Abstract
Objectives:
This study examined differences in sleep patterns by race, ethnicity, and socioeconomic status (SES) among children with Obstructive Sleep Apnea Syndrome (OSAS), and linkages between sleep patterns and neurobehavioral functioning.
Method:
We used baseline data from the Childhood Adenotonsillectomy Study (CHAT), a multicenter, single-blind, randomized controlled trial designed to evaluate the efficacy of early adenotonsillectomy versus watchful waiting with supportive care for children with OSAS. Participants included children with OSAS (ages 5.0 to 9.9 years). SES indicators were obtained via questionnaire and geocoding (ArcGIS version 10.1). Caregivers and teachers reported on child inattention/impulsivity and executive functioning. Nighttime sleep duration and variability were measured using five-night sleep diaries.
Results:
Black children experienced shorter nighttime sleep duration than White children, by about 25 minutes, as well as greater sleep duration variability, while sleep duration was more variable in children of “other” racial and ethnic backgrounds versus White children. Of the socioeconomic correlates, only lower family income was associated with sleep duration variability. A short and more variable nighttime sleep duration were each associated with caregiver-rated child inattention and impulsivity. Greater sleep duration variability was linked to greater teacher-rated, but not caregiver-rated, executive functioning impairments.
Conclusions:
Compared to White children with OSAS, Black children with OSAS experience a shorter and more variable nighttime sleep duration. Having a short and/or variable sleep duration may increase risk for neurobehavioral impairments in youth with OSAS, underscoring the potential benefits of sleep health promotion in the context of OSAS care.
Keywords: Disparities, neurobehavior, obstructive sleep apnea syndrome, race, sleep health, socioeconomic status
1. Introduction
Healthy sleep is crucial to optimal child physical, neurobehavioral, and socioemotional development.1,2 A large body of research has demonstrated pediatric sleep health disparities by race, ethnicity, and socioeconomic status (SES).3-5 These pediatric sleep health disparities exist along the continuum of sleep disordered breathing severities, from primary snoring to obstructive sleep apnea syndrome (OSAS). OSAS is characterized by airway obstruction that disrupts ventilation during sleep and is associated with a wide range of poor health outcomes, including increased markers of inflammation,6 metabolic conditions,7 and cardiovascular morbidities.8 Untreated OSA is also associated with neurobehavioral impairments including inattention and diminished executive functioning.9,10 OSAS affects 1.2% to 5.7% of all children,11,12 with an increased prevalence and diagnostic severity among non-Latinx Black/African American (hereafter, ‘Black’) children relative to their non-Latinx White peers (hereafter, ‘White’).3 Beyond racial and ethnic disparities, children from lower-SES homes and/or neighborhoods are also more likely to experience OSAS,3,13 and those from lower-SES homes experience increased OSAS severity.14 Interestingly, previous research has shown that including neighborhood in addition to family-level SES attenuates the observed association between child race and OSAS severity.15
Poor sleep health (e.g., short sleep duration, variable sleep timing)1 and other behavioral sleep concerns (e.g., insomnia symptoms) are highly prevalent in childhood and often co-occur among children clinically referred for OSAS.16,17 Up to 55% of school-aged children obtain insufficient sleep, or less sleep than is recommended by age, on school nights.18 A number of studies have found that insufficient sleep is linked to worse executive functioning skills, emotional regulation, and academic performance.2 Although there is less research available, variable sleep timing and duration are also associated with inattention and behavioral challenges.19 Consistent with noted disparities in OSAS, insufficient sleep is more prevalent among Black compared to White youth and in those of lower-SES versus higher-SES backgrounds.5,20,21 As race and ethnicity are socio-political constructs and do not reflect genetic or other biological differences,22 sociocultural factors including racism at multiple levels (i.e., systemic, institutional, or structural; interpersonal; internalized)23,24 and differential access to care, among other factors, are root causes of these sleep health disparities.4
Despite growing research in the prevalence and outcomes of sleep health disparities, research on disparities in co-occurring sleep problems is needed. There is a surprising lack of research examining sleep patterns, such as sleep duration and regularity, among children with OSAS. In the context of persistent racial, ethnic, and socioeconomic disparities in both pediatric OSAS and sleep health, considering the role of sleep patterns in children with OSAS could identify novel methods to augment OSAS treatment and enhance outcomes. Indeed, research suggests that compared to White children, Black children evidence diminished caregiver- and teacher-reported neurobehavioral improvements following OSAS treatment via adenotonsillectomy.25 Given that more severe pediatric OSAS, assessed via apnea-hypopnea index (AHI), has not been associated with greater neurobehavioral impacts,26,27 it is critical to examine other potentially modifiable factors linked to OSAS outcomes. Insufficient and/or irregular sleep patterns, which are modifiable factors associated with neurobehavioral impairments,2,19 could impact OSAS-related impacts and treatment outcomes but research in this area is limited.
1.1. Current study
The overall goal of this study was to examine differences in child sleep patterns by race, ethnicity, and SES, as well as linkages between child sleep and caregiver- and teacher-rated neurobehavioral functioning in children with OSAS. We focused on nighttime sleep duration and variability as two key aspects of sleep health1 that have documented disparities by race, ethnicity, and SES5,20 In addition, these sleep health domains have been consistently associated with worse caregiver- and teacher-reported child neurobehavioral functioning in children without OSAS,2,19 without corresponding research among children with diagnosed OSAS. We specifically selected caregiver- and teacher-rated neurobehavioral measures rather than observed neurobehavioral task performance as these ratings are more robustly associated with OSAS as well as its treatment.25
Our first study aim was to examine whether there were differences in nighttime sleep duration and variability by child race and ethnicity as well as by neighborhood-level and family-based SES indicators. Based on prior research,5,15 we hypothesized that sleep duration would be shorter and more variable in Black compared to White children, and in children living in lower-SES homes and neighborhoods. In this paper, White children are the reference group for statistical comparisons, however, it is important to clarify that this reference group assignment does not indicate that nighttime sleep duration and variability in White children are the “normative” standard.28 Rather, the use of a White reference group is meant to highlight the relative advantages and privileges afforded to this racial and ethnic group in the US. Our second aim was to identify whether shorter and more variable nighttime sleep duration were each independently associated with worse caregiver- and teacher-rated neurobehavioral functioning in children with OSAS. We hypothesized that shorter and more variable nighttime sleep duration would be associated with worse child neurobehavioral functioning.
