Sleep Health and Quality of Life in Children and Adolescents with NF1: A Biopsychosocial Perspective

1. Introduction

Neurofibromatosis type 1 (NF1) is a genetic condition that results from a pathogenic variant in the NF1 gene encoding neurofibromin. This protein regulates cell growth, division, and differentiation, primarily by acting as a GTPase-activating protein (GAP) that negatively regulates RAS activity [1]. NF1 is associated with a broad range of medical complications, including tumors of the peripheral and central nervous systems (CNS), musculoskeletal changes, and neurological involvement [2]. Cognitive deficits and an increased risk of neurodevelopmental disorders including attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (hereafter referred to as autism) and specific learning disabilities are all common and collectively may contribute to functional impairment and reduced quality of life (QoL) [3,4,5].

Studies across a range of Nf1 animal models including heterozygous and conditional knockouts, demonstrate that the NF1 gene is required for normal regulation of sleep–wake behavior. In the Drosophila knockout model, a short sleep phenotype [6], altered circadian rhythm [7,8], and sleep fragmentation[6], have been observed. Lower arousal threshold and/or increased sleep fragmentation have also been demonstrated in other heterozygous models of NF1 including zebrafish [9], mice [10] and the minipig [11]. Overall, this evidence suggests that the NF1 gene is important for sleep-wake regulation. Characterizing the sleep problems in humans with NF1, however, is challenging. This is not only because sleep is multidimensional, but also because NF1 is a heterogeneous multisystem condition in which sleep disturbance likely reflects a complex interplay of biological, psychological, and socio-demographic factors.

The few studies that have examined sleep in children and adolescents (hereafter referred to as children) with NF1 have relied on subjective measures (self-report or caregiver questionnaires), often only capturing a single sleep construct. Findings regarding the type of sleep disturbance in NF1 are inconsistent. Johnson et al. [12], using the Simonds and Parraga questionnaire, reported elevated parasomnias in NF1 relative to normative data, whereas Licis et al. [13], using the Sleep Disturbance Scale For Children reported greater problems initiating and maintaining sleep, arousal difficulties, sleep–wake transition disorders, and hyperhidrosis in NF1 compared with unaffected siblings. More recently, objective assessment using actigraphy has demonstrated longer sleep onset latency, poorer sleep efficiency and regularity, shorter sleep duration, and greater night-to-night variability in sleep timing in children with NF1 compared with typically developing (TD) controls [14]. These inconsistencies likely reflect the multidimensional nature of sleep health. Studies capturing different constructs in isolation will naturally yield different findings.

Buysse (2014) posits sleep health as a multidimensional construct rather than a unidimensional measure or the absence of a sleep disorder [15]. It is defined as “a multidimensional pattern of sleep-wakefulness, adapted to individual, social, and environmental demands, that promotes physical and mental wellbeing.” Buysse conceptualized sleep as a positive attribute which contributes to the overall health and functioning of an individual or population [15]. This framework, which has been recently adapted to paediatrics [16], identifies six dimensions of sleep critical to psychological and physical well-being: sleep duration, sleep efficiency, sleep timing, daytime sleepiness, sleep quality, and sleep-related behaviors (pre-sleep activities compatible with sleep). Meaningful impairment can occur across these domains without meeting formal diagnostic criteria, underscoring the need for multidimensional assessment of sleep health in both clinical and research contexts.

