Defining states of alertness in a clear and precise manner is key to initiating widespread education on sleep deprivation.
DementIn the course of dealing with the many public policy issues involving sleep, sleep deprivation, and sleep disorders, it has become clear to me that the number one sleep-related problem for our society is sleep debt. Furthermore, public awareness and resolve to deal with the problem are greatly hampered by the lack of a precise and accurate language to describe how we feel in the day in relation to varying amounts of sleep loss at night.
Before attempting to justify these conclusions, we should all be on the same page with regard to the meaning of the term, “sleep debt.” Each of us has a specific sleep requirement—an amount that we must obtain each night on the average. All amounts less than this constitute lost sleep, and the size of the debt is the total of all nights of lost sleep. The available scientific evidence solidly supports this conclusion, and the use of the term sleep debt has gradually become part of the sleep professional’s lexicon. For example, a major symposium at the 2000 meeting of the Associated Professional Sleep Societies was titled “Sleep Debt: New Experimental Evidence for Neurobehavioral and Physiological Effects of Chronic Partial Sleep Loss.” However, this should not obscure the fact that more research is needed. Does sleep loss continue to accumulate when undersleeping is a nightly occurrence over weeks or months? Is there finally an adaptation? Do even a few minutes of undersleeping each day eventually add up to a large sleep debt? Does the deeper, less fragmented sleep that follows sleep deprivation have more recovery value than lighter sleep?
Why sleep debt is the number one problem
First, sleep debt is particularly dangerous because it is invisible. The number of adults, adolescents, and children who clearly understand that lost sleep relentlessly accumulates as a debt is extremely small. People are usually taken completely by surprise when an irresistible tendency to fall asleep overcomes them in a soporific situation (behind the wheel). Perhaps even worse, the sleep disorders associated with “excessive daytime sleepiness” also remain essentially invisible even though there has been a great deal of media attention.
Second, the vast majority of adolescents and adults have some degree of daytime sleepiness. For example, in a study of several hundred individuals who should have represented the most alert segment of the population, Levine et al1 found that only 10% had an Multiple Sleep Latency Test (MSLT) score above 15.
Third, the sleep debt problem is camouflaged by erroneous beliefs and misattributions. In addition, many people seem to be somewhat confused about when they are fatigued versus when they are sleepy and when they are not fatigued or not sleepy. In other words, people need to be more in touch with their feelings.
I have chosen to make this third issue and its role in our sleepy society the focus of the following editorial remarks.
Understanding subjective consequences
Those of us who are interested in the sleep-related determinants of daytime function justifiably regard the development of the MSLT as a major scientific advance.2 This objective measure of waking sleep tendency has shown a satisfyingly predictable relation to daily amounts of sleep over time, which, in turn, has enabled us to understand and quantify the dimension of waking alertness/sleepiness and its nocturnal determinants. Dealing with the subjective side of the alertness/sleepiness coin has not been nearly as satisfactory. Correlation of subjective data with objective data is often poor, and subjective ratings are very susceptible to extraneous influences.
In the earliest compilation of data that related scoring levels on the Stanford Sleepiness Scale (SSS) to daytime sleep latencies, the results were far from perfect.3 In the first place, the mode of constructing the SSS required selecting those descriptors that represented equal intervals.4 As a result, the language is somewhat arbitrary and possibly even a bit arcane. If the language is ignored and subjects regard the seven scale levels as a continuity, the SSS is then almost identical to an analogue sleepiness scale, which requires individuals to rate their subjective sleepiness by checking a point on a line extending from “most alert” to “most sleepy.”
