Speaking of how the media is prone to broadcasting flawed observational studies, an article in yesterday s The New York Times reports on research claiming that eating even a little daily meat may increase a person s risk of Type 2 diabetes. In this purely observational study, researchers from the Harvard School of Public Health analyzed the diet and health habits of about 300,000 individuals, ages 25 to 75, from among three separate groups, all of whom kept dietary diaries starting in 1976. Medical history and information related to diet was updated every two years.
The findings showed that for every daily serving of unprocessed red meat consumed equivalent to a 3.5 ounce serving of steak, less than a typical hamburger a person s risk of Type 2 diabetes went up by 19 percent. Eating 50 grams of processed meat daily (akin to one hot dog or about two strips of bacon) increased that risk by a shocking 51 percent.
The research, published in the American Journal of Clinical Nutrition, also used mathematical models to determine that replacing one serving of meat with nuts could lower the risk of Type 2 diabetes by 21 percent.
ACSH's Dr. Josh Bloom adds, Mathematical models? Please. There is probably a mathematical model proving that you can change your eye color by climbing the Empire State Building. That s about what they re worth.
Though the authors concede that the people who ate the most meat were also more likely to smoke, be physically inactive, consume more calories, and have a higher body mass index (suggesting excess body weight), Dr. Frank Hu, professor of nutrition and epidemiology at the Harvard School of Public Health and the study s senior author, contends that the risk of Type 2 diabetes was still significantly elevated after accounting for these factors.
To which ACSH's Dr. Gilbert Ross observes, What an atrocious study. He adds, These folks rely on the Nurses Health Study and the Health Professionals Follow-Up study to crank out article after article by nitpicking through all of the possible data amassed by hundreds of thousands of self-reported diets. If you do enough correlations, as Dr. Gutting pointed out earlier, you ll find enough statistically significant information necessary to publish papers like this one and have the news media report on its misleading results.