TY - JOUR
AU - Biggs, Sarah N
AB - Abstract Background Despite the widespread use of actigraphy in pediatric sleep studies, there are currently no age-related normative data. Objectives To systematically review the literature, calculate pooled mean estimates of actigraphy-derived pediatric nighttime sleep variables and to examine the magnitude of change with age. Methods A systematic search was performed across eight databases of studies that included at least one actigraphy sleep variable from healthy children aged 0–18 years. Data suitable for meta-analysis were confined to ages 3–18 years with seven actigraphy variables analyzed using random effects meta-analysis and meta-regression performed using age as a covariate. Results In total, 1334 articles did not meet inclusion criteria; 87 had data suitable for review and 79 were suitable for meta-analysis. Pooled mean estimates for overnight sleep duration declined from 9.68 hours (3–5 years age band) to 8.98, 8.85, 8.05, and 7.4 for age bands 6–8, 9–11, 12–14, and 15–18 years, respectively. For continuous data, the best-fit (R2 = 0.74) equation for hours over the 0–18 years age range was 9.02 − 1.04 × [(age/10)^2 − 0.83]. There was a significant curvilinear association between both sleep onset and offset with age (p < .001). Sleep latency was stable at 19.4 min per night. There were significant differences among the older age groups between weekday and weekend/nonschool days (18 studies). Total sleep time in 15–18 years old was 56 min longer, and sleep onset and offset almost 1 and 2 hours later, respectively, on weekend or nonschool days. Conclusion These normative values have potential application to assist the interpretation of actigraphy measures from nighttime recordings across the pediatric age range, and aid future research. actigraphy, adolescent, sleep patterns, child, sleep maturation, sleep measurement Statement of Significance Actigraphy provides a greater degree of objectivity of sleep measurement than subjective reports, but despite its widespread use in pediatric sleep, there are currently no age-appropriate normative data available. However, the field has seen a growth in actigraphy studies published over the last decade to measure sleep in healthy children. As such, the current work uses systematic review, meta-analysis, and best-fit line equations to integrate published data and derive normative values for seven key actigraphy nighttime sleep variables across the pediatric age range. These data show typical sleep–wake developmental trends—albeit with shorter sleep than subjective reports—with clear weekday–weekend differences in the sleep patterns of older adolescents, and stability of sleep latency across all ages. Introduction Sleep patterns in children and adolescents are a critical marker of health and well-being. As such, it is important for normative data to be available in relation to measurement of sleep patterns across the pediatric age range. Although it is recognized that the appropriate method for sleep measurement is determined by the goal of the study or assessment, obtaining accurate estimates of sleep from infancy to adolescence—particularly in large cohorts—has inherent challenges. Polysomnography (PSG) is the gold-standard in sleep measurement. It consists of recordings from at least four channels (electroencephalography, electromyography, and two electro-oculography), from which sleep states are manually scored by a trained expert. Although PSG provides detailed information about the stages of sleep, sleep quantity, and sleep quality [1, 2], it is expensive, obtrusive, and is best conducted in a laboratory setting. This makes it difficult to assess sleep in large cohorts and it is virtually impossible to assess sleep using PSG continuously over long periods. One method that is particularly applicable to field-based studies, because it provides more objectivity than questionnaires or sleep diaries, but without the complexity of PSG, is actigraphy [3–5]. Actigraphy is a single-channel behavior monitoring system that exploits the relative absence of movement as one of the fundamental behavioral and physiological markers of sleep and is recognized as an effective and widely accepted tool for recording sleep. However, rather than directly measuring sleep–wake patterns, actigraphy devices infer sleep and wake based on the absence or presence of movement, respectively. The devices are small and unobtrusive, are commonly worn on the wrist like a wristwatch, and record sleep–wake activity continuously within the child’s normal home sleeping environment. Actigraphy causes minimal disruption to sleep [6] and is relatively simple to administer, thereby overcoming some of the key limitations of PSG. The conventional actigraphy unit incorporates an accelerometer-based motion sensor, a microprocessor and memory for data storage. An algorithm (e.g. time-above-threshold, zero-crossing, and digital integration) is applied to accelerometry data to summarize the overall intensity of measured accelerometry data across defined epochs (usually 15, 30, or 60 s) as “activity counts,” which are then recorded on the device [4]. A second algorithm is then used to score sleep and wake on the basis of activity counts. Although these algorithms vary by the population, device, and the site placement, most work in a similar fashion by defining each epoch of recorded activity (using a sliding window) as either sleep or wake by weighting the activity scores of the surrounding minutes [7]. Actigraphy is not without its limitations and certainly cannot replace the preciseness of PSG for sleep–wake discrimination, nor PSG’s ability to discriminate between sleep states. Many of the problems facing accuracy of measurement in children are inherent within the field of actigraphy itself [7]. Misclassification in sleep–wake detection occurs where periods of high activity during sleep are erroneously classified as wake, and low activity during wakefulness is erroneously classified as sleep. Variability is also encountered in different devices, placements, sensitivity settings, and statistical weightings used by different algorithms [7]. Typically, studies validating actigraphy against concurrent PSG recordings in children and adolescents show good agreement for the detection of sleep (good sensitivity), but not wake (poor specificity) [3]. Although combining several forms of sleep measurement would balance the inherent limitations of each, this is often not possible in both research and clinical studies, and as such actigraphy is frequently used in isolation. In pediatric research, actigraphy devices have primarily been used for assessing habitual sleep–wake patterns and sleep quality in healthy participants. Within a clinical setting, they can be useful for discriminating between circadian disorders [4], identifying insomnia and hypersomnia, and identifying significant sleep disturbance in children with chronic medical conditions [5]. Most of our understanding of developmental sleep–wake patterns from birth to the end of adolescence has been derived from parental report (questionnaires or diaries) [8–11]. However, such parental reports are subjective and susceptible to reporting biases when compared with more objective measures. For example, parents overestimate their child’s sleep and underestimate the length of their nighttime awakenings [12–14]. Actigraphy can provide estimates of sleep onset and sleep offset to define both sleep quantity and sleep quality more accurately than diaries or questionnaires; the latter rely on parents to accurately report their child’s sleep–wake variables. These limitations of parental report and widespread use of actigraphy provide clear motivation to establish normative values using actigraphy data from healthy children. Actigraphy devices can support large-scale and longitudinal studies and could provide a rich data source for describing developmental patterns of sleep more accurately than subjective data. However, although no age-related normative values for actigraphy variables across the full pediatric age range exist [7], the field has seen a growth in studies published over the last decade utilizing actigraphy to measure sleep in children. Thus, a panel of pediatric sleep experts from the Australasian Sleep Association was convened to conduct a systematic review of the sleep literature and meta-analysis of normal values of actigraphy-derived sleep variables from studies of healthy children across the pediatric age range. Actigraphy has rarely been used to capture 24 hr sleep–wake patterns (i.e. most recordings in infants and young children do not include daytime napping), and this analysis was restricted to nighttime sleep–wake patterns. The aim of this study was to pool data and conduct a meta-analysis as a first step towards establishing normative estimates of child nighttime sleep assessed by actigraphy and to examine these in relation to sleep–wake developmental trends. More specifically we aimed to clarify the extent to which age influenced key actigraphy sleep variables, and the differences encountered when data were collected over the weekend as opposed to weekdays. Methods Search strategy Articles reporting on pediatric sleep collected via actigraphy and published prior to January 2016 were sourced from PubMed, Ovid Medline, PsycInfo, Scopus, Embase, CINAHL, Web of Science, and Cochrane Library databases by two authors (M.F.-O. and B.C.G.). The search was updated on June 1, 2017. Potentially relevant articles were identified using the keywords sleep and infant (age 0 to 23 months) or child, preschool (age 2 to 5 years) or child (age 6 to 12 years) or adolescent (age 13 to <19) and actigraphy. Titles and abstracts were examined by seven co-authors to extract potentially relevant articles, each of which was downloaded and examined against the inclusion or exclusion criteria as described in Table 1. There are currently many actigraphy instruments on the market, including the newer consumer targeted commercial sleep tracking devices and apps—most of which have not been validated. The emphasis for the data analyses was on studies using actigraphy for research and clinical practice, i.e. studies using devices primarily targeting a consumer market, and were not FDA-approved, were excluded. It is important to note that, within the actigraphy field, there is a lack of standardization of scoring rules for sleep, made difficult by the variety of proprietary hardware used utilizing different sampling rates and algorithms. To facilitate the pooling of actigraphy data for this study, it was impossible to take all these differences into account. Table 1. Systematic review inclusion and exclusion criteria Inclusion criteria  1. The study reported data from nighttime sleep collected via actigraphy  2. Five or more days of actigraphy were used  3. Included children aged 0 to 18 years  4. Study included a nonclinic sample  5. Article must be available in English (either written in English or a translation available)  Exclusion criteria  1. Case reports  2. Review articles  3. Abstract only  4. Dissertations  5. Clinic sample only  6. In studies where clinic patients or special population groups were compared with controls, only the control data were used  7. In studies where an intervention was implemented, only preintervention data were used  8. Where more than one published report used the same study participants, we included the report with the most detailed information  9. The study did not separate nighttime from daytime sleep  Inclusion criteria  1. The study reported data from nighttime sleep collected via actigraphy  2. Five or more days of actigraphy were used  3. Included children aged 0 to 18 years  4. Study included a nonclinic sample  5. Article must be available in English (either written in English or a translation available)  Exclusion criteria  1. Case reports  2. Review articles  3. Abstract only  4. Dissertations  5. Clinic sample only  6. In studies where clinic patients or special population groups were compared with controls, only the control data were used  7. In studies where an intervention was implemented, only preintervention data were used  8. Where more than one published report used the same study participants, we included the report with the most detailed information  9. The study did not separate nighttime from daytime sleep  View Large Table 1. Systematic review inclusion and exclusion criteria Inclusion criteria  1. The study reported data from nighttime sleep collected via actigraphy  2. Five or more days of actigraphy were used  3. Included children aged 0 to 18 years  4. Study included a nonclinic sample  5. Article must be available in English (either written in English or a translation available)  Exclusion criteria  1. Case reports  2. Review articles  3. Abstract only  4. Dissertations  5. Clinic sample only  6. In studies where clinic patients or special population groups were compared with controls, only the control data were used  7. In studies where an intervention was implemented, only preintervention data were used  8. Where more than one published report used the same study participants, we included the report with the most detailed information  9. The study did not separate nighttime from daytime sleep  Inclusion criteria  1. The study reported data from nighttime sleep collected via actigraphy  2. Five or more days of actigraphy were used  3. Included children aged 0 to 18 years  4. Study included a nonclinic sample  5. Article must be available in English (either written in English or a translation available)  Exclusion criteria  1. Case reports  2. Review articles  3. Abstract only  4. Dissertations  5. Clinic sample only  6. In studies where clinic patients or special population groups were compared with controls, only the control data were used  7. In studies where an intervention was implemented, only preintervention data were used  8. Where more than one published report used the same study participants, we included the report with the most detailed information  9. The study did not separate nighttime from daytime sleep  View Large Study selection From each of the articles meeting inclusion or exclusion criteria, the following information and data were extracted by seven co-authors based on published guidelines for actigraphy use in the pediatric population [15] and following consensus agreement: (1) article information (author, year of publication, country and region of origin, and study design); (2) participants (sample size, gender and ratios, mean age and range or standard deviation [SD], and percent Caucasian); (3) actigraph (model, diary use, algorithm used, threshold settings, epoch length, number of recording days, and whether weekday or weekend or both); and (4) sleep variables of interest with a measure of central tendency and variance (bedtime, wake time, sleep onset latency, sleep onset time, sleep offset time, time in bed [TIB], sleep duration, total sleep time [TST], sleep efficiency, waking after sleep onset [WASO], and the frequency of nocturnal waking). In this study, sleep duration, also referred to as sleep period time (SPT) in some studies, is defined as the elapsed time (hours, minutes) between sleep onset and offset, and thus excludes sleep latency. TST, also referred to as true sleep time, was extracted from studies defining this as the amount of time between sleep onset and offset scored as sleep (thus excludes wake). Sleep efficiency was defined by studies as either (TST/TIB) × 100 (expressed as percent) and thus includes sleep latency, or as (TST/SPT) × 100 (expressed as a percent) that excludes sleep latency, and