Introduction

Sleep in Individuals with AS

Angelman Syndrome (AS) is a rare neurodevelopmental disorder affecting approximately 1:10,000–20,000 individuals worldwide (Spruyt et al., 2018). The disorder is caused by the loss of function of the maternal copy of the UBE3A gene, which maps onto chromosome 15q11-q13 (den Bakker et al., 2018; Williams et al., 2010). AS is characterised by motor dysfunction, specifically impaired limb movement; intellectual disability; speech impairment; and a specific behavioral phenotype, distinguished by frequent laughter and sociability (Bird, 2014; Margolis et al., 2015). Individuals with AS often have comorbidities such as epilepsy, autism spectrum disorder (ASD), and severe sleep problems (Didden et al., 2004; Margolis et al., 2015). In other developmental disorders such as ASD, sleep problems co-occur with a range of other comorbid conditions including attention-deficit/hyperactivity disorder, gastrointestinal symptoms, epilepsy, challenging behavior, and feeding and toileting issues (Leader & Mannion, 2016; Leader et al., 2022; Whelan et al., 2022).

Sleep concerns are a prevalent feature in individuals with AS. Research shows that 20–80% of individuals with AS have a diminished sleep need and/or abnormal sleep–wake cycle (Goldman et al., 2011; Williams et al., 2006). Abnormal electroencephalogram (EEG) patterns have been reported in AS participants using polysomnography recording. However, much is still unknown of the irregularities causing sleep problems in the population (Miano et al., 2005; Spruyt et al., 2018). The specific sleep problems experienced vary among individuals but ranges from sleep-onset delay, frequent night-awakenings, early morning waking, decreased sleep duration, daytime sleepiness and distorted sleep leading to reduced sleep quality (Didden et al., 2004; McLay et al., 2019; Williams et al., 2006). Sleep disorders such as insomnia, parasomnias and sleep breathing disorders have been reported (Braam et al., 2008; Bruni et al., 2004).

Research has been conducted assessing the sleep difficulties in AS compared to aged-matched controls, results demonstrated that individuals with AS displayed decreased REM sleep duration while sleep quality was also reduced in comparison to neurotypical (NT) peers (Miano et al., 2004). Sleep problems have been shown to improve throughout the lifespan in individuals with AS (Larson et al., 2015). It was found that 75% of families/caregivers of individuals with AS reported their sleep problems had diminished at a 10 year follow up (Smith et al., 1996). However, a number of families/caregivers still reported sleep disturbance into adulthood, indicating that sleep problems may persist following childhood or adolescence (Bruni et al., 2004; Pelc et al., 2008b).

Implications of Sleep Problems in AS

The implications of sleep problems and their effects on individuals with neurodevelopmental disorders is under-researched, as is the impact on caregiver/family members. Studies have demonstrated that sleep disturbance issues in children, not only negatively impact on the child but also result in reduced sleep duration and sleep quality in parents (Meltzer & Mindell, 2007). Parents of individuals with AS experiencing sleep problems reported higher levels of caregiver stress and low mood (Miodrag & Peters, 2015). Although not reported in AS, sleep problems such as reduced sleep efficiency were predictive of problematic daytime behavior in the form of tantrum behavior and aggressive behaviors in individuals with developmental disorders (Cohen et al., 2018). If sleep problems are left untreated, they can adversely impact behavior and cognitive ability through their effects on learning and social relations (Altevogt et al., 2006; Kronk et al., 2010).

Methods Used to Assess Sleep Problems in AS

The primary means of assessment for sleep problems in AS are subjective in the form of sleep questionnaire surveys, completed by parents/caregivers of children; as well as parental interviews (Pelc et al., 2008b). However, research has utilised semi-objective measures to gather polysomnographic data such as EEG’s and actigraphy recordings to determine sleep–wake patterns (Anderson et al., 2008; Jain et al., 2012; Miano et al., 2004, 2005).

