J. Res. (2011) Regular Research Paper doi: 10.1111/j.1365-2869.2011.00923.x and concordance in juvenile idiopathic arthritis, asthma and healthy children

TERESA M. WARD1 , MARTHA LENTZ2 , GAIL M. KIECKHEFER1 and CAROL A. LANDIS2 1Family and Child Nursing, School of Nursing, University of Washington, Seattle, WA, USA and 2Biobehavioral Nursing and Health Systems, School of Nursing, University of Washington, Seattle, WA, USA

Accepted in revised form 7 April 2011; received 18 March 2011

SUMMARY The aims of this study were to evaluate sensitivity, specificity and accuracy with an epoch-by-epoch comparison of polysomnography (PSG) and actigraphy with activity counts scored at low, medium and high thresholds, and to compare PSG-derived total sleep time (TST), sleep efficiency (SE) and wake after (WASO) to the same variables derived from actigraphy at low, medium and high thresholds in 9- to 11- year-old children with juvenile idiopathic arthritis (JIA), asthma and healthy control children. One night of PSG and actigraphy were recorded. Pairwise group comparisons for sensitivity showed significant differences at the low [TukeyÕs honest significant difference (HSD) P < 0.002], medium (P < 0.001) and high thresholds (P < 0.001) between JIA and asthma groups, and at the high threshold between JIA and controls (P < 0.009). Significant differences were found for specificity at the low (P < 0.001), medium (P < 0.001) and high thresholds (P < 0.001) between JIA and asthma groups, and between JIA and controls (low, P < 0.002: medium, P < 0.002: high, P < 0.008 threshold). PSG TST, WASO and SE were not significantly different among the groups, but significant group differences were found for actigraphy TST, WASO and SE at all three thresholds. Actigraphy showed the least overestimation or underestimation of sleep or wakefulness at the medium threshold for TST and WASO for all three groups. Compared to PSG, actigraphy was most accurate in the identification of sleep from wakefulness in 9- to 11-year-old healthy children, and less accurate in children with JIA and asthma. keywords actigraphy, concordance, polysomnography, school-age children

