Home | How Traditional Tales Reflect Popular Beliefs--a Vivid Example of the Process | WHY THE GODS PREFERS SKEPTICS--JK | DAMNED HUMAN RACE--Mark Twain | Clear Thinking: Use and Abuse of Reason--JK | LOGICAL FALLACIES: a listing | How to Evaluate Medical Discoveries | SOCIAL JUSTICE; 8 STEPS FORWARD--JK | STRANGENESS OF MAN, EXPLAINED--JK | THE RISING IQ PUZZLE--JK | Why Smart People Believer Weird Things--Prof. Shermer | How Thinking Goes Wrong--Prof. Shermer | Contact Me | GOD & EVIL, a witty story | WHY PEOPLE BELIEVE WEIRD THINGS | New Page Title
How to Evaluate Medical discoveries

Enter subhead content here

It was back in the early 70s when I read a medical article on high fiber diet, which I found convincing.  It concluded that the reported health benefits of a high fiber diet were the result of the contravening variables; to wit, eating less meats and other foods that contribute to cancer and the living of a healthful lifestyle.  In other words, the average person in the high-fiber group had a lower cancer incidence because of the contravening variables, and once the two groups were matched for these variables, then the rates of colon cancer were about equal.  The failure to institute proper controls vitiates the claims for fiber being prophylactic. 


The issue of contravening variable goes to the heart of most health studies.  If there is no known reason as to why a certain substance is healthful, then one must look to the contravening variable.  And even when such mechanism exists, the influence of the contravening variables can accentuate the effect, or the putative mechanism might in fact be illusory. 


There is something very suspect of claiming that carbohydrates that we cant digest somehow ward off colon cancer.  First what is the fate of those carbohydrates that pass through usdo they remain undigested, or do they feed intestinal bacteria?  Second, how would a bit of additional bulk in the stool function to prevent?  Third, are the answers to these questions substantiated by research?  Fourth, if so answered has the answered been quantified?  I have not come across research at the molecular level.  Answers at the molecular level do not prove the issue.  Those of faith often in naturalistic health will provide answers on a cellular level.  However, those with the right expertise recognize the fallaciousness of the answers. 


 If one does not bring to the table a deep knowledge of the bio-molecular foundation of life, then one should look to peer review articles in journals of high reputation.  And even then one should be skeptical of findings where the mechanism for the effect is poorly known and where the contravening variables have not be controlled.  


Below is an article illustrating the merit of skepticism when it comes to population research about what is healthful.--JK 





By Harry K. Ziel, M.D.

Of late, the internet represents a great potential for transmission of medical misinformation to an unwary public. As the use of the internet geometrically increases, web surfers intent on learning the most up to date information on disease treatment are exposed to a vast array of unproved therapies. All the traditional safeguards are lost when usually well intentioned authors, enthusiastic to announce their discoveries, pass along their latest findings over the convenient computer network.


A patient with an unusual medical condition like pulmonary interstitial fibrosis or dermatomyositis may surf the web to find any number of sites which discuss his condition. That unsophisticated patient may download, read, and believe that medical researchers have found that a particular diet or herbal medicine is effective in treating his disorder. Desperate with an uncommon disease that is showing no or slow response to traditional allopathic medicine (TAM), the patient is prone to look for and to utilize a number of complementary alternative medical (CAM) nostrums of no value or even a substance that causes harm. These computer web sites are the equivalent of Tijuana locations touting falsely advertised miracle cures that are falsely advertised, like laetrile.*


Excluding internet reporting, transmission of responsible scientific information today falls into the venue of a huge number of scientific journals. A hierarchy of scientific journals exists in each specialty area. Specialty organizations support the publication of most journals in their fields. For instance, the American Chemical Society's publication division supports publication of 27 different journals. In all, publication companies in United States print approximately 16,000 various peer-reviewed journals.

