Running red lights. Compulsive running boosts neuron production in the hippocampus of mice (red label, right), but doesn't make them any smarter.

Running to Distraction

Die-hard joggers now have a new worry. If they're anything like running-addicted mice, they may want to ease off their training regimen to stay sharp. Even though running makes mice sprout new neurons, running compulsively makes them slower learners.

Exercise is known to aid the formation of new neurons in the mammalian hippocampus, a brain area vital in learning and memory, and many researchers had assumed that these extra neurons would translate to more brain power. Behavioral neuroscientist Justin Rhodes, now at Oregon Health and Science University in Portland, set out to test this hypothesis as a grad student at the University of Wisconsin, Madison.

Rhodes and colleagues studied two sets of mice: ordinary ones, and ones that had been bred through many generations to have a predilection for running. They gave half of each group access to an exercise wheel. As expected, the mice that were bred to run averaged up to 12 kilometers a day--two to three times as much as the ordinary mice. And running increased neurogenesis more in the born-to-run mice, as later brain studies showed.

However, all that running contributed nothing to their spatial skills. In an "underwater maze" test requiring animals swimming in darkened water to remember the location of a submerged platform, the born runners who had had wheel access showed little ability to learn the task, performing even worse than their nonrunning relatives. In contrast, the ordinary mice with running experience showed clear improvement with practice in their maze performance, the scientists report in the October issue of Behavioral Neuroscience.

Rhodes says he has evidence that the born-to-run mice are "addicted to exercise": If they are denied access to the running wheel, their brains show "huge amounts of activity" in the same areas as the brains of drug-craving animals in addiction research. With such animals, running "actually is impairing learning," he says. Co-author Fred Gage of the Salk Institute in La Jolla, California, says the findings may have implications for hyperactive kids, perhaps explaining "why the increased activity in hyperactive kids does not lead to cognitive enhancement."

Princeton University neuroscientist Elizabeth Gould says the study adds to evidence that neurogenesis and brain performance aren't as tightly linked as some researchers initially supposed. She points out, for example, that other researchers have shown that stress can enhance learning even as it suppresses neurogenesis.