Fatigue, Sleepiness, and Performance in Simulated Versus Real Driving Conditions Pierre Philip, MD, PhD1,2; Patricia Sagaspe, PhD3; Jacques Taillard, PhD1,2; Cédric Valtat, CRA1; Nicholas Moore, MD, PhD4; Torbjorn Åkerstedt, MD, PhD5; André Charles, PhD3; Bernard Bioulac, MD, PhD1,2 Clinique du Sommeil, CHU Pellegrin, Bordeaux, France; 2CNRS UMR-5543, Université Bordeaux 2, Bordeaux, France; 3Laboratoire de Psychologie, Université Victor Segalen, Bordeaux, France; 4Département de Pharmacologie, Université Victor Segalen, Bordeaux, France; 5Karolinska Institutet, Stockholm, Sweden 1

driving (P = .004) and sleep deprivation (P = .004). Subjects had higher sleepiness scores in the driving simulator (P = .016) and in the sleep restricted condition (P = .001). Fatigue increased over time (P = .011) and with sleep deprivation (P = .000) but was similar in both driving conditions. Conclusions: Fatigue can be equally studied in real and simulated environments but reaction time and self-evaluation of sleepiness are more affected in a simulated environment. Real driving and driving simulators are comparable for measuring line crossings but the effects are of higher amplitude in the simulated condition. Driving simulator may need to be calibrated against real driving in various condition. Keywords: Driving, sleep restriction, simple reaction time, driving simulator, real driving, sleepiness scale, young adult, fatigue Citation: Philip P; Sagaspe P; Taillard J et al. Fatigue, sleepiness, and performance in simulated versus real driving conditions. SLEEP 2005;28(12): 1511-1516.

Study Objectives: To determine whether real-life driving would produce different effects from those obtained in a driving simulator on fatigue, performances and sleepiness. Design: Cross-over study involving real driving (1200 km) or simulated driving after controlled habitual sleep (8 hours) or restricted sleep (2 hours). Setting: Sleep laboratory and open French Highway. Participants: Twelve healthy men (mean age ± SD = 21.1 ± 1.6 years, range 19-24 years, mean yearly driving distance ± SD = 6563 ± 1950 miles) free of sleep disorders. Measurements: Self-rated fatigue and sleepiness, simple reaction time before and after each session, number of inappropriate line crossings from the driving simulator and from video-recordings of real driving. Results: Line crossings were more frequent in the driving simulator than in real driving (P < .001) and were increased by sleep deprivation in both conditions. Reaction times (10% slowest) were slower during simulated

to impair neurobehavioral functioning.14,15 Interaction between these two regulatory processes (circadian and homeostatic) induces a non-linear evolution of sleepiness over time. In a previous study on a freeway16 we showed that performance (reaction time [RT]) measured during the breaks was only marginally impaired by extended real driving (10 hours) when the drivers were rested, but very much impaired when the same drivers were sleep deprived. Self evaluation of performance in the laboratory was significantly more efficient than in the natural environment. Understanding the contribution of the environment to fatigue, sleepiness and performance seems crucial in interpreting driving studies. Several studies have shown that impaired daytime alertness causes an increase in lateral deviations during real driving17-20 and during driving in car simulators.21-25 Driving simulator studies have dominated these types of studies, mainly because of safety, low cost and ease of data collection. On the other hand, some subjects suffer from simulator sickness (and therefore are excluded from the studies) and, intuitively, real life contexts induce much more stimulation than do simulators, even if systematic studies are lacking. This casts some doubts on the interpretation of results obtained in simulators, in particular in relation to fatigue effects. For instance, in real life the majority of sleep related accidents occur after at least one hour of driving26,27 while performance decrements in driving simulators are seen much earlier.24 This could possibly be explained by less stimulation under artificial conditions of driving. To determine whether real life driving would produce results different from those obtained in a simulator we designed a controlled crossover study of real long-distance driving versus simulated driving under normal and sleep restricted conditions.

