A practical approach to the assessment of daytime sleepiness

Excessive daytime sleepiness is a major health problem in the United States. Several surveys indicate that approximately one third of Americans report daytime sleepiness that occurs at least a few days per month, and about 6% report daily symptoms.^1-3 Daytime sleepiness can result in problems with vigilance, cognitive function, memory, concentration, and mood; it often leads to deterioration in school and/or job performance and productivity, social relationships, and driving skills.

The prevalence of driving while drowsy is high in the United States; 23% of persons surveyed in 1998 admitted to having dozed off behind the wheel of a car.³ It has been estimated that 15% to 20% of accidents may be related to sleepiness or fatigue, which causes about 1500 motor vehicle deaths per year.Therefore, an understanding of the causes and differential diagnosis of daytime sleepiness is important for all physicians.

In this article, I will discuss the determinants of sleepiness, the methods commonly used to measure sleepiness, the differential diagnosis, and a practical approach to the diagnosis, with emphasis on which patients need referral to a sleep center.

Sleepiness in the midafternoon is normal; sleepiness is considered abnormal if it occurs at other times of the day or in situations in which alertness is warranted (meetings, lectures, driving). Sleepiness is a physiologic state that is similar to hunger or thirst. The subjective feeling of sleepiness can be reduced by motivation, excitement, exercise, and other competing needs.

As physiologic sleepiness worsens, it becomes more difficult to reduce its impact on overt behav-ior. Therefore, soporific behaviors (such as reading or watching television) unmask sleepiness but do not cause it. A truly alert person does not feel or appear to be sleepy when placed in a low-stimulus environment.

A person's optimal sleep time is defined as the amount of time that allows him or her to maintain alertness throughout the day. For most persons, this is about 7 to 8 hours each night, although the optimal amount is variable.

Even 1 night of sleep deprivation, generally 4 hours of sleep or less, can lead to increased sleepiness on the following day, a situation that most persons have experienced.⁴ However, chronic sleep deprivation, usually less than 6 hours per night over consecutive nights, is a more common cause of daytime sleepiness. According to 2 surveys, approximately 15% of Americans sleep less than 6 hours per night on a regular basis.^3,5

The quality and continuity of sleep are important determinants of sleepiness. Sleep disorders that are characterized by brief arousals, such as obstructive sleep apnea/hypopnea syndrome (OSAHS) and periodic limb movement disorder (PLMD), are chief examples of disorders that affect the quality of sleep. However, other conditions, such as asthma, congestive heart failure, gastroesophageal reflux disease, and various pain syndromes, can also contribute to daytime sleepiness by interfering with sleep.

A biphasic pattern of sleep tendency is observed in studies that test adults for physiologic sleepiness by asking them to sleep at 2-hour intervals in a low-stimulus environment, such as a darkened room.⁶ As expected, the shortest latencies to sleep occur during the early morning hours, 2 to 6 am. However, there are also shorter sleep latencies in the midafternoon, between 2 and 6 pm, which correspond to the time of day that many persons report increased sleepiness (and the time of the "siesta" that is prevalent in many cultures).

Drug side effects should always be investigated as a possible cause of sleepiness.⁷ Many medications can contribute to daytime sleepiness (Table 1).

The first-generation antihistamines, particularly those available over-the-counter (OTC), such as diphenhydramine(Drug information on diphenhydramine), are among the most common causes of sleepiness. However, many patients do not think of these as medications, so they may not mention them when asked about medication use. Thus, it is important to ask patients specifically about the use of OTC antihistamines.Of thenewer, nonsedating antihistamines, only cetirizine has been associated with daytime sleepiness.

Many psychoactive medications cause daytime sleepiness. While the benzodiazepine sedative-hypnotics are generally taken at bedtime to help induce sleep, several of these agents are intermediate- or long-acting (temazepam and estazolam(Drug information on estazolam)) and, therefore, they can cause residual sleepiness during the day. In addition, the anxiolytic benzodiazepines (such as diazepam and alprazolam(Drug information on alprazolam)) can cause sleepiness in some persons.

Several antidepressants, such as amitriptyline(Drug information on amitriptyline), trazodone, and paroxetine, have been associated with daytime sleepiness, as have barbiturates such as phenobarbital(Drug information on phenobarbital). Sedation is a common side effect of the traditional, older antipsychotic medications. Although sedation has been reported for most of the new-generation antipsychotics, it is much less common with these drugs.

In addition, antiepileptic drugs, such as phenytoin(Drug information on phenytoin), carbamazepine(Drug information on carbamazepine), and valproic acid, have been reported to cause sedation. Newer antiepileptic agents, such as gabapentin(Drug information on gabapentin) and topiramate(Drug information on topiramate), can also have sedating effects, although at a lower rate than the other antiepileptic agents.

