Carer-reported sleep disturbance and carer- and teacher-rated executive functioning in children with prenatal alcohol exposure and Fetal Alcohol Spectrum Disorder

ABSTRACT

Children with prenatal alcohol exposure (PAE) and Fetal Alcohol Spectrum Disorder (FASD) have high rates of sleep disturbance and marked difficulties with executive functioning (EF). Sleep disturbance has been associated with poorer EF across development in typically developing children. The contribution of insomnia symptoms and nightmares to EF difficulties in children with PAE and FASD is unclear. The current study examined whether caregiver-reported insomnia symptoms and nightmares predicted difficulties with EF in children with PAE who were assessed at FASD diagnostic clinics. Archival data on 116 children with PAE assessed at FASD diagnostic clinics were extracted from databases. Children were assigned to a preschool-age group (3.1 to 5.9 years, n = 40) and a school-age group (5.9 to 10.9 years, n = 76). Insomnia symptoms and nightmares were measured using items extracted from the Child Behavior Checklist (CBCL) while EF was measured using the caregiver and teacher Behavior Rating Inventory of Executive Function (BRIEF) rating forms. Bootstrapped regression models were used examine the effects of insomnia symptoms and nightmares on domains of EF in each group while adjusting for potential confounds. For preschool children, insomnia symptoms were associated with greater daytime tiredness while nightmares were associated with greater difficulties with Emergent Metacognition according to their teachers. For school-age children, insomnia symptoms predicted greater EF difficulties across most domains according to their caregivers but not teachers. Sleep disturbance may compound EF impairments in children with PAE and should be screened for as part of FASD diagnostic assessment.

KEYWORDS:

• Sleep disturbance
• FASD
• executive functioning
• development
• prenatal alcohol exposure

Alcohol consumed during pregnancy can pass through the placenta and have a teratogenic impact on the development of the fetus (Burd et al., 2012). After birth, children with prenatal alcohol exposure (PAE) exhibit a continuum of adverse outcomes, which can include dysmorphic facial features and body parts, damage to internal organs and bodily systems, brain injury, and neurodevelopmental impairment (Popova et al., 2016; Wozniak et al., 2019). A diagnosis of Fetal Alcohol Spectrum Disorder (FASD) is assigned to children with confirmed PAE and patterns of severe neurodevelopmental impairment (Astley, 2004; Hoyme et al., 2016). In some diagnostic systems, a designation of “At Risk” for FASD is given in the presence of confirmed alcohol exposure when the assessment is inconclusive, impairment is present in fewer than three neurodevelopmental domains, or is present but is not sufficiently severe to meet criteria (Bower & Elliott, 2016; Cook et al., 2016).

The neurodevelopmental consequences of PAE are complex and diffuse, affecting domains such as motor function, cognition, memory, learning, attention, affect regulation, and executive functioning (Bower & Elliott, 2016; Dawe et al., 2023). Impairments across these domains often lead to poor adaptive outcomes in children with FASD, and higher risk of involvement in the justice system (Popova et al., 2019). Better understanding the factors that influence neurodevelopment in children with a diagnosis of FASD or an “At Risk” designation will inform the focus and content of interventions that aim to improve outcomes in this vulnerable population.

An influential factor, and target for intervention, that has potential to influence the neurodevelopmental trajectory of children with FASD is disturbed sleep (Inkelis & Thomas, 2018). Sleep disturbance can be classified into insomnia and parasomnias (Maski & Owens, 2016). Insomnia typically refers to difficulties with initiating sleep, maintaining sleep, and waking early in the morning, and can include behavioral difficulties associated with bedtime. Parasomnias refer to undesirable phenomena that occur before, during, or waking from sleep. Nightmares are one of the most commonly occurring parasomnias in children (El Sabbagh et al., 2023) and can suggest abnormalities in emotional processing (Kovachy et al., 2013). Notably, they are also associated with insomnia symptoms such as greater waking after sleep onset (Rolling et al., 2023).

The direct influence of PAE on circadian and arousal regulation may drive sleep disturbance in children with FASD (Inkelis & Thomas, 2018), along with contributions from exposures to postnatal environmental stressors, such as adversity and trauma (Chandler-Mather et al., 2023). Parents and carers report that children with FASD exhibit poorer sleep than typically developing peers, with between 55% and 85% of children with FASD scoring above the clinical cutoff score for clinical sleep problems compared to between 20% and 45% in control samples of children without FASD (Chen et al., 2012; Dylag et al., 2021; Gerstner et al., 2023; Mughal; Joyce et al., 2020).

Sleep is thought to play a key role in supporting neurodevelopment across childhood (Lokhandwala & Spencer, 2022; Mason et al., 2021). Various functions have been assigned to sleep, including neural consolidation and neuronal waste clearance, as well as the restoral of daytime arousal, that are thought to drive its protective effect on cognitive and behavioral performance and neurodevelopment over time (Tononi & Cirelli, 2014; Turnbull et al., 2013). Several neurodevelopmental domains are thought to be affected by sleep disturbance (Mason et al., 2021).

Particular focus has been paid to the effects of sleep disturbance on the neurodevelopmental domain of executive functioning (Turnbull et al., 2013), which refers to a set of cognitive functions that support the planning and execution of goal-directed behavior, especially in novel or distracting circumstances (Best & Miller, 2010). Executive functions are considered “higher-order” as they leverage multiple, more automatic cognitive and perceptual processes (Miyake & Friedman, 2012). Executive dysfunction is a hallmark feature of FASD (Kingdon et al., 2016; Rasmussen, 2005). Children with FASD perform less well than children without FASD on tasks that tap into executive functioning, and are consistently rated as exhibiting more behaviors indicative of executive dysfunction by their parents and carers relative to the general population using the Behavior Rating Inventory of Executive Function (BRIEF; Cheung et al., 2021; Rai et al., 2017; Rasmussen et al., 2007). Abnormal activation of prefrontal brain areas (Kable & Coles, 2017; Kable et al., 2020) and altered prefrontal network connectivity with parietal and temporal regions (Gómez et al., 2022; O’Hare et al., 2009; Ware et al., 2021) are likely candidates for the neurobiological basis for the executive functioning impairments of children with FASD.

