Multilevel factors associated with sleep duration and bedtime regularity in U.S. children with and without neurodevelopmental disorders: a nationally representative study

Healthy sleep is a biological necessity for pediatric populations, essential to the optimal development and functioning of all bodily systems including cognitive and neurological, immunological, cardiovascular and metabolic systems [1]. Children with neurodevelopmental disorders (NDDs)—conditions that manifest in early childhood due to central nervous system injury and impair one or more areas of functioning (e.g., physical or motor, social, emotional, behavioral) [2]—are substantially more likely to experience sleep disturbances [3, 4]. Up to 75% of children with an NDD report short and/or irregular sleep as compared to 30% of typically developing (TD) children in the general population [5‐7].

There are several reasons to explain why children with NDDs experience poorer sleep. First, many NDDs involve disruptions to neural circuits that regulate circadian rhythms and sleep–wake cycles, including alterations in melatonin production, neurotransmitter function and brain regions responsible for sleep regulation [8]. For example, children with autism spectrum disorder often show atypical melatonin secretion patterns, while those with ADHD frequently exhibit delayed sleep phase syndrome [9, 10]. Second, core symptoms of NDDs can directly interfere with sleep initiation and maintenance. Hyperactivity and impulsivity in ADHD, repetitive behaviors and sensory sensitivities in autism, and motor symptoms in cerebral palsy can all disrupt bedtime routines and sleep quality [11]. Additionally, medications commonly prescribed for NDDs, including stimulants and antipsychotics, can significantly impact sleep architecture and timing [12]. Third, children with NDDs often experience higher rates of comorbid conditions that further compromise sleep, including anxiety, depression, gastrointestinal issues, and seizure disorders [13]. These comorbidities create a complex web of factors that can perpetuate sleep disturbances.

While the association between sleep and development is complex, achieving the recommended hours of sleep on a regular basis is reliably associated with better emotional, neurobehavioral, neurocognitive and cardiometabolic outcomes in both typically developing (TD) children and those with NDDs [14]. However, the consequences of poor sleep may be particularly severe for children with NDDs. Sleep deprivation can impede cognitive function and learning abilities [15]. In children with NDDs, poor sleep health can exacerbate core symptoms of their condition, creating a cycle where NDD symptoms worsen sleep, which in turn intensifies NDD-related difficulties [16, 17]. Another critical consideration in the context of poor sleep health in children is the negative repercussions for family well-being and functioning. In families with children reporting disordered sleep, the likelihood of parental stress is significantly increased and family functioning decreased [18, 19]. This family-level impact underscores the importance of addressing sleep as a modifiable target that can improve outcomes for both the child and the broader family system.

Despite the well-documented sleep disparities between NDD and TD populations, significant gaps remain in our understanding of the multilevel factors that predict sleep outcomes in these groups. The socioecological framework of sleep health posits that individual, family, neighborhood, and community-level factors interact to determine sleep duration and timing [20]. At the individual level, demographic characteristics (age, sex, race/ethnicity) and health behaviors (physical activity, screen time) have been consistently shown to predict sleep health metrics in pediatric populations [21]. Family-level variables include family structure (household composition, parental marital status), family functioning indicators (family meals, family resilience), and socioeconomic factors (income, parental education, employment) that research has linked to sleep health in children [22]. Similarly, community and environmental factors including neighborhood characteristics (safety, social cohesion, amenities) and healthcare access have been shown to influence sleep health in children [23]. To date, this previous research has focused on clinical samples or single diagnostic categories, limiting generalizability to the broader NDD population [24]. Additionally, few studies have systematically examined whether the same predictive factors influence sleep outcomes similarly across NDD and TD groups [25], or whether different intervention targets may be needed for these populations.

Given these knowledge gaps, the primary goal of this study was to use machine learning methods to identify the strongest multilevel predictors of adequate sleep duration and bedtime regularity in a nationally representative sample of TD children and those with NDDs. Multiple variables representing individual, family, and community levels of the socioecological framework were examined to determine whether predictive patterns differ between groups. Results of this theoretically and empirically guided investigation are expected to identify modifiable intervention targets that can improve sleep duration and regularity for both NDD and TD children.