Taken together, existing findings indicate that sleep difficulties are common in NF1, yet their nature and determinants remain poorly understood. A biopsychosocial framework offers a more integrative approach. Within this framework, poor sleep may arise from: (1) intrinsic biological differences related to NF1 associated alterations in neurobiological networks; (2) psychosocial factors such as depression, anxiety, pain, cognitive processing, and behavioral and emotional dysregulation; and (3) contextual factors such as socioeconomic status (SES), parenting practices and technology use [17]. For example, emerging evidence indicates that biological alterations are associated with disrupted sleep in children with NF1, including changes in endogenous melatonin secretion [14] and sleep architecture [18]. Research examining links between sleep and the physical manifestations of NF1, however, remains limited. Chronic pain is common in NF1—often related to plexiform neurofibromas, skeletal complications, and headaches [3]—and is a plausible contributor to sleep disruption. Clarifying the sleep–pain relationship in children with NF1 is clinically important given evidence of bidirectional effects [19]. Recent studies also suggest associations between sleep disturbance and ADHD symptoms in NF1 [14,20]. To date, no study has integrated neurodevelopmental traits, cognition, social-emotional symptoms, and pain within a single biopsychosocial study to examine their unique associations with multidimensional sleep health in NF1.

Because the same biopsychosocial factors that we hypothesize are associated with sleep in NF1 (e.g. pain, neurodevelopmental traits, mental health symptoms) also have well-established links to functioning and health-related quality of life (HR-QoL) [4,21,22], it is important to determine whether sleep health explains unique variance in HR-QoL beyond these established predictors. In the general population, sleep duration and sleep quality are consistently linked to QoL [23,24,25], such that shorter sleep and poorer sleep quality are linked to reduced physical, emotional and social functioning. Given that sleep difficulties are amendable to intervention, understanding their contribution to functional outcomes and QoL in NF1 has direct clinical relevance. To date, however, no study has examined this relationship.

This study had three aims. First, we aimed to describe multi-dimensional sleep health characteristics in children with NF1 and compare these domains to a typically developing (TD) control group. Second, we examined the associations between sleep health and biopsychosocial factors in NF1. Third, we examined whether sleep quality and duration predicted HR-QoL alongside established predictors of HR-QoL in children with NF1.

2. Materials and Methods

2.1. Participants and Procedures

Participants were 131 children and adolescents with NF1 recruited through outpatient clinics at The Children’s Hospital at Westmead in Sydney and the Royal Children’s Hospital, in Melbourne. Children aged 6-16 years who met diagnostic criteria for NF1 were included. Those who had received chemotherapy or radiation therapy in the prior 12 months were excluded. A TD control group aged 6-16 years (n = 71), was recruited through research and community networks. Children with a history of developmental delay, neurodevelopmental disorder, or neurological or psychological conditions were excluded. Eligible participants completed a neuropsychological assessment and caregivers completed surveys and questionnaires. Participants were also asked to wear a wrist-worn accelerometer that measures rest/activity cycles for one week and complete a sleep diary the same week. Either the Actiwatch 2 or the Geneactiv accelerometer were provided and data scored using a standard algorithm which has been a validated in paediatric populations and validated across the two devices [26,27]. Participants with fewer than 5 days of data were excluded. Group differences on individual actigraphy-derived outcomes (sleep efficiency, total sleep time, sleep timing) from this dataset have been reported on previously [14]. Here, we adopted a conceptually distinct approach by dichotomising key sleep health measures, consisting of objective actigraphy data and novel subjective questionnaire data, aggregating them into a composite sleep health score as in accordance with the paediatric sleep health framework [16]. The study was approved by Sydney Children’s Hospital Network Ethics Committee. Written informed consent was obtained for all participants. Figure 1 shows the flow of participants into the study and participant exclusion.

2.2. Measures

Sleep Health Composite

We created a multidimensional sleep health composite score capturing six dimensions: sleep duration, sleep timing, sleep quality, daytime sleepiness, sleep efficiency, and sleep behaviors as defined by the sleep health paediatric framework [16]. Each dimension was converted into a binary variable, where 1 represents an unfavourable sleep characteristic and 0 indicates a favourable characteristic (see Table 1). Cutoffs were based on previous literature and informed by clinical and empirical cut points [28,29,30]. The sleep health composite required at least 5 of 6 components to be present. Prorated sums were then calculated, with higher scores indicating poorer sleep health. Further details regarding variable coding are provided in supplementary material (Table S1).