represent a subjectÕs customary sleep at home. Nevertheless, INTRODUCTION PSG is the Ôgold standardÕ in sleep research and in the clinical Polysomnography (PSG) is the most accurate method to evaluation of suspected sleep disorders. measure sleep stages and wakefulness. It requires that partic- In contrast to PSG, actigraphs, worn at the wrist or ankle, ipants either come to a sleep laboratory or be connected to are relatively unobtrusive, inexpensive and can record subjectsÕ portable PSG equipment at home, creating a considerable sleep and wake patterns for multiple days and nights in the burden to participants and increasing study costs. Although home environment. Wrist actigraphy has become a widely used the most accurate sleep measure, PSG data are usually technology for estimating sleep–wake patterns in community- obtained for 1 or 2 consecutive nights in a laboratory setting based research (Berger et al., 2008; Littner et al., 2003; and thus provides only a limited view of sleep, which may not Morgenthaler et al., 2007; Sadeh and Acebo, 2002). The actigraph is a battery-operated device worn on the non- Correspondence: Professor Carol A. Landis DNSc, RN, FAAN, Box 357266, Biobehavioral Nursing and Health Systems, University of dominant wrist and programmed for continuous limb move- Washington, Seattle, WA, 98195-7266, USA. Tel.: 206 616 1908; ment data collection 24 h a day for consecutive days or weeks fax: 206 543 4771; e-mail: [email protected] at a time. Several types of actigraphs are commercially Published 2011 This article is a US Government work and is in the public domain in the USA 1 2 T. M. Ward et al. available and differ in the type of motion sensor used for the difficulty breathing in asthma. Varying disease-related symp- detection and recording of limb movements. Most modern toms could influence movement patterns of the children actigraphs use accelerometers and have an event marker, throughout the night. PSG studies in JIA have shown sleep which a participant presses when the actigraph is removed or fragmentation (e.g. sleep stage shifts, more and replaced, and at and waketime. Limb movement awakenings, increased limb movements), , mild sleep- (activity) counts can be recorded and stored in the actigraph in disordered breathing and daytime sleepiness (Lopes et al., varying (e.g. 15 s, 30 s or 1 min) or fixed epoch lengths. 2008; Passarelli et al., 2006; Ward et al., 2008, 2010; Zamir Activity counts are downloaded from the actigraph to a et al., 1998) but less is known about sleep as measured by computer for processing and scoring into epochs of sleep and actigraphy in JIA. Compared to healthy school-aged children, wake using algorithms that are unique to the particular PSG studies in asthma have shown decreased total sleep time, software associated with the device. The software and scoring more arousals and awakenings (Ramagopal et al., 2008; Sulit algorithms for actigraphs vary in how the activity counts are et al., 2005), and actigraphy studies have shown decreased summarized for each epoch (Jean-Louis et al., 2001; Pollak total night-time sleep, more daytime napping and higher et al., 1998; Sadeh and Acebo, 2002; Sadeh et al., 1995). Some activity counts during sleep (Kieckhefer et al., 2008; Sadeh software programs provide operator-determined activity count et al., 1995). Without studies of concordance between PSG and thresholds to specify for the discrimination of epochs of sleep actigraphy in JIA or asthma it is unclear, for example, which and wake (Cole et al., 1992; Jean-Louis et al., 1999; Kushida threshold for scoring actigraphy would yield the most sensitive, et al., 2001). The differences in the devices and algorithms can specific and accurate measure of sleep and of nocturnal pose significant challenges when comparing actigraph-derived wakefulness. Further, given the known developmental changes sleep and wake parameters across different study populations. in sleep as children mature, concerns about the ability of The validity of actigraphy to discriminate epochs of sleep actigraphy to accurately evaluate sleep quality in school-aged and wake compared to PSG has been studied most extensively children (Spruyt et al., 2011), increasing use of actigraphy in in adults (Kripke et al., 2010; Littner et al., 2003; Morgent- community-based clinical studies and a lack of reports in the haler et al., 2007; Van de Water et al., 2011). Actigraphy literature comparing actigraphy to PSG in children with agrees best with PSG in its detection of sleep duration in chronic illness, there is a need for further validation studies. healthy subjects, but is less accurate in recognition of nocturnal In 9- to 11-year-old children with JIA, asthma and healthy wakefulness, especially in adults with various sleep disorders control children, the aims of this study were: (1) to evaluate (Morgenthaler et al., 2007; Paquet et al., 2007; Sadeh and sensitivity, specificity and accuracy with an epoch-by-epoch Acebo, 2002), in infants (Insana et al., 2010) and in school- comparison of (PSG and actigraphy with activity counts aged children (Spruyt et al., 2011). Actigraphy is best used to scored at low, medium and high thresholds and (2) to compare derive measures of sleep and wake patterns (e.g. total sleep and PSG-derived total sleep time (TST), sleep efficiency (SE) and wake time, number of awakenings) over multiple sleep cycles wake after sleep onset (WASO) to the same variables derived (Morgenthaler et al., 2007), but not for estimates of time to fall from actigraphy at low, medium and high thresholds. asleep (sleep latency) or final awakening (Lichstein et al., 2006). Validity is increased when actigraphic data are supple- mented with sleep diaries that can help the analyst control for METHODS artifacts in data collection, e.g. removal of the device (Kushida Participants et al., 2001), or when subjects are contacted and reminded to press the event marker to denote wake and sleep times or times This study is a secondary analysis of data obtained from two of actigraph removal (Tsai et al., 2008). research projects that examined sleep quality in children with The American Academy of Sleep has issued JIA (Ward et al., 2008, 2010) and in children with and without practice parameters on the use of actigraphy, which includes asthma (Kieckhefer et al., 2008, 2009). Human subjects use of them in children with sleep disorders and other pediatric approval for the studies was obtained from the Institutional