Evaluating subjective sleepiness in research projects has been relatively productive because protocols are highly standardized, subjects are thoroughly instructed in what to do, and many potentially confounding influences are eliminated. On the other hand, I believe there is a huge problem in the real world in communicating about levels of sleepiness/alertness with patients, health professionals, and the general public. In a nutshell, the current situation is this. MSLT scores have become extremely meaningful to sleep researchers and a range of assumptions about individuals with a certain score is readily made. If an individual’s mean MSLT score is 2, we know that person is sleep deprived or may have a sleep disorder and is certainly impaired. Likewise, if an individual’s score is 18, a satisfyingly predictable but opposite range of assumptions is readily made. Someone who rates himself or herself as 3 or 4 on the SSS is simply a puzzle.
Sleepiness vs Drowsiness
Perhaps not surprisingly, the number one complaint among college students is “tiredness.” We have surveyed more than 7,000 undergraduates at several universities and have also surveyed about 1,000 high school students in the surrounding area. Analysis of the survey data has yielded several important results:

  • college students and high school students feel tired more often during the day than they feel sleepy. They also feel tired on more days per week than they feel sleepy;
  • the majority of respondents believe that tiredness is not the same as sleepiness, and that tiredness has a different cause. In my opinion, the most impressive example of confusion about tiredness and sleepiness is captured by the statement “I’m too tired to fall asleep.” In the college population, this misattribution is commonplace because students do not equate tiredness with being sleep deprived and they are mainly unaware that strong clock-dependent alerting in the late evening can prevent falling asleep. Thus, the very sleep deprived, tired student tries to sleep and cannot, and erroneously labels this situation as being “too tired to fall sleep.” Although this is essentially nonsense, it certainly serves to foster the spurious belief that feeling tired and feeling sleepy are two entirely different conditions with different causes;
  • the words “sleepiness” and “drowsiness” mean the same thing to about three quarters of the students, but when they are not regarded as equal, drowsiness is more often felt to be stronger, more intense, or closer to sleep than sleepiness;
  • when asked to select the physical signs of sleepiness or drowsiness, the great majority (approximately 80-90%) select symptoms involving the eyes (“heavy eyelids,” “can’t keep my eyes open,” “eyelids close without intending,” “can’t focus,” and “eyes burn”).

Because college students (and possibly many other adults) do not attribute the feeling of tiredness to inadequate sleep, they inevitably underestimate their personal sleep requirement and they inevitably undervalue adequate sleep in determining how they feel in the daytime.
Daytime Consequences
Ever since Mary Carskadon, MD, and I carried out our first study of partial sleep restriction in the Stanford Summer Sleep Camp and showed that the daytime consequences accumulate,5 I have exhorted students to get more sleep with only marginal success. More recently, I have carried out an exercise for credit in my Stanford University undergraduate course, Sleep and Dreams, which requires students to lower their sleep debt by getting as much extra sleep as possible. As expected, students who are able to obtain large amounts of extra sleep no longer feel sleepy in the daytime. But more important, they no longer feel tired. Furthermore, if these same students, for whatever reason, stop obtaining adequate sleep, they soon start feeling tired again. From this, we have concluded that “feeling tired” is entirely the consequence of sleep debt. Certainly this conclusion is quite plausible in a young and physically healthy student population.
In 1995, we carried out a pilot study in which we asked subjects to describe their feelings continuously over a relatively brief interval of time. The designated interval began immediately after subjects carried out enough physical exercise to make them “wide-awake” and ended when they fell asleep. The subjects generally described changes in the way they felt in terms of somatic associations. From these observations, we derived the generalization that people can identify a moment when they become aware that they must make an effort to stay awake. This moment usually occurs within a few minutes from the onset of sleep. In addition, it is at this point that most subjects said they had become sleepy or drowsy.
We then consulted Webster’s Unabridged Dictionary and found that the definition of “drowsy” included “half asleep” and “sluggish” whereas the definition of sleepy included “inclined to sleep” and “languorous.” These definitions, though somewhat overlapping, seem to suggest that the likelihood of falling asleep should be a little greater in a drowsy person as opposed to a sleepy person. Even so, as we found in our surveys, most people consider the two words to have exactly the same meaning.