Treatment/Intervention for Sleep Problems in AS

Egan et al. (2020) conducted a systematic review on interventions to improve sleep in individuals with AS. The findings demonstrated pharmacological interventions as the most commonly prescribed treatment, followed by combined interventions and behavioral treatment. Of the treatments, pharmacological interventions were associated with both positive and mixed results; whereas combined interventions (consisting of both pharmacological and behavioral interventions) demonstrated more positive outcomes, as did behavioral interventions when implemented solely. Previous research identified melatonin as the most frequently prescribed medication for sleep problems in individuals with AS (Blackmer & Feinstein, 2016); whereas, environmental accommodations and sleep/wake schedule modifications comprised behavioral interventions (Allen et al., 2013; Summers et al., 1992).

Current Review

A lack of research warrants the need to investigate the specific sleep problems that exist in AS. It is important to assess the screening tools/methods utilised by clinicians to diagnose these problems prior to intervention. A sleep profile must first be established in order to select an appropriate intervention/treatment. Evaluation of the different assessment tools used within studies provides an insight into the utility and efficacy of each measure. If an effective assessment tool is employed, the specific sleep problem can be identified which may better aid in the selection process for a suitable treatment.

The current systematic review aimed to characterise the sleep problems identified and reported by families/caregivers of individuals with AS, explore how sleep problems were assessed and how sleep problems were treated/managed. This review aimed to update and expand upon the reviews conducted by Spruyt et al. (2018) and Egan et al. (2020), with the inclusion of studies published subsequent to 2016. Additionally, the type of assessment tool selected was analysed and compared between studies to identify their effect on the outcome variables while age was also explored as an independent variable and its impact on sleep problems. The sleep treatments/interventions prescribed were identified.

Method

Eligibility Criteria

This review was conducted according to the preferred reporting items of systematic reviews and meta-analysis (PRISMA) guidelines. The following inclusion and exclusion criteria were used to search, screen, and select the articles. Inclusion criteria for the review were (1) empirical studies involving participants with AS (of all ages), (2) articles that measured sleep problems in the population, (3) articles with a specifically stated objective for their measurement, (4) articles published after the year 2016 as this was an expansion of the work conducted by Spruyt et al. (2018), and (5) those published in the English language. Exclusion criteria then were as follows: (1) papers that did not discuss individuals with AS, (2) articles that did not measure sleep problems, (3) articles published before the year 2016, (4) articles without a specifically stated objective for their measurement, (5) reports, reviews in general, meta analyses or poster presentations, (6) papers describing sleep tools, and (7) articles not written in the English language.

Search Procedures

A search strategy was made up of two keyword areas with the MeSH terms ‘Angelman*’ OR ‘Angelman Syndrome’ AND ‘Sleep*’. Three electronic databases were utilised to conduct the searches: PubMed, Psychology and Behavioral Sciences Collection, and PsycINFO; with key search terms applied to all. Searches were performed using a custom date range for articles published between 01/01/2016 and 01/08/2020. The database search initially resulted in 344 potential publications, however, this was then decreased to 90 when the publication date was determined.

Study Selection

After exporting all publications to a reference management software, the 90 titles and abstracts were screened for eligibility by both raters independently. This resulted in the removal of a further 73 articles based on the exclusion criteria mentioned previously. There was 85.9% agreement from the authors for the inclusion of the articles. Upon second screening of abstracts, it was found that three of the studies were review articles; deeming them ineligible for inclusion. These were subsequently removed, leaving 14 remaining studies and 100% agreement from the authors. The remaining 14 studies were analysed in detail by both raters independently. Finally, seven of the fourteen studies were subsequently chosen by the authors, satisfying inclusion criteria. The other seven studies were removed based on their focus on biology of sleep, circadian ontology, and pharmacological treatments. The full procedure is presented in the flow diagram (see Fig. 1).