2. Method
This study was conducted using baseline data from the Childhood Adenotonsillectomy Study (CHAT),25 a multicenter, single-blind, randomized controlled trial designed to evaluate the efficacy of early adenotonsillectomy (eAT) versus watchful waiting with supportive care (WWSC). The study was approved by the Institutional Review Board of each participating institution. Informed consent was obtained from caregivers, and assent from children ages 7 years and older. The CHAT study design has been previously described in detail.25
2.1. Participants
Participants included children ages 5.0 to 9.9 years who were recruited from pediatric sleep centers, general pediatric clinics and otolaryngology clinics. Inclusion criteria were caregiver report of snoring and polysomnography (PSG) indicating AHI ≥ 2 events per hour of sleep or an obstructive apnea index ≥ 1 per hour, and an otolaryngology evaluation showing that the child was a candidate for adenotonsillectomy. Exclusion criteria included AHI > 30, obstructive apnea index > 20, oxygen saturation < 90% for ≥ 2% total sleep time, significant health problems, medication use for psychiatric or attention deficit hyperactivity disorder, developmental delays requiring school accommodations, recurrent tonsillitis, body mass index (BMI) z-score ≥ 3, and any known genetic, craniofacial, or neurologic disorders likely to affect the airway, cognition, or behavior.
2.2. Measures
2.2.1. Sociodemographic information.
Caregivers completed questionnaires to report on child age, race, ethnicity, sex, prematurity, obesity, asthma, secondhand smoke exposure at home, family income, and maternal education. As in prior CHAT research,15,25 child race and ethnicity were categorized into White, Black, and Other racial and ethnic groups.
2.2.2. Family SES.
Two dichotomized variables used in previous CHAT research15 were selected to represent indicators of family SES. These variables were (1) a caregiver-reported family income of $29,999 or less and (2) a caregiver-reported maternal education of high school or less.
2.2.3. Neighborhood SES.
To assess neighborhood level indicators of lower-SES (i.e., poverty rate; high school dropout rate; proportion of families with related children headed by single females; and proportion of civilian, noninstitutionalized, working age (aged 16–64) males unemployed or not in the labor force), each participant’s address was geocoded using ArcGIS version 10.1 (Environmental Systems Research Institute, Redlands, CA) and joined to American Community Survey 2006–2010 data at the U.S. census tract level to extract area-level sociodemographic data. Lower-SES neighborhoods were defined as neighborhoods with at least three of the four characteristics defined by previous literature:13,29 (1) High percentage of people living in poverty (>28.0%), (2) high percentage of families with related children headed by single females (>21.8%), (3) high percentage of 16- to 19-year-olds who are not enrolled in school and not high school graduates (>17.6%), and (4) High percentage of civilian, noninstitutionalized males 16–64 years old who are unemployed or not in the labor force (>41.2%). A detailed description of geocoding procedures is published elsewhere.15
2.2.4. Nighttime sleep duration and variability.
Caregivers completed a 5-night sleep journal for their child, which included what time their child went to bed and woke up. Nighttime sleep duration was estimated via sleep diary information, as described in prior CHAT research.30 We used the average weighted nighttime sleep duration over all nights of diary data, which adjusted for the number of sleep diary nights completed (i.e., 1 to 5 nights).30 Nighttime sleep duration variability was calculated at the within person level by constructing a robust regression model in which daily nighttime sleep duration was the outcome of interest and fixed effects were set on the level of individual subject, to generate standard errors that reflected average variation for each subject. The number of daily observations varied by subject, from 1 to all 5 nights of diary data.
2.2.5. Neurobehavioral outcomes.
To assess inattention and impulsivity, caregivers and teachers completed the Conners’ Rating Scale Revised: Long Version Global Index (Conners). We used the Conners overall T-scores reflecting the restless-impulsive and emotional lability factor sets.31 Caregivers and teachers of children were given assessments for executive functioning using the Behavior Rating Inventory of Executive Function (BRIEF) Global Executive Functioning T-score, which is comprised of measures of behavioral regulation (e.g., attention, task shifting), working memory, and metacognition.32
2.3. Analytic approach
Statistical analyses were conducted with Stata 17MP (StataCorp, State College, TX) with two-sided tests of hypotheses and a p-value <.05. Descriptive analyses included tabulation of count and frequency for categorical variables and mean, standard deviation, and interquartile range (IQR) for continuous variables. We conducted tests of normal distribution (i.e., Shapiro-Wilk test) to determine the extent of skewness for continuous data.
We performed a series of analyses to identify covariates to include in final models. First, we used robust linear regression to examine univariate associations by aim between dependent variables (sleep and neurobehavioral functioning), independent variables of interest (child race and ethnicity, family income, maternal education of high school or less, and lower neighborhood SES), and covariates previously included in CHAT study analyses.15,25 Potential covariates for Aim 1 were child age (5-7 years or 8-10 years), gender, history of prematurity, asthma, obesity, and secondhand smoke exposure. Covariates for Aim 2 included Aim 1 covariates and the SES indicators. Next, each of the independent variables and covariates were entered into multivariable stepwise regression models to identify the most parsimonious selection of covariates for finalized models. Stepwise results informed subsequent finalized robust regression models by aim. Robust regression was applied for all final models as this method accommodates significant outliers and/or influential observations. The regression coefficient estimates with 95% confidence interval (95% CI) and the associated p-value were reported.