Biopsychosocial Factors

Mental Health was assessed using the Depression and Anxiety Scale T scores from the Behavior Assessment System for Children Version 3 (BASC-3) Parent Rating Scale (PRS) [34], using either the child (6-11 years) or adolescent (12-21 years) form. Given Depression and Anxiety T scores were highly correlated in this NF1 cohort (r = .69), and to avoid multicollinearity, a composite measure of mental health was created by summing and averaging the T scores. Higher scores indicated poorer mental health.

Cognition was assessed with the Full- Scale Intelligence Quotient (FSIQ) from the Wechsler Intelligence Scale for Children 5^th Version Australian and New Zealand Edition (WISC-V A&NZ) [35].

Neurodevelopmental traits were assessed using the ADHD Inattentive and ADHD Hyperactive/Impulsive subscale T scores from the Conners 3 Parent Questionnaire [36] and the Restricted and Repetitive Behaviors T Score and Social Interactions and Social Communication T Score from the Social Responsiveness Scale-Version 2 (SRS-2) [37]. Given the substantial overlap among these traits in NF1 [38], and to reduce multicollinearity (r = .65 - .88), these symptoms were combined into a single neurodevelopmental trait composite by summing and averaging the four T scores. Higher T scores indicated more elevated neurodevelopmental traits.

Biological variables included pain, assessed using the Parent Proxy measure of the Paediatric Quality of Life Neurofibromatosis Type 1 Module (PedsQL-NF1) [39]. The young child (5-7 years), child (8-12 years), and adolescent (13-18 years) versions were used. We selected the pain subscale which has been well-validated in paediatric NF1 cohorts [40]. Scores range from 0-100 with lower scores indicating increased subjective pain. Physical complications of NF1 were reported by parents and classified as a skeletal abnormality (osseous lesion, scoliosis, or sphenoid wing dysplasia), one or more plexiform neurofibroma, an optic pathway glioma (OPG), or CNS abnormality (other than OPG e.g. epilepsy, cerebral tumor, hydrocephalus). Medication use was also recorded and classified as ADHD medication or other CNS medications (e.g. anticonvulsants, antidepressants, analgesics).

Sociodemographic variables included age, sex and socioeconomic status (SES), which were controlled for in all analyses. The Index of Relative Socio-Economic Advantage and Disadvantage from Socio-Economic Indexes for Areas, a standard measure that ranks postal codes based on average income, education and occupational level was used to assess SES [41]. Decile scores range from 1-100 with higher scores indicating higher SES.

Health Related Quality of Life

The Paediatric Quality of Life Inventory (PedsQL) Version 4.0 Generic Core Scales [42] was used to assess HR-QoL. The Total Scale Score, calculated as the mean of all 23 items, was used as the overall generic HR-QoL. Scores range from 0 to 100, with higher scores indicating better HR-QoL. Parent proxy reports for children and adolescents were completed.

2.2.3. Data Analysis

All analyses were conducted using IBM SPSS Statistics software [43] or JASP [44]. Group differences in demographic and clinical characteristics between children with NF1 and TD controls were examined using independent t-tests for continuous variables or chi-square tests for categorical variables. Statistical analyses for all aims controlled for age, sex and SES. For Aim 1, an analysis of covariance (ANCOVA) was conducted to examine group differences in sleep health. Assumptions of ANCOVA were assessed prior to analysis including normality, homogeneity of variance, and homogeneity of regression slopes. To account for minor violations of heteroscedasticity, bootstrapping with 5,000 resamples was used to generate 95% confidence intervals and p-values. Interaction terms between group and each covariate were included to test the assumption of homogeneity of regression slopes. Where significant interactions were identified, these were retained in the model and probed using estimated marginal means to examine simple effects (i.e. group differences at each level of the covariate). Effect sizes are reported as partial eta squared (ηp²). Exploratory analyses were also performed to examine group differences in impairment rates for each binary sleep health dimension. Firth penalised logistic regression was performed with each sleep health dimension as the dependent variable and group, age, sex and SES entered. Odds ratios and 95% confidence intervals are reported. Group mean differences in sleep behavior subdomains (Children’s Sleep Hygiene Scale) [33] were also examined with separate ANCOVAs.

