- Email: [email protected]
Pilot study on the validity of the pupillographic sleepiness test in children and adolescents
Sleep Medicine 15 (2014) 720–723
Contents lists available at ScienceDirect
Sleep Medicine journal homepage: www.elsevier.com/locate/sleep
Brief Communication
Pilot study on the validity of the pupillographic sleepiness test in children and adolescents Michael S. Urschitz a,b,⇑,1, Katrin Heine b,1, Lea Mendler b, Tobias Peters c, Barbara Wilhelm c, Christian F. Poets b a b c
Unit of Pediatric Epidemiology, Institute of Medical Biostatistics, Epidemiology, and Informatics, University Medical Center of the Johannes Gutenberg University Mainz, Germany Working Group on Pediatric Sleep Medicine, University Children’s Hospital Tuebingen, Germany Pupil Research Group, Centre for Ophthalmology, University Hospital Tuebingen, Germany
a r t i c l e
i n f o
Article history: Received 4 November 2013 Received in revised form 12 February 2014 Accepted 16 February 2014 Available online 13 April 2014 Keywords: Central nervous activation Alertness Hypersomnia Sleepiness Pupillography Multiple sleep latency test
a b s t r a c t Objective: To report preliminary validation data for the pupillographic sleepiness test (PST) in children and adolescents. Methods: Twelve patients (13.1 ± 4.4 years of age) underwent the multiple sleep latency test (MLST) and three PSTs at 09:00, 11:00, and 13:00 on one single day. Correlations were tested between mean sleep latency and gender-adjusted z-values of the natural logarithm of the pupillary unrest index (zlnPUI). Results: Spearman’s correlation (P-value) between the zlnPUI values obtained at 09:00 and 11:00 with the MSL was rS = 0.641 (0.025) and r = 0.553 (0.062). Conclusion: There was satisfactory agreement between PST and the MLST, which is similar to what is found in adults. The PST may be promising for the evaluation of daytime sleepiness in children and adolescents, and should be further evaluated in future studies. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Excessive daytime sleepiness (EDS) is a widespread problem in children and adolescents and can have major negative effects on performance, health, and safety [1]. Evaluations usually include some kind of structured sleep history, sleep logs, sleep questionnaires, and objective tests for the state domain of EDS (eg the multiple sleep latency test [MSLT] [2]. Lowenstein et al. were the first to observe typical fluctuations in pupil size in sleepy adults [3], and the first clinical application of pupillography was in narcolepsy [4]. The pupillographic sleepiness test (PST) is a standardized, accurate, and reliable physiological test for the level of EDS in adults [5] and has also been discussed for the laboratory assessment of EDS in children [6,7]. However, feasibility and accuracy of the PST are unknown in pediatric patients, hampering its widespread pediatric use. As part of an interdisciplinary project (TUPEDS: Tuebingen Project on EDS in Childhood), we had demonstrated the feasibility of ⇑ Corresponding author. Address: Unit of Pediatric Epidemiology, Institute of Medical Biostatistics, Epidemiology, and Informatics, University Medical Center of the Johannes Gutenberg University Mainz, Obere Zahlbacher Str. 69, 55131 Mainz, Germany. Tel.: +49 6131 17 3122; fax: +49 6131 17 2968. E-mail address: [email protected] (M.S. Urschitz). 1 M.S. Urschitz and K. Heine contributed equally to this manuscript. http://dx.doi.org/10.1016/j.sleep.2014.02.008 1389-9457/Ó 2014 Elsevier B.V. All rights reserved.
the PST in a field setting and reported preliminary reference values for the pupillary unrest index (PUI), the main sleepiness parameter of the PST [8]. In the current report from TUPEDS, we present preliminary validation data for the PST. 2. Methods 2.1. Study design A prospective diagnostic test pilot study was performed with the MSLT as reference standard and the PST as index test. Patients underwent a standard overnight polysomnography in a sleep laboratory, followed by multiple PST and MSLT assessments the next day. The study design was approved by the Ethics Committee of Tuebingen University Hospital; written informed parental and child’s consent were obtained. The pilot study was stopped after 12 patients, and an interim analysis was performed. 2.2. Subjects Eligible subjects were children referred to the sleep disorders center of the University Children’s Hospital Tuebingen between October 2011 and January 2013. Inclusion criteria were: (i) 6–18 years of age and (ii) history of EDS and/or referral for the
721
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723
evaluation of EDS. Exclusion criteria were: (i) developmental or cognitive disorder impairing patient’s compliance and (ii) restricted pupillary motility.
starting at 09:30, shortly after performance of the PST. Manual scoring was performed by one of the authors (M.S.U.), who was blinded to PST results. The sleep latency to the first epoch of sleep was calculated on each nap attempt and the mean sleep latency (MSL) computed across all naps, assigning 20 min for those periods in which no sleep was obtained. A shorter MSL corresponded to more sleepiness.
