The behavioral phenotype of Rubinstein–Taybi syndrome: A scoping review of the literature

Abstract Rubinstein–Taybi syndrome (RTS) is a rare genetic syndrome associated with growth delay, phenotypic facial characteristics, microcephaly, developmental delay, broad thumbs, and big toes. Most research on RTS has focused on the genotype and physical phenotype; however, several studies have described behavioral, cognitive, social, and emotional characteristics, elucidating the behavioral phenotype of RTS. The reporting of this review was informed by PRISMA guidelines. A systematic search of CINAHL, Medline, and PsychINFO was carried out in March 2021 to identify group studies describing behavioral, cognitive, emotional, psychiatric, and social characteristics in RTS. The studies were quality appraised. Characteristics reported include repetitive behavior, behaviors that challenge, intellectual disability, mental health difficulties, autism characteristics, and heightened sociability. Findings were largely consistent across studies, indicating that many characteristics are likely to form part of the behavioral phenotype of RTS. However, methodological limitations, such as a lack of appropriate comparison groups and inconsistency in measurement weaken these conclusions. There is a need for multi‐disciplinary studies, combining genetic and psychological measurement expertise within single research studies. Recommendations are made for future research studies in RTS.

Cognitive characteristics include intellectual disability (ID), ranging from mild to severe, difficulties with short term memory, delayed speech, and poor attention (Hennekam et al., 1992;Stevens et al., 1990;Waite et al., 2016). An IQ >70 has been reported some RTS individuals with a EP300 variant, although very few individuals with this genetic variant have been described (Fergelot et al., 2016). Further research has shown a wider IQ range  in individuals with the CREBBP variant using shorter non-verbal assessment tools that rely less on language, attention, and motor skills .
Behavioral characteristics associated with RTS include hyperactivity, impulsivity, and repetitive behaviors (e.g., repetitive speech and body stereotypy); some particular repetitive behaviors may be more frequent in RTS compared with other rare genetic conditions and autism spectrum disorder (ASD) (Waite et al., 2015). Age-related changes have also been described in RTS, with reports that mood difficulties and temper tantrums increase with age (Hennekam et al., 1992).
Behavioral, cognitive, and emotional characteristics associated with RTS may differ dependent the pathogenic variant and further research is needed to examine genotype-phenotype correlations.

| Establishing the behavioral phenotype of Rubinstein-Taybi syndrome
The term behavioral phenotype was introduced by Nyhan (1972) who argued observed behaviors were integral to genetic conditions and emphasized organic etiology. Since then, more widely accepted definitions have been introduced by Dykens (1995), who conceptualizes a behavioral phenotype as the increased likelihood of individuals with a particular condition displaying a behavior or set of behaviors relative to individuals who do not have that condition (Dykens, 1995); and O'Brien (2006) who describes it as a distinctive pattern of social, linguistic, cognitive, and motor observations normally associated with a biological or genetic disorder. Cognitive and emotional characteristics are often included under the umbrella term "behavioral phenotype" despite not being directly observable, as these characteristics can be indirectly measured and have been demonstrated to influence behavior (Flint, 1996;Waite et al., 2014).
Describing the behavioral phenotype associated with a genetic syndrome is of importance to families, carers, and individuals with a genetic syndrome. For example, when supporting their child, parents request further information on topics that align with core phenotypic characteristics associated with their child's syndrome (Pearson et al., 2018). A thorough description of the behavioral phenotype of a syndrome improves understanding and helps clinicians develop targeted advice (Waite et al., 2014). Furthermore, describing behavioral phenotypes informs the development of interventions and can improve decisions about how to adapt the environment to suit a person. There are numerous examples of how behavioral phenotype research has improved practice, such as research in Cornelia de Lange syndrome (CdLS). Self-injurious behavior is prevalent in CdLS, as it is in many rare genetic syndromes; however, in CdLS self-injury has been specifically associated with gastro-esophageal reflux. This knowledge has led to international clinical recommendations for the assessment of selfinjury in CdLS and subsequent treatment (Kline et al., 2018).
This scoping review aims to describe the behavioral phenotype of RTS by identifying literature that comments on the behavioral, cognitive and social characteristics of RTS. Mental health problems will be included in this review as they are often associated with cognitive, emotional, social and behavioral profiles (Waite et al., 2014). The literature will be summarized followed by an evaluation of the quality of the methodology applied in the studies for the purpose of drawing conclusions about the behavioral, cognitive, social, and psychiatric profile of RTS. Recommendations for further research will be identified.

