Neighborhood Disadvantage, Quality of Life, and Symptom Burden in Children with Mild Sleep-disordered Breathing

Abstract

Rationale

Neighborhood disadvantage (ND) has been associated with sleep-disordered breathing (SDB) in children. However, the association between ND and SDB symptom burden and quality of life (QOL) has not yet been studied.

Objectives

To evaluate associations between ND with SDB symptom burden and QOL.

Methods

Cross-sectional analyses were performed on 453 children, ages 3–12.9 years, with mild SDB (habitual snoring and apnea–hypopnea index < 3/h) enrolled in the PATS (Pediatric Adenotonsillectomy Trial for Snoring) multicenter study. The primary exposure, neighborhood disadvantage, was characterized by the Child Opportunity Index (COI) (range, 0–100), in which lower values (specifically COI ⩽ 40) signify less advantageous neighborhoods. The primary outcomes were QOL assessed by the obstructive sleep apnea (OSA)-18 questionnaire (range, 18–126) and SDB symptom burden assessed by the Pediatric Sleep Questionnaire–Sleep-related Breathing Disorder (PSQ-SRBD) scale (range, 0–1). The primary model was adjusted for age, sex, race, ethnicity, maternal education, recruitment site, and season. In addition, we explored the role of body mass index (BMI) percentile, environmental tobacco smoke (ETS), and asthma in these associations.

Results

The sample included 453 children (16% Hispanic, 26% Black or African American, 52% White, and 6% other). COI mean (standard deviation [SD]) was 50.3 (29.4), and 37% (n = 169) of participants lived in disadvantaged neighborhoods. Poor SDB-related QOL (OSA-18 ⩾ 60) and high symptom burden (PSQ-SRBD ⩾ 0.33) were found in 30% (n = 134) and 75% (n = 341) of participants, respectively. In adjusted models, a COI increase by 1 SD (i.e., more advantageous neighborhood) was associated with an improvement in OSA-18 score by 2.5 points (95% confidence interval [CI], −4.34 to −0.62) and in PSQ-SRBD score by 0.03 points (95% CI, −0.05 to −0.01). These associations remained significant after adjusting for BMI percentile, ETS, or asthma; however, associations between COI and SDB-related QOL attenuated by 23% and 10% after adjusting for ETS or asthma, respectively.

Conclusions

Neighborhood disadvantage was associated with poorer SDB-related QOL and greater SDB symptoms. Associations were partially attenuated after considering the effects of ETS or asthma. The findings support efforts to reduce ETS and neighborhood-level asthma-related risk factors and identify other neighborhood-level factors that contribute to SDB symptom burden as strategies to address sleep-health disparities.

Clinical trial registered with www.clinicaltrials.gov (NCT 02562040).

Pediatric sleep-disordered breathing (SDB), which includes habitual snoring and obstructive sleep apnea (OSA), is highly prevalent: 7–15% of children experience habitual snoring, whereas 1–5% experience OSA (1, 2). Importantly, SDB is associated with significant morbidity, including neurobehavioral impairments (1–3). The pathophysiology of SDB includes upper airway inflammation and abnormalities in upper airway muscle tone and is commonly associated with adenotonsillar hypertrophy. Obesity, asthma, and environmental tobacco smoke (ETS) exposure also have been associated with SDB (4–8), suggesting the potential importance of environmental risk factors in this disorder (9). The finding that SDB disproportionately affects children living in households and neighborhoods with low socioeconomic status (SES) (or “neighborhood disadvantage”) (10–13) further supports the need to consider the role of environmental risk factors in SDB disparities.

