I decided to look into some of the research on sleep and sleep deprivation when I started this polyphasic experiment, and some of the findings I came across are very, very surprising, and the opposite to what you might expect, based on common wisdom. This is a quick review of what I learned. I’ve tried to keep this readable and yet I have been forced to keep some jargon in. I’d really appreciate any feedback on the writing style here…if it’s too jargony, dry or boring let me know by posting a comment. Likewise if it is readable, I’d like to hear that too. Also, to keep this readable I have not added any references in, but post a comment if you wish to know the reference for a study I mention and I’ll tell you.
Most of the concerns I have come across regarding polyphasic sleep have been based implicitly on what is called the recuperation theory of sleep. This theory proposes that during our waking hours, the stability of the body is affected by our activities, and that going to sleep is what restores the body and mind to a fit state and a normal balance. This is what most people think sleep is for. Surprisingly though, the research evidence does not suggest that this is the purpose of sleep.
For instance, if this was true, we would expect a person who had not slept for many days to stay asleep for a long time when they eventually did sleep, since the body would be well out of balance and would require major recuperation. But this is not the case; such people actually tend to oversleep for just one night and then return to their normal schedule, as was the case with Randy Gardner, who broke the world record for wakefulness in 1965. After over 11 days awake, Randy slept for 14 hours, before returning to a normal eight-hour schedule.So theoretically, we do not use sleep to rejuvenate, at least not primarily.
So theoretically, sleep-deprivation should not necessarily cause many ill effects. But that’s theory, what about in practice? Some studies have been done on sleep-deprivation in humans. With only 3-4 hours of sleep-deprivation, we are sleepier, our mood is a little off, and we perform poorly on tests of vigilance and concentration. After a few days of sleep-deprivation, we experience microsleeps – short periods of sleep lasting a couple of seconds, that can occur when standing, or even driving.
However there is controversy over the physiological effects of sleep-deprivation, some researchers say there is no convincing evidence of adverse changes, and although there are some studies that report adverse effects, there are many more showing no negative effects to a range of tests, including on tests of IQ and even physical strength.
What about reducing sleep, not removing it altogether?
So there are some problems associated with sleep deprivation, but perhaps they are not as harmful as you may have thought. This is full sleep deprivation though. What do we know about people who simply sleep less, polyphasically or monophasically?Starting with monophasic sleep reduction, there have only been two studies published. This is because they take a lot of time and money to do, being long-term, and it is no doubt hard to recruit participants for them. The first saw participants reduce sleep to 5.5 hours per night, for 60 days, and suffer only a slight reduction in auditory vigilance when they were given an extensive series of mood, medical and performance tests. So 5.5 hours a night seems to be quite acceptable. In the second study, people reduced their sleep gradually to either 4.5, 5, or 5.5 hours per night, for one year. Although they experienced daytime sleepiness, there were again no detriments found on mood, medical or performance tests throughout. Perhaps a daytime nap would have removed even the sleepiness?In terms of polyphasic sleep, I have not managed to get hold of Stampi’s book “Why we nap” which describes the research on polyphasic sleep. It is out of print now, I can’t access the online versions of it with my ATHENS account, and it is not in my university’s library. However I understand it is available at the British Library, so I may yet get it out and review it at a later time. I hear that polyphasic sleep was replicated in several experiments, which suggests random assignment to me (to prevent just testing people who might be naturally more able to adjust to new sleep patterns). They found no detriments on performance tests, however I do not know how long the experiments ran for.
Why isn’t reduced sleep harmful?
