As part of the efforts to better capture the considerable heterogeneity within autism and to provide a more nuanced and detailed picture of the experiences of autistic individuals, investigations are increasingly moving beyond symptoms traditionally viewed as ‘core’ (Astle & Fletcher-Watson, 2020). A key finding to emerge from these investigations is a potential difference in temporal synchrony tendencies between autistic and neurotypical individuals. Temporal synchrony can be defined as two or more actions integrated in time and/or individuals entrained to each other in social interaction. This can include cross-domain integration within an individual, e.g. audio-visual, audio-motor, visuo-tactile, or visuo-motor integration, or it can refer to interpersonal integration during social interaction, e.g. the alignment of gestures, facial expressions, and language during communication. Thus defined, differences in temporal synchrony tendencies have been proposed to underlie some of the observed differences between typically developing and autistic individuals and to be a promising intervention target (e.g. McNaughton & Redcay, 2020). The goal of the present study is to help evaluate these claims by systematically reviewing the evidence on the nature, extent, and generalisability of temporal synchrony differences between typically developing and autistic individuals, as well as interventions targeting temporal synchrony in autistic individuals.

It has been proposed that during the course of an interaction, individuals become temporally synchronised to each other, meaning that both verbal and non-verbal behaviour during communication align in terms of the phase and frequency or period (McGrath & Kelly, 1986). While there is ample empirical evidence supporting these propositions in neurotypical development (Condon & Sander, 1974; Petitto et al., 2000, 2001), studies on temporal synchrony in autism have identified differences in temporal constructs such the flow and timing of conversations (Wimpory, 2015). The first review study to report on temporal processing and management skills in autistic individuals suggested that there were processing difficulties in various temporal patterns (millisecond, interval, circadian) and difficulties in audio-visual temporal integration (Jurek et al., 2019). Another review study reported inter- and intra-personal asynchrony in autistic individuals, suggesting an altered experience of time and atypical integration of audial, visual, and motor sensory inputs (Bloch et al., 2019). Temporal deviations in various sensory inputs are most likely to influence behaviour related to the ability to establish and maintain synchronous, reciprocal engagement with others, and therefore, may contribute to social-communication differences in autism (McNaughton & Redcay, 2020). Indeed, it has been proposed that diagnostic practices would benefit from assessing a wider presentation of symptoms associated with autism, and measures of interpersonal alignment could serve as implicit measures aiding diagnosis (Bloch et al., 2019).

While reviews to date have provided important insights and helped provide an expanded focus on autistic symptomatology that goes beyond traditionally ‘core’ symptoms, they have done so only in limited categories of temporal synchrony or using methods that do not involve systematically identifying and evaluating all the primary studies in this area (e.g. Jurek et al., 2019; McNaughton & Redcay, 2020). On the basis of a scoping search, it was found that temporal synchrony has been empirically studied in numerous domains and with a diversity of paradigms. Aspects of temporal synchrony that have been examined in the context of autism to date include audio-visual, audio-motor, visuo-tactile, visuo-motor, social-motor, and conversational synchrony and within these, a variety of paradigms have been employed. For example, studies examining temporal synchrony tendencies across the auditory and visual domains have used preferential looking, illusion susceptibility, and temporal order judgment paradigms, while studies examining conversational synchrony have used methods such as the coding of dyadic interactions for gestural synchrony.

Given the volume and diversity of research in this area, and the potential implications for illuminating autistic experiences and informing interventions, a systematic review would thus be beneficial to provide an overview of the evidence concerning various types of temporal synchrony, and to assess the quality of the available data. Another advantage of a systematic review of this topic is that it can identify common methodological weaknesses and evidence gaps that can provide the basis for recommendations to help strengthen knowledge in this field. It can also identify the generalisability and the limits of the findings of asynchrony in autism and define the domains and tasks in which it is in evidence.

