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

are one of the leading causes of death in developed nations, and sleepy people are responsible for many of them. A remarkably large proportion of vehicle accidents are the direct or indirect result of tired drivers losing concentration or falling asleep at the wheel. A few examples may give a flavour of the carnage they cause. In March 1994 near Barstow in California, a pickup truck carrying 20 people veered off the road and crashed into a culvert after the driver apparently fell asleep at the wheel. The driver survived but 12 passengers died. In July 1995 near Roquemaure in France the driver of a bus carrying Spanish students from Amsterdam to Barcelona seemed to nod off then wake abruptly as his bus scraped a passing truck. He lost control and the bus swerved wildly before rolling over several times. The accident caused 22 deaths and 32 injuries. In February 2001 a sleep-deprived driver caused the Selby rail disaster in the UK, after he fell asleep at the wheel and his vehicle crashed onto a railway line. Ten train passengers died. The driver, Gary Hart, admitted getting no sleep the night before the crash, but claimed he could still drive safely. He was sent to prison.

      Scientists have judged that sleepiness is a factor in at least 10 per cent of fatal car crashes in the USA and more than 50 per cent of fatal crashes in which a truck driver is killed. A 1994 report by the US National Commission on Sleep Disorders concluded that driver fatigue contributed to 54 per cent of all vehicle accidents in the USA. A comparable situation applies in the UK and elsewhere. Research concluded that at least 10 per cent of vehicle accidents in the UK are related to sleepiness, though some experts have put the figure much higher. Two large surveys in England found that sleepiness was a causal factor in 16 per cent of accidents to which the police were summoned and at least 20 per cent of accidents on motorways. Driving on a motorway is generally more monotonous than driving on a minor road, and monotony heightens the risk that a tired driver will fall asleep at the wheel. Half the drivers involved in these sleep-related accidents were men under the age of 30 and many of the accidents involved truck drivers, company cars or workers returning home from night shifts.

      For every sleepy driver who actually crashes there are uncounted numbers who have had near misses. A large survey by the British Transport Research Laboratory found that 29 per cent of drivers had come close to falling asleep at the wheel within the previous year, while other research established that at least 5 per cent of middle-aged male drivers had actually fallen asleep while driving on several occasions. Not surprisingly, drivers suffering from moderate or severe daytime sleepiness are at least twice as likely to have a vehicle accident.

      The official statistics tend to underestimate the true extent of the sleepiness problem, and it is easy to see why. Drivers who survive crashes are naturally reluctant to admit that they dozed off at the wheel, even if they recollect doing it. And it is difficult to prove legally that sleepiness caused a crash (especially if the driver is dead). Unlike alcohol, drugs or mechanical defects in a vehicle, sleepiness leaves few evidential traces. You can easily measure how drunk someone is at the roadside immediately after an accident. But measuring sleepiness is neither quick nor easy, and in practice it is simply not done.

      The systematic under-reporting of fatigue was highlighted by accident statistics for Italian highways. Over the period from 1993 to 1997 the Italian authorities officially ascribed the cause as sleep in only 3 per cent of accidents. However, by analysing the data in more depth, researchers were able to estimate that sleepiness had probably contributed to about 22 per cent of accidents. If true, this means that the official statistics had underestimated the hazard of driver sleepiness by a factor of seven.

      The time of day is a major element in the relationship between sleepiness and accidents. Thanks to our natural circadian rhythms, we all feel sleepier at certain times in the 24-hour cycle (usually in the early hours of the morning and again in the afternoon) regardless of how much sleep we have had. As expected, sleepiness-related vehicle accidents occur most often in the early hours of the morning and in the afternoon, during these natural peaks in sleepiness. Older drivers are particularly susceptible to afternoon sleepiness, whereas younger drivers are more prone to crashing late at night or in the early hours. When researchers analysed the data for accidents in which the driver had been injured or killed (excluding those involving alcohol) they found that young drivers were between five and ten times more likely to crash late at night than during the morning.

      Driving late at night poses a risk for train drivers as well. In one study, scientists monitored train drivers while they drove the same route, both by day and at night. The train drivers felt much sleepier when driving at night, and physiological measurements mirrored their subjective feelings. Their brain waves, heart rates and eye movements at night were all characteristic of sleepy people. Four of the 11 drivers who were monitored admitted to dozing off during the night journey and two of them failed to respond to signals. Sleepiness has almost certainly caused numerous rail crashes over the years, but again the official statistics have systematically underestimated its importance.

      The sleepiness experienced by many drivers is partly a product of natural cìrcadian variations in wakefulness. But much of the blame rests with simple lack of sleep. And when lack of sleep is combined with driving at odd hours, the effect can be lethal. Prevailing social attitudes towards this issue are frankly perverse. Many parents think nothing of packing their family into a car and then driving long distances to a holiday destination while they are seriously tired. They would be horrified at the thought of doing this while drunk, but the effects of tiredness and alcohol on their ability to drive safely are strikingly similar, as we shall see in the next chapter. The number of drivers involved is large. In August 1996 French researchers assessed the extent of sleep deprivation among drivers during the holiday season, by randomly stopping two thousand cars at tollbooths and interviewing the drivers. It transpired that half of them had slept less than they would normally have done during the previous 24 hours. On average, these happy holidaymaking drivers had slept for 3.4 hours less than normal.

      Many long-haul truck drivers get insufficient sleep, with potentially serious consequences for their performance and safety. When investigators studied truck drivers working in the USA and Canada they found that the drivers spent on average slightly more than five hours a day in bed and got slightly less than five hours’ sleep. This was much less than their self-reported ideal of more than seven hours a day. Nearly half the truck drivers augmented their sleep by napping, but the naps were not sufficient to compensate. Video and EEG brainwave recordings revealed that more than half the drivers had at least one period of drowsiness while they were driving, and two actually fell asleep at the wheel.

      Even changing the clocks can be dangerous. The switch to daylight-saving time each spring reduces the length of one night by one hour, which slightly disrupts sleep routines for the next few nights. The extra sleepiness caused by even this apparently trivial disturbance is enough to generate a statistically significant seasonal rise in traffic accidents. Fatal accidents peak on the day immediately following the changeover. Alcohol-related accidents also rise during the week after the clocks change, probably because the effects of alcohol and sleepiness reinforce each other.

      You might think that changing clocks in the opposite direction each autumn would have the reverse effect, but you would be wrong. When researchers analysed 21 years of US vehicle accident statistics, they found that the switch back from daylight-saving time each autumn was also accompanied by an increase in fatal accidents – despite the fact that in this case everyone got an extra hour in bed. The likely explanation is that many people anticipated the extra hour in bed by staying up even later the night before. This was borne out by the fact that the rise in fatal accidents around the autumn changeover was most marked just before the clocks changed, and especially in the early hours of the morning.

      One of the most alarming aspects of daytime sleepiness is that we can fall asleep briefly without even noticing. You are unlikely to be aware that you have slept unless your sleep has lasted for at least a couple of minutes. Tired people can therefore fall asleep at the wheel of a speeding vehicle for tens of seconds at a time and never even know. Researchers measured this phenomenon by waking volunteers after daytime naps of varying durations and asking them if they had been asleep. (Their sleep was confirmed by objective physiological measures.) After one minute of sleep only 15 per cent of subjects had any awareness that they had been asleep, and only 35 per cent were aware even after five minutes of sleep. The upshot is that so-called microsleeps,