| METHOD
The reporting of this scoping review aligns with the standards of the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) (Tricco et al., 2018).

| Search strategy
A search of CINAHL, Medline, and PsycINFO was carried out on 3rd March 2021 and included search terms relevant to the name of the syndrome and the cognitive, behavioral, and emotional phenotype (for a full list of search terms see Table 1). Truncations (*) were used to ensure alternative word endings were included and to allow for variations in spelling. The "AND" and "OR" functions were used to combine relevant search terms, and the advanced search function was used for phenotypic characteristics.

| Selection of studies
A total of 507 articles were identified. Following the removal of duplicates, 483 articles remained, and these were then screened by title and abstract (Stage 1 screening). Table 2 outlines the exclusion criteria used during the selection of studies. At Stage 1 screening, a total of 435 studies were excluded, including the following: studies where the cognitive, behavioral, emotional, psychiatric, or social phenotype of RTS, or genotype-phenotype correlations, were not the primary focus of the study (n = 333), case studies (n = 88), animal studies (n = 1), and book chapters (n = 1). The full texts of the remaining 48 studies were accessed at Stage 2, and 24 were deemed appropriate to include in the review due to containing results detailing the cognitive, behavioral, emotional, psychiatric, or social characteristics of RTS (see Figure 1). Several case series reporting on genotype-phenotype correlations were included due to (1) of the focus on of genotypephenotype correlations in the abstract indicating the possibility of aggregated group-level data, (2) full-text screening identifying that these papers reported aggregated data at group level, and (3) the importance of delineating characteristics associated with EP300 and CREBBP pathogenic variants (see Table 2 for criteria). The articles and reference list of the final 24 papers were backward searched, and additional two genotype-phenotype papers were added to the review as they provided aggregated group-level data, resulting in 26 papers.

| Data extraction and synthesis
Data extracted from each paper included demographic information, recruitment strategy, information on the genetic confirmation of the RTS sample, comparison groups, measures included, and key findings arising from the papers. These data were recorded using a table by a primary researcher. A second researcher reviewed the accuracy of data extraction for 25% of the studies. Any errors or omissions were highlighted and amended by the second reviewer. The results in this review are presented in a narrative method and the themes were derived from discussion between researchers and behavioral phenotype literature.

| Quality review
The literature was appraised using a quality framework adapted from a meta-analysis on the prevalence of ASD in rare genetic syndromes . The adapted framework focuses on five domains tailored to genetic syndrome research that reflect key threats to internal and external validity: sample identification (e.g., via syndrome support groups), level of confirmation diagnosis (e.g., clinical diagnosis, genetic testing), inclusion of comparison groups, properties of behavioral/psychological measures, and properties of cognitive assessments (see Table 3 for full criteria). Each domain was scored from 1 (poor) to 4 (excellent) producing a total score for each paper. If a paper did not include behavioral assessment or cognitive assessment, the respective domain was not scored. The total score was then divided by the highest possible score (20 for 5 domains, 16 for 4 domains) to produce a final score ranging from 0 (lowest possible score) to 1 (highest possible score). All scores were reported to two decimal places (see Table 4). A second researcher also completed quality ratings using the chosen quality framework for 25% of the papers to confirm the reliability of the ratings. An excellent level of inter-rater reliability (96.67%) was achieved.
T A B L E 2 Exclusion criteria used in selection of papers Stage 1: Abstract search exclusion criteria Case studies, reviews/meta-analyses, books, chapters Case Series, unless abstract eluded to the possibility of aggregated genotype-phenotype data.
Not peer-reviewed Non-human studies Behavioral, emotional, cognitive, psychiatric, social characteristics, or genotype-phenotype correlations are not the main focus of the study.

