Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Nov 21.
Published in final edited form as: Int J Hyg Environ Health. 2025 Aug 5;269:114638. doi: 10.1016/j.ijheh.2025.114638

Air pollution exposures and adverse childhood experiences in relation to sleep health in middle childhood☆,☆☆

Jonika B Hash a, Logan C Dearborn b, Christine T Loftus b,*, Catherine J Karr c, Adam A Szpiro d, Emily S Barrett e, Kaja Z LeWinn f, Ruby Nguyen g, Paul E Moore h, Brent Collett i, Amanda N Noroña-Zhou j, Nicole R Bush j, Sheela Sathyanarayana i
PMCID: PMC12633831  NIHMSID: NIHMS2104123  PMID: 40768818

Abstract

Purpose/aims:

Sleep health is an understudied but potentially important outcome of joint air pollution and psychosocial stress exposures in children. This study examined children’s sleep health outcomes in relation to air pollution (PM2.5, NO2, O3; aim 1), adverse childhood experiences (ACEs; aim 2), and air-pollution-by-ACEs interactions (aim 3).

Methods:

Participants were from ECHO-PATHWAYS, a three-cohort consortium. Aim 1 included 1166 participants across the three cohorts, and aims 2 and 3 included a subset of 719 participants from a single cohort. PM2.5 (μg/m3), NO2 (ppb), and O3 (ppb) were estimated during early infancy (0–6 months) and early childhood (6 months–6 years) using geocoded residential histories and spatiotemporal prediction models. Children’s lifetime exposures to 8 different types of ACEs were measured via parent report at child age 8–9 years. Sleep disturbance and sleep-related impairment outcomes were measured via children’s self-report at age 8–9 years. Analyses included linear regressions, adjusting for a priori-selected confounders.

Results:

Aim 1 results showed that, for every 1 IQR increase in early infancy NO2, children scored 0.31 (95 % CI 0.01, 0.61) points lower on sleep-related impairment. Aim 3 results showed that, for every additional ACE, the difference in sleep-related impairment per IQR increase in early infancy and early childhood NO2 was 0.43 (95 % CI 0.08, 0.78) and 0.41 (95 % CI 0.08, 0.73), respectively (psinteractions = 0.02). No other associations were observed.

Conclusion:

We found little evidence of associations, with the exception of suggestive evidence for associations of NO2 and NO2-by-ACE interactions with sleep-related impairment.

Keywords: Adverse childhood experiences, Air pollution, Particulate matter, Gaseous components, Sleep health, Children

1. Introduction

Sleep health is a concept characterized by aspects of sleep that are relevant to all individuals (Buysse, 2014; Meltzer et al., 2021), and as such, is of public health importance (Hale et al., 2020). In children, sleep health is defined by six dimensions including sleep duration, efficiency (ease of initiating and maintaining sleep), daytime sleepiness/alertness, timing of sleep, subjective satisfaction/quality, and sleep-related behaviors (Meltzer et al., 2021). Sleep health is foundational for children’s broader health and development, including weight regulation (Magee and Hale, 2012; Matricciani et al., 2019; Quist et al., 2016), cardiometabolic health (Agostini and Centofanti, 2021; Quist et al., 2016), school performance (Agostini and Centofanti, 2021; Chaput et al., 2016), and emotional regulation (Chaput et al., 2016; Matricciani et al., 2019). In alignment with the public health importance of sleep health, the United States (U.S.) Department of Health and Human Services Healthy People (2030) initiative aims to increase the proportion of adults, high schoolers, and children who obtain sufficient sleep (U.S. Department of Health and Human Services, n.d.).

Environmental exposures, including air pollution, are increasingly recognized as potential contributors to poor sleep health (Billings et al., 2020; Cao et al., 2021; Liu et al., 2020, 2021; Mayne et al., 2021). Air pollution is made up of a mixture of components, including gases such as carbon monoxide (CO), sulfur dioxide (SO2), ozone (O3), and nitrogen dioxide (NO2); particulate matter of various sizes (coarse, fine, and ultrafine); metals such as lead (Pb); and organic compounds such as volatile organic compounds (VOCs; Costa et al., 2014). While the toxicological properties of specific air pollutants vary (Brumberg and Karr, 2021), air pollution is generally implicated as a neurotoxicant (Costa et al., 2014, 2019); and sleep, which is fundamentally an activity of the brain (Dahl, 1996), may be vulnerable to the neurotoxic effects of air pollution (Cao et al., 2021; Liu et al., 2020, 2021; Mayne et al., 2021). A small but growing number of studies in children show that air pollution is associated with difficulties initiating and maintaining sleep (Abou-Khadra, 2013; Gui et al., 2024; Gui et al., 2024; Lawrence et al., 2018), sleep duration (Bose et al., 2019; Gui et al., 2024; Wang, Gueye-Ndiaye, et al., 2025), variability in sleep duration (He et al., 2023), and daytime sleepiness (Cai et al., 2022). The most well-studied pollutant of potential concern for sleep health in children is particulate matter (coarse and fine) (Abou-Khadra, 2013; Bose et al., 2019; Cai et al., 2022; Gui et al., 2024; He et al., 2023; Lawrence et al., 2018), but NO2 (Lawrence et al., 2018; Wang, Gueye-Ndiaye, et al., 2025), SO2 (Lawrence et al., 2018), and O3 (Gui et al., 2024; Lawrence et al., 2018) may also be of concern. Mechanisms are not well understood, but air pollution could potentially target the neural regulation of sleep via oxidative stress, neuroinflammation, and direct passage of ultrafine particles into the brain (Cao et al., 2021; Costa et al., 2014, 2017, 2019; Genc et al., 2012; Payne-Sturges et al., 2019). Air pollution may also induce alterations in neurotransmitters involved in regulating sleep (Chuang et al., 2018; Liu et al., 2021).

Psychosocial stressors, in parallel, also contribute to poor sleep health (Brown et al., 2022; Kajeepeta et al., 2015; Oh et al., 2018) and could amplify the neurotoxicity of air pollutants (Block et al., 2012; Gee and Payne-Sturges, 2004; Morello-Frosch and Shenassa, 2006; Padula et al., 2020). According to chemical-by-non-chemical stress interaction frameworks, psychosocial stressors take a chronic “wear and tear” toll on neural, endocrine, and immune systems – a concept termed allostatic load (Danese and McEwen, 2012) – and this “wear and tear” could, in turn, lower the body’s defenses against environmental toxicants (Gee and Payne-Sturges, 2004; Morello-Frosch and Shenassa, 2006). Psychosocial stressors and air pollution may also operate on the same or similar mechanistic pathways, including those involved in the neural, endocrine, and immune systems that are vulnerable to allostatic load (Block et al., 2012; McEwen and Tucker, 2011; Snow et al., 2018), and exposure to both types of stressors could act as “multiple hits” that compromise these systems and exacerbate risk (Cory-Slechta, 2005).

Adverse childhood experiences (ACEs), including maltreatment, household dysfunction, and neighborhood/community-level stressors (Cronholm et al., 2015; Felitti et al., 1998), are types of psychosocial stressors that have the potential to trigger allostatic load (Danese and McEwen, 2012; Finlay et al., 2022; McEwen and Tucker, 2011). ACEs, like air pollution, disrupt neurodevelopment (Shonkoff et al., 2009) and are associated with a broad range of poor mental health outcomes across the lifespan (Anda et al., 2006; Chapman et al., 2004; Petruccelli et al., 2019) including, but not limited to, poor sleep (Brown et al., 2022; Chapman et al., 2013; Chapman et al., 2011; Hambrick et al., 2018; Harada et al., 2021; Hash et al., 2019; Kajeepeta et al., 2015; Lewis-de Los Angeles, 2022; Lin et al., 2022; Oh et al., 2018; Rojo-Wissar et al., 2021; Sterling et al., 2021; Wang et al., 2016; Yu et al., 2022). It is possible that joint exposure to ACEs and air pollution could heighten risk for poor sleep health in childhood, but no studies to our knowledge have tested this possibility. Evidence of air pollution-by-ACEs interactions is accumulating in other similar literatures, however. A review by Padula and colleagues (2020), for example, identified several studies in which prenatal psychosocial stress exposures exacerbated risks associated with prenatal air pollution exposures, including risks of poorer birth outcomes, poorer offspring neurodevelopment, and offspring asthma.

Additionally, the neurotoxicity of air pollution may vary by age of exposure (Costa et al., 2014, 2019). During the first 6 months of life, the neural networks that regulate children’s sleep may be more vulnerable to neurotoxicants, as this is a time in which sleep architecture undergoes profound developmental changes. Key changes include a shift from rapid eye movement (REM)-like sleep at sleep onset to non-rapid eye movement (NREM) sleep at sleep onset, as well as the emergence of distinct NREM sleep stages (Bathory and Tomopoulos, 2017; Lenehan et al., 2022; Mindell and Owens, 2015; Sheldon, 2014). These changes are dramatic and could indicate a uniquely vulnerable window of central nervous system reorganization and maturation (Sheldon, 2014). Additionally, the blood-brain barrier, which protects the central nervous system from environmental toxicants (Block et al., 2012), may not be fully developed at birth (Costa et al., 2014; Lam et al., 2015).

Taken together, the above literatures raise concern for sleep-related neurotoxicity of air pollution, particularly during the first 6 months of life, and for ACEs to amplify the potential sleep-related neurotoxicity of air pollution. Despite these concerns, relationships between air pollution and sleep health in children are still poorly understood, with only a handful of studies to date examining air pollution in relation to children’s sleep health outcomes in general (Abou-Khadra, 2013; Bose et al., 2019; Cai et al., 2022; Gui et al., 2024; Gui et al., 2024; He et al., 2023; Lawrence et al., 2018; Wang, Gueye-Ndiaye, et al., 2025), and no studies examining air pollution-by-ACEs interactions in relation to sleep health in children. To address these gaps, we examined air pollutants of concern (fine particulate matter/PM2.5, NO2, O3) and ACEs in relation to children’s sleep health outcomes in middle childhood among a large sociodemographically diverse sample. Specifically, we examined independent associations of the air pollutants (Aim 1; “air pollution models”), independent associations of ACEs (Aim 2; “ACEs models”), and air pollution-by-ACEs interactions (Aim 3; “interaction models”) in relation to children’s sleep health. Additionally, given the potential for distinct sleep-related developmental vulnerability to air pollution, we examined air pollution exposures during the first 6 months of life (“early infancy” window) and, separately, exposures from 6 months to 6 years (“early childhood” window representing chronic, long-term exposure), to assess for a sensitive period. We hypothesized that greater air pollution exposures and more ACEs would be associated with poorer sleep health outcomes. We also hypothesized that associations between greater air pollution exposures and poorer sleep health would be stronger at higher levels of ACEs than at lower levels of ACEs.

