Why We Sleep

by Matthew Walker

We will return to the issue of sleep and psychiatric illness several times in the course of this book, but schizophrenia deserves special mention at this juncture. Several studies have tracked neural development using brain scans every couple of months in hundreds of young teenagers as they make their way through adolescence. A proportion of these individuals went on to develop schizophrenia in their late teenage years and early adulthood. Those individuals who developed schizophrenia had an abnormal pattern of brain maturation that was associated with synaptic pruning, especially in the frontal lobe regions where rational, logical thoughts are controlled—the inability to do so being a major symptom of schizophrenia. In a separate series of studies, we have also observed that in young individuals who are at high risk of developing schizophrenia, and in teenagers and young adults with schizophrenia, there is a two- to threefold reduction in deep NREM sleep.fn17 Furthermore, the electrical brainwaves of NREM sleep are not normal in their shape or number in the affected individuals. Faulty pruning of brain connections in schizophrenia caused by sleep abnormalities is now one of the most active and exciting areas of investigation in psychiatric illness.fn18

  Adolescents face two other harmful challenges in their struggle to obtain sufficient sleep as their brains continue to develop. The first is a change in their circadian rhythm. The second is early school start times. I will address the harmful and life-threatening effects of the latter in a later chapter; however, the complications of early school start times are inextricably linked with the first issue—a shift in circadian rhythm. As young children, we often wished to stay up late so we could watch television, or engage with parents and older siblings in whatever it was that they were doing at night. But when given that chance, sleep would usually get the better of us, on the couch, in a chair, or sometimes flat out on the floor. We’d be carried to bed, slumbering and unaware, by those older siblings or parents who could stay awake. The reason is not simply that children need more sleep than their older siblings or parents, but also that the circadian rhythm of a young child runs on an earlier schedule. Children therefore become sleepy earlier and wake up earlier than their adult parents.

  Adolescent teenagers, however, have a different circadian rhythm from their young siblings. During puberty, the timing of the suprachiasmatic nucleus is shifted progressively forward: a change that is common across all adolescents, irrespective of culture or geography. So far forward, in fact, it passes even the timing of their adult parents.

  As a nine-year-old, the circadian rhythm would have the child asleep by around nine p.m., driven in part by the rising tide of melatonin at this time in children. By the time that same individual has reached sixteen years of age, their circadian rhythm has undergone a dramatic shift forward in its cycling phase. The rising tide of melatonin, and the instruction of darkness and sleep, is many hours away. As a consequence, the sixteen-year-old will usually have no interest in sleeping at nine p.m. Instead, peak wakefulness is usually still in play at that hour. By the time the parents are getting tired, as their circadian rhythms take a downturn and melatonin release instructs sleep—perhaps around ten or eleven p.m., their teenager can still be wide awake. A few more hours must pass before the circadian rhythm of a teenage brain begins to shut down alertness and allow for easy, sound sleep to begin.

  This, of course, leads to much angst and frustration for all parties involved on the back end of sleep. Parents want their teenager to be awake at a “reasonable” hour of the morning. Teenagers, on the other hand, having only been capable of initiating sleep some hours after their parents, can still be in their trough of the circadian downswing. Like an animal prematurely wrenched out of hibernation too early, the adolescent brain still needs more sleep and more time to complete the circadian cycle before it can operate efficiently, without grogginess.

  If this remains perplexing to parents, a different way to frame and perhaps appreciate the mismatch is this: asking your teenage son or daughter to go to bed and fall asleep at ten p.m. is the circadian equivalent of asking you, their parent, to go to sleep at seven or eight p.m. No matter how loud you enunciate the order, no matter how much that teenager truly wishes to obey your instruction, and no matter what amount of willed effort is applied by either of the two parties, the circadian rhythm of a teenager will not be miraculously coaxed into a change. Furthermore, asking that same teenager to wake up at seven the next morning and function with intellect, grace, and good mood is the equivalent of asking you, their parent, to do the same at four or five a.m.