Therefore, the present information gap will be filled by reviewing different aspects of temporal synchrony in autism with the aim of answering whether, and in which contexts, the research findings indicate atypical temporal synchrony in autism.

Methods

A systematic review protocol was developed following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) recommendations (Moher et al., 2015) prior to beginning the review. The only deviations from the original protocol were that data on participant sex/gender was not mentioned in the data extraction section of the protocol, and neural temporal synchrony studies were excluded from the review.

Search Strategy

Two main search methods were used: (a) database search and (b) examination of reference lists.

  1. a)

    The database search involved key concepts of the study defined as ‘autism’, ‘temporal’, and ‘synchrony’. To identify relevant studies, synonyms were used in search terms (Table 1). Boolean operators (‘AND’, ‘OR’, and ‘NOT’) were used to manage search strings and to combine key concepts. The operators were tailored to the syntax of each of the databases used: EBSCO (which combines MEDLINE, Child Development and Adolescent Studies, and CINAHL), OVID (which combines PsycINFO, PsycArticles and EMBASE), Web of Science, and Scopus.

    An example of the search string used and adapted to other databases was (Autism OR Asperger OR autistic disorder OR Autism spectrum disorder OR autistic syndrome disorder OR ASD OR pervasive developmental disorder) AND (temporal OR time OR timing OR chrono* OR pattern) AND (synchron* OR dyssynchron* OR asynchron* OR desynchron* OR entrain* OR align*).

  2. b)

    Reference lists of both the relevant papers included in the review and literature reviews identified through database searches were screened to identify additional pertinent works.

Table 1 Example search terms
Full size table

Data Extraction

The bibliographic software programme Zotero was used to manage and store relevant studies. Duplicate records were removed using the same software. Search results were independently evaluated for eligibility by two reviewers based on the study title, abstract, and full text.

Studies were selected according to the criteria below:

  1. i.

    Participants with a diagnosis of autism based on either Diagnostic and Statistical Manual of Mental Health Disorders (DSM) or International Classification of Diseases (ICD) criteria and/or individuals who met the research and/or clinical cut-off points of a validated screening tool.

  2. ii.

    Participants from all age groups

  3. iii.

    Studies that were reported in English

  4. iv.

    Temporal synchrony investigated in various contexts to be discussed later in coherent groups

  5. v.

    Both randomised and non-randomised studies including intervention studies

  6. vi.

    There was no restriction on the publication year

Studies were excluded according to the criteria below:

  1. i.

    Animal studies

  2. ii.

    Grey literature identified as dissertations/theses, abstracts, unpublished reports and ongoing trials

  3. iii.

    Cohort studies and other non-experimental designs

  4. iv.

    Non-empirical articles, including case reports, book chapters, letters, editorials, and opinion pieces

  5. v.

    Studies investigating neural temporal synchrony

  6. vi.

    Studies for which it was not possible to obtain the full text

The selection process was completed when an agreement was reached between the first author and the second reviewer. In the case of a difference of opinion, the issue was resolved by involving a third reviewer. The final decision on the inclusion or exclusion of studies (along with reasons for exclusion) was documented for each study. The initial literature search returned 2000 recordsFootnote 1 at the time of search, which was reduced to 675 after manually excluding duplicates. An additional 21 records were identified through reference lists of relevant papers. Following the preliminary screening of the titles and the eligibility criteria, 602 records were excluded; a further 94 records were excluded after screening the abstracts and/or full-texts based on the relevance of data to the current study (Fig. 1). Three studies were excluded after the data extraction: Feldstein et al. (1982) owing to the lack of a control group; Massaro and Bosseler (2002) for not reporting the autism assessment method; and Tavano et al. (2014) for not reporting the results of the paradigm relevant to the review.

