Exercise and cognition - I'm still skeptical

I've been skeptical of articles purporting to show a relationship between exercise and cognitive performance. It's not that I don't like exercise (I love it), but I just couldn't see the evolutionary physiology -- and I don't have much faith in case-control studies.

Now I'm moving from more to less skeptical, but I'm still skeptical

Study shows why exercise boosts brainpower - CNN.com

... Tests on mice showed they grew new brain cells [in response to exercise] in a brain region called the dentate gyrus, a part of the hippocampus that is known to be affected in the age-related memory decline that begins around age 30 for most humans.

The researchers used magnetic resonance imaging scans to help document the process in mice -- and then used MRIs to look at the brains of people before and after exercise.

... They recruited 11 healthy adults and made them undergo a three-month aerobic exercise regimen.

They did MRIs of their brains before and after. They also measured the fitness of each volunteer by measuring oxygen volume before and after the training program.

Exercise generated blood flow to the dentate gyrus of the people, and the more fit a person got, the more blood flow the MRI detected, the researchers found.

"The remarkable similarities between the exercise-induced cerebral blood volume changes in the hippocampal formation of mice and humans suggest that the effect is mediated by similar mechanisms," they wrote.

"Our next step is to identify the exercise regimen that is most beneficial to improve cognition and reduce normal memory loss, so that physicians may be able to prescribe specific types of exercise to improve memory," Small said.

I'm still skeptical, but a bit less so.

It was a pretty small study, and it would be hard to separate the effects of exercise and sleep (since sleep improves due to exercise). On the other hand, it's almost plausible that a mechanism that evolved for muscle learning might secondarily benefit other dependent processes (such as recall) ... but it still seems fishy to me.

It might turn out that the benefit is a one time plus for people going from unfit to fit, and then it's done. Or it might turn out that you need to learn new motor actions to benefit, so for most of us taking up snowboarding would help but bicycling wouldn't.

If there is any effect, I'm betting it's small and that sleeping 7+ hours a night is more important.

On the other hand, if my dementia progresses any more quickly every activity will represent new muscle learning opportunities.