INTRODUCTION THOUGH FATIGUE AND SLEEPINESS AT THE WHEEL ARE WELL-KNOWN RISK FACTORS FOR TRAFFIC ACCIDENTS,1-3 MANY DRIVERS COMBINE SLEEP deprivation and extensive driving.4,5 Because of these conflicts between physiological needs and social or professional activities6,7 understanding the human limits of fatigue/sleepiness and sleep deprivation are becoming key issues in accident prevention. The two concepts “fatigue” and “sleepiness” are similar, but not identical. Fatigue is usually seen as a gradual and cumulative process due to sustained activity and associated with a disinclination towards effort, eventually resulting in reduced performance efficiency.8 It has been described in driving episodes which require sustained attention for long periods of time.9 Fatigue is eliminated by a period of rest. Sleepiness denotes a difficulty in remaining awake.10 This symptom is related to circadian and homeostatic influences and disappears after sleep,11 but not after rest. Extended time awake and/or sleep restriction increase sleep pressure and generate cumulative sleepiness12,13 which is known Disclosure Statement This was not an industry supported study. Drs. Philip, Sagaspe, Taillard, Valtat, Moore, Akerstedt, Charles, and Bioulac have indicated no financial conflicts of interest. Submitted for publication April 2004 Accepted for publication August 2005 Address correspondence to: Dr. P. Philip, Clinique du Sommeil, CHU Pellegrin, Place Amélie Raba-Léon, 33076 Bordeaux Cedex, France; Tel: 33 5 56 79 55 13; Fax: 33 5 56 79 48 06; E-mail: [email protected] SLEEP, Vol. 28, No. 12, 2005

1511

Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al

AM,

11:00 AM to 12:45 PM, 1:15 PM to 3:00 PM, 3:15 PM to 5:00 5:15 PM to 7:00 PM, and 7:15 PM to 9:00 PM. The driving took place on a straight highway on a weekday with light traffic and in fair weather. The conditions were virtually the same for all drivers. Subjects were instructed to maintain a constant speed (130 kph; 80 mph) and not to cross the lane markings except to pass a slower vehicle. During the whole experiment, a professional driving instructor monitored the driving speed and was ready to take over control of the car (equipped with dual controls) if the subject started losing control of the vehicle. Subjects unable to continue driving were driven back to the rest area and carried out ratings and the performance test at the next scheduled time. If the subject felt he could continue the drive after the break, he was encouraged to do so. The car used for the experiment was equipped with a video camera that filmed and recorded the road.30 Because sleep-related accidents frequently happen with a single car driving off the road and hitting an obstacle with no reaction from the driver,26 inappropriate line crossings were used as the main criterion for quantifying driving impairment after sleep restriction. In the middle of each driving session, subjects were asked to rate their sleepiness on the Karolinska Sleepness Scale (KSS).31 At the end of each driving session (at 10:45 AM, 12:45 PM, 3:00 PM, 5:00 PM, 7:00 PM, 9:00 PM) subjects had a 15-minute break to carry out tests and ratings. Before the first driving session and during each of these breaks, subjects were asked to rate their instantaneous fatigue (“describe how fatigued you are now”) on a 100-mm visual analogue scale and to perform a 10-minute cognitive test. At the end of this 15-minute break, driving was resumed. At 1:00 PM, subjects had an additional 15-minute break for a quick lunch, taken after the tests. An inappropriate line crossing was counted each time the car crossed 1 of the lateral highway lane markers, as evidenced by video-recording analysis. Effects of passing another car or some other necessary driving action were disregarded.

METHODS

PM,

Subjects The study was approved by the local ethics committee (consultative committee for the protection of persons participating in biomedical research, CCPPRB Bordeaux A). Twelve healthy men (mean age ± SD = 21.1 ± 1.6 years, range 19-24; mean yearly driving distance ± SD = 10,500 ± 3120 km [6563 ± 1950 miles]) participated after providing written informed consent. Subjects were recruited by announcement in the hospital and university and were paid €400 for the whole experiment. Inclusion Criteria All subjects underwent a clinical interview with a sleep specialist and a nocturnal polysomnography to rule out any sleep disorders. Because sleep duration and sleep efficiency are crucial in sleep-restriction protocols, we used actimeters (Actiwatch®, Cambridge Neurotechnology, Cambridge, UK)28 to quantify our volunteers’ sleep duration. This device monitors body movements and allows calculation of mean nocturnal sleep episodes and of nocturnal awakenings. Time in bed was also computed as the time difference between going to bed in the evening and getting up in the morning. Sleep efficiency was calculated as the ratio of time asleep to the time in bed, in percentage.29 To rule out any sleep-wake schedule disorders each subject was monitored for 7 days before being included in the study. Subjects were included if they had a mean sleep efficiency of at least 85% during the 7 days of recordings. Study Design The study was a randomized open cross-over study using a balanced latin square design, with all subjects having 4 driving periods: in the driving simulator after normal sleep and after sleep deprivation and on the open road after normal sleep and after sleep deprivation. The order of these was random. At least 3 days elapsed between 2 successive sessions.

