Excessive Daytime Sleepiness and Obstructive Sleep Apnea Syndrome

Arousals and excessive daytime sleepiness 

In normal subjects, brief arousals from sleep produce EDS and impaired daytime performance [10], [11]. In subjects with obstructive sleep apnea, arousal from sleep is believed to be an essential mechanism for re-establishing airway patency [12]. The apneic and hypopneic events in OSAS are associated with frequent arousals that lead to sleep fragmentation [12], [13], and thereby EDS [14]. Also, arousals can occur without apneas, hypopneas, or hypoxemia, and this leads to the concept of respiratory effort–related arousals. Respiratory effort–related arousals are defined as arousals occurring during and interrupting a succession of loud snores. It has been shown that respiratory effort– related arousals occur mostly in association with a rise in inspiratory effort secondary to increased upper airway resistance, resulting in arousals [6]. Short electroencephalogram (EEG) arousals (defined as detectable lightening of the EEG with alpha and beta waves for 3 seconds or more) were subsequently defined in 1992, as efforts were made to define the minimum degree of sleep disturbance that was needed to result in EDS [15].

Investigators have shown no consistent significant correlations between arousals and EDS [16], [17], [18], [19]. A recent study by Goncalves and coworkers [20] showed a stronger correlation between short EEG arousals (as per American Sleep Disorders Association [ASDA] criteria [15]) and severity of OSAS than those previously published, but the polysomnographic measures did not correlate with the subjective measurements of sleepiness. Even when other investigators reduced the length of EEG changes required to define an EEG arousal to 1.5 seconds, it was only minimally better at predicting objective daytime sleepiness [21].

Arousals could be poor correlates of objective EDS for various reasons. First, the reasons for daytime sleepiness could be from many factors other than recurrent arousals, such as sleep deprivation, drugs, xanthines, or cytokines. Second, there is night-to-night variation in sleep apnea severity [22], and a single hospital sleep study may not fully represent what happens normally, because the symptom is a cumulative result of variable sleep fragmentation. Third, approximately 28% of apneas and hypopneas are not terminated by visible cortical arousals [23]. This could be caused by the limitations of visual scoring in detecting all arousals, which might be better appreciated by EEG power spectral analysis [24]; it could be related to the fact that apnea and hypopnea may not necessarily lead to an EEG arousal, but only brain-stem activation without involvement of neuronal network above the thalamus. Further validation is needed to compare this marker with EDS severity. Fourth, integrating EDS with other EEG markers of sleep disruption, like cyclical alternating patterns (CAP) and detection of respiratory cycle–related EEG changes in OSAS subjects [25], [26], may improve evaluation of EDS. Finally, subjects may have difficulties recognizing their own degree of sleepiness, as shown in many sleep-deprivation or fragmentation studies, but this is mostly associated with denial of sleepiness.

Apnea-hypopnea index and excessive daytime sleepiness 

The severity of OSAS is most frequently gauged by measuring the AHI using the overnight polysomnogram. The AHI is used to reflect the severity of sleep fragmentation, but investigators have not been able consistently to demonstrate an association between AHI and EDS when EDS was assessed by the multiple sleep latency test (MSLT) or other measures [16], [17], [19], [21], [27]. Other studies have shown inability to correlate sleep study indices with daytime measures of sleepiness, where the r value rarely rises above 0.4 (ie, less than 20% of symptoms across a sleep clinic population seem explicable on the basis of the sleep study, the primary diagnostic tool).

Slow wave activity and excessive daytime sleepiness 

OSAS leads to disruption of normal sleep architecture with deprivation of rapid eye movement (REM) sleep and stages 3 and 4 non–rapid eye movement (NREM) sleep, although their sleep efficiency seems to be unchanged [16], [28]. Slow wave activity is considered an objective marker for homeostatic process. It has a broader definition that includes slow wave sleep stage criteria (including slow waves in stage 2 sleep) and a broader frequency distribution (0.75–4.5 Hz). Heinzer and coworkers [29] showed a decrease in slow wave activity across the night with OSAS patients leading to EDS. Slow wave activity increased after nasal continuous positive airway pressure (CPAP) treatment, but there was a lack of correlation with MSLT posttreatment.