populations, including children with chronic illnesses (Littner Review Boards at Seattle ChildrenÕs Hospital (JIA study) and et al., 2003; Morgenthaler et al., 2007). Agreement between the University of Washington (asthma study). From 2002 to actigraphy and PSG has been reported in a study of infants 2007, children 9–11 years of age with JIA (n = 34), with (Insana et al., 2010), healthy school-aged children <9 years asthma (n = 19) and controls (n = 18) were studied. Children old (Spruyt et al., 2011) and in children with sleep-disordered in either study were excluded if they had a diagnosis of a breathing (Hyde et al., 2007), but to our knowledge studies psychiatric condition, diabetes, cancer or cystic fibrosis or were have not reported similar comparisons in healthy children diagnosed with an upper respiratory infection or acute 9–11 years old, or in children with chronic illnesses such as exacerbation of asthma within the past 2 weeks. arthritis or asthma. Chronic illnesses vary in underlying disease mechanisms and manifestation of symptoms that may have General procedures differential effects on sleep and wake patterns. For example, joint inflammation and pain in JIA represent different path- For both studies, each child and a parent slept in the ophysiological processes compared to bronchoconstriction and University of Washington School of Nursing sleep research Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. Polysomnography and actigraphy concordance 3 laboratory. They were scheduled to arrive at the laboratory epochs (JIA) or 30-s epochs (asthma and controls), inspected approximately 2–3 h prior to the childÕs usual bedtime. visually and screened of artifacts prior to running the scoring A schedule for bedtime and rise time was established based program. For the JIA data, prior to applying the scoring on a childÕs usual school-night schedule, except during summer algorithm the raw activity counts in 15-s epochs were months, when most children followed a similar schedule every combined to create 30-s epochs. night of the week. PSG and actigraphy were recorded Activity counts were scored using the Actiware software simultaneously for each child using standard laboratory version 3.4 (Mini-Mitter Philips Respironics, Inc.). The Acti- protocols. In the JIA study average bedtime was 21:25 hours ware software scoring algorithm to discriminate activity counts and rise time was 07:13 hours, in the asthma study average into epochs of sleep and wake computes a weighted sum of bedtime was 21:15 hours and rise time was 07:30 hours. activity in the current epoch, preceding four epochs and the following four epochs (Mini-Mitter Philips Respironics, Inc.). This software has three different sensitivity thresholds. If the Polysomnography total activity count is above the defined threshold the epoch is Both studies used the same laboratory PSG protocols. Details scored as wake, and if the total activity count is equal to or for the PSG recordings have been published previously (Ward below the sensitivity threshold the epoch is scored as sleep. et al., 2008, 2010). In brief, electrodes for electro-oculogram, The threshold sensitivity values include: (1) low, defined as 80 electrocardiogram, electromyelogram and leg movements were activity counts per epoch; (2) medium, defined as 40 activity placed according to laboratory protocols based on published counts per epoch; and (3) high, defined as 20 activity counts standards (Rechtschaffen and Kales, 1968). Electroencephalo- per epoch (Table 1). gram electrodes were positioned at two frontal (F7, F8), two The actigraphy sleep interval was set manually for each central (C3, C4) and two occipital (O1, O2 locations (interna- record using PSG lights out and on for sleep onset and offset, tional 10–20 system of measurement) and linked to reference respectively. Sleep onset was defined as the first 10-min period electrodes placed over each mastoid bone (A1, A2). Nasal in which no more than one epoch was scored as mobile, and airflow was monitored with a pressure cannula placed in the sleep offset was defined as the last 10-min period in which no nose (Pro-Tech Services, Inc., Mukilteo, WA, USA) and more than one epoch was scored as mobile. Actigraphy respiratory effort was measured by piezo respiratory effort parameters of interest included (1) total sleep time, defined as bands placed around the chest and abdomen (Pro-Tech the number of minutes scored as sleep between sleep onset and Services Inc.). Oxygen saturation was measured from the left sleep offset during the sleep interval (similar to PSG sleep or right index finger with a pulse oximeter (Nonin XPod, Nonin period time); (2) sleep efficiency, defined as the ratio of total Med, Plymouth, MN, USA). Electrophysiological signals were sleep time ⁄ time in ; and (3) wake after sleep onset, defined recorded and digitized using the Somnologica data acquisition as the percentage of scored total wake time during the sleep recording system (A10 recorders; Embla, Broomfield, CO, interval multiplied by 100. We calculated the three actigraphy USA) and displayed and stored on a desktop (Dell Pentium III) sleep parameters using each of the threshold sensitivities. computer. All digitized data were acquired and stored unfil- tered and displayed continuously in 30-s intervals during each PSG and actigraphy data processing for analysis recording. PSG data from each laboratory night were scored (Somnological software version 3.1.2) manually into 30-s PSG and actigraphy data were synchronized to National epochs of wake and sleep stages by an experienced technologist Institute of Standards and Technology (NIST) using the same according to standard criteria (Rechtschaffen and Kales, 1968). computerized system for each recording session so that each 30-s PSG parameters of interest included: (1) total sleep time epoch of PSG could be matched epoch-by-epoch with actigra- defined as the amount of time in non-rapid eye movement phy for comparison. Each scored PSG record was exported (NREM) Stages 1–4 and REM; (2) sleep efficiency expressed as from Somnologica as a text file and read into spss (version 15.0; a ratio of total sleep time ⁄ time in bed (e.g. from lights off to SPSS Inc, Chicago, IL, USA) and wake and sleep stage data lights on); and (3) wake after sleep onset expressed as a percentage of wake during sleep period time (e.g. time from lights out until final awakening). Table 1 Measures for actigraphic analysis