  Sadly, neither society nor our parental attitudes are well designed to appreciate or accept that teenagers need more sleep than adults, and that they are biologically wired to obtain that sleep at a different time from their parents. It’s very understandable for parents to feel frustrated in this way, since they believe that their teenager’s sleep patterns reflect a conscious choice and not a biological edict. But non-volitional, non-negotiable, and strongly biological they are. We parents would be wise to accept this fact, and to embrace it, encourage it, and praise it, lest we wish our own children to suffer developmental brain abnormalities or force a raised risk of mental illness upon them.

  It will not always be this way for the teenager. As they age into young and middle adulthood, their circadian schedule will gradually slide back in time. Not all the way back to the timing present in childhood, but back to an earlier schedule: one that, ironically, will lead those same (now) adults to have the same frustrations and annoyances with their own sons or daughters. By that stage, those parents have forgotten (or have chosen to forget) that they, too, were once adolescents who desired a much later bedtime than their own parents.

  You may wonder why the adolescent brain first overshoots in their advancing circadian rhythm, staying up late and not wanting to wake up until late, yet will ultimately return to an earlier timed rhythm of sleep and wake in later adulthood. Though we continue to examine this question, the explanation I propose is a socio-evolutionary one.

  Central to the goal of adolescent development is the transition from parental dependence to independence, all the while learning to navigate the complexities of peer-group relationships and interactions. One way in which Mother Nature has perhaps helped adolescents unbuckle themselves from their parents is to march their circadian rhythms forward in time, past that of their adult mothers and fathers. This ingenious biological solution selectively shifts teenagers to a later phase when they can, for several hours, operate independently—and do so as a peer-group collective. It is not a permanent or full dislocation from parental care, but as safe an attempt at partially separating soon-to-be adults from the eyes of Mother and Father. There is risk, of course. But the transition must happen. And the time of day when those independent adolescent wings unfold, and the first solo flights from the parental nest occur, is not a time of day at all, but rather a time of night, thanks to a forward-shifted circadian rhythm.

  We are still learning more about the role of sleep in development. However, a strong case can already be made for defending sleep time in our adolescent youth, rather than denigrating sleep as a sign of laziness. As parents, we are often too focused on what sleep is taking away from our teenagers, without stopping to think about what it may be adding. Caffeine also comes into question. There was once an education policy in the US known as “No child left behind.” Based on scientific evidence, a new policy has rightly been suggested by my colleague Dr. Mary Carskadon: “No child needs caffeine.”

  SLEEP IN MIDLIFE AND OLD AGE

  As you, the reader, may painfully know; sleep is more problematic and disordered in older adults. The effects of certain medications more commonly taken by older adults, together with coexisting medical conditions, result in older adults being less able, on average, to obtain as much sleep, or as restorative a sleep, as young adults.

  That older adults simply need less sleep is a myth. Older adults appear to need just as much sleep as they do in midlife, but are simply less able to generate that (still necess
ary) sleep. Affirming this, large surveys demonstrate that despite getting less sleep, older adults reported needing, and indeed trying, to obtain just as much sleep as younger adults.

  There are additional scientific findings supporting the fact that older adults still need a full night of sleep, just like young adults, and I will address those shortly. Before I do, let me first explain the core impairments of sleep that occur with aging, and why those findings help falsify the argument that older adults don’t need to sleep as much. These three key changes are: (1) reduced quantity/quality, (2) reduced sleep efficiency, and (3) disrupted timing of sleep.

  The postadolescent stabilization of deep-NREM sleep in your early twenties does not remain very stable for very long. Soon—sooner than you may imagine or wish—comes a great sleep recession, with deep sleep being hit especially hard. In contrast to REM sleep, which remains largely stable in midlife, the decline of deep NREM sleep is already under way by your late twenties and early thirties.