Fig. 1
figure 1

Flow diagram depicting the studies identified and selected for review

Full size image

Methodological Quality

The first and second authors independently assessed the quality of the seven studies using an adapted version of the Downs and Black Quality Checklist (Downs & Black, 1998) in line with the studies chosen and their applicable items (see Table 2). This checklist has been shown to evaluate the strengths and weaknesses of health care studies. Four subscales were used in the adapted version which consisted of 12 items. The authors modified the checklist due to the irrelevance of certain assessment items in relation to the evaluation of clinical trials. The subscales consisted of ‘reporting’—7 items (e.g. clarity of hypothesis, findings etc.), ‘external validity’—1 item (representative of entire population), ‘bias’—3 items (e.g. appropriateness of statistical tests, accuracy of measures etc.) and ‘power’ -1 item (sufficient power to detect clinically important effect). Items could be marked either ‘yes’ (1 mark) or ‘no’ (0 marks) with a total score calculated at the end. One item was an exception in that it had two parts i.e. 2 marks were available, depending on whether the statistical power was stated as being at least 80%. A maximum score of 13 could be attained with higher scores indicating higher overall quality. Interrater reliability was calculated for each subscale (see Table 1). Using recommendations by Munn and colleagues (2010), studies were grouped via their scores for ease of comparison between raters. Studies were considered to be of low methodological quality if they scored > 60%, studies scoring 60–74.9% were considered moderate while studies scoring > 75% were considered high methodological quality.

Table 1 Quality ratings from rater 1 and 2 using the adapted Downs and Black Checklist (1998)
Full size table

Data Extraction

Data extraction for all included studies was completed by the second author. Each study was summarised in terms of (a) participant characteristics (number of participants, sample, age range, gender) of both AS participants and other participants (if applicable), (b) authors of study, (c) year of study, (d) study design used, (e) sleep measure used, (f) treatment of sleep problems, (g) results and (h) conclusions drawn from the study. A summary of the extracted data is presented in Tables 2 and 3.

Table 2 Participant demographics for the included studies
Full size table
Table 3 Study design, aims, results and conclusions
Full size table

Reliability and Inter-Rater Agreement

Reliability measures among the authors were collected for different steps of the process. In order to ensure accuracy, results were compared by both authors and provided in tables. Screening and detailed analysis was completed independently by both authors (see Table 1). Quality Assessment was then taken independently (see Table 1) for each of the seven studies with 71.4% agreement on the standard of the studies (low/moderate/high) between the authors.

Results

The search strategy and inclusion/exclusion criteria were systematically synthesised, determining the seven studies included in this review.

Study Sample

Following examination of the seven studies, a total of 266 parents/caregivers of individuals with AS participated. Each of the seven studies specified gender with an even distribution of male (n = 133, 50%) and female (n = 133, 50%) participants (Abel & Tonnsen, 2017; Agar et al., 2020; Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2017; Trickett et al., 2018; Trickett et al., 2019). Of the seven studies, an age range of 8 months-45 years was established, with a mean age of 12.5 years; recorded for six studies. The study by Abel and Tonnsen (2017) did not specify the mean age of participants diagnosed with AS but recorded a range of 8 months to 3 years and 9 months, with a median age of 18 months. Four of the studies (Abel & Tonnsen, 2017; Agar et al., 2020; Trickett et al., 2018, 2019) also included NT controls and individuals with other genetic and developmental disorders. Studies included in this review differed considerably in terms of sample size and age of participants.

Study Methodology

Of the seven studies included for review, six were based on quantitative analysis while one used qualitative analysis (see Table 3). Within the studies, both subjective and semi-objective sleep assessment tools were utilised, 86% of studies implemented subjective measures which comprised parental/caregiver reports/interviews and sleep questionnaires. The sleep questionnaires implemented are presented in Table 4. Two of the studies implemented semi-objective sleep measures consisting of actigraphy and polysomnography/EEG recordings (Sueri et al., 2017; Trickett et al., 2019); while one study used paired video and actigraphy analysis (Agar et al., 2020). Both measures are considered semi-objective as actigraphy recordings rely on parents/caregivers to provide exact bedtime/wake-times for accuracy; while EEG’s are interpreted by neurologists and results may differ by expertise or experience.