There are multiple theoretical models of executive functioning that aim to capture the components and structure (Jurado & Rosselli, 2007). A prominent model of executive functioning identifies three distinct, yet overlapping abilities (Miyake et al., 2000): inhibitory control (sometimes referred to as “inhibition;” the ability to suppress goal-irrelevant processing or prepotent responses), working memory (or “updating;” the ability to hold information in mind and manipulate it in a goal-directed manner), and shifting (or “task switching;” to direct attention between different tasks with different stimuli, goals, or rules). Other conceptualizations and models of executive functioning have been proposed (Doebel, 2020; Zink et al., 2021), including one that suggests that these (e.g., inhibitory control), along with other basic executive functions (e.g., self-monitoring), act as components that are integrated to serve more global abilities (e.g., the ability to regulate behavior; Behavior Regulation) (Gioia et al., 2002). In addition, executive functions have been taxonomized based on whether they are deployed in more emotionally or motivationally salient contexts, such as those involved in controlling emotional responses (“hot” functions), compared to those that involve regulating behavior in less salient conditions (“cold” functions) (Zelazo & Carlson, 2012).

Poorer sleep has been associated prospectively with poorer executive functioning in typically developing children (Beaugrand et al., 2023; Bernier et al., 2010, 2013, 2021; Friedman et al., 2009; Philbrook et al., 2022; Sun et al., 2022; Taveras et al., 2017). Poorer sleep is associated with reduced gray matter volumes, and abnormalities in the activation of the prefrontal cortex (Kocevska et al., 2017; Muzur et al., 2002), which is the cortical area that underpins executive functioning (Yuan & Raz, 2014). Slow oscillation activity that occurs during NREM sleep is thought to drive the consolidation of neural pathways in the brain (Tononi & Cirelli, 2014), and disruption to this process across childhood may contribute to altered prefrontal network formation that underlies executive dysfunction (Mason et al., 2021). In addition, deficits in REM sleep can lead to problems with emotional reactivity and “hot” executive functioning, likely owing to the effect of REM neural activity on resetting limbic networks (Walker, 2009).

Although the association between sleep problems and executive functioning has been studied extensively in neurotypical children (Bernier et al.,2013, 2021; Nelson et al., 2021; Sun et al., 2022), the contribution of sleep problems to executive dysfunction in children with FASD is less understood. In a study of the relationship between sleep and cognitive outcomes that included direct tasks that tapped into executive functioning in 25 children with FASD, there was an unexpected positive relationship between later bed times and superior performance on a working memory task (digit span backwards), while sleep outcomes were unrelated to a measure of inhibitory control (Mughal, Hill, et al., 2020). However, the relatively small sample size, and limited assessment of executive functioning, may have obscured any detrimental impact of sleep problems in that study.

In contrast to the results found by Mughal, Hill, et al. (2020), a significant relationship between sleep problems and poorer executive functioning was recently reported by Gerstner et al. (2023) using the BRIEF caregiver rating form (Gioia et al., 2002). The relationship between executive functioning and caregiver-rated sleep problems was examined using the Sleep Disturbance Scale for Children (SDSC; Bruni et al., 1996) in a sample of 53 children aged between three and 17 years with FASD. Children who were classified as having sleep disturbance as rated on the SDSC were rated higher on the BRIEF Behavior Regulation Index, indicating greater difficulties with shifting from one task to another and modulating negative emotions via inhibitory control compared to children without sleep disturbance. However, there was no significant effect on the Metacognition Index, which reflects difficulties with planning, initiating, organizing, monitoring, and working memory.

A lack of convergence in findings between studies that have used direct and caregiver report measures of executive functioning reflects the generally weak correlation detected between these two assessment methods (Cheung et al., 2021) and may reflect measurement of different aspects of executive functioning (Wallisch et al., 2018). Sleep problems might also impact the deployment of executive functioning in more motivationally salient (or “hot”) conditions, as reflected in the BRIEF Behavior Regulation Index (Gerstner et al., 2023), relative to executive functioning deployed in less motivationally salient (or “cold”) contexts as indexed by the BRIEF Metacognitive Index (Gerstner et al., 2023) and the laboratory inhibitory control and working memory tasks (Mughal, Hill, et al., 2020). Ratings of how executive functions are deployed in a child’s potentially distracting everyday environment, afforded by behavior rating scales such as the BRIEF (Gioia et al., 2003), may offer more ecologically valid assessments of executive control compared to information gathered from tasks administered in typically controlled clinical environments. Further, there is poor correspondence between carer and teacher reports of executive functioning (Muñoz & Filippetti, 2021; Schneider et al., 2020), suggesting that the deployment of developing executive functions by children differs across clinic, home, and preschool and school contexts in typically developing children. The relationship between sleep disturbance and teacher-reported executive functioning has not been investigated empirically in children with FASD thus far.

An examination of the relationship between sleep problems and executive functioning across early childhood may clarify the developmental impact of poor sleep in young children with FASD. The development of executive functioning over childhood is protracted, reflected by the extended maturation of the prefrontal cortex (Fiske & Holmboe, 2019). During the toddler and preschool years, executive functions such as inhibitory control continue to develop, a process that begins from infancy (Posner & Rothbart, 2000), while more complex abilities such as working memory and task shifting differentiate themselves from inhibitory control from four to six years (Best & Miller, 2010). These basic executive functions then consolidate across middle childhood and adolescence and become more distinct from each other (Lee et al., 2013). Other models of executive functioning, such as the one built into the BRIEF assessments, posit that a greater number of executive functions are activated in early childhood from five years (e.g., Organization of Materials, Task-Monitoring, Self-Monitoring) (Gioia et al., 2015) relative to the preschool years (Gioia et al., 2003). Sleep disturbance might therefore differentially affect executive functions depending on when sleep disturbance occur across the developmental trajectory (Sun et al., 2022). Overall, investigation of the context and timing of the effects of sleep disturbance on executive functioning in children with PAE and FASD has been limited.

