and cholesterol was "readily extractable," irrespective of whether atherosclerosis developed or was inhibited.

Analysis of the lipid content of the aorta of rabbits not fed cholesterol, whether diabetic or non-diabetic, and from alloxan-diabetic rabbits fed cholesterol but protected from the development of atherosclerosis, showed that there was no significant difference in lipid content or composition among the animals of these groups. When atherosclerosis developed following cholesterol feeding, the lipid composition of the aortas was essentially the same in both control and diabetic animals. The deposited lipid consisted predominantly of cholesterol with small and fairly constant proportions of other lipids that did not vary significantly regardless of the quantities of these other lipids present in the circulating blood. In the less severe lesions the proportion of ester cholesterol was greater than that of free cholesterol, but in advanced lesions the reverse was true.

The following conclusions are drawn concerning the pathogenesis of experimental cholesterol atherosclerosis in the rabbit:

1. Instability of cholesterol in the blood rather than hypercholesterolemia, per se, is the general condition responsible for the deposition of this substance in the arterial walls.

2. Of the two factors considered to be responsible for the stability of the lipids in the blood, the interrelations of the lipids appear to be more important than their relation to the serum proteins, at least in so far as the development of experimental cholesterol atherosclerosis is concerned.

The importance of these conclusions in relation to the pathogenesis of human atherosclerosis is discussed.