In my experience, people who report waking up in the middle of the night most commonly do so around 3 a.m. This alone may not be very significant, but many people with insomnia say that they have a hard time falling back asleep. We usually attribute this to being an awakening that occurs after the deep sleep of the night has ended and the time of increasing dream sleep has started. At the end of a sleep cycle, we are almost awake and it may not take much, even a full bladder or a sound, to fully wake us.
Some people may fall back asleep and not even remember the brief awakening. Others may take a trip to the bathroom and quickly return to sleep. If that doesn’t occur, it is usually because excessive thinking or negative cognitive processes begin. Worrying about not sleeping results in further arousal, which makes it difficult to fall back asleep, which causes more worry that leads to more arousal, and so on. After an hour or so of tossing and turning the person will probably fall back asleep until their alarm goes off.
The relationship between excessive nocturnal mental activity and over-arousal that leads to difficulty staying asleep seems clear. But what if there is more going on?
There is good reason to believe that this explanation is not the full story. The evidence comes from two different sources: One is historical and the other is based in research done in the 1990s. We will return to this evidence in a bit.
I remember being in graduate school and being told that humans don’t sleep like wolves or cats. As anyone who has a feline companion knows full well, frequent naps are a part of a cats' daily routine. Humans, as everyone knows, are typically awake for 16 hours straight and then asleep for eight—one long period of wakefulness and one long period of sleep.
Seems simple enough.
But in certain areas of the world people have a siesta, a regular afternoon rest period that often includes a nap. And they seem to do just fine. In fact, this may be a better way of living than knocking back a 16-ounce energy drink and going for a brisk walk around the office to try and get through the afternoon slump that frequently happens around 2 p.m. or so. If you have traveled to a developing country, you may have noticed that people seem to be able to nap whenever they can, and that they seem to find this refreshing.
If humans seem to need a break for sleep in the middle of the day’s waking period, what about needing to be awake for a period of time in the middle of the night’s sleep period? After all, we generally consider a brief awakening for a trip to the bathroom to be fairly normal. But how normal is it to be able to quickly fall back asleep?
As I mentioned above, there is evidence, and from very different directions, that quickly falling back to sleep after a middle-of-the-night awakening may not be natural for humans. It could be that, left to our natural inclinations in a 24-hour day, humans would have a period of wakefulness, an afternoon siesta, another period of wakefulness, a period of sleep, a period of wakefulness in the middle of the night, and another period of sleep. It may be biologically normal to be up in the middle of the night.
This idea was first introduced to the public, as far as I can determine, in an excellent 2012 BBC article. The article briefly reviews several of the major issues involved in what is known as segmented sleep. I was introduced to it at a seminar during the 2013 national sleep conference entitled, “One Sleep or Two? The History and Science of Segmented Sleep.”
Ekirch (2001, 2006, 2016), has done important work in documenting different sleep patterns in the preindustrial world. The historical evidence indicates that people in the Middle Ages were up for an hour or more in the middle of the night and thought of sleep as occurring in two segments: first sleep and second sleep. In many ways, this makes sense because being awake during the night has certain advantages. At that time, one could stoke the fire, check the defenses, have sex, and tell tall tales. We don’t think much about it now, but before gas and electric lighting, the night could be a frightening time. Think about the dangers that might await you if you ventured out in the night without a good source of illumination. In the country, you could fall in a ditch, be attacked by an animal, be attacked by robbers, or simply never be heard from again. In the city, you could also be attacked and never heard from again. Getting up and making sure your residence was warm and safe made a lot of sense.
When the industrial revolution occurred, it became necessary to regiment our lives. In the preindustrial era there was little reason to stay up after dark—most people had no easy source of illumination beyond a candle or cooking fire and, again, it was scary outside. With the development of factories and the need for regular work shifts, people had to get their sleep when they could—usually at night. So, by staying up throughout the day, sleep drive was sufficient to get deep enough into sleep that people stayed asleep throughout the night. We are all familiar by now with the impact that light at night has on our circadian rhythm, and how we now stay up later and sleep less than in the past.
It makes biological sense to have segmented sleep. If the mid-afternoon siesta has some roots in our evolutionary history, it may be that it served to conserve energy resources and help our ancestors lay low during the hot mid-day sun on the savannas of Africa where weak, slow, formerly tree-inhabiting humans were very vulnerable. At night, they would have been extremely vulnerable in deep sleep, so a period of wakefulness during which they could make sure they were safe made sense.
It's not just historical data that supports the idea of segmented sleep. Further evidence emerged from work on circadian rhythm. Wehr (1991, 1992) conducted experiments showing that humans could take on a biphasic sleep pattern. This occurred when experimental participants were exposed to “winter” conditions with a forced schedule of 10 hours of light and 14 hours of total dark. (Note that total dark is not something most of us are used to. A look around almost any modern bedroom will find many electronic sources of light.) When this schedule was imposed, participants began to naturally adopt a sleep schedule in which they slept for four hours, awakened for one to three hours, and then slept for another four. It seems that, under the right environmental conditions, this could be a natural sleep pattern for humans.
Exposure to light is the most important factor in lengthening our days and then (more or less) sleeping through the night. Our sleep patterns have certainly been affected by the invention of electric lighting and, before that, but to a lesser degree, by gas lights such as those used to light the streets of Paris since the 1600s.
The upshot is that we still have a lot to learn about sleep and insomnia. For individuals who wake up in the middle of the night, the realization that their awakening may just be a throwback to an earlier sleep pattern may reduce some of the frustration and anxiety they feel. This is part of the cognitive behavioral technique of cognitive restructuring, or reducing stress by having more accurate information about sleep. It is also possible that we need to think about other patterns of sleep and not confine ourselves to assuming that the regimented, factory-shift-driven model we’ve all come to know is what we should to strive for.
There is likely more than one way for the Sandman to welcome us to the Land of Nod.
Ekirch, A. R. (2001). Sleep we have lost: Pre-industrial slumber in the British Isles. The American Historical Review, 106 (2), April, 2001, pp. 343 – 386.
Ekirch, A. R. (2006). At Day’s Close: Night in Times Past. New York: W. W. Norton & Company, Inc.
Ekirch, A. R. (2016). Segmented sleep in preindustrial societies. Sleep, 39 (3), p. 715 -716.
Wehr, T.A. (1991). The durations of human melatonin secretion and sleep respond to changes in daylength (photoperiod). Journal of Clinical Endocrinology and Metabolism, 73(6), pp. 1276 - 1280.
Wehr, T.A. (1992). In short photoperiods, human sleep is biphasic. Journal of Sleep Research, 1(2), pp. 103 - 107.