Data for this study were collected from the National Survey of Children’s Health (NSCH), an annual household survey by the U.S. Census Bureau assessing the health and well-being of non-institutionalized children ages 0–17 years. U.S. households with children are randomly selected from a nationwide address list. A screener determines eligibility, and one child is randomly selected for an age-specific survey (0–5, 6–11, or 12–17 years). A parent or guardian familiar with the child’s condition completes the survey, which can be administered via mail, phone, or online. Publicly available data from the 2016–2017, 2018–2019, and 2020–2021 datasets were used. Weighted response rates were: 40.7% (2016), 37.4% (2017), 43.1% (2018), 42.4% (2019), 42.4% (2020), and 40.3% (2021). Analyses were weighted for selection and nonresponse bias. Detailed survey design and ethical approval information are available elsewhere [26].

Participants in this study were aged 9–17 years and had complete data for study variables. The omnibus dataset with the combined datasets from years 2016–2017, 2018–2019, and 2020–2021 had 123,215 cases. Including children beginning at age 9 years in the analytic sample maximized the likelihood that the presence and subsequent diagnosis of any neurological and developmental diagnosis would have been made since diagnosis by age 8 years is likely [27]. Participants with missing data for any of the study variables (n = 37,201) were excluded, yielding a final analytic sample of 86,014. Participants with missing data on any of the study variables did not significantly differ demographically (sex and race) from the analytic sample, suggesting minimal selection bias on these key characteristics.

Given the large sample size and statistical power to detect significance, we emphasize only associations with moderate effect sizes (Cohen’s d ≥ 0.3). In both NDD and TD groups, bedtime regularity was moderately and significantly associated with adequate sleep duration (Tables 1 and 2).

For children with NDDs, the following were moderately and significantly associated with bedtime regularity: adequate sleep, younger age, food sufficiency, physical activity, lower screen time, absence of depression, better maternal mental health, and frequent family meals (Table 1). Among TD children, similar factors were associated with bedtime regularity: adequate sleep, younger age, food sufficiency, physical activity, lower screen time, better maternal mental health, frequent family meals, and living in a supportive neighborhood (Table 2).

Findings from this investigation revealed that a greater proportion of children with NDDs had parent reports of inadequate sleep duration (34.0% vs. 30.9%) and irregular bedtimes (18.8% vs. 13.7%) compared to TD children. For both groups, having more family meals together and achieving healthy sleep in the other domain (duration or bedtime regularity) were significantly associated with better sleep outcomes. Additionally, current depressive symptoms were linked to lower odds of adequate sleep in TD children and reduced bedtime regularity in both groups. These findings identify potential intervention targets to improve sleep in children aged 9–17 years, with or without NDDs.

Across the multivariable models of parent-reported sleep duration and bedtime regularity, the "other" sleep health variable emerged as the most strongly associated independent variable with each outcome in the current samples. Buysee’s framework of sleep health conceptualizes sleep duration and regularity as distinct, but related constructs for adults [29], while Meltzer and colleagues posit that regularity of sleep (that encompasses bedtime regularity) is inherent to the broader category of sleep-related behaviors that promote or inhibit sleep health in pediatric populations [30]. From a clinical and public health perspective, these findings support the common practice of first-line sleep health interventions for both TD children and those with NDDs. Indeed, sleep hygiene and cognitive behavioral therapy for insomnia approaches provide psychoeducation about child sleep needs, and behavioral modification to promote consistent and, as needed, earlier bedtimes, to allow for sufficient sleep duration [31, 32].

Consistent with Meltzer and colleagues framework for pediatric sleep health that emphasizes the importance of sleep related behaviors [30], data from this investigation showed that family meals together on more days of the week increased the odds of parent-report adequate and regular sleep in both groups. These results align with an earlier daily-diary study conducted in a nationally representative sample of 2454 TD children that showed each additional hour spent eating family meals during weekdays was associated with a 0.37 hours increase in total sleep time for younger children and a 0.57 hours increase in total sleep time for older children [33]. Noteworthy about the current study is that it extends this evidence to children with NDDs.

Family environment factors such as family function, connectedness and routines may help explain the strong link between regular family meals and achieving parent-reported adequate and bedtime regularity in children with and without NDDs. Previous studies have found that more regular family meals are associated with higher family functioning [34], stronger family bonds and connectedness [35], and enhanced feelings of inclusion among children with intellectual disabilities [36]. A large body of literature has reliably shown that in families with high function, characterized by positive and stable parental behaviors and feelings of connectedness between children and parents, children tend to demonstrate fewer sleep disturbances and better sleep health [37]. Relatedly, regular family meals support a robust, predictable rhythmicity in the family home that is a cornerstone of resilience, stability, self-regulation and positive outcomes in children [38‐41]. Homes with frequent family meals may be more likely to have other robust rhythms such as structured bedtime routines, which are strongly linked to healthier sleep behaviors in children and adolescents [22, 42‐44]. Behavioral and communication problems that serve as a defining characteristic for many children with NDDs [45] can make it particularly challenging for parents of children with NDDs to establish family routines (e.g., bedtime routines, regular meal times). Data from the current study underscore the need for pediatric sleep health interventions to address family routines, which may be enhanced by better family functioning. In the context of families with children who have NDDs, these approaches may need to offer more intensive support and consideration of behavioral and communication challenges [46].