For aim 2, a multiple linear regression analysis was conducted to examine the association between biopsychosocial predictor variables (neurodevelopmental traits composite, mental health composite, FSIQ, and PedsQL-NF1 pain) and the sleep health composite. Predictors were selected a priori based on theoretical relevance within a biopsychosocial framework[17]. Categorical variables (e.g. sex) were entered using dummy coding. All other variables were continuous. Variables were reversed (where appropriate) and standardized for regression. Spearman rho correlations were also performed to examine relationships between sleep health dimensions and biopsychosocial predictors. Spearman correlations were exploratory and presented using unadjusted p-values. For aim 3, a multiple linear regression was conducted to examine predictors (e.g. mental health composite, neurodevelopmental traits composite, sleep quality, sleep duration, and pain) of PedsQL Generic Total Score in the NF1 group. Prior to interpretation of results, all assumptions of multiple linear regression were evaluated. Assumptions of logistic regression were evaluated using the Box-Tidwell procedure where possible. Visual inspection of scatterplots indicated the assumptions of linearity and homoscedasticity were met. Examination of histograms and normal Q-Q plots of standardised residuals suggested residuals were approximately normally distributed. Independence of errors was supported by a Durbin-Watson statistic within acceptable limits. Multicollinearity was not observed, with variance of inflation factors within acceptable ranges. Standardized residuals did not exceed ±3, indicating no evidence of problematic outliers. Standardised beta coefficients (β), unstandardised coefficients (B), confidence intervals, and R² values were reported to describe effect sizes and model fit.

3. Results

3.1. Sample Characteristics

Table 2 and Figure 2 show descriptive statistics for the groups, including percentages for dichotomous variables; means and standard deviations for continuous variables. Corresponding descriptive statistics for participants with available sleep health data (N=168) are provided in Supplementary Material (Table S2).

3.2. Group Difference in Sleep Health

An ANCOVA examining group differences in sleep health revealed a significant main effect of group F(1,163) = 12.68, p <.001, ηp² = .07, and a significant main effect of age (F(1,163) = 12.13, p =.001, ηp²=.07). SES was not significant (p =.396). The interaction between group and sex was significant, F(1,163) =5.26, p = .023, ηp² = .03, indicating that group differences in sleep health were dependent on sex. Follow-up analyses probing the interaction indicated that within males, participants in the NF1 group had significantly poorer sleep health scores compared to controls (mean difference = -1.18, CI [-1.68,-.69], p <.001, ηp² = .20). In contrast, there was no significant group difference in females (mean difference = -.11, CI [ -.74,.51), p = .713, ηp²= .002). These results suggest that sex moderated the relationship between group and sleep health, such that group differences are most pronounced in males, but attenuated in females (Figure 3a).

3.3. Group Difference in Sleep Health Dimensions

Compared to their peers, children with NF1 showed higher frequency of impairment across all sleep health dimensions including sleep quality, sleep duration, sleep efficiency, sleep timing, daytime sleepiness and sleep behaviors (Figure 3b). Following adjustment for age, sex and SES, children with NF1 had greater odds of impairment in sleep quality, daytime sleepiness, sleep efficiency and sleep duration compared to unaffected peers. (Figure 3c).

Given the sleep behavior dimension has multiple subdomains, we examined group differences for each subdomain (Figure 3d). Children with NF1 demonstrated poorer cognitive behaviors (e.g. cognitive hyperarousal before bed; p = .002), emotional behaviors (e.g. emotional state at bedtime such as worrying in bed (p = .011), and difficulties with bedtime routine (p =.020) when compared to unaffected children (Figure 3c).

3.4. Biopsychosocial Correlates of Sleep Health in NF1

The multiple linear regression model predicting sleep health in children and adolescents with NF1 was statistically significant F(7,95) = 8.53, p < .001, explaining 38.60% of the variance in sleep health (R² = .39) representing a large effect size (Table 3). Examination of individual predictors indicated that male sex was associated with poorer sleep health. Neurodevelopmental traits and pain were also significant predictors of poor sleep health in NF1. Age, mental health, SES, and FSIQ were not significantly associated with sleep health in children with NF1 once adjusting for other variables in the model.