Table 4 Sleep measures used and sleep problems assessed
Full size table

Dependent Variables

Outcome variables included (a) sleep problems (i.e. sleep duration, sleep efficiency, sleep-onset, daytime sleepiness and awakenings). For the purpose of this review, the most commonly used terminology across the seven studies was adopted to categorise and define sleep problems. Further aims were to analyse the assessment tools used to determine sleep problems and identify any treatments/interventions prescribed to manage sleep difficulties.

Sleep Problems

When appraising the broad content of the studies, four studies examined whether respondents had experienced any general issue in relation to sleep, 87.2% (n = 143) of parent/caregiver respondents reported a difficulty (Abel & Tonnsen, 2017; Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2017).

Sleep duration was reported to range between 480 and 660 (M = 538) minutes in AS (Abel & Tonnsen, 2017; Agar et al., 2020; Pereira et al., 2020; Trickett et al., 2019). Of the studies included for review, two studies reported sleep duration for NT peers, ranging between 497 and 728 (M = 613) minutes (Abel & Tonnsen, 2017; Trickett et al., 2019). Abel and Tonnsen (2017) reported sleep duration was atypically shorter in the AS group compared to NT as demonstrated by Cohens d (d = 1.22), while Trickett et al., (2019) did not report any significant differences between total sleep duration in the AS and NT (p = 0.66) groups. Increased variability in total sleep time of individuals with AS was reported when compared with peers (Sueri et al., 2017; Trickett et al., 2019).

Trickett et al. (2019) reported a sleep efficiency rate of 84% in the NT group, sleep efficiency was reduced among AS individuals with a rate of 78% reported. This finding was replicated by Agar et al. (2020), with parents/caregivers reporting a sleep efficiency rate of 78% among AS individuals.

Across the three studies that reported the duration of sleep onset latency (Abel & Tonnsen, 2017; Pereira et al., 2020; Trickett et al., 2019), sleep-onset latency ranged between 15–19 min (M = 18) in AS individuals; whereas NT peers ranged between 15–30 min (M = 22.5). Of those reporting frequency, 64.9% (n = 87) reported difficulties with sleep-onset latency (Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2017).

Trickett et al. (2019) reported that 7 children with AS had a nap ranging between 5 and 240 min at least once a day during a 7-day assessment period; while only three NT children had a nap during the day ranging between 20 and 100 min. Daytime somnolence was more frequent among those scoring above clinical cut-off for indicative gastro-oesophageal reflux (GORD) symptoms (Trickett et al., 2018).

Night-awakenings occurring once per night was experienced by 76.1% (n = 134) of those in the AS group; whereas night-awakenings of more than once per night were less frequent, affecting 15.0% (n = 26) of individuals (Abel & Tonnsen 2017; Agar et al., 2020; Pereira et al., 2020; Trickett et al., 2017, 2019). Agar et al. (2020) found an average wake time of 104 min after sleep onset, and Trickett et al. (2019) indicated greater variability in mean wake time. A negative correlation was reported between age and night-waking, suggesting night-waking decreases with age (Trickett et al., 2018).

Sueri et al., 2017, reported that sleep problems subsided in 12.8% (n = 5) of respondents with a sleep disorder in childhood while spike-wave index (SWI) was stable during adolescence and adulthood. SWI quantifies the frequency of spikes in EEG recordings which is considered the percentage of non-REM sleep occupied by spike-waves (Tassinari et al., 2000). Pereira et al. (2020) also reported a decrease in sleep problems from childhood to adulthood.

