George Mann’s Editorial on ‘Diet-Heart: End of an Era?’ A Reply
The title and the introduction to George Mann’s intemperate New England Journal editorial leave no doubt where he stood on the diet-heart relationship in the 1970s. He has never budged from this stand. And myriad strikes to the heart of “Diet-Heart” follow his opening salvo.
He begins with a dour opinion that, “A generation of research on the diet-heart question has ended in disarray” (Mann 1977, 644). In fact, the diet-heart hypothesis and multiple studies on the question were hardly in disarray in 1977. Clinical studies were corroborative of laboratory and population studies; the laboratory evidence had established by direct experiment a clear and mathematical relationship between dietary fatty acid saturation and serum lipoprotein levels. And the first generation of prospective epidemiological studies had come to fruition in the early 1970s, indicating an exponential relationship between increasing serum cholesterol levels and subsequent coronary event rates among healthy individuals and whole populations and cultures.
George Mann, in those days, carried the imprimatur of the Harvard School of Public Health’s Department of Nutrition as well as a central place as investigator in the Framingham Study. But, as this editorial illustrates, he typically ignored the tenets of good diplomacy by speaking loudly and carrying a big stick, and usually distinguished poorly between personal attack on colleagues and scientific criticism. Moreover, he was no epidemiologist, despite some interesting field studies he conducted among Mau Mau herders with high-fat diets and low serum levels.
In the second sentence of the editorial he states that, despite wide official recommendations for dietary change “…the [coronary] epidemic continues unabated, cholesteremia in the population is unchanged, and clinicians are unconvinced of efficacy.” To the contrary, the age-specific, coronary heart disease death rate had been in steady decline in the U.S., on the order of 3 per cent per year, since the late 1960s. In addition, average U.S. cholesterol levels had been documented to decline in national surveys. And most clinicians knowledgeable in the field, including those associated with the American Heart Association’s recommendations were convinced of a strong diet-heart relationship, if not all were sanguine about the effectiveness of dietary strategies among their patients. The disarray over diet-heart in 1977 was mainly, it seems, in Dr. Mann’s mindset.
Mann quotes one of the originators of the diet-heart hypothesis, E.H. (Pete) Ahrens, Jr. of the Rockefeller Institute that: “It is not proven that dietary modification can prevent arteriosclerotic disease in man,” taken from the latter’s current Congressional testimony. This was quite true, but Ahrens’s own NIH Panel recommendation for a definitive trial did not convince the Heart Institute that, though a definitive diet-heart trial was required for proof, it was feasible to conduct.
Mann found most disturbing of all: “the vibrant certainty of scientists claiming to be authorities on these matters…” (ibid., 644). Quoting the British biostatistician, Major Greenwood, along with Voltaire, about the uncommonness of common sense, he concludes his introduction by writing: “These limitations [presumably of common sense] must explain the lost generation of misguided and fruitless preoccupation with the diet-heart hypothesis” (ibid.).
Mann attributes the beginning of the diet-heart era to John Gofman and his Index of Atherogenesis in 1950. He cites the Cooperative Lipoprotein Study Group on which he served, that measurement of Gofman’s lipoprotein (LP) particles (LDL, HDL, VLDL, etc.) are no more useful than [total serum] cholesterol, and claims erroneously that: “. . subsequent trials have failed to show more than a trivial effect of diet on cholesterolemia. . ” (ibid.). In fact, by that time, Keys, Grande, and Anderson in Minnesota, and Hegsted and colleagues at Harvard, among others, had firmly established the quantitative predictive relationship of diet change to serum cholesterol change.
Subsequent inflation of the diet-heart hypothesis Mann attributes to Ancel Keys’s 1953 Mt. Sinai address about the ecologic correlation of diet fat and coronary disease. This report he derogates as “now a classroom demonstration [of naïve interpretation]… Nevertheless, in a few years some combination of the urgent needs of health agencies, oil-food companies and ambitious scientists, had transformed that fragile hypothesis into treatment dogma . . . A low-fat, low-cholesterol diet became as automatic in their treatment advice as a polite goodbye” (ibid.).
Mann proceeds to attack the 1967 lipid phenotypic classification of Fredrickson, Levy and Lees at NIH as “enormously popular and profitable although…poorly reproducible, without much genetic reality, and no more helpful in management than a quick, cheap and accurate measurement of blood cholesterol” (ibid.). He finds irony in new findings that high density lipoprotein (HDL) is inversely correlated to coronary disease incidence, in that Barr’s similar claims for HDL back in 1951 had been thrown out like the baby, while “keeping the bathwater” (ibid.).
