The difficulties of feeding free-living research subjects is illustrated once again in a study by Sacks et al reported in this issue (p 640). The object was to observe the effects on blood lipids and blood pressure (BP) from feeding beef to vegetarians. Vegetarians, who usually exhibit low plasma cholesterol levels and BP, should be ideal subjects for the study of the isolated effect of eating beef. The difficulties encountered, however, nicely illustrate the problems produced when dissimilar food items are exchanged in control and experimental diets.
Four recent studies,1-4 in addition to that of Sacks et al, explored the effects of meat or other animal products on plasma cholesterol. The experimental conditions differed among all of the studies. Two studies1,4 other than that of Sacks et al suggest a hypercholesterolemic effect of meat or other animal products, while two do not.2,3 Examination of each study may