2. Methods

2.1. Study design, setting, and participants

The study sample for this analysis was drawn from the ECHO prenatal and early childhood pathways (ECHO-PATHWAYS) study, which includes participants from three pregnancy cohorts: The Conditions Affecting Neurocognitive Development and Learning in Early Childhood (CANDLE) study, The Infant Development and Environment Study (TIDES), and the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth (PW-GAPPS) study. Detailed information about these three cohorts is provided elsewhere (Barrett et al., 2014; GAPPS, n.d.; LeWinn et al., 2022; Sontag-Padilla et al., 2015), but we briefly summarize pertinent details below.

The CANDLE cohort includes 1503 participants recruited during the second trimester of pregnancy. Participants were eligible if they resided in Shelby County of Memphis, Tennessee; were 16–40 years of age; spoke English; and had a singleton, low medical risk pregnancy that they planned to deliver at one of five hospitals in Shelby County. The CANDLE cohort was recruited between 2006 and 2011 from obstetric sites, including a hospital obstetric clinic and community obstetric practices.

The TIDES cohort includes 900 participants enrolled during the first 13 weeks of pregnancy. Participants were eligible if they were ≥18 years of age; spoke English or Spanish; were in their first 13 weeks of a singleton pregnancy; intended to give birth at a study hospital; and were not experiencing a serious threat to the pregnancy. The TIDES cohort was enrolled between 2010 and 2012 from four university-based prenatal clinics located in San Francisco, California; Rochester, New York; Minneapolis, Minnesota; and Seattle, Washington.

The PW-GAPPS cohort includes 662 participants who had previously participated in a larger GAPPS pregnancy cohort study. PW-GAPPS participants were initially recruited into the larger GAPPS study at various times during pregnancy from two study hospitals located in Seattle, Washington and one hospital located in Yakima, Washington; were older than 18 years of age at the time of pregnancy; spoke English; had consented to be contacted about PW-GAPPS; and had previously shared demographics, health information, and biospecimens with the larger GAPPS biorepository between 2011 and 2016.

Data from these three cohorts were harmonized, allowing for pooled three-cohort (the PATHWAYS cohort) analyses. All data relevant to the present analyses were available in the pooled PATHWAYS cohort, with the exception of ACEs data, which was unique to the CANDLE cohort. Thus, based on data availability, we drew the present study sample from the pooled PATHWAYS cohort for aim 1 (air pollution models) and from the CANDLE cohort only for aims 2 and 3 (ACEs and interaction models, receptively). Specifically, participants from the pooled PATHWAYS cohort (N = 3065) were included in the present study sample if, for aim 1, they had completed sleep questionnaires at age 8–9 years, were not missing primary covariate information, and had valid air pollution exposure data derived between ages 0–6 months (n = 1166). Those who additionally completed a childhood ACEs questionnaire when their child was aged 8–9 years were included in aims 2 and 3 (n = 719; CANDLE cohort only). When assessing air pollution exposures averaged between 6 months and 6 years, the population consisted of 1122 participants for the aim 1 air pollution models and 695 participants for the aim 3 interaction models (Fig. 1).

Fig. 1.

Fig. 1.

Open in a new tab

Study participant inclusion flowchart.

Note. ACEs = adverse childhood experiences. Pooled PATHWAYS = combined sample of CANDLE, TIDES, and PW-GAPPS cohorts: CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study; TIDES = The Infant Development and Environment Study; PW-GAPPS = the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study. Aims 2 and 3 are from the CANDLE cohort only. For the Aim 1 pooled PATHWAYS sample, the 242 excluded due to missingness on main covariates were missing one or more of the following: child race (n = 19), screen use (n = 54), parental education (n = 5), household income (n = 96), household size (n = 74), attention-deficit/hyperactivity disorder medication use (n = 71), and child ethnicity (n = 19). For the Aims 2 and 3 CANDLE-only sample, the 98 missing on main covariates were missing one or more of the following: child race (n = 1), screen use (n = 22), parental education (n = 1), household income (n = 64), household size (n = 42), attention-deficit/hyperactivity disorder medication use (n = 5), and child ethnicity (n = 1).

a sample sizes are provided with respect to availability of air pollution data during the early infancy period (0–6 months). Sample sizes for analyses examining air pollution exposure during the early childhood period (6 month− 6 years) were n = 1122 and n = 695 for the aim 1 air pollution and aim 3 interaction models, respectively.

All research activities for this secondary data analysis were approved by the Institutional Review Board (IRB) of the University of Washington and local IRBs at each data collection site, and all participants provided informed consent at the time of enrollment in each cohort.

2.2. Outcome assessment

2.2.1. Sleep disturbance

At age 8–9 years, children completed the Patient-Reported Outcomes Measurement Information System (PROMIS) Sleep Disturbance 4a questionnaire (Bevans et al., 2018; Forrest et al., 2018; PROMIS, 2023). The PROMIS Sleep Disturbance 4a questionnaire is a 4-item retrospective (about the past 7 days) child self-report tool indicating difficulty initiating sleep, difficulty maintaining sleep, and subjective sleep quality. Items are scored on a 5-point rating scale ranging from 1 = never to 5 = always. Positively-worded items are reverse scored, and items are summed for a total raw score (range 4–20). Higher scores indicate greater sleep disturbance. The PROMIS Sleep Disturbance 4a questionnaire was developed and tested in accordance with the National Institutes of Health PROMIS program approach, and the sleep disturbance items demonstrate content, structural, and convergent validity (Bevans et al., 2018; Forrest et al., 2018). Specifically, extensive testing supports the PROMIS Sleep Disturbance 4a questionnaire’s content validity (via parent, child, and expert input; cognitive debriefing), structural validity (via factor analysis), and convergent validity (via correlations with other parent- and child-reported sleep health and fatigue questionnaires) (Forrest et al., 2018).

2.2.2. Sleep-related impairment

At age 8–9 years, children completed the PROMIS Sleep-Related Impairment 4a questionnaire (Bevans et al., 2018; Forrest et al., 2018; PROMIS, 2023). The PROMIS Sleep-Related Impairment 4a questionnaire is a 4-item retrospective (about the past 7 days) child self-report tool indicating daytime sleepiness and effects of sleep on daytime functioning (cognitive functioning and activities). Items are scored on a 5-point scale ranging from 1 = never to 5 = always, and items are summed for a total raw score (range 4–20). Higher scores indicate greater sleep-related impairment. As with the PROMIS Sleep Disturbance questionnaire, development of the PROMIS Sleep-Related Impairment questionnaire followed the National Institutes of Health PROMIS program approach, and the sleep-related impairment items demonstrate content, structural, and convergent validity via extensive testing in a manner similar to the PROMIS Sleep Disturbance questionnaire discussed above (Bevans et al., 2018; Forrest et al., 2018).

2.3. Exposure assessment

2.3.1. Air pollution (PM2.5, NO2, and O3)

PM2.5 (μg/m3), NO2 (ppb), and O3 (ppb) exposures were each estimated using predictions of outdoor ambient pollutant concentrations from geocoded residential histories using separate validated, point-resolution spatiotemporal models (Keller et al., 2015; Kirwa et al., 2021; Wang et al., 2018). In brief, models utilized measured pollutant concentrations and over 200 linked geographic covariates from a combination of over 4000 external research campaign and regulatory monitors using dimension reduction via partial least squares. Observed time series of regulatory pollutant data were used to estimate time trends using a spatiotemporal generalization of universal kriging. Pollutant concentrations at point locations and a two-week resolution were averaged from date of birth to 6 months and from 6 months to 6 years to characterize exposure in both the early infancy and early childhood windows, respectively. In the case of participant change of address in a given window, the average exposure was calculated as time-weighted across all reported addresses in that period. Although there is often a trade-off in air pollution modeling between spatial and temporal resolution, with many daily models utilizing a grided temporal surface that may average across small-scale changes in pollution concentrations related to roadways and other geographies (Wang, Young, et al., 2025), we opted to prioritize spatial resolution given our aggregation of exposure periods across months and years.

2.3.2. Adverse childhood experiences (ACEs)

At child age 8–9 years, parents reported their child’s lifetime history of adversities on an 8-item retrospective questionnaire adapted from the Centers for Disease Control’s National Survey of Children’s Health (Kwong and Hayes, 2017). Parents endorsed (0 = no, 1 = yes) whether their child had ever experienced (“thinking about your child’s entire lifetime”) each of 8 types of ACEs: 1) mental illness of a household member, 2) substance abuse in the household, 3) incarceration of a household member, 4) caregiver divorce or separation, or loss of a caregiver, 5) domestic violence, 6) witnessing or directly experiencing neighborhood violence, 7) economic hardship, such as difficulty covering food or housing, and 8) racial or ethnic discrimination. Scores were summed for a total ACE score indicating the number of types of ACEs experienced (range 0–8).

2.4. Covariates

The selection of model covariates was performed a priori and was guided by the research literature (Belmon et al., 2019; Franco et al., 2020; Guglielmo et al., 2018; Newton et al., 2020; Nixon et al., 2008). Covariates included study site, child sex assigned at birth (male, female), child age at the time of the sleep outcome assessment (years), highest level of the pregnant parent’s education at enrollment (< high school, high school or equivalent, 2- or 4-year college degree, graduate or professional degree), region- and inflation-adjusted household income at child age 4–6 years (U.S. dollars), household size at child age 4–6 years (<4, 4, 5, >5), child race (reported by the parent as American Indian or Alaska Native, Asian, Black, Native Hawaiian or other Pacific Islander, White, multiracial, and another race; collapsed into Black, White, and another race, due to small ns in some categories), child ethnicity (reported by the parent as Hispanic or non-Hispanic), summer v. other season at the time of the sleep outcome assessment, urban v. rural census classification at the time of the sleep outcome assessment, average screen use in the past 30 days (<1 h, 1 h, 2 h, 3 h, 4 h, ≥5 h), report of attention-deficit/hyperactivity disorder (ADHD) medication use in the past 30 days (yes, no), date of birth splines (natural splines with 1 degree of freedom for each year), and the neighborhood deprivation index (NDI). In the aim 1 air pollution and aim 3 interaction models, we used the NDI averaged over the same time periods as the air pollutants (0–6 months and 6 months–6 years); and for interpretability and consistency, we used the NDI averaged over the 6 month–6 year time period in the aim 2 ACEs models. We included race and ethnicity as confounding variables because, due to systemic racism, minoritized racial and ethnic populations experience disproportionately greater exposures to environmental health stressors (Kaufman and Hajat, 2021; Lane et al., 2022), psychosocial stressors (Kim et al., 2020; Merrick et al., 2018), and sleep health disparities (Guglielmo et al., 2018; Johnson et al., 2019; Smith et al., 2019). We used the rural v. urban census classification as a proxy for noise exposure.