Autonomic activation and excessive daytime sleepiness 

Autonomic activation has attracted attention as a potential index to quantify sleep disturbance in OSAS [30] because this method identifies even minor disturbing events that occur without visible EEG arousals [31]. There was an increase in EDS in one study, as measured by maintenance of wakefulness test (MWT) in normal subjects when autonomic activations were repeatedly induced in the absence of visible EEG arousals [32]. Subsequently, in OSAS patients, Bennett and coworkers [33] described autonomic activations as predictors of EDS by showing a significant correlation between the autonomic activation index and pretreatment objective sleepiness, and between the autonomic activation index and nasal CPAP responsive objective sleepiness. On the contrary, a recent study [34] showed that activation of brainstem (and thereby autonomic nervous system), without cortical arousal through subthreshold auditory stimulation, did not induce any change in MSLT or subjective sleepiness. Autonomic nervous system changes can be obtained with different types of stimuli; some of them may only involve simple reflexes with relay in medulla and lower brainstem, whereas others are associated with involvement of higher nervous system structures including the cortex. Investigation of autonomic nervous system stimulations, EEG arousal, and importance of autonomic nervous system stimulation necessary to induce sleepiness measured objectively has not been done. To relay only on an autonomic nervous system response to determine cortical arousal is erroneous; some type of EEG analysis is needed.

Hypoxemia and excessive daytime sleepiness 

OSAS also leads to repetitive oxygen desaturation, but this variable has not been considered to be of significant importance for EDS [35]. Colt and coworkers [36] showed no change in objective improvement in EDS as measured by MSLT when OSAS patients were treated with nasal CPAP, irrespective of the presence or absence of nocturnal hypoxemia. Their results lend further support to the hypothesis relating EDS to sleep fragmentation. Additionally, many patients with COPD or other lung diseases who present with chronic hypoxemia, which may tend to be more severe during rapid eye movement sleep, showed no correlation with EDS.

Snoring and excessive daytime sleepiness 

Snoring is also associated with EDS, although it could simply be a marker for obstructive sleep apnea. Young and colleagues [2] found that snorers with respiratory disturbance index ≤5 were substantially more likely to report EDS than were nonsnorers with respiratory disturbance index ≤5. Subsequent work by Gottlieb and coworkers [37] showed a significant association between snoring and sleepiness that was independent of the association between AHI and sleepiness. Snoring may be an independent cause of excess sleepiness, through mechanisms that remain to be elucidated [6]. In the recent past studies of snorers were performed with nasal cannula and pressure transducer, and with esophageal pressure. It was shown that these subjects presented with upper airway resistance syndrome when appropriately investigated, but this syndrome is not presented here.

Cyclic alternating pattern 

CAP is formed by electrocortical events that recur at regular intervals in the range of seconds during NREM sleep [Table 1] [25]. These events are clearly distinguishable from the background EEG rhythm as abrupt frequency shifts or amplitude changes. Two phases (A and B) are present that are part of a CAP cycle and recur within 2 to 60 seconds [see Table 1]. When neither of the phases (A and B) is identifiable, sleep has reached a new stable state [38]. Phase A is identified by transient events typically observed in NREM sleep. It includes EEG patterns of higher voltage; slower frequency; and faster lower voltage than the background EEG (with an increase in amplitude by at least one third compared with the background EEG). It is an activation phase lasting 2 to 60 seconds. Phase B follows phase A; it is the interval between two phases A with duration of 2 to 60 seconds, and has been defined by decreased EEG amplitude with EEG evidence of stages 1 to 2 NREM. Phase A has been subdivided into three subtypes. Subtype A1 is marked by a predominance of synchronized EEG activity and less than 20% of desynchronization of the EEG (fast frequency and low amplitude), such as delta bursts, K complex sequences, vertex waves, and polyphasic bursts (of slow and fast EEG rhythms). Subtype A2 is scored in the presence of 20% to 50% of desynchronized EEG activity with a predominance of polyphasic bursts. Subtype A3 is scored when at least 50% of the EEG activity is comprised of low amplitude fast rhythms, such as K-alpha complexes, ASDA-defined arousals, and polyphasic bursts [25].