3.5. How Do NF1 Complications Influence Sleep Health in NF1

Given the relationship between poorer sleep health and elevated pain and neurodevelopment traits in NF1, further exploratory analyses were performed to examine whether sleep health differed between those with and without various NF1 physical complications. Children with familial (inherited) NF1 had poorer sleep health (t =-2.61, p = .01) compared to those with sporadic NF1 (de_novo NF1 variation). There was no difference in sleep health between those with and without a skeletal complication, a plexiform neurofibroma, a CNS abnormality or an OPG (all p > .05). In those with an ADHD diagnosis, sleep health was similar across those who took ADHD medication (n=26) and those with ADHD who did not (n=24. p > .05). In contrast, children who were taking other CNS medications (n=22) had poorer sleep health (M =3.01, SD = 1.28) compared to those who do not take CNS medications (n=89, Mean =2.44 SD = 1.14) F =4.31, p = .04).

3.6. Biopsychosocial Correlates of Individual Sleep Health Dimensions in NF1

We next examined the relationship between biopsychosocial factors and individual sleep health dimensions in children with NF1 (Figure 5). Children with elevated pain had lower sleep efficiency (p =.002), greater daytime sleepiness (p <.001), and worse sleep behaviors (p <.001). Children with elevated neurodevelopmental traits had greater daytime sleepiness, poorer sleep quality and poorer sleep behaviors (all p <.001). Elevated mental health was also associated with greater daytime sleepiness (p <.005), and worse sleep behaviors (p <.005). A higher FSIQ was associated with better sleep behaviors (p = .040). Male sex was associated with lower sleep efficiency (p = .006) and shorter sleep duration (p =.027).

3.7. Health Related Quality of Life and Sleep in NF1

Children with NF1 had substantially lower parent-reported health-related quality of life (HR-QoL) compared with TD controls, even after adjusting for age, sex and SES (adjusted mean 65.7 vs 87.9; p < .001) representing a large effect (ηp² = .25). A linear regression model was conducted to examine predictors of HR-QoL in children with NF1 (Table 4). The overall regression model was statistically significant explaining 69.7% of the variance in HR-QoL representing a large effect size (Table 4). Findings demonstrate poorer mental health, elevated neurodevelopmental traits, greater pain, shorter sleep duration and male sex predicts poorer HR-QoL in children with NF1. Although univariate regression showed both sleep quality and duration were strongly associated with HR-QoL, only sleep duration remained significant after adjustment for other predictors, indicating a unique contribution of sleep duration to HR-QoL in NF1.

4. Discussion

Using Buysse’s (2014) sleep health framework which emphasizes the importance of multiple sleep characteristics for overall health and functioning, this study examined whether sleep health – including sleep timing, sleep quality, daytime sleepiness, efficiency, duration and sleep-related behavior - is affected in children and adolescents with NF1. Compared to their typically developing peers, children with NF1 demonstrated poorer sleep health overall. While higher rates of impairment were evident across the six sleep health dimensions, impairments in sleep quality, daytime sleepiness, sleep efficiency and sleep duration, derived from clinical cutoffs, emerged as the strongest contributors. Poor sleep quality was a predominant complaint occurring in approximately 30% of children with NF1, a rate 5 times greater compared to unaffected children. Almost 78% of children with NF1 were found to have impaired sleep efficiency, suggesting most children with NF1 experience difficulties initiating and maintaining sleep based on objective assessment. As a result, nearly half the cohort were not getting a sufficient amount of sleep at night based on guidelines from the National Sleep Foundation [28].