Driving Simulation We used the Divided Attention Steering Simulator (Stowood Scientific Instruments, Oxford, UK) based on a tracking task and visual detection of digits located at the 4 points near the sides of a computer screen (see Figure 1). The software reproduces a winding road shown as lines on the screen, and the tasks are to use the steering wheel to keep the front of the car in the middle of the road and not to go over the edge of the road. The steering Figure 1 was : wheel of a computer game type but was modified to full size (Grandprix 1, ThrustmasterTM, Thrustmaster, Croydon, UK) and

Sleep Schedules The subjects were instructed to maintain a regular sleep-wake schedule and were monitored by actimetry during the 3 days preceding each experimental session. No stimulants of any kind were allowed during the study. During the rested condition, subjects were monitored in the laboratory from 9:00 PM to 8:30 AM and were in bed from 11:00 PM to 7:00 AM. In the sleep-restriction condition, subjects were monitored in the laboratory from 9:00 PM to 8:30 AM the next day but were allowed to sleep only from 11:00 PM to 1:00 AM. The duration of sleep was monitored by actimetry during both conditions. Driving Sessions Real Driving Starting at 9:00 AM after the night of full or partial sleep, subjects drove 6 identical 200-km (125-mile) segments on a separated-lanes highway (100 km [62.5 miles] in one direction and 100 km in the other). Each driving session lasted 105 minutes, from 9:00 AM to 10:45 SLEEP, Vol. 28, No. 12, 2005

Figure 1—Screen display of the Divided Attention Steering Simulator. 1512

Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al

Rested Condition

Sleep Restricted Condition

Inappropriate line crossings (Real driving)

Inappropriate line crossings (Simulated driving)

Rested Condition 350 300 250 200 150 100 50 0 9h-10h45

11h-12h45

13h15-15h

15h15-17h

Time of day

17h15-19h

Sleep Restricted Condition

20 16 12 8 4 0

9h-10h45

19h15-21h

11h-12h45

13h15-15h

15h15-17h

17h15-19h

19h15-21h

Time of day

Figure 2—Total number of inappropriate line crossings during simulated driving in the rested versus sleep-restricted condition.

Figure 3—Total number of inappropriate line crossings during real driving in the rested versus sleep-restricted condition.

fixed to a table. The steering wheel was equipped with buttons on the right and left sides. The computer program sampled the lateral deviation from the center of the road at 10 Hz. If the driver was unable to hold the car between the edges of the road for 15 seconds, the drive stopped automatically. The simulated driving sessions lasted the same time as the real driving sessions (105 minutes), and the sessions started at the same times as the real driving sessions. In addition to the tracking test, subjects were asked to monitor 4 digits located at the sides of the computer screen. These digits changed every 8 to 10 seconds in a pseudorandom order, and the subject was instructed to click on the appropriate (left or right) button every time he saw the number 2 appear at the side of the screen. The software calculated the mean deviation from the center of the road and its standard deviation. This constitutes 1 of the outcome variables and has shown clear sensitivity to fatigue in a previous study.32 The standard deviation of distance from the “ideal curve” was also used as a dependent variable. The time taken to respond to the relevant digit, that is, the RT, was calculated as the time elapsed from the presentation of the digit to the button press. The slope across time of this RT was also calculated, with zero slope indicating a constant performance level. When the performance deteriorates during the test, the slope becomes positive. To reduce any learning effect, each subject first performed two 10-minute training sessions.