Table 1. Summary of cyclic alternating pattern phases

	CAP phase	Characteristics
	Phase A	2–60 seconds
		Transient events
		EEG patterns of higher voltage, slower frequency, and faster-lower voltage than the background EEG (increase in amplitude by at least one third)
	Subtype A1	Predominance of synchronized EEG activity
		<20% desynchronization of EEG activity
		Delta bursts, K complex sequences, vertex waves, and polyphasic bursts
	Subtype A2	20%–50% of desynchronized EEG activity
		Predominance of polyphasic bursts
	Subtype A3	≥50% of desynchronized EEG activity
		Comprised of low-amplitude fast rhythms
		K-alpha complexes, ASDA-defined arousals, and polyphasic bursts
	Phase B	2–60 seconds
		Interval between two phases A
		Decreased EEG amplitude with EEG evidence of stages 1–2 NREM

Abbreviations: ASDA, American Sleep Disorders Association; CAP, cyclic alternating pattern; EEG, electroencephalogram; NREM, non–rapid eye movement.

The use of ASDA short EEG arousals and CAP scorings is increasingly becoming a popular way to evaluate sleep structure unlike the simple sleep staging evaluation based on at least 15 seconds out of one 30-second epoch that is currently being done using the Rechtschaffen and Kales method of scoring. CAP is a condition of NREM sleep instability [25].

Patients with OSAS have increases in CAP rate (>80% at the expense of non-CAP) with apneas and hypopneas usually occurring during phase B. Sleep fragmentation is associated with a significant enhancement of CAP phase A, especially A3 (that correlates strongly with ASDA EEG arousals), and more than 90% of the respiratory events are noted during CAP [39].

The enhanced resolution offered by the CAP, non-CAP perspective can shed light on the complex interactions between arousal rhythmicity, apneas-hypopnea, sleep disruption, and thereby its role in EDS and the effects of calibrating CPAP to the previously mentioned end points [40]. Further studies are required to explore the advantages of this sleep scoring approach in diagnosis and clinical management of OSAS patients with EDS, including nasal CPAP calibration.

Respiratory cycle–related EEG change 

Respiratory cycle-related EEG change is a quantification of computer-based signal analysis that detects subtle cortical responses (otherwise not detected by visual inspection) caused by increased work of breathing in sleep-disordered breathing subjects that might otherwise not be picked up during the nonapneic or hypopneic respiratory cycles. Using this method, Chervin and coworkers [41] showed that in subjects with sleep-disordered breathing the tendency for sigma (13–15 Hz) electroencephalographic power to vary with each respiratory cycle predicted next-day sleepiness as measured by the MSLT. Further investigation is necessary for complete validation of this method.

Objective measures 

Normally, the initial step in investigating the etiology of EDS is the performance of the overnight polysomnogram test, which can confirm sleep-related breathing disorders. This also helps to rule out other disorders that can potentially cause EDS.

Subjective measures 

Quality of life 

Impaired QOL is common among patients with OSAS [67], [68], [69], [70]. These patients may experience moodiness, anxiety, lack of motivation, and compromised performance in social function, all of which may contribute to a lower QOL. EDS is a major symptom of OSAS, and seems to be related to a decreased QOL for patients with OSAS [71]. Although the progression of EDS symptoms and their effect on QOL have not been investigated in detail, and although it has been suggested that QOL is dependent on the patient's depressive status, several studies examining the association between EDS and QOL have shown that EDS, whether measured subjectively or objectively, was associated with decreased QOL [72], [73]. Sleep deprivation can also lead to endocrine and metabolic changes associated with diabetes and weight gain; it is not uncommon for obese patients to have lower sleep efficiency and increased daytime sleepiness as measured by ESS [74], but these symptoms get worse when associated with OSAS. A large epidemiologic study of overweight subjects found that those with symptoms of OSAS reported poorer perceived health, lower economic income, increased odds of having had psychiatric care, multiple divorces, and impaired work performance compared with overweight subjects without symptoms of OSAS [75].

Functional impairments are normally mediated by symptom severity. QOL scales, such as the medical outcomes study Short Form-36, have been used to study the effect of OSAS on overall health. The Short Form-36 evaluates physical, emotional, and social functions; pain; general health; vitality; and mental health. This scale has revealed considerable impairments in samples of patients with OSAS, with vitality being most affected indicating a greater affect of OSAS on sleepiness [76].