Studies of mixed diagnoses if RTS is not commented on separately
Stage 2: Abstract search exclusion criteria All criteria above with the addition of: Genotype-phenotype paper that did not comment on emotional, cognitive, psychiatric or social characteristics in full text T A B L E 1 Search terms used for the identification of relevant articles Syndrome search terms "Rubinstein-Taybi syndrome," "Rubinstein Taybi syndrome," "Rubinstein Taybi," "Rubinstein-Taybi," "Broad Thumb Hallux" "16p13.3" Cognitive, behavioral, and emotional phenotype search terms [behavio* or psychiatr* or psycholog* or emotion or mood or "mental health" or social* or Autism or Autistic or "Autis* Spectrum Disorder" or ASD or Cogniti* or "executive function" or "attention deficit hyperactivity disorder" or ADHD or intelligen* or intellectual* or IQ or "mental illness" or "adaptive function" or psychosocial or affect* or hyperactiv* or impulsiv* or overactiv* or "repetitive behavio*" or aggression or aggress* or "problem behavio*" or "challenging behavio*"] 3 | RESULTS

| Summary of participants and study quality
The results obtained from the papers and the quality ratings are summarized in Table 4 Fergelot et al., 2016;Negri et al., 2016). Conversely, studies that were rated highly on the quality of the methodology for assessing behavior and associated characteristics typically did not score as highly on diagnostic confirmation due to not conducting genetic testing as part of the study (e.g., Ellis et al., 2020;Waite et al., 2015). There was also a lack of comparison groups in 14 of 26 studies; however, when comparison groups were included, half were high-quality matched concurrent samples (6/12 studies).

| Behavioral characteristics
Ten of the twenty-six studies presented findings focusing on selfstimulatory or repetitive behaviors, aggressive behavior, and selfinjurious behavior.
3.2.1 | Self-stimulatory/repetitive behavior F I G U R E 1 PRISMA flowchart (Page et al., 2021) (N = 50) displayed "unusual behaviors," which are reported as being primarily self-stimulatory in nature, including rocking, spinning, and hand flapping. In a later study (N = 44), which covered a larger age range, repetitive speech was reported in 57% of children and 84.6% of adults; however, repetitive movements appeared to occur in fewer adults (38.5%) compared with children (77.4%) (Boer et al., 1999), as did adherence to strong routines. Neither study compared these findings to typically developing (TD) individuals, individuals with other rare genetic syndromes or ID of heterogeneous etiology.
Studies that included comparison groups have confirmed higher levels of self-stimulatory/repetitive behaviors in RTS. When comparing children with RTS to a comparison group of TD children matched for developmental ability and chronological age, Galéra et al. (2009) found that children with RTS (N = 39) scored significantly higher on the items: "flaps arms/hands when excited"; "makes odd/fast movements with fingers/hands"; and "pleased by movements/keeps doing them." When matched for ability to children with ID of heterogenous etiology, it has also been found that children (N = 3) with RTS displayed significantly more self-stimulatory behaviors compared with the children without RTS, although the sample size in this study was small (Gotts & Liemohn, 1977).
Comparisons between RTS and other neurodevelopmental condi-  Studies can only be classified into a category if all of the participants were tested using the outlined method. For instance, if only 50% of participants were genetically tested as part of the study, the study cannot receive a score of 4 and will receive a score of 3.

T A B L E 4 Summary of all studies and quality scoring
Authors ( However, RTS had lower levels of restricted conversation, repetitive phrases and echolalia compared with FXS and ASD. The RTS group also showed lower levels of adherence to routine and hand stereotypy than individuals with FXS and lower levels of cleaning than in ASD.
However, when matched for age, adaptive ability and verbal ability many of these differences were no longer statistically significant, although the RTS group still showed body stereotypy significantly more frequent than the DS group.

