INSTRUMENTATION & METHODOLOGY Autonomic Arousal Index: an Automated Detection Based on Peripheral Arterial Tonometry Giora Pillar,1 Amir Bar,2 Arie Shlitner,1 Robert Schnall,2 Jacob Shefy,2 and Peretz Lavie1 1Sleep Laboratory, Bruce Rappaport Faculty of Medicine, Technion - Israel Institute of Technology, Haifa, Israel; 2Itamar Medical Ltd., Caesarea, Israel 46.2±14.4 years, BMI 28.5±5.4 kg/m2). All underwent a whole night PSG with simultaneous PAT recording. The PSG recordings were blindly manually analyzed for arousals based on American Academy of Sleep Medicine (AASM) criteria, while PAT was scored automatically. There was a significant correlation between PSG and PAT arousals (R=0.82, p<0.0001) with a good agreement across a wide range of values, with a ROC curve having an area under the curve (AUC) of 0.88. We conclude that automated analysis of the peripheral arterial tonometry signal can detect EEG arousals from sleep, in a relatively quick and reproducible fashion. Key words: Sleep; autonomic nervous system; arousals; peripheral arterial tonometry; sympathetic activation. INTRODUCTION nomic parameters such as heart rate and blood pressure, which are not necessarily associated with any visible EEG changes. This has led several authors to suggest that markers of autonomic activation (e.g., continuous blood pressure monitoring or pulse transit time) should be included in routine polysomnographic (PSG) recordings to assess the true extent of sleep fragmentation in the form of autonomic arousal index (AAI).13-17 Piston et al. reported a good correlation between PTT and EEG frequency shifts in response to external stimuli in normal subjects.14 The PTT could also, to some extent, detect sleep-disordered breathing events.13 Argod et al. found a reasonable agreement between standard scoring to PTT in detecting nonapneic obstructive respiratory events but reported a very high interobserver variability in the scoring of both (30%-37%).17 The recently developed technique of peripheral arterial tonometry (PAT) has been shown to be a sensitive marker of both tonic18 and transient sympathetic arousal during sleep.19 Here, we sought to examine and validate the efficacy of an automatic analysis of this signal along with the PAT-derived pulse rate in the detection of arousals from sleep. SLEEP FRAGMENTATION IN THE FORM OF FREQUENT “BRIEF AROUSALS” HAS BEEN IMPLICATED IN DAYTIME IMPAIRMENT OF COGNITIVE AND PSYCHOMOTOR PERFORMANCE, PARTICULARLY IN SLEEP RELATED BREATHING DISORDERS.1,2 Thus, the number of arousals is a useful marker of sleep quality, which is independent of traditional sleep-quality markers such as sleep efficiency and sleep latency. Based on the AASM criteria, an arousal from sleep is scored when there is a notable EEG shift toward a higher frequency for at least 3 seconds but no more than 15 seconds during all non-REM (NREM) stages of sleep, assuming sleep is recorded prior to and following the event for at least 10 seconds. Since EEG alpha waves or mixed frequency waves are common during rapid-eye movement (REM) sleep, the definition of arousal during REM sleep relies on a combination of EEG arousal and increased EMG or body movements.3 Since these criteria are not always clear-cut and are commonly difficult to determine, large inter-scorer variability in scoring arousals from sleep has been reported.4,5 Thus, an automatic and reliable method to detect arousals has been sought. It is well documented that arousals from sleep are associated with transient sympathetic activation.6-12 The term “autonomic arousal” is used to denote transient changes during sleep in auto- METHODS Subjects The study group consisted of 96 adult subjects, with 11 being healthy volunteers and 85 being patients referred with suspected obstructive sleep apnea (OSA) (Table 1). The average age was 46 years, with the patients older than the healthy volunteers, and the average BMI was 28.5kg/m2 with the patients heavier than the healthy volunteers. Twenty-three percent of the patients were hypertensive, and 5% were known to have ischemic heart disease. None of the patients had congestive heart failure. All the patients were drawn from the Technion Sleep Disorders Center Disclosure Statement Consultants or Employees of Itamar-Medical Ltd. Submitted for publication October 2001 Accepted for publication March 2002 Address correspondence to: Peretz Lavie, PhD, Sleep Laboratory, Faculty of Medicine, Gutwirth Building, Technion-Israel Institute of Technology, Haifa 32000, ISRAEL SLEEP, Vol. 