Fig. 1
figure 1

Flow Chart of the Systematic Review Based on PRISMA Guidelines

Full size image

Data from each selected study was extracted both by the author and the second reviewer using a piloted extraction form. The extracted data are presented in two tables per category of temporal synchrony. The ‘study characteristics’ tables include information regarding study names and locations, sample sizes, study designs, autism assessment, comparator groups, temporal synchrony assessment, study paradigms, results, findings, and finally, the group main effects were reported in partial eta squares (\({n}_{p}^{2})\) when available. The ‘participant characteristics’ tables include information regarding participants’ mean age and mean IQ, level of functioning (high vs. low).

Each selected study was evaluated regarding the risk of bias and the evaluation was made by two reviewers using an assessment tool selected according to each study design. The quality of the selected randomised controlled studies was assessed by the ‘Risk of Bias’ (RoB 2) tool (Sterne et al., 2019). The tool has five domains assessing bias arising from the randomisation process, due to deviations from intended interventions/exposure, missing outcome data, measurement of the outcome and finally, selection of the reported results. The selected non-randomised controlled studies were assessed by the ‘Risk of Bias in Non-randomised Studies – of Interventions’ (ROBINS-I) tool (Sterne et al., 2016). The assessment was based on pre-intervention, at intervention and post-intervention domains where bias might occur. In order to summarise the quality assessment, risk of bias plots (Figs. 2 and 3) were created using the ‘Risk of Bias Visualisation’ (robvis) web application (McGuinness & Higgins, 2020).

Fig. 2
figure 2

Risk of Bias Assessment of Non-randomised Controlled Studies

Full size image
Fig. 3
figure 3

Risk of Bias Assessment of Randomised Controlled Studies

Full size image

Data Analysis

Meta-analysis was not conducted due to heterogeneity in autism diagnostic assessment, participant age groups, experimental paradigms, and, thus, measurements of outcome. Given these differences between studies, it was judged that it would not be valid to pool effect sizes across studies. Instead, a narrative synthesis approach was used to combine results across studies.

Results

The systematic review of the literature returned 32 articles concerning audio-visual (n = 13), audio-motor (n = 3), visuo-tactile (n = 3), visuo-motor (n = 3), social motor (n = 3), and conversational synchronies (n = 5). Intervention studies (n = 2) on temporal synchrony were also included. Studies confirmed the diagnostic assessment of autism according to one of the following: clinical diagnosis; DSM-III-R; DSM-IV; DSM-V (American Psychiatric Association (1987), 1994, 2013); Autism Diagnostic Observation Schedule (ADOS) (Gotham et al., 2006); Autism Diagnostic Interview-Revised (ADI-R) (Kim et al., 2013); or International Classification of Diseases (ICD)-10 (World Health Organization, 2004). While a majority of the studies focused on autism diagnosis, one of the studies (Magnée et al., 2008) included a Pervasive Developmental Disorders (PDD) experimental group.

An autism sample size of 15 or more was present in 23 studies (71.8%) while an autism sample size less than 15 was present in 8 studies (25%). Half of the studies (n = 16) reported high functioning autism groups and the rest did not report the level of functioning of the participants.

According to the risk of bias assessment of non-randomised (Fig. 2) and randomised (Fig. 3) controlled studies, the overall quality of the included studies was low (n = 25) to moderate (n = 6). Only one study (Kwakye et al., 2011) presented a high risk of bias in three domains (D2; D4; D5).

Audio-visual Synchrony in Autistic Individuals

A majority of empirical studies on temporal synchrony investigated audio-visual synchrony, with thirteen studies in this category meeting inclusion criteria (Tables 2 and 3). The first study using a preferential looking paradigm reported poorer performance of autistic individuals in response to linguistic stimuli (Bebko et al., 2006). Similar results were found in other studies where autistic individuals showed lower scores in audio-visual synchrony tasks that were mainly speech-specific (Righi et al., 2018; Stevenson et al., 2014; Woynaroski et al., 2013). Iarocci et al. (2010) also reported that an autistic participant group showed less visual and more auditory influence on speech perception, providing evidence of the linguistic stimuli impact on audio-visual synchrony in autistic individuals. Another study investigating a preferential looking paradigm with a split-screen video of two identical individuals speaking found that the autistic individuals looked at the synchronous video less than the typically developing group (Grossman et al., 2015).

