How Your Brain Keeps Learning While You Rest

When you’re stuck on a tough problem, it’s not uncommon to take a walk or give your brain a break, only to gain that sudden flash of inspiration or insight you were looking for.

“You have a problem to solve; you cannot really find the right solution. Then you take a nap or do something restful, and it clears up,” says George Dragoi, PhD, professor of psychiatry at Yale School of Medicine. “Something happens there, and suddenly the solution comes to you.”

Dragoi and co-author Baburam Bhattarai, PhD, a former postdoctoral fellow in his lab, have now identified patterns of brain activity associated with this “offline” learning process in a recent study published in Nature Communications.

When humans try to solve a problem or learn a new skill, we do more than just encode new information in the brain. We also draw on past experiences and knowledge, connecting old neurons along new pathways to adapt them to new purposes.

Learning to drive a family car, for example, may be an intensive process over three or four weeks in young adulthood. Later in life, learning to drive a new vehicle—like a large, rented moving truck—requires a much shorter period of learning and adjustment.

“You assimilate the new stuff with the old information,” Dragoi says, “and you kind of learn it again. It’s always a process of adaptation.”

A growing body of research suggests sleep and rest play an important role in this process of adapting and recombining previous experiences or information.

“So there are past experiences and separate pieces of information already present,” says Dragoi, “and in sleep the brain is trying to link and put them all together.”

For the study, the researchers trained rats using a simple maze with six possible locations for a food reward, each associated with a particular scent and flavor cue.

Once the animals had learned to associate each of the six scents with a particular reward location, researchers started swapping the scent-location pairs. Researchers first switched out three of the pairings, then the other three, before eventually changing all six.

Dragoi notes that even when all six scent-location pairs had been changed, there were still consistent non-associative rules the rats could depend on: not to visit the same location twice in one trial, for example, or not to bother with the location that had held the food reward in the previous trial. But those were not sufficient to solve the task.

“This task is associative in nature,” he says. “If they don't apply the associative rules, they're not going to get it right. With six possible locations, the probability is too low.”

Whether they were dealing with three new pairs to learn or six, rats were slow to find the newly changed food rewards at first—but researchers found after a session of rest, they were visiting the correct location much more quickly on average.

“It seemed at first like none of them learned,” Dragoi recalls, “and then they went to this one-hour rest, offline, outside this maze. When they came back, they seemed to know them all. It felt like an ‘aha’ moment had occurred.”

With the connection between rest and rapid learning established, the researchers’ next task was to uncover how the brain reconfigures itself while at rest. One hypothesis involves brief “ripples” of coordinated activity in the hippocampus, the brain’s central processing center for memory. These ripples are associated with spikes of activity in the prefrontal cortex, and scientists only observe them during periods of sleep or quiet, wakeful rest.

“Brain activity in sleep is all over the place,” Dragoi says, “but during these ripples, many neurons are coordinated and firing and you have this high-frequency oscillation, which lasts just 40 to 80 milliseconds.” It’s believed these ripples could represent the rapid rewriting and compiling of brain connections during rest.

When researchers used small electrical signals to suppress the ripples as they started to form, the learning benefits of rest disappeared. “We saw stronger learning was associated with ripples, and then we blocked those ripples and we blocked the rapid associative learning,” says Dragoi. “So we know at least that the ripples are needed for this rapid, insight-like learning.”

The cells and synapses of the brain aren’t arranged in rigid, linear paths for each task and experience. “It's a cognitive map,” Dragoi says. “You can take detours and shortcuts even though you've never taken them before. But your past knowledge and experience are the roads that make it possible.”

He says the study could help pave the way for future therapies targeting learning or memory in humans. “They're not the same neurons across animals,” he allows, “but the principles, the way they coordinate phenomena are transferable from one animal to another.”

Dragoi says in the here and now, the team’s results underscore the important role played by sleep, especially for trainees or those learning new skills. “Younger people, students say to me, ‘I'll sleep when I’m dead,’” he says. “But if you don’t sleep, you're not learning very much.”

He says giving the brain a little time off can help it to do what it does best: make and refine connections.

“Just get a little off the problem and then let the brain kind of do the work for you,” he says. “Rest inspires insight.”