Narcotics can be associated with daytime sleepiness in some persons. Also, alcohol is a major cause of sleepiness.

There have been reports of sleepiness with antihypertensives, such as ß-blockers and centrally acting a-blockers (clonidine).⁷ However, with the exception of clonidine(Drug information on clonidine), no studies have shown objectively that these drugs increase sleepiness.

A normally functioning nervous system, particularly in the areas that control the wakefulness-sleep cycle (primarily the pons, hypothalamus, and basal forebrain), is critical to sleep regulation.⁸Abnormal function of the nervous system can cause daytime sleepiness, generally manifested as narcolepsy or idiopathic hypersomnolence.

Recent research has focused on the role of the neuropeptide hypocretinas a wake-promoting agent. Hypocretin-producing neurons in the hypothalamus project to much of the CNS and heavily innervate many areas of the brain associated with the sleep-wake cycle. A deficiency of hypocretin is now thought to be the major abnormality underlying the pathogenesis of narcolepsy.

There are 2 common methods of measuring daytime sleepiness: the Multiple Sleep Latency Test (MSLT) and the Epworth Sleepiness Scale (ESS).

The MSLT measures the physiologic sleep tendency by measuring the time it takes to fall asleep when one begins a nap in the absence of alerting factors such as noise and light.^9,10 During the test, which is generally performed on the day after a full sleep study, the patient is asked to fall asleep in bed, in a dark, quiet room, 4 or 5 times at 2-hour intervals during the day.

The latency to sleep for each nap is averaged to determine the daytime sleep latency. Generally, mean latencies of greater than 10 to 15 minutesare considered normal; a sleep latency of 5 minutes or less is considered severe or pathologic. In addition, there should be no rapid eye movement (REM) sleep during any of the nap opportunities.

The MSLT is a well-validated test of daytime sleepiness that can detect changes in sleepiness secondary to changes in sleep time,⁴ medications,¹¹ and treatment.¹² The main disadvantage of the MSLT is that it is laboratory-based and has to be performed by trained sleep personnel.

Given this requirement, the primary indication for the MSLT is to evaluate sleepiness in patients in whom sleep apnea is unlikely. In particular, the test is indicated for the evaluation of patients with suspected narcolepsy. Patients with narcolepsy generally show pathologic sleep latencies (less than 5 minutes) and have REM sleep during at least 2 of the 4 or 5 naps. In the absence of cataplexy, these findings on the MSLT are considered diagnostic of narcolepsy.

A modified version of the MSLT that is now frequently used in research studies--the Maintenance of Wakefulness Test (MWT)--measures a patient's ability to stay awake under soporific conditions rather than the time it takes to fall asleep.^13,14 The MWT protocol is similar to that of the MSLT (4 naps, spaced 2 hours apart), except that the patient is instructed to stay awake while sitting upright in a low-stimulus environment.

The MWT is considered to be more clinically relevant because it reflects more closely the challenge patients face in soporific situations of everyday life. It is used frequently in research studies on the effectiveness of treatment strategies such as nasal continuous positive airway pressure in OSAHS¹⁰ and stimulants in narcolepsy.¹⁵ Regulatory agencies occasionally require a normal MWT result before allowing truck drivers or airline pilots who have a diagnosis of OSAHS to return to work.

The ESS was developed as an easy-to-use subjective scale that can measure daytime sleepiness without the requirement for in-laboratory testing.^16-18 In answering the ESS, the patient rates his or her likelihood of dozing off in 8 situations ranging from highly soporific, such as watching television or reading, to those requiring attention, such as talking or attending a meeting (Table 2). Each situation is scored on a scale of 0 (will not doze off) to 3 (high likelihood of dozing off). The scores from each situation are added, which gives a final score between 0 and 24.

The major criticism of the ESS is that it does not correlate well with the MSLT. Traditionally, a score of 10 or above has been used to distinguish between persons with and without daytime sleepiness, although a recent article suggested that a cutoff of 12 may be a more reliable predictor of an increased tendency to fall asleep on the MSLT.¹⁹ A higher ESS score indicates a greater propensity to fall asleep¹⁹--but not necessarily a shorter sleep latency²⁰--on the MSLT. Therefore, on initial evaluation, the ESS value may or may not indicate that the patient is truly sleepy, as measured objectively by the MSLT.

However, several studies show that higher ESS scores decrease with effective treatment of the primary sleep disorder.^12,15,21,22 Therefore, the ESS can provide valuable information to the physician in determining whether the treatment plan has been effective for the patient. It should be noted that while the MWT, MSLT, and ESS all have their uses, there is little or no outcomes-based research that would allow clinicians to use any of them as a reliable predictor of risk for motor vehicle accidents or other future morbidity associated with sleepiness.