Polysomnography Actigraphy Actigraphy Sleep Wake Total Concurrent with PSG, each child wore an actigraph (Acti- watch 64; Mini-Mitter Philips Respironics, Bend, OR, USA) Sleep True sleep (TS) False wake (FW) TS + FW Wake False sleep (FS) True wake (TW) FS + TW placed on the non-dominant wrist during each laboratory Total TS + FS FW + TW TS + FS + night. Actiwatch 64 has an accelerometer that senses the TW + FW occurrence and degree of motion in all directions. This motion is converted into an electric signal and digitally integrated to Sensitivity: TS ⁄ (TS + FS); specificity: TW ⁄ (FW + TW); accuracy: (TS + TW) ⁄ (TS + FS + TW + FW) (Kushida et al., 2001). derive an activity count. Actigraphy data were recorded in 15-s Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. 4 T. M. Ward et al. were collapsed into binary variables (e.g. wake = 0; REM and Finally, the Bland–Altman concordance technique was used sleep Stages 1, 2, 3, 4 = 1). Each scored actigraphy record was to determine if a meaningful agreement could be found exported as a text file from Actiware and read subsequently into between concurrent PSG and each of actigraphy threshold an Excel (Microsoft Office, version 2007) worksheet and also for TST and WASO. Bland–Altman plots represent graphi- read into spss as binary sleep or wake variables. Agreement and cally the difference between two measurement techniques (on disagreement between actigraphy and PSG was assessed for y-axes) against the average of the two techniques (on the each matched 30-s epoch. When the actigraph data agreed with x-axis). Mean difference represents the bias between the two PSG data, the epoch was coded as true sleep (TS) or true wake measures. Standard deviation of the bias (SD) provides an (TW). When the actigraph data did not agree with PSG data it estimate of the scale of the variation of the difference between was scored as false sleep (FS) or false wake (FW). the two measures. For each of the sleep parameter compar- Two sets of analytical methods were used to calculate PSG isons, the average of actigraphy at a specific threshold and and actigraphy agreement using spss. An epoch-by-epoch PSG was plotted on the x-axis, and the difference between agreement analysis provided estimates of sensitivity, specificity actigraphy at a specific threshold and PSG was plotted on the and accuracy parameters. Sensitivity was defined as the y-axis for each group. A positive bias thus indicates an proportion of all epochs scored as sleep by PSG that were overestimation of the variable with actigraphy; a negative bias also scored as sleep by actigraphy (true sleep). Specificity was indicates an underestimation of the variable with actigraphy. defined as the proportion of all epochs scored as wake by PSG that were also scored as wake by actigraphy (true wake). RESULTS Accuracy was the proportion of all PSG scored epochs correctly identified by actigraphy. False-positive fraction Sample characteristics (FPF) was defined as the fraction or probability of actigraphy No group differences were found for age, but sex differences reporting sleep when PSG detects wakefulness (Pepe, 2003). were found (v2 = 18.7, P < 0.001) (see Table 2). As is The second set of analytical methods involved group compar- characteristic of JIA, there were more girls in the JIA group isons of sleep parameters (e.g. TST, SE and WASO) estimated (85%) compared to the asthma (31.6%) and control (38.9%) with PSG and with actigraphy. groups. Of the 34 children with JIA, 27% had pauciarticular disease and 71% had polyarticular disease. Children diagnosed Statistical analyses with asthma were well or partially controlled based on frequency of symptoms and pulmonary function testing Data were analyzed using spss for Windows version 15.0. The (http://www.nhlbi.nih.gov/guidelines/asthma/asthgdln.pdf). first set of analysis was conducted to address group differences in clinical characteristics (e.g. age and sex) using a separate chi-square test for each variable. Statistical significance was set Epoch-by-epoch agreement at P < 0.05 (two-sided). To address aim 1, Table 3 shows the sensitivity, specificity and For the first aim, we examined sensitivity, specificity and accuracy values (mean ± SD) for the epoch-by-epoch agree- accuracy with an epoch-by-epoch comparison between PSG ment between PSG and actigraphy at the low, medium and and actigraphy scored at low, medium, high thresholds among high thresholds for each group. the three groups. Contingency tables were used to count the number of occurrences of true sleep, true wake, false sleep and false wake. Estimates of sensitivity, specificity and accuracy Sensitivity were derived from these occurrences. General linear model Significant main effects were found for threshold (F = 226, (GLM) analyses (i.e. repeated ⁄ related measures analyses) for (2,68) P < 0.001) and group (F = 10.1, P < 0.001), and the the epoch-by-epoch agreement measures (sensitivity, specificity (2,68) group · threshold interaction was also significant (F = and accuracy) were used to evaluate the main effects and (4,136) interactions with threshold as within-subject factor and group Table 2 Demographic and clinical characteristics the between-subject factor. To aid interpretation, GLM analyses (repeated ⁄ related measures) were followed by a series JIA Asthma Controls of univariate one-way analyses of variance (anovas), with (n=34) (n=19) (n=18) pairwise group comparisons executed using TukeyÕs honest Age, years 10.2 ± 0.82 9.8 ± 0.86 9.8 ± 0.71 significant difference (HSD). Sex, n (%)* For the second aim, we examined whether PSG-derived TST, Girls 29 (85.3) 6 (31.6) 7 (38.9) SE and WASO differed from the same variables derived by Boys 5 (14.7) 13 (68.4) 11 (61.1) Ethnicity, n (%) actigraphy at the low, medium and high thresholds among the White 30 (88.2) 10 (52.6) 9 (50) three groups. A series of univariate one-way anovas, with Other 4 (11.8) 9 (47.4) 9 (50) pairwise group comparisons executed using TukeyÕs HSD, were used to evaluate the differences in TST, SE and WASO between Data are mean ± standard deviation or n (%). *v2 = 18.7, P < 0.001. JIA, juvenile idiopathic arthritis. PSG and actigraphy at the low, medium and high thresholds. Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. Polysomnography and actigraphy concordance 5