instructions needed for sampling). Another part of the program calculates the mean of the 10% slowest trials (slowest RT) during the 10-minute task. This value was used as a dependent variable in our study. Data Processing and Analysis Main outcome variables were (1) number of line crossing per subject, session, driving and sleep condition; (2) RT; and (3) selfassessed fatigue and sleepiness. The effects of sleep deprivation, duration of drive, and time of day were tested through 3-way analysis of variance (ANOVA) for repeated measurements. To correct for sphericity, all P values derived from the ANOVA were based on Huynh-Feldt corrected degrees of freedom. Results for line crossings, fatigue, and sleepiness are given as means and SD. Time effect (if significant) was analyzed by repeated contrast (each level versus the next level). Inappropriate line crossings in real versus simulated driving and mean deviation from the center of the road versus line crossings were compared using a Spearman rank correlation test. RESULTS Line Crossings The results from a 3-factor ANOVA showed significant main effects of “driving condition” (F1,10 = 156.184, P < .001) and of “sleep condition” (F1,10 = 60.013, P < .001), with a higher number of inappropriate line crossings for the simulated driving condition than for the real driving condition, and for the sleep-restricted condition versus the rested sleep condition (see Figures 2 and 3). There was no main effect of “time” (F5,50 = 1.274, P = .301), as defined by the evolution of performance across the day. The interaction driving condition × sleep condition was significant (F1,10 = 54.628, P < .001). There was no overall correlation between simulated and real driving in the rested condition. In the sleep-deprived condition, there was an overall correlation between line crossings in real driving and simulated driving (r = 0.655, P = .021). However, this was the result of a correlation found only for the 11- to 13-hour sleep-restricted condition (r = 0.712, P = .008). In the simulated condition, mean deviation from the center of the road and mean RT to the digits were significantly affected only by sleep condition (F = 123.989, P < .001 and F = 90.896, P < .001). In the simulated condition, there was a significant correlation between mean deviation from the center of the road and line crossings for both conditions (Spearman Rank correlation, r

RT Performance RT measurements were used to obtain a performance measure that could be compared between the 2 driving conditions. RT performance is one of the most established and sensitive fatigue indicators.33 A 10-minute RT test was carried out at 8:45 AM, 10:45 AM, 12:45 PM, 3:00 PM, 5:00 PM, 7:00 PM, and 9:00 PM, ie, before the first session and after each driving session, whether real or simulated. The test was run on a PALM personal organizer34,35 and involved a black square displayed on the screen at randomized (2- to 7-second) intervals, during 10 minutes. This yielded 138 stimuli plus a 1-minute training session. The subject’s task was to respond to the stimuli by pressing a key to turn the square off. If no response was given within 1 second, a new interval was started. Pressing the key before the square was displayed, or within 100 milliseconds, caused the response to be discarded and a warning to be displayed. The software that controls the internal clock yields data with at least 0.01-millisecond resolution. (The keyboard is sampled at CPU frequency divided by the number of SLEEP, Vol. 28, No. 12, 2005

1513

Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al

Simulated Driving

Real Driving

Simulated Driving Karolinska Sleepiness Scale

10% slowest RT

1750 1500 1250 1000 750 500 250 0 8h45

10h45

12h45

15h

17h

Rested condition

19h

21h

8h45

Time of day

10h45

12h45

15h

17h

19h

21h

Sleep restricted condition

Figure 4—Impact of real versus simulated driving on 10% slowest reaction time (RT) in the rested and sleep-deprived condition.

9 8 7 6 5 4 3 2 1 0 9h

11h

13h

15h

17h

Rested condition

19h

9h

11h

Time of day

13h

15h

17h

19h

Sleep restricted condition

Figure 5—Impact of real versus simulated driving on sleepiness, based on response to the Karolinska Sleepiness Scale, in the rested and sleep-deprived condition.

= 0.853, P < .001 and r = 0.811, P < .001). In the real driving condition, 2 subjects in the sleep-loss condition had to be driven back to the rest area by the copilot because of exhaustion. Nevertheless, both subjects finished the day of testing. There was no correction in the analysis for these incomplete driving sessions.