25, No. 5, 2002 541 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 Summary: Arousals from sleep are associated with increased sympathetic activation and are therefore associated with peripheral vasoconstriction. We hypothesized that digital vasoconstrictions as measured by peripheral arterial tonometery (PAT), combined with an increase in pulse rate, would accurately reflect arousals from sleep, and can provide an autonomic arousal index (AAI). Based on a previously studied group of 40 sleep apnea patients simultaneously recorded by both polysomnography (PSG) and PAT systems, an automated algorithm using the PAT signal (and pulse rate derived from it) was developed for detection of arousals from sleep. This was further validated in a separate group of 96 subjects (85 patients referred with suspected obstructive sleep apnea and 11 healthy volunteers mean age sleep was scored as an arousal, while during REM sleep, an increase in EMG was required as well. In both cases at least 10 seconds of sleep period was required prior to and following the event in order to be scored as an arousal. Arousal index (ARI) was calculated by dividing the total number of arousals by the number of hours of sleep. Respiratory events were scored according to the AASM guidelines for measurement in clinical research.22 The respiratory disturbance index (RDI) was calculated as the number of apneas plus hypopneas divided by the number of hours of sleep. Periodic limb movements (PLM) were also scored based on AASM criteria, and the PLM index (PLMI) was calculated similarly to the other indices. Table 2—Distribution of the study population according to RDI The Peripheral Arterial Tonometry System All participants underwent a whole night PSG; (EEG 4214; Nihon Kohden, Kogyo Co., Tokyo, Japan) with simultaneous recordings of PAT signal (Site-PAT 200; Itamar- Medical Ltd. Caesarea, Israel). Prior to bedtime, patients completed a sleep questionnaire including physical data (e.g., weight and height), general health condition and medical history, medication usage, sleep habits, and the Epworth sleepiness scale (ESS).20 Lights off were no later than midnight, and lights were on at 06:00. The PAT signal was measured using the Sleep PAT-200TM device (Itamar-Medical LTD, Caesarea, Israel), which is a novel specifically designed noninvasive finger plethysmograph.18,19 The PAT probe consists of an optical sensor housed in a self pressurized finger-mounted shell, which generates a uniform pressure field surrounding the distal two phalanges of the finger (~50mmHg). The optical devices serve to measure pulsatile volume changes while the uniform pressure field unloads the arterial wall tension, improves the dynamic range of the measurement, and helps to reduce artifacts. The signal is further filtered (0.2 to 20Hz) and transferred as an analog input to an auxiliary direct current (DC) input channel of the PSG, where it is recorded simultaneously with the other standard channels. Upon completion of the recordings, the PSG data file, which includes the PAT data alongside the rest of the PSG data, is retrieved for sleep scoring and PAT automatic analysis, using specific software. This software employs algorithms based on weighting two features of the PAT signal, each of which indicates sympathetic surge: amplitude attenuation (reflecting vasoconstriction) and pulserate increase (comparable to heart-rate increase). The automatic algorithm was developed using a previous cohort of 40 patients (training set). The PAT arousal event is scored when one of the following has occurred: 1. The PAT signal amplitude attenuation, relative to baseline, of at least 50% or 2. The PAT signal amplitude attenuation of at least 30%, relative to baseline, with concurrent increase of more than 10% in pulse rate. Pulse-rate changes were derived from the PAT signal itself and the parallelism between a PAT-amplitude decrease and pulse-rate increase was defined using an overlapping window of 30 seconds. Polysomnography Scoring Overnight PSG was performed according to standard laboratory protocol, using computerized PSG with the following channels: two EEG (C3-A2 and O2-A1), EOG, submental EMG, arterial oxygen saturation, nasal-oral airflow (thermistors), electrocardiogram (ECG), chest and abdominal wall motion (piezo electrodes), bilateral anterior tibialis EMG, snoring (microphone placed one meter above the bed, and body position. Sleep was staged according to standard criteria.21 Arousals were defined according to the AASM