(a) Low threshold Table 3 Sensitivity, specificity, accuracy of epoch-by-epoch comparison polysomnography and actigraphy

JIA Asthma Controls (n=34) (n=19) (n=18)

Sensitivity

Low threshold* 0.98 ± 0.01 0.97 ± 0.02 0.98 ± 0.01 Sensitivity Medium threshold 0.96 ± 0.02 0.94 ± 0.03 0.95 ± 0.02 High thresholdà 0.93 ± 0.03 0.88 ± 0.04 0.90 ± 0.04 Specificity Low threshold§ 0.38 ± 0.20 0.60 ± 0.16 0.55 ± 0.15 – Medium threshold 0.51 ± 0.21 0.73 ± 0.13 0.69 ± 0.15 FPF (1-specificity) High threshold** 0.62 ± 0.18 0.80 ± 0.11 0.77 ± 0.15 (b) Medium threshold Accuracy Low threshold 0.88 ± 0.07 0.90 ± 0.04 0.90 ± 0.04 Medium threshold 0.89 ± 0.06 0.89 ± 0.04 0.90 ± 0.03 High threshold 0.88 ± 0.05 0.87 ± 0.04 0.87 ± 0.02

Data are mean ± standard deviation. All P = pairwise compari- son, TukeyÕs honest significant difference. Sensitivity *P < 0.002, juvenile idiopathic arthritis (JIA) versus asthma. P < 0.001, JIA versus asthma. àBoth P < 0.001, JIA versus asthma; JIA versus control. §Both P < 0.002, JIA versus asthma; JIA versus control. – Both P < 0.002, JIA versus asthma; JIA versus control. FPF (1-specificity) **Both P < 0.008, JIA versus asthma; JIA versus control. (c) High threshold

5.8, P < 0.001). Pairwise group comparisons showed signifi- cant differences at the low (TukeyÕs HSD P < 0.002), medium (P < 0.001) and high thresholds (P < 0.001) between JIA and asthma groups. Significant group differences also were found at

the high threshold between and control groups (P < 0.009). Sensitivity

Specificity

Significant main effects were found for threshold (F = (2,68) FPF (1-specificity) 626, P < 0.001) and group (F(2,68) = 11.5, P < 0.001) and the group · threshold interaction also was significant Figure 1. Actigraphy sensitivity versus 1-specificity. Data represent low (a), medium (b) and high (c) actigraphy thresholds. JIA = d, (F(4,136) = 4.3, P < 0.002). Pairwise group comparisons asthma (AS) = h, + = controls. False-positive fraction (FPF) is showed significant differences for specificity at the low defined as the fraction or probability of actigraphy reporting sleep (P < 0.001), medium (P < 0.001) and high thresholds (P < when polysomnography detects wake. The figures show the discrimi- 0.001) between JIA and asthma groups and at the low nation for each subject with respect to threshold. Each point represents (P < 0.002), medium (P < 0.002) and high (P < 0.008) the decision criterion (true positive versus false positive) for each child thresholds between JIA and control groups. at a given threshold.

medium and high thresholds for the three groups. PSG-derived Accuracy TST (F(2,68) = 0.43, P = 0.65), SE (F(2,68) = 1.1, P = 0.36),

Table 3 shows accuracy data for the three groups, which did WASO (F(2,68) = 2.1, P = 0.14) did not differ among the not differ by group at any threshold. Fig. 1(a–c) shows 1- groups. However, significant group differences were found for specificity [false-positive fraction (FPF)] versus sensitivity by actigraphy TST at medium (F(2,68) = 4.0, P < 0.02) and high group, at the low, medium and high thresholds, respectively. (F(2,68) = 5.3, P < 0.008) thresholds; actigraphy SE at the low

The FPF was higher in the JIA group at all thresholds (F(2,68) = 8.2, P < 0.001), medium (F(2,68) = 9.7, P < 0.001) compared to the asthma and control groups, which suggests and high (F(2,68) = 10.7, P < 0.001) thresholds; and actigra- that actigraphy identified sleep when PSG did not. phy WASO at the low (F(2,68) = 13.9, P < 0.001), medium

(F(2,68) = 14.5, P < 0.001) and high (F(2,68) = 14.6, P < 0.001) thresholds. Sleep parameters concordance Bland–Altman plots for TST and WASO between PSG and To address aim 2, Table 4 shows PSG-derived and actigraphy- actigraphy for the low, medium and high thresholds are shown derived TST, SE and WASO with the latter scored at low, in Fig. 2 for the three groups. Compared to the low and high Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. 6 T. M. Ward et al.

Our results highlight the importance of evaluating actigraphy- Table 4 Sleep parameters between polysomnography and acti- graphy determined sleep parameters at each threshold when using a device and software that enables this level of discrimination, as JIA Asthma Controls sensitivity, specificity and accuracy may differ with children (n=34) (n=19) (n=18) with various chronic illnesses. As noted by others, actigraphy Total sleep time, min is not as accurate a measure of sleep parameters compared to Polysomnography 485 ± 48.9 479 ± 66.9 468 ± 75.1 PSG for certain clinical populations (Hyde et al., 2007; Insana Actigraphy et al., 2010; Owens et al., 2009; Paquet et al., 2007; Sitnick Low threshold 512 ± 39.0 481 ± 66.3 482 ± 68.7 Medium threshold* 496 ± 37.9 457 ± 66.2 462 ± 68.9 et al., 2008; Spruyt et al., 2011). Our findings of low specificity High threshold 474 ± 37.0 428 ± 65.0 435 ± 72.3 in the identification of nocturnal wakefulness across all three Sleep efficiency, % groups is consistent with those of others (Hyde et al., 2007; Polysomnography 84 ± 8.9 81 ± 10.1 82 ± 8.4 Paquet et al., 2007; Sitnick et al., 2008). These observations Actigraphy could be attributed to the probability of more sleep epochs Low thresholdà 89 ± 5.2 81 ± 9.3 84 ± 7.5 Medium threshold§ 86 ± 5.7 77 ± 9.9 80 ± 8.1 compared to wake epochs identified during the night-time High threshold– 82 ± 6.3 72 ± 9.9 75 ± 8.9 hours. However, in comparison to the other groups the false- Wake after sleep onset, % positive fraction was higher in the JIA group, which suggests Polysomnography 11.3 ± 8.6 16.2 ± 10.3 14.0 ± 6.0 that actigraphy may be less likely to identify wakefulness in Actigraphy JIA. This finding might be explained by sleep fragmentation Low threshold** 5.0 ± 3.2 12.1 ± 8.1 9.8 ± 3.1 Medium threshold 8.0 ± 4.3 16.5 ± 9.0 13.7 ± 4.3 and brief episodes of wake associated with mild sleep- High thresholdàà 12.1 ± 5.5 21.8 ± 9.2 18.8 ± 5.6 disordered breathing that we have identified in JIA (Ward et al., 2010) and as noted in children with obstructive sleep Data are mean ± standard deviation. All P = pairwise compari- apnea (Hyde et al., 2007). son, TukeyÕs honest significant difference. *P = 0.05 juvenile idiopathic arthritis (JIA) versus asthma.