| Challenging behavior
Nine studies commented on challenging behavior including aggressive behaviors and self-injurious behaviors. The prevalence of challenging behaviors varies across the studies with common behavior problems reported to occur in 25% (N = 40) of individuals with RTS when assessed using the Achenbach Behavior Checklist (Hennekam et al., 1992), and 10% (N = 50) of maladaptive behaviors reported to be moderate to severe (Stevens et al., 1990). Boer et al. (1999) reported high levels of "verbal abuse" in both children and adults with RTS (86.2% and 84.6% respectively; N = 44), whereas other studies report less prevalent aggressive behavior (10.8%; N = 46) (Schorry et al., 2008). The psychometric properties of the measures used to obtain the estimates in these latter two studies is unclear, and this may have impacted results. In contrast, a study of 63 children and adults with RTS using a standardized measure, the Child Behavior Checklist, reported age-related differences with older individuals (>13 years) displaying significantly more aggressive behavior than younger individuals (Yagihashi et al., 2012). Most studies examining aggressive behaviors have been conducted with Type-1 (CREBBP) RTS or a group of RTS participants of unspecified genetic mechanism. One study reported on aggressive behavior in six individuals with the rarer EP300 inactivating mutations (Negri et al., 2016); one participant was reported as showing "aggressiveness"; however, as there was no comparison group, it cannot be concluded from this study whether aggression is a behavioral characteristic that is associated with EP300 inactivating mutations.

| Temper outbursts
"Residual anger" has been noted as a feature of RTS when matched to a heterogenous ID group (Gotts & Liemohn, 1977). However, the number of people with RTS in this study was small (N = 3). Hennekam

T A B L E 4 (Continued)
Authors (

| Self-injurious behaviors
Only three studies commented specifically on self-injurious behavior and the prevalence estimates varied across the studies, ranging from 6.5% to 53.8% (Boer et al., 1999;Schorry et al., 2008;Stevens et al., 2011 however, the age range of the sample is unknown, and the estimate of self-injurious behavior was produced by examining developmental and school performance data. None of the studies reported the topography of self-injurious behavior, and information on the severity and duration of these behaviors was absent. All studies that have reported on aggressive behavior, temper outbursts and self-injurious behavior have used different methods for data collection, none of which included direct observations of challenging behavior. Several studies did not report detailed information regarding the measures or method used to identify challenging behavior (e.g. Negri et al., 2016;Schorry et al., 2008;Stevens et al., 2011).
Although overall the findings point towards the presence of challenging behaviors, it is not possible ascertain a profile of these behaviors without operationalized definitions of challenging behaviors and complete information on the methods used to identify them.

| Cognitive characteristics
Seventeen studies commented on the cognitive characteristics associated with RTS, including seven that noted attentional difficulties, within their results sections (see Table 4 (Fergelot et al., 2016).
However, despite this, it is unclear whether standardized assessment measures of IQ were applied to assess degree of ID as this study used a questionnaire (non-specified) to obtain clinical information.
One study focused on a specific domain of cognitive function and reported impairments in verbal and visuo-spatial working memory across most age groups in people with RTS compared with typically developing children . There were no significant differences between the RTS group and comparison group on a visuospatial working memory task at the youngest developmental age of measurement (3 years old); however, the typically developing group's cross-sectional trajectory had a positive slope with age, whereas this remained flat for the RTS group, suggesting a particular difficulty in visuo-spatial working memory difficulty in RTS.