4. Results
Sociodemographic, sleep, and neurobehavioral characteristics of the study sample are shown in Table 1. A total of 380 children had at least 1 sleep diary entry during the 5-day period. Of these, 2 children (0.5%) had 2 sleep diary entries, 10 (2.6%) had 3 sleep diary entries, 45 (11.8%) had 4 sleep diary entries, and 210 (55.3%) had all 5 sleep diary entries. Average nighttime sleep duration was 9.52 hours (SD = 1.47); 3 participants had a nighttime sleep value ≥3 SD above the mean and were included in robust regression analyses because this method could accommodate these significant outliers. The Shapiro-Wilk test of normality revealed that nighttime sleep duration variability was significantly positively skewed. The median for this variable was 0.59 (range = 0.50 – 3.45), which is approximately 35 minutes. An inverse square function normalized the distribution. The transformed variable was used in Aim 1 analyses.
Table 1.
Sociodemographic, sleep, and neurobehavioral characteristics for the study sample (N = 464)
| Variables | % (n) | Mean (SD) |
|---|---|---|
| Child age (years) | 7.0 (1.4) | |
| Child female gender | 51.7% (234) | |
| Child race: White | 36.0 (167) | |
| Black/African American | 53.7% (249) | |
| Other | 10.3% (48) | |
| Child history of prematurity | 15.0% (68) | |
| Child history of asthma diagnosis | 31.6% (143) | |
| Child BMI ≥ 95% | 33.1% (150) | |
| Secondhand smoke exposure at home | 33.0% (153) | |
| Maternal education ≤ High school | 31.3% (140) | |
| Family income: ≤ $29,999 | 40.3% (182) | |
| Low neighborhood SES | 23.9% (108) | |
| Weighted nightly sleep duration (hours) | 9.45 (1.24) | |
| Sleep time variability | 0.77 (0.53) | |
| Child inattention/hyperactivity (Conner’s global T-score) | ||
| Caregiver-rated | 52.17 (11.45) | |
| Teacher-rated | 54.75 (12.81) | |
| Child executive functioning impairments (BRIEF global T-score) | ||
| Caregiver-rated | 49.94 (11.31) | |
| Teacher-rated | 56.54 (12.06) | |
Note. Conner’s = Conner’s Rating Scale Revised: Long Version; BRIEF = Behavior Rating Inventory of Executive Function; SES = socioeconomic status.
4.1. Aim 1: Associations among sleep, race, ethnicity, and SES
Table 2 shows preliminary univariate robust models for associations between nighttime sleep duration and each of the independent variables (child race and ethnicity, maternal education, family income, and neighborhood SES) and covariates. Older child age and obesity were associated with shorter nighttime sleep duration in univariate models but only child age emerged as a significant covariate in stepwise analyses. The final robust regression model (Table 2) that included child age as a covariate showed a significantly shorter sleep duration in Black compared to White children (β = −0.42, 95% CI [−0.71, −0.13], p=.005), by approximately 25 minutes on average (Figure 1). Average nighttime sleep duration for children of other (i.e., non-White and non-Black) racial and ethnic backgrounds did not significantly differ from White children (β = −0.20, 95% CI [−0.59, 0.19], p = .316). Sleep duration did not vary by family SES indicators (i.e., family income and maternal education) or by neighborhood SES (Table 2).
Table 2.
Race, ethnicity, family SES and neighborhood SES correlates of nighttime sleep duration and nighttime sleep variability
| Univariate associations for each independent variable and covariate |
Nighttime sleep duration | Nighttime sleep variability | ||||
|---|---|---|---|---|---|---|
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Race and ethnicity1 | ||||||
| Black | −0.48 (−0.73, −0.24) | 0.13 | <.001 | 0.22 (−0.05, 0.48) | 0.13 | .104 |
| Other | −0.19 (−0.57, 0.21) | 0.20 | .340 | 0.30 (−0.12, 0.71) | 0.21 | .158 |
| Lower family income | 0.19 (−0.05, 0.43) | 0.12 | .115 | 0.27 (0.02, 0.53) | 0.13 | .037 |
| Maternal high school education or less | 0.17 (−0.08, 0.43) | 0.13 | .178 | −0.05 (−0.32, 0.22) | 0.14 | .712 |
| Lower neighborhood SES | −0.35 (−0.62, −0.07) | 0.14 | .013 | 0.11 (−0.19, 0.42) | 0.16 | .467 |
| Female gender | −0.14 (−0.38, 0.09) | 0.12 | .226 | 0.07 (−0.18, 0.32) | 0.13 | .577 |
| Age 8-10 years | −0.35 (−0.61, −0.10) | 0.13 | .007 | 0.04 (−0.24, 0.32) | 0.14 | .791 |
| Secondhand smoke | −0.17 (−0.43, 0.08) | 0.13 | .186 | −0.17 (−0.44, 0.10) | 0.14 | .208 |
| Prematurity | 0.04 (−0.28, 0.36) | 0.16 | .817 | −0.17 (−0.51, 0.17) | 0.17 | .328 |
| Asthma | −0.15 (−0.40, 0.10) | 0.13 | .248 | −0.09 (−0.36, 0.18) | 0.14 | .533 |
| Obesity | −0.28 (−0.53, −0.04) | 0.12 | .023 | −0.09 (−0.35, 0.18) | 0.13 | .521 |
| Final models after stepwise analyses2 | Nighttime sleep duration | Nighttime sleep variability | ||||
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Race and ethnicity1 | ||||||
| Black | −0.42 (−0.71, −0.13) | 0.15 | .005 | 0.38 (0.07, 0.69) | 0.16 | .018 |
| Other | −0.20 (−0.59, 0.19) | 0.20 | .316 | 0.47 (0.04, 0.89) | 0.22 | .034 |
| Lower family income | −0.01 (−0.27, 0.25) | 0.13 | .940 | 0.48 (0.19, 0.76) | 0.14 | .001 |
| Maternal high school education or less | −0.04 (−0.27, 0.27) | 0.14 | .978 | −0.11 (−0.41, 0.19) | 0.15 | .454 |
| Lower neighborhood SES | −0.06 (−0.37, 0.26) | 0.16 | .714 | 0.04 (−0.30, 0.39) | 0.18 | .800 |
| Age 8-10 years | −0.27 (−0.53, −0.01) | 0.13 | .041 | -- | -- | -- |
Note.