2.5. Statistical analyses

Descriptive statistics were used to examine the study population characteristics within the pooled PATHWAYS and CANDLE only samples. Linear regression with robust standard error estimators was used to examine associations between continuous exposures and sleep outcomes in all models. Multiplicative interaction terms were utilized to assess effect modification in interaction models. A staged modeling approach that included a minimally-adjusted and a main model was utilized for covariate adjustment in order to assess the sensitivity of results to the level of adjustment. The minimally-adjusted model included child age and child sex (all aims), and due to multi-site representation in the pooled PATHWAYS sample, the aim 1 air pollution models additionally included study site in the minimally-adjusted model. In the main model, we adjusted for all potential confounders (predictors of the outcome likely to be a cause or correlate of the exposure) and precision variables (predictors of the outcome) for which we had high-quality data. Specifically, the main model for all aims included the covariates from the minimal model plus parental education, household income adjusted for annual regional price parity (non-normally distributed; treated as continuous to allow for interaction with household size), household size, a term interacting household income with household size, child race, child ethnicity, summer v. other season, urban v. rural census classification, average screen use in the past 30 days, report of ADHD medication use in the past 30 days, and the NDI. To account for long-term temporal trends in air pollution, date of birth splines (with 1 degree of freedom per year of birth) were additionally included in the main models with air pollution as an exposure variable (aim 1 air pollution and aim 3 interaction models).

Several sensitivity analyses were conducted to assess the robustness of the results to our modeling approach. Sensitivity analyses for the aim 1 air pollution models included: 1) assessment of potential co-pollutant confounding, by mutually adjusting for PM2.5, NO2, and O3 in the same model; 2) assessment of potential confounding by associations in other exposure windows, by co-adjusting for the air pollution exposure in both the early infancy and early childhood windows, for each pollutant; 3) evaluation of robustness of findings to site and cohort inclusion, by means of a leave-one-out analysis with the systematic omission of each cohort, as well as a CANDLE only analysis using the aim 3 CANDLE only sample, 4) a model removing ADHD medication, and 5) general additive models exploring potential non-linear relationships between air pollution and sleep health. Sensitivity analyses for the aim 2 ACEs models included reparameterization of reported ACEs into score categories of 0, 1, 2, 3, and 4+ ACEs. Likewise, for the aim 3 interaction models, we repeated analyses using categorical ACEs. Additionally, for all three aims, we conducted a sensitivity analysis that excluded children with neurodevelopmental conditions, as sleep problems among children with neurodevelopmental conditions are common, complex, and likely multifactorial in etiology (Belli et al., 2022; Gregory and Sadeh, 2016). We defined neurodevelopmental conditions as ever having a diagnosis of autism spectrum disorder (ASD), ever having a diagnosis of ADHD, and/or having a t-score of >63 on the Child Behavioral Checklist (CBCL) internalizing or externalizing problem scales at age 8–9 years. All sensitivity analyses were assessed adjusting for the main set of covariates. All analyses were conducted using R 4.2.2.

3. Results

3.1. Sample characteristics

The left column of Table 1 shows characteristics for the aim 1 (pooled PATHWAYS) sample. This sample was comprised of 62.0 % CANDLE, 28.4 % TIDES, and 9.6 % PW-GAPPS participants. Mean child age at the time of outcome assessment was 8.82 years (SD = 0.66), and about half (48.2 %) of the children were male. Parents reported their child’s race as 42.6 % Black, 45.5 % White, and 11.9 % another race. Of the children, 5.6 % were Hispanic per parent report. The right column of Table 1 shows characteristics for the aims 2 and 3 (CANDLE only) sample. Mean child age of this sample was 8.78 years (SD = 0.73) at the time of outcome assessment, and 48.3 % were male. Parents reported their child’s race as 64.3 % Black, 29.1 % White, and 6.7 % another race. Of the children, 2.9 % were Hispanic per parent report.

Table 1.

Study sample characteristics.

Aim 1: Pooled PATHWAYS n = 1166
 Aims 2 & 3: CANDLE only n = 719
M/n SD/% M/n SD/%
Cohort
 CANDLE 723 62.0 % 719 100.0 %
 TIDES 331 28.4 %
 PW-GAPPS 112 9.6 %
Child age at time of outcome assessment, yrs 8.82 0.66 8.78 0.73
Child sex, male 562 48.2 % 347 48.3 %
Child race
 Black 497 42.6 % 462 64.3 %
 White 530 45.5 % 209 29.1 %
 Asian 26 2.2 % <5
 American Indian or 6 0.5 % <5
 Alaska Native
 Two or more 81 6.9 % 35 4.9 %
 Another 26 2.2 % 7 1.0 %
Child ethnicity, Hispanic 65 5.6 % 21 2.9 %
Parental education at enrollment
 < High school 78 6.7 % 65 9.04 %
 High school or equivalent 402 34.5 % 332 46.18 %
 College degree 384 32.9 % 229 31.85 %
 Graduate or professional degree 302 25.9 % 94 13.07 %
Household income at child age 4–6 years, US dollars 67364 55085 39803 28499
Household size at child age 4–6 years
 2–3 260 22.30 % 157 21.84 %
 4 465 39.88 % 260 36.16 %
 5 258 22.13 % 168 23.37 %
 6+ 183 15.69 % 134 18.64 %
Season at time of outcome assessment, summer 334 28.64 % 247 34.35 %
Screen use in past 30 days
 < 1 h 237 20.33 % 97 13.49 %
 1 h 201 17.24 % 109 15.16 %
 2 h 410 35.16 % 263 36.58 %
 3 h 93 7.98 % 70 9.74 %
 4 h 121 10.38 % 91 12.66 %
 5 or more hours 104 8.92 % 89 12.38 %
ADHD medication use in past 30 days 81 6.95 % 54 7.51 %
Census designation at time of outcome assessment, urban 1103 94.60 % 687 95.55 %
Neighborhood deprivation index (NDI)
 Early infancy (0–6 months) 0.12 0.84 0.35 0.85
 Early childhood (6 months-6 years) 0.07 0.78 0.29 0.78
Open in a new tab

Note. Pooled PATHWAYS = combined sample of CANDLE, TIDES, and PW-GAPPS cohorts: CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study; TIDES = The Infant Development and Environment Study; PW-GAPPS = the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study. ADHD = Attention-Deficit/Hyperactivity Disorder. The pooled PATHWAYS n corresponds to the final sample size for the 0–6 month air pollution exposure window.

3.2. Sleep health outcome descriptive statistics

The mean sleep disturbance and sleep-related impairment scores were 8.8 (SD = 4.0) and 8.9 (SD = 4.1), respectively, in the pooled PATHWAYS cohort. In CANDLE, the mean sleep disturbance and sleep-related impairment scores were 8.8 (SD = 4.2) and 9.6 (SD = 4.4), respectively (Table 2).

Table 2.

Descriptive statistics for the middle childhood sleep health outcomes.

Aim 1: Pooled PATHWAYS
 Aims 2 & 3: CANDLE only
n M SD Min Q25 Q50 Q75 Max n M SD Min Q25 Q50 Q75 Max
Sleep disturbance 1162 8.8 4.0 4.0 5.0 8.0 11.0 20.0 719 8.8 4.2 4.0 5.0 8.0 12.0 20.0
Sleep-related impairment 1162 8.9 4.1 4.0 6.0 8.0 12.0 20.0 719 9.6 4.4 4.0 6.0 9.0 12.0 20.0
Open in a new tab

Note. Pooled PATHWAYS = combined sample of CANDLE, TIDES, and PW-GAPPS cohorts: CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study; TIDES = The Infant Development and Environment Study; PW-GAPPS = the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study.

3.3. Air pollution and ACEs exposure descriptive statistics

In the pooled PATHWAYS cohort, the median early infancy exposures to air pollution were 9.81 μ g/m3 (IQR = 2.56), 8.90 ppb (IQR = 4.32), and 24.83 ppb (IQR = 7.34) for PM2.5, NO2, and O3, respectively, and the median early childhood exposures were 8.84 μ g/m3 (IQR = 2.07), 8.77 ppb (IQR 3.33), and 26.32 ppb (IQR = 1.88) for PM2.5, NO2, and O3, respectively. In the CANDLE cohort only, the median early infancy exposures were 10.50 μ g/m3 (IQR = 1.70), 8.71 ppb (IQR = 3.67), and 26.13 ppb (IQR = 7.39) for PM2.5, NO2, and O3, respectively, and the median early childhood exposures were 9.28 μ g/m3 (IQR = 1.09), 9.08 ppb (IQR 2.55), and 26.61 ppb (IQR = 1.51) for PM2.5, NO2, and O3, respectively (Table 3). Appendix Table A1 shows Pearson’s correlations between the air pollution windows, for each pollutant, in the pooled PATHWAYS and CANDLE only samples.

Table 3.

Descriptive statistics for the air pollution exposures.

Aim 1: Pooled PATHWAYS
 Aims 2 & 3: CANDLE only
n M SD Median IQR n M SD Median IQR
Early infancy (0–6 months)
 PM2.5, ug/m3 1166   9.51 2.14   9.81 2.56 719 10.58 1.32 10.50 1.70
 NO2, ppb 1166   9.09 3.41   8.90 4.32 719   8.68 2.71   8.71 3.67
 O3, ppb 1166 24.92 4.75 24.83 7.34 719 26.05 4.07 26.13 7.39
Early childhood (6 months-6 years)
 PM2.5, ug/m3 1122   8.48 1.49   8.84 2.07 695   9.32 0.66   9.28 1.09
 NO2, ppb 1122   8.73 2.63   8.77 3.33 695   9.04 2.16   9.08 2.55
 O3, ppb 1122 25.91 2.48 26.32 1.88 695 26.74 1.12 26.61 1.51
Open in a new tab

Note. Pooled PATHWAYS = combined sample of CANDLE, TIDES, and PW-GAPPS cohorts: CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study; TIDES = The Infant Development and Environment Study; PW-GAPPS = the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study.

Descriptive statistics of ACEs exposures in the CANDLE cohort showed that just over half of the sample (54.5 %) reported no ACEs, 22.8 % reported one ACE, 12.0 % reported two ACEs, 4.9 % reported three ACEs, and 5.8 % reported four or more ACEs. The mean ACE score was 0.9 (SD = 1.3; Table 4).

Table 4.

Descriptive statistics of the ACEs exposures in the CANDLE cohort.

CANDLE cohort N = 719
M/n SD/%
ACE type
 Mental illness of a household member 143 19.9 %
 Household substance abuse   59   8.2 %
 Incarceration of a household member   64   8.9 %
 Caregiver divorce, separation, or loss 290 40.3 %
 Domestic violence 190 26.4 %
 Neighborhood violence   34   4.7 %
 Economic hardship   78 10.8 %
 Racial or ethnic discrimination   26   3.6 %
ACE score   0.9   1.3
 0 ACEs 392 54.5 %
 1 ACE 164 22.8 %
 2 ACEs   86 12.0 %
 3 ACEs   35   4.9 %
 4+ ACEs   42   5.8 %
Open in a new tab

Note. ACEs = adverse childhood experiences. CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study.