Table 2 Studies on audio-visual synchrony and autism
Full size table
Table 3 Participant characteristics of studies on audio-visual synchrony and autism
Full size table

While the findings indicate speech- and linguistically influenced performance in audio-visual synchrony in autistic groups, there seems to be evidence for temporal processing differences between autistic participants and comparator groups from studies where temporal order judgment tasks were adapted. It was found that autistic individuals tended to show less sensitivity in judgments of audio-visual temporal order (de Boer-Schellekens et al., 2013), reported the flash-beep illusion effect over an extended range of stimulus onset asynchronies (Foss-Feig et al., 2010), and had higher thresholds on the auditory task relative to typically developed groups, evidencing reduced synchronous tendencies in auditory temporal processing (Kwakye et al., 2011). Such tendencies of a PDD group on complex phonological processes in audio-visual interactions were also reported (Magnée et al., 2008). However, van der Smagt et al. (2007) did not find such differences on temporal order judgment tasks between autistic and typically developed groups.

Furthermore, another autistic group was reported to have preferences that are less synchronous when audio-visual tasks involve human faces and voices (as opposed to non-human stimuli, such as bouncing balls; Mongillo et al., 2008). Grossman et al. (2009) found no differences between autistic and typically developed groups in the accuracy of onset asynchrony detection in video clips of human faces and speech.

In summary, the evidence suggests that autistic individuals’ performance in various audio-visual tasks are different and less synchronous when compared to typically developed individuals. The majority of the evidence seems to suggest that audio-visual synchrony was affected by speech and linguistic related tasks resulting in a poorer performance of autistic individuals.

Audio-motor Synchrony in Autistic Individuals

Three studies evaluated audio-motor synchrony in autistic individuals (Tables 4 and 5). There was agreement between the three studies that the autistic groups’ performance was poorer in audio-motor synchrony tasks (Morimoto et al., 2018; Whyatt & Craig, 2013; Zapata-Fonseca et al., 2019). It was suggested that the autistic group showed more variability in temporal processing parameters than neurotypical individuals, and that temporal processing instability was related to altered motor performance (Morimoto et al., 2018). Autistic individuals were also found to perform poorer than both receptive-language and non-verbal control groups (Whyatt & Craig, 2013), and to be less inclined to sustain mutual interaction in general (Zapata-Fonseca et al., 2019). Empirical evidence on audio-motor synchrony thus suggests that autistic individuals showed poorer performance on audio-motor synchrony tasks when compared with typically developed individuals.

Table 4 Studies on audio-motor synchrony and autism
Full size table
Table 5 Participant characteristics of studies on audio-motor synchrony and autism
Full size table

Visuo-tactile Synchrony in Autistic Individuals

Three studies evaluated visuo-tactile synchrony in autistic individuals (Tables 6 and 7). All three studies used a rubber hand illusion paradigm and reported similar outcomes: autistic individuals’ performance was poorer when compared to neurotypical individuals (Cascio et al., 2012; Greenfield et al., 2015; Ropar et al., 2018). When compared with chronological age-matched controls, autistic individuals showed similar performance to younger mental-age controls: reduced embodiment of the spatially incongruent, but temporally congruent, hand at shorter temporal delays (Ropar et al., 2018). Thus, the available evidence on visuo-tactile synchrony suggests poorer task performance in autistic participant groups, indicating delayed visuo-tactile integration.