these definitions were reported separately accordingly. Only values extracted from the data for a particular sleep variable as presented were included. If the age range of included participants was not given, the range was estimated from the SD. If the article presented data stratified only by male or female, or only by weekday or weekend or nonschool day, results were pooled according to formulae for combining groups for sample size, mean, and SD as appropriate [16]. Within a single meta-analysis and where data from a longitudinal study were eligible, we took one time point to avoid bias by overweighting; the first time point was chosen based on the rationale that the participant information would match most closely to this time point. Where there was uncertainty about participant overlap in studies, authors were contacted to provide clarification. Some studies with participant overlap provided data for different sleep variables, meaning the overlapping studies could be retained as part of the database of articles contributing to data syntheses. This involved four articles from two cohorts [17–20]. Data analyses All analyses were performed using Stata/IC 14.2 (Statacorp LP, College Station, TX). Meta-analysis was performed using the mean and standard error (SE) for each study to produce the pooled estimate mean and the 95% confidence intervals (CI) using fixed or random-effects models as appropriate and weighted by inverse variance. As most data describing variability were given as the SD, the SE was calculated before pooling the data by using the number of participants for each study. A shortage of data for appropriately integrating data from infants aged 0–2 years into narrower developmental age bands meant that nine studies were excluded from the meta-analysis, a posteriori (as detailed below). Studies that were eligible for meta-analysis were stratified into five age categories (3–5, 6–8, 9–11, 12–14, and 15–18 years) for each sleep variable (depending on the supply of data). Forest plots were used as a graphical display of the strength of the individual studies, age groups, and entire analysis contributing to the pooled mean estimates. Between studies, heterogeneity was assessed using the Cochran Q statistic, and inconsistency measured by I2 (the percentage of total variation across studies that is due to heterogeneity rather than chance). A p > .01 for the Q statistic, or I2 < 50 per cent, was taken as an indicator of low heterogeneity to allow a fixed-effect model in meta-analysis, otherwise a random effect model was used [21]. Meta-regression analysis (weighted by inverse variance) was conducted to determine age group effects for each sleep variable. We also used fractional polynomials together with a random effects model as appropriate to perform meta-regression analysis (weighted by inverse variance) using mean study age as a continuous variable. Age was considered the main potential source of variability in all meta-regression models. Study region was also investigated and the Variance Accounted For (VAF) factor used in meta-regression models to indicate the proportion of total variance explained by these covariates. A covariate was considered a potential source of variability if p < .20. Unless otherwise stated, a p-value of <.05 was considered as statistically significant. Results Systematic review and meta-analysis: included/excluded studies Figure 1 illustrates the study selection process and a summary of the reasons for exclusion. We identified 87 articles (85 studies, 95 separate age cohorts/datasets) containing data from healthy children from 1423 publications (after duplicates were removed) that met our inclusion or exclusion criteria (Table 1). Thirty-eight studies were cross-sectional, 16 longitudinal (used first time point only), 25 case–control (control data used only), 4 experimental (sleep restriction or extension and 1 quasi-experimental comparing low stress versus high stress exam time; baseline data only used for these designs), and 2 randomized controlled trials. Studies were undertaken in 17 different countries across five different regions (Northern America, Europe, Asia, Middle East, South America, and Oceania). The number of children per dataset ranged from 20 to 515 (median 50) with a median gender M:F ratio of 1.0. Three studies used male participants only, one study female participants only, and in three studies the gender distribution was unstated. The age ranges were <4 years in 80 per cent of studies. Ten studies did not provide age range information, and thus, it was estimated from the standard deviation. In four studies, an estimated range could not be calculated. One author was contacted to confirm the age range [22], one study was categorized as preschool, and two were infant studies with several other age points included allowing us to assume a narrow band [23, 24]. Thirty-five studies (41 per cent) gave an ethnicity breakdown; participants were predominantly Caucasian (median 69 per cent). Forty-three studies (50.5 per cent) stated the days of the week they collected actigraphy, nine studies stated that they only collected weekday data, usually Sunday or Monday to Thursday or Friday with one study stating that the 5 day data collection crossed the weekend. Few studies reported on whether or not the weekdays were school days. Figure 1. View largeDownload slide Flow chart of the systematic review process and meta-analysis. The systematic review process included articles with children aged 0–18 years. The meta-analysis of studies included children aged 3–18 years, i.e. nine articles (0–2 years) were excluded because of the limited supply of data to pool into developmentally appropriate age groups. Figure 1. View largeDownload slide Flow chart of the systematic review process and meta-analysis. The systematic review process included articles with children aged 0–18 years. The meta-analysis of studies included children aged 3–18 years, i.e. nine articles (0–2 years) were excluded because of the limited supply of data to pool into developmentally appropriate age groups. Only one study did not report the actigraphy device [25]. Of the others, 36 (43 per cent) used Ambulatory Monitoring Inc. devices (Ardsley, New York), 30 (36 per cent) used Mini-Mitter devices (Phillips-Respironics, Bend, Oregon), 8 (10 per cent) used Cambridge Actiwatch (Cambridge, UK), 7 (8 per cent) reported using ActiGraph devices (ActiGraph LLC, Pensacola, FL), and three studies used other brands: Lifecorder (Suzuken, Nagoya, Japan), SenseWear (BodyMedia, Pittsburgh, PA), and Somnowatch (Somnomedics, Randersacker, Germany). Devices were mainly wrist-worn (ankle/shin in young infants), one arm [26], and three were waist-worn [22, 27, 28]. One study compared waist and wrist data; wrist data only were used [29]. Variables incorporating waking were not extracted from waist-worn devices, given potentially poorer concordance with these wrist-worn outputs [29]. TST data from the arm-worn device were used because this device has been validated against PSG for TST [26]. Wherever applicable, sensitivity analyses were performed to determine the effect on mean estimates of excluding data from waist-worn devices. All variables extracted were related to overnight sleep, i.e. they did not include napping or 24 hr sleep data. Meta-analysis: excluded studies Eight stand-alone and one cross-sectional study contained data from infants aged 0–2 years and thus were excluded from the meta-analysis (Table 2). Five were cross-sectional, two longitudinal, one case–control, and one randomized controlled trial. The 79 included studies are summarized in Table 3. Table 2. Summary of studies contributing to systematic review, but data excluded from meta-analysis a posteriori First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Volkovich, 2015 [71]  Israel  L-sub  NA  0.25  29  TST, WASO  Galland, 2017 [72]  New Zealand  RCT  NA  0.5  138  TST, S-on, S-off, SE-spt, WASO  Newland, 2016 [24]  USA  L-sub  NA  0.67  132  SE-spt, WASO  Acebo, 2005† [23]  USA  X  NA  1.0  24  SD, TST, S-on, S-off, SE-spt  Montgomery-Downs, 2006 [73]  USA  X-sub  NA  1.2  20  TST, S-on, S-off, SE-spt  Sadeh, 2007 [74]  Israel  C-C  NA  1.3  48  SD, S-on, WASO  Acebo, 2005† [23]  USA  X  NA  1.5  29  SD, TST, S-on, S-off, SE-spt  Kohyama, 2007 [75]  Japan  X  NA  1.9  204  SD, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.0  22  SD, TST, S-on, S-off, SE-spt  Molfese, 2015 [76]  USA  X-sub  NA  2.5  60  TST, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.5  21  SD, TST, S-on, S-off, SE-spt  LeBourgeois, 2013 [77]  USA  X  NA  2.8  45  SD, S-on, S-off  First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Volkovich, 2015 [71]  Israel  L-sub  NA  0.25  29  TST, WASO  Galland, 2017 [72]  New Zealand  RCT  NA  0.5  138  TST, S-on, S-off, SE-spt, WASO  Newland, 2016 [24]  USA  L-sub  NA  0.67  132  SE-spt, WASO  Acebo, 2005† [23]  USA  X  NA  1.0  24  SD, TST, S-on, S-off, SE-spt  Montgomery-Downs, 2006 [73]  USA  X-sub  NA  1.2  20  TST, S-on, S-off, SE-spt  Sadeh, 2007 [74]  Israel  C-C  NA  1.3  48  SD, S-on, WASO  Acebo, 2005† [23]  USA  X  NA  1.5  29  SD, TST, S-on, S-off, SE-spt  Kohyama, 2007 [75]  Japan  X  NA  1.9  204  SD, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.0  22  SD, TST, S-on, S-off, SE-spt  Molfese, 2015 [76]  USA  X-sub  NA  2.5  60  TST, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.5  21  SD, TST, S-on, S-off, SE-spt  LeBourgeois, 2013 [77]  USA  X  NA  2.8  45  SD, S-on, S-off  Study design: C-C = Case-Comparison/Control (Comparison/control data only used); L = Longitudinal (if results of more than one time point represented, first time point only used); X = Cross-sectional; -sub = subsample. Actigraphy variables: SD = Sleep Duration; SE-spt = Sleep Efficiency calculated using SPT as the denominator; S-on = Sleep Onset; S-off = Sleep Offset; TST = Total Sleep Time; WASO = Waking After Sleep Onset. †Same study contributing different age groups. View Large Table 2. Summary of studies contributing to systematic review, but data excluded from meta-analysis a posteriori First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Volkovich, 2015 [71]  Israel  L-sub  NA  0.25  29  TST, WASO  Galland, 2017 [72]  New Zealand  RCT  NA  0.5  138  TST, S-on, S-off, SE-spt, WASO  Newland, 2016 [24]  USA  L-sub  NA  0.67  132  SE-spt, WASO  Acebo, 2005† [23]  USA  X  NA  1.0  24  SD, TST, S-on, S-off, SE-spt  Montgomery-Downs, 2006 [73]  USA  X-sub  NA  1.2  20  TST, S-on, S-off, SE-spt  Sadeh, 2007 [74]  Israel  C-C  NA  1.3  48  SD, S-on, WASO  Acebo, 2005† [23]  USA  X  NA  1.5  29  SD, TST, S-on, S-off, SE-spt  Kohyama, 2007 [75]  Japan  X  NA  1.9  204  SD, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.0  22  SD, TST, S-on, S-off, SE-spt  Molfese, 2015 [76]  USA  X-sub  NA  2.5  60  TST, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.5  21  SD, TST, S-on, S-off, SE-spt  LeBourgeois, 2013 [77]  USA  X  NA  2.8  45  SD, S-on, S-off  First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Volkovich, 2015 [71]  Israel  L-sub  NA  0.25  29  TST, WASO  Galland, 2017 [72]  New Zealand  RCT  NA  0.5  138  TST, S-on, S-off, SE-spt, WASO  Newland, 2016 [24]  USA  L-sub  NA  0.67  132  SE-spt, WASO  Acebo, 2005† [23]  USA  X  NA  1.0  24  SD, TST, S-on, S-off, SE-spt  Montgomery-Downs, 2006 [73]  USA  X-sub  NA  1.2  20  TST, S-on, S-off, SE-spt  Sadeh, 2007 [74]  Israel  C-C  NA  1.3  48  SD, S-on, WASO  Acebo, 2005† [23]  USA  X  NA  1.5  29  SD, TST, S-on, S-off, SE-spt  Kohyama, 2007 [75]  Japan  X  NA  1.9  204  SD, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.0  22  SD, TST, S-on, S-off, SE-spt  Molfese, 2015 [76]  USA  X-sub  NA  2.5  60  TST, S-on, S-off  Acebo, 2005† [23]  USA  X  NA  2.5  21  SD, TST, S-on, S-off, SE-spt  LeBourgeois, 2013 [77]  USA  X  NA  2.8  45  SD, S-on, S-off  Study design: C-C = Case-Comparison/Control (Comparison/control data only used); L = Longitudinal (if results of more than one time point represented, first time point only used); X = Cross-sectional; -sub = subsample. Actigraphy variables: SD = Sleep Duration; SE-spt = Sleep Efficiency calculated using SPT as the denominator; S-on = Sleep Onset; S-off = Sleep Offset; TST = Total Sleep Time; WASO = Waking After Sleep Onset. †Same study contributing different age groups. View Large Table 3. Summary of studies contributing to meta-analyses First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Acebo, 2005† [23]  USA  X  3–5  3  21  SD, TST, S-on, S-off, SE-spt  Williams, 2014 [22]  NZ  L  3–5  3.0  216  SD  Goodlin-Jones, 2008‡ [19]  USA  L-nc  3–5  3.4  69  S-on, S-off, SE-tib  Anders, 2011‡ [20]  USA  L-nc  3–5  3.4  69  SL, WASO  Acebo, 2005† [23]  USA  X  3–5  4.0  24  SD, TST, S-on, S-off, SE-spt  Elmore-Staton, 2012 [78]  USA  X  3–5  4.0  29  TST, SE-tib  Lam, 2011 [79]  USA  X  3–5  4.3  59  TST, S-on, S-off  Ishihara, 2014 [33]  Japan  X-sub  3–5  4.8  72  SD, S-on, S-off, SL  Hatzinger, 2012 [80]  Switzerland  C-C  3–5  4.9  35  SD, TST, SE-spt, SL, WASO  Kushnir, 2011 [81]  Israel  C-C  3–5  4.9  30  SD, TST, SE-spt  Acebo, 2005† [23]  USA  X  3–5  5.0  28  SD, TST, S-on, S-off, SE-spt  Cairns, 2014 [82]  USA  L  3–5  5.0  34  SD, S-on, S-off, SE-spt  Iwasaki, 2010 [83]  Japan  X  3–5  5.0  47  SD, S-on, S-off, SL  Werner, 2008 [32]  Switzerland  X  3–5  5.9  50  SD, TST, S-on, S-off, WASO  Ravid, 2009 [84]  Israel  X  6–8  6.0  92  TST, SE-tib, SL  Souders, 2009 [31]  USA  C-C  6–8  7.1  40  SD, TST, S-off, SE-spt, WASO  Burt, 2013 [85]  Canada  X  6–8  7.7  56  SD, TST, S-on, S-off, SL  Michels, 2013 [86]  Belgium  L-sub  6–8  7.9  165  TST, SE-tib, SL, WASO  Hvolby, 2008 [87]  Denmark  C-C  6–8  8.0  97  TST, SL  Pesonen, 2014 [34]  Finland  L-sub  6–8  8.15  188  SD, SE-tib, SL  Bagheri, 2015 [66]  Teheran  X  6–8  8.3  111  SD, SE-spt  Markovich, 2015 [88]  Canada  X-sub  6–8  8.6  30  SD, TST, SL, WASO  Kelly, 2014a§ [18]  USA  L  6–8  8.68  176  TST, SE-spt, WASO  Keller, 2011§ [17]  USA  L  6–8  8.68  176  S-on, S-off, SL  Wiebe, 2013 [89]  Canada  C-C  6–8  8.7  46  SD, SL, WASO  Alfano, 2015 [35]  USA  C-C  6–8  8.7  36  SD, TST, S-off, SL  Holley, 2011 [90]  UK  X  6–8  8.8  91  SD, TST, SE-tib  Gruber, 2011 [91]  Canada  C-C Exp  6–8  8.8  32  TST, SE-spt  Ashworth, 2013 [92]  UK  C-C  9–11  9.2  51  SD, TST, SE-spt, SL, WASO  Holley, 2014 [93]  UK  C-C  9–11  9.3  50  SD, TST, SE-spt  Gruber, 2000 [25]  Israel  C-C  9–11  9.4  64  SD, TST, S-on, SE-tib  Geiger, 2010|| [94]  Switzerland  X  9–11  9.4  60  SD  Kelly, 2014b [95]  USA  L  9–11  9.4  282  TST, SE-spt  Moreau, 2013 [96]  Canada  C-C  9–11  9.6  41  TST, SE-tib, SL,WASO  Corkum, 2001|| [97]  Canada  C-C  9–11  9.7  25  SD, S-on, S-off, SL  Hvolby, 2011 [98]  Denmark  C-C  9–11  9.8  21  TST, SL  Vriend, 2012 [99]  Canada  X  9–11  9.8  32  SD, S-on, S-off, SE-tib  McNeil, 2015 [27]  Canada  X  9–11  10.0  515  SD, S-off  Fallone, 2002 [100]  US  Exp  9–11  10.2  78  SD, TST  Yamakita, 2014|| [101]  Japan  X-sub  9–11  10.2  42  SD, TST, S-off  Tzischinsky, 2006 [30]  Israel  C-C  9–11  10.2  25  SD, TST, S-on, S-off, SE-spt, SL,WASO  Martoni, 2016 [102]  Switzerland  X  9–11  10.2  115  TST, SE-tib, SL, WASO  Burszstein, 2006 [103]  Israel  C-C  9–11  10.3  108  SD, TST, S-on, S-off  Hjorth, 2012 [29]  Denmark  X  9–11  10.3  62  TST, SE-tib  Barreira, 2015 [28]  Canada  X  9–11  10.3  34  SD, S-on, S-off  Vincent, 2017 [26]  Australia  X  9–11  10.4  65  TST  Gaina, 2004||,¶ [68]  Japan  X  9–11  10.8  41  SD, TST, S-on, S-off, SE-tib, SL, WASO  Allick, 2006|| [104]  Sweden  C-C  9–11  10.9  32  SD, TST, S-on, S-off, SE-tib, SL, WASO  Filardi, 2016 [105]  Italy  C-C  9–11  