3.4. Aim 1 air pollution model results: associations between air pollution and sleep health (pooled PATHWAYS cohort)

We observed an inverse association between NO2 exposure during early infancy and sleep-related impairment in middle childhood in the main model: for every 1 IQR increase in NO2 exposure during early infancy, children scored − 0.31 (95 % CI −0.61, − 0.01) points on the sleep-related impairment questionnaire (Table 5). The relationship between NO2 and sleep-related impairment was similar for the early childhood exposure period, although this relationship did not reach significance (β = − 0.29, 95 % CI −0.62, 0.04). We did not observe associations between NO2 and the sleep disturbance outcome, or between the other air pollutants (O3 and PM2.5) and either sleep health outcome after controlling for the full set of covariates.

Table 5.

Aim 1 results of associations between air pollution exposures (during early infancy and early childhood periods) per IQR and sleep health outcomes in middle childhood (pooled PATHWAYS cohort analyses and CANDLE only sensitivity analyses).

IQR Sleep Disturbance
 Sleep-Related Impairment
Model 1
 Model 2
 CANDLE only
 Model 1
 Model 2
 CANDLE only
beta 95 % CI beta 95 % CI beta 95 % CI beta 95 % CI beta 95 % CI beta 95 % CI
Infancy (0–6 months)a
 PM2.5 (ug/m3) 2.56 −0.17 −0.72,0.37 −0.33 −0.89,0.23   0.15 −0.88,1.18   0.15 −0.35,0.65 −0.16 −0.66,0.34   0.35 −0.76,1.46
 NO2 (ppb) 4.32 −0.02 −0.34,0.29 −0.17 −0.52,0.18 −0.16 −0.82,0.51   0.12 −0.16,0.40 −0.31 −0.61,-0.01 −0.47 −1.15,0.22
 O3 (ppb) 7.34 −0.15 −0.57,0.28 −0.07 −0.49,0.36   0.26 −0.44,0.96 −0.34 −0.75,0.06 −0.10 −0.51,0.30 −0.14 −0.89,0.62
Childhood (6 months–6 years)b
 PM2.5 (ug/m3) 2.07   0.78 −0.06,1.63   0.44 −0.49,1.36   0.17 −2.88,3.22   0.88 0.17,1.59 −0.04 −0.75,0.67 −1.10 −4.04,1.84
 NO2 (ppb) 3.33   0.13 −0.19,0.45 −0.09 −0.46,0.29   0.10 −0.53,0.74   0.18 −0.12,0.48 −0.29 −0.62,0.04 −0.50 −1.15,0.16
 O3 (ppb) 1.88 −0.07 −0.35,0.20   0.08 −0.23,0.38 −0.52 −1.15,0.11 −0.08 −0.31,0.16   0.17 −0.07,0.41   0.13 −0.59,0.85
Open in a new tab

Note. Pooled PATHWAYS = combined sample of CANDLE, TIDES, and PW-GAPPS cohorts: CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study; TIDES = The Infant Development and Environment Study; PW-GAPPS = the PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study. Model 1 = minimally-adjusted model in the pooled PATHWAYS cohort, adjusted for child age, child sex, and study site. Model 2 = main model in the pooled PATHWAYS cohort, adjusted for all covariates included in model 1 plus parental education, household income, household size, a term interacting household income with household size, child race, child ethnicity, season, urbanicity, average screen use, ADHD medication use, the neighborhood deprivation index, and date of birth splines. CANDLE only = sensitivity analysis that includes the Model 2 covariates (with the exception of the study site covariate) restricted to the CANDLE cohort only. IQR is for main pooled PATHWAYS cohort.

a

n = 1166 for Models 1 and 2; n = 719 for CANDLE only.

b

n = 1122 for Models 1 and 2; n = 719 for CANDLE only.

Findings from sensitivity analyses were generally similar to the primary model (see Table 5 for CANDLE only analysis; Appendix Table A2 and Appendix Figures A1–A2 show additional sensitivity analyses). However, in the CANDLE only analysis, the magnitude of the inverse association between NO2 exposure during early infancy and sleep-related impairment in middle childhood increased to − 0.47 and was more imprecise (Table 5). Adjustment for the early childhood exposure window and exclusion of children with neurodevelopmental concerns attenuated the association between NO2 exposure in early infancy and sleep-related impairment in middle childhood (Appendix Table A2). We also observed a significant association between higher O3 exposure during early infancy and lower sleep-related impairment in the multipollutant model (Appendix Table A2). We did not find robust evidence of non-linear associations between air pollution and sleep health in the general additive models (Appendix Figure A2).

3.5. Aim 2 ACEs model results: associations between ACEs and sleep health (CANDLE cohort)

We did not observe significant associations between continuous ACEs and sleep health outcomes after controlling for the full set of covariates, although estimates were in the expected direction (Table 6).

Table 6.

Aim 2 results of associations between ACEs exposures and sleep health outcomes in middle childhood (CANDLE cohort).

n Sleep Disturbance
 Sleep-Related Impairment
Model 1
 Model 2
 Model 1
 Model 2
beta 95 % CI beta 95 % CI beta 95 % CI beta 95 % CI
ACE score 719 0.22 0.00, 0.44 0.14 −0.09, 0.36 0.16 −0.09, 0.4 0.09 −0.15, 0.32
Open in a new tab

Note. ACEs = adverse childhood experiences. CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study. Model 1 = minimally-adjusted model, adjusted for child age and sex. Model 2 = main model, adjusted for all covariates included in model 1 plus parental education, household income, household size, a term interacting household income with household size, child race, child ethnicity, season, urbanicity, average screen use, ADHD medication use, and the neighborhood deprivation index.

Sensitivity analyses reparametrizing ACEs into categories showed a similar pattern of findings, with the exception of an association between the 3 ACE category (v. 0 ACEs) and lower sleep-related impairment (β = − 1.36, 95 % CI −2.61, − 0.11). This should be interpreted with caution, however, as the 3 ACE category was comprised of a small group of participants (n = 35). Sensitivity analyses excluding children with neurodevelopmental concerns also showed similar findings (Appendix Table A3).

3.6. Aim 3 interaction model results: associations between air pollution-by-ACEs interactions and sleep health (CANDLE cohort)

We observed a significant air pollution-by-ACEs interaction for NO2 exposure during both the early infancy (pinteraction = 0.02) and early childhood (pinteraction = 0.02) periods on the sleep-related impairment outcome (Table 7). To better understand the nature of these interactions, we plotted the beta coefficient for the relationship between NO2 exposure and sleep-related impairment in early infancy (Fig. 2a) and early childhood (Fig. 2b) as a function of the continuous ACE score. The pattern of the relationship was similar for the two NO2 exposure periods: At lower levels of ACEs, there was an inverse relationship between NO2 and sleep-related impairment, but at higher levels of ACEs, the relationship between NO2 and sleep-related impairment was positive and increased in magnitude as the number of ACEs increased. The 95 % confidence interval of the beta coefficient for the relationship between NO2 exposure and sleep-related impairment did not cross zero at higher levels of ACEs in the early infancy NO2 exposure period model and at lower levels of ACEs in the early childhood NO2 exposure period model. There were no other significant air pollution-by-ACEs interactions.

Table 7.

Aim 3 results of associations between air pollution-by-ACEs interactions and sleep health outcomes in middle childhood (CANDLE cohort).

n Sleep Disturbance
 Sleep-Related Impairment
Model 1
 Model 2
 Model 1
 Model 2
beta 95 % CI p beta 95 % CI p beta 95 % CI p beta 95 % CI p
Early infancy (0–6 months)
 PM2.5 x ACEs 719 −0.36 −0.83, 0.12 0.14 −0.38 −0.84, 0.09 0.11 −0.31 −0.81, 0.18 0.22 −0.20 −0.67, 0.28 0.42
 NO2 x ACEs 719 −0.05 −0.44, 0.33 0.78 −0.01 −0.39, 0.36 0.94   0.33 −0.01, 0.67 0.06   0.43 0.08, 0.78 0.02
 O3 x ACEs 719   0.05 −0.35, 0.45 0.82   0.02 −0.38, 0.43 0.91 −0.33 −0.81, 0.14 0.17 −0.25 −0.73, 0.24 0.32
Early childhood (6 months-6 years)
 PM2.5 x ACEs 695 −0.12 −0.91, 0.67 0.76 −0.04 −0.85, 0.77 0.91 −0.06 −0.91, 0.79 0.89   0.09 −0.80, 0.98 0.84
 NO2 x ACEs 695 −0.01 −0.35, 0.33 0.95   0.01 −0.34, 0.35 0.97   0.33   0.01, 0.65 0.05   0.41 0.08, 0.73 0.02
 O3 x ACEs 695 −0.05 −0.36, 0.27 0.77 −0.04 −0.37, 0.28 0.80 −0.17 −0.52, 0.17 0.32 −0.27 −0.61, 0.08 0.13
Open in a new tab

Note. ACEs = adverse childhood experiences. CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study. Model 1 = minimally-adjusted model, adjusted for child age and sex. Model 2 = main model, adjusted for all covariates included in model 1 plus parental education, household income, household size, a term interacting household income with household size, child race, child ethnicity, season, urbanicity, average screen use, ADHD medication use, the neighborhood deprivation index, and date of birth splines. Betas, 95 % CIs, and p values are for the interaction term.

Fig. 2.

Fig. 2.

Open in a new tab

Interaction Between NO2 Exposure (During Early Infancy and Early Childhood Periods) and ACEs on Sleep-Related Impairment in Middle Childhood (CANDLE cohort).

Note. ACEs = adverse childhood experiences. CANDLE = The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study. Panel a shows interaction between ACEs and NO2 exposure during the early infancy period (0–6 months). Panel b shows interaction between ACEs and NO2 exposure during the early childhood period (6 months–6 years). Y-axis is the beta coefficient for the relationship between NO2 and sleep-related impairment. X-axis is the ACE score with a jittered rug plot showing the density of children with each discrete number of ACEs. Shaded gray areas show the 95 % confidence interval; darker gray shading indicates where the confidence interval does not cross zero. Models adjust for child age, child sex, parental education, household income, household size, a term interacting household income with household size, child race, child ethnicity, season, urbanicity, average screen use, ADHD medication use, the neighborhood deprivation index, and date of birth splines.