2.3. PST A detailed description of the PST is given elsewhere [5,8]. In short, the PST was performed according to standard operating procedures established in adults [9]. The procedures were not modified for application in children. Measures were obtained three times at 2 h intervals starting at 09:00. Recordings of spontaneous pupillary oscillations were acquired for 11 min by infra-red video pupillography in a fully darkened and quiet room [5]. During the test, subjects fixated a red light source at the lens aperture of the video camera in an upright position. To block ambient light fully, subjects wore goggles with black lenses and infrared filters. Pupillary oscillations were registered, quantified, and analyzed automatically by the system [10]. The entire 11 min recording was divided into eight segments each comprising 82 s recording time and 2048 data points. Before off-line analysis, missing data values were replaced by linear interpolation. The amount of interpolation was calculated as percentage of recording time. A sufficient recording quality was defined as interpolation <40% of recording time [8]. Based on the pupil diameter, the PUI (mm/min) was calculated, which corresponded to a low pass filtering of the data set [10]. In the present study, PUI trend data were printed out and visually inspected for artifacts by two of the authors (B.W., T.P.). Artifacts were handled and PUI manually recalculated according to a previously described protocol without knowledge of MSLT results [8]. Higher PUI values corresponded to more sleepiness. Immediately after each PST, subjective sleepiness was investigated using a German version of the Stanford Sleepiness Scale (SSS) [11]. A higher SSS value corresponded to more sleepiness.
2.5. Statistical analysis Descriptive statistics including mean, standard deviation (SD), and range were used to summarize demographic and clinical characteristics. To obtain a normal distribution, the natural logarithm of the PUI (lnPUI) was calculated. To account for gender differences in children, lnPUI values were transformed into lnPUI z-values (z-lnPUI) using the mean and SD of previously obtained pediatric reference values (i.e. 2.01 ± 0.43 for boys and 1.93 ± 0.43 for girls) and the formula: z-lnPUI = (lnPUI mean)/SD [8]. In accordance with the MSL, intra-individual means were calculated for lnPUI, z-lnPUI, and SSS. Time-of-day variations were assessed by scatter plots and fitting lines of simple linear regression. Agreement of the mean z-lnPUI and the three single z-lnPUI values obtained at 09:00, 11:00, and 13:00 with the MSL and the three single SSS values was evaluated using Spearman’s rank correlation coefficient, rS. Due to the exploratory character of the pilot study, no sample size calculation and no statistical hypothesis test was performed. Hence, P-values for rS were calculated for illustrative purposes rather than hypothesis testing. All analyses were done with statistical software (IBM SPSS Statistics 20). 3. Results Twelve children (five boys, seven girls) with a mean ± SD age (range) of 13.1 ± 4.4 (7–18) years were enrolled and 36 PST recordings successfully performed (Table 1). No recording was terminated prematurely. Interpolation ranged from 1.2% to 67.1% of recording time (mean ± SD, 14.1 ± 15.3). Only one recording had an interpolation of >40%; this child was extremely sleepy,
2.4. MSLT The MSLT was performed according to a current guideline [12]. It included three to five 20 min nap attempts at 2 h intervals Table 1 Clinical characteristics of study patients (n = 12). No.
Sex
Age (years)
Diagnosis
Mean sleep latency (min)
SOREMPs (no.)
Mean SSS
1
Female
11
Narcolepsy with cataplexy
1.1
>2
4.7
2
Female
11
5.1
1
4.0
3
Female
17
2.5
0
3.3
4
Female
7
Unclear excessive daytime sleepiness Suspected idiopathic hypersomnia Obstructive sleep apnea
20.0
0
2.3
5
Male
18
Insufficient sleep syndrome
6.5
1
3.3
6
Female
14
Obstructive sleep apnea
1.6
0
4.0
7
Male
12
8.4
0
2.7
8
Female
18
Suspected narcolepsy without cataplexy Chronic fatigue
16.8
0
3.0
9
Male
14
Insufficient sleep syndrome
2.8
2
3.3
10
Male
10
17.0
0
2.7
11
Female
13
Periodic limb movement disorder Narcolepsy with cataplexy
8.0
3
3.0
12
Male
10
Habitual snoring
20.0
0
2.0
lnPUI (z-lnPUI) 09:00
11:00
13:00
2.74 (+1.88) 3.00 (+2.49) 3.49 (+3.63) 2.17 (+0.57) 1.42 ( 1.37) 3.23 (+3.02) 2.50 (+1.14) 1.76 ( 0.40) 2.71 (+1.64) 2.47 (+1.08) 1.59 ( 0.80) 2.25 (+0.55)
2.63 (+1.63) 2.95 (+2.37) 2.38 (+1.05) 2.12 (+0.45) 1.20 ( 1.88) 2.52 (+1.37) 2.32 (+0.72) 1.89 ( 0.10) 2.85 (+1.95) 2.40 (+0.91) 1.36 ( 1.33) 2.36 (+0.81)
2.53 (+1.40) 2.97 (+2.41) 2.44 (+1.19) 2.25 (+0.75) 0.72 ( 2.99) 2.41 (+1.12) 1.74 ( 0.62) 1.96 (+0.06) 2.06 (+0.12) 2.76 (+1.75) 1.65 ( 0.65) 2.44 (+1.00)
SOREMPs, sleep-onset rapid eye movement periods; SSS, Stanford Sleepiness Scale; PUI, pupillary unrest index; ln, natural logarithm.
