Uncategorized

Learning while you sleep – How the brain stores important information

A new study shows how people learn in their sleep and how sleep phases promote this. Two different sleep phases during deep sleep seem to play an important, complementary role in learning. One improves overall performance, while the other stabilizes what the sleeping person learned the day before.

How can people learn in their sleep?

young woman in bed with sleeping mask and book

Scientists have long known that good sleep works wonders for the ability to learn new things. What was less clear, however, is the role of different stages of sleep. In particular, there has been controversy over the relative contributions of rapid eye movement (REM) sleep, when most dreams occur, and non-REM sleep, which is largely dreamless. A study by psychologists from the Institute of Cognitive, Linguistic and Psychological Sciences at Brown University provides important clues that could help resolve the debate. Most of her experiment focuses on visual learning. It suggests that it is not one level that is important than the other in learning new skills, but that neurochemical processing plays both an essential and complementary role.

fell asleep while studying with books and pencil next to alarm clock

The team found that while the non-REM phase improves the performance of newly acquired skills by restoring flexibility, REM sleep stabilizes those improvements. In addition, this sleep phase prevents the new information from being overwritten by subsequent learning. Most REM sleep occurs in the last hours of sleep in such a way that the finding reinforces the importance of not shortening these later stages. When people sleep at night, there are many sleep cycles. REM sleep occurs at least three, four, five times, especially in the later part of the night. In addition, people need a lot of REM sleep so that they can better remember what they have learned. Therefore this sleep should take place continuously.

Benefits of sleep for learning

scanning of brain activity while sleeping with electrodes and sensors

Psychologists previously identified two distinct benefits of sleep for learning. They often refer to the first benefit as “offline performance enhancement”. This is that the pre-bedtime learning improves after sleep without additional training. The second benefit, considered resilience to disturbance, protects skills learned before bed from being disturbed or overridden by subsequent learning after waking up. To take advantage of both, there is a trade-off between flexibility and stability. Learning during the day involves the formation of new synapses, which are the electrical connections between nerve cells. Accordingly, strengthening existing synapses depends on repeated use. As people learn while they sleep, the brain seems to streamline its operations in order to work more efficiently.

sign showing in sleep learning symbolic representation

One leading hypothesis suggests that the brain reactivates synapses that become stronger as the day progresses or all of them indiscriminately weaker. This restores the flexibility or plasticity of the brain’s local connections and wider networks to improve overall performance. At the same time, the brain has to stabilize important synapses during sleep in order to prevent what was learned the previous day from being erased by new learning experiences.

The researchers used electrodes that were taped to the subjects’ eyelids and scalp. This enabled them to determine when the test subjects entered different stages of sleep. They also used a technique called magnetic resonance spectroscopy. This allows the relative concentrations of two neurotransmitters (glutamate and gamma-aminobutyric acid (GABA)) that process visual information to be measured in parts of the brain. All in all the results suggest, that both phases of sleep are essential to learning new things. While the brain is “offline”, non-REM sleep improves performance on newly learned tasks. However, without REM sleep to stabilize memories, those gains are lost.