Close your eyes, go to sleep, and let your brain shore up memories of the places you’ve visited recently and the routes you took to get there.
That’s the implication of the latest study to explore sleep’s role in learning and recall.
A team of neuroscientists led by Philippe Peigneux of the University of Liège in Belgium studied volunteers who, during an afternoon session, learned to navigate between landmarks in a computer-generated virtual town. Participants who later that night during a particular stage of deep sleep displayed pronounced activity in the hippocampus—a brain structure already tied to memory—also showed superior recall the next day of travel routes in the virtual town.
“Enhanced hippocampus activity during slow wave sleep reflects the processing of memory traces, which eventually leads to an improvement in performance the next day,” the scientists conclude in the Oct. 28 Neuron. Peigneux’s group used a positron emission tomography (PET) scanner to measure blood flow throughout the brains of 24 men ages 18 to 30 while they were awake or while they slept.
Earlier investigations in rodents had indicated that specific patterns of hippocampal activity in animals learning spatial tasks reappear during slow-wave sleep, a stage of deep sleep, but not during rapid eye movement (REM) sleep.
To establish how the virtual-town task affected waking-brain activity, the researchers had 12 volunteers use a joystick to explore the virtual town for 4 hours. They then immediately underwent PET scans while trying to quickly retrace routes between various landmarks in the virtual town. Men who best remembered how to get from one place to another displayed substantially greater hippocampus activity than did men with poorer recall.
Another six participants completed the same training and memory testing and then wore nets of electrodes on their heads while sleeping that night. Those contraptions enabled the researchers to monitor volunteers’ sleep stages, providing guidance about when to administer PET scans to the slumberers. Furthermore, six men who didn’t undergo any virtual-town training also received brain monitoring and PET scans while they slept.
The volunteers who had explored the virtual town displayed elevated hippocampus activity during slow-wave sleep, compared with the untrained men. Trained participants who showed the most pronounced hippocampus activity while sleeping were the best the next day at remembering routes.
The hippocampus activity in trained participants reflects neural operations involved in spatial memory, Peigneux’s team argues. Previous research had shown that other types of memory tasks activate brain areas outside the hippocampus, the researchers note.
The new results fit with a theory, proposed by Robert Stickgold of Harvard Medical School in Boston and others, that slow-wave sleep mainly fosters memories of personally experienced places and events, such as the virtual town, whereas REM sleep facilitates a slower process of acquiring memories for complex actions and procedures.
Robert P. Vertes of Florida Atlantic University in Boca Raton regards Peigneux’s report and others like it (SN: 10/11/03, p. 228: Restoring Recall: Memories may form and reform, with sleep) as insufficient for establishing a connection between sleep and memory. Well-practiced tasks, such as navigating in a virtual town, stimulate brain activity so strongly that it rebounds later when people sleep and after they wake up. It’s more like a neural echo than a memory, Vertes suggests.