| Emotional and psychiatric characteristics
Thirteen studies discussed psychiatric or emotional difficulties in individuals with RTS (see Table 4). Levitas and Reid (1998) completed a psychiatric assessment and reported on the characteristics of 13 adults with RTS. It was identified that 8/13 of the sample had a 'mood disorder' and 4/13 were identified as having tics or OCD. A further study with a larger sample reported that 31% of adults with RTS had received a psychiatric diagnosis, mostly OCD, anxiety or depression (Stevens et al., 2011). In studies using standardized questionnaire measures, "internalising behavioural difficulties" were reported in 41.7% of RTS participants <20 years based on the Child Behavior Checklist (Ajmone et al., 2018). Age-related differences were also reported in one study, with older individuals with RTS (>14 years) displaying higher levels of anxiety, depression, nervousness and fearfulness compared with younger individuals (≤13 years) with RTS (Yagihashi et al., 2012). Anxiety has been reported in individuals with RTS carrying EP300 inactivating mutations as well as those with CREBBP mutations Pérez-Grijalba et al., 2019).  Rutter et al., 2003). Waite et al. (2015) reported that individuals with RTS on average had a moderate score on the SCQ; although, it was also reported that scores on the SCQ were likely elevated due to repetitive behavior in RTS rather than social-communication difficulties, and that characteristics associated with autism may be dissociated in RTS. Two studies included in this review reported on autism in small groups of individuals carrying novel EP300 variants. Autism and 'autism like' behaviors were reported in 3/8 and 1/6 participants in these studies respectively (L opez et al., 2018;Negri et al., 2016).
However, the absence of standardized measures or autism and small sample sizes means these findings should be interpreted with caution.

| Sociability and social interest
Social skills were examined in 12 studies, in which heightened levels of sociability or enhanced social skills were reported in 6 studies.
Individuals with RTS showed higher levels of social competence compared with TD children; social contact and interest were found to be significantly higher in the RTS group (Galéra et al., 2009;Hennekam et al., 1992). These findings are consistent across most studies measuring social characteristics, with over-friendliness reported in 77.3% of children (Boer et al., 1999) and that the RTS group "accepts social contacts readily" and significantly more than the matched comparison group (Gotts & Liemohn, 1977).

| DISCUSSION
The results confirm that several behavioral, cognitive, social and psychiatric characteristics appear to be present in individuals with RTS, including repetitive behavior, challenging behavior, ID, heightened sociability, mood disorders, and anxiety. A key limitation is the heterogeneity of assessment methods used across the studies to measure these areas. With some studies providing minimal information about the measures used and some using non-standardized measures, it has highlighted the need for a more robust and uniform methodology using direct and indirect tools. A further limitation that made it difficult to interpret the findings is the heterogeneity in the age of the participants across the studies, with some studies including only children and others including adults in their sample. Furthermore, some studies were dated, and some had small samples, which made identifying specific phenotype-genotype correlations very difficult. Moreover, the lack of longitudinal studies does not allow for a natural trajectory of the behavioral phenotype of RTS to be established. Although similar characteristics are often reported across studies, the lack of contrast groups in over half of the studies limits the conclusions that can be drawn regarding whether these characteristics are more likely to be displayed in someone with RTS relative to someone who does not have RTS.

| Behavioral characteristics
The results showed variability in the prevalence estimates for repetitive movements (31% and 77.4%) (Boer et al., 1999;Galéra et al., 2009). This may partly be explained by the different tools used across the studies, as they vary in the way in which they identify and mea- Prevalence estimates for challenging behavior also varied across the studies and this may be explained by the lack of a clear definition of challenging behavior. Without a shared understanding of behaviors that are deemed to be challenging, it is difficult to measure the presence of those behaviors in individuals with RTS. Emerson (1995) described challenging behavior as "behaviours of such an intensity, frequency or duration that the physical safety of the person or others is likely to be placed in serious jeopardy." Research studies that have focused on the epidemiology of challenging behavior identified specific behaviors that are considered to fall within the category of challenging behavior, including aggression, self-injurious behavior and property destruction (Borthwick-Duffy, 1994;Kiernan & Qureshi, 1993;Qureshi, 1994;Qureshi & Alborz, 1992). Some of the studies included in this review measured aggression and self-injurious behavior, however other studies did not. For example Hennekam et al. (1992) reported that common behavior problems occurred in 25% of individuals with RTS; however, the tool employed in this study (Achenbach Behavior Checklist;Achenbach & Edelbrock, 1983) did not measure aggression, self-injurious behavior or destruction of property. The types of behaviors captured by this tool include "wets bed," "thumb-sucking," "picks nose," and "temper tantrums." Although these behaviors can be of concern, they do not necessarily fall within the definition of challenging behavior.
The studies in this review focused solely on documenting the prevalence of challenging behavior. None of the studies documented the etiology; however, this is particularly important in enhancing our understanding of the factors associated with challenging behavior in individuals with RTS. Challenging behaviors have been shown to serve as a communicative function in individuals with ID (Durand & Merges, 2001;Mirenda, 1997;Richman et al., 2001) suggesting that this is an attempt by the individual to communicate something, such as needing help, requesting access to an object or activity, or communicating dislike for something (Bopp et al., 2004;Carr et al., 2002;Horner, 2000;Kincaid et al., 2002). Exploring the functions of challenging behavior in individuals with RTS including the role of communication is imperative in allowing appropriate interventions and support to be offered.