The current study aims to investigate whether sleep disturbance is associated with executive functioning in children aged 3 to 10 years with confirmed PAE who have received a diagnosis of FASD or a designation of “At Risk” of FASD (Bower & Elliott, 2016). Clinical data on caregiver rated sleep disturbance and executive functioning were extracted from FASD diagnostic clinic databases and used to examine the relationship between sleep and executive functioning. Items from a common clinical questionnaire completed by carers were used to categorize children as having a frequent insomnia symptom and nightmares. As per the extensive literature demonstrating clear associations between insomnia and poor executive functioning, we hypothesized that insomnia symptoms would be associated with broad executive functioning difficulties across domains (Bernier et al., 2013; Bruni et al., 2020; Nelson et al., 2021; Sun et al., 2022; Tesfaye et al., 2021). Given the relatively limited preexisting literature on the effects of nightmares, as opposed to parasomnias more broadly, examination of the effect of nightmares on executive functioning was exploratory. Examination of the influences of developmental range (preschool- and school-age children) and rating context (home and preschool/school) were also exploratory.

Materials and method

Procedure and sample

Specialist services providing diagnostic assessment of FASD in young children (aged 3–11 years) have been established across Australia. The current data are drawn from two, government-funded FASD diagnostic clinics in South-East Queensland and in metropolitan Victoria from 2018 to 2022. All children in this sample had confirmed prenatal alcohol exposure and had undergone diagnostic assessment for FASD at one of these clinics. For the purposes of this study, children were grouped according to age into a preschool group (age 3 years, 1 month − 5 years, 10 months) and a school-age group (5 years, 11 months − 10 years, 11 months). Institutional and hospital ethical review processes included approval to share and use de-identified data collected at the clinics for research projects. Written consent was provided by caregivers and guardians at each site as per Human Research Ethical Committee requirements.

Measures

Demographic and clinical characteristics

Demographic characteristics reported in this study include age, sex (assigned male or female), involvement in child protective services, and relationship to the child of the primary carer (biological parent, kinship carer, foster carer, adopted legal guardian). Clinical characteristics reported include FASD diagnostic status: FASD with three Sentinel Facial Features (SFF; FASD w/3 SFF), FASD with less than three SFF (FASD <3 SFF), or designated as “At Risk” of FASD. Children were assessed by registered psychologists, along with assistance from pediatricians and other allied health professionals (e.g., occupational therapists and speech pathologists), across ten neurodevelopmental domains using clinical interview, direct neuropsychological assessment, and questionnaires. As per the Australian Guide (Bower & Elliott, 2016), children received a FASD diagnosis if they had confirmed prenatal alcohol exposure and were assessed to have at least three neurodevelopmental domains of impairment classified as severe (2 or more standard deviations below the mean or less than the 3^rd percentile). Children received an “At Risk” diagnosis if they had confirmed prenatal alcohol exposure and exhibited neurodevelopment impairments below the severity threshold for a full diagnosis of FASD (mild or moderate impairments) (see Dawe et al., 2023 for more information on diagnostic processes). Co-occurring diagnoses, including ADHD, were assessed using information collected from clinical interviews and a semi-structured interview (the Kiddie Schedule for Affective Disorders and Schizophrenia; KSADS) (Kaufman et al., 1997) based on DSM-5 criteria (American Psychiatric Association Ed., 2013). Information on sleep medication use obtained through information gathered from the clinical interview with the caregiver and/or from collateral clinical records provided to the assessors. Sleep medication use was defined as children taking medications in the evenings in order to manage the circadian or arousal system to induce sleep, or otherwise explicitly stated in the child’s file to be used for sleep (based on criteria used by Efron et al., 2014). Data on stimulant medication use was also extracted from clinic databases.

Sleep disturbance and daytime tiredness

Ratings on items from the Child Behavior Checklist (CBCL) 1.5–5 years and 6–18 years forms (Achenbach & Rescorla, 2001) were used to construct indicators of insomnia symptoms and nightmares. The 100-item CBCL 1.5–5 years (preschool group) and 113-item CBCL 6–18 years (school-age group) forms are well-established measures of behavioral and emotional problems that are commonly used in pediatric assessment and clinical settings. Both forms have items relating to developmentally relevant sleep problems. For each item, caregivers are asked to rate the frequency of a given behavior problem for their child in the last 6 months on a three-point scale: “Not True” (0), “Somewhat or Sometimes True” (1), and “Very True or Often True” (2).

Preschool children were classified as having a frequent insomnia symptom if their carer rated the occurrence of one of these items as “Very True or Often True:” “has trouble getting to sleep,” “wakes often during the night,” and “sleeps less than most kids during day and/or night.” One frequent insomnia symptom was selected as a cutoff as this is sufficient to meet insomnia disorder criteria (Maski & Owens, 2016). These items were selected as indicators of insomnia because they describe symptoms of insomnia (difficulty with sleep onset, maintenance, and/or duration) and because ratings on these items have been shown to correlate significantly with actigraphy measures of sleep duration and efficiency (Bélanger et al., 2014). Preschool children were classified as having nightmares if they were rated as “Somewhat or Sometimes True” or “Very True or Often True” on the “nightmares” item.

School-age children were classified as having a frequent insomnia symptom if their carer rated the occurrence of these items as “Very True or Often True:” “trouble sleeping” and “sleeps less than most kids.” These items also map onto symptoms of insomnia and correlate significantly with other validated scales and objective indicators of sleep duration and efficiency (Gregory et al., 2011). Both items correlate significantly with multiple validated scales of insomnia symptoms and the “trouble sleeping” item is significantly associated with having a psychophysiological insomnia diagnosis (Becker et al., 2015). Given the distribution of nightmares in the school-age group, children in this age range were classified as having nightmares “sometimes” if they were rated as “Somewhat or Sometimes True” and frequent if they were rated as “Very True or Often True” on the “nightmares” item.