guidelines.3 Any EEG frequency shift for at least 3 seconds but no more than 15 seconds during NREM The PSG recordings were manually analyzed for sleep stages, respiratory disturbances, movements, and arousals based on ASDA criteria, with the scorer being completely blinded to the PAT signal. To determine scoring reliability, a second scorer scored 20 records independently (randomly selected). The correlation between the results of the two scorers with respect to the total number of AASM-defined arousals was 0.96. This correlation is substantially higher than the previously reported ones, possibly as the two scorers have worked together for a long period of time (>15years). The scored data file, also including the Breathing Disorder Severity Normal (RDI: 0-10) Mild (RDI:11-20) Moderate (RDI: 21-40) Severe (RDI: 40+) Control Patients Overall N % N % N % 7 63.6 21 24.7 28 29.2 4 36.4 18 21.1 22 22.9 0 0 19 22.4 19 19.8 0 0 27 31.8 27 28.1 population and were referred for evaluation because of presumed OSAS. Normal volunteers were without any complaints of sleep disruption, daytime sleepiness, or snoring and were free of any disease or use of any medication. They were recruited via advertisements in the Technion’s Faculty of Medicine. The suspected OSA patients were excluded if they were on alpha-blocking medications or in unstable clinical condition. The study was approved by the Rambam Medical Center IRB, and patients signed an informed consent form prior to being studied. Protocol SLEEP, Vol. 25, No. 5, 2002 542 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 Table 1—Age and body habitus characteristics of the study group. Control Patients Overall Age N 11 85 96 (years) Mean 22.3 49.3 46.2 STD 6.6 12.0 14.4 Height Mean 172.8 179.9 172.0 (cm) STD 6.8 10.3 9.9 Weight Mean 65.9 87.4 85.0 (kg) STD 10.5 18.9 19.4 BMI Mean 22.0 29.3 28.5 (Kg/m2) STD 2.9 5.1 5.4 ASDA-ARI vs. PAT-ARI 140 120 80 60 40 r=0.82 20 0 0 20 40 60 80 100 120 140 PAT-ARI Figure 1—As can be seen, a very high and statistically significant correlation (R=0.82, p<0.0001) was found between the PAT-ARI (PAT-based arousal index) and the ASDA-ARI (arousal index based on ASDA criteria). The red dots represent patients with PLMs. recorded raw PAT signal, was then retrieved for automatic identification of PAT-based arousal events. This analysis takes into consideration the sleep staging as interpreted by the PSG scorer and processes the PAT signal within the sleep periods only. The average ESS score for the whole group was 9 ± 6.4, with 18% of participants reporting severe sleepiness (ESS>16), and 48% reporting virtually no sleepiness (ESS<8). The distribution of the study population according to RDI is presented in Table 2. The average RDI for the whole group was 30 ± 28 events per hour. As can be seen, the group had a wide range of sleep-disordered breathing, distributed from normal breathing (primarily 7 of the 11 normal volunteers, but also 21 of the referred patients) to severe OSAS. The average ASDA-ARI (EEG arousals) for the whole group was 27 ± 24 arousals per hour while the average PAT-AAI for the group was 29 ± 19 arousals per hour. Ten of the patients had also PLM, with an average PLM index of 32 ± 23 movements per hour. Figure 1 displays a scatter graph of the ASDA-ARI (gold standard) versus the PAT-AAI values for the whole study population with the calculated correlation coefficient. There was a good and statistically significant correlation between the two parameters (R=0.82, p<0.0001). There was also a significant correlation between RDI and both measures of arousals (0.92 for ASDA-ARI and 0.80 for PAT-AAI, p<0.05 for both). Analyzing the correlations between ASDA-ARI and PAT-AAI separately for the subjects with no OSA (RDI of 10 or less) and OSA patients revealed correlation coefficients of 0.58 and 0.84 respectively (p<0.05 for both). Figure 2 displays the Bland-Altman plot for ASDA-ARI and the PAT-AAI for the study population. There was a good agreement between the two parameters across a wide range of arousal indexes. Figure 3 shows the ROC curve for identifying patients with pathologic AAI with the PAT (defined by a threshold of 20 arousals per hour of sleep23) when taking the ASDAbased scoring as “gold standard”, with an AUC of 0.88. Since PLM may be associated with subcortical autonomic arousals, occasionally without meeting criteria to be scored as Statistical Analysis Analysis was performed to assess the correlation between the PAT-based AAI (PAT-AAI) and the ASDA-based arousal