10.9  21  TST, SE-tib, SL, WASO  Bagley, 2016 [106]  USA  L  9–11  11.3  210  SD, TST, SE-spt  Greenfeld, 2011 [107]  Israel  C-C  9–11  11.4  61  TST, SE-tib, SL  Tremaine, 2010|| [108]  Australia  X  9–11  11.5  54  TST, S-on, WASO  Aroro, 2013|| [109]  UK  X  12–14  12.0  225  SD  Rigney, 2015 [110]  Australia  RCT  12–14  12.2  87  TST, S-off, SE-tib, SL  Marco, 2011|| [111]  USA  L  12–14  12.6  155  SD, TST, S-on, S-off  Amin, 2005 [112]  USA  C-C  12–14  12.6  40  TST, SE-spt, WASO  Beebe, 2007|| [113]  USA  C-C  12–14  12.6  22  SD, S-on, S-off, SE-spt  Guedes, 2016 [114]  Brazil  X-sub  12–14  13.2  37  SD  Auger, 2011 [115]  USA  C-C  12–14  13.7  22  TST, S-on, S-off, SE-tib  Anderson, 2009|| [116]  USA  L-sub  12–14  13.7  236  SD, SE-tib  Beijamini, 2008 [67]  Brazil  X  12–14  13.8  34  SD, S-on, S-off  Gaina, 2004||,¶ [68]  Japan  X  12–14  14.2  42  SD, TST, S-on, S-off, SE-tib, SL, WASO  Tavernier, 2016 [117]  USA  X-sub  12–14  14.37  77  TST, SE-tib, SL, WASO  Chen, 2011 [118]  USA  X  12–14  14.5  44  SD, SL, WASO  Huynh, 2015|| [119]  Canada  C-C  12–14  14.7  20  SD, TST, S-off, SE-tib, SL,WASO  Law, 2012 [120]  USA  L-sub  12–14  14.8  60  TST, SE-tib, WASO  Fobian, 2016|| [121]  USA  X-sub  12–14  14.9  55  SD, S-on, S-off, SE-tib  Carskadon, 1998|| [122]  USA  L  15–18  15.0  32  TST, S-on, S-off  Jiang, 2011 [123]  China  Exp  15–18  15.0  20  SD  Dewald, 2014|| [124]  Netherlands  Expq  15–18  15.1  175  TST, S-on, S-off, SE-tib, SL  Tham, 2015 [125]  USA  C-C  15–18  15.3  50  SD, WASO  Dewald, 2012# [126]  Australia  X  15–18  15.5  236  SD, TST, SE-tib, SL  Baum, 2014 [127]  USA  Exp  15–18  15.5  50  SD, S-on, S-off  Baker, 2013|| [128]  Australia  C-C  15–18  15.5  27  SD, S-on, S-off, SL  Malone, 2016 [129]  USA  X-sub  15–18  15.5  68  SD  Matthews, 2014|| [130]  USA  X  15–18  15.7  250  TST  El-Sheikh, 2016 [131]  USA  L-sub  15–18  15.9  252  SD  Bei, 2013|| [36]  Australia  X  15–18  16.2  146  TST, SE-tib, SL  Park, 2016 [132]  USA  X  15–18  16.4  315  TST  Rodriguez-Colon, 2015 [133]  USA  L-sub  15–18  16.7  322  TST, SE-tib  Astill, 2013 [134]  Netherlands  X  15–18  17.6  24  TST, SE-tib, SL, WASO  Tonetti, 2015 [135]  Italy  X  15–18  18.1  36  TST, S-off, SE-tib  First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Acebo, 2005† [23]  USA  X  3–5  3  21  SD, TST, S-on, S-off, SE-spt  Williams, 2014 [22]  NZ  L  3–5  3.0  216  SD  Goodlin-Jones, 2008‡ [19]  USA  L-nc  3–5  3.4  69  S-on, S-off, SE-tib  Anders, 2011‡ [20]  USA  L-nc  3–5  3.4  69  SL, WASO  Acebo, 2005† [23]  USA  X  3–5  4.0  24  SD, TST, S-on, S-off, SE-spt  Elmore-Staton, 2012 [78]  USA  X  3–5  4.0  29  TST, SE-tib  Lam, 2011 [79]  USA  X  3–5  4.3  59  TST, S-on, S-off  Ishihara, 2014 [33]  Japan  X-sub  3–5  4.8  72  SD, S-on, S-off, SL  Hatzinger, 2012 [80]  Switzerland  C-C  3–5  4.9  35  SD, TST, SE-spt, SL, WASO  Kushnir, 2011 [81]  Israel  C-C  3–5  4.9  30  SD, TST, SE-spt  Acebo, 2005† [23]  USA  X  3–5  5.0  28  SD, TST, S-on, S-off, SE-spt  Cairns, 2014 [82]  USA  L  3–5  5.0  34  SD, S-on, S-off, SE-spt  Iwasaki, 2010 [83]  Japan  X  3–5  5.0  47  SD, S-on, S-off, SL  Werner, 2008 [32]  Switzerland  X  3–5  5.9  50  SD, TST, S-on, S-off, WASO  Ravid, 2009 [84]  Israel  X  6–8  6.0  92  TST, SE-tib, SL  Souders, 2009 [31]  USA  C-C  6–8  7.1  40  SD, TST, S-off, SE-spt, WASO  Burt, 2013 [85]  Canada  X  6–8  7.7  56  SD, TST, S-on, S-off, SL  Michels, 2013 [86]  Belgium  L-sub  6–8  7.9  165  TST, SE-tib, SL, WASO  Hvolby, 2008 [87]  Denmark  C-C  6–8  8.0  97  TST, SL  Pesonen, 2014 [34]  Finland  L-sub  6–8  8.15  188  SD, SE-tib, SL  Bagheri, 2015 [66]  Teheran  X  6–8  8.3  111  SD, SE-spt  Markovich, 2015 [88]  Canada  X-sub  6–8  8.6  30  SD, TST, SL, WASO  Kelly, 2014a§ [18]  USA  L  6–8  8.68  176  TST, SE-spt, WASO  Keller, 2011§ [17]  USA  L  6–8  8.68  176  S-on, S-off, SL  Wiebe, 2013 [89]  Canada  C-C  6–8  8.7  46  SD, SL, WASO  Alfano, 2015 [35]  USA  C-C  6–8  8.7  36  SD, TST, S-off, SL  Holley, 2011 [90]  UK  X  6–8  8.8  91  SD, TST, SE-tib  Gruber, 2011 [91]  Canada  C-C Exp  6–8  8.8  32  TST, SE-spt  Ashworth, 2013 [92]  UK  C-C  9–11  9.2  51  SD, TST, SE-spt, SL, WASO  Holley, 2014 [93]  UK  C-C  9–11  9.3  50  SD, TST, SE-spt  Gruber, 2000 [25]  Israel  C-C  9–11  9.4  64  SD, TST, S-on, SE-tib  Geiger, 2010|| [94]  Switzerland  X  9–11  9.4  60  SD  Kelly, 2014b [95]  USA  L  9–11  9.4  282  TST, SE-spt  Moreau, 2013 [96]  Canada  C-C  9–11  9.6  41  TST, SE-tib, SL,WASO  Corkum, 2001|| [97]  Canada  C-C  9–11  9.7  25  SD, S-on, S-off, SL  Hvolby, 2011 [98]  Denmark  C-C  9–11  9.8  21  TST, SL  Vriend, 2012 [99]  Canada  X  9–11  9.8  32  SD, S-on, S-off, SE-tib  McNeil, 2015 [27]  Canada  X  9–11  10.0  515  SD, S-off  Fallone, 2002 [100]  US  Exp  9–11  10.2  78  SD, TST  Yamakita, 2014|| [101]  Japan  X-sub  9–11  10.2  42  SD, TST, S-off  Tzischinsky, 2006 [30]  Israel  C-C  9–11  10.2  25  SD, TST, S-on, S-off, SE-spt, SL,WASO  Martoni, 2016 [102]  Switzerland  X  9–11  10.2  115  TST, SE-tib, SL, WASO  Burszstein, 2006 [103]  Israel  C-C  9–11  10.3  108  SD, TST, S-on, S-off  Hjorth, 2012 [29]  Denmark  X  9–11  10.3  62  TST, SE-tib  Barreira, 2015 [28]  Canada  X  9–11  10.3  34  SD, S-on, S-off  Vincent, 2017 [26]  Australia  X  9–11  10.4  65  TST  Gaina, 2004||,¶ [68]  Japan  X  9–11  10.8  41  SD, TST, S-on, S-off, SE-tib, SL, WASO  Allick, 2006|| [104]  Sweden  C-C  9–11  10.9  32  SD, TST, S-on, S-off, SE-tib, SL, WASO  Filardi, 2016 [105]  Italy  C-C  9–11  10.9  21  TST, SE-tib, SL, WASO  Bagley, 2016 [106]  USA  L  9–11  11.3  210  SD, TST, SE-spt  Greenfeld, 2011 [107]  Israel  C-C  9–11  11.4  61  TST, SE-tib, SL  Tremaine, 2010|| [108]  Australia  X  9–11  11.5  54  TST, S-on, WASO  Aroro, 2013|| [109]  UK  X  12–14  12.0  225  SD  Rigney, 2015 [110]  Australia  RCT  12–14  12.2  87  TST, S-off, SE-tib, SL  Marco, 2011|| [111]  USA  L  12–14  12.6  155  SD, TST, S-on, S-off  Amin, 2005 [112]  USA  C-C  12–14  12.6  40  TST, SE-spt, WASO  Beebe, 2007|| [113]  USA  C-C  12–14  12.6  22  SD, S-on, S-off, SE-spt  Guedes, 2016 [114]  Brazil  X-sub  12–14  13.2  37  SD  Auger, 2011 [115]  USA  C-C  12–14  13.7  22  TST, S-on, S-off, SE-tib  Anderson, 2009|| [116]  USA  L-sub  12–14  13.7  236  SD, SE-tib  Beijamini, 2008 [67]  Brazil  X  12–14  13.8  34  SD, S-on, S-off  Gaina, 2004||,¶ [68]  Japan  X  12–14  14.2  42  SD, TST, S-on, S-off, SE-tib, SL, WASO  Tavernier, 2016 [117]  USA  X-sub  12–14  14.37  77  TST, SE-tib, SL, WASO  Chen, 2011 [118]  USA  X  12–14  14.5  44  SD, SL, WASO  Huynh, 2015|| [119]  Canada  C-C  12–14  14.7  20  SD, TST, S-off, SE-tib, SL,WASO  Law, 2012 [120]  USA  L-sub  12–14  14.8  60  TST, SE-tib, WASO  Fobian, 2016|| [121]  USA  X-sub  12–14  14.9  55  SD, S-on, S-off, SE-tib  Carskadon, 1998|| [122]  USA  L  15–18  15.0  32  TST, S-on, S-off  Jiang, 2011 [123]  China  Exp  15–18  15.0  20  SD  Dewald, 2014|| [124]  Netherlands  Expq  15–18  15.1  175  TST, S-on, S-off, SE-tib, SL  Tham, 2015 [125]  USA  C-C  15–18  15.3  50  SD, WASO  Dewald, 2012# [126]  Australia  X  15–18  15.5  236  SD, TST, SE-tib, SL  Baum, 2014 [127]  USA  Exp  15–18  15.5  50  SD, S-on, S-off  Baker, 2013|| [128]  Australia  C-C  15–18  15.5  27  SD, S-on, S-off, SL  Malone, 2016 [129]  USA  X-sub  15–18  15.5  68  SD  Matthews, 2014|| [130]  USA  X  15–18  15.7  250  TST  El-Sheikh, 2016 [131]  USA  L-sub  15–18  15.9  252  SD  Bei, 2013|| [36]  Australia  X  15–18  16.2  146  TST, SE-tib, SL  Park, 2016 [132]  USA  X  15–18  16.4  315  TST  Rodriguez-Colon, 2015 [133]  USA  L-sub  15–18  16.7  322  TST, SE-tib  Astill, 2013 [134]  Netherlands  X  15–18  17.6  24  TST, SE-tib, SL, WASO  Tonetti, 2015 [135]  Italy  X  15–18  18.1  36  TST, S-off, SE-tib  Study design: C-C = Case-Comparison/Control (Comparison/control data only used); Exp = Experimental (sleep restriction, extension or both, baseline data only used); Expq = Quasiexperimental; L = Longitudinal (if results of more than one time point represented, first time point only used); X = Cross-sectional; -sub = subsample. Actigraphy variables: SD = Sleep Duration; SE-tib = Sleep Efficiency calculated using TIB as the denominator; SE-spt = Sleep Efficiency calculated using SPT as the denominator; S-on = Sleep Onset; S-off = Sleep Offset; TST = Total Sleep Time; WASO = Waking After Sleep Onset. ||Dataset included weekday or weekend splits suitable for analyses. †,¶Same studies contributing different age groups to the meta-analyses. ‡,§Same cohorts contributing different actigraphy variables to the meta-analyses. #Study included Australian and Netherland participants; Australian cohort only used because of overlap with Dewald, 2014 [126]. View Large Table 3. Summary of studies contributing to meta-analyses First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Acebo, 2005† [23]  USA  X  3–5  3  21  SD, TST, S-on, S-off, SE-spt  Williams, 2014 [22]  NZ  L  3–5  3.0  216  SD  Goodlin-Jones, 2008‡ [19]  USA  L-nc  3–5  3.4  69  S-on, S-off, SE-tib  Anders, 2011‡ [20]  USA  L-nc  3–5  3.4  69  SL, WASO  Acebo, 2005† [23]  USA  X  3–5  4.0  24  SD, TST, S-on, S-off, SE-spt  Elmore-Staton, 2012 [78]  USA  X  3–5  4.0  29  TST, SE-tib  Lam, 2011 [79]  USA  X  3–5  4.3  59  TST, S-on, S-off  Ishihara, 2014 [33]  Japan  X-sub  3–5  4.8  72  SD, S-on, S-off, SL  Hatzinger, 2012 [80]  Switzerland  C-C  3–5  4.9  35  SD, TST, SE-spt, SL, WASO  Kushnir, 2011 [81]  Israel  C-C  3–5  4.9  30  SD, TST, SE-spt  Acebo, 2005† [23]  USA  X  3–5  5.0  28  SD, TST, S-on, S-off, SE-spt  Cairns, 2014 [82]  USA  L  3–5  5.0  34  SD, S-on, S-off, SE-spt  Iwasaki, 2010 [83]  Japan  X  3–5  5.0  47  SD, S-on, S-off, SL  Werner, 2008 [32]  Switzerland  X  3–5  5.9  50  SD, TST, S-on, S-off, WASO  Ravid, 2009 [84]  Israel  X  6–8  6.0  92  TST, SE-tib, SL  Souders, 2009 [31]  USA  C-C  6–8  7.1  40  SD, TST, S-off, SE-spt, WASO  Burt, 2013 [85]  Canada  X  6–8  7.7  56  SD, TST, S-on, S-off, SL  Michels, 2013 [86]  Belgium  L-sub  6–8  7.9  165  TST, SE-tib, SL, WASO  Hvolby, 2008 [87]  Denmark  C-C  6–8  8.0  97  TST, SL  Pesonen, 2014 [34]  Finland  L-sub  6–8  8.15  188  SD, SE-tib, SL  Bagheri, 2015 [66]  Teheran  X  6–8  8.3  111  SD, SE-spt  Markovich, 2015 [88]  Canada  X-sub  6–8  8.6  30  SD, TST, SL, WASO  Kelly, 2014a§ [18]  USA  L  6–8  8.68  176  TST, SE-spt, WASO  Keller, 2011§ [17]  USA  L  6–8  8.68  176  S-on, S-off, SL  Wiebe, 2013 [89]  Canada  C-C  6–8  8.7  46  SD, SL, WASO  Alfano, 2015 [35]  USA  C-C  6–8  8.7  36  SD, TST, S-off, SL  Holley, 2011 [90]  UK  X  6–8  8.8  91  SD, TST, SE-tib  Gruber, 2011 [91]  Canada  C-C Exp  6–8  8.8  32  TST, SE-spt  Ashworth, 2013 [92]  UK  C-C  9–11  9.2  51  SD, TST, SE-spt, SL, WASO  Holley, 2014 [93]  UK  C-C  9–11  9.3  50  SD, TST, SE-spt  Gruber, 2000 [25]  Israel  C-C  9–11  9.4  64  SD, TST, S-on, SE-tib  Geiger, 2010|| [94]  Switzerland  X  9–11  9.4  60  SD  Kelly, 2014b [95]  USA  L  9–11  9.4  282  TST, SE-spt  Moreau, 2013 [96]  Canada  C-C  9–11  9.6  41  TST, SE-tib, SL,WASO  Corkum, 2001|| [97]  Canada  C-C  9–11  9.7  25  SD, S-on, S-off, SL  Hvolby, 2011 [98]  Denmark  C-C  9–11  9.8  21  TST, SL  Vriend, 2012 [99]  Canada  X  9–11  9.8  32  SD, S-on, S-off, SE-tib  McNeil, 2015 [27]  Canada  X  9–11  10.0  515  SD, S-off  Fallone, 2002 [100]  US  Exp  9–11  10.2  78  SD, TST  Yamakita, 2014|| [101]  Japan  X-sub  9–11  10.2  42  SD, TST, S-off  Tzischinsky, 2006 [30]  Israel  C-C  9–11  10.2  25  SD, TST, S-on, S-off, SE-spt, SL,WASO  Martoni, 2016 [102]  Switzerland  X  9–11  10.2  115  TST, SE-tib, SL, WASO  Burszstein, 2006 [103]  Israel  C-C  9–11  10.3  108  SD, TST, S-on, S-off  Hjorth, 2012 [29]  Denmark  X  9–11  10.3  62  TST, SE-tib  Barreira, 2015 [28]  Canada  X  9–11  10.3  34  SD, S-on, S-off  Vincent, 2017 [26]  Australia  X  9–11  10.4  65  TST  Gaina, 2004||,¶ [68]  Japan  X  9–11  10.8  41  SD, TST, S-on, S-off, SE-tib, SL, WASO  Allick, 2006|| [104]  Sweden  C-C  9–11  10.9  32  SD, TST, S-on, S-off, SE-tib, SL, WASO  Filardi, 2016 [105]  Italy  C-C  9–11  10.9  21  TST, SE-tib, SL, WASO  Bagley, 2016 [106]  USA  L  9–11  11.3  210  SD, TST, SE-spt  Greenfeld, 2011 [107]  Israel  C-C  9–11  11.4  61  TST, SE-tib, SL  Tremaine, 2010|| [108]  Australia  X  9–11  11.5  54  TST, S-on, WASO  Aroro, 2013|| [109]  UK  X  12–14  12.0  225  SD  Rigney, 2015 [110]  Australia  RCT  12–14  12.2  87  TST, S-off, SE-tib, SL  Marco, 2011|| [111]  USA  L  12–14  12.6  155  SD, TST, S-on, S-off  Amin, 2005 [112]  USA  C-C  12–14  12.6  40  TST, SE-spt, WASO  Beebe, 2007|| [113]  USA  C-C  12–14  12.6  22  SD, S-on, S-off, SE-spt  Guedes, 2016 [114]  Brazil  X-sub  12–14  13.2  37  SD  Auger, 2011 [115]  USA  C-C  12–14  13.7  22  TST, S-on, S-off, SE-tib  Anderson, 2009|| [116]  USA  L-sub  12–14  13.7  236  SD, SE-tib  Beijamini, 2008 [67]  Brazil  X  12–14  13.8  34  SD, S-on, S-off  Gaina, 2004||,¶ [68]  Japan  X  12–14  14.2  42  SD, TST, S-on, S-off, SE-tib, SL, WASO  Tavernier, 2016 [117]  USA  X-sub  12–14  14.37  77  TST, SE-tib, SL, WASO  Chen, 2011 [118]  USA  X  12–14  14.5  44  SD, SL, WASO  Huynh, 2015|| [119]  Canada  C-C  12–14  14.7  20  SD, TST, S-off, SE-tib, SL,WASO  Law, 2012 [120]  USA  L-sub  12–14  14.8  60  TST, SE-tib, WASO  Fobian, 2016|| [121]  USA  X-sub  12–14  14.9  55  SD, S-on, S-off, SE-tib  Carskadon, 1998|| [122]  USA  L  15–18  15.0  32  TST, S-on, S-off  Jiang, 2011 [123]  China  Exp  15–18  15.0  20  SD  Dewald, 2014|| [124]  Netherlands  Expq  15–18  15.1  175  TST, S-on, S-off, SE-tib, SL  Tham, 2015 [125]  USA  C-C  15–18  15.3  50  SD, WASO  Dewald, 2012# [126]  Australia  X  15–18  15.5  236  SD, TST, SE-tib, SL  Baum, 2014 [127]  USA  Exp  15–18  15.5  50  SD, S-on, S-off  Baker, 2013|| [128]  Australia  C-C  15–18  15.5  27  SD, S-on, S-off, SL  Malone, 2016 [129]  USA  X-sub  15–18  15.5  68  SD  Matthews, 2014|| [130]  USA  X  15–18  15.7  250  TST  El-Sheikh, 2016 [131]  USA  L-sub  15–18  15.9  252  SD  Bei, 2013|| [36]  Australia  X  15–18  16.2  146  TST, SE-tib, SL  Park, 2016 [132]  USA  X  15–18  16.4  315  TST  Rodriguez-Colon, 2015 [133]  USA  L-sub  15–18  16.7  322  TST, SE-tib  Astill, 2013 [134]  Netherlands  X  15–18  17.6  24  TST, SE-tib, SL, WASO  Tonetti, 2015 [135]  Italy  X  15–18  18.1  36  TST, S-off, SE-tib  First author, year  Country  Study design  Meta-analytic age group  Mean age  n  Actigraphy variables  Acebo, 2005† [23]  USA  X  3–5  3  21  SD, TST, S-on, S-off, SE-spt  Williams, 2014 [22]  NZ  L  3–5  3.0  216  SD  Goodlin-Jones, 2008‡ [19]  USA  L-nc  3–5  3.4  69  S-on, S-off, SE-tib  Anders, 2011‡ [20]  USA  L-nc  3–5  3.4  69  SL, WASO  Acebo, 2005† [23]  USA  X  3–5  4.0  24  SD, TST, S-on, S-off, SE-spt  Elmore-Staton, 2012 [78]  USA  X  3–5  4.0  29  TST, SE-tib  Lam, 2011 [79]  USA  X  3–5  4.3  59  TST, S-on, S-off  Ishihara, 2014 [33]  Japan  X-sub  3–5  4.8  72  SD, S-on, S-off, SL  Hatzinger, 2012 [80]  Switzerland  C-C  3–5  4.9  35  SD, TST, SE-spt, SL, WASO  Kushnir, 2011 [81]  Israel  C-C  3–5  4.9  30  SD, TST, SE-spt  Acebo, 2005† [23]  USA  X  3–5  5.0  28  SD, TST, S-on, S-off, SE-spt  Cairns, 2014 [82]  USA  L  3–5  5.0  34  SD, S-on, S-off, SE-spt  Iwasaki, 2010 [83]  Japan  X  3–5  5.0  47  SD, S-on, S-off, SL  Werner, 2008 [32]  Switzerland  X  3–5  5.9  50  SD, TST, S-on, S-off, WASO  Ravid, 2009 [84]  Israel  X  6–8  6.0  92  TST, SE-tib, SL  Souders, 2009 [31]  USA  C-C  6–8  7.1  40  SD, TST, S-off, SE-spt, WASO  Burt, 2013 [85]  Canada  X  6–8  7.7  56  SD, TST, S-on, S-off, SL  Michels, 2013 [86]  Belgium  L-sub  6–8  7.9  165  TST, SE-tib, SL, WASO  Hvolby, 2008 [87]  Denmark  C-C  6–8  8.0  97  TST, SL  Pesonen, 2014 [34]  Finland  L-sub  6–8  8.15  188  SD, SE-tib, SL  