White assigned as reference group to reflect relative advantages and privileges afforded to this racial and ethnic group in the United States.
Final robust regression models contained only the covariates that were significant at p<.05 in stepwise analyses.
Figure 1. Marginal means for differences by race and ethnicity in sleep duration.
Note. ***p<.001, ns = not statistically significant at p<.05, vertical whiskers represent 95% confidence intervals.
In preliminary univariate and stepwise regression analyses, no covariates were significantly associated with sleep duration variability (Table 2). The final robust regression model showed significantly more variable sleep duration in Black children (β = 0.38, 95% CI [0.07, 0.69], p =.018) and those of other racial and ethnic backgrounds (β = 0.47, 95% CI [0.04, 0.89], p = .034) compared to White children (Table 2). Regardless of child racial and ethnic background, having a lower family income was also significantly associated with increased sleep duration variability (β = 0.48, 95% CI [0.19, 0.76,] p = .001). Neither maternal education nor neighborhood SES were associated with this sleep outcome.
4.2. Aim 2: Associations among sleep and neurobehavioral outcomes
Preliminary analyses (Table 3) indicated that although lower family income and having a history of asthma were both associated with caregiver-rated child inattention and impulsivity, neither covariate was associated with this outcome in stepwise analyses. Thus, no covariates were selected into the final robust regression model for caregiver-reported child inattention/impulsivity (Table 3), which suggested that both short nighttime sleep duration (β= −0.97, 95% CI [−1.92, −0.03], p = .043) and greater sleep duration variability (β= 2.58, 95% CI [0.22, 4.94], p = .032) were associated with worse caregiver-rated child inattention/hyperactivity. Child race and ethnicity, lower neighborhood SES, secondhand smoke exposure, and obesity were associated with teacher-rated child inattention and impulsivity in univariate analyses (Table 3), but only child race and ethnicity and secondhand smoke exposure were retained as covariates after stepwise analyses. The final robust regression model with these two covariates revealed no significant associations between each of the sleep variables and this neurobehavioral outcome (sleep duration β= 0.07, 95% CI [−1.40, 1.26], p = .916; sleep variability β= 1.99, 95% CI [−1.09, 5.08], p = .204).
Table 3.
Associations of nighttime sleep duration and variability with caregiver- and teacher-rated child inattention/impulsivity
| Univariate associations for each independent variable and covariate |
Caregiver-rated inattention and impulsivity |
Teacher-rated inattention and impulsivity |
||||
|---|---|---|---|---|---|---|
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Nighttime sleep duration | −0.05 (−0.70, 0.69) | 0.35 | .989 | 0.28 (−0.74, 1.31) | 0.52 | .588 |
| Nighttime sleep variability | 1.03 (−0.90, 2.97) | 0.98 | .294 | 2.27 (0.03, 4.51) | 1.14 | .047 |
| Race and ethnicity:1 Black | −0.91 (−2.99, 1.16) | 1.05 | .391 | 4.59 (1.90, 7.29) | 1.37 | .001 |
| Other | −0.20 (−3.55, 3.15) | 1.71 | .906 | 0.94 (−3.72, 5.61) | 2.37 | .689 |
| Lower family income | −2.22 (−4.17, −0.28) | 0.99 | .025 | −1.88 (−4.50, 0.74) | 1.33 | .158 |
| Maternal high school education or less | 0.47 (−1.59, 2.52) | 1.05 | .653 | 0.16 (−2.88, 3.12) | 1.50 | .918 |
| Lower neighborhood SES | −0.60 (−2.85, 1.64) | 1.14 | .597 | 3.31 (0.26, 6.35) | 1.55 | .033 |
| Female gender | −1.78 (−3.67, 0.11) | 0.96 | .065 | 0.07 (−2.56, 2.70) | 1.34 | .958 |
| Age 8-10 years | 0.13 (−2.01, 2.27) | 1.09 | .906 | 1.24 (−1.75, 4.24) | 1.52 | .415 |
| Secondhand smoke | 1.99 (−0.06, 4.03) | 1.04 | .057 | 4.89 (2.11, 7.66) | 1.41 | .001 |
| Prematurity | 1.91 (−0.77, 4.59) | 1.36 | .161 | 1.49 (−2.44, 5.43) | 2.00 | .455 |
| Asthma | 2.11 (0.06, 4.16) | 1.04 | .043 | 2.28 (−0.40, 4.95) | 1.36 | .095 |
| Obesity | 0.98 (−1.05, 3.01) | 1.04 | .346 | 3.44 (0.82, 6.06) | 1.33 | .010 |
| Final models after stepwise analyses2 | Caregiver-rated inattention and impulsivity |
Caregiver-rated inattention and impulsivity |
||||
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Nighttime sleep duration | −0.97 (−1.92, −0.03) | 0.48 | .043 | 0.07 (−1.40, 1.26) | 0.67 | .916 |
| Nighttime sleep variability | 2.58 (0.22, 4.94) | 1.20 | .032 | 1.99 (−1.09, 5.08) | 1.57 | .204 |
| Race and ethnicity:1 Black | -- | -- | -- | 3.34 (0.35, 6.33) | 1.52 | .029 |
| Other | -- | -- | -- | 0.47 (−4.36, 5.29) | 2.50 | .849 |
| Secondhand smoke exposure | -- | 4.32 (1.25, 7.38) | 1.56 | .006 | ||
Note. Caregiver- and teacher-rated child inattention/hyperactivity indexed by caregiver- and teacher-rated Conner’s T scores.