index derived from the PSG scoring (ASDA-ARI) using Pearson correlation coefficient and a Bland-Altman plot. In order to evaluate the efficacy of the PAT automatic analysis as a potential tool for diagnosing a disorder of arousals from sleep, a receiver operating characteristic (ROC) curve based on a threshold of AAI=2023 for abnormality was plotted and its area under the curve (AUC) was calculated. This curve joins points that represent the sensitivity and specificity for all of possible PAT-AAI thresholds on an X-Y plane (x=1-specificity, y=sensitivity) for all possible values of the PAT-AAI thresholds. The area under the ROC curve (i.e., AUC) is considered a measure of the overall efficacy of the score–an AUC value of 0.5 indicates a nonsignificant score for separating normals from patients with pathologic AAI, and a value close to 1.0 indicates a very efficient score. In addition, the correlation of both the ASDA-ARI and PATAAI with ESS was performed to examine whether ASDA-based arousals or PAT-based autonomic arousals, or both, correlates with subjective sleepiness. RESULTS Of the 96 sleep studies, complete data were available for 94. In 2 patients (2.1%), the data had to be discarded because of technical problems. SLEEP, Vol. 25, No. 5, 2002 543 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 ASDA-ARI Score 100 Bland Altman Presentation of ASDA-ARI vs. PAT-ARI 60 20 0 -20 -40 -60 0 20 40 60 80 100 120 140 Mean of (ASDA-ARI) and (PAT-ARI) Figure 2—As can be seen, across a wide range of arousals frequencies, there was a good agreement between PAT-ARI (PAT-based arousal index) and the ASDAARI (arousal index based on ASDA criteria). The red dots represent patients with PLMs. ASDA-based arousals shown in the EEG recording,24-27 the statistical analysis was repeated after adding PLM events that were not associated with ASDA-arousals to the total “gold standard” arousals. Such events were noted in 10 patients. This resulted in an improvement of the correlation between ASDA-based arousals and PAT-AAI (R=0.88, p<0.0001), and the calculated AUC minimally increased to 0.89. Finally, although significant, there was only a poor correlation between the arousal index (either by standard criteria or by PAT) and subjective sleepiness as assessed by ESS (R=0.43, and R=0.35, respectively, p<0.001 for both). scoring ASDA arousals from sleep by experts from 14 sleep laboratories.5 In another study Loredo et al evaluated various types of arousals and showed good interscorer reproducibility in scoring PLMs or respiratory events accompanied arousals characterized by increased EMG. Poor reproducibility was demonstrated in the scoring of “classic” ASDA-defined arousals.4 Thus, an automated or other well-defined method of detecting sleep fragmentation seems warranted. Two additional methods based on sympathetic activation have been attempted previously. The PTT demonstrated some promising results initially, but has not been compared in large-scale studies to standard measures and has been shown to have the disadvantage of high interobserver variability.17 Arousal and inspiratory blood pressure changes were used by Davies et al15 as a potential marker of disturbed sleep and disordered breathing in sleep. Although the blood-pressure profile appeared to be useful to identify patients with OSA, this method has not been further investigated. One of the advantages of the PAT method is its’ automatic computer-based analysis, which makes it reproducible, objective, and time saving. All three methods of comparisons between the ASDA and PAT-based AAI revealed reasonably good agreement, suggesting that the PAT based algorithm may accurately reflect sleep fragmentation. Although we have not quantified time saving using PAT automated scoring in comparison with PSG-based manual ones, we estimate that this method has saved approximately 20 to 50 minutes per record. Other potential advantages of this system are its simplicity and minimal required technical intervention in patient preparations and recording. Also, the applied pressure to the finger of approximately 50 mmHg did not cause any discomfort to the subjects. DISCUSSION The primary finding of this study is that the standard ASDAbased “EEG arousals” can be reasonably accurately predicted by measuring “autonomic arousals” at the level of the finger. This is consistent with the observation that arousals from sleep are associated with sympathetic activation and, therefore, can be readily measured by the PAT. Utilizing two important