Table 6 Studies on visuo-tactile synchrony and autism
Full size table
Table 7 Participant characteristics of studies on visuo-tactile synchrony and autism
Full size table

Visuo-motor Synchrony in Autistic Individuals

Visuo-motor synchrony in autism was evaluated by three studies (Tables 8 and 9; Edey et al., 2019; Kawasaki et al., 2017; Marsh et al., 2013). These studies used different paradigms to assess synchrony in autistic individuals but two reported similar results. During tasks in which participants were expected to show temporally synchronous motor behaviour with given visual stimuli, autistic individuals showed greater difficulty (Kawasaki et al., 2017) and less tendency to cooperate simultaneously and synchronously (Marsh et al., 2013), compared to neurotypical individuals. However, Edey et al. (2019) reported that autistic participants exhibited fewer temporal errors relative to their neurotypical counterparts in the visuo-motor synchrony task. Thus overall, these studies suggested less synchronous performance in autistic individuals.

Table 8 Studies on visuo-motor synchrony and autism
Full size table
Table 9 Participant characteristics of studies on visuo-motor synchrony and autism
Full size table

Social Motor Synchrony in Autistic Individuals

Three studies investigated social motor synchrony in autistic individuals in comparison to neurotypicals using different experimental conditions (Tables 10 and 11). Autistic individuals demonstrated less synchronisation in both spontaneous and intentional interpersonal coordination when they were asked to swing pendulums with others (Fitzpatrick et al., 2016). They also exhibited fewer stable patterns of social synchronisation ability, while their performance of motor movements was slower in both spacing and timing (Fitzpatrick et al., 2017). However, in another study no difference between autistic and neurotypical individuals was found when participants were asked to perform movements that had been introduced earlier in synchrony with the experimenter (Fitzpatrick et al., 2013).

Table 10 Studies on social motor synchrony and autism
Full size table
Table 11 Participant characteristics of studies on social motor synchrony and autism
Full size table

In summary, one study on social motor synchrony found no difference between autism and typically developed group performances, but two suggested reduced temporal synchrony in the social motor performance of autistic individuals.

Conversational Synchrony in Autistic Individuals

Five studies investigated conversational synchrony in autistic individuals compared to neurotypical subjects (Tables 12 and 13). It was reported that autistic individuals synchronised gestures less closely with co-occurring speech (de Marchena & Eigsti, 2010), and interpersonal synchrony was more observable in neurotypical individuals (Georgescu et al., 2020). Furthermore, spontaneous verbal and non-verbal gestures were also reported to be less likely in autistic individuals (Hobson & Lee, 1998). However, two studies did not find any difference in performance between autistic and neurotypical participants. Autistic and typically developed groups were found to be equally able to detect physical differences on two given faces when temporally desynchronised (Shah et al., 2016). The tendency to align on a pragmatically conditional aspect of language was also found to be similar between autistic and typically developed groups (Branigan et al., 2016). In summary, a majority of the studies reported a reduced synchronous tendency in gesture performance and interpersonal synchrony in autism.

Table 12 Studies on conversational synchrony and autism
Full size table
Table 13 Participant characteristics of studies on conversational synchrony and autism
Full size table

Intervention Studies on Temporal Synchrony in Autistic Individuals

Two intervention studies on temporal synchrony were included in the systematic review. Griffioen et al. (2020) used six weekly sessions of 30-min-long dog-assisted therapy with autistic individuals and individuals with Down’s syndrome. Intervention activities included psychomotor and socialisation skills that would ensure aligned motor action between the child and the therapy dog. While the results showed a significant increase in synchronous interactions between autistic children and their therapy dogs, the number of participants (n = 10) in the study limits the generalisability of the findings.

Another study provided 10 h/week of classroom intervention, which provided a supplementary curriculum targeting engaged imitation, joint attention and affect sharing (Landa et al., 2011). Additionally, 38 h of parent education and 1.5 h of home-based parent training was also given. At the end of the intervention, Landa et al. (2011) reported a significant increase in socially engaged imitation with eye contact in autistic toddlers. The study provided evidence for development in temporal synchrony in dyadic interactions.

Both of the intervention studies included in the systematic review concluded an increased performance in social temporal synchrony in autistic individuals after the intervention.