  As you enter your fourth decade of life, there is a palpable reduction in the electrical quantity and quality of that deep NREM sleep. You obtain fewer hours of deep sleep, and those deep NREM brainwaves become smaller, less powerful, and fewer in number. Passing into your mid- and late forties, age will have stripped you of 60 to 70 percent of the deep sleep you were enjoying as a young teenager. By the time you reach seventy years old, you will have lost 80 to 90 percent of your youthful deep sleep.

  Certainly, when we sleep at night, and even when we wake in the morning, most of us do not have a good sense of our electrical sleep quality. Frequently this means that many seniors progress through their later years not fully realizing how degraded their deep-sleep quantity and quality have become. This is an important point: it means that elderly individuals fail to connect their deterioration in health with their deterioration in sleep, despite causal links between the two having been known to scientists for many decades. Seniors therefore complain about and seek treatment for their health issues when visiting their GP, but rarely ask for help with their equally problematic sleep issues. As a result, GPs are rarely motivated to address the problematic sleep in addition to the problematic health concerns of the older adult.

  To be clear, not all medical problems of aging are attributable to poor sleep. But far more of our age-related physical and mental health ailments are related to sleep impairment than either we, or many doctors, truly realize or treat seriously. Once again, I urge an elderly individual who may be concerned about their sleep not to seek a sleeping pill prescription. Instead, I recommend you first explore the effective and scientifically proven non-pharmacological interventions that a doctor who is board certified in sleep medicine can provide.

  The second hallmark of altered sleep as we age, and one that older adults are more conscious of, is fragmentation. The older we get, the more frequently we wake up throughout the night. There are many causes, including interacting medications and diseases, but chief among them is a weakened bladder. Older adults therefore visit the bathroom more frequently at night. Reducing fluid intake in the mid- and late evening can help, but it is not a cure-all.

  Due to sleep fragmentation, older individuals will suffer a reduction in sleep efficiency, defined as the percent of time you were asleep while in bed. If you spent eight hours in bed, and slept for all eight of those hours, your sleep efficiency would be 100 percent. If you slept just four of those eight hours, your sleep efficiency would be 50 percent.

  As healthy teenagers, we enjoyed a sleep efficiency of about 95 percent. As a reference anchor, most sleep doctors consider good-quality sleep to involve a sleep efficiency of 90 percent or above. By the time we reach our eighties, sleep efficiency has often dropped below 70 or 80 percent; 70 to 80 percent may sound reasonable until you realize that, within an eight-hour period in bed, it means you will spend as much as one to one and a half hours awake.

  Inefficient sleep is no small thing, as studies assessing tens of thousands of older adults show. Even when controlling for factors such as body mass index, gender, race, history of smoking, frequency of exercise, and medications, the lower an older individual’s sleep efficiency score, the higher their mortality risk, the worse their physical health, the more likely they are to suffer from depression, the less energy they report, and the lower their cognitive function, typified by forgetfulness.fn19 Any individual, no matter what age, will exhibit physical ailments, mental health instability, reduced alertness, and impaired memory if their sleep is chronically disrupted. The problem in aging is that family members observe these daytime features in older relatives and jump to a diagnosis of dementia, overlooking the possibility that bad sleep is an equally likely cause. Not all old adults with sleep issues have dementia. But I will describe evidence in chapter 7 that clearly shows how and why sleep disruption is a causal factor contributing to dementia in mid- and later life.