Sleep Assessment

There was great variability among the studies in relation to the assessment tools selected and the combination of assessment tools used. Across the seven studies, only two authors used the same sleep measure, see Table 4. Trickett et al. (2017, 2018) both used the Modified Simonds and Parraga Sleep Questionnaire (MSPSQ; Wiggs & Stores, 1996) but their outcomes differed. Trickett et al. (2017) provided much more detail in relation to the characterisation of sleep problems, reporting the frequency of settling difficulties, night-awakenings and daytime somnolence; whereas Trickett et al. (2018) provided the effect sizes for night-waking and daytime somnolence only.

Of the seven studies assessing sleep-onset, only three studies reported the exact frequency of difficulties (Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2017) and the duration of sleep-onset latency (Abel & Tonnsen, 2017; Pereira et al., 2020; Trickett et al., 2019). Although, two studies used an actigraphy, the duration of sleep-onset latency was only reported for one (Trickett et al., 2019); while Agar et al. (2020) neglected to report this. Whereas, two studies using only subjective measures reported the duration of sleep-onset latency (Abel & Tonnsen, 2017; Pereira et al., 2020).

Sleep Treatments

Medication to target sleep difficulties was used either currently or historically by 67.0% (n = 177) of all AS respondents across the seven studies. Behavioural interventions were used less frequently than pharmacological interventions and only one of the studies included in this review implemented behavioural intervention (Trickett et al., 2017). Medication tends to be prescribed from a very young age in the AS population, with 17% (n = 3) of infants and toddlers assessed by Abel and Tonnsen (2017) already having been administered sleep-related medications. More broadly, as per informant reports 65.3% (n = 111) of all parents/caregivers of children (including infants and toddlers) and young people with AS reported using medication to aid sleep. This excluded the studies by Sueri et al. and Pereira et al. as they were the only studies that included adult participants. When distinguishing between medications, melatonin was the primary medication administered for sleep problems. Melatonin use ranged from 23 to 71% in each of the studies assessing the type of medication prescribed (Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2017, 2018, 2019). Two studies reported the ‘helpfulness’ of medication in aiding sleep, with 66% and 77% of the sample deeming it effective when used alone/in combination (Trickett et al., 2017, 2019). Abel and Tonnsen (2017) reported that participants taking medication had reported less daytime sleeping than those not taking medication. Small numbers of behavioral interventions were reported by Trickett et al. (2017) which varied from planned-ignoring to interventions targeting an increase in the child’s communicative ability.

Evaluation of Sleep Methodology

The majority of the studies (n = 4, 57%) reported reliability and/or social validity (Agar et al., 2020; Sueri et al., 2017; Trickett et al., 2017, 2019). Agar et al. (2020) observed nineteen behaviours which were identified and coded through caregiver report questionnaires that measured target areas for exploration (i.e. pain, challenging behaviour etc.) with high reliability and validity. For example, ‘aggression’ and ‘destructive behaviour’ were derived from the Challenging Behaviour Qustionnaire (CBQ; Hyman et al., 2002 as cited in Agar et al., 2020). Additionally, behaviours were coded using ObsWin (Martin et al., 2000), a programme that allows users to code operationally defined occurrences of behaviour in real-time (e.g. jerking, rocking etc.), as well as continuous behaviours (e.g. duration of laying down). Agar also assessed inter-rater reliability with an independent second observer coding 22% of clips, reliability was deemed ‘substantial’. Sueri et al., (2017) ensured reliability by independently evaluating EEG recordings. Inter-interviewer agreement was sought in one study using audio recordings, raw positive agreement (i.e. when the same code was applied to data by both the primary author and original interviewer) was 35% and raw negative agreement (i.e. when multiple codes were identified, the primary author and original interviewer determined which codes were irrelevant) was 86% (Trickett et al., 2017), disagreements were mutually resolved.