Simulated Driving

Fatigue (VAS)

100

Reaction Time Mean RTs (10% slowest), shown in Figure 4, were significantly longer during simulated driving (670 milliseconds vs 533 milliseconds, P = .004), and after sleep restriction (713 milliseconds vs 490 milliseconds, P = .004). There was no change of RT over driving time in either condition. There was no interaction between sleep and driving condition, showing that the effects of sleep restriction were similar in either condition. There was no interaction of either sleep condition or driving condition with driving time. Significant main effects were found for “driving condition” (F1,11 = 12.981, P = .004), with slower RTs for the simulated driving condition, and for “sleep condition” (F1,11 = 13.083, P = .004), with slower RTs for the sleep-restricted condition. No main effect of “time” (F6,66 = 1.657, P = .216) was found. The interaction driving condition × sleep condition was not significant (F1,11 = 0.022, P = .884), indicating that the effect of sleep restriction is the same under simulated and real driving conditions. The interaction driving condition × time was not significant (F6,66 = 2.364, P = .074) either. The interaction sleep condition × time was not significant (F6,66 = 2.246, P = .112 ), indicating a lack of difference in performance over time in the rested and sleep-restricted condition.

Real Driving

90 80 70 60 50 40 30 20 10 0 8h45

10h45

12h45

15h

17h

Rested condition

19h

21h

8h45

Time of day

10h45

12h45

15h

17h

19h

Sleep restricted condition

Figure 6—Impact of real versus simulated driving on fatigue, based on a visual analog scale (VAS) in the rested and sleep-deprived condition.

fatigue across time. Higher fatigue was seen at 10:45 AM than at 8:45 AM (F1,11 = 7.340, P = .020) (see Figure 6). No interaction was found. DISCUSSION Our study shows that driving (line crossings), RT performance, and sleepiness were unaffected by 12 hours of real and simulated driving but strongly affected by partial sleep restriction combined with extended wakefulness. Interestingly, 2 hours of sleep, probably rich in slow-wave activity, did not help sustain performance even at the very beginning of the trip. The results were similar for fatigue, except for a lack of effect for driving time. Simulated driving affected RT performance and sleepiness negatively but did not affect fatigue. The effects of sleep loss on driving, RT performance, and subjective sleepiness are very similar to those in our previous study.19 The effects of sleep loss on simulated driving are well established,25 as are the effects on subjective sleepiness25 and RT performance.36 Very little previous work is available on the effects on real-life driving, but one study30 has shown increased lane drifting in real-life driving in sleepy subjects. The present study, however, used lane crossing as a dependent variable and was carried out for a much longer duration than have been previous studies. One important finding in this study, which also confirmed the results of our previous study,19 is the lack of effect of the duration of the drive on performance and sleepiness. This goes against the common notion that the driving time is an important factor that needs to be controlled in order to promote road safety. It should be emphasized, however, that the present study permitted short active breaks every 2 hours (to test RT of our subjects). This may

Sleepiness From the KSS The results from the 3-factor ANOVA showed a significant main effect of “driving condition” (F1,11 = 8.036, P = .016) and of “sleep condition” (F1,11 = 58.294, P = .001), with higher sleepiness scores in the driving simulator and in the sleep-restricted condition. Neither the effect of time or the interaction terms were significant (see Figure 5). Fatigue From the VAS The results from the 3-way ANOVA showed no significant main effect of “driving condition” (F1,11 = 3.012, P = .111), but that of “sleep condition” was significant (F1,11 = 31.275, P = .000), with higher fatigue in the sleep-restricted condition (65.4 vs 39.5), as was that of “time” (F6,66 = 5.613, P = .011), with increasing SLEEP, Vol. 28, No. 12, 2005