Although prior studies have reported associations between neighborhood disadvantage and objective measures of OSA severity (10–13), little is known about the associations between neighborhood-level factors and SDB symptom burden and related quality of life (QOL) in children. Understanding these associations could help identify modifiable neighborhood-level factors that could be improved to benefit SDB-related child outcomes. Furthermore, no studies have included indices of neighborhood disadvantage beyond socioeconomic factors such as poverty and female-headed households. Thus, we aimed to investigate the relationship between multiple indices of neighborhood disadvantage (ND), including social, economic, environmental, and health-related factors relevant to SDB symptom burden, and related QOL among children with mild sleep-disordered breathing (mSDB) enrolled in PATS (Pediatric Adenotonsillectomy Trial for Snoring), a multicenter clinical trial designed to evaluate the effect of early adenotonsillectomy compared with watchful waiting with supportive care to treat mSDB. We hypothesized that, after adjustment for individual-level and household-level risk factors, increased ND would be associated with greater SDB symptom burden and poorer QOL.

Methods

Study Design and Population

We analyzed baseline data collected from the PATS multicenter trial. A detailed description of the recruitment, design, and screening visit process for PATS was reported previously (14). Briefly, children were recruited between 2016 and 2021 from pediatric sleep centers from seven geographic sites in the United States (Boston, MA; Philadelphia, PA; Cincinnati, OH; Cleveland, OH; Ann Arbor, MI; Dallas, TX; and Norfolk, VA) that represented the demographics and diversity of the sites (14). Children ages 3–12.9 years with mSDB were included if parents reported habitual child snoring, they had tonsillar hypertrophy (⩾2 on Brodsky scale 0–4) (15), they were deemed to be an appropriate candidate for adenotonsillectomy by an otolaryngologist, and they had a baseline polysomnogram that demonstrated an obstructive apnea index of <1/h and obstructive apnea–hypopnea index (AHI) of <3/h, with no oxyhemoglobin saturation <90% associated with an obstructive event (14). Children with recurrent tonsillitis, previous tonsillectomy, severe obesity (body mass index [BMI] z-score ⩾ 3), and/or a severe chronic health condition (e.g., sickle cell disorders, severe cardiopulmonary disorders, etc.) were excluded (14).

We performed a cross-sectional analysis of PATS baseline data to test our hypothesis that ND would be associated with SDB-related sleep symptoms and QOL, excluding six children with incomplete outcome data or geocode information. The study protocol was approved by a central institutional review board at the Children’s Hospital of Philadelphia, to which the other sites agreed to cede oversight. Caregivers gave informed consent, and children ⩾7 or 8 years old (depending on the institution) provided informed assent. Data from the Boston site were not included, because that site enrolled only one participant and was closed early.

Main Exposure

Each participant’s residential address was geocoded using the U.S. Census Tract database and Geographic Information System software (ArcGIS Pro 2.8.7 software; Environmental Systems Research Institute). The primary exposure was the census-level neighborhood SES index from the Child Opportunity Index (COI 2.0). The COI comprises 29 variables across three domains (education, health/environment, and social/economic opportunity) and is calculated using data from several sources (diversitydatakids.org) (16, 17). Nationally adjusted total COI scores range from 0 to 100 and are categorically characterized as very low (0–20), low (20–40), moderate (40–60), high (60–80), and very high (80–100). Higher COI scores indicate more advantageous neighborhoods. The distribution for the 29 items that comprise the COI is presented in Table E1 in the data supplement. Neighborhood disadvantage was defined as a COI score ⩽ 40 (including very low or low categories).

Primary Outcomes

The outcomes included scores on the validated caregiver-reported Obstructive Sleep Apnea-18 (OSA-18) QOL questionnaire to reflect SDB-related QOL and the Pediatric Sleep Questionnaire–Sleep-related Breathing Disorder (PSQ-SRBD) scale to index symptom burden. The OSA-18 is a disease-specific QOL survey that captures symptoms across five domains: sleep disturbance, physical suffering, emotional distress, daytime problems, and parent/caretaker concerns (18). With a Likert 7-point scale, caregivers rate the perceived frequency of 18 OSA-related problems ranging from 1 (none of the time) to 7 (all the time). Scores on each item are summed to produce a total score ranging from 18 to 126. Higher scores correspond to poorer SDB-related QOL, with a score ⩾60 signifying a clinically meaningful negative impact of SDB on QOL (19–21). The PSQ-SRBD scale is a 22-item questionnaire that captures symptoms, each individually and together validated as predictors of OSA. The instrument includes three subscales: snoring, daytime sleepiness, and hyperactive behaviors/inattention. The PSQ-SRBD is commonly used to assess SDB risk in pediatric patients but is also increasingly used to assess symptom burden (22). The total score ranges from 0 to 1. Higher scores correspond to greater SDB symptoms, and scores ⩾0.33 indicate clinically relevant high symptom burden and suggest elevated risk for OSA in children.