Cognitive decline 

Reports have described an association between OSAS and cognitive decline, and some have implicated EDS as the primary component of this association [77]. Patients with EDS secondary to OSAS can suffer from poor concentration and memory disturbance, making it difficult to function productively and efficiently on a day-to-day basis. A recent study showed two thirds of new patients with OSAS reported difficulties in work efficiency and performing new tasks [78]. Another study of industrial workers described negative effects of EDS on workers' well-being including higher rates of work accidents, less job satisfaction, and higher drug usage [79]. Patients may try to compensate for daytime sleepiness with behavioral adjustments, such as napping or taking stimulants.

Motor vehicle accidents 

Also of major concern is sleepiness as an important cause of motor vehicle accidents, which often result in injury and death and cost billions of dollars each year. Habitual sleepiness while driving caused by a chronic condition, such as OSAS, occurs more frequently in middle age, whereas younger drivers more commonly experience sporadic sleepiness caused by sleep deprivation, alcohol, or drug abuse [80]. Drivers with OSAS may be two to seven times more likely to have a motor vehicle accident compared with normal drivers [81], [82]. Research using the Psychomotor Vigilance Task to study reaction time and response in subjects with sleep-disordered breathing has shown lapses, slow response time, and variability in response time comparable with actions of sleep-deprived or alcohol-impaired subjects [52]. Interestingly, although drivers may know when they are sleepy, they cannot reliably predict when they have become impaired [83], [84].

Previous studies have found that sleepiness measured by the ESS or sleep latency measured by the MSLT does not predict which drivers with OSAS will or will not have a motor vehicle accident [81], [85], [86], [87], [88], [89]. Additionally, neither of these tests measure sleepiness while driving, making it difficult to measure a possible association of EDS with automobile accident risk. For these reasons it has been suggested that asking about excessive sleepiness while driving rather than asking about overall sleepiness may suggest which breathing-disordered sleepy drivers are at higher risk of having a motor vehicle accident [80]. Currently, there are no clinical methods that identify which drivers with OSAS are at higher or lower risk of causing an accident [90]. The ability subjectively to assess a patient's ability to drive safely may be complicated if the patient is reluctant to admit struggling with daytime sleepiness in fear of losing one's license. This may lead to difficulties with assessment of EDS based on subjective measures. A physician responsible for the care of a patient with OSAS should address the issue of safe driving; however, despite the advice of a physician, patients with excessive sleepiness may often continue to drive in traffic, highlighting the importance of preventive efforts that focus on increasing the awareness of the dangers of sleepiness behind the wheel.

Obesity, sleepiness, and obstructive sleep apnea 

Many obese patients suffer from OSAS and EDS. It has even been suggested that there is a continuous interaction between weight increase, OSAS, and sleepiness. Any subject over 25 kg/m2 is considered overweight, and any small increase in body mass index above 2 standard deviations of the norm (ie, above 25 kg/m2) increases the chance of abnormal breathing particularly when supine and during REM sleep when weight distribution is android (ie, involving the abdomen). There is also occurrence of disturbances of metabolic and inflammatory factors that may again impact breathing during sleep. Such impairments may have a low key effect on sleep initially, but may have negative feedback on alertness, and may lead to sleep fragmentation and change in normal daytime activity levels with feedback on activity, food intake, and ultimately on weight. Presence of feedback loops with negative impact on both weight and alertness have been suggested; however, further research on these feedback loops with progressive worsening of signs and symptoms is needed.

Nasal continuous positive airway pressure 

Nasal CPAP therapy has emerged as the primary treatment for OSAS. It works either by increasing the intraluminal pressure in the pharynx, and thereby providing a mechanical pneumatic stent of the upper airway [91], [92], or by increasing the lung volume, which then mediates the upper airway stabilizing effect [93] (and probably both factors are active). When used properly and consistently CPAP is the treatment of choice for symptoms of EDS secondary to OSAS. There is still debate, however, regarding the efficacy of this modality in treating symptoms of EDS. Currently, Medicare covers CPAP devices for patients with an AHI of ≥15, or patients with sequelae of OSAS and an AHI of 5 to 14. Such sequelae include EDS, impaired cognition, hypertension, coronary artery disease, cerebrovascular accident, mood disorders, and insomnia.

