In the Beginning
Almost two hundred years ago, the methodology for investigating the occurrence and movement of infectious diseases in populations was born. It happened in London during the cholera epidemic of 1836 with the work of the English physician John Snow (1, p.246). It was an era in which epidemics of infectious diseases caused by yet unidentified “things” were decimating populations throughout Europe.
Dr. Snow studied the eating and living habits of patients who had cholera and neighboring townspeople who did not have cholera for the purpose of identifying the similarities and differences between the groups. It is noteworthy that Dr. Snow examined people who did not get cholera as well as those who did. As a result of the current disregard in medical research of this practice of looking at both sick and non-sick people, important lessons available from subjects who are resistant to disease remain unrecognized.
Dr. Snow ultimately found a strong association between cases of cholera and a public well into which sewage was found to be draining. Dr. Snow requested that the pump handle of the offending well be removed. The immediate cessation of new cases of cholera was his proof that sewage-contaminated water was a cause of cholera.
Epidemiology of Infectious Diseases
Dr. Snow’s success stimulated the search for associations of every variety in the study of epidemics and led to the procedure that came to be known as epidemiology (from epidemic-ology). Within a few decades of Dr. Snow’s discovery, it became common knowledge that the “things” that caused infectious diseases were microscopic organisms called germs. Thus the science of microbiology expanded, and epidemiology flourished as a major tool in investigations of infectious diseases.
First Signs of Trouble
It soon became apparent in the early days of microbiology that epidemiology frequently failed to produce reliable results when matching infectious organisms with their respective diseases. All varieties of infections were being attributed to weird kinds of exposures. Obviously, the newly found practice of epidemiology was as dreadfully misunderstood and misused in its early days as it has been in the recent decades of research in the nutritional science that formed the basis for the Dietary Guidelines for Americans (DGA).
The obvious conclusion from the failures of epidemiology throughout its history is that epidemiology has been regarded from its inception by novice practitioners as merely a simple matter of biostatistical maneuvering. They forgot or neglected to do the most important part of epidemiology that was demonstrated by Dr. Snow; they forgot to “remove the pump handle!”
The Koch Solution
Fortunately, the early epidemiologists recognized that a major problem was assuming that association equaled cause and thereby neglecting to prove the cause with applicable hard science. It was decided that they had to adopt a standard procedure for confirming causality of associations they found. To this end, Robert Koch, a German Physician framed Koch’s Postulates in the late 1800s. The postulates were as follows (2, p.195):
- The organism responsible for the disease must be present in every case of the disease.
- The organism must be able to be isolated from the patients and grown in pure culture.
- The cultured organism must be able to cause the disease by exposure of a healthy host.
- The organism must be recovered from the experimentally infected host.
Koch’s postulates served well with few exceptions until advancements in microbiological science revealed some incompatibilities with the wording. Through the years, the postulates were appropriately modified. With today’s modern microbiological techniques, they are not routinely used or required. However, the postulates’ historical importance and theoretical significance for proving causality are still acknowledged and honored.
Nutritional Science Discovers Epidemiology
It must be recalled that epidemiology was born of a need to understand epidemics of infectious diseases. Well before its discovery by nutritional scientists, epidemiology had developed as a well-matured discipline of inestimable value in the study and causation of infectious diseases. However, Koch’s rules for proving that an association is causal cannot be applied directly to effects of environmental chemicals, noninfectious agents, or nutrients because of the differences between microorganisms and all other chemical substances in how they cause disease (2, p. 195ff). Epidemiology was neither designed for study of noninfectious diseases nor was it ever considered to be useful for study of nutritional diseases. In fact, the concept of “nutritional disease” itself was nonexistent. How did nutritional science find epidemiology?
The Role of Pellagra
It was not until the great pellagra pandemic, circa 1908, that the medical community and general public were finally forced to accept the fact that serious or even lethal disease could be caused by a nutritional deficiency (1, p.250). Although pellagra was known to be associated with high consumption of maize (corn), a staple food in poor countries, it was declared to be an infectious disease because Southern pride was offended by the suggestion that the disease was associated with poverty and malnutrition.
Over its course, the pellagra epidemic produced more than 3 million cases and caused 100,000 deaths in 15 southern states. By 1914, the epidemiological and clinical studies of Dr. Joseph Goldberger, an officer in the Public Health Service, showed that the disease could be reversed and prevented by dietary means. However, it was not until the 1940s, when it finally became accepted that pellagra was caused by a deficiency of niacin, that the outbreak was finally controlled. The fascinating story of the social and political forces that seriously interfered with medical research and with the investigation of the epidemic were finally published in 1992 (3).