Covariates

Children’s characteristics included age (yr), sex (male/female), and caregiver-reported race (American Indian or Alaska Native; Asian [including Asian and Indian subcontinent]; Black or African American; Native Hawaiian or other Pacific Islander; White; more than one race; unknown or not reported) and ethnicity (Hispanic), which was dichotomized to compare all minoritized children to White children (reference group). Caregiver-reported maternal education was used as a proxy household SES and dichotomized as “more than a high school degree” and “high school degree or less.” (23) BMI percentile was calculated with participants’ measured height and weight and adjusted for age, sex, and height (24). Urinary cotinine level (ng/ml) was assayed in duplicate at a central laboratory (Abnova Cotinine ELISA kits). Participants were considered exposed to ETS if the average of the two values was ⩾5 ng/ml (25) (n = 82) or if the child’s main caregiver reported smoking at least one cigarette a day (n = 46; when urine cotinine value was unavailable). Asthma was identified by a score of ⩾5 on the eight-item caregiver-reported ISAAC (International Study of Asthma and Allergies in Childhood) questionnaire (95% sensitivity and 100% specificity for identifying asthma) (26).

Statistical Analysis

We used linear regression models to evaluate the association between COI total and subscale scores (independent variables) in separate models for OSA-18 and PSQ-SRBD (dependent variables). In multivariable models, we adjusted for age, sex, minoritized race and ethnicity, maternal education, recruitment site, and season (model 1). Models 2–4 included variables from model 1, with the addition of potential mediators or confounders (BMI percentile, ETS, and asthma) in separate models to explore whether and to what extent each factor may explain associations between ND and OSA-18 or PSQ-SRBD in separate models. Model 5 included the variables in model 1 plus asthma and ETS. The percentage change in COI coefficient estimates from model 1 compared with models including health-related variables were calculated. A change in the coefficient of ⩾10% was considered consistent with confounding or mediation.

Secondary analysis included evaluating the outcomes of the PSQ-SRBD subscales, including snoring, excessive daytime sleepiness, and behaviors (inattentive and/or hyperactive) in models described above. Sensitivity analysis included evaluating model 1 without maternal education as a covariate. We repeated analysis in children without harmonized variable for ETS with parent-reported smoking for missing urinary cotinine measures in 44 children (i.e., ETS was defined as urine cotinine concentration). In addition, we evaluated for any difference in association by recruitment time point before versus after the coronavirus disease (COVID-19) pandemic.

All tests were two-sided at a level of 0.05. Statistical analysis was conducted using R Statistical Software (Version 4.2.1) using the base functions and packages “lubridate” and “season.”

Results

Participant Characteristics

Characteristics of study participants are presented in Table 1. The analytic sample included 453 children from diverse racial and ethnic backgrounds (16% Hispanic, 26% Black or African American, 52% White, and 6% other) with a mean age (standard deviation [SD]) of 6.6 (2.3) years. Nineteen percent (n = 85) of participants reported that the highest level of maternal education was high school. The mean COI (SD) was 50.3 (29.4) (i.e., moderate), and 37% (n = 169) of participants lived in disadvantaged neighborhoods (COI ⩽ 40). Groups defined by COI level differed by race/ethnicity, maternal education, BMI percentile, asthma, and environmental tobacco smoke (P < 0.05) for each univariate comparison.

Table 1.