Depression was found to be associated with a reduced likelihood of parent reported adequate sleep and bedtime regularity in typically developing children, and bedtime regularity in those with NDDs. These data align with evidence showing a bi-directional association between depressive symptoms and sleep disturbances in children and adolescents [47]. For example, poor and disturbed sleep health in childhood has been shown to independently predict a greater likelihood of increased depressive symptoms and depression in adolescence [48, 49]. Conversely, depression is predictive of subsequent disturbed sleep [50]. This result that depression was associated with bedtime regularity has implications for chronotherapy. Regular sleep is a function of a robust circadian rhythm that regulates patterns of alertness and sleepiness in each  24-hour cycle. Individuals with an evening chronotype or greater eveningness have a propensity for completing the bolus of activity later in the day as well as having later bed- and wake-times. While a shift toward greater “eveningness” is developmentally appropriate in adolescents [51], there is evidence that depression during childhood predicts a greater preference for eveningness post-puberty [52]. Moreover, greater eveningness preference longitudinally predicts increases in depression symptoms [52]. As a manifestation of circadian disruption and abnormal circadian rhythms, delayed and irregular sleep is especially common in children with NDDs including autism and Down’s syndrome as compared to TD children [53, 54]. Mechanisms to explain the higher prevalence of circadian dysregulation in children with NDDs include altered melatonin-cortisol rhythms, genetic polymorphisms affecting clock genes, and differences in light sensitivity that impact circadian entrainment [55, 56]. Given the critical role of circadian stability in mood regulation, these findings reinforce the need to target sleep regularity and circadian alignment in pediatric sleep health interventions, particularly for children with depression and NDDs.

Findings from this study have clinical and public health implications for pediatric sleep health interventions targeting both TD children and those with NDDs. Given the higher prevalence of inadequate sleep duration and irregular bedtimes in children with NDDs, targeted interventions addressing behavioral and environmental factors, such as family routines and sleep hygiene, are essential [30]. The strong association between sleep duration and regularity underscores the need for comprehensive strategies prioritizing both aspects of sleep health [29]. Additionally, the link between family meal frequency and improved sleep in both cohorts highlights the role of structured routines in shaping sleep behaviors [57]. While both NDD and TD children showed similar patterns with family meals and sleep variables emerging as top predictors, some important differences suggest the need for tailored approaches. Food sufficiency appeared more prominently in NDD models, indicating that socioeconomic interventions addressing food security may be particularly important for this population, likely reflecting the additional economic burden faced by families requiring specialized care and services. Conversely, neighborhood social support was more influential for TD children, suggesting that community-level interventions may be less critical than family-level factors for children with developmental challenges. Incorporating family-based routine interventions may be beneficial for all children’s sleep health, regardless of NDD status [58]. The relationship between depression and poor sleep suggests integrating mental health support into sleep interventions [59]. Given circadian dysregulation’s role in sleep disturbances, particularly in children with NDDs, interventions emphasizing chronotherapy and melatonin regulation may be crucial [60]. These findings suggest that while universal family-centered approaches around meal routines may benefit all children, targeted interventions addressing specific socioeconomic needs may be necessary for optimal outcomes in NDD populations. Results from the current study converge with Meltzer and colleagues’ framework [30] and support multifaceted, family-centered, and targeted interventions that address sleep health.

In conclusion, this study shows differences in sleep health between TD children and those with NDDs, suggesting the need for targeted interventions. The strong association between sleep duration and bedtime regularity suggests that comprehensive strategies addressing both factors are essential. Findings also underscore the importance of family routines, particularly regular family meals, and potential social connectedness among family members in promoting healthy sleep behaviors. Moreover, the link between depression and poor sleep highlights the need to integrate mental health support into sleep interventions. Directions for future research in this area include the conduct of prospective, longitudinal studies, using culturally and regionally diverse samples to ascertain the temporal associations of key determinants of device-estimated sleep behaviors. Studies that explore tailored approaches that incorporate chronotherapy and family-centered interventions to improve sleep health in children, particularly those with NDDs, are also warranted.