Relationship between low cholesterol and disease. Evidence from epidemiological studies and preventive trials.

This review has addressed the concern that the reduction of the risk of coronary heart disease and other cardiovascular disorders by means of reducing serum cholesterol levels might, at the same time, increase the risk of developing noncardiovascular conditions. This concern is justified because, at face value, prospective epidemiological studies indicate that total mortality and mortality from a number of specific causes of death increase below a threshold value of serum cholesterol (the J-shaped curve phenomenon) and intervention studies to date have mostly failed to show a decrease of total mortality in the experimental group, despite a fall in coronary heart disease mortality (suggesting an increase in noncardiac mortality). Detailed scrutiny of the evidence from prospective studies provides no convincing evidence that low cholesterol levels are associated with an excess of deaths from cancer, except possibly to a minor degree; excess mortality from other causes which account for only a minority of all deaths, are in all likelihood confounded by influences that, in themselves, increase the probability of dying. Preventive trials that have been designed to test whether reduction of serum cholesterol will lower heart disease risk have, so far, provided insufficient numbers of noncardiovascular deaths to test, in addition, whether serum cholesterol lowering is accompanied by an increase in such deaths. According to available evidence, both the apparent excess of noncardiac deaths and total mortality could be attributable to chance. In the light of these findings from prospective studies and preventive trials, as well as the cross-cultural, ecological data that have been summarized, the total evidence provides no justification for depriving either high-risk individuals or populations at high risk as a whole of the benefits of serum cholesterol reduction for preventing heart disease on the grounds that such reduction may increase the risk of noncardiovascular conditions. This latter risk is far from being established and, if present, from all the available evidence is much smaller than the risk of withholding protection from persons and populations exposed to elevated coronary heart disease risk. In addition to the conclusions from epidemiological and intervention studies, a paramount question relates to the biological plausibility of causal connections between low or lowered cholesterol and the hazard of noncardiovascular disorders. Assessment of currently available data provides no evidence for the existence of mechanisms that might explain such a link. Countries in which average serum cholesterol levels are still low, as in Japan, face a different problem.(ABSTRACT TRUNCATED AT 400 WORDS)