Sample characteristics, overall and by COI, among participants of the Pediatric Adenotonsillectomy Trial for Snoring study

	Overall (N = 453)	COI
		Very Low to Low: COI ⩽ 40 (n = 169)	Moderate to High: COI > 40 (n = 284)
Demographics
Age, yr, mean (SD)	6.6 (2.3)	6.6 (2.2)	6.6 (2.4)
Sex
Male	225 (50)	78 (46)	147 (52)
Female	228 (50)	91 (54)	137 (48)
Race and ethnicity*
Black or African American	118 (26)	88 (52)	30 (11)
Hispanic	74 (16)	42 (25)	32 (11)
White	234 (52)	32 (19)	202 (71)
Other†	26 (6)	7 (4)	19 (7)
Household SES
Maternal education
Greater than HS degree	368 (81)	115 (68)	253 (89)
HS degree or less	85 (19)	54 (32)	31 (11)
Health factors
BMI percentile, mean (SD)	63.4 (31.2)	70.6 (26.9)	59.1 (32.7)
Asthma (ISAAC score ⩾ 5)*	72 (16)	36 (21)	36 (13)
Environmental tobacco smoke	88 (19)	54 (32)	34 (12)
Sleep symptoms
OSA-18, mean (SD)	52.0 (17.6)	54.5 (17.1)	50.5 (16.1)
OSA-18 ⩾ 60	134 (30)	58 (34)	76 (27)
PSQ-SRBD, mean (SD)	0.5 (0.2)	0.5 (0.2)	0.4 (0.2)
PSQ-SRBD ⩾ 0.33	341 (75)	134 (79)	207 (73)
AHI‡	0.77 (0.71)	0.80 (0.69)	0.75 (0.73)
Recruitment site
Ann Arbor, MI	84 (19)	15 (9)	69 (24)
Cincinnati, OH	84 (19)	28 (17)	56 (20)
Cleveland, OH	64 (14)	23 (14)	41 (14)
Dallas, TX	88 (19)	50 (30)	38 (13)
Norfolk, VA	66 (15)	19 (11)	47 (17)
Philadelphia, PA	67 (15)	34 (20)	33 (12)

Definition of abbreviations: AHI = apnea–hypopnea index; BMI = body mass index; COI = Child Opportunity Index; HS = high school; ISAAC = International Study of Asthma and Allergies in Childhood questionnaire; OSA-18 = Obstructive Sleep Apnea-18 questionnaire; PSQ-SRBD = Pediatric Sleep Questionnaire–Sleep-related Breathing Disorder Scale; SD = standard deviation; SES = socioeconomic status.

Data are presented as n (%) unless otherwise noted.

^*Missing data: race and ethnicity n = 1, asthma n = 1.

^† Other includes American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, more than one race, and unknown or not reported—due to paucity of participants in these categories in our sample.

^‡ AHI defined as respiratory events associated with 3% desaturation or arousals.

Associations between SDB-related QOL and Neighborhood Factors

Thirty percent (n = 134) of participants had an elevated OSA-18 score (⩾60), indicating a significant SBD-related negative impact on QOL, including 34% (n = 58/169) of those with COI ⩽ 40 and 27% (n = 76/284) of those with COI > 40 (Table 1). Figure 1 provides the estimated β coefficients and 95% confidence intervals (CIs) from models of the association between COI and OSA-18, adjusted for covariates in model 1. A neighborhood COI score increase of 1 SD was associated with an improvement in QOL on the OSA-18 score by 2.48 points (95% CI, −4.34 to −0.62; P = 0.01). Significant associations were also observed between the OSA-18 score and the COI subdomains (neighborhood-level education, health/environment, and social/economic opportunities), such that more advantageous neighborhood subdomains were linked to better QOL.

Figure 1.