Daytime tiredness was also measured for a subset of children using an item from teacher-report on the CBCL1–1.5 Caregiver-Teacher report form and the Teacher Report Form (TRF) for school-age children (Achenbach & Rescorla, 2001). Preschool-age children (n = 22) were classified as exhibiting daytime tiredness if their teacher rated “Overtired” as “Sometimes true” or “Often true.” School-age children (n = 61) were classified as exhibiting daytime tiredness if their teacher rated them displaying “Overtired without good reason” as “Sometimes true” or “Often true.”

Executive functioning

Preschool children were assessed using the BRIEF-Preschool Version (BRIEF-P) parent and teacher versions (based on the day-care or preschool context) (Gioia et al., 2003). The BRIEF-P contains five scales: Inhibit, Emotional Control (EC), Shift, Working Memory (WM), Plan/Organize (PO). These scales are collapsed into three indexes: the Inhibitory Self-Control Index (ISCI; consists of Inhibit and Shift scales), the Flexibility Index (FI; Shift and EC), and the Emergent Metacognition Index (EMI; WM and PO).

School-age children were assessed using the BRIEF-Second Edition (BRIEF-2) parent and teacher versions (Gioia et al., 2015). The BRIEF-2 contains nine scales: Inhibit, Shift, EC, WM, PO, Self-Monitor, Initiate, Task-Monitor, and Organization of Materials. These scales collapse into three indexes: the Behavior Regulation Index (BRI; Inhibit and Self-Monitor), the Cognitive Regulation Index (CRI; Initiate, WM, PO, Task-Monitor, and Organization of Materials), and the Emotion Regulation Index (ERI; Shift and EC). Both the BRIEF-P and the BRIEF-2 contain a global measure of executive functioning, the Global Executive Composite (GEC) score.

Both measures have demonstrated acceptable to good internal consistency, moderate test-retest reliability, and convergent and discriminant validity (Gioia et al., 2003). The BRIEF-P and BRIEF-2 are regularly used in the assessment of FASD in children (Bower & Elliott, 2016; Rasmussen et al., 2007) and in studies examining the association between sleep and executive functioning in other pediatric populations (e.g., Esbensen & Hoffman, 2018; Joyce et al., 2020).

Analyses

Descriptive statistics (means and standard deviations of continuous variables and counts of binary or categorical variables) for variables were presented for the total sample and by preschool and school-age groups. Chi-square tests were used to examine the distribution of children across covariates and insomnia and nightmares groups. Welch two-sample t-tests was used to examine differences in BRIEF indexes by the presence of a frequent insomnia symptom and nightmares. Analysis of Variance (ANOVA) was used as an omnibus test to detect group differences by nightmare frequency for the school-age children, followed-up with Welch t-tests to examine differences between groups. Clinical indexes on the BRIEF-P and BRIEF-2 parent and teacher forms that demonstrated statistically significant differences by insomnia and nightmare status were selected to be modeled using bootstrapped regressions.

Bootstrapped linear regression (2000 replications) was used to estimate the effect of having a frequent insomnia symptom and nightmares on BRIEF clinical indexes in separate models. Each model controlled for potential confounding factors that were found to covary with having either an insomnia symptom or nightmares (see Section 3.1). The potentially confounding variables examined were: child age, child sex (binary: reference was female), placement in out of home care (binary: reference was in care of biological parents), stimulant use (binary: reference was no stimulant use), and sleep medicine use (binary: reference was no use of sleep medication). The 95% confidence intervals for the sleep disturbance model coefficient and the R² for each model were reported (Field et al., 2012). If the 95% confidence interval of a model coefficient did not cross zero, it was considered statistically significant (Field et al., 2012). Bootstrapping was used to estimate the effects because it produces estimates that are robust to violations of parametric assumptions (Field et al., 2012) and has been shown to produce more reliable estimates of regression coefficients when estimating parameters based on relatively small samples (Freedman, 1981).

Results

One hundred and sixteen cases were extracted from FASD diagnostic clinics across South-East Queensland (n = 79) and metropolitan Victoria (n = 37). Forty children were in the preschool group (aged 3.08 to 5.92 years) and 76 were in the school-age group (aged 5.92 to 10.93 years). The characteristics of the entire clinical sample and age-based subgroups are presented in Table 1. There were 37 (31.9%) children with PAE using sleep medication. Of these, 19 (16.3%) were prescribed melatonin, 11 (9.5%) were prescribed clonidine, 2 (1.7%) were prescribed risperidone, 1 was using antihistamine (0.9%) and one was using St John’s Wort (0.9%). There were 3 children (2.6%) who were prescribed a combination of sleep medications: one was prescribed both melatonin and clonidine; one was prescribed both melatonin and risperidone; and the third was prescribed melatonin, clonidine, and amitriptyline.

Table 1. Descriptive statistics for total sample and by age group.

		Age group
Characteristic	Total (N = 116)	Preschool group (n = 40)	School-age group (n = 76)
Sex (male) (n (%))	60 (51.7%)	21 (52.5%)	39 (51.3%)
Age (years) (mean (SD))	6.5 (1.7)	4.7 (0.8)	7.4 (1.3)
Care type (n (%))
Biological parent	32 (27.6%)	17 (42.5%)	15 (19.7%)
Kinship care	33 (28.5%)	4 (10.0%)	29 (38.2%)
Foster care	39 (33.6%)	16 (40.0%)	23 (30.3%)
Adopted parent	12 (10.3%)	3 (7.5%)	9 (11.8%)
FASD Diagnosis (n (%))
At risk of FASD	32 (27.6%)	7 (17.5%)	25 (32.9%)
FASD with < 3 sentinel facial features	64 (55.2%)	25 (62.5%)	39 (51.3%)
FASD with 3 sentinel facial features	20 (17.2%)	8 (20.0%)	12 (15.8%)
Comorbid ADHD diagnosis (n (%))	74 (63.8%)	21 (52.5%)	53 (68.7%)
Stimulant medication use (n (%))	37 (31.9%)	2 (5.0%)	35 (46.1%)
Sleep medication use (n (%))	37 (31.9%)	10 (25.0%)	27 (35.5%)

