Dream a little dream of recall

People who have nap-time dreams about a task that they’ve just practiced get a big memory boost on the task upon awakening, Harvard researchers report.

Those who dream about anything else have no such enhanced recall, the team reports in a paper published online April 22 in Current Biology. Neither do those who stay awake, even if they think about the task.

“I was startled by this finding,” says study coauthor Robert Stickgold, a cognitive neuroscientist at Harvard Medical School. “Task-related dreams may get triggered by the sleeping brain’s attempt to consolidate challenging new information and to figure out how to use it.”

His new findings elaborate on research suggesting that sleep generally enhances memory and learning (SN: 4/28/07, p. 260).

Dreaming about a demanding undertaking doesn’t cause enhanced memories for that experience, Stickgold emphasizes. Rather, memory-fortifying brain processes during sleep cause the dreams, he proposes. During slumber, Stickgold posits, a structure called the hippocampus integrates recently learned information, such as how to navigate a virtual maze, while other brain regions apply this information to related but broader situations, such as how to navigate a maze of job application forms.

That’s a “tempting speculation,” remarks physiological psychologist Jan Born of the University of Lübeck in Germany. Stickgold’s idea has much potential for fostering advances in dream research, Born says.

Stickgold’s group focused on dreams that occur during non–rapid eye movement, or NREM, sleep. Previous studies found links between chemical and electrical activity in the brain during NREM sleep and better learning by rats and people. Neural activity sparked by recent learning has not been observed during rapid eye movement, or REM, sleep, which often includes especially vivid and bizarre dream elements.

In the new investigation, 99 college students age 18 to 30 spent an hour practicing a virtual maze task on a computer. In a series of trials, volunteers navigated through a complex, three-dimensional maze, starting from a different spot each time. They were instructed to remember the location of a particular tree in the maze.

For the first 90 minutes of a five-hour break from practicing, students were assigned either to take a nap or to engage in quiet activities such as watching videos.

Nappers’ electrical brain activity was monitored with scalp sensors. Experimenters questioned the students about their dreams just before they fell asleep, after one minute of continuous NREM sleep and at the end of the nap period. Volunteers who stayed awake recounted their thoughts at the start, middle and end of the 90-minute session.

After lunch and a period of quiet activity, participants reentered the virtual maze at random spots and were asked to find the tree that they had previously tried to remember.

Those who had dreamed of the experimental task — four of 50 nappers — found the tree much faster than they had in initial trials. These individuals described dreams such as seeing people at particular locations in a maze or hearing music that had played in the lab during testing.

All of the volunteers who dreamed about the maze had performed relatively poorly during pre-nap training, Stickgold notes. Memory processes invoked by the sleeping brain may respond most strongly to challenges perceived as difficult and important to solve, he suggests.

Stickgold’s group is now designing a more exciting maze task intended to elicit task-related dreams in a larger proportion of volunteers.

The researchers also plan to examine whether people who have REM dreams about a maze task during a full night’s slumber navigate that maze better the next day.