All of this suggests that our commonplace English language is deficient when it comes to expressing the subjective nuances of quantitatively different states of sleep deprivation. Recognizing this, I have recommended elsewhere6 that the subjective consequences of sleep debt can be divided into two categories. One category should be the feeling of drowsiness and the other should be the feeling of sleepiness. Based again on survey data, the term drowsiness should be applied to the period of wakefulness immediately prior to sleep onset when a person will soon fall asleep unless a conscious effort is made to avoid it. The onset of the drowsy state is defined as the moment of the first awareness that the eyelids feel heavy and a conscious effort is required to keep them open and to stay attentive. The second category refers to the period immediately prior to the onset of the drowsy state. During this time, people are definitely not completely wide-awake and at peak alertness. They are much more likely to say they are feeling tired, fatigued, lethargic, or unmotivated. We would say these feelings are the same as feeling sleepy. Accordingly, when we are neither unambiguously wide-awake nor unambiguously drowsy, we are sleepy.
Levels of Alertness
Although the aforementioned terms may not be the words people choose to label their subjective states, defining these three easily recognized levels of alertness is consonant with the fundamental principle underlying the SSS, other sleepiness scales, as well as the MSLT. This principle states that subjective sleepiness/alertness is a continuum. Most important, distinguishing these three subjective states allows us to define a detectable internal warning of impending disaster—a signal that absolutely requires a response. Drowsiness is red alert. Get off the road. Get out of harm’s way. Do it now.
As the field of sleep research and sleep medicine has become more involved with safety and public policy issues, communicating about fatigue and sleepiness has become much more important and greatly needing refinement. In 1990, the National Commission on Sleep Disorders Research launched a study that included interviews with 602 truck drivers and overnight testing of a subset of 200. It is quite certain that most of these drivers carried large sleep debts both because of inadequate amounts of sleep and because the majority were found to have obstructive sleep apnea. However, for the most part they denied having any serious problem with fatigue and sleepiness.
The questions pertaining to the symptoms of sleep disorders were highly standardized. Other questions that were originally intended merely to satisfy the curiosity of the researchers were less rigidly standardized and accordingly not reported. However, one result from the latter was profoundly surprising. In response to questions aimed at understanding why truckers decided to stop driving, 82% gave answers clearly indicating they stopped driving only after they fell asleep. The most common responses were “My head drops and I have a startle,” or “I see something in the road that isn’t there.” I have worried about this irrational and dangerous decision-making process ever since we obtained the data. Waiting to fall asleep in order to tell if one should stop driving is contrary to every principle of safety on the road.
Drowsy Driving
Sleep professionals are now engaged in a public policy campaign to stamp out “drowsy driving” without stating exactly what we mean by drowsy driving. It is clear that a sleep-deprived person can sometimes function at an adequate level depending, of course, on the degree of sleep deprivation, and the demands of the task. What is also abundantly clear is that there are countless people who would have shown a relatively low (sleepy) score on the MSLT if they had been tested immediately before they drove several hundred miles without having an accident. They would be indignant if we told them they could not drive because of their low score and would have to cancel their trip.
In this article, I wish to suggest an alternative that is better and much more likely to be accepted by truckers. I believe that drivers can be taught to step back from the abyss. Rather than falling asleep, the signal to stop driving will be the moment of the onset of drowsiness. Drivers can readily learn to recognize this signal. With the proper training, they will respond immediately and never drive drowsy.
Because under ordinary circumstances, the drowsy state occupies a relatively small amount of time, the prohibition of driving while drowsy would not be a great and impractical curtailment of driving time. It would, therefore, surely be a more acceptable restriction than being prohibited from driving if someone has not fulfilled their nightly sleep requirement, or has lost some arbitrary amount of sleep, or does not feel completely wide-awake and energetic.
As we continue to study the subjective side of the coin, we become more and more certain that the premonitory moment of drowsiness is easily detectable. The challenge is to break through a lifetime of disregarding this signal, or erroneously attributing it to such things as a heavy meal, or monotony. Everyone, not just drivers, must learn that there is one and only one cause of drowsiness. If you feel drowsy at any time of the day or night, you are sleep deprived.
