Everyone knows that going to bed too late can leave you less alert the next day. But recent claims have gone further, promoting the idea that regularly getting too little sleep raises the risk of Alzheimer鈥檚 disease and even shrinks the brain.
Now, this idea has been challenged by one of the most comprehensive analyses of brain scanning studies to date. So how much do we really need to worry about getting enough shut-eye?
Most health bodies, such as those in the US and the UK, advise that adults generally need between 7 and 9 hours of sleep a night. And, thanks to the recent availability of wrist-worn sleep tracking devices, it has never been easier to know if you are meeting this quota.
Advertisement
It is understandable if some people obsess over getting enough sleep. Widely covered animal research 鈥� including in New 杏吧原创 颅鈥� shows that at night, the brain鈥檚 cleaning system ramps up, ridding it of toxic compounds linked to Alzheimer鈥檚 disease, such as a protein called beta-amyloid. And large population studies have generally found that people with unusually short or long sleep durations have worse health by several measures, including memory loss and brain shrinkage, one of the hallmarks of Alzheimer鈥檚.
Typically, there is a roughly U-shaped curve for the relationship between hours of slumber and the chances of a bad health outcome, such as dementia, and an inverted U-shaped curve for measures where a higher value signifies better health, such as brain volume seen on an MRI scan.
But such population studies can only find correlations between sleep duration and health, they can鈥檛 tell us if poor sleep is causing the health problem. Only a randomised trial could do that, but these are almost impossible to do in this case, because few people would agree to change their sleeping habits for any length of time in the name of science, says at the University of Oslo in Norway.
Fjell and his team have now tried to take a deeper look, through a series of studies that used brain volume as a proxy for brain health. First, they looked at brain volume in relation to sleep duration at a single point in time, using existing data on about 47,000 people. Here, they found an inverted U-shaped curve, although the highest brain volume was linked with a surprisingly low 6.5 hours of sleep a night.
The team then carried out a further analysis that tracked about 4000 people for up to 11 years. In this case, there was no correlation between sleep duration at the beginning of the study and brain shrinkage over this period. 鈥淚t would be very surprising to see this, if short sleep had a negative effect on the brain,鈥� says Fjell.
The correlation in the first analysis could be explained by brain shrinkage causing sleep disturbance, rather than the other way round. Alternatively, it could just reflect stable differences between people rather than the result of brain shrinkage, such as that people with naturally smaller or larger brains tend to sleep less for some unknown reason, he says.
The researchers also did a third kind of analysis, using genetic data gathered in one of the studies, for about 30,000 people. This found that those who are genetically predisposed to have either short or long sleep durations don鈥檛 have smaller brain volumes than more moderate sleepers. Put together, this set of results challenges the idea that failing to get enough sleep shrinks the brain, says Fjell.
While Fjell doesn鈥檛 advise anyone to deliberately change their sleeping habits as a result of the findings, he believes that there is much natural variation in people鈥檚 innate sleep requirements and 鈥� unless prevented by circumstances 鈥� the brain will make sure it gets the necessary amount. 鈥淲e speculate that you have a homeostatic drive towards the sleep you need,鈥� he says.
鈥淚f you are feeling tired all day, then you probably sleep too little, and that can have negative consequences,鈥� says Fjell. 鈥淏ut as long as you feel fine during the day, I wouldn鈥檛 worry about whether your sleep is 6, 7 or 8 hours.鈥�
The right kind of sleep
It is unlikely that the team鈥檚 conclusions will convince everyone. at the University of California, Berkeley, who wrote a book called Why We Sleep 鈥� which advises people to get 8 hours a night 鈥� says the key measure of brain health is neuron density, not total brain volume. 鈥淭hose two measures are not perfectly correlated with each other, which tells us that they measure different things,鈥� he says.
Walker also says the crucial measure of sleep quality is how long people spend in deep sleep 鈥� when brainwaves become slower 鈥� rather than their total quantity of sleep. This is supported by at the University of Rochester in New York, who discovered the brain鈥檚 waste-clearing system, which is called the glymphatic system.
Certainly, it is during slow-wave sleep that amyloid clearance from the brain intensifies. However, most such sleep happens during the first 4 hours of the night, Fjell points out. 鈥淚t is unclear whether more sleep than that would further facilitate this clearance,鈥� he says. 鈥淔rom the 鈥榗leaning system鈥� theory of sleep, it does not follow that long sleep is beneficial.鈥�
Fjell and his team didn鈥檛 look at the duration of slow-wave sleep in their study 鈥� indeed, gathering that data on this sort of scale would be a Herculean task, as it requires people to spend the night in a sleep laboratory, hooked up to electrodes on their scalp.
Nedergaard says this is an important piece of work nevertheless, due to its finding that, in the single-time-point analysis, larger brain volume was linked with a mere 6.5 hours of sleep a night. 鈥淭he surprise is that the optimal sleep duration is shorter than what is currently recommended,鈥� she says.
This result suggests, therefore, that even if you are concerned about total sleep duration, you may not need to worry about falling short of 8 hours a night. 鈥淵ou should not lie in bed tossing and turning to meet [that] target,鈥� says at the National University of Singapore, who says the findings are credible due to the enormous amount of data analysed. 鈥淭his study is quite reassuring.鈥�
Nature Human Behaviour