Carer-reported sleep problems

The pattern of ratings for each sleep problem item are presented in Table 2. There were 18 preschool-age children (45.0%) and 31 school-age children (40.79%) with a frequent insomnia symptom. Preschool children with a frequent insomnia symptom were more likely to be taking sleep medication (44.4% vs 9.1%; X² (1, N = 40) = 4.85, p = .028), which was used as a covariate in regression analyzes. School-age children with a frequent insomnia symptom were more likely to have a diagnosis of ADHD (83.9% vs 60.0%; X² (1, N = 76) = 3.89, p = .049), and to be taking stimulant medication (71.0% vs 28.9%; X² (1, N = 76) = 11.44, p = .001), taking sleep medication (58.1% vs 20.0%; X² (1, N = 76) = 10.01, p = .002), and residing in out of home care (93.5% vs 71.1%; X² (1, N = 76) = 4.50, p = .034), which were all in turn used as covariates in regression analyzes of BRIEF-2 outcomes for school-age children. There were 15 (37.5%) preschool children with occasional or frequent nightmares. They did not exhibit any significant differences by covariates compared to preschool children without nightmares. There were 46 school-age children with occasional nightmares (60.53%) and 10 (13.16%) with frequent nightmares. Similarly, they did not exhibit any significant differences by covariates compared to school-age children without nightmares.

Table 2. Ratings on CBCL items related to sleep problems.

Sleep problem	Not true n (%)	Sometimes true n (%)	Often true n (%)
Preschool group - CBCL 1.5–5 years (n = 40)
Doesn’t want to sleep alone	16 (40.0%)	10 (25.0%)	14 (35.0%)
Has trouble getting to sleep	16 (40.0%)	13 (32.5%)	11 (27.5%)
Nightmares	25 (62.5%)	11 (27.5%)	4 (10.0%)
Resists going to bed at night	12 (30.0%)	16 (40.0%)	12 (30.0%)
Sleeps less than most kids during day and/or night	23 (57.5%)	6 (15.0%)	11 (27.5%)
Talks or cries out in sleep	22 (55.0%)	9 (22.5%)	9 (22.5%)
Wakes often during the night	20 (50.0%)	11 (27.5%)	9 (22.5%)
Overtired	10 (25.0%)	19 (47.5%)	11 (27.5%)
School-age group - CBCL 6–18 years (n = 76)
Nightmares	20 (26.3%)	46 (60.5%)	10 (13.2%)
Overtired without good reason	50 (65.8%)	17 (22.4%)	9 (11.8%)
Sleeps less than most kids	47 (61.8%)	13 (17.1%)	16 (21.1%)
Sleeps more than most kids during day and/or night	66 (86.8%)	5 (6.6%)	5 (6.6%)
Talks or walks in sleep	45 (59.2%)	23 (30.3%)	8 (10.5%)
Trouble sleeping	34 (44.7%)	16 (21.1%)	26 (34.2%)
Wets the bed	44 (57.9%)	14 (18.4%)	18 (23.7%)

Preschool-age children with a frequent insomnia symptom were significantly more likely to be rated as displaying daytime tiredness by their teacher than those without a frequent insomnia symptom (66.66% vs 15.38%; X² (1, N = 22) = 4.03, p = .045). Those with and without nightmares exhibited no significant differences in teacher-reported tiredness (66.66% vs 25.00%; X² (1, N = 22) = 1.72, p = .190). School-age children with a frequent insomnia symptom showed no significant differences in teacher-reported daytime tiredness than those without a frequent symptom (40.74% vs 44.12%; X² (1, N = 61) < 0.01, p = .997). Similarly, there were no significant differences between the rates of tiredness between those never displaying nightmares (35.29%) and those displaying them sometimes (48.57%), or frequently (33.33%; X² (1, N = 61) = 1.20, p = .550).

Associations between caregiver-reported sleep disturbance and executive functioning outcomes

In the preschool-age group (see Table 3), those with a frequent insomnia symptom had significantly greater FI (t(35.90) = 2.67, p = .011), ISCI (t(36.57) = 2.06, p = .047), and GEC scores (t(37.07) = 2.28, p = .029) than those without a frequent insomnia symptom according to their carers. Those with nightmares had significantly greater EMI scores according to their teachers than those with nightmares endorsed as never or sometimes occurring (t(31.80) = 2.27, p = .030).

Table 3. BRIEF-P carer and teacher index scores by the presence of a frequent insomnia symptom and nightmares.

		Frequent insomnia symptom			Nightmares
	Total sample	No (n = 22)	Yes (n = 18)	p	No (n = 25)	Yes (n = 15)	p
Carer form (N = 40)
Flexibility Index (FI)	68.28 (15.46)	62.77 (14.13)	75 (14.64)	.011	65.52 (15.49)	72.87 (14.75)	.148
Emergent Metacognition Index (EMI)	78.12 (15.26)	74.5 (15.38)	82.56 (14.30)	.097	77.72 (15.00)	78.8 (16.2)	.832
Inhibitory Self-Control Index (ISCI)	74.30 (15.53)	69.91 (14.99)	79.67 (14.85)	.047	73.68 (15.47)	75.33 (16.12)	.749
General Executive Composite (GEC)	78.17 (16.50)	73.09 (16.01)	79.67 (15.29)	.029	76.92 (16.13)	80.27 (17.46)	.542
		Frequent insomnia symptom			Nightmares
	Total sample	No (n = 16)	Yes (n = 18)	p	No (n = 20)	Yes (n = 14)	p
Teacher form (N = 34)
Flexibility Index (FI)	69.03 (13.85)	65.62 (15.31)	79.67 (12.04)	.18	66.1 (13.6)	73.21 (13.59)	.143
Emergent Metacognition Index (EMI)	79.53 (14.56)	77.5 (17.90)	79.67 (11.03)	.452	75.35 (15.91)	85.5 (10.15)	.044
Inhibitory Self-Control Index (ISCI)	72.00 (15.83)	68.94 (17.53)	79.67 (14.09)	.294	69.45 (16.79)	75.64 (14.13)	.268
General Executive Composite (GEC)	78.68 (15.66)	75.88 (18.93)	79.67 (12.08)	.333	74.55 (17.02)	84.57 (11.65)	.065