| Cognitive characteristics
The findings across the studies show that most individuals with RTS fall within the moderate ID range Hennekam et al., 1992;Schorry et al., 2008); however, there was variability in the extent of cognitive impairment reported across studies. For example, Schorry et al. (2008) reported that 44.3% of their sample had an IQ below 50, 53.2% with an IQ between 50 and 75 and 2.5% with an IQ above 75. Another study reported a mean IQ of 35.6 (range 25-79) and a sharp decline in IQ as age increased (Hennekam et al., 1992).  (Fergelot et al., 2016). However, despite this, it is unclear whether standardized assessment measures of IQ were applied to assess degree of ID as this study used a questionnaire (non-specified) to obtain clinical information.
Most studies that included individuals with EP300 pathological variants had small sample sizes, apart from Fergelot et al. (2016), so the ability to compare characteristics between those with EP300 to CREBBP variants is limited. The findings on EP300 are also inconsistent across studies and are based on unstandardised/non-specified assessments; however, there is some evidence indicating the EP300 variant may be associated with less severe ID and lower rates of autism relative to the CREBBP variant, and that individuals with the EP300 variant may also experience anxiety .
These findings indicate the importance of conducting genetic testing to confirm an RTS diagnosis; however, a number of the studies confirmed diagnosis through the presence clinical characteristics or by participants reporting that a diagnosis had previously been confirmed by a clinical geneticist or pediatrician (e.g., Stevens et al., 1990;Waite et al., 2016). While genetic testing is the gold standard within research studies, it is possible that the practicalities of conducting these tests within behavioral research settings is a barrier to the inclusion of these tests. In addition, the genetic mechanism leading to RTS cannot be identified in all individuals and, therefore, clinical features are still essential for the confirmation of the presence of RTS (Stevens, 2019

| Social characteristics
The results showed a high prevalence of ASD characteristics in individuals with RTS. These characteristics include restricted preferences, sensitivity to noise, difficulties with unexpected change and selfstimulatory behaviors (Stevens et al., 2011). These findings are not unexpected as research has shown higher rates of ASD in rare genetic conditions compared the general population ; however, it is important to recognize that none of the studies used comprehensive observational assessments to identify ASD, such as the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000); therefore, we cannot say with certainty whether ASD is more  (Crawford et al., 2015;Crawford et al., 2017;Waite et al., 2015) used the SCQ (Rutter et al., 2003), which is a well-validated tool that identifies ASD characteristics in individuals with ID. The measure has high concurrent validity with the ADOS, as the total score on the SCQ is strongly related to the total score on the ADOS (Berument et al., 1999;Lord et al., 1994), therefore making it a more suitable tool to be used to assess ASD in individuals with RTS. shyness, social anxiety and gaze avoidance, emotion sensitivity and willingness to interact may also be preserved (Cornish et al., 2007;Hall et al., 2006;Turk & Graham, 1997). Research has also shown heightened levels of ASD in individuals with CdLS based on the total ADOS score; however, domain and item specific analysis indicate individuals with CdLS show more eye contact and gestures, and less repetitive behavior and stereotyped speech than the ASD group (Moss et al., 2012). These findings, along with reports of prolonged eye gaze and heightened social anxiety in CdLS (Collis et al., 2006;Goodban, 1993), suggest that the profile of social impairments in CdLS may be different to that observed in ASD. With regard to RTS, recent research conducted following this review has indicated the benefit of examining autism characteristics and social characteristics at this level of fine-grained description indicating nuanced differences to those observed in ASD Ellis et al., 2021;Taupiac et al., 2021).
Delineation of the profile of ASD in rare genetic syndromes clearly demonstrates how subtle differences in phenomenology can be obscured when the presence or absence of ASD is estimated solely from clinical cutoff scores. The use of questionnaires to assess ASD may have inflated prevalence estimates in RTS due to the high frequency of repetitive behavior in the syndrome. Many individuals with RTS may have met the cut off for ASD due to the presence of repetitive behavior alone.

