Counting Sheep: The Science and Pleasures of Sleep and Dreams. Paul Martin

of only 6.2 hours. Similar conclusions have emerged from other countries. To give just a few illustrations, a survey in Poland found that adults there slept for 7.1 hours on workday nights, while Korean university students averaged 6.7 hours. When the Japanese Ministry of Health conducted a large survey they discovered that almost two thirds normally slept less than seven hours a night, and more than a quarter slept less than six hours. Worse still, half of Japanese high school students reported sleeping six hours or less on weekdays.

      The research leaves little doubt that most adults in the USA, UK and other industrialised nations get substantially less than eight hours’ sleep most nights of the week, and many get less than seven. But does that matter? How much should we sleep?

      The mythical inhabitants of Sir Thomas More’s idyllic island state of Utopia accorded sleep the priority it truly deserves. They slept for eight solid hours each night. Of the remaining 16 hours, work accounted for only six. (Sensible people.) The Utopians worked for three hours before noon then ate lunch; after lunch they rested for two hours, worked for another three hours then ate supper. They went to bed at about eight in the evening and slept for eight hours. The rest of each day they did as they pleased. Not quite the urban chic lifestyle, perhaps, but a refreshingly different perspective on life.

      Conventional wisdom still holds that we need about eight hours of sleep a night. For most of us, however, the reality falls far short of the Utopian ideal, except perhaps at weekends and when we are on holiday. But is an average of, say, six or seven hours a night enough? Obviously, we cannot use Thomas More’s sixteenth-century fantasy as a scientific yardstick. So how do we judge the adequacy of sleep?

      One approach is to ask people whether they think they are getting enough sleep. A recent study of more than twelve thousand adults did just that, and found that 20 per cent of them felt they were not getting sufficient sleep. Unsurprisingly, one of the factors most strongly associated with insufficient sleep was working long hours. Further evidence came from the Planet Project, said to be the largest opinion poll ever carried out. In 2000, an Internet-based survey was conducted with 1.26 million people in 251 countries. When asked how much sleep they needed in order to feel rested, 47 per cent of people replied eight hours or more. But when asked how much sleep they actually got, only 15 per cent reported sleeping eight hours or more, while 8 per cent said they got less than five hours a night. Many respondents reported having gone without sleep for long periods, whereas a mere 6 per cent said they had never missed a night’s sleep.

      Six or seven hours of sleep a night is probably not enough for many people on a long-term basis. The experimental evidence suggests that the underlying sleep tendency for a typical healthy adult – that is, the amount of sleep they would take if completely liberated from work schedules and other constraints – is more like eight or eight and a half hours a night. This implies a shortfall of around an hour and a half each night. A shortfall of this size almost certainly matters. Restricting someone’s sleep by an hour and a half for just one night will measurably reduce their daytime alertness. The cumulative effects, when sleep is short-changed night after night, are far more pervasive. We shall be looking at the psychological and physical consequences of chronic sleep deprivation in the next two chapters.

      A substantial nightly shortfall in sleep is difficult to sustain for more than five or six days in succession without sleepiness seriously impairing daytime alertness, mood and performance. This aspect of human biology might conceivably have contributed to the establishment of the seven-day week, comprising five or six days of work followed by one or two days of legitimised rest, as a standard unit of time. Most humans have been following the seven-day pattern for thousands of years. The seven-day week dates back to the pre-Biblical Sumerian and Babylonian civilisations. Right from the outset, one day of the week was deemed to be a day of rest and recreation. The Babylonians originally named their seven days after the five visible planets plus the Sun and the Moon, but the choice of seven otherwise has no objective basis in astronomy or any obvious feature of the physical environment. The Romans later adopted the seven-day week, dabbled with eight, then reverted to seven. The seven-day week appeared in the Bible, again with the inherent concept of a day of rest. According to the Biblical account of the Creation, God laboured for six days and rested on the seventh. Nowadays, people who work long hours and sleep for only six or seven hours a night really do need that extra time in bed at weekends to stave off sleep deprivation.

      The thesis that many people are chronically sleep-deprived is not without its sceptics, however. Yvonne Harrison and Jim Horne at Loughborough University have argued that we all have the capacity to sleep more than we usually do, but only in the way that we can carry on eating after our physiological need for food has been satisfied. In support of their sceptical position, Harrison and Horne have cited, for example, an experiment in which healthy young adults slept for up to ten hours a night for two weeks, getting an extra hour or so of sleep each night. This additional sleep produced some improvements in their reaction times and a slight reduction in daytime sleepiness. However, there were no significant improvements in the volunteers’ subjective ratings of their own mood or sleepiness.

      Scientists are paid to be sceptical, and counterblasts are an essential part of scientific debate. Nonetheless, such arguments must be set against all the other strands of evidence showing that chronic sleep deprivation is a real and widespread phenomenon. We have considered some of that evidence and there is more to come. First, though, what about you?

Are you sleep-deprived?

      ‘… Are you going to bed, Holmes?’

      ‘No: I am not tired. I have a curious constitution. I never remember feeling tired by work, though idleness exhausts me completely.’

      Sir Arthur Conan Doyle, The Sign of Four (1890)

      There is no universal, one-size-fits-all figure for the right amount of sleep. Individuals differ considerably in their sleep requirements. The conventional standard of eight hours a night is somewhat arbitrary, though not far off the mark for most people. But the best yardstick for you is your own preferred sleep duration. You can measure this by conducting a simple, if time-consuming, experiment on yourself. You will need two weeks of complete freedom from the tyranny of work schedules and alarm clocks, which means you will probably have to wait until your next long holiday (and possibly much longer if you have small children).

      What you do is this. Every night for about two weeks, go to bed at approximately the same time. Make a note of the exact time just before you turn off the light to go to sleep. Then sleep to your heart’s content until you wake spontaneously the next morning – not with the aid of an alarm clock. Note down the time when you woke up and, ideally, a more precise time for when you think you fell asleep the night before. (All being well, this should have been within 10–20 minutes of turning out the light.) Your time of waking should be when you first became fully conscious, not when you eventually stumbled out of bed. Then calculate how long you spent sleeping. The experiment will work much better if you do not roll over after waking and doze for a further hour or two before getting up.

      Ignore the figures for the first few days, because you will probably be sleeping longer to compensate for your prior sleep deficit. We tend to feel sleepy on relaxing holidays and at weekends because reality has finally caught up with us; when the usual pressures and stimulation that keep us going through the working week are suddenly removed, we sleep more to catch up on the backlog. This extra sleeping is only transitory, however. Once you have caught up and reached equilibrium you should start to feel livelier and more energetic during the day. Unfortunately, by that time the holiday has usually ended and the normal regime of late nights and early mornings resumes.

      After a few days, your nightly sleep duration should settle down to a stable figure. Take the average over the final few days: this represents your preferred sleep duration. It should be somewhere between seven and nine hours, although not everyone fits this pattern. Unless you are elderly or have an unusually relaxed lifestyle, your preferred sleep duration will probably be longer than the time you normally spend sleeping. If the difference is very large – say, two or three hours a night – you could be storing up serious trouble