Sleep Measures

The Modified Epworth Sleepiness Scale (MESS; Johns, 1992) has demonstrated good test–retest reliability and has been validated among individuals with neurodevelopmental differences (Gringras et al., 2012; Williams et al., 2008). The MSPSQ was compared with the Children’s Sleep Habits Questionnaire (CSHQ; Owens et al., 2000) to assess their utility in characterising sleep difficulties in children with neurodevelopmental differences, namely autism (Johnson et al., 2012). Results demonstrated that both measures had adequate internal consistency and were highly correlated. A systematic review assessing a number of subjective sleep measures applied an evidence-based assessment rating of ‘well-established’ to the Brief Infant Sleep Questionnaire (BISQ; Sadeh, 2004) and Pediatric Sleep Questionnaire (PSQ; Chervin et al., 2000) while the Family Inventory of Sleep Habits (FISH; Malow et al., 2009) received a rating of ‘promising’ (Lewandowski et al., 2011). ‘Well-established’ ratings required sound psychometric properties and were required to be used by two/more investigators; while ‘promising’ ratings were given to measures that had moderate/vague psychometric properties.

Discussion

Current Research

This systematic review aimed to extend upon and update the research conducted by Spruyt et al. (2018) and Egan et al. (2020) with the inclusion of studies published post-2016. This review identified and characterised sleep-specific problems in the AS population as reported by family members/caregivers while examining the assessment tools used and treatments/interventions prescribed.

Sleep duration and sleep efficiency were reduced compared with age-matched controls which was in line with that reported by Miano et al. (2004). The findings suggest night-awakenings are a significant sleep related issue as reported by families/caregivers of individuals with AS, with the majority of respondents reporting at least one nightly occurrence, supporting previous literature (Didden et al., 2004; McLay et al., 2019).

Evidence from previous research illustrates that the onset of sleep problems tend to begin in infancy. Results from Trickett et al. (2017) indicated that the majority of sleep problems reported (awake all night, decreased sleep duration, early morning waking and daytime sleepiness) by parents were present from birth or developed within the first two years of the child’s life. One-hundred percent of parents of children with AS in the 16 months to three years bracket reported that their child had experienced a sleep problem. Abel and Tonnsen (2017) suggested that problems initiating and maintaining sleep which are present in 48–70% of older children with AS emerge in infancy but are targeted proactively by parents.

Sleep problems appear to subside somewhat with age in individuals with AS, as informants of adults with AS report lower levels of frequency and intensity of sleep disturbance (Larson et al., 2015). Trickett et al. (2018) report similar findings in relation to age as night waking scores were negatively correlated with age in the AS group. Sueri et al. (2017) reported that sleep problems present at childhood had disappeared in five of the 39 participants with AS at final evaluation. These findings are indicative of previous research which demonstrate that although sleep problems can improve, they can also persist into adulthood for some individuals (Bruni et al., 2004; Pelc et al., 2008a, b).

This review demonstrates that subjective measures (i.e. sleep questionnaires and parental/caregiver interviews) were the most common method of sleep assessment used across the selected studies, cited in five of the seven studies included for review (Abel & Tonnsen, 2017; Pereira et al., 2020; Sueri et al., 2017; Trickett et al., 2018, 2019). Similarly, subjective measures were used most frequently in the review by Spruyt et al. (2018). This suggests that subjective measures of sleep difficulties remain a more favoured means of assessment in terms of their frequency of implementation.

Of the seven studies reviewed, only two studies used the same sleep questionnaire (Trickett et al., 2018, 2019). The MSPSQ was used in isolation by Trickett et al. (2018), they reported descriptive statistics specifically effect size for night-awakening, daytime somnolence and sleep disorders. However, Trickett et al. (2019) used a further three sleep questionnaires and an actigraphy. Their study provided more quality and depth by reporting descriptive statistics (i.e. frequency and duration) for a higher number of sleep problems i.e. sleep-onset latency, sleep duration, night-awakenings and sleep efficiency. Although, the MSPSQ was used in two studies, there was a greater number of sleep outcomes observed in the study by Trickett and colleagues (2019) compared to the 2018 study when the MSPSQ was used as a singular measure. This finding suggests that sleep measures used in isolation are limited in the depth and quality of sleep assessment and may need to be supplemented by additional measures to assess more sleep difficulties.

