University of Minnesota

Famous Polemics on Diet-Heart Theory

The Seven Countries Study was apparently motivated, in part, from the 1955 expert meeting on atherosclerosis at the World Health Organization in Geneva. There Ancel Keys presented his diet-lipid-heart disease hypothesis with his usual confidence and bluntness. He was taken aback by the vigorous challenge from Sir George Pickering who asked him, in this approximation: “If you would be so kind, Professor Keys, what do you consider the single best piece of evidence to support your diet-heart idea?”

Keys fell for the trick by this Oxford debating expert and answered the question with a piece of evidence. The assembled experts were quickly able to diminish the significance of that bit of evidence. Keys had lost his opportunity to present the totality of evidence in a rational argument.

The next year an article by Yerushalmy and Hilleboe appeared in what was a direct, and for Keys, painful sequel to the Geneva meeting. We learn from their footnote that: “The authors became interested in this question (diet vs. national mortality data) during a meeting of a WHO study group on atherosclerosis and ischemic heart disease in Geneva, Switzerland, November 1955. At that meeting statements were made about the association between heart disease and mortality and fat in the diet. As it later developed, these were based on a few selected countries and were of questionable validity. On their return to the United States, the authors reviewed the available data carefully, and the results indicated that the subject required further study.”

I recall well the reaction in Minnesota and in epidemiological circles to this article and the challenge it represented for proponents of the diet-heart idea. The article seemed to magnetize the pre-existing polar elements of the scientific community: the lines of force were drawn. Meanwhile, proponents were already reviewing data from several North American cohort studies showing the short-term predictive relationship of serum cholesterol level in health to risk of subsequent coronary events. Moreover, several investigators were involved in feeding experiments of the effects of fatty acid composition of diet on serum cholesterol level. Keys, meanwhile, had made new comparisons of diet and blood lipids among casual samples of men in South Africa and Japan, to test the relation in contrasting populations.

The tone of the Yerushalmy-Hilleboe “methodologic note” was authoritative, pedantic, and patronizing, beginning with: “The evidence which has been presented for the existence of a relationship between diet and heart disease, however, is, for the most part, derived from indirect methods of study. In these indirect methods the primary unit of observation is the group; in the direct method the primary unit of observation is the individual. . . .It is well known that the indirect method merely suggests that there is an association between the characteristics studied and mortality rates and, further, that no matter how plausible such an association may appear, it is not in itself proof of a cause-effect relationship.”

The authors go on to editorialize and scold the naïve: “The quotation and repetition of the suggestive association soon creates the impression that the relationship is truly valid, and ultimately it acquires status as a supporting link in a chain of presumed proof.” After admitting, however, that such an indirect method of study may be valuable, they again wax pedantic: “unfortunately, however, if you superficially accept and do not properly augment it, the indirect method has many weaknesses. The most serious is the fact that the apparent association often proves to be the result of non-pertinent extraneous factors. Therefore, it is always necessary to probe further, to go beyond the simple, apparent association and to investigate related variables.”

They illustrate this with the figure from Keys’s 1953 Mount Sinai Hospital lecture showing the relationship in six different countries between coronary death rate and the proportion of fat calories in national diet data. And they quote Keys referring to the same evidence, interpreting the data in strong terms: “No other variable in the role of lifestyle besides fat calories in the diet is known which shows anything like such a consistent relationship to the mortality rate from coronary or degenerative heart disease.”

Yerushalmy and Hilleboe again patronize: “Clearly it has become important to determine whether these statements reflect the known or ascertainable facts. But to do so, it is essential to consider questions of methodology, namely the elementary operations and procedures [ed. italics] that are possible, to determine whether the association is valid.” The masters scold the school child and then go on to describe and carry out the “elementary” procedures required to test the validity and specificity of the relationship.

They conclude, “….the evidence from 22 countries for which data are available indicates that the association between the percentage of fat calories available for consumption in the national diets and mortality from arteriosclerotic and degenerative heart disease is not valid; the association is specific neither for dietary fat nor for heart disease mortality. Clearly this tenuous association cannot serve as much support for the hypothesis which implicates fat as an etiologic factor in arteriosclerotic and degenerative heart disease.” Pow!

They then lecture us (and Keys) on proof and association: “Because it is well understood that an association is at best only a small part of the etiologic story, there is a tendency to deal with associations uncritically or even superficially. Investigators must remember that evidence, which is not inherently sound, cannot serve even for partial support. Evidence for an association must be carefully scrutinized and scientifically weighed. As in any other study, the primary data must be evaluated for relevance, accuracy, and internal consistency, and the result must be tested for validity. Whether it is used as a guide by the investigator himself, or offered as a link in a chain of supporting evidence or a hypothesis, the association itself must be valid, that is, specific for the variables on the study. If the association is not specifically related to the variables but rather affects a relationship with extraneous or irrelevant factors, it is worse than useless.”

The critique was devastating. But by this time, two years following the WHO meeting, Ancel Keys was conducting pilot studies for the Seven Countries Study. It would become his effective reply to the questions of specificity and validity and a well-conceived response to these seriously embarrassing criticisms of his ecologic associations and early diet-heart propositions. (Henry Blackburn)


Yerushalmy, J. and Hilleboe, H., July 1957. Fat in the diet and mortality from heart disease; a methodologic note. New York State Journal of Medicine, (14): 2343-2354.