β estimates for the relationship between neighborhood factors, Child Opportunity Index (COI) total score and subdomains, and sleep outcomes. Each variable was analyzed in separate models and adjusted for age, sex, minoritized race/ethnicity background, maternal education, recruitment site, and season. The β coefficient reflects a 1–standard deviation change in the COI scores and three subdomains (29.4, 23.3, 27.3, and 30.7 points, respectively).

Associations between SDB Symptom Burden and Neighborhood Factors

Seventy-five percent (n = 341) of participants had an elevated PSQ-SRBD total score (⩾0.33), signifying elevated SDB-related symptoms, including 79% (n = 134/169) of those with COI ⩽ 40 and 73% (n = 207/284) of those with COI > 40 (Table 1). Figure 1 provides the estimated β coefficients from models of the association between COI score and SDB symptom burden on the PSQ-SRBD, adjusted for covariates in model 1. Similar to the QOL analyses, more advantageous neighborhood-level opportunities were associated with less SDB symptom burden. Specifically, for every increase of 1 SD in neighborhood COI score, the PSQ-SRBD score scale improved by 0.03 points (β = 0.028; 95% CI, −0.05 to −0.01; P = 0.01). Comparable associations were observed for the COI subdomains.

Associations between PSQ-SRBD Subscales and Neighborhood Factors

In the adjusted analysis, the effect size of the Sleepiness Subscale was similar to results with the PSQ-SRBD Total Scale in model 1, with a decrease in score by 0.03 points (β = 0.035; 95% CI, −0.073 to −0.003; P = 0.07) for every 1-SD increase in the COI score. The estimate was −0.02 for the other individual subscales for Snoring and Behavior (β = 0.021; 95% CI, −0.054 to 0.011; P = 0.20; and β = 0.022; 95% CI, −0.060 to 0.015; P = 0.25, respectively) (Table E2).

Relationship with Health-related Factors

Table 2 shows the changes in variable estimates after adjustment for BMI percentile, ETS, and asthma, which are factors that may potentially influence associations with SDB symptom burden and QOL and are often associated with neighborhood disadvantage. After controlling for ETS or asthma, the association between total COI score and OSA-18 score attenuated but remained statistically significant. Specifically, the magnitude of the coefficient relating COI to OSA-18 declined by 23% (from −2.48 to −1.92) and by 10% (from −2.48 to −2.23), respectively, after considering ETS or asthma. In contrast, the inclusion of BMI percentile did not change the strength of the observed association between COI score and OSA-18 score. For symptom burden, the inclusion of ETS in the model resulted in a significant decrease of 14% in the magnitude of association between COI and PSQ-SRBD (from −0.028 to −0.024), and for asthma, it attenuated the association by 10% (from 0.028 to 0.025). The inclusion of BMI percentile did not change the strength of the association between COI and PSQ-SRBD.

Table 2.

Estimated β coefficients for the strength of association between COI, OSA-18 questionnaire, and Pediatric Sleep Questionnaire–Sleep-related Breathing Disorder scores in models adjusted for potential confounders or mediators

Models	OSA-18* β (95% CI), P Value	% Change (from Model 1)	PSQ-SRBD† β (95% CI), P Value	% Change (from Model 1)
	Unadjusted		−2.04 (−3.56 to −0.52), 0.01		—	−0.026 (−0.043 to −0.009), 0.003	—
Primary adjusted (model 1†)	−2.48 (−4.34 to −0.62), 0.01	—	−0.028 (−0.048 to −0.007), 0.009	—
Additionally adjusted COI models
Model 2: model 1 + BMI percentile	−2.48 (−4.35 to −0.61), 0.01	0	−0.027 (−0.048 to −0.006), 0.01	2
Model 3: model 1 + ETS	−1.92 (−3.85 to 0.02), 0.05	23	−0.024 (−0.045 to −0.002), 0.03	14
Model 4: model 1 + asthma	−2.23 (−4.08 to −0.38), 0.02	10	−0.025 (−0.045 to −0.004), 0.02	10
Model 5: model 1 + ETS, asthma	−1.69 (−3.61 to 0.23), 0.09	32	−0.022 (−0.043 to −0.001), 0.05	21

Definition of abbreviations: BMI = body mass index; CI = confidence interval; COI = Child Opportunity Index; ETS = environmental tobacco smoke; OSA-18 = Obstructive Sleep Apnea-18 questionnaire; SD = standard deviation; PSQ-SRBD = Pediatric Sleep Questionnaire–Sleep Related Breathing Disorder Scale.