Real Driving

10

1514

Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al

have prevented an accumulation of sleepiness. The results may have been different if the maximum “legal” driving time between breaks (5 hours) had been used. The importance of the break should be addressed in a separate study. Possibly, the presence of the copilot could have helped counteract sleepiness, but it should be emphasized that he did not interact with the driver other than in a situation of danger. Obviously, for safety reasons, we could not let our subjects drive alone under sleep deprivation. Incidentally, most other simulator studies show a rather strong increase in sleepiness and performance impairment during the first 20 to 40 minutes,37 presumably as a response to the relative monotony of the task. Our results do not show an increase of line crossings in the first 2 hours of real driving (in rested and sleep-deprived conditions), confirming that premature performance impairment in a simulated environment does not match with real driving. Our simulated and real driving sessions were quite different in terms of light exposure, proprioceptive feedback or mental stimulation, and the number of line crossings calculated in different ways, which could partially explain that line crossings were much more frequent in the simulated condition. Because we normalized our data through Z scores in each condition, we believe that these differences between driving condtions have been controlled and that our results are valid. Indeed, the measure of RT during the stops, which was performed in exactly the same condition of testing (duration, light exposure, scoring) agrees with the results observed in line crossings. Interestingly, ratings of fatigue showed a slightly different pattern apart from the clear response to sleep loss. Thus, fatigue increased with time driving but not with condition. This may be in line with the notion that fatigue is mainly a disinclination due to extended activity, as has been suggested by Grandjean.8 The reason for the negative effect of the simulator condition on RT performance and sleepiness is, very likely, the lower level of stimulation in the simulator—not much visuomotor input, a rather simple task, and no danger associated with errors. Still, there was a rather similar response of RT and sleepiness ratings to sleep loss in the 2 driving conditions. The general direction of the changes was also similar for line crossings, but there were no correlations between the 2 driving conditions in the rested condition and only partial correlations in the sleep-deprived condition. The observations seem to suggest that a simulator of the type used in the present study will produce results that may not be generalized to real-life driving, except perhaps on a group level. Our subjects were young and healthy, and it may be that older healthy subjects or people suffering from medical disorders might be more sensitive to simulator effects. In a previous study, we noticed that some subjects were not able to perform the simulated task.32 Many studies have also shown that subjects suffer from simulator sickness.38 This is a major issue, especially for testing older individuals, and seems to indicate that one should use real driving to quantify driving impairment in the most representative populations. One might also speculate that a driving task involving decision making (like stopping at a signal or passing a car in a dangerous situation) could be differently affected in the simulator versus real driving. Sleepiness in the middle of the drive was not a good predictor of future driving impairment, as we have shown in a previous study.19 Possibly only sleepiness at the exact time of driving has a direct correlation with actual driving performance. SLEEP, Vol. 28, No. 12, 2005

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 1515

Connor J, Norton R, Ameratunga S, et al. Driver sleepiness and risk of serious injury to car occupants: population based case control study. BMJ 2002;324:1125. Horne JA, Reyner LA. Sleep related vehicle accidents. BMJ 1995;310:565-7. Philip P, Vervialle F, Le Breton P, Taillard J, Horne JA. Fatigue, alcohol, and serious road crashes in France: factorial study of national data. BMJ 2001;322:829-30. Mitler MM, Miller JC, Lipsitz JJ, Walsh JK, Wylie CD. The sleep of long-haul truck drivers. N Engl J Med 1997;337:755-61. Philip P, Taillard J, Guilleminault C, Quera Salva MA, Bioulac B, Ohayon M. Long distance driving and self-induced sleep deprivation among automobile drivers. Sleep 1999;22:475-80. Gaba DM, Howard SK. Fatigue among clinicians and the safety of patients. N Engl J Med 2002;347:1249-55. Rajaratnam SM, Arendt J. Health in a 24-h society. Lancet 2001;358(9286):999-1005. Grandjean E. Fatigue in industry. Br J Ind Med 1979;36:175-86. Lal SK, Craig A. A critical review of the psychophysiology of driver fatigue. Biol Psychol 2001;55:173-94. Dement WC, Carskadon MA. Current perspectives on daytime sleepiness: the issues. Sleep 1982;5 Suppl 2:S56-66. Dijk DJ, Czeisler CA. Contribution of the circadian pacemaker and the sleep homeostat to sleep propensity, sleep structure, electroencephalographic slow waves, and sleep spindle activity in humans. J Neurosci 1995;15:3526-38. Carskadon MA, Dement WC. Effects of total sleep loss on sleep tendency. Percept Mot Skills 1979;48:495-506. Carskadon MA, Dement WC. Cumulative effects of sleep restriction on daytime sleepiness. Psychophysiology 1981;18:107-13. Froberg JE. Twenty-four-hour patterns in human performance, subjective and physiological variables and differences between morning and evening active subjects. Biol Psychol 1977;5:119-34. Powell NB, Schechtman KB, Riley RW, Li K, Troell R, Guilleminault C. The road to danger: the comparative risks of driving while sleepy. Laryngoscope 2001;111:887-93. Philip P, Sagaspe P, Taillard J, et al. Fatigue, sleep restriction, and performance in automobile drivers: a controlled study in a natural environment. Sleep 2003;26:277-80. O’Hanlon JF, Vermeeren A, Uiterwijk MM, van Veggel LM, Swijgman HF. Anxiolytics’ effects on the actual driving performance of patients and healthy volunteers in a standardized test. An integration of three studies. Neuropsychobiology 1995;31:81-8. O’Hanlon JF, Volkerts ER. Hypnotics and actual driving performance. Acta Psychiatr Scand Suppl 1986;332:95-104. Philip P, Sagaspe P, Moore N, et al. Fatigue, sleep restriction and driving performance. Accid Anal Prev 2005;37:473-8. Ramaekers JG, O’Hanlon JF. Acrivastine, terfenadine and diphenhydramine effects on driving performance as a function of dose and time after dosing. Eur J Clin Pharmacol 1994;47:261-6. Haraldsson PO, Carenfelt C, Laurell H, Tornros J. Driving vigilance simulator test. Acta Otolaryngol 1990;110:136-40. George CF, Boudreau AC, Smiley A. Simulated driving performance in patients with obstructive sleep apnea. Am J Respir Crit Care Med 1996;154:175-81. Lenne MG, Triggs TJ, Redman JR. Time of day variations in driving performance. Accid Anal Prev 1997;29:431-7. George CF. Vigilance impairment: assessment by driving simulators. Sleep 2000;23 Suppl 4:S115-8. Reyner LA, Horne JA. Evaluation “in-car” countermeasures to sleepiness: cold air and radio. Sleep 1998;21:46-50. Sagberg F. Road accidents caused by drivers falling asleep. Accid Anal Prev 1999;31:639-49. McCartt AT, Rohrbaugh JW, Hammer MC, Fuller SZ. Factors associated with falling asleep at the wheel among long-distance truck Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al