Both P < 0.04 JIA versus asthma, JIA versus control. Epoch-by-epoch agreement àP < 0.001 JIA versus asthma. § P < 0.001 JIA versus asthma; P < 0.04,JIA versus control. Concordance between actigraphy and PSG for sensitivity to – P < 0.001 JIA versus asthma; P < 0.02 JIA control. measure sleep was high for the three groups. This finding is not **P < 0.001 JIA asthma; P = 0.007JIA versus control. P < 0.001 JIA versus asthma; P = 0.005JIA control. surprising, as the probability of actigraphy identifying sleep ààP < 0.001 JIA versus asthma; P = 0.003 JIA versus control. epochs compared to wake epochs are high during the night and improve with longer sleep duration. Unlike sensitivity, specificity was lower in JIA children. thresholds, actigraphy showed the least amount of overesti- As mentioned above, this finding may be related to mild sleep- mation or underestimation in the identification of TST and disordered breathing that we recently reported in JIA (Ward WASO at the medium threshold for all three groups. Mean et al., 2010). When children with arthritis are awake they may difference for TST at the medium threshold was 11 min for move less during sleep, particularly if they have painful joints JIA, )20 min for asthma and )8 min for controls. Mean leading actigraphy identification of sleep when the child is difference for WASO at the medium threshold was )3 min for actually awake. Alternatively, as there were more girls in the JIA, 0.61 min for asthma and )0.42 min for controls. JIA group compared to asthma and control groups, group In contrast, TST at the high threshold showed the least differences in actigraphy recognition of wakefulness might be underestimation in the JIA group ()10.9 min) but the most related to girls but not boys being more likely to lie awake underestimation in the asthma group ()50.1 min). At the high without moving, as has been noted in adult women with threshold, WASO was only slightly overestimated (0.75 min) (Lichstein et al., 2006). Unfortunately, we did not in the JIA group, but it was overestimated (5.8 min) in the have a sufficient sample to evaluate gender differences within asthma group. the groups. JIA and asthma are chronic conditions that probably impact sleep quality differently, and more research is necessary to replicate our findings and explore reasons behind DISCUSSION group differences. Previous studies in infants and children have The findings from this study are the first report of PSG and also shown that actigraphy is less accurate in identifying wake actigraphy concordance in 9- to11-year-old children with JIA, episodes in certain clinical populations (Insana et al., 2010; asthma and healthy controls. Overall, findings showed that Owens et al., 2009; Sitnick et al., 2008; Spruyt et al., 2011). actigraphy had the least overestimation or underestimation of sleep or wakefulness at the medium threshold for TST and Sleep parameters concordance WASO for all three groups. However, compared to PSG, actigraphy was accurate in the identification of sleep and The Bland–Altman plots highlight the close concordance wakefulness in 9- to 11-year-old healthy children at all three between PSG and actigraphy for both TST and WASO at thresholds, and less accurate in children with JIA and asthma. the medium threshold in healthy children. This finding is Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. Polysomnography and actigraphy concordance 7

(a)

(b)