The detrimental effects of poor sleep health, particularly insufficient sleep quantity and quality, on children’s cognitive, emotional, and overall functioning are well established [45,46]. The current findings extend this evidence to NF1, demonstrating that both sleep quality and quantity contributed to health-related quality of life (HR-QoL) in this population. Collectively, this evidence reinforces the clinical imperative to routinely assess sleep in NF1 care and the importance of proactive sleep health education for families. All patients with NF1 would benefit from screening of their sleep at routine clinic visits. If screening is positive, patients may benefit from a more detailed assessment of their sleep. Unfortunately, evidence-based recommendations for the screening and management of sleep disorders in patients with NF1 are limited due to the paucity of studies evaluating sleep disorders and therapeutic interventions in this population.

The current study identifies a number of biopsychosocial correlates of sleep health in NF1 (Figure 6). In terms of biological factors, males experienced shorter sleep duration and lower sleep efficiency than females. Although, sex differences in sleep are found in the general population [47], they are often linked to puberty [48], emerging in adolescence with mixed findings on which sex is impacted more. In this study, sex differences were independent of age, suggesting they manifest prior to puberty in NF1. Some have suggested that an increased engagement with screens and gaming in males might contribute to sex differences in sleep in children from general population [48]. We did not find a significant relationship between sex and sleep behaviors in NF1. Converging preclinical evidence from various NF1 animal models suggests a neurobiological basis for the sex-based findings of reduced sleep depth, lower arousal thresholds and increased fragmentation in males. Male Nf1-null flies show shorter sleep duration compared to wildtype males, a difference not found in females [6]. Interestingly, sex differences in neuronal dopamine, a neurotransmitter that regulates the generation of sleep-wake states [49], have also been shown in studies involving Nf1-conditional knock out mice, with males exhibiting decreased hippocampal dopamine compared to females [50]. Despite this, sex is typically not found to be a significant determinant of the behavioral phenotype in NF1 [51]. Taken together, these findings suggest boys with NF1 represent a group at particular risk for poor sleep health, and highlight the need for future studies to consider sex-specific mechanisms that may modulate or influence sleep circuitry and arousal systems in NF1.

The strongest correlate of sleep health in NF1 was elevated neurodevelopmental traits , specifically ADHD and autism. More specifically, children with elevated neurodevelopmental traits experienced more difficulties with sleep behaviors, poorer sleep quality, and reduced sleep efficiency. This aligns with recent evidence which has shown an ADHD diagnosis to be associated with sleep disturbance and sleep impairment (using the PROMIS) in youth with NF1 [20]. The mechanisms explaining the link between sleep and neurodevelopmental traits in NF1 are likely complex and are yet to be identified. NF1 is implicated in several Ras pathways including RAF/MEK/ERK,[1] Akt/mTOR and cAMP[52], as well as GABAergic and dopaminergic signalling processes. These pathways have been implicated in increased risk for neurodevelopmental outcomes in NF1 animal and human studies [52], as well as sleep regulation and circadian functioning in animal models of NF1 [6,9,10,53,54,55]. Emerging research, although still in its early stages, suggests that alterations in neuroplasticity and an imbalance in excitation-inhibition may be one mechanism contributing to the increased risk of sleep and circadian disruption in children with neurodevelopmental conditions [56]. Future research focusing on clarifying shared neurobiological pathways of sleep and neurodevelopment, including arousal-regulation pathways, are needed in humans with NF1.

Our study found that sleep health was similarly impaired across children with NF1, irrespective of physical burden of the condition, including skeletal abnormalities, plexiform neurofibromas, or CNS abnormalities. This aligns with preclinical evidence showing that zebrafish with one wildtype NF1 allele exhibit disrupted sleep at a developmental stage prior to the establishment of manifestations such as tumors [9] and may suggest that the poor sleep health in NF1 is not secondary to NF1 physical complications.

