Effects of sleep disturbance on neuropsychological functioning in patients with pediatric brain tumor

While the effects of sleep on cognition in typically developing children are well established, there is a paucity of research in patients with pediatric brain tumor (PBT), despite their increased risk for sleep-related disturbances. The aim of this retrospective study was to examine the impact of sleep factors on patient-reported outcome (PRO) measures, including adaptive and executive functioning within this population. 133 patients with PBT (52% male) ages 5–23 (x̄ = 12.8 years; SD = 4.5 years) underwent neuropsychological evaluation, including assessment of adaptive and executive functioning. Subjective sleep concerns, nocturnal sleep duration, and daytime sleep behavior were also collected and compared to age-based guidelines. Nearly 30% of patients endorsed subjective sleep concerns, while the sample as a whole reported reduced nocturnal sleep duration (approximately one hour below age-based recommendations). Despite the expectation for monophasic sleep by age five, nearly half of the sample reported consistent daytime napping. Inadequate sleep predicted decreased adaptive functioning, F(1, 56) = 4.23, p = .022 (R2 = .07), as well as increased symptoms of executive dysfunction, F(1, 108) = 3.51, p = .003 (R2 = .03). Patients with PBT demonstrate several aspects of abnormal sleep, which are associated with poorer long-term PROs. Further exploration of diagnostic, treatment-related, and demographic variables will be needed to better understand these relationships among patients with PBT in order to inform appropriate interventions.


Introduction
Sleep is essential for daily functioning and is necessary for multiple aspects of development, as well as energy conservation [1] and brain waste clearance [2]. Sleep also helps modulate cognition, disease, immune response, and psychological state [1][2][3]. Although adequate sleep is known to be imperative for optimal functioning, average sleep duration (i.e., length of sleep) from infancy through adolescence has steadily declined within the general population over the last 20 years [4,5]. The consequences of insufficient sleep or reduced sleep duration in children have been associated with a wide range of adverse physical and mental health outcomes [5], as well as deficits in cognition [6][7][8][9][10], and executive functioning (EF; higher order thinking skills) in particular [9]. Increased risk and incidence of psychiatric disorders are also observed [11,12], as sleep and mental health appear to be bidirectional processes [13]. Accordingly, sleep difficulties in children are associated with poorer quality of life and secondary familial effects (e.g., maternal stress) [14]. Sex and race/ethnicity are important factors to examine as well, considering prior studies reveal that girls consistently demonstrate greater sleep duration and shorter sleep latency (i.e., amount of time to fall asleep) than boys [15], while ethnic/racial minority [16] is identified as a risk factor for disrupted sleep.
While daytime napping is important for early cognition, transition to monophasic sleep by age five is a developmental marker of brain maturation [17]. As children age, lower sleep quality, longer sleep latency, more arousals, and increased time spent in lighter sleep may be associated with napping behaviors due to circadian misalignment, as the light-dark cycle and endogenous circadian rhythms are at risk of deviation [18]. Additionally, according to the twoprocess model of sleep regulation, napping behaviors can interfere with circadian regulatory processes by delaying the accumulation of sleep pressure necessary for successful sleep initiation [19].
Medical conditions increase the risk of sleep disruption in children [20]. Pediatric cancer and associated treatments significantly interfere with multiple aspects of sleep, such as sleep latency [21], sleep duration [22], daytime sleepiness [23], sleep habits (e.g., disturbed sleep-wake rhythm) [24], and fatigue [25]. Pediatric patients both with and without central nervous system involvement of their cancer experience greater disturbances related to sleep initiation and maintenance compared to typically developing peers [26]. However, patients with pediatric brain tumor (PBT) in particular are at increased risk for sleep-related deficits, as the location of the tumor (e.g., with disruption to the hypothalamic-pituitary region) and associated focal treatments (e.g., resection, cranial irradiation) may disrupt circadian rhythms [27][28][29]. Moreover, increased daytime sleepiness and longer sleep duration in survivors of PBT can persist years after treatment into adulthood [25].
Neurocognitive impairment (including executive dysfunction) has been found to be correlated with fatigue and sleep problems in long-term female survivors of childhood leukemia, though not with their male counterparts [30]. Additionally, recent investigation of survivors of PBT with neurocognitive complaints in Amsterdam revealed 48% presented with sleep problems, which was associated with parent-reported executive dysfunction as measured by the Behavior Rating Inventory of Executive Function (BRIEF) [31]. While the need for adequate sleep to function properly is well documented, the association between inadequate sleep and adaptive functioning (AF) deficits among patients with PBT has yet to be explored, despite poor adaptive functioning being a known late effect of pediatric cancer [32].
Given the nature of how sleep needs and patterns shift drastically throughout the developmental period, the primary goal of this retrospective exploratory study was to further investigate sleep factors (i.e., subjective sleep concerns, nocturnal sleep duration, daytime napping patterns) across different age ranges of patients with PBT, as well as examine the impact of these sleep factors on reported functioning.