^*The β coefficient reflects a 1-SD change in the COI score (29.4 points).

^† Model 1: adjusted for age, sex, minoritized race and ethnicity background, maternal education, recruitment site, and season.

Sensitivity analysis included removing maternal education from model 1, defining ETS with objective measure of urine concentration only, and adding a time point variable (before vs. after COVID) to model 1, and showed no substantive change in associations (Table E3).

Discussion

This study evaluated the relationship between a multidimensional index of neighborhood disadvantage and measures of SDB-related QOL and symptom burden in children with mild SDB. Our findings reveal that greater neighborhood disadvantage across education, health, and environment and social and economic subdomains is associated with poorer SDB-related QOL and increased SDB symptom burden in children with mSDB from multiple centers in the United States. These associations persisted after adjustment for maternal education, minoritized race and ethnicity background, and several factors common in low-income neighborhoods (i.e., BMI percentile, ETS, and asthma). The magnitude of identified associations was not high enough to suggest that major impact on a given individual is likely or common, but their existence on a population level could implicate important societal influences that in aggregate—if reflective of causal underlying relationships—could have substantive adverse impact.

Prior studies have shown an association between neighborhood disadvantage and increased SDB prevalence, as well as an association with greater OSA severity (10–13). In CHAT (Childhood Adenotonsillectomy Trial), a multicenter clinical trial designed to evaluate early adenotonsillectomy in children with OSA, cross-sectional analysis of nearly 800 children revealed that neighborhood-level indicators of lower SES, including higher poverty rate, the proportion of single female–headed households with children, and the unemployment rate, were associated with increased severity of OSA as measured by a higher AHI (12). Other studies also reported decreased green space access, low ambient air quality (i.e., traffic-related pollution, distance to major roadways), and increased chemical pollution (from industrial plants, hazard waste dumps, etc.) to be associated with more severe OSA (10–13, 27–29).

Our findings add to the existing literature by highlighting associations between neighborhood disadvantage and SDB-related functioning, including symptom burden and QOL impact measured by validated instruments. This finding is important, as polysomnography indices do not fully reflect the clinical manifestations and symptom burden experienced by children with SDB (30). Therefore, patient-reported measures of symptom burden and QOL are key outcomes that are increasingly being used clinically in the evaluation of patients (23, 31, 32) and in conjunction with polysomnography to determine treatment efficacy (19, 20).

Although previous studies have used census tract information to describe socioeconomic features of neighborhoods (i.e., poverty rates, proportions of female-headed households, population densities, median family incomes, etc.), our study used the COI measure, which includes additional neighborhood factors specific to child development, health, and well-being. For example, the COI includes measures such as access to quality early childhood education, green space, healthy food, toxin-free environments, and walkability (diversitydatakids.org), among other factors (16, 17). The inclusion of factors relevant to child development can yield crucial information for neighborhood and policy-related health promotion efforts. The sum of evidence suggests that neighborhood disadvantage results in increased OSA prevalence and severity through differential access to material resources and adverse environments. Efforts to improve the neighborhood environment could benefit both QOL and symptom burden among children with mSDB.

The relationship between SDB-related QOL and neighborhood disadvantage substantially decreased after we controlled for ETS and, to a lesser extent, asthma. This suggests that potentially modifiable child and family exposures may contribute to QOL and symptom burden in children with mSDB. The observed association of ETS with symptoms and QOL in children with mSDB is consistent with findings that ETS was associated with an approximately 20% higher obstructive AHI in children with OSA (33). The mechanisms linking ETS and SDB are not well understood but could relate to inflammation, as suggested by a mediation analysis in toddlers showing that 18% of the relationship between ETS and SDB was explained by chronic inflammation as measured by serum C-reactive protein levels (34).