Two studies implemented semi-objective measures in the form of actigraphy and video recording, they reported descriptive data for sleep-onset latency, sleep duration, sleep efficiency, night awakening and duration of wake time after sleep onset (Agar et al., 2017; Trickett et al., 2019). Specific to those applying semi-objective measures was the reporting of wake time (after sleep onset) duration which ranged between 76–103 min. This outcome was not achievable for studies using only subjective measures of assessment, this is assumed to be due to the difficulty of accurately recalling the exact moment of waking and subsequently recording the duration of time spent sleeping.

Sueri et al. (2017) was the only study included for review that utilised polysomnography data (i.e. EEG), however, only half of participants had sleep EEG’s and comparisons were unable to be drawn between infancy/childhood and point of intervention. However, the sleep problems assessed and identified did not differ greatly to that reported in studies utilising subjective sleep assessment tools. Although, it is important to note that SWI is only observable via EEG recording, specifically the evolution of SWI. In addition, EEG recording can assess for continuous spike and waves which are associated with non-convulsive status epilepticus (Pelc et al., 2008a). This is of importance as approximately 80–90% of individuals with AS develop epilepsy and analysis of such is beyond the scope of subjective measures of sleep (Fiumara et al., 2010). However, a limiting factor of EEG recording is the cost of implementation and interpretation is required by neurologists.

Spruyt et al. (2018) provided a descriptive synthesis of the sleep assessment tools used across the studies included in their review. However, the present review is distinct in that it provides an analysis of the differing sleep outcomes in relation to the assessment used. However, the findings in relation to sleep assessment should be considered with caution due to the high variability of sleep measures selected and the combination of sleep measures utilised across studies which affects the ability to generate accurate comparisons.

Sleep treatments/intervention were assessed in each of the seven studies reviewed. However, treatment effectiveness was only subjectively evaluated by parents/caregivers in one study (Trickett et al., 2017). For example, parents/caregivers were asked to rate interventions using a 4-point scale (i.e. helpful, not helpful, unsure, only helpful to fall asleep); due to the descriptive nature of the study, it was not feasible to gather pre and post intervention measures for different sleep constructs (i.e. sleep quality etc.). Considering the majority of studies used a subjective sleep assessment, this may also account for the limited information in relation to intervention effectiveness.

Over half of the population across all studies included for review reported the use of pharmacological intervention for sleep problems. This finding demonstrates the extent of sleep issues with the majority of individuals requiring medical intervention to combat sleep related problems.

Similar to that reported by Egan et al. (2020), the studies included in this review predominantly identified pharmacological intervention with only one study reporting use of a behavioural intervention (Trickett et al., 2017). This suggests that pharmacological intervention remains a frequent intervention/treatment for sleep problems in individuals with AS. When distinguishing between medications, melatonin remains the primary medication administered for sleep problems which is consistent with similar reports (Egan et al., 2020; Spruyt et al., 2018). Two of the studies included for review, reported the percentage of NT peers using medication to aid sleep (Trickett et al., 2018, 2019). Of the NT controls assessed, only one individual (N = 47, 1.5%) had been prescribed pharmacological intervention for sleep compared with 66.5% (n = 177) of the total AS sample (N = 266). This finding demonstrates that medical intervention to treat sleep difficulties is more frequent among AS individuals than NT controls. However, these findings warrant further research to evaluate the differences that exist between these populations as only two of the studies included for review provided such data for NT peers.

Trickett et al. (2017) reported on fourteen behavioral interventions which varied from timed-ignoring to interventions targeting an increase in the child’s communicative ability. Parents generally indicated behavioral interventions to be more efficacious than medication across participants in this study; a result which is representative of findings by Egan et al. (2020) who reported behavioral interventions to be more consistent across studies. Although the review by Egan et al. (2020) provided support for behavioural interventions, the more recent studies included in this review did not reflect a shift toward behavioural intervention.