drivers. Accid Anal Prev 2000;32:493-504. 28. Delafosse JY, Leger D, Quera-Salva MA, Samson O, Adrien J. [Comparative study of actigraphy and ambulatory polysomnography in the assessment of adaptation to night shift work in nurses]. Rev Neurol (Paris) 2000;156:641-5. 29. Kushida CA, Chang A, Gadkary C, Guilleminault C, Carrillo O, Dement WC. Comparison of actigraphic, polysomnographic and subjective assessment of sleep parameters in sleep-disordered patients. Sleep Med 2001;2:389-96. 30. Ramaekers JG, Uiterwijk MM, O’Hanlon JF. Effects of loratadine and cetirizine on actual driving and psychometric test performance, and EEG during driving. Eur J Clin Pharmacol 1992;42:363-9. 31. Akerstedt T, Gillberg M. Subjective and objective sleepiness in the active individual. Int J Neurosci 1990;52:29-37. 32. Philip P, Taillard J, Klein E, et al. Effect of fatigue on performance measured by a driving simulator in automobile drivers. J Psychosom Res 2003;55:197-200. 33. Balkin TJ, Bliese PD, Belenky G, et al. Comparative utility of instruments for monitoring sleepiness-related performance decrements in the operational environment. J Sleep Res 2004;13:219-27. 34. Gillberg M, Kecklund G, Akerstedt T. Relations between performance and subjective ratings of sleepiness during a night awake. Sleep 1994;17:236-41. 35. Philip P, Taillard J, Sagaspe P, Valtat C, Sanchez-Ortuno M, Moore N, Charles A, Bioulac B. Age, performance and sleep deprivation. J. Sleep Res. 2004;13:105-110. 36. Dinges DF, Pack F, Williams K, et al. Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4-5 hours per night. Sleep 1997;20:267-7. 37. Reyner LA, Horne JA. Suppression of sleepiness in drivers: combination of caffeine with a short nap. Psychophysiology 1997;34:7215. 38. Min BC, Chung SC, Min YK, Sakamoto K. Psychophysiological evaluation of simulator sickness evoked by a graphic simulator. Appl Ergon 2004;35:549-56.

SLEEP, Vol. 28, No. 12, 2005

1516

Fatigue and Sleepiness in Simulated vs Real Driving—Philip et al