Patients
Archival data were used to conduct a retrospective study investigating sleep and neuropsychological outcomes in patients with PBT ages five and older who underwent evaluation through the Cancer and Blood Disease Institute (CBDI) at Children's Hospital Los Angeles (CHLA) between September 2014 and February 2020. Exclusionary criteria included intellectual disability, stroke, extreme prematurity, autism spectrum disorder, and insufficient sleep data. The study was approved by the CHLA Institutional Review Board as part of a larger project on neuropsychological sequelae of pediatric cancer.

Measures
Sleep data (e.g., estimated sleep duration and presence of napping/sleep concerns) were collected through clinical interview and background history form. Age-based recommendations for sleep duration were obtained from the National Sleep Foundation (Table 1).
Patient-reported outcome (PRO) measures were selected for this study given their ecological validity, as they have been shown to be especially sensitive to functional impact of cancer-related impairment and capture the nuances of how neuropsychological concerns manifest in real-word environments [33].

The General Adaptive Composite (GAC) of the Adaptive Behavioral Assessment System-Third Edition (ABAS-3) and Global Executive Composite (GEC) of the BRIEF-Preschool Version, Second Edition, or Adult Version (BRIEF-P, BRIEF-2, BRIEF-A)
were used to assess AF and EF, respectively. The ABAS-3 has a mean (x) Standard Score (SS) of 100, with a standard deviation (SD) of 15. GAC scores range from 40 to 120, with higher scores reflecting better functioning. Test-retest reliability is consistent with other behavior rating scales (0.86) and GAC is strongly correlated with both verbal/nonverbal intelligence (0.91) and alternate measurement of adaptive functioning (0.80). BRIEF GEC scores are presented as T-Scores (x̄ = 50; SD = 10), with higher scores reflecting increased concerns. GEC scores have strong internal consistency (0.95-0.97) and test-retest reliability (0.88-0.94), as well as substantial evidence of validity based on numerous convergent and discriminant relationships with similar and dissimilar constructs respectively. As is common practice, caregiver-report was used as a proxy for PRO depending on age.

Statistical analyses
Analyses were conducted using IBM SPSS Statistics, Version 27. An alpha level of 0.05 was used to indicate statistical significance (with < 0.01 noted when applicable). Demographic, sleep-related, and PRO variables were analyzed using x̄ and SD for continuous variables, as well as frequencies and percentages for categorical variables. Chi-square analyses were conducted to identify differences in categorical sleep factors based on demographic differences (e.g., sex, ethnicity, age group). One-sample t-tests were used to examine whether PRO measures differed from age expectations, while linear regression analyses were conducted to explore reduced nocturnal sleep duration as a predictor of significant functional concerns. Potential outliers were calculated using a 2.2 multiplier given increased validity associated with this approach [35]. One case was identified as an outlier and excluded from associated analyses. Nonparametric analyses (Mann-Whitney, Kruskal Wallis) were utilized in the case of violated normality assumptions.

PRO measures
As expected, overall AF was poorer than expected for patients with PBT, with age-based normative scores just about one SD below the mean on average (x̄ GAC SS = 85.8).
A relatively similar downward trend in functioning was identified across AF domains, with the most noticeable concerns in practical AF (x̄ Conceptual SS = 87.6; x̄ Social SS = 91.3, x̄ Practical SS = 85.3). However, for the sample as a whole, no significant EF concerns were evident (x̄ GEC T-score = 51.5). See Table 3 for findings by age group.

Subjective sleep concerns
118 patients provided information regarding subjective sleep concerns and of those, nearly 28% reported concerns. Subjective sleep concerns were most notable among young adults (44% of patients 18-23 years of age) and least notable among preschoolers (17% of patients five years of age). However, no significant differences in reported sleep concerns were identified between age groups. See Table 3 for additional frequencies.
Patients with subjective sleep concerns were also reported to have increased executive dysfunction as compared to those without subjective sleep concerns (U = 573, p = 0.002). More specifically, EF for patients without subjective sleep concerns was noted to be average (x̄ GEC T-score = 49.9), whereas those with subjective sleep concerns were approximately 2/3 SD above the mean (x̄ GEC T-score = 56.2). While this latter value does not exceed the BRIEF cutoff for clinical significance, it statistically exceeds normal levels in the general population [t(109) = 1.68, p = 0.048], and likely represents a trend toward increased executive dysfunction in these patients. While AF was lower than expected in the sample as a whole, there were no significant differences between patients with and without subjective sleep concerns.