Figure 2. Bland–Altman plots between polysomnography and actigraphy for total sleep time (TST) (a) and wake after sleep onset (WASO) (b). Data represent the mean bias ± standard deviation which provides an estimate of the degree of variation of the difference between poly- somnography and actigraphy measures of the TST and WASO. Juvenile idiopathic arthritis (JIA) = d, asthma (AS) = h, + = controls. inconsistent with a recent study that found poor concordance the JIA and healthy children, the low threshold for TST had between PSG and actigraphy for both TST and WASO at the the least amount of bias (3.6 min) in the asthma group. Future medium threshold in 4- to 9-year-old healthy children who studies examining PSG and actigraphy concordance in chil- wore the same Actiwatch device (Actiwatch 64; Mini-Mitter dren with chronic conditions may want to include videosom- Philips Respironics) (Spruyt et al., 2011). The difference in nography to understand the nature of the movements recorded concordance may be attributed to the inclusion of younger by the actigraph. Our findings highlight that changes in sleep children and ⁄ or the use of a 1-min epoch sampling in the are probably not uniform across pediatric chronic conditions, Spruyt et al. study. The manufacturer recommends use of the and that recommendations regarding thresholds for scoring medium threshold. However, further studies are needed on actigraphy from healthy children are not adequate for assum- PSG and actigraphy concordance at low, medium and high ing the same quality of validity of actigraphy data in children thresholds to determine which threshold for scoring actigraphy with chronic conditions. would yield the most sensitive, specific and accurate measure of sleep and of nocturnal wakefulness in healthy children of Limitations different ages and to assess the impact of different recorded epoch lengths on these measures. There are several limitations worth noting in this study. First, In JIA and asthma, the Bland–Altman analysis revealed the small number of asthma (n = 19) and control (n = 18) striking differences in the underestimation of sleep scored at children compared to the JIA (n = 34) children restricts the the high threshold. The asthma group showed nearly five times generalizability of the findings and the types of data analysis the amount of bias compared to the JIA group. This finding that could be conducted. Secondly, only one type of actigraph may be explained by more night-time movement, differing device was studied. Future studies could compare various disease mechanisms and ⁄ or disease-related symptoms (fre- actiwatch devices. Additional comparison studies are needed quent night cough) in asthma children. The difference in TST between PSG and actigraphy concordance in different pediat- and WASO for the JIA children may also reflect particular ric chronic conditions compared to healthy control children. distinct mechanisms, low disease activity or a combination of In conclusion, there is a paucity of data on PSG and disease-related symptoms (pain, fatigue) and medication use. actigraphy concordance in children with chronic conditions. Unlike the JIA children, the asthma children had an under- The extent of sleep disturbances as measured by actigraphy in estimation of TST at the medium threshold and an underes- pediatric chronic conditions has not been well studied. Given timation of WASO at the low and medium thresholds. Unlike the unobtrusiveness and cost-effectiveness of actigraphy in Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. 8 T. M. Ward et al. pediatric chronic illness, it is important for clinicians and Kieckhefer, G. M., Ward, T. M., Tsai, S. Y. and Lentz, M. J. researchers to understand the differences in scoring threshold Nighttime sleep and daytime patterns in school age children algorithms and their strengths and limitations. Although with and without asthma. J. Dev. Behav. Pediatr., 2008, 29: 338–344. Kieckhefer, G. M., Lentz, M. J., Tsai, S. Y. and Ward, T. M. Parent– actigraphy may not be as sensitive for certain populations, it child agreement in report of nighttime respiratory symptoms and remains an ecologically useful device to measure aspects of sleep disruptions and quality. J. Pediatr. Health Care, 2009, 23: 315– sleep and wake. Additional research on PSG and actigraphy 326. concordance will provide more nuanced use in children with Kripke, D. F., Hahn, E. K., Grizas, A. P. et al. Wrist actigraphic chronic health conditions. scoring for sleep laboratory patients: algorithm development. J. Sleep Res., 2010, 19: 612–619. Kushida, C. A., Chang, A., Gadkary, C., Guilleminault, C., Carrillo, DECLARATIONS OF INTEREST O. and Dement, W. C. Comparison of actigraphic, polysomno- graphic, and subjective assessment of sleep parameters in sleep- None for all the authors. disordered patients. Sleep Med., 2001, 2: 389–396. Lichstein, K. L., Stone, K. C., Donaldson, J. et al. Actigraphy validation with insomnia. Sleep, 2006, 29: 232–239. ACKNOWLEDGEMENTS Littner, M., Kushida, C. A., Anderson, W. M. et al. Standard of Practice Committee of the American Academy of . The authors thank the children and families who helped with Practice parameters for the role of actigraphy in the study of sleep this research. We thank Robert Burr MSEE, PhD, Research and circadian rhythms: an update for 2002. Sleep, 2003, 26: 337–341. Professor, for his assistance with the data analysis and Lopes, M. C., Guilleminault, C., Rosa, A., Passarelli, C., Roizenblatt, S. and Tufik, S. Delta sleep instability in children with