Asthma is a common chronic inflammatory respiratory disorder and has a bidirectional relationship with SDB (4, 5). Exposure to environmental pollutants promotes upper and lower airway inflammation, decreases airway patency, and increases both asthma symptoms and the frequency of respiratory infections (34–36). Chronic exposure to poor air quality may exacerbate SDB in a similar fashion. It is, therefore, plausible that many neighborhood risk factors that are associated with asthma (e.g., air pollution, pest and allergen exposures, ETS, and obesogenic environments) also increase upper airway inflammation, which may worsen SDB symptoms (37).

Both asthma and ETS are more prevalent in communities with neighborhood disadvantage (38–40). If there is an underlying cause-and-effect relationship, these findings would suggest that strategies focused on reduction in ETS and other asthma-related risk factors could decrease SDB-related symptoms. However, the persistence of an association between COI and symptoms after adjustment for both ETS and asthma suggests that there are additional neighborhood features that contribute to SDB symptoms.

Additional mechanisms that explain the relationship between neighborhood disadvantage and SDB are not well understood; however, potential factors may include less access to healthy food, limited opportunities for safe physical activity, and exposure to poor air quality, all of which may promote nasal and upper/lower airway inflammation and reduce airway patency (13, 34–38). Diets higher in red or processed meats and lower in whole grains are proinflammatory and are associated with increased OSA risk in adults (41). Low neighborhood walkability and decreased physical activity are associated with increased severity of OSA in adults (42–44). Increased physical activity in sedentary adults is associated with decreased OSA severity independent of change in body weight (45). Air pollution, whether measured by a composite index of multiple pollutants (46) or by neighborhood traffic patterns (27, 47), is suggested to be a risk factor for SDB. In the United States, neighborhood disadvantage and toxic environmental exposures (i.e., air pollutants and ETS) are higher in minoritized communities (40, 48). This suggests that racist and discriminatory historical policies that promoted neighborhood segregation continue to affect marginalized communities and contribute to health disparities (49). Further research is needed to determine if there are similar associations between neighborhood walkability, physical activity, and diets with SDB in children and whether these features of the neighborhood may, in part, explain the higher SDB symptom burden observed in children from disadvantaged neighborhoods.

Strengths and Limitations

Study strengths include a diverse study population of participants from multiple geographic regions and the use of validated caregiver-reported measures of symptoms and disease-specific QOL. We focused on children with snoring and mild SDB, which is prevalent in children and can lead to behavioral morbidity (50). The study is limited by its cross-sectional analysis, although it is unlikely that there is reverse causality (i.e., childhood SDB symptoms are not likely to influence residential characteristics). Despite using a validated, multidimensional index of neighborhood disadvantage, we did not have direct measures of indoor and outdoor air quality. Participants in this study were symptomatic children who were candidates for adenotonsillectomy; therefore, it is possible that the relationship is different in less-symptomatic snoring children. As this analysis focuses on baseline PATS data, it remains unknown how ND impacts symptom burden and QOL scores after treatment with adenotonsillectomy. Future studies are needed to determine whether ND leads to poorer QOL outcomes in children with mSDB who undergo adenotonsillectomy.

Conclusions

ND is associated with poorer SDB-related QOL and greater symptom burden in children with mild SDB across multiple U.S. cities, even after accounting for multiple child and family factors. The observed relationship was attenuated after considering asthma and ETS exposure, suggesting opportunities to target ETS exposure and asthma-related risk factors as strategies for decreasing sleep health disparities and possibly to combine asthma and SDB prevention initiatives. Further research is needed to more clearly define the specific neighborhood environmental risk factors (e.g., chemical pollutants, housing factors, green space) that may worsen SDB-related symptoms and clarify the mechanistic pathways underlying these associations.