In each specialty area of medicine, a hierarchy of journals has developed over time. The most prestigious and selective medical journal is The New England Journal of Medicine (NEJM), published since 1812 by the Massachusetts Medical Society. Only one out of 10 articles submitted survives the NEJM peer review process. NEJM editors send articles submitted to anonymous reviewers, experts in the fields covered by the articles to be considered for publication. These highly responsible referees must reject, accept with recommended revision, or rarely accept without revision all articles that clear the editors' primary review.

A close second to the NEJM is the Journal of the American Medical Association (JAMA). The American Medical Association has published the JAMA since 1883. The JAMA also accepts about one in 10 articles submitted. For the fields of general medicine, these two journals have attained the first tier status in the hierarchy of medical journal reporting.

Journal editors today virtually require that all the authors have subjected their data to sophisticated statistical evaluation. One of the referees evaluating the article is a statistician who must peruse and agree with the testing of the data. First tiered medical journals require that authors follow set formats. Editors insist on pertinent references supporting views the authors present. Authors must disclose their sources of financial support. Purposals to perform research must first clear investigational review boards (IRBs). Before the research protocol starts, IRBs insist on meaningful studies, check on adequate size of patient cohorts to attain statistically significant conclusions, require patient consents which enumerate all patient risks as well as provision for study termination should patient injury become manifest.

Lower tiered journals tend to accept articles which meet with ever lower standards of quality review. The lowest tiered journals may accept and publish nearly all articles submitted. Obviously, one needs to scrutinize and be more wary of conclusions tabulated in the lower tiered journals. Unpublished internet findings have the least degree of peer review and accordingly should bear the greatest skepticism. Without any ability to review the material and methods, the statistical evaluation of the data, the quality of internet study conclusions are highly problematic.


High quality peer reviewed reports can not be accepted fully, even if data appear to be statistically valid. Unrecognized biases often contaminate data. Statistical validation simply suggests truth. Only multiple studies, all statistically valid, each concluding similar findings, biologically plausible, with increasing exposure correlating with increasing effect will point to a causal rather than a casual association between cause and effect. The best studies are prospective in which patients are randomly assigned to a study group or a control group. Both the researchers and the patients are unaware (blinded) into which study or control group patients are enlisted. Study medications and placebos look alike. Researchers check compliance in taking medication by inspecting patient logs and remaining medication during each visit. Prospective studies are long and expensive.

As an alternative to prospective studies, researchers often choose to perform retrospective case control studies which are far faster and less expensive to conduct than prospective studies. Selection of control patients randomly chosen and matched by age, ethnicity, parity, Ponderal index, socioeconomic status etc. are paramount to avoid bias in retrospective case control studies. Simple observational studies, sometimes matched with historical controls, provide information of a far lessor quality on which physicians must sometime decide therapy when no better information is available.

To illustrate why multiple studies, each coming to the similar conclusions, must be the gold standard for scientific decision making, one needs only to point to the recent NEJM article which refutes the long held belief that a high fiber diet was protective against colon cancer. In 1971, Denis Burkitt first reported that Africans who ate a high fiber diet had a low incidence of colon cancer. A 1992 meta-analysis done by Howe et al of 13 case control studies documented both a protective effect of fiber against colon cancer as well as a dose-response relationship (greater fiber use resulted in less colon cancer incidence). Thun et al in 1992 and Steinmetz et al in 1994 both showed an inverse relationship between high fiber intake and colon cancer occurrence indicating protection from colon cancer from fiber intake. Because of other associated benefits of high fiber intake, i.e., reduced incidence of diverticulosis, less coronary artery atherosclerosis, lower incidence of hypertension, and less frequent type 2 - non insulin dependent diabetes, high fiber became a highly encouraged prophylactic disease intervention. The fiber bandwagon was rolling along!

Hints that fiber was ineffective in colon cancer prevention however arose from four publications. Responsible were DeCosse et al in 1989, McKeown-Eyssen et al in 1994, MacLennan et al in 1995, and Platz et al in 1997.