Bagheri, 2015 [66]  Teheran  X  6–8  8.3  111  SD, SE-spt  Markovich, 2015 [88]  Canada  X-sub  6–8  8.6  30  SD, TST, SL, WASO  Kelly, 2014a§ [18]  USA  L  6–8  8.68  176  TST, SE-spt, WASO  Keller, 2011§ [17]  USA  L  6–8  8.68  176  S-on, S-off, SL  Wiebe, 2013 [89]  Canada  C-C  6–8  8.7  46  SD, SL, WASO  Alfano, 2015 [35]  USA  C-C  6–8  8.7  36  SD, TST, S-off, SL  Holley, 2011 [90]  UK  X  6–8  8.8  91  SD, TST, SE-tib  Gruber, 2011 [91]  Canada  C-C Exp  6–8  8.8  32  TST, SE-spt  Ashworth, 2013 [92]  UK  C-C  9–11  9.2  51  SD, TST, SE-spt, SL, WASO  Holley, 2014 [93]  UK  C-C  9–11  9.3  50  SD, TST, SE-spt  Gruber, 2000 [25]  Israel  C-C  9–11  9.4  64  SD, TST, S-on, SE-tib  Geiger, 2010|| [94]  Switzerland  X  9–11  9.4  60  SD  Kelly, 2014b [95]  USA  L  9–11  9.4  282  TST, SE-spt  Moreau, 2013 [96]  Canada  C-C  9–11  9.6  41  TST, SE-tib, SL,WASO  Corkum, 2001|| [97]  Canada  C-C  9–11  9.7  25  SD, S-on, S-off, SL  Hvolby, 2011 [98]  Denmark  C-C  9–11  9.8  21  TST, SL  Vriend, 2012 [99]  Canada  X  9–11  9.8  32  SD, S-on, S-off, SE-tib  McNeil, 2015 [27]  Canada  X  9–11  10.0  515  SD, S-off  Fallone, 2002 [100]  US  Exp  9–11  10.2  78  SD, TST  Yamakita, 2014|| [101]  Japan  X-sub  9–11  10.2  42  SD, TST, S-off  Tzischinsky, 2006 [30]  Israel  C-C  9–11  10.2  25  SD, TST, S-on, S-off, SE-spt, SL,WASO  Martoni, 2016 [102]  Switzerland  X  9–11  10.2  115  TST, SE-tib, SL, WASO  Burszstein, 2006 [103]  Israel  C-C  9–11  10.3  108  SD, TST, S-on, S-off  Hjorth, 2012 [29]  Denmark  X  9–11  10.3  62  TST, SE-tib  Barreira, 2015 [28]  Canada  X  9–11  10.3  34  SD, S-on, S-off  Vincent, 2017 [26]  Australia  X  9–11  10.4  65  TST  Gaina, 2004||,¶ [68]  Japan  X  9–11  10.8  41  SD, TST, S-on, S-off, SE-tib, SL, WASO  Allick, 2006|| [104]  Sweden  C-C  9–11  10.9  32  SD, TST, S-on, S-off, SE-tib, SL, WASO  Filardi, 2016 [105]  Italy  C-C  9–11  10.9  21  TST, SE-tib, SL, WASO  Bagley, 2016 [106]  USA  L  9–11  11.3  210  SD, TST, SE-spt  Greenfeld, 2011 [107]  Israel  C-C  9–11  11.4  61  TST, SE-tib, SL  Tremaine, 2010|| [108]  Australia  X  9–11  11.5  54  TST, S-on, WASO  Aroro, 2013|| [109]  UK  X  12–14  12.0  225  SD  Rigney, 2015 [110]  Australia  RCT  12–14  12.2  87  TST, S-off, SE-tib, SL  Marco, 2011|| [111]  USA  L  12–14  12.6  155  SD, TST, S-on, S-off  Amin, 2005 [112]  USA  C-C  12–14  12.6  40  TST, SE-spt, WASO  Beebe, 2007|| [113]  USA  C-C  12–14  12.6  22  SD, S-on, S-off, SE-spt  Guedes, 2016 [114]  Brazil  X-sub  12–14  13.2  37  SD  Auger, 2011 [115]  USA  C-C  12–14  13.7  22  TST, S-on, S-off, SE-tib  Anderson, 2009|| [116]  USA  L-sub  12–14  13.7  236  SD, SE-tib  Beijamini, 2008 [67]  Brazil  X  12–14  13.8  34  SD, S-on, S-off  Gaina, 2004||,¶ [68]  Japan  X  12–14  14.2  42  SD, TST, S-on, S-off, SE-tib, SL, WASO  Tavernier, 2016 [117]  USA  X-sub  12–14  14.37  77  TST, SE-tib, SL, WASO  Chen, 2011 [118]  USA  X  12–14  14.5  44  SD, SL, WASO  Huynh, 2015|| [119]  Canada  C-C  12–14  14.7  20  SD, TST, S-off, SE-tib, SL,WASO  Law, 2012 [120]  USA  L-sub  12–14  14.8  60  TST, SE-tib, WASO  Fobian, 2016|| [121]  USA  X-sub  12–14  14.9  55  SD, S-on, S-off, SE-tib  Carskadon, 1998|| [122]  USA  L  15–18  15.0  32  TST, S-on, S-off  Jiang, 2011 [123]  China  Exp  15–18  15.0  20  SD  Dewald, 2014|| [124]  Netherlands  Expq  15–18  15.1  175  TST, S-on, S-off, SE-tib, SL  Tham, 2015 [125]  USA  C-C  15–18  15.3  50  SD, WASO  Dewald, 2012# [126]  Australia  X  15–18  15.5  236  SD, TST, SE-tib, SL  Baum, 2014 [127]  USA  Exp  15–18  15.5  50  SD, S-on, S-off  Baker, 2013|| [128]  Australia  C-C  15–18  15.5  27  SD, S-on, S-off, SL  Malone, 2016 [129]  USA  X-sub  15–18  15.5  68  SD  Matthews, 2014|| [130]  USA  X  15–18  15.7  250  TST  El-Sheikh, 2016 [131]  USA  L-sub  15–18  15.9  252  SD  Bei, 2013|| [36]  Australia  X  15–18  16.2  146  TST, SE-tib, SL  Park, 2016 [132]  USA  X  15–18  16.4  315  TST  Rodriguez-Colon, 2015 [133]  USA  L-sub  15–18  16.7  322  TST, SE-tib  Astill, 2013 [134]  Netherlands  X  15–18  17.6  24  TST, SE-tib, SL, WASO  Tonetti, 2015 [135]  Italy  X  15–18  18.1  36  TST, S-off, SE-tib  Study design: C-C = Case-Comparison/Control (Comparison/control data only used); Exp = Experimental (sleep restriction, extension or both, baseline data only used); Expq = Quasiexperimental; L = Longitudinal (if results of more than one time point represented, first time point only used); X = Cross-sectional; -sub = subsample. Actigraphy variables: SD = Sleep Duration; SE-tib = Sleep Efficiency calculated using TIB as the denominator; SE-spt = Sleep Efficiency calculated using SPT as the denominator; S-on = Sleep Onset; S-off = Sleep Offset; TST = Total Sleep Time; WASO = Waking After Sleep Onset. ||Dataset included weekday or weekend splits suitable for analyses. †,¶Same studies contributing different age groups to the meta-analyses. ‡,§Same cohorts contributing different actigraphy variables to the meta-analyses. #Study included Australian and Netherland participants; Australian cohort only used because of overlap with Dewald, 2014 [126]. View Large Sleep variables Forest plots for meta-analyses (overall data and stratified by age groups from 3 to 18 years) together with heterogeneity statistics are given in Supplementary Figures S1 to S8. Data from infants <3 years of age were not stratified by age group, because of the paucity of data to provide appropriate developmental bands. Overall study heterogeneity and inconsistency were significant and high for all actigraphy-derived variables as shown within the significance of the Q statistic (all p < .0001) and I2 value (all >98 per cent). Thus, all models were analyzed using random effects. Summary statistics for meta-analyses are given in Table 4. Where there were significant age-group effects, meta-regression results are given using 9–11 years age group data as the reference representing the mid-range, and comprising data pooled from the largest number of studies and participants across all sleep variables. Table 4. Mean estimates statistics for sleep variables by age groups and comparisons Actigraphy Variable  n Datasets (Subjects)  Range of study means  Pooled mean estimate (95% CI)  Estimated difference (∆) from Ref age  P  Sleep Duration    Hours  Hours  Hours     3-5 years  10 (557)  9.18 to 10.20  9.68 (9.4, 9.97)  0.83 (0.41, 1.25)  <.001   6-8 years  8 (1598)  8.10 to 9.86  8.98 (8.53, 9.43)  0.13 (-0.32, 0.58)  .573   9-11 years  15 (1367)  7.87 to 9.52  8.85 (8.62, 9.08)  Reference  -   12-14 years  10 (870)  7.23 to 8.75  8.05 (7.72, 8.38)  -0.80 (-1.22, -0.38)  <.001   15-18 years  7 (703)  6.50 to 8.35  7.40 (6.90, 7.90)  -1.45 (-1.92, -0.97)  <.001  TST   3-5 years  8 (276)  7.68 to 9.33  8.64 (7.83, 8.68)  0.55 (0.08, 1.03)  .022   6-8 years  10 (815)  7.36 to 8.50  8.24 (7.83, 8.65)  0.17 (-0.26, 0.61)  .430   9-11 years  19 (1423)  7.30 to 9.30  8.07 (7.88, 8.26)  Reference  -   12-14 years  8 (503)  6.00 to 8.70  7.15 (6.57, 7.74)  -0.90 (-1.38 -0.43)  <.001   15-18 years  9 (1536)  6.40 to 7.53  7.02 (6.77, 7.27)  -1.05 (-1.50, -0.60)  <.001  Sleep Onset    Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     3-5 years  9 (404)  21:06 to 22:12  21:31 (21:18, 21:44)  -0:33 (-0:56, -0:10)  .006   6-8 years  -  -  -  -  -   9-11 years  9 (415)  21:49 to 22:34  22:04 (21:55, 22:13)  Reference  -   12-14 years  6 (275)  22:42 to 23:46  23:09 (22:51, 23:28)  1:05 (0:39, 1:30)  <.001   15-18 years  4 (284)  22:58 to 24:16  23:27 (22:58, 23:56)  1:22 (0:53, 1:52)  <.001  Sleep Offset   3-5 years  9 (404)  6:40 to 7:31  7:07 (6:56, 7:18)  0:09 (-0:14, 0:33)  .411   6-8 years  4 (308)  6:05 to 7:16  6:48 (6:09,7:28)  -0:09 (-0:40, 0:21)  .534   9-11 years  9 (854)  5:58 to 7:28  6:57 (6:42, 7:12)  Reference  -   12-14 years  8 (437)  6:36 to 7:52  7:17 (7:00, 7:36)  0:20 (-0:04, 0:45)  .092   15-18 years  5 (320)  6:55 to 8:00  7:21 (6:58, 7:45)  0:24 (-0:04, 0:52)  .110  Sleep EfficiencyTIB    %  %       3-14 years  23 (1702)  70.3 to 94.9  86.3 (84.4, 88.2)      Sleep EfficiencySPT   3-14 years  16 (1183)  79.2 to 97.0  88.3 (85.9, 90.6)      Sleep Latency    Minutes  Minutes       3-18 years  33 (2420)  3.8 to 45  19.4 (16.6, 22.1)      WASO   3-18 years  28 (975)  7 to 109  55 (43, 68)      Actigraphy Variable  n Datasets (Subjects)  Range of study means  Pooled mean estimate (95% CI)  Estimated difference (∆) from Ref age  P  Sleep Duration    Hours  Hours  Hours     3-5 years  10 (557)  9.18 to 10.20  9.68 (9.4, 9.97)  0.83 (0.41, 1.25)  <.001   6-8 years  8 (1598)  8.10 to 9.86  8.98 (8.53, 9.43)  0.13 (-0.32, 0.58)  .573   9-11 years  15 (1367)  7.87 to 9.52  8.85 (8.62, 9.08)  Reference  -   12-14 years  10 (870)  7.23 to 8.75  8.05 (7.72, 8.38)  -0.80 (-1.22, -0.38)  <.001   15-18 years  7 (703)  6.50 to 8.35  7.40 (6.90, 7.90)  -1.45 (-1.92, -0.97)  <.001  TST   3-5 years  8 (276)  7.68 to 9.33  8.64 (7.83, 8.68)  0.55 (0.08, 1.03)  .022   6-8 years  10 (815)  7.36 to 8.50  8.24 (7.83, 8.65)  0.17 (-0.26, 0.61)  .430   9-11 years  19 (1423)  7.30 to 9.30  8.07 (7.88, 8.26)  Reference  -   12-14 years  8 (503)  6.00 to 8.70  7.15 (6.57, 7.74)  -0.90 (-1.38 -0.43)  <.001   15-18 years  9 (1536)  6.40 to 7.53  7.02 (6.77, 7.27)  -1.05 (-1.50, -0.60)  <.001  Sleep Onset    Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     3-5 years  9 (404)  21:06 to 22:12  21:31 (21:18, 21:44)  -0:33 (-0:56, -0:10)  .006   6-8 years  -  -  -  -  -   9-11 years  9 (415)  21:49 to 22:34  22:04 (21:55, 22:13)  Reference  -   12-14 years  6 (275)  22:42 to 23:46  23:09 (22:51, 23:28)  1:05 (0:39, 1:30)  <.001   15-18 years  4 (284)  22:58 to 24:16  23:27 (22:58, 23:56)  1:22 (0:53, 1:52)  <.001  Sleep Offset   3-5 years  9 (404)  6:40 to 7:31  7:07 (6:56, 7:18)  0:09 (-0:14, 0:33)  .411   6-8 years  4 (308)  6:05 to 7:16  6:48 (6:09,7:28)  -0:09 (-0:40, 0:21)  .534   9-11 years  9 (854)  5:58 to 7:28  6:57 (6:42, 7:12)  Reference  -   12-14 years  8 (437)  6:36 to 7:52  7:17 (7:00, 7:36)  0:20 (-0:04, 0:45)  .092   15-18 years  5 (320)  6:55 to 8:00  7:21 (6:58, 7:45)  0:24 (-0:04, 0:52)  .110  Sleep EfficiencyTIB    %  %       3-14 years  23 (1702)  70.3 to 94.9  86.3 (84.4, 88.2)      Sleep EfficiencySPT   3-14 years  16 (1183)  79.2 to 97.0  88.3 (85.9, 90.6)      Sleep Latency    Minutes  Minutes       3-18 years  33 (2420)  3.8 to 45  19.4 (16.6, 22.1)      WASO   3-18 years  28 (975)  7 to 109  55 (43, 68)      P values in bold represent significant values TIB Sleep efficiency operationalized as: (total sleep time (TST) /time in bed (TIB)) × 100. SPT Sleep efficiency operationalized as: (total sleep time (TST) /sleep period time (SPT) × 100. View Large Table 4. Mean estimates statistics for sleep variables by age groups and comparisons Actigraphy Variable  n Datasets (Subjects)  Range of study means  Pooled mean estimate (95% CI)  Estimated difference (∆) from Ref age  P  Sleep Duration    Hours  Hours  Hours     3-5 years  10 (557)  9.18 to 10.20  9.68 (9.4, 9.97)  0.83 (0.41, 1.25)  <.001   6-8 years  8 (1598)  8.10 to 9.86  8.98 (8.53, 9.43)  0.13 (-0.32, 0.58)  .573   9-11 years  15 (1367)  7.87 to 9.52  8.85 (8.62, 9.08)  Reference  -   12-14 years  10 (870)  7.23 to 8.75  8.05 (7.72, 8.38)  -0.80 (-1.22, -0.38)  <.001   15-18 years  7 (703)  6.50 to 8.35  7.40 (6.90, 7.90)  -1.45 (-1.92, -0.97)  <.001  TST   3-5 years  8 (276)  7.68 to 9.33  8.64 (7.83, 8.68)  0.55 (0.08, 1.03)  .022   6-8 years  10 (815)  7.36 to 8.50  8.24 (7.83, 8.65)  0.17 (-0.26, 0.61)  .430   9-11 years  19 (1423)  7.30 to 9.30  8.07 (7.88, 8.26)  Reference  -   12-14 years  8 (503)  6.00 to 8.70  7.15 (6.57, 7.74)  -0.90 (-1.38 -0.43)  <.001   15-18 years  9 (1536)  6.40 to 7.53  7.02 (6.77, 7.27)  -1.05 (-1.50, -0.60)  <.001  Sleep Onset    Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     3-5 years  9 (404)  21:06 to 22:12  21:31 (21:18, 21:44)  -0:33 (-0:56, -0:10)  .006   6-8 years  -  -  -  -  -   9-11 years  9 (415)  21:49 to 22:34  22:04 (21:55, 22:13)  Reference  -   12-14 years  6 (275)  22:42 to 23:46  23:09 (22:51, 23:28)  1:05 (0:39, 1:30)  <.001   15-18 years  4 (284)  22:58 to 24:16  23:27 (22:58, 23:56)  1:22 (0:53, 1:52)  <.001  Sleep Offset   3-5 years  9 (404)  6:40 to 7:31  7:07 (6:56, 7:18)  0:09 (-0:14, 0:33)  .411   6-8 years  4 (308)  6:05 to 7:16  6:48 (6:09,7:28)  -0:09 (-0:40, 0:21)  .534   9-11 years  9 (854)  5:58 to 7:28  6:57 (6:42, 7:12)  Reference  -   12-14 years  8 (437)  6:36 to 7:52  7:17 (7:00, 7:36)  0:20 (-0:04, 0:45)  .092   15-18 years  5 (320)  6:55 to 8:00  7:21 (6:58, 7:45)  0:24 (-0:04, 0:52)  .110  Sleep EfficiencyTIB    %  %       3-14 years  23 (1702)  70.3 to 94.9  86.3 (84.4, 88.2)      Sleep EfficiencySPT   3-14 years  16 (1183)  79.2 to 97.0  88.3 (85.9, 90.6)      Sleep Latency    Minutes  Minutes       3-18 years  33 (2420)  3.8 to 45  19.4 (16.6, 22.1)      WASO   3-18 years  28 (975)  7 to 109  55 (43, 68)      Actigraphy Variable  n Datasets (Subjects)  Range of study means  Pooled mean estimate (95% CI)  Estimated difference (∆) from Ref age  P  Sleep Duration    Hours  Hours  Hours     3-5 years  10 (557)  9.18 to 10.20  9.68 (9.4, 9.97)  0.83 (0.41, 1.25)  <.001   6-8 years  8 (1598)  8.10 to 9.86  8.98 (8.53, 9.43)  0.13 (-0.32, 0.58)  .573   9-11 years  15 (1367)  7.87 to 9.52  8.85 (8.62, 9.08)  Reference  -   12-14 years  10 (870)  7.23 to 8.75  8.05 (7.72, 8.38)  -0.80 (-1.22, -0.38)  <.001   15-18 years  7 (703)  6.50 to 8.35  7.40 (6.90, 7.90)  -1.45 (-1.92, -0.97)  <.001  TST   3-5 years  8 (276)  7.68 to 9.33  8.64 (7.83, 8.68)  0.55 (0.08, 1.03)  .022   6-8 years  10 (815)  7.36 to 8.50  8.24 (7.83, 8.65)  0.17 (-0.26, 0.61)  .430   9-11 years  19 (1423)  7.30 to 9.30  8.07 (7.88, 8.26)  Reference  -   12-14 years  8 (503)  6.00 to 8.70  7.15 (6.57, 7.74)  -0.90 (-1.38 -0.43)  <.001   15-18 years  9 (1536)  6.40 to 7.53  7.02 (6.77, 7.27)  -1.05 (-1.50, -0.60)  <.001  Sleep Onset    Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     3-5 years  9 (404)  21:06 to 22:12  21:31 (21:18, 21:44)  -0:33 (-0:56, -0:10)  .006   6-8 years  -  -  -  -  -   9-11 years  9 (415)  21:49 to 22:34  22:04 (21:55, 22:13)  Reference  -   12-14 years  6 (275)  22:42 to 23:46  23:09 (22:51, 23:28)  1:05 (0:39, 1:30)  <.001   15-18 years  4 (284)  22:58 to 24:16  23:27 (22:58, 23:56)  1:22 (0:53, 1:52)  <.001  Sleep Offset   3-5 years  9 (404)  6:40 to 7:31  7:07 (6:56, 7:18)  0:09 (-0:14, 0:33)  .411   6-8 years  4 (308)  6:05 to 7:16  6:48 (6:09,7:28)  -0:09 (-0:40, 0:21)  .534   9-11 years  9 (854)  5:58 to 7:28  6:57 (6:42, 7:12)  Reference  -   12-14 years  8 (437)  6:36 to 7:52  7:17 (7:00, 7:36)  0:20 (-0:04, 0:45)  .092   15-18 years  5 (320)  6:55 to 8:00  7:21 (6:58, 7:45)  0:24 (-0:04, 0:52)  .110  Sleep EfficiencyTIB    %  %       3-14 years  23 (1702)  70.3 to 94.9  86.3 (84.4, 88.2)      Sleep EfficiencySPT   3-14 years  16 (1183)  79.2 to 97.0  88.3 (85.9, 90.6)      Sleep Latency    Minutes  Minutes       3-18 years  33 (2420)  3.8 to 45  19.4 (16.6, 22.1)      WASO   3-18 years  28 (975)  7 to 109  55 (43, 68)      P values in bold represent significant values TIB Sleep efficiency operationalized as: (total sleep time (TST) /time in bed (TIB)) × 100. SPT Sleep efficiency operationalized as: (total sleep time (TST) /sleep period time (SPT) × 100. View Large Overnight sleep duration This was the most commonly reported actigraphy variable found in 57 (67 per cent) of all studies (0 to 18 years). Forest plots for the meta-analysis are illustrated in Supplementary Figure S1 and data summarized in Table 4 and stratified by the five age groups suitable for meta-analysis over 3 to 18 years. Age group alone contributed 67.5 per cent (F[4,45] = 25.3; p < .001) of the VAF to overnight sleep