In the school-age group (see Table 4), those with a frequent insomnia symptom had significantly greater BRI (t(72.05) = 4.25, p < .001), CRI (t(71.14) = 3.53, p < .001), ERI (t(67.86) = 4.16, p < .001), and GEC (t(73.61) = 4.77, p < .001) index scores than those without a frequent insomnia symptom. There were significant differences in CRI scores among those with no, some, and frequent nightmares according to their carers. School-age children with frequent nightmares had significantly higher CRI scores than those with nightmares endorsed as never (t(27.83) = 3.03, p = .005) or sometimes (t(24.55) = 3.69, p = .001) occurring.

Table 4. BRIEF-2 carer and teacher index scores by the presence of a frequent insomnia symptom and nightmares.

		Frequent insomnia symptom			Nightmares
	Total sample	No (n = 45)	Yes (n = 31)	p	No (n = 20)	Sometimes (n = 46)	Frequent (n = 10)	p
Carer form (N = 76)
Behavior Regulation Index (BRI)	75.16 (9.63)	71.76 (9.45)	80.10 (7.63)	<.001	73.35 (7.88)	74.96 (10.09)	79.70 (10.06)	.231
Cognitive Regulation Index (CRI)	69.83 (8.79)	67.16 (8.78)	73.71 (7.34)	.001	68.05 (10.18)	69.22 (8.38)	76.20 (4.54)	.041
Emotion Regulation Index (ERI)	75.39 (10.08)	71.82 (9.45)	80.58 (8.72)	<.001	73.40 (8.32)	75.70 (10.63)	78.00 (10.90)	.480
General Executive Composite (GEC)	76.29 (9.69)	72.58 (9.56)	81.68 (7.05)	<.001	75.20 (9.53)	75.76 (10.27)	80.90 (5.93)	.268
		Frequent insomnia symptom			Nightmares
	Total sample	No (n = 41–42)	Yes (N = 31)	p	No (n = 19)	Sometimes (n = 43–44)	Frequent (n = 10)	p
Teacher form
Behavior Regulation Index (BRI) (N = 73)	69.15 (12.94)	68.50 (11.75)	70.03 (14.55)	.620	68.11 (10.97)	70.25 (13.16)	66.30 (15.92)	.635
Cognitive Regulation Index (CRI) (N = 72)	67.90 (10.37)	68.10 (9.24)	67.65 (11.86)	.856	66.37 (10.57)	68.14 (10.32)	69.80 (10.90)	.686
Emotion Regulation Index (ERI) (N = 73)	72.52 (13.59)	71.48 (12.86)	73.94 (14.61)	.448	72.26 (11.58)	73.30 (13.99)	69.60 (16.13)	.742
General Executive Composite (GEC) (N = 72)	71.81 (11.63)	71.56 (9.98)	72.13 (13.68)	.839	70.68 (10.48)	72.49 (11.93)	71.00 (13.33)	.834

Indexes that demonstrated significant differences by the presence of a frequent insomnia symptom or nightmares were selected for regression analysis (see Table 5). For BRIEF indices that differed significantly by whether children had an insomnia symptom or nightmares, bootstrapped regressions were conducted to examine whether differences held after controlling for potential confounding factors that were found to covary with having an insomnia symptoms or nightmares (see Section 3.1). For most outcomes, dummy coded insomnia and nightmare variables were not entered as predictors into the same models given that they impacted different BRIEF indexes (see Tables 3 and 4). However, there was one index (BRIEF-2 CRI) where there was a significant difference in scores by having an insomnia symptom and by having nightmares. For this model, both sleep problems were entered together to determine which sleep problem had an impact on BRIEF-2 CRI scores.

Table 5. 95% CI of bootstrapped regression coefficients for the effects insomnia and nightmares on selected BRIEF clinical index scores.

	BRIEF-P Carer			BRIEF-P Teacher	BRIEF-2 Carer
	ISCI	FI	GEC	EMI	BRI	ERI	CRI	GEC
Preschool age group
Frequent insomnia symptom^b	−6.863–14.694	−3.302–17.331	−5.900–16.929	–	–	–	–	–
Nightmares ^a	–	–	–	1.69–18.99	–	–	–	–
School age group
Frequent insomnia symptom ^c	–	–	–	–	1.399–9.792	−1.574–8.664	−1.756–7.368	2.302–11.578
Nightmares ^c
Sometimes	–	–	–	–	–	–	−4.8508–5.0510	–
Frequent	–	–	–	–	–	–	1.605–12.395	–

^aModel unadjusted.

^bModels adjusted sleep medication use.

^cModels adjusted for ADHD comorbidity, stimulant and sleep medication use, and out of home care status.

For the preschool group, controlling for sleep medication use (which covaried with having a frequent insomnia symptom) abolished its effect on carer-reported BRIEF-P ISCI, FI, and GEC scores (coefficient estimates spanned zero). There were no potential confounds of having nightmares detected, so the model of the effect of nightmares on BRIEF-P CRI scores was unadjusted. Having nightmares was a significant predictor of teacher-reported BRIEF-P CRI scores such that they increase them between 1.69 and 18.99 (95% CI) relative to those who did not have nightmares. For the school-age group, having a frequent insomnia symptom significantly increased BRIEF-2 BRI scores by 1.399–9.792 and significantly increased GEC scores by 2.302–11.578. Controlling for ADHD comorbidity, stimulant and sleep medication use, and placement in out of home care (which covaried with having a frequent insomnia symptom) abolished the effect of having a frequent insomnia symptom on carer-reported BRIEF-2 EMI scores. For the model of BRIEF-2 CRI scores, having frequent nightmares significantly increased scores by 1.605–12.395 relative to children without nightmares, while having nightmares occasionally did not significantly increase scores. The effect of having a frequent insomnia symptom in this model was also non-significant.