sympathetic-related parameters—digital vasoconstriction and pulserate changes, we could develop an automatic and reliable algorithm for identifying autonomic arousals throughout sleep. Since the PAT analysis is performed automatically, this is a relatively simple, objective, and reproducible method, which can be a very useful and important channel added to the standard PSG. Brief arousals from sleep may impair cognitive and psychomotor performance, particularly in sleep-related breathing disorders.1,2 However, EEG arousals, as currently defined, are not easily reproducible primarily due to a considerably large interscorer variability. Drinnan et al found large disagreement in SLEEP, Vol. 25, No. 5, 2002 544 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 Delta of (ASDA-ARI) and (PAT-ARI) 40 1.0 0.9 0.8 Sensitivity 0.7 0.6 0.5 0.3 0.2 AUC=0.88 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1 - Specificity Figure 3—ROC curve for identifying pathologic arousals from sleep (threshold =20 arousals per hour of sleep) based on PAT vs. standard criteria. As can be seen, the area under the curve is 0.88, yielding potentially high sensitivity and specificity in diagnosing pathological arousal frequency by PAT, when the gold standard is the ASDA criteria based on PSG. Both the EEG and the autonomic arousals were poorly correlated with subjective sleepiness, probably demonstrating either poor judgment of the patients to subjectively assess their sleepiness level, or an objective limitation of the EEG arousal index or AAI to predict it. This is not surprising, as previous studies have also failed to show good correlation between arousals from sleep and daytime sleepiness.13,28,29 As this study showed good agreement between the PAT-defined AAI and the EEG-defined arousals index, the poor correlation between the PAT-AAI and subjective daytime sleepiness assessment were also expected. When originally introduced the ESS demonstrated good correlation with objective sleepiness measures (MSLT),20 but later studies failed to replicate this.30 Although the 0.35 correlation between the PAT-AAI and the ESS is somewhat small, it should be kept in mind that PAT-AAI is a state measure reflecting a specific night while the ESS is a trait measure reflecting a much longer period. The ability of the PAT to predict daytime sleepiness needs further evaluation, probably with more objective modalities such as MSLT (which is also a state measure). It has been suggested that arousals from sleep are associated with increased sympathetic activation.6-10 Therefore, it is not surprising that we have found a good match between EEG arousals and autonomic arousals. Furthermore, this good agreement suggests that sympathetic activation is indeed a result of arousals from sleep rather than other potential triggers such as hypoxemia or hypercapnia. It has been previously suggested that both hypoxemia and hypercapnia, acting via chemoreflexes, may result in an acute increase in sympathetic activity to blood vessels in the muscles.7,31-34 Thus, the increased sympathetic activation SLEEP, Vol. 25, No. 5, 2002 in patients with OSA could well be the result of abnormal arterial gas levels. Furthermore, several studies, using direct intraneural measurements of sympathetic traffic to blood vessels in the muscles have found that patients with sleep apnea have very high levels of sympathetic flow, even during normoxic wakefulness, which may explain their increased risk of arterial hypertension.34,35 However, the blood-gas changes are not the only potential factor contributing to the increased sympathetic activation in sleep apnea. In their classic dog model, Brooks et al found that acoustic-stimulated arousals from sleep resulted in abrupt increases in blood pressure, although during apnea-stimulated arousals the high blood pressure persisted to wakefulness.36,37 In another study aiming at determining the effects of arousals from sleep per se on the sympathetic outputs to the cardiovascular system, Horner et al reported that these resulted in phasic sympathetic activation, as assessed by increased heart rate and blood pressure.6 Although we have not quantified blood gas in this study, our finding that peripheral arterial vasoconstrictions with increased pulse rate correlated well with arousals from sleep (determined by EEG) both in a wide severity range of OSA patients and in normal healthy volunteers, supports the finding that arousals from sleep per se mediate the increased sympathetic activation, without hypoxemia or hypercapnia, as in the normal volunteers these arousals are not accompanied by blood-gas changes. This is consistent with the finding from the Sleep Health Heart Study that arousals from sleep were associated (although weakly) with hypertension.38 Of note, Morrel et al39 reported that an index of sleep fragmentation was significantly associated with awake systolic blood pressure in subjects with 545 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 0.4 REFERENCES 1. Martin SE, Engleman HM, Deary IJ, Douglas NJ. The effect of sleep fragmentation on daytime function. Am J Respir Crit Care Med 1996;153:1328-32. 