duration. The mean estimate for sleep duration at the reference age of 9–11 years (8.85 hr) was similar to the 6–8 year pooled estimate (8.98 hr) but significantly longer than the age 3–5 years age group and significantly shorter than the older age groups (Table 4). Other age group comparisons revealed significantly longer overnight sleep duration amongst younger age groups (3–5 vs 6–8; p = .006), whereas the sleep duration at age 15–18 years was 39 min shorter compared with the 12- to 14-years-old data (p = .013). Sensitivity analyses omitting each of the three studies utilizing waist-worn devices [22, 27, 28] made very little difference to mean estimates (maximum difference; mean estimate 3 min longer in 3–5 years age group). Total sleep time Forest plots for the meta-analysis are illustrated in Supplementary Figure S2 and data summarized in Table 4. Age group contributed 50.6 per cent (F[4,49] = 13.8; p < .001) of the VAF to TST. Mean estimates for TST at the reference age of 9–11 years (8.07 hr) were not significantly different to mean estimates for the 6- to 8-year-old data but shorter than the 3–5 years age group (Table 4). TST at the reference age was however significantly shorter than older age groups (12–14 and 15–18 years). Other age group comparisons revealed no significant differences in TST between age groups 3–5 and 6–8 years (p = .153) or between older age groups (12–14 years vs 15–18 years; p = .596). A sensitivity analysis removing data from the one arm-worn device [26] resulted in a trivial change to the mean estimate (0.6 min shorter for the 9–11 years age group). Sleep onset Fewer studies reported on this sleep variable although data were available across all age categories. There were only two studies in the 6–8 years age group for this variable; therefore, this age group was omitted from the meta-analyses (Table 4, Supplementary Figure S3). Age group alone contributed 81.2 per cent (F[3,24] = 33.5; p < .001) of the VAF to sleep onset. Using 9–11 years as the reference group, the pooled mean estimate for sleep onset at 22:04 was 33 min, later than the 3–5 year age group, and significantly earlier than at older ages (12–14 and 15–17 years; 65 min and 82 min, respectively). There was no significant difference between sleep onset at 12–14 years and sleep onset at 15–18 years (p = .259). Two datasets came from waist-worn devices, both in the 9–11 years age group [27, 28]. Sensitivity analyses removing each in turn resulted in a negligible change to sleep onset (maximum 2 min earlier). Sleep offset Forty-four studies reported sleep offset time across the full age range, and 34 within the age groups for meta-analyses (Supplementary Figure S4 and Table 4). Age group contributed 8.9 per cent (F[4,30] = 1.7; p = .176) of the VAF to sleep offset, still considered a potential source of variability due to p < .20. In the 9–11 years age group, sleep offset occurred at a similar time (6:57 am) to the two younger age groups (Table 4), and although sleep offset was earlier than the older age groups (20 and 24 min for 12–14 and 15–18 year age groups, respectively), the differences were not significant. Further between age group comparisons yielded no significant differences in sleep offset between the 12–14 and 15–18 years old (p = .780). Sensitivity analyses removing each of the waist–worn device data [27, 28] in turn (both 9 and 11 years age group) resulted in a negligible change to sleep onset (maximum 3 min later). Sleep efficiency Most studies reporting on sleep efficiency in this meta-analysis provided operative details; two studies were omitted because this information was not provided and attempts to contact the authors were unsuccessful [30, 31]. Across the full age range, 53 datasets (56 per cent) reported on sleep efficiency: 29 used TIB as the denominator and 24 used SPT. Pooling data for a cross-comparison could only be made between age groups 3 to 14 years because no datasets in the 15–18 year age group reported sleep efficiency using SPT as the denominator (Table 4). By contrast, no datasets in children younger than 3 years used TIB as the denominator, whereas 7 used SPT (although this young age is not included in meta-analysis). Forest plots for sleep efficiency operationalized as a function of TIB and SPT from ages 3–14 are shown in Supplementary Figures S5 and S6, respectively. In meta-analyses, there were no significant age effects for either, and therefore, the pooled mean estimates represent the overall data and are estimated at 86.3 per cent where TIB is used as the denominator and 88.3 per cent when SPT was the denominator for ages 3–14 (Table 4). Sleep latency Thirty-three datasets contributed to the pooled mean estimate for sleep latency from 2420 participants (Table 4 and Supplementary Figure S6). Age group did not contribute to the VAF (p = .261). Sleep latencies across studies ranged from to 3.8 to 45 min. There were no significant differences in the pooled mean estimates for any combination of age groups shown in Supplementary Figure S7, suggesting this variable remains stable across the pediatric age range. The pooled mean estimate for sleep latency across all age groups was 19.4 min (95% CI: 16.6, 22.1). Wake after sleep onset Twenty-eight studies provided data related to WASO. The majority of data were contained to the 9–11 years age group. Age groups contained too few numbers (two age groups with n = 2) to reliably categorize by age. The pooled mean estimate across all ages was 55 min (95% CI: 43, 68) as given in Table 4 and Supplementary Figure S8. Nocturnal wake frequency Studies rarely provided definitions of what determined a “waking bout” to enable the calculation of waking frequency, or were inconsistent; some studies stating at least 5 min of continuous awake were required to calculate a single bout, whereas others required at least 1 min. These inconsistencies precluded any meaningful pooling of data for this variable. Weekday–weekend/nonschool day differences Twenty-two datasets included data split by weekday and weekend or nonschool day, although four were excluded from meta-analyses because of the lack of data in their young age groups [32–35]. All nonschool data were weekend data, excluding one study with data extracted from the start of a school vacation period [36]. The final dataset included 18 studies (asterisked in Table 3) with a mean age range of 9.4 to 16.2 years spanning three age categories (9–11, 12–14, and 15–18 years). Meta-regression analyses revealed no significant weekday–weekend differences in the 9–11 year age group for sleep duration, TST, sleep onset, or sleep offset (Table 5). For the 11–14 years age group, significant weekday–weekend differences were encountered for sleep offset and sleep duration only; children slept 40 min longer during the weekend compared with weekdays and woke 1 h 40 min later. The difference was made up by sleep onset being 52 min later at weekends but this difference did not reach significance (p = .099). Significant weekday–weekend differences for the 15–18 years age group were found across three of the four sleep variables; TST was 56 min (0.93 hr) longer, sleep onset just under 1 hr (57 min) later, and sleep offset almost 2 hr later (1 hr 53 min) at the weekend compared with the weekday (Table 5). Table 5. Weekday–weekend difference: mean estimates by sleep variable and age group   n datasets (subjects)  Ref#  Pooled Mean Estimates  Difference (∆)  P  Weekday (95% CI)  Weekenda (95% CI)  Sleep duration      Hours  Hours  Hours     9-11 years  5 (254)  94,97,101,68,104  8.69 (8.13, 9.26)  8.83 (8.35, 9.30)  0.12 (-0.67, 0.92)  .718   12-14 years  7 (755)  109,111,113,116,68,119,121  7.80 (7.36, 8.24)  8.48 (7.97, 9.00)  0.67 (0, 1.35)  .048   15-18 years  4 (234)  122,126,128,130  7.88 (7.59, 8.16)  8.69 (8.21, 9.18)  0.83 (-0.02, 1.7)  .054  TST   9-11 years  4 (169)  101,68,104,108  8.29 (7.65, 8.94)  8.54 (7.85, 9.24)  0.25 (-0.92, 1.42)  .616   12-14 years  3 (217)  111,68,119  6.60 (5.50, 7.70)  7.19 (6.47, 7.92)  0.57 (-1.27, 2.42)  .414   15-18 years  4 (602)  122,124,130,36  6.66 (6.20, 7.09)  7.58 (7.27, 7.89)  0.93 (0.17, 1.72)  .025  Sleep Onset      Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     9-11 years  4 (152)  97,68,104,108  21:52 (21:38, 22:10)  22:38 (22:00, 23:16)  0:45 (-0:06, 1:37)  .075   12-14 years  3 (219)  111,113,68  23:00 (22:18, 23:42)  23:54 (23:44, 24:05)  0:52 (-0:15, 1:59)  .099   15-18 years  3 (234)  122,124,128  23:00 (22:35, 23:25)  23:57 (23:17, 24:36)  0:57 (0:00, 1:54)  .049  Sleep Offset   9-11 years  5 (194)  97,101,68,104,108  6:47 (6:27, 7:08)  7:18 (6:28, 8:09)  0:30(-0:37, 1:38)  .334   12-14 years  4 (239)  111,113,68,119  6:51 (6:31, 7:12)  8:32 (7:56, 9:09)  1:40(0:53, 2:27)  .002   15-18 years  3 (234)  122,124,128  6:55 (6:16, 7:34)  8:47 (7:52, 9:41)  1:53 (0:31, 3:18)  .018    n datasets (subjects)  Ref#  Pooled Mean Estimates  Difference (∆)  P  Weekday (95% CI)  Weekenda (95% CI)  Sleep duration      Hours  Hours  Hours     9-11 years  5 (254)  94,97,101,68,104  8.69 (8.13, 9.26)  8.83 (8.35, 9.30)  0.12 (-0.67, 0.92)  .718   12-14 years  7 (755)  109,111,113,116,68,119,121  7.80 (7.36, 8.24)  8.48 (7.97, 9.00)  0.67 (0, 1.35)  .048   15-18 years  4 (234)  122,126,128,130  7.88 (7.59, 8.16)  8.69 (8.21, 9.18)  0.83 (-0.02, 1.7)  .054  TST   9-11 years  4 (169)  101,68,104,108  8.29 (7.65, 8.94)  8.54 (7.85, 9.24)  0.25 (-0.92, 1.42)  .616   12-14 years  3 (217)  111,68,119  6.60 (5.50, 7.70)  7.19 (6.47, 7.92)  0.57 (-1.27, 2.42)  .414   15-18 years  4 (602)  122,124,130,36  6.66 (6.20, 7.09)  7.58 (7.27, 7.89)  0.93 (0.17, 1.72)  .025  Sleep Onset      Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     9-11 years  4 (152)  97,68,104,108  21:52 (21:38, 22:10)  22:38 (22:00, 23:16)  0:45 (-0:06, 1:37)  .075   12-14 years  3 (219)  111,113,68  23:00 (22:18, 23:42)  23:54 (23:44, 24:05)  0:52 (-0:15, 1:59)  .099   15-18 years  3 (234)  122,124,128  23:00 (22:35, 23:25)  23:57 (23:17, 24:36)  0:57 (0:00, 1:54)  .049  Sleep Offset   9-11 years  5 (194)  97,101,68,104,108  6:47 (6:27, 7:08)  7:18 (6:28, 8:09)  0:30(-0:37, 1:38)  .334   12-14 years  4 (239)  111,113,68,119  6:51 (6:31, 7:12)  8:32 (7:56, 9:09)  1:40(0:53, 2:27)  .002   15-18 years  3 (234)  122,124,128  6:55 (6:16, 7:34)  8:47 (7:52, 9:41)  1:53 (0:31, 3:18)  .018  P values in bold represent significant values. aAll weekend days excluding one dataset (start of school vacation) [36]. View Large Table 5. Weekday–weekend difference: mean estimates by sleep variable and age group   n datasets (subjects)  Ref#  Pooled Mean Estimates  Difference (∆)  P  Weekday (95% CI)  Weekenda (95% CI)  Sleep duration      Hours  Hours  Hours     9-11 years  5 (254)  94,97,101,68,104  8.69 (8.13, 9.26)  8.83 (8.35, 9.30)  0.12 (-0.67, 0.92)  .718   12-14 years  7 (755)  109,111,113,116,68,119,121  7.80 (7.36, 8.24)  8.48 (7.97, 9.00)  0.67 (0, 1.35)  .048   15-18 years  4 (234)  122,126,128,130  7.88 (7.59, 8.16)  8.69 (8.21, 9.18)  0.83 (-0.02, 1.7)  .054  TST   9-11 years  4 (169)  101,68,104,108  8.29 (7.65, 8.94)  8.54 (7.85, 9.24)  0.25 (-0.92, 1.42)  .616   12-14 years  3 (217)  111,68,119  6.60 (5.50, 7.70)  7.19 (6.47, 7.92)  0.57 (-1.27, 2.42)  .414   15-18 years  4 (602)  122,124,130,36  6.66 (6.20, 7.09)  7.58 (7.27, 7.89)  0.93 (0.17, 1.72)  .025  Sleep Onset      Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     9-11 years  4 (152)  97,68,104,108  21:52 (21:38, 22:10)  22:38 (22:00, 23:16)  0:45 (-0:06, 1:37)  .075   12-14 years  3 (219)  111,113,68  23:00 (22:18, 23:42)  23:54 (23:44, 24:05)  0:52 (-0:15, 1:59)  .099   15-18 years  3 (234)  122,124,128  23:00 (22:35, 23:25)  23:57 (23:17, 24:36)  0:57 (0:00, 1:54)  .049  Sleep Offset   9-11 years  5 (194)  97,101,68,104,108  6:47 (6:27, 7:08)  7:18 (6:28, 8:09)  0:30(-0:37, 1:38)  .334   12-14 years  4 (239)  111,113,68,119  6:51 (6:31, 7:12)  8:32 (7:56, 9:09)  1:40(0:53, 2:27)  .002   15-18 years  3 (234)  122,124,128  6:55 (6:16, 7:34)  8:47 (7:52, 9:41)  1:53 (0:31, 3:18)  .018    n datasets (subjects)  Ref#  Pooled Mean Estimates  Difference (∆)  P  Weekday (95% CI)  Weekenda (95% CI)  Sleep duration      Hours  Hours  Hours     9-11 years  5 (254)  94,97,101,68,104  8.69 (8.13, 9.26)  8.83 (8.35, 9.30)  0.12 (-0.67, 0.92)  .718   12-14 years  7 (755)  109,111,113,116,68,119,121  7.80 (7.36, 8.24)  8.48 (7.97, 9.00)  0.67 (0, 1.35)  .048   15-18 years  4 (234)  122,126,128,130  7.88 (7.59, 8.16)  8.69 (8.21, 9.18)  0.83 (-0.02, 1.7)  .054  TST   9-11 years  4 (169)  101,68,104,108  8.29 (7.65, 8.94)  8.54 (7.85, 9.24)  0.25 (-0.92, 1.42)  .616   12-14 years  3 (217)  111,68,119  6.60 (5.50, 7.70)  7.19 (6.47, 7.92)  0.57 (-1.27, 2.42)  .414   15-18 years  4 (602)  122,124,130,36  6.66 (6.20, 7.09)  7.58 (7.27, 7.89)  0.93 (0.17, 1.72)  .025  Sleep Onset      Clock time (hh:mm)  Clock time (hh:mm)  Time ∆ (hh:mm)     9-11 years  4 (152)  97,68,104,108  21:52 (21:38, 22:10)  22:38 (22:00, 23:16)  0:45 (-0:06, 1:37)  .075   12-14 years  3 (219)  111,113,68  23:00 (22:18, 23:42)  23:54 (23:44, 24:05)  0:52 (-0:15, 1:59)  .099   15-18 years  3 (234)  122,124,128  23:00 (22:35, 23:25)  23:57 (23:17, 24:36)  0:57 (0:00, 1:54)  .049  Sleep Offset   9-11 years  5 (194)  97,101,68,104,108  6:47 (6:27, 7:08)  7:18 (6:28, 8:09)  0:30(-0:37, 1:38)  .334   12-14 years  4 (239)  111,113,68,119  6:51 (6:31, 7:12)  8:32 (7:56, 9:09)  1:40(0:53, 2:27)  .002   15-18 years  3 (234)  122,124,128  6:55 (6:16, 7:34)  8:47 (7:52, 9:41)  1:53 (0:31, 3:18)  .018  P values in bold represent significant values. aAll weekend days excluding one dataset (start of school vacation) [36]. View Large Overall age-related trends Age-related trends were also described using fractional polynomials across the full age range (0–18 years; studies summarized in Tables 2 and 3). These were calculated using the mean of the actigraphy sleep variable, and mean age and weighting (1/SE) for each study to derive the best-fit equation. Figure 2 shows the scatter plots and best-fit lines illustrating developmental patterns: the decline in sleep duration, TST, and sleep efficiency (TIB as denominator) with age, and later sleep onset and sleep offset with increasing age. Only those sleep variables with statistically significant relationships with age are shown in Figure 2. Equations for the best-fit lines as a function of age and correlations for each sleep variable are given as follows:  Overnight Sleep Duration (h)=9.02−1.04 ×[(age10)2−0.83];R2=0.74  Overnight TST Total Sleep Time (h)=8.17−0.66×[(age10)2−0.87];R2=0.60  Sleep Onset (24 h decimal)=21.68+1.04 ×[(age10)3−0.60];R2=0.82  Sleep Offset (24 h decimal)=6.98+0.09 ×[(age10)3−0.62];R2=0.10  Sleep Efficiency (%)=85.8−4.5 ×[(age10)−0.5−0.16];R2=0.12 Figure 2. View largeDownload slide Meta-regression plots of actigraphy sleep variables regressed against age using fractional polynomials. Shaded area represents 95% confidence intervals and bubble size is proportional to the weight of the study. Equations for best-fit estimates are given in text. Figure 2. View largeDownload slide Meta-regression plots of actigraphy sleep variables regressed against age using fractional polynomials. Shaded area represents 95% confidence intervals and bubble size is proportional to the weight of the study. Equations for best-fit estimates are given in text. A significant (p < .001) negative curvilinear relationship between overnight sleep duration and age was found such that overnight sleep duration as shown in Figure 2 declines at the rate of approximately 5.0 min per year across ages 2 and 6, whereas a steeper decline is shown across ages 12 and 16 (approximately 17.5 min per year). TST followed a