Nocturnal sleep
For the sample as a whole, reported nocturnal sleep duration was approximately 1 h below National Sleep Foundation age recommendations (x̄ = 0.98 h). This varied significantly by age, H(3) = 11.15, p = 0.011, with young adults obtaining the largest deviation from recommended sleep duration (x̄ = 1.83 h less); however, this does not take into consideration daytime nap duration since typical young adults are expected to achieve monophasic sleep. Adolescents and school-aged children achieved 0.98 and 0.71 h less than recommended for their ages, respectively. Nocturnal sleep was not interpretable for preschoolers based on age recommendations, which allow for daytime napping.
Nocturnal sleep and executive dysfunction were inversely related in that longer sleep duration predicted better EF within the sample as a whole, F(1, 108) = 3.51, p = 0.003 (R 2 = 0.03). Reduced duration also significantly predicted poorer general AF in the overall sample, F(1, 56) = 4.23, p = 0.022 (R 2 = 0.07). However, when further post-hoc analyses were conducted to compare patients with and without endorsed sleep concerns, these findings were only remarkable in conjunction with reported sleep disturbances for both EF,

Daytime napping
65 patients (47%) reported daytime napping, which is not considered developmentally appropriate after age five. There were no significant differences in napping behavior between age groups, though young adults (ages 18-23) notably had the largest frequency of daytime napping. See Table 3 for additional daytime napping frequencies.

Discussion
The purpose of this study was to investigate the effects of sleep on neuropsychological functioning in patients with PBT. Sleep factors (i.e., subjective sleep concerns, daytime napping, nocturnal sleep duration) across different age ranges within this population were explored, as well as their effect on reported AF and EF outcomes. With regard to sleep factors, insufficient sleep duration was prevalent among patients with PBT. Reported subjective sleep concerns were found in nearly 28% of the sample, as compared to 20-25% in the general population [36]. Moreover, nearly half (47%) of patients with PBT ages five and older were napping regularly, despite napping being developmentally atypical by this age [17]. Napping behavior did not vary significantly between age groups, as napping was more prevalent than expected across all age ranges; however, young adults (ages 18-23) surprisingly had the largest frequency of daytime napping (60%). This is particularly interesting, as napping behaviors would be expected to decrease as one ages.
Consistent with prior literature, there was a downward trend in AF for patients with PBT, particularly for the practical domain [32], as well as increased symptoms associated with executive dysfunction as compared to typically developing peers [31]. In terms of clinical significance, mean AF scores for this sample were nearly one SD below the normative mean and did not differ between patients with and without reported sleep concerns, whereas EF scores were not clinically elevated (though approximately 2/3 SD above the mean for patients with subjective sleep concerns in particular). Inadequate sleep duration was associated with poorer AF and EF outcomes among patients with reported sleep concerns.
The above findings highlight the importance of addressing sleep-related concerns following diagnosis of PBT. Given increased risk for subjective sleep disturbances, as well as insufficient nocturnal sleep duration and increased need for daytime napping among this population, early identification of sleep problems can allow more timely provision of supports, such as appropriate referrals and therapies. Additionally, given the relationship between sleep factors and functional outcomes (both in terms of higher order thinking abilities and practical daily living skills), screening for sleep-related concerns can help inform treatment planning and ensure accommodations, modifications, and interventions to mitigate these sequelae. This allows clinicians to provide psychoeducation about sleep needs and benefits, as well as behavioral strategies related to sleep hygiene. In severe cases, referral to a therapist who specializes in behavioral sleep medicine may be warranted. School-based supports may also be implemented to address fatigue (informally or through a section 504 Plan or Individualized Education Program), such as monitoring a student's fatigue, structuring schedules with tests or more demanding coursework when most alert, or modification of assignments as needed.
The primary limitation of the current study relates to the retrospective design, which did not allow predefined sleep criteria or collection of certain data outside of the predetermined dataset (e.g., medication use). As a result, analyses were restricted, leading to a more exploratory approach. With regard to medications, the vast majority of patients seen in this clinic are not prescribed antiepileptic medications or steroids beyond the initial post-resection  period (prior to typical neuropsychological evaluation), though lack of data related to the rare patients who are taking medication that may impact sleep is still a study limitation and further research should investigate the role of medication in sleep disruption for patients with PBT. As for sleep data, given the nuances of both organic and inorganic sleep disturbances, utilizing validated measures of sleep (e.g., Child Sleep Habits Questionnaire) in combination with direct measurement (e.g., actigraphy) is recommended to obtain sleep duration, latency, and maintenance in future studies. It would also be beneficial to examine potential differences in tumor presentation (e.g., location), treatment-related factors (e.g., radiation), and additional demographic variables (e.g., socioeconomic status), particularly since associations described in the current study do not indicate causal relationships. More specifically, further investigation of factors contributing to the relationship between inadequate sleep and poor functional outcomes would help clarify future directions for identification and treatment.

Conclusion
Compared to the normative population, the above findings show patients with PBT experience shorter nocturnal sleep duration and greater occurrence of sleep concerns, with sleep factors associated with poorer adaptive and executive functioning (as reported on PRO measures). Future studies should aim to explore additional diagnostic, treatment, and demographic factors to better understand the association between sleep and neuropsychological outcomes in this population.
Author contributions JAC and AMW designed the study. JAC and AS completed data collection. AMW performed statistical analyses. JAC, PTF, and AMW interpreted findings and prepared the manuscript. All authors approved final manuscript.
Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability Summary data supporting results reported in this article can be requested by emailing the senior author at awhitaker@ chla.usc.edu.

Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval This retrospective study was approved by the institutional review board of Children's Hospital Los Angeles, Los Angeles, CA.