Dr. Gary Gutting, a professor of philosophy at the University of Notre Dame, is tired of opening the paper only to read about yet another new study describing the purported health benefits of various vitamins, supplements, or lifestyle habits. In an article for The New York Times, Dr. Gutting explains to readers the differences between observational studies and randomized controlled trials. The latter, considered the gold standard of proving cause and effect, are studies designed to investigate a specific theory by randomly assigning subjects to receive whatever substance or treatment is being tested. On the other hand, observational studies only review information without having imposed any intervention, and thus can only suggest an association, not causation. Unfortunately, most of the research reported on by the media is observational, since the media s interest in attracting attention so often compels reporters to highlight sensational but flawed or preliminary studies. Such reports, as Dr. Gutting points out, can at best suggest directions for further research instead of establishing reliable conclusions.
What s the harm, one might ask, in publishing such research, even if it's only preliminary? Well, as Dr. Gutting aptly points out, the problem is that most of these observational studies make recommendations for the prevention of future diseases instead of remedies for current ailments; therefore, even the small risks associated with following their advice outweigh the probability of achieving any health benefit. From this point of view, the media s continual focus on observational studies is counterproductive and only causes people undue stress as they try and keep up with the latest, often tenuous, findings.
As ACSH s Dr. Elizabeth Whelan points out, The media is there to entertain us, and they rely on these observational studies to accomplish just that.
ACSH s Dr. Gilbert Ross further explains that studies lacking any medical intervention cannot come up with a cause-and-effect relationship. Instead, they simply try to correlate a behavior with some outcome. He adds, Unfortunately, my guess is that about three-quarters of the studies the media reports on are observational rather than interventional and controlled.
ACSH s Dr. Josh Bloom points out a particularly egregious example. About a month ago, he notes, there were headlines claiming that diet soda consumption is linked to obesity. Some even said that the soda caused obesity. But after looking at the study for a minute, the real answer became obvious: people who are already obese drink more diet soda. In this case, as in so many others, correlation and causation became ridiculously confused.
