Parts of the brain may sleep more soundly at night after being exercised by challenging tasks, researchers have found. Moreover, they say, this localized deep sleep improves performance the next day.
Even after decades of studying sleep, scientists can't fully explain why sleep is needed, or what goes on in the brain while we're snoring away. Electroencephalograms (EEGs) reveal five stages of sleep, and researchers suspected that the deepest of these is a period of slow, rolling waves of electrical activity caused when groups of neurons fire in synchrony and then go silent. But little is known about what the oscillations do or why they sometimes vary in strength.
Neurobiologist Giulio Tononi of the University of Wisconsin, Madison, and colleagues created a clever test to help tease out some answers. He asked two groups of people to click targets on a computer screen using a mouse, a task known to stimulate neural activity in a specific part of the brain's right hemisphere. The researchers made the test more challenging for one group by changing the way the cursor responded to the mouse, forcing the group to compensate in order to successfully click the target. That night, both groups were monitored with EEGs during their slow wave sleep.
Surprisingly, there was a difference. In the compensating group, slow waves emanating from the brain area exercised by the test were an average of 25% stronger than waves in subjects with well-behaved mouses. And the compensating group scored an average of 11% better the next morning when tested on performance on the original task. This improvement vanished, however, when the researchers repeated the experiment in another group of subjects who didn't sleep between the training and testing periods, the team reports in the 7 June issue of Nature.
The research demonstrates for the first time that slow wave strength increases in localized areas of the brain after training and helps consolidate memory, says neurophysiologist Mircea Steriade of University Laval in Quebec, Canada. "I never thought it would be possible to [demonstrate] this," he says. "It's a tremendous paper."