Mean lnPUI
Mean zlnPUI
2.63
+1.64
2.97
+2.42
2.77
+1.96
2.18
+0.59
1.11
2.08
2.72
+1.84
2.19
+0.41
1.87
0.15
2.54
+1.23
2.55
+1.25
1.53
0.92
2.35
+0.79
722
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723
Fig. 1. Distribution of sleepiness parameters stratified by time of day.
repeatedly closed his eyes, and had high lnPUI (3.23) and z-lnPUI values (+3.02). However, his results seemed valid and the respective recording remained in the analysis. Considering all 36 recordings, SSS, sleep latency, lnPUI, and z-lnPUI ranged from 1 to 6 points, from 0.5 to 20 min, from 0.72 to 3.49, and from 2.99 to +3.63, respectively. In 13, three, and two recordings, z-lnPUI was >+1, >+2, and >+3, respectively. Considering all 12 children, mean SSS, MSL, mean lnPUI, and mean z-lnPUI ranged from 2.0 to 4.7 points, from 1.1 to 20 min, from 1.11 to 2.97, and from 2.08 to +2.42. Six children had MSL < 8 min. These children had higher mean lnPUI (mean ± SD, 2.46 ± 0.67) and higher mean z-lnPUI values (+1.17 ± 1.64) compared to children with a mean sleep latency of P8 min (mean lnPUI, 2.11 ± 0.36; mean z-lnPUI, +0.33 ± 0.76). On a descriptive basis, all parameters indicated that daytime sleepiness was decreasing with increasing time of day (Fig. 1). This was evident for lnPUI (decreasing), SSS (decreasing), and sleep latency (increasing), thereby showing some agreement in the time-of-day variation. rS (P-value) between the three paired z-lnPUI SSS values obtained at 9:00, 11:00, and 13:00 was 0.656 (0.020) for 9:00, 0.205 (0.523) for 11:00, and 0.311 (0.325) for 13:00. rS (P-value) for the mean z-lnPUI and the three single z-lnPUI values obtained at 9:00, 11:00, and 13:00 with the MSL was 0.501 (0.097) for the mean z-lnPUI, 0.641 (0.025) for the z-lnPUI obtained at 09:00, 0.553 (0.062) for the z-lnPUI obtained at 11:00, and 0.242 (0.449) for the z-lnPUI obtained at 13:00.
In theory, the propensity to fall asleep and pupillary instability are linked via the level of tonic central nervous activation in the locus coeruleus [15]. Within this theoretical framework an association between sleep latency and PUI may be expected. The few correlation studies in healthy adults found moderate agreement between MSLT and PST, depending on the definition of sleep latency (N1 or N2 sleep) [16] and the pupillographic variable used (pupil diameter, coefficient of variation, power of the 0.1–0.8 Hz spectrum, or PUI) [17]. In the latter study, the median of the intra-individual rS was 0.46 for the agreement between sleep latency and PUI (interquartile range, 0.38 to 0.69 [17]). In sleepy adults, rS between sleep latency and PUI was 0.44 [18] and 0.40 [19]. Variation in agreement with time of day has been reported only in one study [20]; the weakest rS ( 0.04) was reached at 14:00, corresponding to the observation in the present study. We therefore conclude that the PST should be performed in the morning or to be predictive of results of the MSLT. In conclusion, the present study is largely in line with the adult literature showing moderate agreement between MSLT and PST with considerable variation across the time of day. Hence, accuracy and precision of the PST in children may be comparable with what is seen in adults. This may encourage the scientific community to investigate further the usability and prognostic ability of the PST in children. Funding sources None.
