During the decades after World War II the main risk factors for atherosclerotic cardiovascular diseases (CVD) were identified through numerous epidemiologial follow-up studies.1 Many other observations from animal studies to ecological comparisons have supported the hypothesis that these risk factors are, indeed, causal. In the classical research chain, the “final proof” would come from randomized trials.
Since risk factors are closely related to certain behaviors, in particular dietary habits, smoking and physical activity, such a study is, however, very difficult in practice. Double-blind studies are not feasible. It is not feasible to expect that large numbers of randomized individuals would change their behaviors over many years, in the absence of other changes, while randomized control individuals from the same neighborhood would not make such changes.