Wright and coworkers [94] did a systematic review of the literature in 1997 looking at the treatment of OSAS with CPAP. Although the review consisted of studies that were often poorly designed (except for one randomized controlled trial with a small sample size) with nasal CPAP and control groups often not comparable at baseline, they still strongly suggested that CPAP may be effective in reducing EDS. There have been several randomized controlled trials, including sham CPAP as placebo-control, that consistently show that in patients with symptomatic mild to severe OSAS with daytime sleepiness, CPAP reduced subjective and objective measures of EDS and improved QOL [95], [96], [97], [98], [99], [100], [101], [102], [103], [104], [105], [106]. Patients with OSAS who are experiencing EDS may benefit more from treatment with nasal CPAP than patients without EDS. Interestingly, when Barbé and coworkers [8] looked at patients with severe OSAS (with AHI >30 per hour) with no daytime sleepiness, they found 6 weeks of nasal CPAP treatment had no significant effect on objective and subjective measures of daytime sleepiness. The lack of substantial improvement in daytime sleepiness could be related to a “floor effect”; the score was already low at baseline and there is little room for it to decrease further. The results of this study may not be reliable because of the small sample size. A recent meta-analysis was performed by Patel and coworkers [107] looking at all published randomized controlled trials of nasal CPAP in patients with OSAS. They specifically addressed the effects of nasal CPAP on subjective and objective sleepiness, and found that CPAP therapy across diverse populations of patients with OSAS resulted in significant improvement in subjective (decreases ESS by 2.94 points) and objective (increases sleep latency by 0.93 minute) measures of sleepiness. Effectiveness of nasal CPAP for OSAS patients with EDS may be overestimated if negative trials were unpublished and thereby not included in this meta-analysis.

Compliance is an issue in administering nasal CPAP; patients need to maintain long-term use of the treatment for it to be effective. Kribbs and coworkers [108] reported that just one night without treatment was sufficient to reverse improvements, despite a reduced AHI compared with pretreatment levels. Although high success rates with long-term use of nasal CPAP [109] have been reported, most of these studies were subjective, and described rates based on patient self-reports. Later objective studies used time clocks built into CPAP machines to record use and found that rates of CPAP use are actually lower [110], [111] and often irregular, rarely meeting prescribed levels [108], [112]. There are few long-term objective studies on nasal CPAP compliance.

There are several reasons for intolerance of nasal CPAP. First, patients using CPAP frequently report nasal discomfort. Often little effort is made to treat the nose, impaired by deviated septum and very commonly enlarged nasal inferior turbinates. Radiofrequency treatment of the nasal turbinates and aggressive treatment of nasal allergies help to decrease nocturnal disruption and residual sleepiness in OSAS patients. Second, problems have been encountered in calibrating nasal CPAP, as described by Thomas [40]. Calibration of nasal CPAP is performed by looking at the nasal flow curve, which does not detect the presence of small sleep disruptions that can only be seen with CAP scoring. Subsequently, inadequate treatment of OSAS with nasal CPAP slowly leads to the reappearance of complaints of tiredness and sleepiness. This is particularly true with autotitrating equipment: this equipment has several drawbacks. First, it is often set up with a very wide range of pressures (eg, 4–20 cm H2O) and such settings increases the risk of overshoot of pressure and thereby arousals. Second, it reacts slowly to the need for increasing pressures and this fragments sleep. Last but not least, OSAS is associated, at least in a certain number of patients, with permanent neurologic lesions in the upper airways that have been well demonstrated since the early 1990s [113], [114]. Treatment with nasal CPAP or bilevel CPAP never cures the permanent lesions. Slow evolution of these lesions over time related to the presence of local neuropathy is expected, and subsequently there is a need to reset pressures at higher levels. This was found in one of the authors' studies exploring patients with clear OSAS at entry (AHI ≥30 events per hour) [115]. This finding indicates the need to recalibrate nasal CPAP on a regular basis (probably every 2 years if no symptom occurs before). Recalibration should be done after interruption of CPAP for 1 night and even more so for 2 or 3 nights. This leads to increased nocturnal sleep disruption, initially demonstrated with CAP scoring during the night, and allows for better recalibration of the machine. Permanent neurologic lesions alone can lead to persistence of EDS, as explained later.

