Over the past twenty to thirty years, a welter of data, sometimes conflicting, has accumulated relating to the epidemiology of coronary heart disease (CHD). This information has been collected by a variety of methods including the analysis of international statistics, international field studies (both clinical and pathologic), transectional national studies of racial, ethnic, occupational and economic groups, and longitudinal studies of living and dietary habits with their associated morbidity and mortality patterns.
INCIDENCE OF CORONARY HEART DISEASE IN POOR VERSUS WEALTHY NATIONSDespite a number of major and minor discrepancies which have helped to confuse the issue, these studies have clearly and irrefutably established that among the peoples of poor and/or technologically primitive countries there is very little clinical or pathologic CHD whereas among the peoples of wealthy, technologically advanced nations there is a very high incidence (1). Furthermore, the incidence of CHD appears to be roughly proportional to per capita income and leisure time (l), the countries with the most luxurious way of life having the most CHD. The United States ranks first on both counts. In the poorer countries, the incidence of CHD in the wealthy, sedentary upper classes is comparable to that seen in wealthier nations, despite its rarity in the total population (1-5).Ancel Keys has been very active in this field. With numerous co-workers, he has carried out studies in the United States, Spain (6), South Africa (7), Italy (3), and Japan (4, 8). Other groups have made comparative studies in Africa (9-11), Central America (2) , Israel (12, 13), Korea, India and various other countries. In each instance, these studies have confirmed the relationship between a high standard of living and a high incidence of CHD on the one hand, and between poverty or a primitive way of life and a minimal incidence of CHD on the other.
LONGITUDINAL STUDIESFurther evidence relating a high standard of living to increased CHD morbidity and mortality comes from an appraisal of longitudinal statistics. In