Fuchs et al in the January 21, 1999 issue of the NEJM reported from the Nurses' Health Study begun in 1976 that no protective association existed secondary to the use of high fiber diets from colon cancer or from premalignant adenomas known to precede colon cancer development. Their meticulous study refuted a belief held for the past 28 years. The Nurses' Health Study is an ongoing prospective study of 88,757 women conducted by a highly regarded research team at Harvard, one member of which team, Walter Willett, has made dietary influence on disease development his life's work.

Shari Roan, writing in the January 25, 1999 Los Angeles Times, was quick to jump both on and off the Nurse's Health Band Wagon saying that the report shows that one can not trust a "lone study". Her assessment of where we stand in understanding the causes and prevention of colon adenomas and cancer is just what the public needs to hear.

A true skeptic must say, "I still see no gold standard met. The story of colon cancer prevention is complicated by too many other factors than fiber ingestion. The skeptic requires many more studies involving complex carbohydrates and sugars, carcinogens derived from high temperature cooking , ingestion of smoked fish and meats, camplobacter and other enteric pathogens, genetic predilection, as well as pesticides and other contaminants from foodstuffs and water to begin to find the solution to the colon cancer causes and prevention's. It's not a crime to admit one does not know.


The skeptic who steps outside the parade to await the final float before he steps back into line will experience the fewest upsets. The cocky drum major who heads a parade is most conspicuous if he leads followers who are all out of step. Scientific proof may take generations before discovery.


1.Burkitt D P. Epidemiology of cancer of the colon and rectum. Cancer 1971; 28:3-13.
2. DeCosse J J, Miller H H, Lesser M L. Effect of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenomatous polyposis. J Natl Cancer Inst 1989; 81, 1290-1297.
3. Fuchs C S, Gioannucci E L, Colditz B A, Hunter J H, Stampfer M J, Rosner B, Speizer F E, Willett W C. Dietary Fiber and the Risk of Colorectal Cancer and Adenoma in Women. N Engl. J Med 1999; 340: 169-176.
4. Howe G R, Benito E, Castelleto R, et al. Dietary intake of fiber and decreased risk of cancers of the colon and rectum: evidence from the combined analysis if 13 case-control studies, J Natl Cancer Inst 1992;84:1887-1896.
5. MacLennan R, Macrae F, Bain C, et al. Randomized trial of intake of fat, fiber, and beta carotene to prevent colorectal adenomas: the Australian Polyp Prevention Project. J Natl Cancer Inst 1995, 87:1760-1766.
6. McKeown-Eyssen G E, Bright-See E, Bruce W R, Jazmaji V. A randomized trial of a low fat high fibre diet in the recurrence of colorectal polyps: Toronto Polyp Prevention Group. J Clin Epidemol. 1994; 47:525-536.
7. Platz E A, Giovannucci E, Rimm E B, et al. Dietary fiber and distal colorectal adenoma in men. Cancer Epidemiol Biomarkers Prev 1997; 6: 661-670.
8. Steinmetz K A, Kushi L H, Bostick R M, Folsom A R, Potter J D. Vegetables, fruit, and colon cancer in the Iowa Woman's Health Study. Am J Epidemiol 1994; 139: 1-15.
9. Thun M J, Calle E E, Namboodiri M M, et al. Risk factors for fatal colon cancer in a large prospective study. J Natl Cancer Inst !992; 84: 1491-1500


*Laetrile is a good point of example for there is nothing peculiar about the chemical contents of peach kernels that support strongly a medicinal claim.  The traces of cyanide are not medicinal and the other substances are found in the common almondto which the peach is a member of the family.  Speculation as to some natural curative effect of a chemical without a demonstration of its biochemical mechanism is mere product hype.  When a credentials is added to a name, this does not entail that the experts knowledge is in fact scientifically sound.  There are a few who have earned such initials attached to a name how have a religious belief in quackery, and there are many more who have obtained such letters from institutions that do not qualify them to practice medicine or be employed at a university as a research scientist.  Their articles have an odor like that of laetrile, rotten almonds, and the results can be as deadly as cyanide.  For the failure to obtain proper treatment of a medical condition has accounted for many early deaths.     


Enter supporting content here