White assigned as reference group to reflect relative advantages and privileges afforded to this racial and ethnic group in the United States.
Final robust regression models contained only the covariates that were significant at p<.05 in stepwise analyses.
Table 4 displays univariate analyses and final robust regression models for caregiver- and teacher-rated child executive functioning impairments. A lower family income, secondhand smoke exposure, and obesity were associated with greater caregiver-rated child executive functioning impairments in univariate models. However, obesity was the only covariate retained in stepwise analyses. In the final robust regression model, which covaried for obesity, neither nighttime sleep duration (β= −0.34, 95% CI [−1.25, 0.58], p = .469) nor sleep duration variability (β= 1.40, 95% CI [−1.13, 3.93], p = .276) were significantly associated with this outcome. A lower family income and obesity were each linked to teacher-rated child executive functioning impairments in univariate analyses; only obesity was retained in stepwise analyses. In the final robust regression model, greater sleep duration variability was significantly associated with greater teacher-rated child executive functioning impairments (β= 4.19, 95% CI [0.85, 7.52], p = .014). There was no significant association between nighttime sleep duration and this outcome (β= −0.74, 95% CI [−2.14, 0.66], p = .296).
Table 4.
Associations of nighttime sleep duration and variability with caregiver- and teacher-rated child executive functioning impairments
| Univariate associations for each independent variable and covariate |
Caregiver-rated executive functioning impairments |
Teacher-rated executive functioning impairments |
||||
|---|---|---|---|---|---|---|
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Nighttime sleep duration | −0.08 (−0.86, 0.69) | 0.39 | .830 | 0.25 (−0.88, 1.38) | 0.57 | .661 |
| Nighttime sleep variability | 1.17 (−0.97, 3.31) | 1.09 | .283 | 3.74 (1.10, 6.38) | 1.34 | .006 |
| Race and ethnicity:1 Black | −1.23 (−3.53, 1.06) | 1.17 | .290 | 2.95 (−0.02, 5.90) | 1.50 | .050 |
| Other | 1.36 (−2.38, 5.09) | 1.90 | .475 | −1.67 (−6.69, 3.35) | 2.55 | .514 |
| Lower family income | −2.25 (−4.40, −0.10) | 1.10 | .041 | −2.83 (−5.73, 0.07) | 1.50 | .056 |
| Maternal high school education or less | −0.21 (−2.48, 2.06) | 1.16 | .856 | −3.27 (−6.39, −0.15) | 1.59 | .040 |
| Lower neighborhood SES | −0.33 (−2.80, 2.13) | 1.26 | .790 | 3.01 (−0.42, 6.44) | 1.75 | .085 |
| Female gender | −2.00 (−4.11, 0.11) | 1.08 | .063 | −2.57 (−5.3, 0.20) | 1.41 | .069 |
| Age 8-10 years | 0.35 (−2.01, 2.71) | 1.20 | .772 | 2.97 (−0.16, 6.11) | 1.60 | .063 |
| Secondhand smoke | 3.85 (1.61, 6.09) | 1.14 | .001 | 2.91 (−0.13, 5.97) | 1.55 | .061 |
| Prematurity | 2.48 (−0.45, 5.41) | 1.49 | .096 | 0.10 (−4.11, 4.30) | 2.14 | .964 |
| Asthma | 1.79 (−0.49, 4.07) | 1.16 | .123 | 1.52 (−1.43, 4.49) | 1.51 | .316 |
| Obesity | 2.62 (0.39, 4.86) | 1.14 | .021 | 5.17 (2.27, 8.09) | 1.48 | .001 |
| Final models after stepwise analyses2 | Caregiver-rated executive functioning impairments |
Teacher-rated executive functioning impairments |
||||
| β (95% CI) | SE | p | β (95% CI) | SE | p | |
| Nighttime sleep duration | −0.34 (−1.25, 0.58) | 0.46 | .469 | −0.74 (−2.14, 0.66) | 0.71 | .296 |
| Nighttime sleep variability | 1.40 (−1.13, 3.93) | 1.29 | .276 | 4.19 (0.85, 7.52) | 1.70 | .014 |
| Obesity | 2.53 (0.09, 4.96) | 1.23 | .042 | 4.84 (1.72, 8.00) | 1.60 | .002 |
Note. Caregiver- and teacher-rated child executive functioning impairments indexed by caregiver- and teacher-rated BRIEF T scores.
White assigned as reference group to reflect relative advantages and privileges afforded to this racial and ethnic group in the United States.
Final robust regression models contained only the covariates that were significant at p<.05 in stepwise analyses.
5. Discussion
This study examined differences in nighttime sleep duration and variability by child race, ethnicity, and SES indicators in children with OSAS, as well as associations between these sleep patterns and neurobehavioral outcomes. Consistent with prior pediatric sleep health disparities research5,33 and in partial support of our hypotheses, Black children experienced a shorter nighttime sleep duration, by about 25 minutes, compared to White youth, as well as more variable sleep timing. Of the SES indicators, however, only lower family income was linked to sleep duration variability. Having a short and more variable nighttime sleep duration was associated with increased caregiver-rated child inattention/hyperactivity, whereas only a more variable sleep duration was linked to greater teacher-reported child executive functioning impairments. No other linkages emerged between sleep patterns and caregiver-reported executive functioning or teacher-reported inattention and impulsivity.