Although overt physical NF1 complications were not associated with sleep health in this cohort, pain remained an important correlate of sleep difficulties. One possible explanation for why sleep was associated with pain, but not the physical manifestations of NF1, could be that pain reported in this study arises from mechanisms not captured by these broad complication groupings, such as headache pain,⁷ restless leg pain,²⁰ or altered nociceptive processing. Whatever the cause of pain in these children, we found elevated subjective pain was associated with lower sleep efficiency, shorter sleep duration, greater daytime sleepiness, and poorer sleep behaviors in children with NF1. These associations may reflect a pathway whereby pain-related distress and pre-sleep cognitive/emotional arousal (e.g., worrying about sleep) contribute to difficulties initiating and/or maintaining sleep, resulting in shorter sleep and increased sleepiness the following day. Given the cross-sectional design, however, we cannot determine the directionality of these relationships. Indeed, a systematic review that included different pain conditions in paediatric populations support a bidirectional relationship between sleep and pain [57]. Various mechanisms have been proposed to explain this bidirectional relationship, including disruption to endogenous pain modulation, where pain thresholds are lowered due to insufccient sleep [19]. For example, studies that have manipulated sleep in healthy individuals have shown that partial sleep deprivation can impair endogenous pain modulation. Restriction of sleep over a 3 week period led to increases in pain, decreased pain habituation and increased temporal summation suggesting exposure to chronic insufficient sleep may increase vulnerability to chronic pain by alterations in pain sensitization [58]. Interestingly, evidence in the minipig model of NF1 (Nf1+/^ex42del), suggest that there may be biologically linked pathways to pain and sleep with nociceptive signalling linked to pain behaviors and sleep quality [11]. Overall, our findings indicate that pain is an important correlate of poorer sleep health in NF1, highlighting the importance of integrated assessment and management of pain and sleep in NF1.

Our study has several clinical implications and raise some important avenues for future research. Targeting pre-sleep arousal, particularly in those with NF1 + ADHD/autism, may provide a potential treatment route for poor sleep efficiency (especially sleep onset issues). Cognitive behavioral therapy for insomnia (CBT-I) is a structured evidence based approach that has been shown to reduce pre-sleep arousal in those with adults with insomnia [59], and has been shown to improve sleep quality and onset in adolescents with emerging evidence in school aged children [60]. Behavioral sleep interventions which focus on sleep hygiene, progressive muscle relaxation, and stimulus controls are also a promising avenue for future research and may assist those with elevated pain or difficulties with sleep behaviors. Indeed, a number of systematic reviews support the use of behavioral sleep interventions in children [61] as well as those affected by ADHD with moderate effect sizes reported [62]. The absence of any published trials, targeting sleep disruption in NF1, highlight an urgent and unmet clinical need, particularly given the functional consequences for HR-QoL identified in the current study.

An important limitation of the current study is that the threshold used to define impaired sleep timing irregularity (i.e. > 1SD in midpoint sleep variability) and daytime sleepiness was empirically derived rather than based on an established paediatric clinical cutoff. Although this approach is consistent with efforts to operationalise multidimensional sleep health) [30], it may have reduced sensitivity to more subtle between group differences in sleep timing irregularity which have been shown previously [14] and thus should be interpreted as indicating relative, rather than clinically validated irregularity.

In conclusion, sleep disruption is markedly more common in children and adolescents with NF1 than their neurotypical peers, supporting the value of a multidimensional sleep health framework in this population. Sleep difficulties in NF1 were not limited to a single domain, but were evident across domains including sleep quality, sleep efficiency, sleep duration, and sleep-related behaviors, highlighting the broad and clinically meaningful nature of sleep disruption. The findings also suggest that poorer sleep health in NF1 is associated with multiple interacting biopsychosocial factors, with male sex, elevated neurodevelopmental traits, and pain emerging as the strongest correlates.

Sleep duration uniquely contributed to health-related quality of life beyond established predictors, indicating that sleep may represent a modifiable and under-recognised treatment target in NF1. Together, these findings support the routine assessment of sleep as part of comprehensive clinical care for children and adolescents with NF1 and underscore the need for future longitudinal and intervention studies to clarify mechanisms, identify children at greatest risk, and determine whether improving sleep health can enhance broader developmental and functional outcomes.