4. Conclusions In the present pilot study, there was satisfactory agreement between the MSLT and PST in children, depending on the PST variable and time of day. The best agreement was found for PST results obtained at 09:00 and 11:00. The intra-individual mean of three PST recordings may not have better agreement compared to one single PST recording performed at 09:00. These results suggest that higher PUI values are indeed linked to an increased propensity to fall asleep and therefore associated with a shorter sleep latency, particularly in the morning and forenoon. To our knowledge, this is the first validation study on the PST in children, which suggests that PST variables reflect sleepiness in children in the same manner as in adults. A complement to the MSLT is desirable because MSLT sleep latency shows a large overlap between healthy subjects and subjects with sleep disorders [12]. Some authors have even questioned its status as a gold standard for the evaluation of sleepiness [13]. In addition, it is costly and time-consuming, lacks sufficient normative data in pediatric patients, and suffers from motivational influences and the last nap effect [14]. Consequently, a feasible, convenient, and time-saving complementary method that is also less dependent on motivation would be a major improvement.
Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: http://dx.doi.org/10.1016/j.sleep.2014.02.008.
References [1] Millman RP. Excessive sleepiness in adolescents and young adults: causes, consequences, and treatment strategies. Pediatrics 2005;115:1774–86. [2] Carskadon MA, Dement WC, Mitler MM, Roth T, Westbrook PR, Keenan S. Guidelines for the multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep 1986;9:519–24. [3] Lowenstein O, Feinberg R, Loewenfeld IE. Pupillary movements during acute and chronic fatigue. Invest Ophthalmol 1963;2:138–57. [4] Yoss RE, Moyer NJ, Ogle KN. The pupillogram and narcolepsy. A method to measure decreased levels of wakefulness. Neurology 1969;19:921–8. [5] Wilhelm B, Wilhelm H, Lüdtke H, Streicher P, Adler M. Pupillographic assessment of sleepiness in sleep-deprived healthy subjects. Sleep 1998;21:258–65. [6] Kotagal S, Goulding PM. The laboratory assessment of daytime sleepiness in childhood. J Clin Neurophysiol 1996;13:208–18. [7] Urschitz MS. Assessing objective daytime sleepiness in children and adults: do we have appropriate instruments? Sleep Med 2013;14:812–3.
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723 [8] Urschitz MS, Heine K, Brockmann PE, Peters T, Durst W, Poets CF, et al. Subjective and objective daytime sleepiness in school children and adolescents. Results of a community-based study. Sleep Med 2013;14:1005–12. [9] Wilhelm B, Koerner A, Heldmaier K, Moll K, Wilhelm H, Luedtke H. Normal values of the pupillographic sleepiness test in male and female subjects aged 20 to 60 years. Somnologie 2001;5:115–20. [10] Lüdtke H, Wilhelm B, Adler M, Schaeffel F, Wilhelm H. Mathematical procedures in data recording and processing of pupillary fatigue waves. Vision Res 1998;38:2889–96. [11] Hoddes E, Dement WC, Zarcone V. The history and use of the Stanford Sleepiness Scale. Psychophysiology 1971;9:150. [12] Littner MR, Kushida C, Wise M, Davila DG, Morgenthaler T, Lee-Chiong T, et al. Practice parameters for clinical use of the multiple sleep latency test and the maintenance of wakefulness test. Sleep 2005;28:113–21. [13] Johns MW. Sensitivity and specificity of the multiple sleep latency test (MSLT), the maintenance of wakefulness test and the epworth sleepiness scale: failure of the MSLT as a gold standard. J Sleep Res 2000;9:5–11. [14] Kothare SV, Kaleyias J. The clinical and laboratory assessment of the sleepy child. Semin Pediatr Neurol 2008;15:61–9.
723
[15] Samuels ER, Szabadi E. Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function Part I: principles of functional organisation. Curr Neuropharmacol 2008;6:235–53. [16] Kraemer S, Danker-Hopfe H, Dorn H, Schmidt A, Ehlert I, Herrmann WM. Timeof-day variations of indicators of attention: performance, physiologic parameters, and self-assessment of sleepiness. Biol Psychiatry 2000;48: 1069–80. [17] Danker-Hopfe H, Kraemer S, Dorn H, Schmidt A, Ehlert I, Herrmann WM. Timeof-day variations in different measures of sleepiness (MSLT, pupillography, and SSS) and their interrelations. Psychophysiology 2001;38:828–35. [18] McLaren JW, Hauri PJ, Lin SC, Harris CD. Pupillometry in clinically sleepy patients. Sleep Med 2002;3:347–52. [19] Yamamoto K, Kobayashi F, Hori R, Arita A, Sasanabe R, Shiomi T. Association between pupillometric sleepiness measures and sleep latency derived by MSLT in clinically sleepy patients. Environ Health Prev Med 2013;18:361–7. [20] Pohl A. Tagesschläfrigkeit – Vergleich verschiedener Testverfahren untereinander und mit Parametern des Nachtschlafs bei Patienten mit schlafbezogenen Atmungsstörungen. Marburg: Görich & Weiershäuser Verlag; 2003.