  A more immediate, though equally dangerous, consequence of fragmented sleep in the elderly warrants brief discussion: the nighttime bathroom visits and associated risk of falls and thus fractures. We are often groggy when we wake up during the night. Add to this cognitive haze the fact that it is dark. Furthermore, having been recumbent in bed means that when you stand and start moving, blood can race from your head, encouraged by gravity, down toward your legs. You feel light-headed and unsteady on your feet as a consequence. The latter is especially true in older adults whose control of blood pressure is itself often impaired. All of these issues mean that an older individual is at a far higher risk of stumbling, falling, and breaking bones during nighttime visits to the bathroom. Falls and fractures markedly increase morbidity and significantly hasten the end of life of an older adult. In the footnotes, I offer a list of tips for safer nighttime sleep in the elderly.fn20

  The third sleep change with advanced age is that of circadian timing. In sharp contrast to adolescents, seniors commonly experience a regression in sleep timing, leading to earlier and earlier bedtimes. The cause is an earlier evening release and peak of melatonin as we get older, instructing an earlier start time for sleep. Restaurants in retirement communities have long known of this age-related shift in bedtime preference, epitomized (and accommodated) by the “early-bird special.”

  Changes in circadian rhythms with advancing age may appear harmless, but they can be the cause of numerous sleep (and wake) problems in the elderly. Older adults often want to stay awake later into the evening so that they can go to theater or the movies, socialize, read, or watch television. But in doing so, they find themselves waking up on the couch, in a movie theater seat, or in a reclining chair, having inadvertently fallen asleep mid-evening. Their regressed circadian rhythm, instructed by an earlier release of melatonin, left them no choice.

  But what seems like an innocent doze has a damaging consequence. The early-evening snooze will jettison precious sleep pressure, clearing away the sleepiness power of adenosine that had been steadily building throughout the day. Several hours later, when that older individual gets into bed and tries to fall asleep, they may not have enough sleep pressure to fall asleep quickly, or stay asleep as easily. An erroneous conclusion follows: “I have insomnia.” Instead, dozing off in the evening, which most older adults do not realize is classified as napping, can be the source of sleep difficulty, not true insomnia.

  A compounding problem arrives in the morning. Despite having had trouble falling asleep that night and already running a sleep debt, the circadian rhythm—which, as you’ll remember from chapter 2, operates independently of the sleep-pressure system—will start to rise around four or five a.m. in many elderly individuals, enacting its classic earlier schedule in seniors. Older adults are therefore prone to wake up early in the morning as the alerting drumbeat of the circadian rhythm grows louder, and corresponding hopes of returning back to sleep diminish in tandem.

  Making matters worse, the strengths of the circadian rhythm and amount of nighttime melatonin released also decrease the older we get. Add these things up,
and a self-perpetuating cycle ensues wherein many seniors are battling a sleep debt, trying to stay awake later in the evening, inadvertently dozing off earlier, finding it hard to fall or stay asleep at night, only to be woken up earlier than they wish because of a regressed circadian rhythm.

  There are methods that can help push the circadian rhythm in older adults somewhat later, and also strengthen the rhythm. Here again, they are not a complete or perfect solution, I’m sad to say. Later chapters will describe the damaging influence of artificial light on the circadian twenty-four-hour rhythm (bright light at night). Evening light suppresses the normal rise in melatonin, pushing an average adult’s sleep onset time into the early-morning hours, preventing sleep at a reasonable hour. However, this same sleep-delaying effect can be put to good use in older adults, if timed correctly. Having woken up early, many older adults are physically active during the morning hours, and therefore obtain much of their bright-light exposure in the first half of the day. This is not optimal, as it reinforces the early-to-rise, early-to-decline cycle of the twenty-four-hour internal clock. Instead, older adults who want to shift their bedtimes to a later hour should get bright-light exposure in the late-afternoon hours.

  I am not, however, suggesting that older adults stop exercising in the morning. Exercise can help solidify good sleep, especially in the elderly. Instead, I advise two modifications for seniors. First, wear sunglasses during morning exercise outdoors. This will reduce the influence of morning light being sent to your suprachiasmatic clock that would otherwise keep you on an early-to-rise schedule. Second, go back outside in the late afternoon for sunlight exposure, but this time do not wear sunglasses. Make sure to wear sun protection of some sort, such as a hat, but leave the sunglasses at home. Plentiful later-afternoon daylight will help delay the evening release of melatonin, helping push the timing of sleep to a later hour.