Under ordinary circumstances, vigorous physical activity can completely abolish any trace of feeling tired and, of course, of feeling sleepy. However, when a severely sleep deprived person who is feeling wide-awake because of vigorous physical activity then sits quietly in a warm room to attend a lecture, or sits behind the wheel of a car for a boring freeway drive, there will be a rapid progression from wide-awake to sleepy to drowsy to falling asleep. The amount of time required to complete this progression is a function of the size of the sleep debt. At worst, the progression to falling asleep can be almost instantaneous. A small sleep debt can also cause drowsiness when stimulation from the environment is at its nadir and clock-dependent alerting has subsided. If sleep debt is zero, drowsiness never occurs.
We do not know at the present time how small the carryover sleep debt must be so that an individual is wide-awake and energetic all day long regardless of whether their situation is inactive and boring or entails a great deal of physical activity.
I do not honestly believe that all persons whose first language is English can be made to substitute “I am sleepy” for “I am tired.” However, I am certain they can be made to understand that sleep deprivation and sleep disorders collectively are the number one cause of “feeling tired” or being fatigued. They can also learn that in the absence of adequate stimulation and their own active resistance, they will inevitably progress toward falling asleep. They can learn that the feeling of drowsiness, which we have precisely defined, means that sleep is imminent. Finally, they can learn that in any situation where falling asleep has an unacceptably negative consequence, drowsiness is red alert. It is probably better to assert that sleepy and drowsy are synonymous. Thus, driving drowsy or sleepy is completely unacceptable. Driving tired is not possible if the “drowsiness is red alert” warning is always heeded.
As more and more people enjoy getting rid of their tiredness and fatigue by lowering their sleep debt or obtaining effective treatment of their sleep disorder, the problem of driving drowsy will be greatly reduced or even eliminated.
Conclusion
In summary, sleep deprivation can be quantified and the strength of the tendency to fall asleep at a specific time can be objectively measured. This is the scientific foundation upon which to build a safer, more rested, and more alert and energetic populace. The key to success for such a public safety initiative will be widespread education utilizing the terminology of more precise and clearly defined states of alertness.
It will be a huge step forward when everyone is aware of the following:

  • they clearly know when they are wide-awake, energetic, and not tired;
  • they clearly know that feeling tired means they have not obtained enough sleep or have a sleep disorder, and have accumulated a sleep debt of sufficient size to undermine their waking life;
  • they clearly know that the occurrence of drowsiness in the daytime means their sleep debt is large, and unless some action is taken immediately, drowsiness will soon be followed by sleep; and
  • they clearly understand the potentially horrible consequences of becoming drowsy in hazardous situations, and respond immediately the moment drowsiness occurs. Drowsiness is red alert.

William C. Dement, MD, PhD, is the Lowell W. and Josephine Q. Berry Professor of Psychiatry and Behavioral Sciences at Stanford University School of Medicine and director of Stanford Sleep Disorders Clinic and Research Center, Stanford, Calif.

References
1. Levine B, Zorick F, Roth T. Daytime sleepiness in young adults. Sleep. 1988;111:39-46.
2. Carskadon MA, Dement WC. Daytime sleepiness: quantification of a behavioral state. Neuroscience Biobehavioral Review. 1987;11:307-317.
3. Dement WC. Daytime sleepiness and sleep attacks. In: Guilleminault C, Dement WC, Passouant P, eds. Narcolepsy. Flushing, NY: Spectrum Publications; 1976:684.
4. Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC. Quantification of sleepiness: a new approach. Psychophysiology. 1973;10:431-436.
5. Carskadon M, Dement WC. Cumulative effects of sleep restriction on daytime sleepiness. Psychophysiology. 1981;18:107-113.
6. Dement WC, Vaughan C. The Promise of Sleep. New York: Random House Inc; 1999.