Contents lists available at ScienceDirect
Sleep Medicine journal homepage: www.elsevier.com/locate/sleep
Brief Communication
Pilot study on the validity of the pupillographic sleepiness test in children and adolescents Michael S. Urschitz a,b,⇑,1, Katrin Heine b,1, Lea Mendler b, Tobias Peters c, Barbara Wilhelm c, Christian F. Poets b a b c
Unit of Pediatric Epidemiology, Institute of Medical Biostatistics, Epidemiology, and Informatics, University Medical Center of the Johannes Gutenberg University Mainz, Germany Working Group on Pediatric Sleep Medicine, University Children’s Hospital Tuebingen, Germany Pupil Research Group, Centre for Ophthalmology, University Hospital Tuebingen, Germany
a r t i c l e
i n f o
Article history: Received 4 November 2013 Received in revised form 12 February 2014 Accepted 16 February 2014 Available online 13 April 2014 Keywords: Central nervous activation Alertness Hypersomnia Sleepiness Pupillography Multiple sleep latency test
a b s t r a c t Objective: To report preliminary validation data for the pupillographic sleepiness test (PST) in children and adolescents. Methods: Twelve patients (13.1 ± 4.4 years of age) underwent the multiple sleep latency test (MLST) and three PSTs at 09:00, 11:00, and 13:00 on one single day. Correlations were tested between mean sleep latency and gender-adjusted z-values of the natural logarithm of the pupillary unrest index (zlnPUI). Results: Spearman’s correlation (P-value) between the zlnPUI values obtained at 09:00 and 11:00 with the MSL was rS = 0.641 (0.025) and r = 0.553 (0.062). Conclusion: There was satisfactory agreement between PST and the MLST, which is similar to what is found in adults. The PST may be promising for the evaluation of daytime sleepiness in children and adolescents, and should be further evaluated in future studies. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Excessive daytime sleepiness (EDS) is a widespread problem in children and adolescents and can have major negative effects on performance, health, and safety [1]. Evaluations usually include some kind of structured sleep history, sleep logs, sleep questionnaires, and objective tests for the state domain of EDS (eg the multiple sleep latency test [MSLT] [2]. Lowenstein et al. were the first to observe typical fluctuations in pupil size in sleepy adults [3], and the first clinical application of pupillography was in narcolepsy [4]. The pupillographic sleepiness test (PST) is a standardized, accurate, and reliable physiological test for the level of EDS in adults [5] and has also been discussed for the laboratory assessment of EDS in children [6,7]. However, feasibility and accuracy of the PST are unknown in pediatric patients, hampering its widespread pediatric use. As part of an interdisciplinary project (TUPEDS: Tuebingen Project on EDS in Childhood), we had demonstrated the feasibility of ⇑ Corresponding author. Address: Unit of Pediatric Epidemiology, Institute of Medical Biostatistics, Epidemiology, and Informatics, University Medical Center of the Johannes Gutenberg University Mainz, Obere Zahlbacher Str. 69, 55131 Mainz, Germany. Tel.: +49 6131 17 3122; fax: +49 6131 17 2968. E-mail address: [email protected] (M.S. Urschitz). 1 M.S. Urschitz and K. Heine contributed equally to this manuscript. http://dx.doi.org/10.1016/j.sleep.2014.02.008 1389-9457/Ó 2014 Elsevier B.V. All rights reserved.
the PST in a field setting and reported preliminary reference values for the pupillary unrest index (PUI), the main sleepiness parameter of the PST [8]. In the current report from TUPEDS, we present preliminary validation data for the PST. 2. Methods 2.1. Study design A prospective diagnostic test pilot study was performed with the MSLT as reference standard and the PST as index test. Patients underwent a standard overnight polysomnography in a sleep laboratory, followed by multiple PST and MSLT assessments the next day. The study design was approved by the Ethics Committee of Tuebingen University Hospital; written informed parental and child’s consent were obtained. The pilot study was stopped after 12 patients, and an interim analysis was performed. 2.2. Subjects Eligible subjects were children referred to the sleep disorders center of the University Children’s Hospital Tuebingen between October 2011 and January 2013. Inclusion criteria were: (i) 6–18 years of age and (ii) history of EDS and/or referral for the
721
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723
evaluation of EDS. Exclusion criteria were: (i) developmental or cognitive disorder impairing patient’s compliance and (ii) restricted pupillary motility.
starting at 09:30, shortly after performance of the PST. Manual scoring was performed by one of the authors (M.S.U.), who was blinded to PST results. The sleep latency to the first epoch of sleep was calculated on each nap attempt and the mean sleep latency (MSL) computed across all naps, assigning 20 min for those periods in which no sleep was obtained. A shorter MSL corresponded to more sleepiness.
