Causation, Correlation, Dogma, Weight, and Health
Posted by Miriam Gordon on August 10, 2009
After acquiring the book almost a year ago, I (again) started reading Gary Taubes’ book entitled Good Calories, Bad Calories. Based on what I’ve read so far, and knowing Gary Taubes’ background, I believe it’s a very scholarly work, and very thoroughly researched. From the title, it’s obvious that this book considers the scientific evidence for specific types of diets and how they affect body weight regulation.
In the first part of the book, in order to draw parallels with current scientific evidence for the “epidemics of obesity and diabetes,” Taubes puts forth a detailed historical analysis of the decades-long debate on whether dietary fat intake was the definitive cause of, rather than merely correlated with, the rapid rise in the incidence of cardiovascular disease in the developed world that started in the 1920s (when consumption of red meat increased considerably). The upshot of this work is to point out what happens when a researcher, who becomes prominent for various reasons, influences health care policy even when his/her scientific data are far from conclusive. In this particular case, Taubes discusses the work of the prominent physiologist Ancel Keys, who was convinced based on his research that the observed rise in incidence of heart disease was due to increased blood cholesterol levels, which was in turn due to increased total dietary fat intake. Keys was apparently a formidable character who felt very strongly that his data conclusively proved this hypothesis, and was very quick to strongly criticize those who opposed his theories. Throughout this time period, and even into the 1960s, there were many prominent researchers who had serious reservations about Keys’ theories, based on scientific analysis of his data as well as their own. Nevertheless, because Keys was so forceful, Taubes brings various elements to show that the media picked up Keys’ theories, and physicians who were faced with an alarming medical mystery began to recommend low fat diets to their patients, despite serious controversy over Keys’ data.
In such a scenario, the question becomes one of correlation versus causation, i.e., depending on how solid the scientific evidence is for any given observed public health phenomenon, one might be able to say there is a CORRELATION of an observed public health trend with disease, rather than being able to state, through a solid base of scientific evidence, that the observed trend CAUSES the disease. There is considerable scientific evidence demonstrating that the physical attribute of fatness does not conclusively indicate bad health, and that many “obese” individuals are metabolically healthy.
One reason for this phenomenon of correlation overpowering the media is that it provides a solid message to address what appears to be an alarming trend. Often, people don’t have the patience to wait for conclusive scientific evidence to be produced when faced with a potentially scary scenario. When scientific evidence that contradicts the popular theory is published, it tends to be ignored, because it contradicts what has become DOGMA. This is what has happened with “obesity” research. Taubes skillfully points out that when scientifically observed correlations are not thoroughly researched scientifically, and they become socially accepted as dogma, real scientific progress breaks down.