Sensitivity analyses reparametrizing the ACE score into categories showed inconsistent patterns of findings but also detected additional potential air pollution-by-ACEs interactions for the sleep-related impairment and sleep disturbance outcomes (Appendix Figure A3). These should be interpreted with caution, however, as the strata that appear to be accounting for these interactions have small sample sizes (n = 35 and n = 42 in the 3 and 4 ACE categories, respectively). Sensitivity analyses excluding children with neurodevelopmental concerns showed patterns of findings similar to the main models (Appendix Table A4), although the interactions that were significant in the main models were not significant in the models excluding children with neurodevelopmental concerns.

4. Discussion

This study examined independent and interactive associations of air pollution exposure (PM2.5, NO2, and O3 during early infancy and early childhood) and ACEs on children’s sleep health in middle childhood. Overall, we found little evidence of associations, with the exception of suggestive evidence for associations of NO2 and NO2-by-ACE interactions with the sleep-related impairment outcome. Specifically, the aim 1 air pollution models (pooled PATHWAYS cohort) indicated that, contrary to our hypothesis, greater NO2 exposure associated with lower sleep-related impairment in middle childhood. This association was small in magnitude but was significant for NO2 exposure in the early infancy period and marginally significant for the early childhood period. The aim 3 interaction models (CANDLE cohort only) were suggestive of ACEs as a modifier of the association between NO2 and sleep-related impairment for both the early infancy and early childhood NO2 exposure periods - a novel observation, as this study is the first to test air pollution-by-ACEs interactions with respect to sleep health outcomes in children. The other air pollution-by-ACEs interactions tested here were not supported which, in the context of being the first study to test such interactions, contributes novel null results. We also did not observe significant independent associations between NO2 and the sleep disturbance outcome, between the other air pollutants (PM2.5, O3) and either sleep health outcome, or between ACEs and either sleep health outcome, adding mixed findings to existing literatures supporting such independent associations of air pollution and ACEs with sleep health in children (Abou-Khadra, 2013; Brown et al., 2022; Cai et al., 2022; Chapman et al., 2013; Chapman et al., 2011; Gui et al., 2024; Gui et al., 2024; Hambrick et al., 2018; Harada et al., 2021; Hash et al., 2019; He et al., 2023; Kajeepeta et al., 2015; Lawrence et al., 2018; Lewis-de Los Angeles, 2022; Lin et al., 2022; Oh et al., 2018; Rojo-Wissar et al., 2021; Sterling et al., 2021; Wang, Gueye-Ndiaye, et al., 2025; Wang et al., 2016; Yu et al., 2022). We discuss these findings in greater detail below.

For aim 1, we observed that greater NO2 exposure, particularly during the first 6 months of life in the pooled PATHWAYS cohort, associated with lower sleep-related impairment in middle childhood. This observation runs contrary to our hypothesis and that of prior studies reporting associations of greater NO2 exposure with increased odds of a potential sleep disorder (Lawrence et al., 2018) and short sleep duration (Wang, Gueye-Ndiaye, et al., 2025). It may be that this finding is spurious, as we conducted a large number of comparisons. However, one potential explanation should this finding not be spurious could be related to how prepubertal children can paradoxically appear more alert and active when they are actually sleepy (Givan, 2004; Owens et al., 2020). Although overall, it is difficult to interpret why greater NO2 would associate with lower sleep-related impairment. Further research into NO2 exposures and sleep health outcomes in prepubertal children would help clarify potential associations.

The aim 3 interaction analyses in the CANDLE cohort suggested that the relationship between NO2 and sleep-related impairment was qualified by cumulative lifetime exposure to adversity for both the early infancy and early childhood NO2 exposure periods. The overall pattern of the interaction, which was similar for the two NO2 exposure periods, showed that, at lower levels of ACEs, greater NO2 exposure associated with lower sleep-related impairment, but at higher levels of ACEs, greater NO2 exposure associated with greater sleep-related impairment. We interpret these findings as suggestive of a synergistic relationship between NO2 exposure in general (birth–6 years) and cumulative lifetime ACEs with respect to the outcome of sleep-related impairment in middle childhood. Yet just why the direction of association between NO2 exposure and sleep-related impairment changed between lower and higher levels of ACEs is puzzling. Drawing on chemical-by-non-chemical stress interaction frameworks, it may be that higher ACE scores – as indicators of accumulated exposures to distinct types of highly stressful psychosocial experiences (Anda et al., 2010) – amplify the sleep-related neurotoxicity of NO2 (Block et al., 2012; Cory-Slechta, 2005; Gee and Payne-Sturges, 2004; McEwen and Tucker, 2011; Morello-Frosch and Shenassa, 2006). However at lower levels of ACEs, the association between greater NO2 exposure and lower sleep-related impairment is – as discuss above – more difficult to interpret. It is also noteworthy that few children in this sample reported very high ACE scores, and thus the NO2-by-ACE interaction at higher levels of ACEs is limited by a small sample size and should be interpreted with caution. Additionally, similar to the aim 1 findings, we cannot rule out the possibility that the significant interaction findings were spurious. Further investigation into NO2-by-ACEs interactions with respect to sleep health outcomes would be informative.

It is also worthy of comment that NO2 may not be the relevant exposure for sleep-related impairment but, rather, one or more other closely related exposures. NO2 is a proxy for traffic-related air pollution (TRAP; Loftus et al., 2020), and it may be that other specific components of TRAP (e.g., ultrafine particles), mixtures of TRAP components, and/or traffic-related noise are the more important exposures for sleep-related impairment, not NO2 itself. Although we included a rural v. urban variable as a proxy for noise, we cannot rule out residual confounding.

We did not find associations between other air pollutant and sleep outcome pairs, adding predominately null findings to an otherwise growing body of literature supporting such associations. Findings from prior studies indicate that not just NO2 but also other air pollutants, including particulate matter and O3, associate with sleep health in children (Abou-Khadra, 2013; Cai et al., 2022; Gui et al., 2024; Gui et al., 2024; He et al., 2023; Lawrence et al., 2018). Differences in exposure or outcome assessments, the magnitude of pollutant exposures measured, ability to adjust for potential confounders, and/or population studied may account for these discrepant findings. Given these mixed findings, and given that the literature about air pollution and sleep health in children is still relatively small, further research is needed to better understand linkages between specific air pollutants and sleep health outcomes in children.

We also did not find an independent association between children’s ACEs and either sleep health outcome, although effect estimates were in the expected direction. This is in contrast to multiple studies showing that more ACEs associate with poorer sleep health across the lifespan (Brown et al., 2022; Chapman et al., 2013; Chapman et al., 2011; Hambrick et al., 2018; Harada et al., 2021; Hash et al., 2019; Kajeepeta et al., 2015; Lewis-de Los Angeles, 2022; Lin et al., 2022; Oh et al., 2018; Rojo-Wissar et al., 2021; Sterling et al., 2021; Wang et al., 2016; Yu et al., 2022). During middle childhood specifically, prior studies have shown that higher ACE scores associate with caregiver-reported insufficient sleep duration (Harada et al., 2021; Lin et al., 2022), sleep problems (Hambrick et al., 2018), and overall sleep disturbances (Lewis-de Los Angeles, 2022). Differences in measurement methods, number of ACEs experienced, ability to adjust for potential confounders, and/or population studied may again explain these discrepant findings.

Finally, we did not find strong evidence for the first 6 months of life as a sensitive period. Rather, our findings are more suggestive of an association between NO2 exposure in general from birth–6 years and sleep-related impairment in middle childhood. This interpretation is supported by the aim 1 air pollution analyses in the pooled PATHWAYS cohort, in which sleep-related impairment significantly associated with NO2 exposure during the early infancy period and marginally so during the early childhood period, as well as the aim 3 interaction analyses in the CANDLE cohort, in which ACEs interacted significantly with NO2 for both exposure periods. We also cannot rule out the possibility that NO2 exposure in one period reflected exposure in the other period, or that stability in exposure across time mattered for the sleep-related impairment outcome. The correlation between NO2 exposure in the early infancy and early childhood periods was strong (rs = 0.64 and 0.60 in the pooled PATHWAYS and CANDLE only samples, respectively), indicating similar measurements across the two exposure periods.

4.1. Limitations

There are some limitations that deserve comment. First, we utilized child self-report measures of sleep health. These measures are validated and are important for capturing children’s perceptions and lived experiences about their own sleep health (Forrest et al., 2018), but they offer only one perspective. Different perspectives may be gathered via parents or caregivers. The PROMIS sleep disturbance and sleep-related impairment questionnaires are available in parent proxy report (for children ages 5–17 years) and child self-report (for children ages 8–17 years) versions. The parent proxy and child self-report PROMIS sleep measures are highly but not perfectly correlated (rs = 0.71–0.77), indicating that the two versions offer mostly overlapping but also some unique information (Forrest et al., 2018). The child self-report version is preferred and recommended (Health Measures, 2023), but future inquiry could consider collecting both parent proxy and child self-report sleep measures together in the same study. Separately but relatedly, future studies could also consider gathering objective measures of sleep health using actigraphy or polysomnography. Objective measures assess aspects of sleep that are different from the PROMIS sleep health questionnaires, and as expected, the pediatric PROMIS sleep health questionnaires do not correlate with polysomnography-obtained sleep parameters (Forrest et al., 2018). Thus, future studies could consider multi-informant (parent/caregiver and child) and multi-method (subjective and objective) sleep health measures. A second limitation is that we did not have air pollution data at child age 6–9 years, limiting our ability to examine air pollution exposures at points in time closer to the sleep outcome assessment. Air pollution exposures occurring closer in time with the assessment of sleep could yield different, and potentially stronger, associations. Third, the timing of our air pollution exposure assessment did not align exactly with our ACEs exposure assessment. We measured air pollution in early infancy (0–6 months) and early childhood (6 months–6 years), while we measured cumulative lifetime exposure to distinct types of ACEs from birth to middle childhood (8–9 years). Although the ACEs assessment period overlapped with both air pollution exposure periods, we cannot comment on exactly when ACEs occurred with respect to the air pollution exposures. Importantly, however, both the ACEs and air pollution measurement periods preceded or were concurrent with that of the sleep health outcomes. Fourth, we measured outdoor ambient air pollution, not indoor air pollution. Infants and young children spend the majority of their time indoors (Early Childhood Scientific Council on Equity and the Environment, 2025), and according to the U.S. Environmental Protection Agency (EPA), indoor air pollutants are approximately two to five times greater than that of outdoor air pollutants (U.S. EPA, 2019). Indoor air quality is thus likely highly pertinent and should be considered in future studies. At the same time, measurement of outdoor air pollution remains informative, not only for the times in which children go outside, but also because an estimated 50 % of outdoor air pollution, on average, travels to indoor spaces; and because infants and young children spend most of their time indoors, the majority of their exposure to outdoor air pollution occurs indoors (Allen and Macomber, 2022; Early Childhood Scientific Council on Equity and the Environment, 2025). Finally, as is almost always the case with observational studies, our ability to adjust for potential confounders was limited by data availability and quality. There may be residual confounding that we were unable to account for.