2.3. PST A detailed description of the PST is given elsewhere [5,8]. In short, the PST was performed according to standard operating procedures established in adults [9]. The procedures were not modified for application in children. Measures were obtained three times at 2 h intervals starting at 09:00. Recordings of spontaneous pupillary oscillations were acquired for 11 min by infra-red video pupillography in a fully darkened and quiet room [5]. During the test, subjects fixated a red light source at the lens aperture of the video camera in an upright position. To block ambient light fully, subjects wore goggles with black lenses and infrared filters. Pupillary oscillations were registered, quantified, and analyzed automatically by the system [10]. The entire 11 min recording was divided into eight segments each comprising 82 s recording time and 2048 data points. Before off-line analysis, missing data values were replaced by linear interpolation. The amount of interpolation was calculated as percentage of recording time. A sufficient recording quality was defined as interpolation <40% of recording time [8]. Based on the pupil diameter, the PUI (mm/min) was calculated, which corresponded to a low pass filtering of the data set [10]. In the present study, PUI trend data were printed out and visually inspected for artifacts by two of the authors (B.W., T.P.). Artifacts were handled and PUI manually recalculated according to a previously described protocol without knowledge of MSLT results [8]. Higher PUI values corresponded to more sleepiness. Immediately after each PST, subjective sleepiness was investigated using a German version of the Stanford Sleepiness Scale (SSS) [11]. A higher SSS value corresponded to more sleepiness.
2.5. Statistical analysis Descriptive statistics including mean, standard deviation (SD), and range were used to summarize demographic and clinical characteristics. To obtain a normal distribution, the natural logarithm of the PUI (lnPUI) was calculated. To account for gender differences in children, lnPUI values were transformed into lnPUI z-values (z-lnPUI) using the mean and SD of previously obtained pediatric reference values (i.e. 2.01 ± 0.43 for boys and 1.93 ± 0.43 for girls) and the formula: z-lnPUI = (lnPUI mean)/SD [8]. In accordance with the MSL, intra-individual means were calculated for lnPUI, z-lnPUI, and SSS. Time-of-day variations were assessed by scatter plots and fitting lines of simple linear regression. Agreement of the mean z-lnPUI and the three single z-lnPUI values obtained at 09:00, 11:00, and 13:00 with the MSL and the three single SSS values was evaluated using Spearman’s rank correlation coefficient, rS. Due to the exploratory character of the pilot study, no sample size calculation and no statistical hypothesis test was performed. Hence, P-values for rS were calculated for illustrative purposes rather than hypothesis testing. All analyses were done with statistical software (IBM SPSS Statistics 20). 3. Results Twelve children (five boys, seven girls) with a mean ± SD age (range) of 13.1 ± 4.4 (7–18) years were enrolled and 36 PST recordings successfully performed (Table 1). No recording was terminated prematurely. Interpolation ranged from 1.2% to 67.1% of recording time (mean ± SD, 14.1 ± 15.3). Only one recording had an interpolation of >40%; this child was extremely sleepy,
2.4. MSLT The MSLT was performed according to a current guideline [12]. It included three to five 20 min nap attempts at 2 h intervals Table 1 Clinical characteristics of study patients (n = 12). No.
Sex
Age (years)
Diagnosis
Mean sleep latency (min)
SOREMPs (no.)
Mean SSS
1
Female
11
Narcolepsy with cataplexy
1.1
>2
4.7
2
Female
11
5.1
1
4.0
3
Female
17
2.5
0
3.3
4
Female
7
Unclear excessive daytime sleepiness Suspected idiopathic hypersomnia Obstructive sleep apnea
20.0
0
2.3
5
Male
18
Insufficient sleep syndrome
6.5
1
3.3
6
Female
14
Obstructive sleep apnea
1.6
0
4.0
7
Male
12
8.4
0
2.7
8
Female
18
Suspected narcolepsy without cataplexy Chronic fatigue
16.8
0
3.0
9
Male
14
Insufficient sleep syndrome
2.8
2
3.3
10
Male
10
17.0
0
2.7
11
Female
13
Periodic limb movement disorder Narcolepsy with cataplexy
8.0
3
3.0
12
Male
10
Habitual snoring
20.0
0
2.0
lnPUI (z-lnPUI) 09:00
11:00
13:00
2.74 (+1.88) 3.00 (+2.49) 3.49 (+3.63) 2.17 (+0.57) 1.42 ( 1.37) 3.23 (+3.02) 2.50 (+1.14) 1.76 ( 0.40) 2.71 (+1.64) 2.47 (+1.08) 1.59 ( 0.80) 2.25 (+0.55)
2.63 (+1.63) 2.95 (+2.37) 2.38 (+1.05) 2.12 (+0.45) 1.20 ( 1.88) 2.52 (+1.37) 2.32 (+0.72) 1.89 ( 0.10) 2.85 (+1.95) 2.40 (+0.91) 1.36 ( 1.33) 2.36 (+0.81)
2.53 (+1.40) 2.97 (+2.41) 2.44 (+1.19) 2.25 (+0.75) 0.72 ( 2.99) 2.41 (+1.12) 1.74 ( 0.62) 1.96 (+0.06) 2.06 (+0.12) 2.76 (+1.75) 1.65 ( 0.65) 2.44 (+1.00)
SOREMPs, sleep-onset rapid eye movement periods; SSS, Stanford Sleepiness Scale; PUI, pupillary unrest index; ln, natural logarithm.