similar negative curvilinear pattern (p < .001). Across ages 2 and 6, TST declines at the rate of approximately 3.2 min per year, whereas between ages 12 and 16 years, the decline is at a rate of 10.9 min per year. Significant curvilinear associations between sleep onset time and age (p < .001) and between sleep offset time and age (p < .001) were found (Figure 2). Using data points for sleep efficiency across the full 0–18 years age range, sleep efficiency was significantly negatively associated with age in a curvilinear fashion (p = .005) when TIB was used as the denominator in the equation for calculating sleep efficiency. No significant relationship existed when SPT was used (p = .076). Study region as a source of variability Meta-regression analysis weighted by inverse variance was performed to investigate region of origin (Northern America, Europe, Asia, Middle East, South America, and Oceania) as a potential source of heterogeneity for pooled mean estimates from ages 3 to 18 because global differences in sleep according to region have been reported [10, 37–39]. For sleep duration, region of origin was not a source of variability, contributing 2.6 per cent of the VAF (F[5,44] = 0.76; p = .587) to the pooled mean estimate. Similar null findings were found for all other actigraphy variables and controlling for age in all models did not alter the outcome. Discussion This study provides a comprehensive and systematic review of the literature (0–18 years), and meta-analysis of normative data (3–18 years; mean and 95% CIs) for seven key sleep variables commonly used in actigraphy across the pediatric age range: sleep duration, TST, sleep efficiency, sleep latency, sleep onset, sleep offset, and WASO. The results, we believe, will be useful to provide guidance around normal actigraphy values for nighttime sleep measures at any given age. The line equations provide a complementary summary of the age-related trends that may not be evident from age-stratifications alone, with the proviso that these data represent nighttime sleep only, and do not take into account daytime naps contributing to the 24 hr sleep pattern of infants and young children. As far as we are aware this is the only meta-analysis reporting sleep variables derived exclusively from actigraphy. A study published in 2004 analyzed sleep data from age 5 through to adolescence [40] comprised of just five actigraphy studies, and combined actigraphy data with PSG and sleep diaries/logs within the meta-analysis. Since that period, there has been significant growth in studies using actigraphy for sleep assessment. Forty-one studies that had included actigraphy in pediatric populations were identified by Meltzer et al. [15] in 2012, and there are considerably more 5 years on. Our study results confirm the inverse relationship between overnight sleep duration, TST, and age across the pediatric age range reported within other meta-analytic reviews using combined PSG and actigraphy, or parent-report [39, 40]. The pooled estimate for sleep duration in the 3–5 years age group was 9.68 hr, with the caveat that this was nighttime sleep duration and only studies reporting nighttime sleep separate from daytime sleep were included. Across the other age groups, estimates for nighttime sleep duration were 8.98 hr in 6–8 years old, 8.85 hr in 9–11 years old, 8.05 hr in 12–14 years old, and 7.40 hr in the 15–18 years age group. These objective measures show that the hours of sleep older children are getting (operationalized as sleep onset to offset) are below the recommended American Academy of Sleep Medicine (AASM) [41] and National Sleep Foundation (NSF) [42] guidelines of 9–11/12 hr for children aged 6 to 12 years, and 8–10 hr for teenagers. As far as we are aware, although actigraphy data have been used amongst the relevant scientific literature guiding sleep recommendations, the majority of studies examined in the AASM guidelines at least have been from parental report [41]. Given studies comparing actigraphy with parent-reported sleep have consistently found that parents overestimate their child’s sleep [12, 23, 32, 43], it is not surprising that the pooled estimates fall short of the recommendations. This parental overestimation can also be demonstrated between meta-analytic data; calculations using the best-fit equations for overnight sleep duration derived from parent report [39] versus actigraphy data from this study demonstrate that parent reports overestimate overnight sleep duration by 53 and 25 min in 5 and 7 years old, respectively, with a small difference calculated for 10 years old. It is important to note, however, that because outcome measurements such as daytime function have not been examined, it is not possible to infer that these normative values reflect sleep need, nor are they sufficient to guide sleep recommendations. The scientific literature guiding sleep recommendations is based around good or healthy sleep outcomes to promote optimal health in children. Extracted data in this study were not examined in relation to good or healthy outcomes for actigraphy measures. Rather, these data provide valuable information on typical nighttime sleep in normal children and adolescents as measured by actigraphy. We anticipate that these data will help our understanding of changes in actigraphy nighttime sleep variables that occur as children age and guide research utilizing actigraphy in large population studies. Sleep onset data showed typical developmental trends across the age groups with later onset observed with increasing age. Sleep offset across most age groups remained relatively stable across younger age groups with a tendency for later sleep offset in older age groups; 20 to 24 min later in 12–14 and 15–18 years age groups, respectively, compared with the 9–11 years age group confirming self-report data of sleep timing across similar age groups [44]. These differences could reflect later start schedules as adolescents transition into later schooling years or college [45], or a predominance of later sleep offset at weekends since our pooled mean estimates combine both weekday and weekend data. A subgroup of 17 studies (18 datasets) provided weekday–weekend data splits spanned across ages 9.4 to 16.2. No weekday–weekend differences were evident in the 9–11 years age group, but the 12–14 years data supported the emergence of a teenage shift in sleep habits with significantly later sleep offset at weekends, and a nonsignificant tendency towards later sleep onset at weekends. Clear differences were found in older adolescents (15–18 years), where at weekends, both sleep onset and offset times were later and sleep was longer compared with the weekday. The finding was typical for this age group where there is a well-recognised biological shift in sleep preference to go to bed later and to wake much later than their adult and or child counterparts [46, 47]. Sleep efficiency was operationalized in two ways, with some using TIB as the denominator, and others’ SPT. Both definitions are recognized in actigraphy and PSG recordings, although recommendations for consistent operational definitions have recently been made particularly in recognition of the fact that TIB, before sleep onset, does not always include sleep-related activities [48]. For pediatric actigraphy, Meltzer et al. [15] recommend that the denominator in the sleep efficiency equation be TIB as the primary operative, and SPT be used as an alternative. The TIB variable relies on the operator (in most cases, parents) to log the time or use the actigraphy event marker. This disqualifies studies not recording or omitting the TIB measure, for example, studies using fully automated algorithms for analyzing large-scale research studies [49, 50]. Thus, use of SPT in defining sleep efficiency remains a practical alternative but accurate reporting is critical for evaluating and comparing studies. The pooled mean estimates of sleep efficiency using both operative definitions (86.3%–88.3%), and up to age 14 years inclusive, are within the range as recommended by the recent NSF evidence–based recommendations of good sleep quality across the lifespan [51]. However, using continuous data within the full dataset from infancy up to age 18 years, we found a significant negative curvilinear relationship between sleep efficiency and age when the operating definition for sleep efficiency was TIB. No such relationship existed when SPT was used, although no data were available beyond 15 years. The lower values of sleep efficiency in the later teenage years fit with what we know about adolescent sleep architecture. Across adolescence, EEG changes occur as a function of age and pubertal development resulting in less slow wave sleep (N3) and more lighter stages of sleep (N1, N2, and REM sleep) [52]. Within actigraphy, these lighter stages with more motor activity could render more sleep epochs misclassified as wake, providing a potential explanation for the lower sleep efficiency recorded by actigraphy in adolescents’ sleep. The data documented here support stability of sleep latency across the child and adolescent age range described previously in meta-analytic studies by Ohayon et al. [40]. The study by Ohayon et al. [40] collated sleep variable estimates across all ages from age 5 and found that sleep latency increased modestly but significantly across the lifespan, but apparent only when data from young adults were compared with older adults. The recent NSF sleep–quality recommendations include a sleep latency of ≤15 min as an appropriate measure for indexing good sleep quality across all pediatric age groups [51]. Sleep latencies between 16 and 30 min were judged to be good sleep quality, and latencies 45–60 min indicative of poor sleep quality. In the current study, our pooled mean estimate of 19.4 min from 3 to 18 years and collated from studies of healthy individuals fits within these recommendations. Although the region of origin is considered a likely source of data variability given several studies have reported regional differences in a variety of sleep parameters [10, 37–39], our data did not support this. A meta-analysis of subjectively reported sleep duration in adolescents showed a tendency towards shorter sleep duration in studies from Northern American countries compared with those of European origin, and Asian adolescents’ bedtimes were later than peers from North America and Europe resulting in less TST on school nights [10]. Notably, few cohorts from Asia contributed to our dataset (5/85, 6 per cent) and the majority were from Japan. This contrasts with the meta-analysis of subjectively reported data where Asian studies represented 20 per cent (7/34) of the dataset and included samples from five Asian countries [39]. In that meta-analysis, significantly shorter sleep duration was found in studies originating from Asian as opposed to Caucasian/non-Asian countries, the reasons for which remain unknown. Thus, inclusion of more samples from Asia in this meta-analysis may have changed pooled estimates of sleep. The actigraphy variable of WASO generated variable data and the frequency of waking data could not be reliably pooled due to a lack of consistency in how waking bouts were defined. From earliest records [53], WASO is one of the least accurate of actigraphy measures with a wakefulness detection specificity of less than 60 per cent in more than half of pediatric validation studies [15]. This translates to errors across many sleep–wake parameters. Although 7 day actigraphy protocols are recommended to obtain at least five nights of useable data [54], not all achieved this. A major challenge, particularly in younger age groups, is the need for parents to concurrently complete daily diaries. These data are needed to document bed times and rise times as well as artifact (nonwear time, periods of inactivity), in order to assist scoring programs. Not only is this challenging to comply with [55], separate sleep places commonly practised within western societies make this a difficult task. There may also be unique challenges to the reliability of actigraphy measurement in respect of gender. In particular, boys of younger and adolescent age groups have been found to move more during sleep than girls [56, 57], leading to inflated WASO (relative to PSG measurements) [58]. Although all age ranges and both genders were well represented in the present review, not enough studies reported gender separately to evaluate gender differences. Although small gender differences in sleep patterns have been reported in adolescents through both subjective and objective assessments [57, 59, 60], the evidence for clear gender-based differences in sleep patterns in younger school-aged children prior to the onset of puberty is less consistent. Some survey study results suggest that girls may need more sleep than boys [61, 62], or that gender sleep patterns may differ only on nonschool nights [63]. Studies considered or included in this meta-analysis that addressed gender reported no differences in actigraphy-derived sleep variables of TST, sleep efficiency, and sleep latency in children aged 5–12 years [64], nor in sleep duration (which includes wake epochs) in 8–10 years old [65]. By contrast, one study found significant gender differences in sleep duration and sleep efficiency in 6–10 years old (boys had 15 min longer sleep duration, but poorer sleep efficiency; 82 per cent boys vs 85 per cent girls) [66]. Another study in adolescents aged 13 to 16 years found that there was no effect of gender on sleep onset, sleep offset, or sleep duration [67]. However, the most comprehensive actigraphy study addressing gender conducted by Gaina et al. [68] in 13- to 14-years-old Japanese children found several differences including lower sleep efficiency in boys on both weekdays and the weekend, and shorter TST on weekends. Further objective studies are required in this area to truly understand gender differences in sleep patterns and potential underlying factors as currently the literature appears contradictory. Socioeconomic status and race/ethnicity would be worthwhile to investigate in future studies as others factors that moderate the relationship between sleep and age. Furthermore, as the adolescent age range crosses pubertal development, tighter age bands are recommended to further clarify gender differences. There were inherent limitations in this systematic review and meta-analysis. Although many studies of healthy individuals and across several different age ranges and cultures were included, interpretation of the mean estimates is limited by the failure to be able to control for important confounders such as ethnicity, sex, socioeconomic status, and family factors. The data must be interpreted as applicable to nighttime sleep only. Twenty-four hour sleep–wake actigraphy recordings are rare; thus, the mean estimates may underestimate daily sleep—particularly in infants and younger children with considerable daytime napping. Most studies included were nonrandomized and consequently stratification by key confounders was inconsistent. Also, there was substantial study heterogeneity, but it was not possible to investigate other major potential sources of variability beyond age and region. Within the technology itself, different instruments, sensitivity settings, device placement, and scoring algorithms used to derive actigraphy data would also contribute to study heterogeneity. For waist-worn devices, differences with wrist actigraphy have been noted with variables incorporating waking indices [29], and for that reason, those data were not extracted. Furthermore, sensitivity analyses made