| Clinical implications
The findings from across the studies indicate that repetitive behavior and behaviors that challenge are likely to be specific features of RTS, thus highlighting the need for appropriate support for individuals who display these behaviors. There are no intervention studies for challenging behavior in individuals with RTS, however, there are effective interventions and clinical guidance available for behaviors that challenge in ID populations (National Institute for Health and Care Excellence, 2018). Challenging behavior has been found to be more likely in individuals who have an increased need of assistance and those who have restricted receptive and expressive communication (Emerson et al., 2001;Emerson & Bromley, 1995), so supporting the development of communication from an early age and providing increased mobility support, may help toward preventing and managing behaviors that challenge.
The use of augmentative and alternative communication (AAC) strategies including aided modalities such as PECS (Bondy & Frost, 1994) and unaided modalities such as Makaton (Walker, 1987) have been recommended for use with individuals with ID (Beukelman & Mirenda, 2013). Both unaided and aided modalities of AAC have been successfully taught to individuals with ID and severe communication difficulties (Kagohara et al., 2013;Lancioni et al., 2013;Schwartz & Nye, 2006;Sutherland et al., 2010;Wendt, 2009). Early input from speech and language therapy for individuals with RTS would therefore be very beneficial in supporting the development of communication and subsequently reducing behaviors that challenge. This will have a positive impact on the quality of life of those with RTS.
Interventions for repetitive behaviors may not be necessary unless the behavior is having a significant impact on quality of life.
However, if adherence to routines becomes problematic some interventions that have been developed for other conditions may be appropriate (e.g. Bull et al., 2017). Finally, several studies have suggested that anxiety may occur in RTS. There are very few validated interventions for anxiety in people with severe to profound ID (Vereenooghe et al., 2018), however, behavioral strategies for anxiety may be able to be adapted for this group. In those with mild to moderate ID, behavioral strategies or adapted CBT may be appropriate (Hatton, 2002;Jahoda et al., 2017). There is guidance available on supporting individuals with learning disabilities who are experiencing anxiety using low-intensity CBT (Dagnan et al., 2015). Some of the adaptations suggested for individuals with learning disabilities may be appropriate for individuals with RTS, including adjusting the length of the therapy session; providing support when filling in outcome measures; using easy read resources; focusing on behavioral aspects of an intervention; and finally considering inviting carers/family members to the session if the individual feels this would be beneficial (Dagnan et al., 2015).

| LIMITATIONS OF THIS REVIEW
Although this was a systematic search, it is possible that some publications were missed if they were not listed in the identified databases.
Furthermore, due to initially screening papers based on title and abstract, it is also possible that some papers commented on the behavioral phenotype of RTS in the full text but were screened out.
Despite this, this review provides a useful overview of the status of the RTS literature, particularly regarding methodological issues that may preclude accurate identification of syndrome characteristics.

| CONCLUSION
Research on RTS to date has made some progress in describing the behavioral phenotype of RTS. This review has highlighted the need for further research to replicate findings, to address the inconsistencies across studies and the lack of comparison groups. The varying methodology used to measure the behavioral phenotype of RTS has drawn attention to the importance of using standardized assessment tools that are appropriate for individuals with rare genetic conditions.
It may be useful to create a standard criterion of instruments that are suitable for use to improve the overall quality of the research and to allow for a clearer comparison of the research findings. A thorough understanding the behavioral, cognitive, and emotional characteristics of RTS will allow for appropriate interventions to be developed and trialed to ensure that evidence-based support is developed to help those with the condition and their families. Greenhill reviewed data charting and analysis for accuracy. All authors