Residual daytime sleepiness 

Even with regular use of nasal CPAP, residual daytime sleepiness may persist in patients with OSAS, and reasons for this are not always clear. Many patients who use nasal CPAP regularly may experience insufficient sleep syndrome, which leads to cumulative partial sleep loss and impaired alertness and performance during the day [116].

Interruption of nasal CPAP for 1 night and even more so for 2 or 3 nights before recalibration leads to increased nocturnal sleep disruption and reappearance of sleepiness. Patients using nasal CPAP for only part of the night experience abnormal breathing when CPAP is not being used. Appropriate nocturnal nasal CPAP usage is often described as at least 4 hours per night, but there is no evidence that partial usage of CPAP during the night is sufficient to avoid reappearance of sleepiness in the long run. There is no data supporting the notion of complete treatment of sleepiness with usage of nasal CPAP for only a part of sleep, and thereby recommending a minimum cutoff point of time of usage is speculative.

Persistence of residual EDS despite regular nasal CPAP use and appropriate calibration with monitoring of sleep EEG may also be seen. One study found an association between abdominal obesity and persistence of EDS [117]. Nasal CPAP is a poor treatment of chest bellows syndrome, related to clear abdominal obesity, particularly during rapid eye movement sleep. In this case, bilevel PAP gives a better approach, but residual sleepiness also has been seen with bilevel PAP. Two hypotheses have been raised in this regard. First, the potential association of obesity with EDS: obesity leads to metabolic and inflammatory changes that in themselves have an impact on sleep; however, good studies are lacking to support this plausible hypothesis. Second, the presence of central nervous system lesions that lead to persistence of EDS. Gora and coworkers [118] presented some preliminary data indicative of abnormal evoked potential during the NREM sleep in OSAS patients. A more definitive study was performed by Afifi and coworkers [119], which clearly showed the absence of evoked responses to inspiratory occlusion stimuli during NREM sleep. The absences of these responses may be related to local neuropathy, but secondary neuronal degenerative evolution also could have happened. Another possibility is that repetitive hypoxemia, arousals, and cerebral blood flow changes associated with sleep-disordered breathing for years could account for these permanent central nervous system changes, particularly in an individual with a certain genetic background. For example, the presence of the APOE-4 gene is considered to be a risk factor for neurodegenerative disease, and is currently shown to be present in a series of OSAS patients [120].

Sleepiness may also be indicative of other risk factor associations, such as a coexisting undiagnosed sleep disorder, and antihypertensive or other drug use. Also, as studies are looking at the potential association between obesity and EDS, other causes of EDS apart from sleep disordered breathing should be considered at, such as the association of EDS with mood disorders (eg, depression), metabolic factors like diabetes, and age.

Modafinil 

Modafinil, 2-[(diphenylmethyl)-sulfinyl] acetamide, promotes wakefulness and is currently used in treatment of daytime sleepiness in adults suffering from narcolepsy. Use of amphetamine-like drugs for the treatment of EDS secondary to OSAS is limited because of psychiatric disturbances, interference with night sleep, and the addictive nature of these drugs [121]. Alternatively, modafinil is a nonamphetaminic drug that has fewer side effects, does not interfere with night-time sleep, and has lower potential for abuse [122], [123].

This drug has no effect on the AHI, but has been shown to reduce objective sleepiness. Six prospective studies report the effects of modafinil on EDS in OSAS patients. Five studies [124], [125], [126], [127], [128] have reported the effects of modafinil (200–400 mg/day) on residual sleepiness in subjects with OSAS, despite compliance with CPAP [Table 3]. Across studies, modafinil consistently improved subjective and objective sleepiness, QOL, and vigilance compared with placebo. Dinges and Weaver [129] reported that the frequency of lapses of attention during Psychomotor Vigilance Task performance was significantly decreased, and both the median and slowest reaction times were significantly improved in subjects treated with nasal CPAP and modafinil compared with those treated with nasal CPAP and placebo. It is important to appreciate, however, that most (75%) subjects with severe sleepiness at baseline had multiple sleep latency times of <10 minutes on modafinil, despite effective CPAP and good compliance with therapy. Prescribing modafinil should not lessen the concerns of continued risk for driving-related motor vehicle accidents caused by sleepiness in patients with OSAS.