Mean lnPUI
Mean zlnPUI
2.63
+1.64
2.97
+2.42
2.77
+1.96
2.18
+0.59
1.11
2.08
2.72
+1.84
2.19
+0.41
1.87
0.15
2.54
+1.23
2.55
+1.25
1.53
0.92
2.35
+0.79
722
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723
Fig. 1. Distribution of sleepiness parameters stratified by time of day.
repeatedly closed his eyes, and had high lnPUI (3.23) and z-lnPUI values (+3.02). However, his results seemed valid and the respective recording remained in the analysis. Considering all 36 recordings, SSS, sleep latency, lnPUI, and z-lnPUI ranged from 1 to 6 points, from 0.5 to 20 min, from 0.72 to 3.49, and from 2.99 to +3.63, respectively. In 13, three, and two recordings, z-lnPUI was >+1, >+2, and >+3, respectively. Considering all 12 children, mean SSS, MSL, mean lnPUI, and mean z-lnPUI ranged from 2.0 to 4.7 points, from 1.1 to 20 min, from 1.11 to 2.97, and from 2.08 to +2.42. Six children had MSL < 8 min. These children had higher mean lnPUI (mean ± SD, 2.46 ± 0.67) and higher mean z-lnPUI values (+1.17 ± 1.64) compared to children with a mean sleep latency of P8 min (mean lnPUI, 2.11 ± 0.36; mean z-lnPUI, +0.33 ± 0.76). On a descriptive basis, all parameters indicated that daytime sleepiness was decreasing with increasing time of day (Fig. 1). This was evident for lnPUI (decreasing), SSS (decreasing), and sleep latency (increasing), thereby showing some agreement in the time-of-day variation. rS (P-value) between the three paired z-lnPUI SSS values obtained at 9:00, 11:00, and 13:00 was 0.656 (0.020) for 9:00, 0.205 (0.523) for 11:00, and 0.311 (0.325) for 13:00. rS (P-value) for the mean z-lnPUI and the three single z-lnPUI values obtained at 9:00, 11:00, and 13:00 with the MSL was 0.501 (0.097) for the mean z-lnPUI, 0.641 (0.025) for the z-lnPUI obtained at 09:00, 0.553 (0.062) for the z-lnPUI obtained at 11:00, and 0.242 (0.449) for the z-lnPUI obtained at 13:00.
In theory, the propensity to fall asleep and pupillary instability are linked via the level of tonic central nervous activation in the locus coeruleus [15]. Within this theoretical framework an association between sleep latency and PUI may be expected. The few correlation studies in healthy adults found moderate agreement between MSLT and PST, depending on the definition of sleep latency (N1 or N2 sleep) [16] and the pupillographic variable used (pupil diameter, coefficient of variation, power of the 0.1–0.8 Hz spectrum, or PUI) [17]. In the latter study, the median of the intra-individual rS was 0.46 for the agreement between sleep latency and PUI (interquartile range, 0.38 to 0.69 [17]). In sleepy adults, rS between sleep latency and PUI was 0.44 [18] and 0.40 [19]. Variation in agreement with time of day has been reported only in one study [20]; the weakest rS ( 0.04) was reached at 14:00, corresponding to the observation in the present study. We therefore conclude that the PST should be performed in the morning or to be predictive of results of the MSLT. In conclusion, the present study is largely in line with the adult literature showing moderate agreement between MSLT and PST with considerable variation across the time of day. Hence, accuracy and precision of the PST in children may be comparable with what is seen in adults. This may encourage the scientific community to investigate further the usability and prognostic ability of the PST in children. Funding sources None.