Discussion

The aim of this study was to examine the association between sleep disturbance (both insomnia symptoms and nightmares) and executive functioning in preschool and school age children with prenatal alcohol exposure (PAE), who had either a diagnosis of FASD or a designation of “At Risk” of FASD. Preschool-age children with FASD or “At Risk” of FASD and who also had occasional or frequent nightmares, were rated by their preschool and/or day care teachers as having significantly greater difficulties with executive functions related to sustaining information in working memory and planning and organizing activities (the Emergent Metacognition Index) compared to children without nightmares. This finding is in line with research examining carer-rated parasomnias, which encompass nightmares and other abnormal sleep-wake behaviors (e.g., sleep walking and talking), and that have been associated with poorer carer-rated executive functioning in preschool children from the general population (Bruni et al., 2020). Indeed, carer-rated nightmares are associated with alterations to objective markers of sleep, including reduced sleep efficiency because of increases in arousal (Kovachy et al., 2013; Kushnir & Sadeh, 2011; Rolling et al., 2023), and lower sleep quality might in turn impact prefrontal circuitry that supports executive functioning (Jones & Harrison, 2001; Ma et al., 2015; Turnbull et al., 2013). Of note, the presence of nightmares in children with FASD is common (Mughal et al., 2021), and these have been recently linked with exposure to early childhood adversity, in particular events that are likely perceived as threatening by the child (Chandler-Mather et al., 2023). A proxy measure of exposure to trauma and adversity was used in the current study and placement in out of home care/engagement in the child protection system was included as a confound in the present analyzes.

The results also demonstrated that preschool-age children with a frequent insomnia symptom exhibited significantly greater difficulties with FI, ISCI, and GEC scores, and were more likely to display daytime tiredness according to their teachers, than those without a frequent insomnia symptom. However, these effects were negated when sleep medication use was controlled for. It is plausible that a relatively small sample size in the preschool age group contributed to wide confidence intervals for the effects of having a frequent insomnia symptom after controlling for sleep medication use. Overall, there was a general pattern suggesting preschool-age children with or “At Risk” of FASD with a frequent insomnia symptom were more tired during the day according to their teachers and had greater difficulties with executive functioning generally, although the limited sample size for preschool children with PAE and FASD produced unreliable effect sizes (wide confidence intervals) and meant that testing these associations in a mediation model was not possible.

The effects of having a frequent insomnia symptom on carer-reported executive functioning difficulties were more pronounced for school-age children. Those with a frequent insomnia symptom were rated as having greater difficulties with all executive functioning domains. Most of these associations held after controlling for factors that covaried with having a frequent insomnia symptom: ADHD comorbidity, stimulant and sleep medication use, and out of home care status. The association between having a frequent insomnia symptom and ERI scores, suggesting that control of emotional responses or “hot” executive functioning may be better explained by confounding factors in children with or “At risk” of FASD.

The broad negative impact of sleep disturbance on executive functioning in school-age children is consistent with findings from prior studies reporting that sleep problems predict poorer executive functioning in childhood when using both direct measures of executive functioning (Bernier et al., 2010, 2013, 2021; Friedman et al., 2009), and self/caregiver reports (e.g., BRIEF; Sun et al., 2022; Taveras et al., 2017). Similarly, cross-sectional studies have detected associations between caregiver reported sleep disturbance and poorer executive functioning in school-age children (Didden et al., 2002; McCann et al., 2018; Sciberras et al., 2015). Insomnia symptoms have also been associated with reductions in prefrontal cortical volumes in children (Cheng et al., 2021; Kocevska et al., 2017) and adults (Joo et al., 2013; Li et al., 2018), which underpin executive functioning. Relevant to the current study, children with greater insomnia symptoms assessed using the CBCL 1.5–5 years longitudinally from infancy to 6 years, the assessment instrument in the current study, exhibited a thinner dorsolateral prefrontal cortex at 7 years, an area that underpins executive control (Kocevska et al., 2017). Thus, these findings contribute to mounting converging evidence that insomnia symptoms are associated with difficulties with engaging executive functioning and with altered underlying neural circuitry.

School-age children with a frequent insomnia were not rated as exhibiting occasional or frequent tiredness by their teachers, suggesting that fatigue might not mediate the effects of having caregiver-reported insomnia symptoms and nightmares on executive control in older children with PAE and FASD. Previous studies examining the mediating effect of daytime tiredness on the relationship between sleep and executive control found that it did not mediate the effect of poorer sleep quality on poorer working memory performance (McCann et al., 2018), indicating other potential mediating pathways. Disruption of homeostatic pruning of neural pathways connecting prefrontal networks may be a candidate mechanism (Tononi & Cirelli, 2006). Given that children with heavy PAE are significantly more likely to exhibit chronic sleep problems across childhood (Chandler-Mather et al., 2021), chronic sleep disturbance may lead to disrupted pruning and consolidation in prefrontal areas over time. This longitudinal trajectory of sleep disturbance might also explain more pronounced impacts of insomnia in the school-age compared to preschool-age children. Future longitudinal studies of sleep and brain function and structure in children with FASD would bear this question out.

School-age children with frequent nightmares were found to have greater carer-reported difficulties on the CRI, which taps into holding information in working memory and planning, initiating, and monitoring tasks, after controlling for having a comorbid insomnia symptom and factors listed above that covaried with having an insomnia symptom. Those who had them occasionally did not have significantly greater difficulties with CRI. Having nightmares impacted similar domains of executive functioning in preschool-children, those that tap working memory and planning abilities. Occasional nightmares are considered normative in school-age children (El Sabbagh et al., 2023), whereas frequent nightmares may lead to consistently fragmented sleep and arousal (Kovachy et al., 2013) which may have marked detrimental effects on these particular domains of executive functioning.