very little difference to any of the mean estimates incorporating waist- or arm-worn devices. Other limitations include the strategy used to combine two reported subgroups into a single group (e.g. studies that only provided weekday and weekend data separately, or male and female data separately) using the pooled standard deviation. Although this strategy is recommended, it does provide a slight underestimate of the desired standard deviation [16]. Furthermore, although calculating a pooled mean estimate for each sleep variable by age group was a practical way to estimate differences, it did come with potential limitations due to the small number of studies in some age groups, and the random-effects analysis, where our estimate of the error may itself have been unreliable. Related to this, we were unable to include infants aged 0–2 years in the meta-analysis due to the limited number of studies and the dramatic and rapid changes in sleep over this developmental period. For the majority of studies, actigraphy nighttime sleep estimates were collected at one point in time. Few longitudinal studies were extracted from the literature search, and data from only one point of time could be included in the meta-analysis to minimize bias. A qualitative assessment of the studies included in this meta-analysis was not deemed possible given the heterogeneity of study designs. Data were extracted specific for the objectives of the meta-analysis and were not always the objectives of the original studies. Finally, some of the early age–category datasets and weekday–weekend or nonschool days pooled data from just three studies. Although this is acceptable within meta-analytic techniques, four studies are recommended before a meta-analysis for an outcome in a systematic review settles to reach 10 per cent of the final estimate [69]. However, our inclusion of the curvilinear fits and bubble plots provide a way of overcoming the limitations of the age stratifications in the meta-analysis. We identified other shortcomings that need addressing in the literature. In particular, more studies are required within the younger age groups where the data up to age 3 were scarce, and from more regions across the globe, particularly Asia. More prospective large-scale longitudinal studies are required to provide richer sources of data and may represent a better estimate to reduce heterogeneity evident within this meta-analysis. We recommend that future studies report weekend and weekday data, as well as the combined data. Similarly, studies reporting data from males and females should report the data separately and combined to aid future meta-analyses and to better understand the discrepancies related to these factors. Although all studies required at least one sleep variable of interest to be included in this review, the majority failed to report all key sleep actigraphy variables, highlighting a further area of improvement in reporting standards to include all key variables in research reporting, similar to recommendations for individual actigraphy reports [70]. In conclusion, to the best of our knowledge, this is the largest meta-analysis to establish actigraphy-derived normative values for overnight sleep parameters across the pediatric age range (3–18 years), and the first to include children younger than 5 years of age. Results can be generalized to healthy pediatric populations to assist the interpretation of actigraphy measures in recordings from individuals, and be useful for research purposes. The review has also defined the extent of the research and highlighted gaps in the literature and areas for further improvement to enable a better understanding of sleep across development, and across different cultures and ethnicities. Finally, it is well known that the pediatric actigraphy field lacks established rules and standards for sleep measurement [15]. The variability in data encountered in this study highlights a need for that to occur. Supplementary Material Supplementary material is available at SLEEP online. Notes Conflict of interest statement. None declared. References 1. Rechtschaeffen A et al.   A Manual of Standardized Terminology, Techniques, and Scoring Systems for Sleep Stages of Human Subjects . Washington: NIH Publ. 204, US Government Printing Office; 1968. 2. Berry RB et al.  ; American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications, Version 2.2 . Darien, IL: American Academy of Sleep Medicine; 2015. 3. Meltzer LJ et al.   A comparison of actigraphy scoring rules used in pediatric research. Sleep Med . 2011; 12( 8): 793– 796. Google Scholar CrossRef Search ADS PubMed  4. Ancoli-Israel S et al.   The role of actigraphy in the study of sleep and circadian rhythms. Sleep . 2003; 26( 3): 342– 392. Google Scholar CrossRef Search ADS PubMed  5. Morgenthaler T et al.  ; Standards of Practice Committee; American Academy of Sleep Medicine. Practice parameters for the use of actigraphy in the assessment of sleep and sleep disorders: an update for 2007. Sleep . 2007; 30( 4): 519– 529. Google Scholar CrossRef Search ADS PubMed  6. Scholle S et al.   First night effect in children and adolescents undergoing polysomnography for sleep-disordered breathing. Clin Neurophysiol . 2003; 114( 11): 2138– 2145. Google Scholar CrossRef Search ADS PubMed  7. Galland B et al.   Challenges and emerging technologies within the field of pediatric actigraphy. Front Psychiatry . 2014; 5: 99. Google Scholar CrossRef Search ADS PubMed  8. Davis KF et al.   Sleep in infants and young children: part one: normal sleep. J Pediatr Health Care . 2004; 18( 2): 65– 71. Google Scholar CrossRef Search ADS PubMed  9. Iglowstein I et al.   Sleep duration from infancy to adolescence: reference values and generational trends. Pediatrics . 2003; 111( 2): 302– 307. Google Scholar CrossRef Search ADS PubMed  10. Gradisar M et al.   Recent worldwide sleep patterns and problems during adolescence: a review and meta-analysis of age, region, and sleep. Sleep Med . 2011; 12( 2): 110– 118. Google Scholar CrossRef Search ADS PubMed  11. Carskadon MA et al.   Association between puberty and delayed phase preference. Sleep . 1993; 16( 3): 258– 262. Google Scholar CrossRef Search ADS PubMed  12. Short MA et al.   Estimating adolescent sleep patterns: parent reports versus adolescent self-report surveys, sleep diaries, and actigraphy. Nat Sci Sleep . 2013; 5: 23– 26. Google Scholar CrossRef Search ADS PubMed  13. Short MA et al.   Time for bed: parent-set bedtimes associated with improved sleep and daytime functioning in adolescents. Sleep . 2011; 34( 6): 797– 800. Google Scholar CrossRef Search ADS PubMed  14. Holley S et al.   A comparison of actigraphy and parental report of sleep habits in typically developing children aged 6 to 11 years. Behav Sleep Med . 2010; 8( 1): 16– 27. Google Scholar CrossRef Search ADS PubMed  15. Meltzer LJ et al.   Use of actigraphy for assessment in pediatric sleep research. Sleep Med Rev . 2012; 16( 5): 463– 475. Google Scholar CrossRef Search ADS PubMed  16. Higgins JPT, Green S eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 . [updated March 2011]. The Cochrane Collaboration, 2011. http://handbook.cochrane.org/chapter_7/table_7_7_a_formulae_for_combining_groups.htm. Accessed 17 June, 2017. 17. Keller P et al.   Children’s emotional security and sleep: longitudinal relations and directions of effects. J Child Psychol Psychiatry . 2011; 52( 1): 64– 71. Google Scholar CrossRef Search ADS PubMed  18. Kelly RJ et al.   Reciprocal relations between children’s sleep and their adjustment over time. Dev Psychol . 2014; 50( 4): 1137– 1147. Google Scholar CrossRef Search ADS PubMed  19. Goodlin-Jones BL et al.   Sleep patterns in preschool-age children with autism, developmental delay, and typical development. J Am Acad Child Adolesc Psychiatry . 2008; 47( 8): 930– 938. Google Scholar CrossRef Search ADS PubMed  20. Anders TF et al.   Six-month sleep-wake organization and stability in preschool-age children with autism, developmental delay, and typical development. Behav Sleep Med . 2011; 9( 2): 92– 106. Google Scholar CrossRef Search ADS PubMed  21. Higgins JP et al.   Measuring inconsistency in meta-analyses. BMJ . 2003; 327( 7414): 557– 560. Google Scholar CrossRef Search ADS PubMed  22. Williams SM et al.   Do more active children sleep more? A repeated cross-sectional analysis using accelerometry. PLoS One . 2014; 9( 4): e93117. Google Scholar CrossRef Search ADS PubMed  23. Acebo C et al.   Sleep/wake patterns derived from activity monitoring and maternal report for healthy 1- to 5-year-old children. Sleep . 2005; 28( 12): 1568– 1577. Google Scholar CrossRef Search ADS PubMed  24. Newland RP et al.   Goodness of fit between prenatal maternal sleep and infant sleep: associations with maternal depression and attachment security. Infant Behav Dev . 2016; 44: 179– 188. Google Scholar CrossRef Search ADS PubMed  25. Gruber R et al.   Instability of sleep patterns in children with attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry . 2000; 39( 4): 495– 501. Google Scholar CrossRef Search ADS PubMed  26. Vincent GE et al.   Temporal and bidirectional associations between physical activity and sleep in primary school-aged children. Appl Physiol Nutr Metab . 2017; 42( 3): 238– 242. Google Scholar CrossRef Search ADS PubMed  27. Mcneil J et al.   Objectively-measured sleep and its association with adiposity and physical activity in a sample of Canadian children. J Sleep Res . 2015; 24( 2): 131– 139. Google Scholar CrossRef Search ADS PubMed  28. Barreira TV et al.   Identifying children’s nocturnal sleep using 24-h waist accelerometry. Med Sci Sports Exerc . 2015; 47( 5): 937– 943. Google Scholar CrossRef Search ADS PubMed  29. Hjorth MF et al.   Measure of sleep and physical activity by a single accelerometer: can a waist-worn Actigraph adequately measure sleep in children? Sleep Biol Rhythms . 2012; 10( 4): 328– 335. Google Scholar CrossRef Search ADS   30. Tzischinsky O et al.   Sleep-wake cycles in obese children with and without binge-eating episodes. J Paediatr Child Health . 2006; 42( 11): 688– 693. Google Scholar CrossRef Search ADS PubMed  31. Souders MC et al.   Sleep behaviors and sleep quality in children with autism spectrum disorders. Sleep . 2009; 32( 12): 1566– 1578. Google Scholar CrossRef Search ADS PubMed  32. Werner H et al.   Agreement rates between actigraphy, diary, and questionnaire for children’s sleep patterns. Arch Pediatr Adolesc Med . 2008; 162( 4): 350– 358. Google Scholar CrossRef Search ADS PubMed  33. Ishihara K et al.   The reliability and validity of the Japanese version of the Children’s ChronoType Questionnaire (CCTQ) in preschool children. Chronobiol Int . 2014; 31( 9): 947– 953. Google Scholar CrossRef Search ADS PubMed  34. Pesonen AK et al.   Continuity and change in poor sleep from childhood to early adolescence. Sleep . 2014; 37( 2): 289– 297. Google Scholar PubMed  35. Alfano CA et al.   Subjective-objective sleep comparisons and discrepancies among clinically-anxious and healthy children. J Abnorm Child Psychol . 2015; 43( 7): 1343– 1353. Google Scholar CrossRef Search ADS PubMed  36. Bei B et al.   Actigraphy-assessed sleep during school and vacation periods: a naturalistic study of restricted and extended sleep opportunities in adolescents. J Sleep Res . 2014; 23( 1): 107– 117. Google Scholar CrossRef Search ADS PubMed  37. Mindell JA et al.   Cross-cultural differences in infant and toddler sleep. Sleep Med . 2010; 11( 3): 274– 280. Google Scholar CrossRef Search ADS PubMed  38. Mindell JA et al.   Cross-cultural differences in the sleep of preschool children. Sleep Med . 2013; 14( 12): 1283– 1289. Google Scholar CrossRef Search ADS PubMed  39. Galland BC et al.   Normal sleep patterns in infants and children: a systematic review of observational studies. Sleep Med Rev . 2012; 16( 3): 213– 222. Google Scholar CrossRef Search ADS PubMed  40. Ohayon MM et al.   Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. Sleep . 2004; 27( 7): 1255– 1273. Google Scholar CrossRef Search ADS PubMed  41. Paruthi S et al.   Consensus statement of the American Academy of sleep medicine on the recommended amount of sleep for healthy children: methodology and discussion. J Clin Sleep Med . 2016; 12( 11): 1549– 1561. Google Scholar CrossRef Search ADS PubMed  42. Hirshkowitz M et al.   National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health . 2015; 1( 1): 40– 43. Google Scholar CrossRef Search ADS PubMed  43. Nelson TD et al.   Estimating child sleep from parent report of time in bed: development and evaluation of adjustment approaches. J Pediatr Psychol . 2014; 39( 6): 624– 632. Google Scholar CrossRef Search ADS PubMed  44. Crowley SJ et al.   A longitudinal assessment of sleep timing, circadian phase, and phase angle of entrainment across human adolescence. PLoS One . 2014; 9( 11): e112199. Google Scholar CrossRef Search ADS PubMed  45. Wolfson AR et al.   Middle school start times: the importance of a good night’s sleep for young adolescents. Behav Sleep Med . 2007; 5( 3): 194– 209. Google Scholar CrossRef Search ADS PubMed  46. Crowley SJ et al.   Estimating dim light melatonin onset (DLMO) phase in adolescents using summer or school-year sleep/wake schedules. Sleep . 2006; 29( 12): 1632– 1641. Google Scholar CrossRef Search ADS PubMed  47. Taylor DJ et al.   Sleep tendency during extended wakefulness: insights into adolescent sleep regulation and behavior. J Sleep Res . 2005; 14( 3): 239– 244. Google Scholar CrossRef Search ADS PubMed  48. Reed DL et al.   Measuring sleep efficiency: what should the denominator be? J Clin Sleep Med . 2016; 12( 2): 263– 266. Google Scholar CrossRef Search ADS PubMed  49. Tudor-Locke C et al.   Fully automated waist-worn accelerometer algorithm for detecting children’s sleep-period time separate from 24-h physical activity or sedentary behaviors. Appl Physiol Nutr Metab . 2014; 39( 1): 53– 57. Google Scholar CrossRef Search ADS PubMed  50. Taylor BJ et al.   Targeting sleep, food, and activity in infants for obesity prevention: an RCT. Pediatrics . 2017; 139( 3): e20162037. Google Scholar CrossRef Search ADS PubMed  51. Ohayon M et al.   National Sleep Foundation’s sleep quality recommendations: first report. Sleep Health . 2017; 3( 1): 6– 19. Google Scholar CrossRef Search ADS PubMed  52. Baker FC et al.   Age-Related differences in sleep architecture and electroencephalogram in adolescents in the National Consortium on alcohol and neurodevelopment in adolescence sample. Sleep . 2016; 39( 7): 1429– 1439. Google Scholar CrossRef Search ADS PubMed  53. Mullaney DJ et al.   Wrist-actigraphic estimation of sleep time. Sleep . 1980; 3( 1): 83– 92. Google Scholar CrossRef Search ADS PubMed  54. Acebo C et al.   Estimating sleep patterns with activity monitoring in children and adolescents: how many nights are necessary for reliable measures? Sleep . 1999; 22( 1): 95– 103. Google Scholar CrossRef Search ADS PubMed  55. Sadeh A. Commentary: comparing actigraphy and parental report as measures of children’s sleep. J Pediatr Psychol . 2008; 33( 4): 406– 407. Google Scholar CrossRef Search ADS PubMed  56. Sadeh A et al.   Sleep patterns and sleep disruptions in school-age children. Dev Psychol . 2000; 36( 3): 291– 301. Google Scholar CrossRef Search ADS PubMed  57. Gaina A et al.   Gender and temporal differences in sleep-wake patterns in Japanese schoolchildren. Sleep . 2005; 28( 3): 337– 342. Google Scholar PubMed  58. Johnson NL et al.   Sleep estimation using wrist actigraphy in adolescents with and without sleep disordered breathing: a comparison of three data modes. Sleep . 