Application of Epidemiology to Nutritional Science
The spectacular success of epidemiology in the instance of the nutritional disease pellagra suggested it for application in investigation of noninfectious diseases. In the absence of information to the contrary, it can be assumed that the newly popular epidemiology was well employed and accepted in the mid- to late-20th Century, which happened to be an active period of vitamin discovery and research.
This quiet but scientifically productive period provided a friendly environment for both the research community and epidemiology. The research community was given time to familiarize itself with this relatively new tool for research and popularize its applications; epidemiology was presented with the opportunity to serve a new area of research interest, namely nutritional science. Fortunately for epidemiology, the field of nutritional science it served was the relatively simple one related to vitamin requirements and deficiencies.
Epidemiology’s good fortune was due to the fact that the diseases that resulted from vitamin deficiencies had very much in common with those of infectious origin; they dealt primarily with single agents. When a disease is caused by a single agent such as a vitamin or microorganism, an association of the agent with the disease is easier to detect because it not confounded by the presence or absence of other variables. Thus, epidemiology faced few problems in investigations of beneficial and detrimental effects of individual vitamins.
Origin of Flawed Nutritional Science
After the end of WWII, Ancel Keys, an American multidiscipline scientist (4), entered the scene. He authored the lipid hypothesis, the label for the diet-heart connection, and suddenly nutritional science was recognized to be tremendously more complicated. In hindsight, Ancel keys with his lipid hypothesis can now be identified as the ultimate architect of flawed nutritional science.
In 1953, Keys published an analysis of data from six countries that showed a direct, correlation between mortality from coronary heart disease (CHD) and percentage of calories from dietary fat. He did not, however, include data from all 22 countries that were available to him at the time. If data from all 22 countries had been used, there would have been no straight line, and no reasonable correlation would have been possible. Keys also ignored a well-established epidemiologic principle by claiming that his straight- line association between CHD and dietary fat proved a causal relationship (5).
Keys’ lipid hypothesis study soon became known in professional epidemiology circles as a textbook example of how not to conduct an epidemiological study: it did not include all available pertinent data; it selected data that supported its thesis; and it claimed a cause-effect relationship where there was only a mere association (6). Despite the many exceptions that existed to his hypothesis and the criticisms of his epidemiology, Keys persisted and made validating the lipid hypothesis his life’s work.
Despite Keys’ blatant misapplication of epidemiology, his peers supported him wholeheartedly. The research of Ancel Keys and colleagues became widely publicized and popular. The nutrition community enthusiastically adopted the lipid hypothesis. By the early 1970s, the lipid hypothesis was fully accepted as fact by the nutrition community. Keys’ colleagues shored up Keys’ arguments with innumerable studies using equally flawed epidemiology. Only a few scientists called attention to his misuse of accepted epidemiologic methods.
Little by little, the lipid hypothesis was reworked and refined to eventually produce an ultimately-proven-to-be-faulty nutritional program that was designed to prevent cardiovascular diseases. It became known as the heart-healthy diet and specified to be low in fat, high in carbohydrates, and restricted in animal products. In less than two decades, the heart-healthy diet became national policy.
The Politicization of Nutritional Science
Official sanction was given to the heart-healthy diet with the entry of politics into the field of nutrition in 1977. It occurred with the publication of Dietary Goals for the United States (1, p. 14ff). It is of extreme yet unappreciated significance that government control shifted the focus of nutritional science from choosing food components that contained healthful nutrients to avoiding intakes of food components linked to chronic disease. Personal eating preferences and habits had never before in recorded history been the subject of governmental action.
After a decade or two of bureaucratic central planning, it became obvious that the long and gradual change in the diet composition created by the bureaucracy was accompanied by a strikingly steady climb in the incidence of obesity and cardiovascular diseases that the heart-healthy diet was enacted to prevent. With time, the inventory of nutritional diseases began to expand and the numbers continued up the path to epidemic levels. The fact that there was a strong association between the diet and disease was not lost to members of the scientific community.
Fortunately, a prominent science writer by the name of Gary Taubes became personally concerned about the diet-disease connection and wrote a magnificent, superbly documented volume describing the history of actors and events that led to the public health disaster created by the Dietary Goals for Americans (DGA). Taubes’ book, Good Calories, Bad Calories (GCBC) published in 2007 (7), is now a classic; it is a book that still should be required reading for all members of the medical and nutrition communities.
