The controversy about cholesterol does not stem from ideological differences only. For many years, consensus committees and other evangelists have ignored or misquoted findings unsupportive of the idea that dietary therapy can reduce risk for coronary disease [1]. I wish to correct one of LaRosa's arguments, one which "most dramatically" argues for this idea, to use his own words [2]. The increased coronary risk of Japanese migrants was not associated with dietary fat intake or blood cholesterol level but with cultural factors. In fact, Japanese persons who had become accommodated to a U.S. lifestyle but preferred Japanese food had coronary disease twice as often as did Japanese persons who adhered to Japanese traditions but preferred U.S. food [3].

This example is not an isolated error; many similar misquotations can be found in the reviews used as evidence by the National Cholesterol Education Program [1]. The influence on medical thinking of such writing is best compared with the effect of a computer virus. It is therefore laudable that the American College of Physicians has presented its own guidelines for cholesterol screening, based on evidence rather than wishful thinking.

But even these prudent recommendations can be questioned. It is not self-evident that cholesterol reduction is effective. Bias may have influenced previous trials because most were open, either by design or because side effects or laboratory values showed group affiliation, because only soft end points were affected, and because the degree of cholesterol lowering was unrelated to outcome in and between trials. Absence of dose-response defies the elementary laws of biology, and the success of the recent statin trials must therefore be explained otherwise. Statins inhibit the synthesis of several compounds, some of which may be more injurious to vessels than cholesterol [4, 5]. In accordance, simvastatin prevented cardiovascular diseases in experimental animals independent of cholesterol levels [4, 5]. Thus, the results of the statin trials cannot be extrapolated to cholesterol lowering in general but may hopefully initiate research in more fruitful fields.