5. Conclusion

In sum, the associations tested in this study were overall not supported, with the exception of suggestive evidence for associations of NO2 and NO2-by-ACE interactions with sleep-related impairment.

Supplementary Material

MMC1

Funding

This research is funded by NIH UG3/UH3OD023271 (ECHO-PATHWAYS [PI: Catherine Karr]). The CANDLE study is funded by both the Urban Child Institute and NIH (NIH R01HL109977 [PI: Kecia Carroll]. The TIDES study is funded from the NIH (NIHR01ES25169 [PI: Shanna Swan], UG3/UH3OD023305 [PI: Leonardo Trasande]. Additional funding includes NCATS of NIH: UL1 TR002319. CJK was additionally supported by NIH P30ES007033 (PI: Joel Kaufman) and ESB by P30ES005022. Research reported in this publication was also supported by ECHO PATHWAYS (NIH grants: 1UG3OD023271–01 and 4UH3OD023271–03), ECHO AWARE (UG3OD035528), UW EDGE Center (P30ES00703), the University of Washington Pediatric and Reproductive Environmental Health Scholars program (K12ES033584; MPIs Catherine Karr and Sheela Sathyanarayana), and by grants R56ES026528, R01ES023500, and R01ES02588 from NIEHS and P01AG055367 from NIA. Funding was also provided by Kresge Foundation Grant No. 243365.

Abbreviations:

ACEs

Adverse childhood experiences

CANDLE

The Conditions Affecting Neurocognitive Development and Learning in Early Childhood study

TIDES

The Infant Development and Environment Study

PW-GAPPS

PATHWAYS Global Alliance to Prevent Prematurity and Stillbirth study

PATHWAYS

CANDLE, TIDES, and PW-GAPPS combined

PROMIS

Patient-Reported Outcomes Measurement Information System

PM2.5

f ine particulate matter

NO2

nitrogen dioxide

O3

ozone

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijheh.2025.114638.

Footnotes

CRediT authorship contribution statement

Jonika B. Hash: Writing – review & editing, Writing – original draft, Visualization, Project administration, Methodology, Investigation, Conceptualization. Logan C. Dearborn: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Christine T. Loftus: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Conceptualization. Catherine J. Karr: Writing – review & editing, Supervision, Resources, Methodology, Investigation, Funding acquisition, Conceptualization. Adam A. Szpiro: Writing – review & editing, Methodology, Investigation. Emily S. Barrett: Writing – review & editing, Methodology, Investigation, Funding acquisition. Kaja Z. LeWinn: Writing – review & editing, Methodology, Investigation, Funding acquisition. Ruby Nguyen: Writing – review & editing, Investigation, Funding acquisition. Paul E. Moore: Writing – review & editing, Investigation. Brent Collett: Writing – review & editing, Investigation. Amanda N. Noroña-Zhou: Writing – review & editing, Methodology, Investigation. Nicole R. Bush: Writing – review & editing, Methodology, Investigation, Funding acquisition. Sheela Sathyanarayana: Writing – review & editing, Supervision, Resources, Methodology, Investigation, Funding acquisition, Conceptualization.

☆

Air pollution models for this publication were developed under a STAR research assistance agreements RD831697 (MESA Air), RD-83830001 (MESA Air Next Stage), RD83479601 (UW Center for Clean Air Research), and R83374101 (MESA Coarse), awarded by the U.S. Environmental Protection Agency. This publication has not been formally reviewed by the EPA. The views expressed in this document are solely those of the authors and the EPA does not endorse any products or commercial services mentioned in this publication.

☆☆

The funders had no role in study design; data collection, analysis, or interpretation; decision to publish; or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This manuscript has been reviewed by PATHWAYS for scientific content and consistency of data interpretation with previous PATHWAYS publications.

