One of my colleagues brought to my attention a recent article on the New York Times Well blog. The article reports results of a study in which castrated rats are randomized to do some running on a treadmill, or to be sedentary. The rats’ brains were then tested for DHT and testosterone content, and it was found that the rats who did exercise had more DHT in their hippocampi. You can read the article here. The original study is here.

Here’s my problem, illustrated by an excerpt from the blog post: “But one aspect of the new experiment is already resoundingly clear and reassuring, Dr. McEwen points out. “The exercise in this experiment was quite mild,” he says — the equivalent of jogging at a pace at which someone could speak (or squeak) to a companion. “That’s achievable for most people,” he concludes, “and the evidence suggests that it will improve brain health.”

The study was in castrated rats. The study author goes from rats with no testes, straight to recommending that people jog around, and says that the study provides evidence that for humans, jogging will improve brain health.

One of my colleagues has just written an excellent article on the use of mechanistic reasoning in evidence-based medicine and I’d encourage you to take a look. It tries to solve this problem- that it’s simply inappropriate to suggest that an animal study provides evidence that something ‘works’ in humans. I’m disappointed in the study author that he made this baseless statement, but we all have to do our part in ensuring scientific results are reported accurately and fairly.