Our findings on racial and ethnic disparities in child sleep duration and variability extend prior research to children diagnosed with OSAS, a clinical population that is already at greater risk for adverse health outcomes due to sleep disordered breathing.9,10 A difference of 25 minutes between Black and White youth is clinically meaningful, given experimental research showing that a 20-30-minute increase in nighttime sleep duration is associated with better child neurobehavioral regulation.34 Although there are no established cutoffs for clinically significant sleep duration variability in childhood, existing research suggests that irregular sleep patterns, including duration and bedtime, are linked to increased daytime sleepiness,35 which could contribute to challenges with maintaining attention and focus at school.36
Multilevel social and environmental factors likely contribute to these observed Black-White sleep health disparities.4 For instance, we found that regardless of racial and ethnic background, lower family income was linked to a more variable sleep duration. Studies have shown that caregivers in lower SES homes may experience obstacles to implementing consistent child bedtime routines due to competing demands and stressors such as inflexible work schedules and shift work.37 A growing body of research also suggests that racism and discrimination could result in bedtime hypervigilance and stress, contributing to a short and variable sleep duration in Black compared to White youth. Additional research is needed to examine modifiable factors that could contribute to or buffer against these disparities, including among children with comorbid sleep conditions like OSAS.
Of note, while there was a significant difference in nighttime sleep duration variability when comparing youth of “other” (i.e., non-White and non-Black) racial and ethnic backgrounds to White youth, this finding is challenging to interpret given limited information about youth in this “other” group. Recent research has examined racial and ethnic sleep health disparities among youth of Asian33 and of American Indian/Alaska Native38 backgrounds. Well-powered studies focused on identifying modifiable contributors to these sleep health disparities by specific racial and ethnic groups are crucial for designing culturally responsive sleep health promotion strategies.4 For example, while there is research testing culturally tailored sleep treatment for Black adults with OSAS,39 there very little pediatric research in this regard.
Future research on the potentially amplified effects of comorbid pediatric OSAS and poor sleep health on neurobehavioral functioning could help tailor interventions for youth with OSAS. Both sleep disordered breathing and insufficient and/or irregular sleep are thought to impact neurobehavior via the frontal lobes.40,41 Thus, experiencing both OSAS and poor sleep health could be especially detrimental to neurobehavior, particularly in childhood when there is rapid frontal lobe development. The current study could not compare neurobehavior among youth with short and/or irregular sleep duration but without OSAS, which will be important in future research examining the extent to which comorbid medical and behavioral sleep conditions impact child outcomes. However, we found that in youth with OSAS, a short and more variable nighttime sleep duration is associated with greater caregiver-rated child inattention and impulsivity, while a more variable sleep duration is associated with poorer teacher-rated child executive functioning skills. Given these findings, clinicians treating child OSAS should consider assessing for and addressing short and variable child sleep duration.
At the same time, additional research is needed to establish these linkages between poor sleep health and neurobehavior in youth with OSAS, as we found that associations varied across teacher- and caregiver-reporters and neurobehavioral domains. It is notable that sleep duration and variability were both associated with worse caregiver-rated, but not teacher-rated, child inattention and impulsivity. Caregivers may be more attuned to fluctuations in their child’s behavior, as well as next-day impacts of poor sleep, although these temporal relations could not be assessed in this study. The finding for sleep variability being linked to worse teacher-rated executive functioning could have resulted from teachers’ being more sensitive to differences in child executive functioning skills compared to their peers at school. Caregivers may not be able to make these executive functioning comparisons, and there may also be more demands for executive functioning skills (e.g., following multi-step commands, switching between academic tasks) in a classroom setting. It may also be that sleep duration and variability not as robustly linked to child neurobehavior in the context of OSAS. Poor sleep health may also manifest in greater downstream consequences, such as subsequent academic achievement, rather than impacting neurobehavior,42 however, more research in the context of co-occurring poor sleep health and diagnosed OSAS is necessary.
5.1. Limitations
Study findings are limited by our use of secondary data from the CHAT study, which was designed to assess OSAS and related outcomes among children assigned to eAT versus WWSC. As such, the study was not designed to detect differences in sleep health patterns or to evaluate the extent of racial, ethnic, and socioeconomic disparities. Thus, findings may not be generalizable to youth without co-occurring OSAS and to adolescents, given the age range of children included in CHAT. Furthermore, CHAT excluded youth with AHI > 30 events/hour, and therefore findings may not reflect youth with more severe OSAS, who may experience increased sleep disruption. Of note, however, increased OSAS severity has not been linked to worse neurobehavior, with some research showing increased neurobehavioral sequalae in children with mild sleep disordered breathing.26 Additionally, the neurobehavioral benefits observed in CHAT for children treated with eAT were small overall, and even smaller among Black compared to White youth in the study, which may have further impacted our ability to detect small effects in the current analyses. Importantly, our assessment of sleep duration and variability were limited to caregiver-reported sleep diary data, which may be more indicative of sleep opportunity (i.e., time in bed) rather than sleep duration. Sleep diary data were also limited to 5 days of data collection, with some families completing only one diary entry. Additional longitudinal measures of sleep patterns, including actigraphy, were not included in the CHAT study. Future research should use actigraphy and accompanying child reports to examine sleep health patterns in youth with OSAS, with an assessment of both overnight and daytime sleep over multiple weeks to better reflect sleep patterns across multiple school and weekend days.