Review Strengths and Limitations

This review used systematic, robust methodology to identify and select studies. It also reviewed the quality of studies using a recognised tool. However, the review is limited through only including studies reported in English. Doing so may have excluded valuable studies published in other languages.

Recommendations for Future Research

Of the seven studies selected for the present review, only two studies (Pereira et al., 2020; Sueri et al., 2017) included parent/caregiver reports for adults with AS (Pereira et al., 2020; Sueri et al., 2017), while one study utilised polysomnograpy (Sueri et al., 2017). However, due to the unavailability of infant/childhood EEG recordings, the evolution of SWI could not be explored which limited analysis. Future research could focus on utilising polysomnography at varying developmental milestones (i.e. infancy, toddlerhood, childhood, adolescence etc.) to assess SWI and its evolution across the lifespan.

It may be important to investigate parental/caregiver attitudes toward behavioural intervention for sleep as parents/caregivers would be responsible for implementation of such, consideration should focus on the impact on personal resources (i.e. training required, time, impact on stress etc.). Such research should focus on strengthening the quality of behavoiural intervention with the use of training manuals, fidelity checking (reviewing the extent to which the parent/caregiver adheres to the protocol when delivering the intervention) and the use of semi-objective sleep assessment tools. Previous research assessing the feasibility and fidelity of a manualised behavioural parent training intervention to target sleep difficulties among autistic children reported high treatment fidelity and sleep improvements when compared with a comparison group (Johnson et al., 2013). Future research should further evaluate the efficacy of behavioural interventions in general for the management/treatment of sleep difficulties.

The majority of studies on sleep problems in individuals with AS focused solely on children as their target population while adults were not addressed. Although sleep problems may improve into adulthood, it is not typical for each individual. Therefore, it is limiting that caregiver’s of adults with AS, who may report sleep disturbance are disregarded. More longitudinal studies need to be conducted across the population to give a clear indication of the developmental course of sleep problems in individuals with AS over time.

In the future researchers could improve consistency across studies by developing a standard protocol for measuring sleep in the AS population.

During the literature search, it was evident that studies using clinical trials to evaluate the efficacy of interventions to improve sleep difficulties were limited. Consequently, in the absence of randomised controlled trials, healthcare practitioners are restricted in their ability to make evidence-based decisions on the treatment of sleep problems in AS, making it a worthwhile area of future research.

Conclusions

This review extends the research on sleep in AS by best characterising sleep problems as issues in relation to sleep duration, efficiency, awakenings, daytime somnolence, and sleep-onset. It provides insight into the importance of selecting an appropriate measure(s) when assessing sleep. Although sleep outcomes largely remained consistent across studies, those using a broader range of sleep questionnaires were able to assess and report the outcomes of a greater number of sleep problems. Studies utilising semi-objective measures were able to provide more in-depth information, for example, the duration of night-time waking. Pharmacological treatment remains prevalent among individuals with AS to aid sleep. The general consensus from the literature is that the onset of sleep disorders tend to begin in early childhood in individuals with AS and improve into adulthood. However, the lack of adult respondents across studies, limit this finding. Future research should focus on longitudinal studies to explore how sleep problems develop or alter across time.

This evidence should compel researchers to carry out more detailed analysis of the sleep problems associated with AS and what they are attributed to. The different measures of sleep used (descriptive and EEG) give rise to a broader understanding of the overall sleep topography in the AS population and how it presents. It is imperative that clinicians/researchers are committed to evidence-based practice in order to extend the wellbeing of this population further and to help overcome limitations of current methodologies. They must adopt an integrative approach that maximises current research evidence, clinical expertise and knowledge of AS characteristics to better inform the assessment and intervention of sleep problems in individuals with AS.