This is a valid question, and the answer seems to lie in the adaptiveness of the human brain. The suggestion is not that eight hours is normal and we are able to adapt to other patterns, but that a shorter amount of sleep is optimal, and that we have adapted to an eight-hour monophasic schedule through modern living. When we sleep, there are different stages that we cycle in and out of. I won’t go into too much detail, but there are four stages, the first is where you experience REM sleep and dreams, and the third and fourth are where the majority of the recuperative effects of sleep occur. REM sleep is also of great importance, and your body will not tolerate a lack of it; in such a case you will immediately enter REM sleep the next time you fall asleep.When sleeping we move from the first stage to the second, to the third, then the fourth, then back through the third, second, and to the first again, and repeat. When the overall sleep time is reduced, the efficiency of the sleep that does occur improves. We experience more of the rejuvenating stage three and stage four sleep, up to around the same level as when sleeping for longer. This is most pronounced when reducing sleep gradually, which suggests it might be best to ease into a reduced sleep or polyphasic schedule, to allow the sleep cycles to adapt. During polyphasic sleep, most of the sleep is in the important stages three and four to begin with, then the cycles adapt to include a more ‘normal’ proportion of REM to stage three and four sleep – however, the different stages are reported to occur in separate naps.
So how much do we actually need then?
After reading this research, and living for a month on a mere four hours of sleep per day, I am pretty much convinced that on average, modern humans sleep too much. I’m sure there’s variation between people on how much time we each need to sleep for, and that some people might need their eight hours. I heard this a lot when talking about my polyphasic sleep schedule “I couldn’t do that, I need my eight hours”, people often said. I expect their evidence for this were a few occasions where they were forced to sleep less because of a night out, work hours, or some deadline. I doubt they have systematically experimented with different amounts of sleep for long periods of time, knowing that they might initially experience extra sleepiness as they adapt to their new schedule. But why would they? Ask anyone how much is a healthy about of sleep, and invariably, they will respond “eight hours”, and probably give you a strange look for questioning such an accepted piece of knowledge. But this bit of folk wisdom is a fallacy, like the idea that we need to drink eight glasses of water each day, or that wisdom positively correlates with age: people sleep too much.When I say ‘too much’, surely I mean ‘more than optimal’, right? I mean, you can waste some of your day by oversleeping, but it can’t actually be BAD, can it?In two long-term studies involving a large amount of participants (over 100,000 men and women in the first, and over 82,000 women in the second), the mortality rate was investigated alongside the average sleep time after a 10-year and 14-year period respectively.
The researchers controlled for various factors including age, smoking, snoring, illness, and others. In both studies, they found that sleeping for seven hours was associated with the lowest death rates, and the other sleep-times were then compared to that. I’ve drawn up two graphs so you can see the results. On the bottom axis is number of sleep hours per night, and from there you can see the death rate for that amount of sleep. With seven hours being the lowest, all the other scores are shown as percentages of that.
The first graph shows that the mortality rate was lowest for people sleeping seven hours per night, closely followed by five and six hours per night. Four hours or less, or eight hours or more, were associated with the highest mortality rate, and people sleeping 10 hours or more per night showed a mortality rate that was DOUBLE that of people sleeping seven hours.
The second graph shows similar results. Only women took part in this study, and those sleeping six or seven hours per night showed the lowest mortality rate. Next were five and eight hours, and even higher was nine or more hours, which showed an increase in mortality of over 50%!When I started this experiment, I would often disclaim my actions when talking about it, saying “I’m aware that it’s a bit insane and probably unhealthy, I just want to try it, and anyway, it’s only for a month”. But now, maybe it’s a little closer to sanity than I first thought.
Of course, it’s impossible to say anything about polyphasic sleep schedules from these studies, as the people in them were all monophasic, but they at least say something about our common ‘knowledge’ about sleeping eight hours per night, and especially to the people we all know who get 10 or more hours per night – maybe they should cut down.Anyway, far be it from me to suggest something that might increase human longevity. Maybe I shouldn’t mention this – we’re overcrowded enough on this planet as it is. Instead, maybe we should lionise these people who sleep in for long periods – buy them sedatives, comfortable beds, take them to Sigur Ros concerts, and celebrate them as they heroically sacrifice themselves in order to free up resources for the rest of us. Sleep tight!