Assessing total (24-hour) sleep is especially important in future research as prior work has found that increased daytime sleep in Black compared to White youth offsets observed Black-White disparities in overnight sleep duration.43 This study is also limited in its ability to examine differences between and within other racial and ethnic groups. The present study examined a sample of predominantly White and Black youth and was underpowered to determine differences in the “other” racial and ethnic category, which itself is highly limited given its nonspecific nature. Future research would strongly benefit from clearly identifying racial and ethnic groups, with a focus improving understanding and knowledge of health inequities, increasing inclusivity, and strengthening the results on potential sleep health equity interventions. Finally, as this research is cross-sectional, no causality can be inferred, and future research should examine linkages between poor sleep health and neurobehavior in children with OSAS longitudinally.
5.2. Conclusions
This secondary analysis of baseline CHAT study data extends research on pediatric sleep health disparities by demonstrating that Black-White disparities in nighttime sleep duration and variability, as well as income-related disparities in sleep variability, exist in children with OSAS. A short and more variable sleep duration may exacerbate OSAS-related impacts on child inattention and impulsivity, at least according to caregiver reports, whereas more variable sleep duration could worsen teacher-rated child executive functioning impairments. Although additional research is needed, clinicians should consider assessing for and treating poor sleep health, particularly variable sleep duration, in children with OSAS.
Highlights.
Sleep duration was shorter and more variable in Black vs. White children with OSAS.
Sleep duration and variability were linked to aspects of neurobehavioral functioning.
Sleep patterns are modifiable and should be considered when treating child OSAS.
Acknowledgments:
The authors thank the National Sleep Research Resource for access to the study data.
Funding:
The Childhood Adenotonsillectomy Trial (CHAT) was supported by the National Institutes of Health (HL083075, HL083129, UL1-RR-024134, UL1 RR024989). The National Sleep Research Resource was supported by the National Heart, Lung, and Blood Institute (R24 HL114473, 75N92019R002). R01HL152454 (AAW; IET)
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Clinical Trial Registration: Childhood Adenotonsillectomy Study for Children with OSA (CHAT). ClinicalTrials.gov Identifier #NCT00560859.
References
- 1.Meltzer LJ, Williamson AA, Mindell JA. Pediatric sleep health: It matters, and so does how we define it. Sleep Med Rev. Jun 2021;57:101425. doi: 10.1016/j.smrv.2021.101425 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Matricciani L, Paquet C, Galland B, Short M, Olds T. Children’s sleep and health: A meta-review. Sleep Med Rev. 2019;46:136–150. doi: 10.1016/j.smrv.2019.04/011 [DOI] [PubMed] [Google Scholar]
- 3.Williamson AA, Johnson TJ, Tapia IE. Health disparities in pediatric sleep-disordered breathing. Paediatr Respir Rev. 2022; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jackson CL, Walker JR, Brown MK, Das R, Jones NL. A workshop report on the causes and consequences of sleep health disparities. Sleep. 2020;43(8):zsaa037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Guglielmo D, Gazmararian JA, Chung J, Rogers AE, Hale L. Racial/ethnic sleep disparities in us school-aged children and adolescents: A review of the literature. Sleep Health. 2018;4(1):68–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kheirandish-Gozal L, Gozal D. Pediatric OSA syndrome morbidity biomarkers: the hunt is finally on! Chest. 2017;151(2):500–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gileles-Hillel A, Kheirandish-Gozal L, Gozal D. Biological plausibility linking sleep apnoea and metabolic dysfunction. Nature Reviews Endocrinology. 2016;12(5):290–298. [DOI] [PubMed] [Google Scholar]
- 8.Bhattacharjee R, Kheirandish-Gozal L, Pillar G, Gozal D. Cardiovascular complications of obstructive sleep apnea syndrome: evidence from children. Prog Cardiovasc Dis. 2009;51(5):416–433. [DOI] [PubMed] [Google Scholar]
- 9.Mietchen JJ, Bennett DP, Huff T, Hedges DW, Gale SD. Executive function in pediatric sleep-disordered breathing: a meta-analysis. J Int Neuropsychol Soc. 2016;22(8):839–850. [DOI] [PubMed] [Google Scholar]
- 10.Zhao J, Han S, Zhang J, et al. Association between mild or moderate obstructive sleep apnea-hypopnea syndrome and cognitive dysfunction in children. Sleep Med. 2018;50:132–136. [DOI] [PubMed] [Google Scholar]
- 11.Bixler EO, Vgontzas AN, Lin H-M, et al. Sleep disordered breathing in children in a general population sample: prevalence and risk factors. Sleep. 2009;32(6):731–736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Marcus C, Brooks LJ, Draper KA, Gozal D, Halbower AC, Jones J, Schechter MS, Sheldon SH, Spruyt K, Lehmann C, Shiffman RN. . Clinical practice guideline: Diagnosis and management of childhood obstructive sleep apnea syndrome. Pediatrics. 2012;130:576–584. doi: 10.1542/peds.2012-1671 [DOI] [PubMed] [Google Scholar]
- 13.Spilsbury JC, Storfer-Isser A, Kirchner HL, et al. Neighborhood disadvantage as a risk factor for pediatric obstructive sleep apnea. The Journal Of Pediatrics. 2006;149(3):342–347. [DOI] [PubMed] [Google Scholar]
- 14.Weinstock TG, Rosen CL, Marcus CL, et al. Predictors of obstructive sleep apnea severity in adenotonsillectomy candidates. Sleep. 2014;37(2):261–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang R, Dong Y, Weng J, et al. Associations among neighborhood, race, and sleep apnea severity in children. A six-city analysis. Annals Of The American Thoracic Society. 2017;14(1):76–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Moore M, Bonuck K. Comorbid symptoms of sleep-disordered breathing and behavioral sleep problems from 18–57 months of age: a population-based study. Behav Sleep Med. 2013;11(3):222–230. [DOI] [PubMed] [Google Scholar]