2007; 30( 7): 899– 905. Google Scholar CrossRef Search ADS PubMed  59. Galland BC et al.   Gender differences in sleep hygiene practices and sleep quality in New Zealand adolescents aged 15 to 17 years. Sleep Health . 2017; 3( 2): 77– 83. Google Scholar CrossRef Search ADS PubMed  60. Johnson EO et al.   Epidemiology of DSM-IV insomnia in adolescence: lifetime prevalence, chronicity, and an emergent gender difference. Pediatrics . 2006; 117( 2): e247– e256. Google Scholar CrossRef Search ADS PubMed  61. Tonetti L et al.   Sex difference in sleep-time preference and sleep need: a cross-sectional survey among Italian pre-adolescents, adolescents, and adults. Chronobiol Int . 2008; 25( 5): 745– 759. Google Scholar CrossRef Search ADS PubMed  62. Ekstedt M et al.   Sleep, physical activity and BMI in six to ten-year-old children measured by accelerometry: a cross-sectional study. Int J Behav Nutr Phys Act . 2013; 10: 82. Google Scholar CrossRef Search ADS PubMed  63. Biggs SN et al.   Gender, socioeconomic, and ethnic differences in sleep patterns in school-aged children. Sleep Med . 2013; 14( 12): 1304– 1309. Google Scholar CrossRef Search ADS PubMed  64. Aronen ET et al.   Associations of age and gender with activity and sleep. Acta Paediatr . 2001; 90( 2): 222– 224. Google Scholar CrossRef Search ADS PubMed  65. Chaput JP et al.   Physical activity vs. sedentary time: independent associations with adiposity in children. Pediatr Obes . 2012; 7( 3): 251– 258. Google Scholar CrossRef Search ADS PubMed  66. Bagheri M et al.   Serum ghrelin levels and gender-related indices of body composition in prepubertal children: a cross-sectional study. Eur J Nutr . 2015; 54( 2): 283– 290. Google Scholar CrossRef Search ADS PubMed  67. Beijamini F et al.   Influence of gender on psychomotor vigilance task performance by adolescents. Braz J Med Biol Res . 2008; 41( 8): 734– 738. Google Scholar CrossRef Search ADS PubMed  68. Gaina A et al.   Sleep parameters recorded by Actiwatch in elementary school children and junior high school adolescents: schooldays vs weekends. Sleep Hypn . 2004; 6( 2): 66– 77. 69. Herbison P et al.   Meta-analyses of small numbers of trials often agree with longer-term results. J Clin Epidemiol . 2011; 64( 2): 145– 153. Google Scholar CrossRef Search ADS PubMed  70. Ancoli-Israel S et al.   The SBSM guide to actigraphy monitoring: clinical and research applications. Behav Sleep Med . 2015; 13: S4– S38. Google Scholar CrossRef Search ADS PubMed  71. Volkovich E et al.   Sleep patterns of co-sleeping and solitary sleeping infants and mothers: a longitudinal study. Sleep Med . 2015; 16( 11): 1305– 1312. Google Scholar CrossRef Search ADS PubMed  72. Galland BC et al.   Anticipatory guidance to prevent infant sleep problems within a randomised controlled trial: infant, maternal and partner outcomes at 6 months of age. BMJ Open . 2017; 7( 5): e014908. Google Scholar CrossRef Search ADS PubMed  73. Montgomery-Downs HE et al.   Toddler behavior following polysomnography: effects of unintended sleep disturbance. Sleep . 2006; 29( 10): 1282– 1287. Google Scholar CrossRef Search ADS PubMed  74. Sadeh A et al.   Infant sleep and parental sleep-related cognitions. J Fam Psychol . 2007; 21( 1): 74– 87. Google Scholar CrossRef Search ADS PubMed  75. Kohyama J. Early rising children are more active than late risers. Neuropsychiatr Dis Treat . 2007; 3( 6): 959– 963. Google Scholar PubMed  76. Molfese VJ et al.   Relations between toddler sleep characteristics, sleep problems, and temperament. Dev Neuropsychol . 2015; 40( 3): 138– 154. Google Scholar CrossRef Search ADS PubMed  77. LeBourgeois MK et al.   Circadian phase and its relationship to nighttime sleep in toddlers. J Biol Rhythms . 2013; 28( 5): 322– 331. Google Scholar CrossRef Search ADS PubMed  78. Elmore-Staton L et al.   Preschoolers’ daytime respiratory sinus arrhythmia and nighttime sleep. Physiol Behav . 2012; 107( 3): 414– 417. Google Scholar CrossRef Search ADS PubMed  79. Lam JC et al.   Defining the roles of actigraphy and parent logs for assessing sleep variables in preschool children. Behav Sleep Med . 2011; 9( 3): 184– 193. Google Scholar CrossRef Search ADS PubMed  80. Hatzinger M et al.   Pre-schoolers suffering from psychiatric disorders show increased cortisol secretion and poor sleep compared to healthy controls. J Psychiatr Res . 2012; 46( 5): 590– 599. Google Scholar CrossRef Search ADS PubMed  81. Kushnir J et al.   Sleep of preschool children with night-time fears. Sleep Med . 2011; 12( 9): 870– 874. Google Scholar CrossRef Search ADS PubMed  82. Cairns A et al.   Changes in sleep duration, timing, and quality as children transition to kindergarten. Behav Sleep Med . 2014; 12( 6): 507– 516. Google Scholar CrossRef Search ADS PubMed  83. Iwasaki M et al.   Utility of subjective sleep assessment tools for healthy preschool children: a comparative study between sleep logs, questionnaires, and actigraphy. J Epidemiol . 2010; 20( 2): 143– 149. Google Scholar CrossRef Search ADS PubMed  84. Ravid S et al.   Sleep disturbances are associated with reduced school achievements in first-grade pupils. Dev Neuropsychol . 2009; 34( 5): 574– 587. Google Scholar CrossRef Search ADS PubMed  85. Burt J et al.   Sleep and eating in childhood: a potential behavioral mechanism underlying the relationship between poor sleep and obesity. Sleep Med . 2014; 15( 1): 71– 75. Google Scholar CrossRef Search ADS PubMed  86. Michels N et al.   Children’s sleep and autonomic function: low sleep quality has an impact on heart rate variability. Sleep . 2013; 36( 12): 1939– 1946. Google Scholar CrossRef Search ADS PubMed  87. Hvolby A et al.   Actigraphic and parental reports of sleep difficulties in children with attention-deficit/hyperactivity disorder. Arch Pediatr Adolesc Med . 2008; 162( 4): 323– 329. Google Scholar CrossRef Search ADS PubMed  88. Markovich AN et al.   Validating the children’s sleep habits questionnaire against polysomnography and actigraphy in school-aged children. Front Psychiatry . 2014; 5: 188. Google Scholar PubMed  89. Wiebe S et al.   Sleep and sleepiness in children with attention deficit / hyperactivity disorder and controls. J Sleep Res . 2013; 22( 1): 41– 49. Google Scholar CrossRef Search ADS PubMed  90. Holley S et al.   An hour less sleep is a risk factor for childhood conduct problems. Child Care Health Dev . 2011; 37( 4): 563– 570. Google Scholar CrossRef Search ADS PubMed  91. Gruber R et al.   Impact of sleep restriction on neurobehavioral functioning of children with attention deficit hyperactivity disorder. Sleep . 2011; 34( 3): 315– 323. Google Scholar CrossRef Search ADS PubMed  92. Ashworth A et al.   Cross syndrome comparison of sleep problems in children with Down syndrome and Williams syndrome. Res Dev Disabil . 2013; 34( 5): 1572– 1580. Google Scholar CrossRef Search ADS PubMed  93. Holley S et al.   Executive function and sleep problems in childhood epilepsy. Epilepsy Behav . 2014; 37: 20– 25. Google Scholar CrossRef Search ADS PubMed  94. Geiger A et al.   Association between sleep duration and intelligence scores in healthy children. Dev Psychol . 2010; 46( 4): 949– 954. Google Scholar CrossRef Search ADS PubMed  95. Kelly RJ et al.   Longitudinal relations between parent-child conflict and children’s adjustment: the role of children’s sleep. J Abnorm Child Psychol . 2014; 42( 7): 1175– 1185. Google Scholar CrossRef Search ADS PubMed  96. Moreau V et al.   Sleep of children with attention deficit hyperactivity disorder: actigraphic and parental reports. Behav Sleep Med . 2014; 12( 1): 69– 83. Google Scholar CrossRef Search ADS PubMed  97. Corkum P et al.   Actigraphy and parental ratings of sleep in children with attention-deficit/hyperactivity disorder (ADHD). Sleep . 2001; 24( 3): 303– 312. Google Scholar CrossRef Search ADS PubMed  98. Hvolby A et al.   Use of Ball Blanket in attention-deficit/hyperactivity disorder sleeping problems. Nord J Psychiatry . 2011; 65( 2): 89– 94. Google Scholar CrossRef Search ADS PubMed  99. Vriend JL et al.   Sleep quantity and quality in relation to daytime functioning in children. Child Health Care . 2012; 41( 3): 204– 222. Google Scholar CrossRef Search ADS   100. Fallone G et al.   How well do school-aged children comply with imposed sleep schedules at home? Sleep . 2002; 25( 7): 739– 745. Google Scholar CrossRef Search ADS PubMed  101. Yamakita M et al.   Availability of a simple self-report sleep questionnaire for 9- to 12-year-old children. Sleep Biol Rhythms . 2014; 12( 4): 279– 288. Google Scholar CrossRef Search ADS   102. Martoni M et al.   24-h actigraphic monitoring of motor activity, sleeping and eating behaviors in underweight, normal weight, overweight and obese children. Eat Weight Disord . 2016; 21( 4): 669– 677. Google Scholar CrossRef Search ADS PubMed  103. Bursztein C et al.   Sleep, sleepiness, and behavior problems in children with headache. J Child Neurol . 2006; 21( 12): 1012– 1019. Google Scholar CrossRef Search ADS PubMed  104. Allik H et al.   Sleep patterns of school-age children with Asperger syndrome or high-functioning autism. J Autism Dev Disord . 2006; 36( 5): 585– 595. Google Scholar CrossRef Search ADS PubMed  105. Filardi M et al.   Circadian rest-activity rhythm in pediatric type 1 narcolepsy. Sleep . 2016; 39( 6): 1241– 1247. Google Scholar CrossRef Search ADS PubMed  106. Bagley EJ et al.   Neighborhood economic deprivation and social fragmentation: associations with children’s sleep. Behav Sleep Med . 2016;Dec 9:1–13. 107. Greenfeld M et al.   Sleep disturbances in children recovered from central nervous system neoplasms. J Pediatr . 2011; 159( 2): 268– 72.e1. Google Scholar CrossRef Search ADS PubMed  108. Tremaine RB et al.   Subjective and objective sleep in children and adolescents: measurement, age, and gender differences. Sleep Biol Rhythms . 2010; 8( 4): 229– 238. Google Scholar CrossRef Search ADS   109. Arora T et al.   An investigation into the strength of the association and agreement levels between subjective and objective sleep duration in adolescents. PLoS One . 2013; 8( 8): e72406. Google Scholar CrossRef Search ADS PubMed  110. Rigney G et al.   Can a school-based sleep education programme improve sleep knowledge, hygiene and behaviours using a randomised controlled trial. Sleep Med . 2015; 16( 6): 736– 745. Google Scholar CrossRef Search ADS PubMed  111. Marco CA et al.   Family socioeconomic status and sleep patterns of young adolescents. Behav Sleep Med . 2011; 10( 1): 70– 80. Google Scholar CrossRef Search ADS PubMed  112. Amin R et al.   The relationship between sleep disturbance and pulmonary function in stable pediatric cystic fibrosis patients. Chest . 2005; 128( 3): 1357– 1363. Google Scholar CrossRef Search ADS PubMed  113. Beebe DW et al.   Sleep in overweight adolescents: shorter sleep, poorer sleep quality, sleepiness, and sleep-disordered breathing. J Pediatr Psychol . 2007; 32( 1): 69– 79. Google Scholar CrossRef Search ADS PubMed  114. Guedes LG et al.   Comparison between self-reported sleep duration and actigraphy among adolescents: gender differences. Rev Bras Epidemiol . 2016; 19( 2): 339– 347. Google Scholar CrossRef Search ADS PubMed  115. Auger RR et al.   Light exposure among adolescents with delayed sleep phase disorder: a prospective cohort study. Chronobiol Int . 2011; 28( 10): 911– 920. Google Scholar CrossRef Search ADS PubMed  116. Anderson B et al.   Associations of executive function with sleepiness and sleep duration in adolescents. Pediatrics . 2009; 123( 4): e701– e707. Google Scholar CrossRef Search ADS PubMed  117. Tavernier R et al.   Daily affective experiences predict objective sleep outcomes among adolescents. J Sleep Res . 2016; 25( 1): 62– 69. Google Scholar CrossRef Search ADS PubMed  118. Chen MC et al.   Reduced sleep quality in healthy girls at risk for depression. J Sleep Res . 2012; 21( 1): 68– 72. Google Scholar CrossRef Search ADS PubMed  119. Huỳnh C et al.   Sleep-wake patterns of adolescents with borderline personality disorder and bipolar disorder. Child Psychiatry Hum Dev . 2016; 47( 2): 202– 214. Google Scholar CrossRef Search ADS PubMed  120. Law EF et al.   Daytime and nighttime sleep patterns in adolescents with and without chronic pain. Health Psychol . 2012; 31( 6): 830– 833. Google Scholar CrossRef Search ADS PubMed  121. Fobian AD et al.   Impact of media use on adolescent sleep efficiency. J Dev Behav Pediatr . 2016; 37( 1): 9– 14. Google Scholar CrossRef Search ADS PubMed  122. Carskadon MA et al.   Adolescent sleep patterns, circadian timing, and sleepiness at a transition to early school days. Sleep . 1998; 21( 8): 871– 881. Google Scholar CrossRef Search ADS PubMed  123. Jiang F et al.   Effect of chronic sleep restriction on sleepiness and working memory in adolescents and young adults. J Clin Exp Neuropsychol . 2011; 33( 8): 892– 900. Google Scholar CrossRef Search ADS PubMed  124. Dewald JF et al.   Adolescents’ sleep in low-stress and high-stress (exam) times: a prospective quasi-experiment. Behav Sleep Med . 2014; 12( 6): 493– 506. Google Scholar CrossRef Search ADS PubMed  125. Tham SW et al.   Subjective and objective assessment of sleep in adolescents with mild traumatic brain injury. J Neurotrauma . 2015; 32( 11): 847– 852. Google Scholar CrossRef Search ADS PubMed  126. Dewald JF et al.   The Chronic Sleep Reduction Questionnaire (CSRQ): a cross-cultural comparison and validation in Dutch and Australian adolescents. J Sleep Res . 2012; 21( 5): 584– 594. Google Scholar CrossRef Search ADS PubMed  127. Baum KT et al.   Sleep restriction worsens mood and emotion regulation in adolescents. J Child Psychol Psychiatry . 2014; 55( 2): 180– 190. Google Scholar CrossRef Search ADS PubMed  128. Baker E et al.   An investigation of sleep patterns in adolescents with high-functioning autism spectrum disorder compared with typically developing adolescents. Dev Neurorehabil . 2013; 16( 3): 155– 165. Google Scholar CrossRef Search ADS PubMed  129. Malone SK et al.   Characteristics associated with sleep duration, chronotype, and social jet lag in adolescents. J Sch Nurs . 2016; 32( 2): 120– 131. Google Scholar CrossRef Search ADS PubMed  130. Matthews KA et al.   Sleep in healthy black and white adolescents. Pediatrics . 2014; 133( 5): e1189– e1196. Google Scholar CrossRef Search ADS PubMed  131. El-Sheikh M et al.   Perceived discrimination and youths’ adjustment: sleep as a moderator. J Sleep Res . 2016; 25( 1): 70– 77. Google Scholar CrossRef Search ADS PubMed  132. Park H et al.   Sleep and inflammation during adolescence. Psychosom Med . 2016; 78( 6): 677– 685. Google Scholar CrossRef Search ADS PubMed  133. Rodríguez-Colón SM et al.   Sleep variability and cardiac autonomic modulation in adolescents—Penn State Child Cohort (PSCC) study. Sleep Med . 2015; 16( 1): 67– 72. Google Scholar CrossRef Search ADS PubMed  134. Astill RG et al.   Chronic stress undermines the compensatory sleep efficiency increase in response to sleep restriction in adolescents. J Sleep Res . 2013; 22( 4): 373– 379. Google Scholar CrossRef Search ADS PubMed  135. Tonetti L et al.   Effects of sleep timing, sleep quality and sleep duration on school achievement in adolescents. Sleep Med . 2015; 16( 8): 936– 940. Google Scholar CrossRef Search ADS PubMed  © Sleep Research Society 2018. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
TI - Establishing normal values for pediatric nighttime sleep measured by actigraphy: a systematic review and meta-analysis
JF - SLEEP
DO - 10.1093/sleep/zsy017
DA - 2018-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/establishing-normal-values-for-pediatric-nighttime-sleep-measured-by-1UQgDbSHrD
SP - zsy017
VL - 41
IS - 4
DP - DeepDyve
ER -