References

  1. Abou-Khadra MK, 2013. Association between PM10 exposure and sleep of Egyptian school children. Sleep Breath. 17 (2), 653–657. 10.1007/s11325-012-0738-7. [DOI] [PubMed] [Google Scholar]
  2. Agostini A, Centofanti S, 2021. Normal sleep in children and adolescence. Child Adolescent Psychiatric Clin. North Am 30 (1), 1–14. 10.1016/j.chc.2020.08.011 . [DOI] [PubMed] [Google Scholar]
  3. Allen JG, Macomber JD, 2022. Healthy Buildings: How Indoor Spaces Can Make you sick—or Keep you Well, second ed. Harvard University Press. 10.4159/9780674287457. [DOI] [Google Scholar]
  4. Anda RF, Felitti VJ, Bremner JD, Walker JD, Whitfield C, Perry BD, Dube SR, Giles WH, 2006. The enduring effects of abuse and related adverse experiences in childhood: a convergence of evidence from neurobiology and epidemiology. Eur. Arch. Psychiatr. Clin. Neurosci 256 (3), 174–186. 10.1007/s00406-005-0624-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Anda RF, Butchart A, Felitti VJ, Brown DW, 2010. Building a framework for global surveillance of the public health implications of adverse childhood experiences. Am. J. Prevent. Med 39 (1), 93–98. 10.1016/j.amepre.2010.03.015. [DOI] [PubMed] [Google Scholar]
  6. Barrett ES, Sathyanarayana S, Janssen S, Redmon JB, Nguyen RH, Kobrosly R, Swan SH, 2014. Environmental health attitudes and behaviors: findings from a large pregnancy cohort study. Eur. J. Obstet. Gynecol. Reprod. Biol 176, 119–125. 10.1016/j.ejogrb.2014.02.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bathory E, Tomopoulos S, 2017. Sleep regulation, physiology and development, sleep duration and patterns, and sleep hygiene in infants, toddlers, and preschool-age children. Curr. Probl. Pediatr. Adolesc. Health Care 47 (2), 29–42. 10.1016/j.cppeds.2016.12.001. [DOI] [PubMed] [Google Scholar]
  8. Belli A, Breda M, Di Maggio C, Esposito D, Marcucci L, Bruni O, 2022. Children with neurodevelopmental disorders: how do they sleep? Curr. Opin. Psychiatr 35 (5), 345–351. 10.1097/yco.0000000000000790. [DOI] [PubMed] [Google Scholar]
  9. Belmon LS, van Stralen MM, Busch V, Harmsen IA, Chinapaw MJM, 2019. What are the determinants of children’s sleep behavior? A systematic review of longitudinal studies. Sleep Med. Rev 43, 60–70. 10.1016/j.smrv.2018.09.007. [DOI] [PubMed] [Google Scholar]
  10. Bevans KB, Meltzer LJ, De La Motte A, Kratchman A, Viel D, Forrest CB, 2018. Qualitative development and content validation of the PROMIS pediatric sleep health items. Behav. Sleep Med 17 (5), 1–15. 10.1080/15402002.2018.1461102. [DOI] [PubMed] [Google Scholar]
  11. Billings ME, Hale L, Johnson DA, 2020. Physical and social environment relationship with sleep health and disorders. Chest 157 (5), 1304–1312. 10.1016/j.chest.2019.12.002 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Block ML, Elder A, Auten RL, Bilbo SD, Chen H, Chen JC, Cory-Slechta DA, Costa D, Diaz-Sanchez D, Dorman DC, Gold DR, Gray K, Jeng HA, Kaufman JD, Kleinman MT, Kirshner A, Lawler C, Miller DS, Nadadur SS, Wright RJ, 2012. The outdoor air pollution and brain health workshop. Neurotoxicology 33 (5), 972–984. 10.1016/j.neuro.2012.08.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bose S, Ross KR, Rosa MJ, Chiu YM, Just A, Kloog I, Wilson A, Thompson J, Svensson K, Rojo MMT, Schnaas L, Osorio-Valencia E, Oken E, Wright RO, Wright RJ, 2019. Prenatal particulate air pollution exposure and sleep disruption in preschoolers: windows of susceptibility. Environ. Int 124, 329–335. 10.1016/j.envint.2019.01.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Brown SM, Rodriguez KE, Smith AD, Ricker A, Williamson AA, 2022. Associations between childhood maltreatment and behavioral sleep disturbances across the lifespan: a systematic review. Sleep Med. Rev 64, 101621. 10.1016/j.smrv.2022.101621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Brumberg HL, Karr CJ, 2021. Ambient air pollution: health hazards to children. Pediatrics 147 (6). 10.1542/peds.2021-051484. [DOI] [PubMed] [Google Scholar]
  16. Buysse DJ, 2014. Sleep health: can we define it? Does it matter? Sleep 37 (1), 9–17. 10.5665/sleep.3298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Cai J, Shen Y, Zhao Y, Meng X, Niu Y, Chen R, Quan G, Li H, Groeger JA, Du W, Hua J, Kan H, 2022. Early life exposure to PM2.5 and sleep disturbances in preschoolers from 551 cities of China. Am. J. Respir. Crit. Care Med 10.1164/rccm.202204-0740OC . [DOI] [PubMed] [Google Scholar]
  18. Cao B, Chen Y, McIntyre RS, 2021. Comprehensive review of the current literature on impact of ambient air pollution and sleep quality. Sleep Med. 79, 211–219. 10.1016/j.sleep.2020.04.009. [DOI] [PubMed] [Google Scholar]
  19. Chapman DP, Whitfield CL, Felitti VJ, Dube SR, Edwards VJ, Anda RF, 2004. Adverse childhood experiences and the risk of depressive disorders in adulthood. J. Affect. Disord 82 (2), 217–225. 10.1016/j.jad.2003.12.013. [DOI] [PubMed] [Google Scholar]
  20. Chapman DP, Wheaton AG, Anda RF, Croft JB, Edwards VJ, Liu Y, Sturgis SL, Perry GS, 2011. Adverse childhood experiences and sleep disturbances in adults. Sleep Med. 12 (8), 773–779. 10.1016/j.sleep.2011.03.013. [DOI] [PubMed] [Google Scholar]
  21. Chapman DP, Liu Y, Presley-Cantrell LR, Edwards VJ, Wheaton AG, Perry GS, Croft JB, 2013. Adverse childhood experiences and frequent insufficient sleep in 5 U.S. States, 2009: a retrospective cohort study. BMC Public Health 13, 3. 10.1186/1471-2458-13-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Chaput JP, Gray CE, Poitras VJ, Carson V, Gruber R, Olds T, Weiss SK, Connor Gorber S, Kho ME, Sampson M, Belanger K, Eryuzlu S, Callender L, Tremblay MS, 2016. Systematic review of the relationships between sleep duration and health indicators in school-aged children and youth. Appl. Physiol. Nutr. Metabol 41 (6 Suppl. 3), S266–S282. 10.1139/apnm-2015-0627. [DOI] [PubMed] [Google Scholar]
  23. Chuang HC, Su TY, Chuang KJ, Hsiao TC, Lin HL, Hsu YT, Pan CH, Lee KY, Ho SC, Lai CH, 2018. Pulmonary exposure to metal fume particulate matter cause sleep disturbances in shipyard welders. Environ. Pollut 232, 523–532. 10.1016/j.envpol.2017.09.082. [DOI] [PubMed] [Google Scholar]
  24. Cory-Slechta DA, 2005. Studying toxicants as single chemicals: does this strategy adequately identify neurotoxic risk? Neurotoxicology 26 (4), 491–510. 10.1016/j.neuro.2004.12.007 . [DOI] [PubMed] [Google Scholar]
  25. Costa LG, Cole TB, Coburn J, Chang YC, Dao K, Roque P, 2014. Neurotoxicants are in the air: convergence of human, animal, and in vitro studies on the effects of air pollution on the brain. BioMed. Res. Int 2014, 736385. 10.1155/2014/736385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Costa LG, Cole TB, Coburn J, Chang YC, Dao K, Roque PJ, 2017. Neurotoxicity of traffic-related air pollution. Neurotoxicology 59, 133–139. 10.1016/j.neuro.2015.11.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Costa LG, Cole TB, Dao K, Chang YC, Garrick JM, 2019. Developmental impact of air pollution on brain function. Neurochem. Int 131, 104580. 10.1016/j.neuint.2019.104580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Cronholm PF, Forke CM, Wade R, Bair-Merritt MH, Davis M, Harkins-Schwarz M, Pachter LM, Fein JA, 2015. Adverse childhood experiences: expanding the concept of adversity. Am. J. Prevent. Med 49 (3), 354–361. 10.1016/j.amepre.2015.02.001. [DOI] [PubMed] [Google Scholar]
  29. Dahl RE, 1996. The regulation of sleep and arousal: development and psychopathology. Dev. Psychopathol 8 (1), 3–27. 10.1017/S0954579400006945. [DOI] [Google Scholar]
  30. Danese A, McEwen BS, 2012. Adverse childhood experiences, allostasis, allostatic load, and age-related disease. Physiol. Behav 106 (1), 29–39. 10.1016/j.physbeh.2011.08.019. [DOI] [PubMed] [Google Scholar]
  31. Early Childhood Scientific Council on Equity and the Environment, 2025. Air quality affects early childhood development and health. Working Paper No. 3. www.developingchild.harvard.edu . [Google Scholar]
  32. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, Koss MP, Marks JS, 1998. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The adverse childhood experiences (ACE) study. Am. J. Prevent. Med 14 (4), 245–258. [DOI] [PubMed] [Google Scholar]
  33. Finlay S, Roth C, Zimsen T, Bridson TL, Sarnyai Z, McDermott B, 2022. Adverse childhood experiences and allostatic load: a systematic review. Neurosci. Biobehav. Rev 136, 104605. 10.1016/j.neubiorev.2022.104605. [DOI] [PubMed] [Google Scholar]
  34. Forrest CB, Meltzer LJ, Marcus CL, de la Motte A, Kratchman A, Buysse DJ, Pilkonis PA, Becker BD, Bevans KB, 2018. Development and validation of the PROMIS pediatric sleep disturbance and sleep-related impairment item banks. Sleep 41 (6). 10.1093/sleep/zsy054. [DOI] [PubMed] [Google Scholar]
  35. Franco P, Putois B, Guyon A, Raoux A, Papadopoulou M, Guignard-Perret A, Bat-Pitault F, Hartley S, Plancoulaine S, 2020. Sleep during development: sex and gender differences. Sleep Med. Rev 51, 101276. 10.1016/j.smrv.2020.101276. [DOI] [PubMed] [Google Scholar]
  36. Gee GC, Payne-Sturges DC, 2004. Environmental health disparities: a framework integrating psychosocial and environmental concepts. Environ. Health Perspect 112 (17), 1645–1653. 10.1289/ehp.7074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Genc S, Zadeoglulari Z, Fuss SH, Genc K, 2012. The adverse effects of air pollution on the nervous system. J. Toxicol 2012, 782462. 10.1155/2012/782462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Givan DC, 2004. The sleepy child. Pediatr. Clin 51 (1), 15–31. 10.1016/s0031-3955(03)00175-5. [DOI] [PubMed] [Google Scholar]
  39. Global Alliance to Prevent Prematurity and Still Birth. Global alliance to prevent prematurity and stillbirth. https://www.gapps.org.
  40. Gregory AM, Sadeh A, 2016. Annual research review: sleep problems in childhood psychiatric disorders–a review of the latest science. JCPP (J. Child Psychol. Psychiatry) 57 (3), 296–317. 10.1111/jcpp.12469. [DOI] [PubMed] [Google Scholar]
  41. Gueye-Ndiaye S, Bhaskar S, Rueschman M, Owens J, Gold D, Chen J, Phipatanakul W, Adamkiewicz G, Redline S, 2025. The associations between indoor nitrogen dioxide (NO2) exposure and adverse pediatric sleep outcomes. SLEEP 48. 10.1093/sleep/zsaf090.1069. [DOI] [Google Scholar]
  42. Guglielmo D, Gazmararian JA, Chung J, Rogers AE, Hale L, 2018. Racial/Ethnic sleep disparities in US school-aged children and adolescents: a review of the literature. Sleep Health 4 (1), 68–80. 10.1016/j.sleh.2017.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Gui ZH, Heinrich J, Min Qian Z, Schootman M, Zhao TY, Xu SL, Jin NX, Huang HH, He WT, Wu QZ, Zhang JL, Wang DS, Yang M, Liu RQ, Zeng XW, Dong GH, Lin LZ, 2024a. Exposures to particulate matters and childhood sleep disorders: a large study in three provinces in China. Environ. Int 190, 108841. 10.1016/j.envint.2024.108841. [DOI] [PubMed] [Google Scholar]
  44. Gui ZH, Heinrich J, Morawska L, Zhao TY, Yim SH, Lao XQ, Gao M, Chen DH, Ma HM, Lin LZ, Liu RQ, Dong GH, 2024b. Long-term exposure to ozone and sleep disorders in children: a multicity study in China. Environ. Res 260, 119553. 10.1016/j.envres.2024.119553. [DOI] [PubMed] [Google Scholar]
  45. Hale L, Troxel W, Buysse DJ, 2020. Sleep health: an opportunity for public health to address health equity. Annu. Rev. Publ. Health 41, 81–99. 10.1146/annurev-publhealth-040119-094412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Hambrick EP, Rubens SL, Brawner TW, Taussig HN, 2018. Do sleep problems mediate the link between adverse childhood experiences and delinquency in preadolescent children in foster care? JCPP (J. Child Psychol. Psychiatry) 59 (2), 140–149. 10.1111/jcpp.12802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Harada M, Guerrero A, Iyer S, Slusser W, Szilagyi M, Koolwijk I, 2021. The relationship between adverse childhood experiences and weight-related health behaviors in a national sample of children. Academic Pediat. 21 (8), 1372–1379. 10.1016/j.acap.2021.05.024. [DOI] [PubMed] [Google Scholar]
  48. Hash JB, Oxford ML, Fleming CB, Ward TM, Spieker SJ, Lohr MJ, 2019. Impact of a home visiting program on sleep problems among young children experiencing adversity. Child Abuse Negl. 89, 143–154. 10.1016/j.chiabu.2018.12.016 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. He F, Yanosky JD, Bixler EO, Fernandez-Mendoza J, Chinchilli VM, Al-Shaar L, Vgontzas AN, Liao D, 2023. Short-term and intermediate-term fine particulate air pollution are synergistically associated with habitual sleep variability in adolescents: a cross-sectional study. Environ. Res 227, 115726. 10.1016/j.envres.2023.115726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Health Measures, 2023. Recommendations by age. https://www.healthmeasures.net/explore-measurement-systems/selecting-among-measurement-systems/recommendations-by-age .