- 17.Byars K, Apiwattanasawee P, Leejakpai A, Tangchityongsiva S, Simakajornboom N. Behavioral sleep disturbances in children clinically referred for evaluation of obstructive sleep apnea. Sleep Med. 2011;12(2):163–169. [DOI] [PubMed] [Google Scholar]
- 18.Friel CP, Duran AT, Shechter A, Diaz KM. US children meeting physical activity, screen time, and sleep guidelines. Am J Prev Med. 2020;59(4):513–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Becker SP, Sidol CA, Van Dyk TR, Epstein JN, Beebe DW. Intraindividual variability of sleep/wake patterns in relation to child and adolescent functioning: A systematic review. Sleep Med Rev. 2017;34:94–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Moore M, Kirchner HL, Drotar D, Johnson N, Rosen C, Redline S. Correlates of adolescent sleep time and variability in sleep time: the role of individual and health related characteristics. Sleep Med. 2011;12(3):239–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sheehan C, Powers D, Margerison-Zilko C, McDevitt T, Cubbin C. Historical neighborhood poverty trajectries and child sleep. Sleep Health. 2018;4:127–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Boyd RW, Lindo EG, Weeks LD, McLemore MR. On racism: a new standard for publishing on racial health inequities. Health Affairs Blog. 2020;10 [Google Scholar]
- 23.Jones CP. Levels of racism: a theoretic framework and a gardener's tale. Am J Public Health. 2000;90(8):1212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Paradies Y, Ben J, Denson N, et al. Racism as a determinant of health: a systematic review and meta-analysis. PLoS One. 2015;10(9):e0138511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Marcus CL, Moore RH, Rosen CL, et al. A randomized trial of adenotonsillectomy for childhood sleep apnea. N Engl J Med. 2013;368:2366–2376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yu PK, Radcliffe J, Gerry Taylor H, et al. Neurobehavioral morbidity of pediatric mild sleep-disordered breathing and obstructive sleep apnea. Sleep. 2022;45(5):zsac035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Isaiah A, Spanier AJ, Grattan LM, Wang Y, Pereira KD. Predictors of behavioral changes after adenotonsillectomy in pediatric obstructive sleep apnea: a secondary analysis of a randomized clinical trial. JAMA Otolaryngology–Head & Neck Surgery. 2020;146(10):900–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Andrews K, Parekh J, Peckoo S. How to embed a racial and ethnic equity perspective in research: Practical guidance for the research process. Washington, DC: Child Trends. 2019; [Google Scholar]
- 29.O'Hare WP, Mather M. The growing number of kids in severely distressed neighborhoods: Evidence from the 2000 Census. Citeseer; 2003. [Google Scholar]
- 30.Paruthi S, Buchanan P, Weng J, et al. Effect of adenotonsillectomy on parent-reported sleepiness in children with obstructive sleep apnea. Sleep. 2016;39(11):2005–2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Conners CK. Conners' Rating Scales--revised: CRS-R. Multi-Health Systems North Tonawanda, NJ; 2001. [Google Scholar]
- 32.Gioia GA, Isquith PK, Guy SC, Kenworthy L. Behavior rating inventory of executive function: BRIEF. Psychological Assessment Resources Odessa, FL; 2000. [Google Scholar]
- 33.Yip T, Cheon YM, Wang Y, Deng WQ, Seligson AL. Sociodemographic and environmental factors associated with childhood sleep duration. Sleep Health. 2020; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Gruber R, Cassoff J, Frenette S, Wiebe S, Carrier J. Impact of sleep extension and restriction on children’s emotional lability and impulsivity. Pediatrics. 2012;130(5):e1155–e1161. [DOI] [PubMed] [Google Scholar]
- 35.Ng A, Adjaye-Gbewonyo D, Black L. Regular bedtimes among children aged 5-17 years: United States, 2020. NCHS Data Brief. 2022;437 [PubMed] [Google Scholar]
- 36.Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, Bögels SM. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta-analytic review. Sleep Med Rev. 2010;14(3):179–189. [DOI] [PubMed] [Google Scholar]
- 37.Lee S, Hale L, Berger LM, Buxton OM. Maternal perceived work schedule flexibility predicts child sleep mediated by bedtime routines. Journal of Child And Family Studies. 2019;28(1):245–259. [Google Scholar]
- 38.Palimaru A, Dong L, Brown R, et al. Mental health, family functioning and sleep in cultural context among American Indian/Alaska Native urban youth: A mixed methods study. Soc Sci Med. 2022;292:114582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Seixas AA, Trinh-Shevrin C, Ravenell J, Ogedegbe G, Zizi F, Jean-Louis G. Culturally tailored, peer-based sleep health education and social support to increase obstructive sleep apnea assessment and treatment adherence among a community sample of blacks: study protocol for a randomized controlled trial. Trials. 2018;19(1):1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Isaiah A, Ernst T, Cloak CC, Clark DB, Chang L. Associations between frontal lobe structure, parent-reported obstructive sleep disordered breathing and childhood behavior in the ABCD dataset. Nature Communications. 2021;12(1):1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Krause AJ, Simon EB, Mander BA, et al. The sleep-deprived human brain. Nature Reviews Neuroscience. 2017;18(7):404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Tonetti L, Fabbri M, Filardi M, Martoni M, Natale V. Effects of sleep timing, sleep quality and sleep duration on school achievement in adolescents. Sleep Med. 2015;16(8):936–940. [DOI] [PubMed] [Google Scholar]
- 43.James S, Chang A-M, Buxton OM, Hale L. Disparities in adolescent sleep health by sex and ethnoracial group. SSM-population health. 2020;11:100581. [DOI] [PMC free article] [PubMed] [Google Scholar]