chronic Shao-Yu Tsai PhD, RN, Assistant Professor, National Taiwan arthritis. Braz. J. Med. Biol. Res., 2008, 41: 938–943. University for her assistance with the data entry. We thank Morgenthaler, T., Alessi, C., Friedman, L. et al. Standards of Practice Linda Peterson, Research Coordinator, Dr Laurie Beitz MD Committee; American Academy of Sleep Medicine. Practice param- and the staff in the Rheumatology Clinic for recruiting the eters for the use of actigraphy in the assessment of sleep and sleep participants. We thank Ernie Tolentino, Laboratory Manager, disorders: an update for 2007. Sleep, 2007, 30: 519–529. Owens, J., Sangal, B., Sutton, V. K., Bakken, R., Allen, A. J. and and the sleep laboratory staff, James Rothermel, Taryn Kelsey, D. Subjective and objective measures of sleep in children Jenkins, David Krizan and Paul Wilkinson for recording, with attention-deficit ⁄ hyperactivity disorder. Sleep Med., 2009, 10: processing and scoring of the sleep data. We also thank Hieke 446–456. Nuhsbaum for data entry and Salimah Man, Yuen Song, Paquet, J., Kawinska, A. and Carrier, J. Wake detection capacity of Tuyet Nguyen. Sarah Shapro, and Whitney Jewell for helping actigraphy during sleep. Sleep, 2007, 30: 1362–1369. Passarelli, C. M., Roizenblatt, S., Len, C. A. et al. A case-control sleep with data collection and processing. This research was study in children with polyarticular juvenile rheumatoid arthritis. supported by grants from the National Institute of Nursing J. Rheumatol., 2006, 33: 796–802. Research, T32 NR00710, NR08136, NR08238, the Center for Pepe, S. M. The Statistical Evaluation of Medical Tests for Classifi- WomenÕs Health and Gender Research, NR04011, the cation and Prediction. Oxford University Press, New York, 2003. National Center for Research Resources (NCRR), M01-RR- Pollak, C. P., Stokes, P. E. and Wagner, D. R. Direct comparison of two widely used activity recorders. Sleep, 1998, 21: 207–212. 00037, the Center for Research on Management of Sleep Ramagopal, M., Scharf, S. M., Roberts, D. W. and Blaisdell, C. J. Disturbances, NR011400 and Pulmonary Center Training Obstructive and history of asthma in snoring children. grant, T72 MC 0007. Sleep Breath., 2008, 12: 381–392. Rechtschaffen, A. and Kales, A. A Manual of Standardized Terminol- ogy, Techniques, and Scoring Systems for Sleep Stages in Human REFERENCES Subjects. Brain Information Service, Brain Research Institute, UCLA, Los Angeles, CA, 1968. Berger, A. M., Wielgus, K. K., Young-McCaughan, S., Fischer, P., Sadeh, A. and Acebo, C. The role of actigraphy in sleep medicine. Farr, L. and Lee, K. A. Methodological challenges when using Sleep Med. Rev., 2002, 6: 113–124. actigraphy in research. J. Pain Symptom Manage., 2008, 36: 191– Sadeh, A., Hauri, P. J., Kripke, D. F. and Lavie, P. The role of 199. actigraphy in the evaluation of sleep disorders. Sleep, 1995, 18: 288– Cole, R. J., Kripke, D. F., Gruen, W., Mullaney, D. J. and Gillin, J. C. 302. Automatic sleep ⁄ wake identification from wrist activity. Sleep, Sitnick, S. L., Goodlin-Jones, B. L. and Anders, T. F. The use of 1992, 15: 461–469. actigraphy to study sleep disorders in preschoolers: some concerns Hyde, M., OÕDriscoll, D. M., Binette, S. et al. Validation of actigraphy about detection of nighttime awakenings. Sleep, 2008, 31: 395–401. for determining sleep and wake in children with sleep disordered Spruyt, K., Gozal, D., Dayyat, E., Roman, A. and Molfese, D. L. breathing. J. Sleep Res., 2007, 16: 213–216. Sleep assessments in healthy school-aged children using actigraphy: Insana, S. P., Gozal, D. and Montgomery-Downs, H. E. Invalidity of concordance with polysomnography. J. Sleep Res., 2011, 20: 223– one actigraphy brand for identifying sleep and wake among infants. 232. Sleep Med., 2010, 11: 191–196. Sulit, L. G., Storfer-Isser, A., Rosen, C. L., Kirchner, L. and Jean-Louis, G., Zizi, F., von Gizycki, H. and Hauri, P. Actigraphic Redline, S. Associations of obesity, sleep-disordered breathing, and assessment of sleep in insomnia: application of the Actigraph wheezing in children. Am. J. Resp. Crit. Care Med., 2005, 171: Data Analysis Software (ADAS). Physiol. Behav., 1999, 65: 659– 659–664. 663. Tsai, S. Y., Labyak, S. E., Richardson, L. P. et al. Actigraphic sleep Jean-Louis, G., Kripke, D. F., Cole, R. J., Assmus, J. D. and Langer, and daytime in adolescent girls with chronic musculoskeletal R. D. Sleep detection with an accelerometer actigraph: comparisons pain. J. Pediatr. Psychol., 2008, 33: 307–311. with polysomnography. Physiol. Behav., 2001, 72: 21–28.

Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res. Polysomnography and actigraphy concordance 9

Van de Water, A. T., Holmes, A. and Hurley, D. A. Objective Ward, T. M., Archbold, K., Lentz, M., Ringold, S., Wallace, C. A. and measurements of sleep for non-laboratory settings as alternatives to Landis, C. A. Sleep disturbance, daytime sleepiness, and neurocog- polysomnography—a systemic review. J. Sleep Res., 2011, 20: 183– nitive performance in children with juvenile idiopathic arthritis. 200. Sleep, 2010, 33: 252–259. Ward, T. M., Brandt, P., Archbold, K. et al. Polysomnography and Zamir, G., Press, J., Tal, A. and Tarasiuk, A. Sleep fragmentation in self-reported sleep, pain, fatigue, and anxiety in children with active children with juvenile rheumatoid arthritis. J. Rheumatol., 1998, 25: and inactive juvenile rheumatoid arthritis. J. Pediatr. Psychol., 2008, 1191–1197. 33: 232–241.

Published 2011 This article is a US Government work and is in the public domain in the USA, J. Sleep Res.