  51. Johnson DA, Jackson CL, Williams NJ, Alcantara C, 2019. Are sleep patterns influenced by race/ethnicity: a marker of relative advantage or disadvantage? Evidence to date. Nat. Sci. Sleep 11, 79–95. 10.2147/NSS.S169312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Kajeepeta S, Gelaye B, Jackson CL, Williams MA, 2015. Adverse childhood experiences are associated with adult sleep disorders: a systematic review. Sleep Med. 16 (3), 320–330. 10.1016/j.sleep.2014.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Kaufman JD, Hajat A, 2021. Confronting environmental racism. Environ. Health Perspect 129 (5), 51001. 10.1289/ehp9511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Keller JP, Olives C, Kim SY, Sheppard L, Sampson PD, Szpiro AA, Oron AP, Lindström J, Vedal S, Kaufman JD, 2015. A unified spatiotemporal modeling approach for predicting concentrations of multiple air pollutants in the multi-ethnic study of atherosclerosis and air pollution. Environ. Health Perspect 123 (4), 301–309. 10.1289/ehp.1408145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Kim HG, Kuendig J, Prasad K, Sexter A, 2020. Exposure to racism and other adverse childhood experiences among perinatal women with moderate to severe mental illness. Community Ment. Health J 56 (5), 867–874. 10.1007/s10597-020-00550-6. [DOI] [PubMed] [Google Scholar]
  56. Kirwa K, Szpiro AA, Sheppard L, Sampson PD, Wang M, Keller JP, Young MT, Kim SY, Larson TV, Kaufman JD, 2021. Fine-scale air pollution models for epidemiologic research: insights from approaches developed in the multi-ethnic study of atherosclerosis and air pollution (MESA air). Curr. Environ. Health Rep 8 (2), 113–126. 10.1007/s40572-021-00310-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Kwong TY, Hayes DK, 2017. Adverse family experiences and flourishing amongst children ages 6–17 years: 2011/12 national survey of Children’s health. Child Abuse Negl. 70, 240–246. 10.1016/j.chiabu.2017.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Lam J, Baello S, Iqbal M, Kelly LE, Shannon PT, Chitayat D, Matthews SG, Koren G, 2015. The ontogeny of P-glycoprotein in the developing human blood-brain barrier: implication for opioid toxicity in neonates. Pediatr. Res 78 (4), 417–421. 10.1038/pr.2015.119. [DOI] [PubMed] [Google Scholar]
  59. Lane HM, Morello-Frosch R, Marshall JD, Apte JS, 2022. Historical redlining is associated with present-day air pollution disparities in U.S. cities. Environ. Sci. Technol. Lett 9 (4), 345–350. 10.1021/acs.estlett.1c01012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Lawrence WR, Yang M, Zhang C, Liu RQ, Lin S, Wang SQ, Liu Y, Ma H, Chen DH, Zeng XW, Yang BY, Hu LW, Yim SHL, Dong GH, 2018. Association between long-term exposure to air pollution and sleep disorder in Chinese children: the seven northeastern cities study. Sleep 41 (9). 10.1093/sleep/zsy122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Lenehan SM, Fogarty L, O’Connor C, Mathieson S, Boylan GB, 2022. The architecture of early childhood sleep over the first two years. Matern. Child Health J 10.1007/s10995-022-03545-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. LeWinn KZ, Karr CJ, Hazlehurst M, Carroll K, Loftus C, Nguyen R, Barrett E, Swan SH, Szpiro AA, Paquette A, Moore P, Spalt E, Younglove L, Sullivan A, Colburn T, Byington N, Sims Taylor L, Moe S, Wang S, Sathyanarayana S, 2022. Cohort profile: the ECHO prenatal and early childhood pathways to health consortium (ECHO-PATHWAYS). BMJ Open 12 (10), e064288. 10.1136/bmjopen-2022-064288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Lewis-de Los Angeles WW, 2022. Association between adverse childhood experiences and diet, exercise, and sleep in pre-adolescents. Academic Pediat. 22 (8), 1281–1286. 10.1016/j.acap.2022.06.007. [DOI] [PubMed] [Google Scholar]
  64. Lin SX, Cheslack-Postava K, McReynolds L, Amsel L, Bresnahan M, Hoven CW, 2022. Adverse childhood experiences and insufficient sleep among U.S. children and adolescents. Academic Pediatrrics 22 (6), 965–971. 10.1016/j.acap.2022.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Liu J, Wu T, Liu Q, Wu S, Chen JC, 2020. Air pollution exposure and adverse sleep health across the life course: a systematic review. Environ. Pollut 262, 114263. 10.1016/j.envpol.2020.114263. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Liu J, Ghastine L, Um P, Rovit E, Wu T, 2021. Environmental exposures and sleep outcomes: a review of evidence, potential mechanisms, and implications. Environ. Res 196, 110406. 10.1016/j.envres.2020.110406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Loftus CT, Ni Y, Szpiro AA, Hazlehurst MF, Tylavsky FA, Bush NR, Sathyanarayana S, Carroll KN, Young M, Karr CJ, LeWinn KZ, 2020. Exposure to ambient air pollution and early childhood behavior: a longitudinal cohort study. Environ. Res 183, 109075. 10.1016/j.envres.2019.109075 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Magee L, Hale L, 2012. Longitudinal associations between sleep duration and subsequent weight gain: a systematic review. Sleep Med. Rev 16 (3), 231–241. 10.1016/j.smrv.2011.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Matricciani L, Paquet C, Galland B, Short M, Olds T, 2019. Children’s sleep and health: a meta-review. Sleep Med. Rev 46, 136–150. 10.1016/j.smrv.2019.04.011. [DOI] [PubMed] [Google Scholar]
  70. Mayne SL, Mitchell JA, Virudachalam S, Fiks AG, Williamson AA, 2021. Neighborhood environments and sleep among children and adolescents: a systematic review. Sleep Med. Rev 57, 101465. 10.1016/j.smrv.2021.101465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. McEwen BS, Tucker P, 2011. Critical biological pathways for chronic psychosocial stress and research opportunities to advance the consideration of stress in chemical risk assessment. Am. J. Publ. Health 101 (S1), S131–S139. 10.2105/ajph.2011.300270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Meltzer LJ, Williamson AA, Mindell JA, 2021. Pediatric sleep health: it matters, and so does how we define it. Sleep Med. Rev 57, 101425. 10.1016/j.smrv.2021.101425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Merrick MT, Ford DC, Ports KA, Guinn AS, 2018. Prevalence of adverse childhood experiences from the 2011–2014 behavioral risk factor surveillance system in 23 states. JAMA Pediatr. 172 (11), 1038–1044. 10.1001/jamapediatrics.2018.2537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Mindell JA, Owens JA, 2015. A Clinical Guide to Pediatric Sleep: Diagnosis and Management of Sleep Problems, third ed. Wolters Kluwer. [Google Scholar]
  75. Morello-Frosch R, Shenassa ED, 2006. The environmental “riskscape” and social inequality: implications for explaining maternal and child health disparities. Environ. Health Perspect 114 (8), 1150–1153. 10.1289/ehp.8930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Newton AT, Honaker SM, Reid GJ, 2020. Risk and protective factors and processes for behavioral sleep problems among preschool and early school-aged children: a systematic review. Sleep Med. Rev 52, 101303. 10.1016/j.smrv.2020.101303 . [DOI] [PubMed] [Google Scholar]
  77. Nixon GM, Thompson JM, Han DY, Becroft DM, Clark PM, Robinson E, Waldie KE, Wild CJ, Black PN, Mitchell EA, 2008. Short sleep duration in middle childhood: risk factors and consequences. Sleep 31 (1), 71–78. 10.1093/sleep/31.1.71 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Oh DL, Jerman P, Silverio Marques S, Koita K, Purewal Boparai SK, Burke Harris N, Bucci M, 2018. Systematic review of pediatric health outcomes associated with childhood adversity. BMC Pediatr. 18 (1), 83. 10.1186/s12887-018-1037-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Owens JA, Babcock D, Weiss M, 2020. Evaluation and treatment of children and adolescents with excessive daytime sleepiness. Clin. Pediatr 59 (4–5), 340–351. 10.1177/0009922820903434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Padula AM, Rivera-Núñez Z, Barrett ES, 2020. Combined impacts of prenatal environmental exposures and psychosocial stress on offspring health: air pollution and metals. Curr. Environ. Health Rep 7 (2), 89–100. 10.1007/s40572-020-00273-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Payne-Sturges DC, Marty MA, Perera F, Miller MD, Swanson M, Ellickson K, Cory-Slechta DA, Ritz B, Balmes J, Anderko L, Talbott EO, Gould R, Hertz-Picciotto I, 2019. Healthy air, healthy brains: advancing air pollution policy to protect children’s health. Am. J. Publ. Health 109 (4), 550–554. 10.2105/ajph.2018.304902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Petruccelli K, Davis J, Berman T, 2019. Adverse childhood experiences and associated health outcomes: a systematic review and meta-analysis. Child Abuse Negl. 97, 104127. 10.1016/j.chiabu.2019.104127. [DOI] [PubMed] [Google Scholar]
  83. PROMIS, 2023. Sleep scoring manual: a brief guide to scoring the PROMIS sleep instruments. https://www.healthmeasures.net/images/PROMIS/manuals/Scoring_Manual_Only/PROMIS_Sleep_Scoring_Manual_05Dec2023.pdf.
  84. Quist JS, Sjodin A, Chaput JP, Hjorth MF, 2016. Sleep and cardiometabolic risk in children and adolescents. Sleep Med. Rev 29, 76–100. 10.1016/j.smrv.2015.09.001. [DOI] [PubMed] [Google Scholar]
  85. Rojo-Wissar DM, Sosnowski DW, Ingram MM, Jackson CL, Maher BS, Alfano CA, Meltzer LJ, Spira AP, 2021. Associations of adverse childhood experiences with adolescent total sleep time, social jetlag, and insomnia symptoms. Sleep Med. 88, 104–115. 10.1016/j.sleep.2021.10.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Sheldon SH, 2014. In: Sheldon SH, Kryger MH, Ferber R, Gozal D (Eds.), Principles and Practice of Pediatric Sleep Medicine, second ed. Saunders. [Google Scholar]
  87. Shonkoff JP, Boyce WT, McEwen BS, 2009. Neuroscience, molecular biology, and the childhood roots of health disparities: building a new framework for health promotion and disease prevention. JAMA 301 (21), 2252–2259. 10.1001/jama.2009.754. [DOI] [PubMed] [Google Scholar]
  88. Smith, Hardy ST, Hale LE, Gazmararian JA, 2019. Racial disparities and sleep among preschool aged children: a systematic review. Sleep Health 5 (1), 49–57. 10.1016/j.sleh.2018.09.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  89. Snow SJ, Henriquez AR, Costa DL, Kodavanti UP, 2018. Neuroendocrine regulation of air pollution health effects: emerging insights. Toxicol. Sci 164 (1), 9–20. 10.1093/toxsci/kfy129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  90. Sontag-Padilla L, Burns RM, Shih RA, Griffin BA, Martin LT, Chandra A, Tylavsky F, 2015. The Urban Child Institute CANDLE Study: Methodological Overview and Baseline Sample Description. RAND Corporation. 10.7249/RR1336. [DOI] [Google Scholar]
  91. Sterling S, Chi F, Lin J, Padalkar P, Vinayagasundaram U, Iturralde E, Young-Wolff K, Metz VE, Herz A, Negusse R, Jackson-Morris M, Espinas P, 2021. Physical, mental health and developmental conditions, and sociodemographic characteristics associated with adverse childhood experiences among young children in pediatric primary care. J. Pediatr. Health Care 35 (5), 491–499. 10.1016/j.pedhc.2021.04.009. [DOI] [PubMed] [Google Scholar]
  92. U.S. Department of Health & Human Services, Office of Deisease Prevention and Health Promotion. Healthy people 2030: sleep. https://health.gov/healthypeople/objectives-and-data/browse-objectives/sleep .
  93. U.S. Environmental Protection Agency, 2019. How big a problem is poor indoor air quality in schools? https://19january2021snapshot.epa.gov/iaq-schools/how-big-problem-poor-indoor-air-quality-schools_html.
  94. Wang Y, Raffeld MR, Slopen N, Hale L, Dunn EC, 2016. Childhood adversity and insomnia in adolescence. Sleep Med. 21, 12–18. 10.1016/j.sleep.2016.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  95. Wang M, Sampson PD, Bechle M, Marshall J, Vedal S, Kaufman JD, 2018. National PM2.5 and NO2 spatiotemporal models integrating intensive monitoring data and satellite-derived land use regression in a universal kriging framework in the United States: 1999–2016. Int. Soc. Environ. Epidemiol. Conf Abstracts 2018 (1). 10.1289/isesisee.2018.O02.02.02. [DOI] [Google Scholar]
  96. Young M, Marshall JD, Piepmeier L, Bi J, Kaufman JD, Szpiro AA, 2025. National PM(2.5) spatiotemporal model integrating intensive monitoring data and land use regression in a likelihood-based universal kriging framework in the United States: 2000–2019. Environ. Pollut 366, 125405. 10.1016/j.envpol.2024.125405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  97. Yu HJ, Liu X, Yang HG, Chen R, He QQ, 2022. The association of adverse childhood experiences and its subtypes with adulthood sleep problems: a systematic review and meta-analysis of cohort studies. Sleep Med. 98, 26–33. 10.1016/j.sleep.2022.06.006. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

MMC1
NIHMS2104123-supplement-MMC1.pdf (961KB, pdf)

RESOURCES