4. Conclusions In the present pilot study, there was satisfactory agreement between the MSLT and PST in children, depending on the PST variable and time of day. The best agreement was found for PST results obtained at 09:00 and 11:00. The intra-individual mean of three PST recordings may not have better agreement compared to one single PST recording performed at 09:00. These results suggest that higher PUI values are indeed linked to an increased propensity to fall asleep and therefore associated with a shorter sleep latency, particularly in the morning and forenoon. To our knowledge, this is the first validation study on the PST in children, which suggests that PST variables reflect sleepiness in children in the same manner as in adults. A complement to the MSLT is desirable because MSLT sleep latency shows a large overlap between healthy subjects and subjects with sleep disorders [12]. Some authors have even questioned its status as a gold standard for the evaluation of sleepiness [13]. In addition, it is costly and time-consuming, lacks sufficient normative data in pediatric patients, and suffers from motivational influences and the last nap effect [14]. Consequently, a feasible, convenient, and time-saving complementary method that is also less dependent on motivation would be a major improvement.
Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: http://dx.doi.org/10.1016/j.sleep.2014.02.008.
References [1] Millman RP. Excessive sleepiness in adolescents and young adults: causes, consequences, and treatment strategies. Pediatrics 2005;115:1774–86. [2] Carskadon MA, Dement WC, Mitler MM, Roth T, Westbrook PR, Keenan S. Guidelines for the multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep 1986;9:519–24. [3] Lowenstein O, Feinberg R, Loewenfeld IE. Pupillary movements during acute and chronic fatigue. Invest Ophthalmol 1963;2:138–57. [4] Yoss RE, Moyer NJ, Ogle KN. The pupillogram and narcolepsy. A method to measure decreased levels of wakefulness. Neurology 1969;19:921–8. [5] Wilhelm B, Wilhelm H, Lüdtke H, Streicher P, Adler M. Pupillographic assessment of sleepiness in sleep-deprived healthy subjects. Sleep 1998;21:258–65. [6] Kotagal S, Goulding PM. The laboratory assessment of daytime sleepiness in childhood. J Clin Neurophysiol 1996;13:208–18. [7] Urschitz MS. Assessing objective daytime sleepiness in children and adults: do we have appropriate instruments? Sleep Med 2013;14:812–3.
M.S. Urschitz et al. / Sleep Medicine 15 (2014) 720–723 [8] Urschitz MS, Heine K, Brockmann PE, Peters T, Durst W, Poets CF, et al. Subjective and objective daytime sleepiness in school children and adolescents. Results of a community-based study. Sleep Med 2013;14:1005–12. [9] Wilhelm B, Koerner A, Heldmaier K, Moll K, Wilhelm H, Luedtke H. Normal values of the pupillographic sleepiness test in male and female subjects aged 20 to 60 years. Somnologie 2001;5:115–20. [10] Lüdtke H, Wilhelm B, Adler M, Schaeffel F, Wilhelm H. Mathematical procedures in data recording and processing of pupillary fatigue waves. Vision Res 1998;38:2889–96. [11] Hoddes E, Dement WC, Zarcone V. The history and use of the Stanford Sleepiness Scale. Psychophysiology 1971;9:150. [12] Littner MR, Kushida C, Wise M, Davila DG, Morgenthaler T, Lee-Chiong T, et al. Practice parameters for clinical use of the multiple sleep latency test and the maintenance of wakefulness test. Sleep 2005;28:113–21. [13] Johns MW. Sensitivity and specificity of the multiple sleep latency test (MSLT), the maintenance of wakefulness test and the epworth sleepiness scale: failure of the MSLT as a gold standard. J Sleep Res 2000;9:5–11. [14] Kothare SV, Kaleyias J. The clinical and laboratory assessment of the sleepy child. Semin Pediatr Neurol 2008;15:61–9.
723
[15] Samuels ER, Szabadi E. Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function Part I: principles of functional organisation. Curr Neuropharmacol 2008;6:235–53. [16] Kraemer S, Danker-Hopfe H, Dorn H, Schmidt A, Ehlert I, Herrmann WM. Timeof-day variations of indicators of attention: performance, physiologic parameters, and self-assessment of sleepiness. Biol Psychiatry 2000;48: 1069–80. [17] Danker-Hopfe H, Kraemer S, Dorn H, Schmidt A, Ehlert I, Herrmann WM. Timeof-day variations in different measures of sleepiness (MSLT, pupillography, and SSS) and their interrelations. Psychophysiology 2001;38:828–35. [18] McLaren JW, Hauri PJ, Lin SC, Harris CD. Pupillometry in clinically sleepy patients. Sleep Med 2002;3:347–52. [19] Yamamoto K, Kobayashi F, Hori R, Arita A, Sasanabe R, Shiomi T. Association between pupillometric sleepiness measures and sleep latency derived by MSLT in clinically sleepy patients. Environ Health Prev Med 2013;18:361–7. [20] Pohl A. Tagesschläfrigkeit – Vergleich verschiedener Testverfahren untereinander und mit Parametern des Nachtschlafs bei Patienten mit schlafbezogenen Atmungsstörungen. Marburg: Görich & Weiershäuser Verlag; 2003.