Table 3. Randomized studies showing improved outcomes for treatment with modafinil in patients residual EDS despite compliance with CPAP

	Study	No. of subjects	Dose/number of days treated	Outcome measures for modafinil group (compared with placebo group)	P-value
	Kingshott et al	32	200 mg × 5 d	Improved alertness as measured by MWT.	0.02
			400 mg × 9 d	No significant improvement with ESS or MSLT.
				FOSQ showed trend of improved vigilance.
	Pack et al	157	200 mg × 7 d	Improved alertness as measured by MWT.	<.001
			400 mg × 21d	Lower ESS score.	0.0001
				Longer sleep latency as measured by MSLT (slightly worsened for placebo).	0.021
	Heitmann et al	24	300 mg × 2 d	Shorter mean reaction time as measured by vigilance test.	0.0023
				Longer mean sleep latency as measured by MSLT.	0.0001
	Black et al	305	2 groups	68% patients showed overall clinical improvement in both treated groups.	<.001
			200 mg × 84 d (12wk)	Improved alertness as measured by MWT.	≤.0001
			400 mg × 84 d (12wk)	Lower ESS score.	<.0001
				Improved FOSQ score.	<.001
	Schwartz et al	125	(12wk)	73% patients showed overall clinical improvement.	0.001
			200 mg week 1	Improved FOSQ score.	<.0001
			400 mg week 2	Lower ESS score.	<.001
			By week 12, 17% of patients were on 200 mg, 19% on 300 mg, and 64% on 400 mg

Abbreviations: ESS, Epworth Sleepiness Scale; FOSQ, Functional Outcomes of Sleep Questionnaire; MSLT, Multiple Sleep Latence Test; MWT, Maintenance of Wakefulness Test.

Adverse events attributed to modafinil include headaches, nervousness, and rhinitis (5%–10% higher likelihood than placebo). A 2-week, placebo-controlled study with one dose (400 mg) of modafinil [124] demonstrated that patients using this stimulant had a modest but significant reduction in their nightly usage of nasal CPAP. This negative effect was not found in two other double-blind studies of longer duration [125], [126], [127], or in the open-label continuation study [128].

Acute administration of modafinil increases arterial blood pressure and heart rate [126]. With exercise, modafinil intake (300 mg) results in a significant increase in mean, systolic, and diastolic pressures and heart rate, compared with placebo. No long-term investigation of cardiovascular outcomes with the use of modafinil has been reported.

In summary, the double-blind, placebo-controlled clinical trials that examined the effectiveness of modafinil in patients compliant with nasal CPAP in treating residual sleepiness found that modafinil subjectively and objectively improved vigilance and sleepiness for as long as 12 weeks. Modafinil, however, does not fully reverse severe baseline sleepiness. It seems that 200 mg daily is as efficacious as 400 mg. There is concern that compliance with nasal CPAP may decrease with modafinil usage [124], and this requires further study. In the interim, patients must be advised of the importance to continue CPAP therapies and physicians must carefully monitor CPAP compliance in this group. The question of the long-term effect of modafinil on the hemodynamic status of OSA patients treated with nasal CPAP remains unresolved. Nevertheless, findings from one study measuring cardiovascular responses to modafinil raise the possibility that modafinil use may increase blood pressure [126]. Careful consideration of an individual patient's health risks (the risk of motor vehicle accidents compared with the risk of cardiovascular morbidities) is required before prescribing modafinil.

Modafinil may also prove useful in patients with mild OSAS, who often go untreated because of the relatively inconvenient nature of current treatments [129], [130]. Although modafinil does seem to enhance alertness, it does not treat the underlying pathophysiology of airway collapse in OSAS and may not be of therapeutic benefit for patients who are not receiving adequate prophylactic treatment for apneas and hypopneas, and may result in serious complications because of long-term OSAS that has gone untreated.