Overall, apart from the effect of nightmares in preschool-age children, significant differences between those with and without insomnia symptoms or nightmares did not emerge in the school context as reported by teachers. Notably, there is poor agreement between carer and teacher BRIEF scores in typically developing children (Muñoz & Filippetti, 2021; Schneider et al., 2020). The ongoing debate around the validity of executive functioning assessment methods that rely on collateral report compared to direct performance on cognitive tasks, with collateral reports (e.g., carers and teachers) may be of relevance here. The former are considered to offer greater ecological validity relative to direct performance, which offers comparatively greater construct validity (Doebel, 2020; Wallisch et al., 2018). Ultimately, these methods may indicate where and how the child is struggling with deploying their executive functioning. Preschool children with or “At risk of” FASD in this sample who have nightmares are having greater difficulties at day-care/preschool with working memory and planning, whereas most school-age children exhibited greater problems engaging executive functioning in general at home, likely before and after school.

This issue of assessment method might explain divergences in findings between the current and previous studies examining sleep and executive functioning in children with FASD. The negative associations between sleep disturbance and global executive functioning in school-age children with PAE are consistent with findings from a previous study in school-age children with FASD (Gerstner et al., 2023). However, the current findings are inconsistent with findings from a prior study of sleep and executive functioning in 29 children with FASD aged 9 years, which detected an unexpected positive association between later bedtimes and better working memory performance, and no association with inhibitory control task (Mughal, Hill, et al., 2020). Both Gerstner et al. (2023) and the current study used the BRIEF scales completed by caregivers based on observations in their everyday environment to measure executive functioning and found significant negative associations between sleep disturbance and executive functioning. By contrast, Mughal, Joyce, et al. (2020) used a select battery of cognitive tasks that included two tasks that tapped executive functioning, digit span backwards (working memory) and commission errors in a continuous response task (inhibitory control). Thus, sleep disturbance in children with FASD may manifest in everyday situations and may be more motivationally salient and distracting, as measured by caregiver-report measures such as the BRIEF, relative to direct tasks administered in structured clinical environments (Doebel, 2020; Wallisch et al., 2018). Future studies incorporating both caregiver-report and direct task-based assessment methods would help to clarify whether the effect of sleep disturbance on executive functioning is moderated by the type of measure used.

The relatively high rates of frequent insomnia symptoms in this sample of preschool (45%) and school-age children (40.79%) with PAE and FASD are similar to other reports of sleep disturbance in this clinical population (Chen et al., 2012; Gerstner et al., 2023; Hayes et al., 2020). Similarly, the high rates of sleep medication use are consistent with previous findings. The rate of frequent insomnia symptoms in children using stimulant medication was markedly high relative to those who were not using stimulants. Some studies report an association between stimulant medication use and sleep disturbance in children with ADHD (Kidwell et al., 2015). Trials of stimulant medication in children with FASD should monitor for side effects of insomnia, especially considering the potential detrimental effects of insomnia symptoms on executive functioning.

Limitations and future directions

Although this study had many strengths, it was not without limitations. First, while studies have shown that sleep items from the CBCL forms correlate with parent diary and actigraphy recordings of sleep (Becker et al., 2015; Bélanger et al., 2014; Gregory et al., 2008), direct measurement of sleep via actigraphy or polysomnography would clarify what physiological features of sleep are associated with impaired executive functioning rather than behavioral markers used in the current study (Gregory et al., 2008). Sleep disordered breathing was not assessed in the current study despite its associations with poorer executive functioning (Joyce et al., 2020) and reports of marked presentations in children with FASD (Chen et al., 2012). Future research should investigate the association between PAE and FASD with sleep disordered breathing and executive functioning.

It must be emphasized that the cross-sectional nature of the study limits the ability to determine the direction of association between sleep problems and executive functioning in the current sample. While prior longitudinal studies support the assertion that sleep has an impact on the later emergence of these areas of functioning (Bernier et al.,2010, 2013, 2021; Friedman et al., 2009; Kahn et al., 2013; Philbrook et al., 2022; Sun et al., 2022; Taveras et al., 2017), it is possible that both clinical sleep problems and difficulties with executive functioning reflect a common underlying impairment with self-regulation in children with FASD (Williams et al., 2017). Future longitudinal research with a FASD sample is needed to bear this question out. Future clinical trials of pharmacological or behavioral sleep interventions aimed at improving sleep in children with FASD could also track changes in executive functioning to determine a more proximal and sequential association between sleep and improved executive functioning.

No measure of trauma was available for the combined study sample. The presence of trauma, FASD, and OOHC are highly inter-related (Chandler-Mather et al., 2023; Flannigan et al., 2021). Given recent findings linking exposure to adverse life events and nightmares in children with FASD (Chandler-Mather et al., 2023), the connection between exposure to exposure to threatening events, nightmares (suggesting difficulties with integrating a traumatic or stressful event), and poorer executive functioning development in young children with FASD presents an area for future research.

Conclusions

Overall, the current study detected high levels of executive functioning difficulties and sleep disturbance in children with FASD. Preschool-age children with PAE and FASD who had nightmares were rated by their teachers as having greater difficulties with working memory and planning and organizing compared to children without nightmares. For school-age children, having a frequent insomnia symptom predicted poorer executive functioning as per carer-report, after controlling for a host of covariates including ADHD comorbidity, sleep and stimulant medication use, and out of home care status. Preschool-age children with a frequent insomnia symptom were rated as exhibiting more daytime tiredness by their preschool/day care teachers relative to those with no or occasional symptoms, whereas school-age children with a frequent insomnia symptom were not. The findings are consistent with the hypothesis that sleep problems compound the deleterious effects of PAE on executive functioning in children with FASD. Given the high prevalence, and association with executive dysfunction, clinicians should investigate and attend to sleep disturbance, both insomnia and nightmares, as part of routine care in children with FASD or potential prenatal alcohol exposure.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by funding for the FASD Consortium led by Prof Dawe from the Drug and Alcohol Program: Fetal Alcohol Spectrum Disorder (FASD) Diagnostic Services and Models of Care Grant Opportunity [H1617G038]; Australian Government Department of Health, Canberra.