Taubes revelations in GCBC confirmed the beliefs of many in public and private life that there is more than ample material in America’s almost century-long history of nutritional research, regulation, and manipulation to write expository volumes reflecting on the what, how, and why of a half century-long flagrant fraud and deception conducted by agencies, institutions, and even a few respected officials charged with setting official nutrition policy. It was little known at the time that another such exposition was already in preparation.
Nina Teicholz, a reporter, journalist, wife, and mother, was busy searching for all the evidence she could find relating to the pros and cons of the DGA low-fat diet. After eight years, Nina’s disturbing and incredible findings were introduced, with little ceremony, to the reading public as The Big Fat Surprise (BFS) (8). Within two years, it received a number of Best Book and Best Seller awards. BFS promises to be, like GCBC, a superbly documented, powerful, and significant document that will become a classic in years to come. More importantly, Nina’s pretty, poised, and very well-informed self has become a popular, charming, and influential talk-show guest.
The future of the DGA is unpredictable at the present time, but we do know from their contributions to the drama before us that Gary Taubes set the stage and Nina Teicholz raised the curtain.
What Went Wrong
The sad state of affairs for epidemiological science today is that, through no fault of its own, it has become the whipping boy for the disastrous health outcomes foisted on the American public by the government-sponsored Dietary Guidelines for Americans. A recent opinion commentary in the Wall Street Journal (9) candidly described the distressing travesty of the government’s most recent release of the 2015-2020 DGA as a “failure for 35 years to check the rise of obesity and diabetes.”
What has brought nutritional science professionals to disparage the venerable discipline of epidemiology that has been and still is a major procedure in their research endeavors and to dismiss its product as “weak science?” It is time to set the record straight and show that epidemiology has not failed nutritional science; nutritional science has failed epidemiology
An in-depth review of the body of literature on which the government-supported diet/heart hypothesis was founded reveals many scientific missteps. The fact that nutritional science was in its infancy fifty years ago may excuse some of the errors apparent in many of the earlier research papers, but such a defense cannot be used today for the scientific missteps that are still occurring in research involving the relationship between diet and disease.
Use of Inappropriate Methodology
There can be little doubt that epidemiology is not an appropriate tool (or at best, an extremely difficult one) for use in nutritional research. The number of macro- and micronutrient variables that exist and the complexity of their interrelationships make for challenging selection of appropriate variables. Are the difficulties worth the time and effort of an outcome that can only tell strength of association?
A provocative but unanswered question is why did the nutrition establishment persist for so many decades in using epidemiological methodology that actually produced few or no results predictive of clinical outcomes? The WSJ opinion commentary (9), for example, reported that an analysis of “several dozen highly cited studies concluded that subsequent clinical trials could reproduce only about 20% of observational findings.” More egregious was the fact that of 52 claims made by prospective epidemiology in nutritional studies, “none – 0% – withstood the scrutiny of subsequent clinical trials.”
The poor performance of epidemiology in nutritional research throughout the decades suggests either epidemiologic incompetence or deliberate misuse of the method. In either event, the conclusions from both are the same but their objectives are vastly different. Deliberate misuse of epidemiology presents very serious ethical issues that are beyond the scope of this discussion, but they may be surmised from the revelations in GCBC and BFS. The means of altering results by varying different elements of the process will be described in the next section.
Methods of Misuse, Manipulation
The benefit from altering epidemiological information is basically twofold: It provides a means for tailor-making the results of research; and it helps cover neglect of proving causation.
The most common flaws found in nutritional studies appear to be incompetent selection of variables, assumption of faulty premises, and confusion between cause and effect. These are described at greater length elsewhere (6), and may be inadvertent or deliberate. Regardless of intent, they can significantly alter the impact of a study.
In recent years, a new technique has been developed for the apparent purpose of deceiving the public into believing that the association in question is strong enough to be accepted as causal. This technique substitutes the traditional word “association” with one that conveys the image of a closer tie. The first such word we noted was “link.”
“Link” was soon followed by the more provocative concept of “risk.” This method for reporting statistical associations no longer expresses the findings as probability of the existence of an association but rather as the risk of the potential event caused by the association actually happening. It is a perfectly acceptable epidemiologic method for comparing the potential for harm among several associations. However, it carries the caveat that the associations must first be shown to be causal.
