Even if the content of your dreams isn't hot or steamy, slipping into rapid eye movement (REM) sleep might still warm you up from the inside, according to a new review.

In nature, warm-blooded creatures with lower body temperatures tend to have longer periods of REM sleep; while those with higher body temperatures, like birds, experience less REM sleep overall.

Neurologist and leading sleep scientist Jerome Siegel, from the University of California Los Angeles, argues the association is noteworthy and should be investigated further.

Siegel argues REM sleep might be sort of "like shivering for the brain" when brain and body temperatures drop too low during non-REM sleep.

During REM the brain becomes highly active, which raises the temperature of the organ. What's more, REM sleep almost always follows non-REM sleep, which is when the brain and body are least active and cold.

"REM sleep might be thought of as a thermostatically controlled brain-heating mechanism, which is triggered by the temperature reduction linked to the reduced metabolism and the decrease in energy consumption in non-REM sleep," writes Siegel.

"Then, REM sleep ends after the amount of REM required to raise brain temperature to close to the waking temperature of the body has occurred."

In fact, that might be why some animals show fluctuations in sleep duration from season to season. The most extreme example of that is hibernation, but even non-hibernating animals like Arctic reindeer sleep 43 percent more in winter than in summer. Humans in hunter-gatherer societies sleep about an hour longer in winter months as well.

Could REM sleep help protect animal brains from the cold while still allowing them crucial rest time?

Siegel thinks it's entirely possible, especially since other hypotheses around REM sleep have proved imperfect.

Some scientists, for instance, have suggested non-REM sleep helps clear toxins from the brain, while REM sleep helps enhance memory and learning, possibly by pruning back neural connections to make the brain more efficient.

But here's the confusing thing: In almost all mammals, non-REM sleep is followed by REM sleep, which is a state of very high brain activity, similar to waking. This would mean that right after toxins and synapses are cleaned up in the brain, they would simply be re-created.

What's more, there's no obvious relationship between REM sleep duration and cognitive power, which suggests its potential role in learning may be overstated. Platypuses, for instance, experience up to 8 hours of REM sleep a night – longer than any other animal, including humans. It's hard to argue the platypus might need this sleep stage for extra brain efficiency.

On the other hand, this bizarre creature is a monotreme – a sort of middle ground between a cold- and warm-blooded animal. Under Siegel's hypothesis, this means the platypus would require more REM sleep to instead maintain a functional brain temperature as it dozes.

REM sleep may therefore have initially evolved as a way for endotherms to keep their brains warm and functional in case they are awakened by a threat.

As mammals that show no signs of undergoing REM sleep, dolphins might be an exception that proves the rule. These anomalies are thought to partake in uni-hemispheric sleep, where only one side of the brain falls asleep at a time. In these exceptional cases, the brain's temperature might not be as easily influenced by sleep because a 'space heater' is still going in one part of the 'room', reducing the need for episodic warm-ups.

Migrating birds, on the other hand, do show some signs of REM sleep, in spite of also partaking in uni-hemispheric sleep. But because this stage of sleep involves both sides of the brain, these kinds of birds only slip into REM for very brief periods of time. As you can imagine, flying with an inactive brain could be very dangerous.

Siegel thinks the uni-hemispheric exception can be tested further in fur seals, which sleep both sides of their brain on land and only one side in the water.

The idea of REM sleep keeping the motor running for animals like us is complicated by the ongoing debate surrounding REM sleep in cold-blooded reptiles, which though yet to be confirmed also can't yet be ruled out.

Sleeping to save energy is crucial, but animals need to make sure they can still awaken to a threat. If Siegel is right, REM sleep might be a hot new solution to an age-old conundrum.

The study was published in The Lancet.