2. Martin SE, Wraith PK, Deary IJ, Douglas NJ. The effect of nonvisible sleep fragmentation on daytime function. Am J Respir Crit Care Med 1997;155:1596-601. 3. EEG arousals: scoring rules and examples. a preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep 1992;15:174-84. 4. Loredo JS, Clausen JL, Ancoli-Israel S, Dimsdale JE. Night-tonight arousal variability and interscorer reliability of arousal measurements. Sleep 1999;22:916-20. 5. Drinnan MJ, Murray A, Griffiths CJ, Gibson GJ. Interobserver variability in recognizing arousal in respiratory sleep disorders. Am J Respir Crit Care Med 1998;158:358-62. 6. Horner RL, Brooks D, Kozar LF, Tse S, Phillipson EA. Immediate effects of arousal from sleep on cardiac autonomic outflow in the absence of breathing in dogs. J Appl Physiol 1995;79:151-62. 7. Somers VK, Dyken ME, Mark AL, Abboud FM. Sympatheticnerve activity during sleep in normal subjects. N Engl J Med 1993;328:303-7. 8. Morgan BJ, Crabtree DC, Puleo DS, Badr MS, Toiber F, Skatrud JB. Neurocirculatory consequences of abrupt change in sleep-state in humans. J Appl Physiol 1996;80:1627-1636. 9. Hornyak M, Cejnar M, Elam M, Wallin BG. Muscle sympathetic nerve activity during sleep in man. Brain 1991;114:1281-1295. 10. Okada H, Iwase S, Mano T, Sugiyama Y, Watanabe T. Changes in muscle sympathetic nerve activity during sleep in humans. Neurology 1991;41:1961-1966. 11. Schneider H, Schaub CD, Chen CA, et al. Effects of arousal and SLEEP, Vol. 25, No. 5, 2002 546 AAI: An Automated Detection—Pillar et al Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 sleep state on systemic and pulmonary hemodynamics in obstructive sleep apnea. J Appl Physiol 2000;88:1084-92. 12. Launois SH, Averill N, Abraham JH, Kirby DA, Weiss JW. Cardiovascular responses to nonrespiratory and respiratory arousals in a porcine model. J Appl Physiol 2001;90:114-120. 13. Pitson DJ, Stradling JR. Autonomic markers of arousal during sleep in patients undergoing investigation for obstructive sleep apnoea, their relationship to EEG arousals, respiratory events and subjective sleepiness. J Sleep Res 1998;7:53-9. 14. Pitson D, Chhina N, Knijn S, van Herwaaden M, Stradling J. Changes in pulse transit time and pulse rate as markers of arousal from sleep in normal subjects. Clin Sci (Colch) 1994;87:269-73. 15. Davies RJ, Vardi-Visy K, Clarke M, Stradling JR. Identification of sleep disruption and sleep disordered breathing from the systolic blood pressure profile. Thorax 1993;48:1242-7. 16. Davies RJ, Belt PJ, Roberts SJ, Ali NJ, Stradling JR. Arterial blood pressure responses to graded transient arousal from sleep in normal humans. J Appl Physiol 1993;74:1123-30. 17. Argod J, Pepin JL, Smith RP, Levy P. Comparison of esophageal pressure with pulse transit time as a measure of respiratory effort for scoring obstructive nonapneic respiratory events. Am J Respir Crit Care Med 2000;162:87-93. 18. Lavie P, Schnall RP, Sheffy J, Shlitner A. Peripheral vasoconstriction during REM sleep detected by a new plethismographic method. Nature Med 2000;6:606. 19. Schnall RP, Shlitner A, Sheffy J, Kedar R, Lavie P. Periodic, profound peripheral vasoconstriction - a new marker of obstructive sleep apnea. Sleep 1999;22:939-946. 20. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 1991;14:540-5. 21. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Los Angeles: Brain Information Service/Brain Research Institute, UCLA, 1968. 22. Sleep-related breathing disorders in adults: Recommendations for syndrome definition and measurement techniques in adults. The report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:667-689. 23. Mathur R, Douglas NJ. Frequency of EEG arousals from nocturnal sleep in normal subjects. Sleep 1995;18:330-333. 24. Sforza E, et al. EEG and cardiac activation during periodic leg movements in sleep: support for a hierarchy of arousal responses. Neurology. 