This new epidemiology (10) absolves epidemiologists from conducting the final epidemiological step of proving that the association is a causal one, which is expensive, time consuming, and requires competence in the appropriate scientific disciplines. Despite the deception it creates, the new epidemiology is employed by eminent scientists whose papers are published in outstanding scientific journals. Thus we have influential scientific papers that report possibilities as probabilities:
… 9.3% of deaths in men and 7.6% in women in these cohorts could be prevented at the end of follow-up if all the individuals consumed fewer than 0.5 servings per day (approx. 42 g/d) of red meat. Conclusions: Red meat consumption is associated with an increased risk of total, CVD, and cancer mortality. Substitution of other healthy protein sources for red meat is associated with a lower mortality risk (11).
The careful inclusion of “associated” does not lessen the impact of “risk” in the public mind. The reporting of the risk of mortality and CVD from consuming red meat when there is no scientific proof of a cause/effect relationship is irresponsible and immoral.
The outstanding leadership of Nina Teicholz in the latest campaign to right the wrongs of the DGA promised for the first time in the history of the DGA that right would prevail. The Secretaries of Agriculture (Ag) and Health and Human Services (HHS), the overseers of the DGA, were forced to the table and directed to produce a 2015-2020 edition of the DGA based on valid science. They gave their promise to Congress that they would do so.
Ag and HHS returned to the table a couple of months later with the final revision that was essentially unchanged from the 2010 DGA claiming that it now was based on valid science. Case closed! Congress was satisfied – a distressing end to a brilliant campaign!
Requirements for Assuring Quality of Guidelines?
The idea of preparing a set of Koch’s postulates for application to the study of noninfectious chemicals was considered several decades ago when toxicity of pesticides first became a matter of concern. The thought was soon abandoned when the difficulties of such an approach were evaluated postulate by postulate (2, p. 195ff).
Although the concept of specific rules governing legitimacy of research findings would be antithetical to the nature of scientific inquiry, an independent review of the DGA by the National Academy of Medicine, as authorized by Congress in response to Nina’s campaign, would be most appropriate. Scientists know when research is honest. However, any effort to condone use of the incompetent epidemiology must be acknowledged as unacceptable.
Let’s Throw the Bums Out
In a very informative post on the history of the DGA (12), Michael Eades prophetically predicted the outcome:
Before we get started, for whatever it’s worth, I don’t think the government should be in the nutrition business, but it is. And barring some sort of revolution, the government will doubtless continue to be in the nutrition business (as well as countless other endeavors they have no business in) for the foreseeable future.
In Comment 50 of the same post, there appears the following paragraph:
Despite it all, there appears to be scant hope that there will be much pure science in the new DGA. For one thing, there is no way that pseudoscience can be rehabilitated to become pure science. And two, is it reasonable to expect authors who are wedded to their own nutritional beliefs (aka believable untruths) to be capable of objectivity?
The comment concludes, “In summary, let us revive in jargon a few well-worn clichés: You cannot make a silk purse out of a sow’s ear, plus a leopard cannot change its spots, so let’s throw the bums out.”
What if There Were No Guidelines?
This thought-provoking, audacious question was asked by one lone voice a little over a year ago (13). It received little or no recognition at the time and is now forgotten. It was written by Adele Hite, an MS/RD, cofounder and director of the Healthy Nation Coalition (14), wife, and mother who is currently enrolled in a doctoral program.
Adele is an unsung heroine in the fight to bring accuracy and integrity into the field of nutritional science. Now is the time not to just revisit Adele’s words, but to memorialize them in a Declaration of Independence for nutritional science as a directive for public action.
It is crucial to note that wholesale adoption of bogus epidemiology by the public is the linchpin of the DGA. If public demand for sound nutritional science is not able to dislodge the linchpin, the massive human feeding experiment foisted on the American public decades ago will continue to take its terrible toll. Let us retire nutritional epidemiology and the DGA committee. Let nutritional science reside in the province of Science.
- Ottoboni A, Ottoboni F. The Modern Nutritional Diseases and How to Prevent Them. Fernley, NV: Vincente Books.
- Frank P, Ottoboni MA. The Dose Makes the Poison: A Plain-Language Guide to Toxicology, 3rd Edition. New York, NY: John Wiley & Sons, 2011.
- Bollet AJ. Politics and pellagra: The epidemic of pellagra in the U.S. in the early 20th The Yale Journal of Biology and Medicine. 1992; 65: 211-221
- Taubes G. Good Calories, Bad Calories. New York, NY: Alfred A. Knopf, 2007.
- Importance of Dietary Animal Fat, The Big Fat Surpise
- New Epidemiology
- Pan A et al. Red Meat Consumption and Mortality. Archives of Internal Medicine. 2012; 172(7): 555. URL: http://www.ncbi.nlm.nih.gov/pubmed/22412075