The series of dietary recommendations from the American Heart Association from the 1950s, culminating in the influential 1970 Inter-Society Report, were lambasted editorially. Mann labeled the scientists involved “committee men” and claimed they had quickly fallen in line (ibid., 645). “The scientific issue was settled by majority votes. Galileo would have flinched. The dietary dogma was a money-maker for segments of the food industry, a fund raiser for the Heart Association, and busy work for thousands of fat chemists” (ibid.). With himself as a prime victim of the established view, he claims: “To be a dissenter was to be unfunded because the peer-review system rewards conformity and excludes criticism” (ibid.).
Mann then shifts from editorial style to a literature review of the “poor lines of evidence” about the diet-heart picture, which led him to conclude that the hypothesis is wrong. For example, among the dietary interviews he supervised in the early years of the Framingham Study no relationship was found between an individual’s 24-hour diet recall data and blood cholesterol level. He fails to understand its probable origins which no way negate the hypothesis. He further concludes from crude death rates that U.S. mortality trends since 1950 “do not support the argument that the expensive dietary propaganda has had an effect on clinical events” (ibid. 645), for which he fails to note that age-specific coronary death rates, in fact, started falling in the late 1960s and persisted to fall until the time of his writing, 1977. And he appears to misinterpret his own citation of the declining level of average serum cholesterol levels from surveys of the U.S. population.
The third line of negative evidence Mann finds from the Primary and Secondary Preventive Trials to date, but he does not discuss the reasons for the failures in the design weaknesses of these trials. He concludes properly, for 1977, that no dietary trial had been shown effective for the prevention or treatment of coronary heart disease, but fails to note that, in fact, the needed definitive primary prevention coronary trial with dietary cholesterol lowering was never carried out.
In his fourth line of evidence, the cholesterol-lowering drug trials, he correctly concludes that as of 1977, the Coronary Drug Project and the WHO Clofibrate Study had not demonstrated clear effect on coronary or total deaths, and concludes that “there is no safe and efficacious drug known for the management of cholesterolemia” (ibid., 646), but fails again to cite that their designs were inadequate to show effects on all-cause deaths. He uses Framingham data to illustrate the predictive advantage of cholesterol level for young men but “the futility of fussing with risk factors in men past 55 years of age” (ibid.).
This was a premature conclusion, at best, as statin trials were later to show. And he rales against screening for hypertension and elevated blood cholesterol in youth “until effective and acceptable treatments are available…” (ibid., 647), a common academic device: take no action based on best overall evidence until specific experimental proof is available, which, unfortunately, is infeasible to obtain.
Mann reviews cursorily and superficially the different coronary picture in primitive populations along with individual differences within affluent society, which provide evidence that elevated serum cholesterol is contributory to mass coronary disease and that its reduction should make a difference. He is, however, prescient in noting that trans fatty acids in diet are importantly hypercholesterolemic, predicting their significance for the public health (ibid., 648).
Finally, Mann launches into his favorite causal theory about coronary disease without, however, as for diet-heart, critiquing it systematically: “The most impressive array of epidemiological evidence suggests that fit and active people are spared the complication of atherosclerosis” (ibid.), based on his idea that the enlargement of their coronary vessels compensates for increasing atherosclerosis.
Mann’s satire turns characteristically personal and biting at the end: “In the meantime, such a distinguished post-coronary scientific savant as Dr. Irvine H. Page likes to quip that he still follows a low-fat, low-cholesterol diet because he has no intention of being the smartest man in the cemetery. The evidence indicates he will not be that” (ibid., 649).
This New England Journal editorial created a furor in the cardiovascular scientific community and was gleefully reprinted in the hundreds of thousands of copies and circulated by commercial interests as well as being cited by all naysayers about Diet-Heart of the time. It also evoked a flurry of letters to the editor with dissections of the author’s reasoning fallacies, factual errors, and inappropriate sarcasm. Today, the world is peopled by very few scientists who challenge the general relationship demonstrated between diet, blood lipoproteins, and coronary disease risk, based on evidence accumulated since Mann’s editorial that effective cholesterol-lowering (LDL) by whatever means reduces risk of heart attack and arrests progression of atherosclerosis. But many of Mann’s arguments, real and specious, are recalled today by on-going sensational journalistic attacks on Ancel Keys’s work and on “Diet-Heart” itself. Indeed, the complexities and mechanisms of the diet-heart relation continue to evolve, inspiring interest and provoking ever-new questions. (Henry Blackburn)
1. Mann, G.V. 1977. Diet-heart. End of an era. New England Journal of Medicine 297: 644-49.