1999;52:786-91. 25. Winkelman JW. The evoked heart rate response to periodic leg movements of sleep. Sleep. 1999 22:575-80. 26. Karadeniz A, Ondze B, Desset A, Billiard M. EEG arousals and awakening in relation with periodic leg movements during sleep. J Sleep Res 2000;9:273-277. 27. Ali NJ, Davies RJ, Fleetham JA, Stradling JR. Periodic movements of the legs during sleep associated with rises in systemic blood pressure. Sleep 1991;14:163-5. 28. Kingshott RN, Engleman HM, Deary IJ, Douglas NJ. Does arousal frequency predict daytime function? Eur Resp J 1998;12:1264-70. 29. Stradling JR, Barbour C, Glennon J, Langford BA, Crosby JH. Prevalence of sleepiness and its relation to autonomic evidence of arousals and increased inspiratory effort in a community based population of men and women. J Sleep Res 2000;9:381-8. 30. Chervin RD, et al. The Epworth Sleepiness Scale may not reflect objective measures of sleepiness or sleep apnea. Neurology. 1999;52:125-31. 31. Somers VK, Zavala DC, Mark AL, Abboud FM. Contrasting effects of hypoxia and hypercapnia on ventilation and sympathetic activity in humans. J Appl Physiol 1989;67:2101-2106. 32. Somers VK, Zavala DC, Mark AL, Abboud FM. Influence of ventilation and hypocapnia on sympathetic nerve responses to hypoxia in normal humans. J Appl Physiol 1989;67:2095-2100. RDI less than one. In some specific circumstances, autonomic arousals can be seen without clear changes in the EEG (“cortical arousals”). One of these conditions is PLM during sleep.24-27 For this reason, we repeated the comparison between the PAT-defined arousals and the ASDA “gold standard” when adding PLM events not associated with classic EEG arousals to the total ASDA-based arousals counting. This resulted in an improvement of the correlation coefficient between the two indexes from 0.82 to 0.88, while the AUC essentially remained unchanged. This result suggests that the EEG-based criteria as nowadays defined (ASDA criteria), are not sensitive enough and some types of sleep fragmentation, such as in the case of PLM, may not be identified. It should be emphasized, however, that autonomic nervous system arousals do not necessarily reach the cortex, and these 2 measures may represent two separate phenomena. While considering this technique as a good measure of arousals from sleep, the population investigated is one limitation of the study. It consisted of healthy volunteers and patients with snoring. One could argue that in other populations, such as insomniacs, arousals from sleep may be associated with a different pattern of autonomic activation. Thus, expanding this study to other populations will add more information regarding their pattern of autonomic arousals. Despite this limitation, we believe that our study supports the concept that sleep fragmentation can be accurately assessed by measurements of peripheral arterial tone at the finger, and by using automated algorithm, an important contribution can be made to the standard PSG by means of reproducibility, time savings, and objectivity. SLEEP, Vol. 25, No. 5, 2002 Downloaded from https://academic.oup.com/sleep/article-abstract/25/5/541/2750145 by guest on 11 June 2020 33. Somers VK, Abboud FM. Chemoreflexes—responses, interactions and implications for sleep apnea. Sleep 1993;16:S30-3; discussion S334. 34. Somers VK, Dyken ME, Clary MP, Abboud FM. Sympathetic neural mechanisms in obstructive sleep apnea. J Clin Invest 1995;96:1897904. 35. Hedner J, Ejnell H, Sellgren J, Hedner T, Wallin G. Is high and fluctuating muscle nerve sympathetic activity in the sleep apnea syndrome of pathogenetic importance for the development of hypertension? J Hypertens 1988;6:S529-S531. 36. Brooks D, Horner RL, Kimoff RJ, Kozar LF, Render-Teixeira CL, Phillipson EA. Effect of obstructive sleep apnea versus sleep fragmentation on responses to airway occlusion. Am J Resp Crit Care Med 1997;155:1609-17. 37. Brooks D, Horner RL, Kozar LF, Render-Teixeira CL, Phillipson EA. Obstructive sleep apnea as a cause of systemic hypertension. Evidence from a canine model. J Clin Invest 1997;99:106-9. 38. Nieto FJ, Young TB, Lind BK, et al. Association of sleep-disordered breathing, sleep apnea, and hypertension in a large communitybased study. Sleep Heart Health Study. JAMA 2000;283:1829-36. 39. Morrell MJ, Finn L, Kim H, Peppard PE, Badr MS, Young T. Sleep fragmentation, awake blood pressure, and sleep-disordered breathing in a population-based study. Am J Resp Crit Care Med 2000;162:2091-96. 547 AAI: An Automated Detection—Pillar et al