Discussion

This is the first study examining temporal synchrony in autism across all domains in which this has been previously examined by means of a systematic review. The systematic review identified the main characteristics of studies and participants as well as the types of synchrony and the study paradigms. Results showed that autistic participants tended to display more temporally asynchronous behaviours when performing tasks that required audio-visual, audio-motor, visuo-tactile, visuo-motor, social motor, and conversational sensory integration. Intervention studies also showed significant improvements to interpersonal and social synchrony in autism when temporal orders were considered. However, sample sizes were generally small and not justified, therefore, future studies should attend to ensuring adequate statistical power, as well as experimental rigour. Taken together, our findings point to the generality of temporal synchrony differences in autistic individuals compared to typically developing individuals. This suggests that temporal synchrony is a promising concept for explaining some of the differences between autistic and typically developing individuals in their daily life functioning and experiences and may also represent a promising diagnostic marker and/or intervention target.

The evidence on audio-visual synchrony in this review suggested poor performance of autistic individuals on tasks, and this was found related to speech and linguistic stimuli (Tables 2 and 3) while studies on conversational synchrony also found asynchronous gesture behaviours related to semantic speech in autistic adolescents (Tables 12 and 13). The findings seemed to provide evidence that autistic individuals experience difficulty in temporally integrating verbal and non-verbal communication that might cause social-communication difficulty (de Marchena & Eigsti, 2010). While social-communication ability is considered to be one of the ‘core deficits’ of autism, and although there are numerous theories on sensory perception in autism (Brock et al., 2002; Happé and Frith, 2006; Mottron et al., 2006), the nature of the links between autism and temporally asynchronous behaviour remains unclear as these theories do not directly address timing aspects of synchronous behaviour. However, it was previously established that poorer temporal acuity across visual and audial information is related to a weaker ability to temporally bind and integrate the two sensory inputs (Stevenson et al., 2014, 2018). Therefore, temporally asynchronous behaviours in autism could result from reduced multisensory integration, as proposed by Stevenson et al. (2014).

In addition to the instability of temporal processing and the lack of temporally synchronous behaviours in audio-, visuo-, and social motor performance in autistic groups, evidence seems to suggest that atypical sensory processing and temporal integration in autism is evident as proposed previously by Wallace and Stevenson (2014). Based on autistic individuals’ performances on rubber hand illusion tasks, studies on visuo-tactile synchrony (Tables 6 and 7) also provided evidence that autistic individuals experience body representation difficulties, which might cause temporally asynchronous behaviours and in turn contribute to social and sensory difficulties within the population (Ropar et al., 2018).

While there were only two intervention studies included in this review (Tables 14 and 15), Landa et al. (2011) provided evidence for improving socially engaged behaviour in autistic toddlers by layering a supplementary curriculum, which is a promising outcome in terms of the plasticity of social-communication development in autism. However, there are limitations to the generalisability and reliability of the dog-assisted therapy by Griffioen et al. (2020) due to its small sample size of 5. Indeed, the interpretability of both of the intervention studies included in this review is limited due to the limited amount of supporting evidence on the topic.

Table 14 Intervention studies on temporal synchrony and autism
Full size table
Table 15 Participant characteristics of intervention studies on temporal synchrony and autism
Full size table

Quality of the Evidence

Our quality assessment identified some studies with very small autistic participant samples of n = 5 (Griffioen et al., 2020) and n = 9 (Fitzpatrick et al., 2016) and many others with sample sizes of only 15–30 and no power calculations to justify such small groups. More attention should thus be paid to statistical power in future studies in order to achieve more definitive results.

It was also established that the reporting of participant characteristics in the studies of temporal synchrony in autism was generally poor. While half of the studies included in this review reported recruiting high-functioning autism groups, the remainder did not report the level of functioning of the participants. More than half of the studies also did not provide information regarding mean IQ scores. It is evident that temporal synchrony performance should be assessed in autistic individuals with low levels of functioning, low IQ scores and/or an intellectual disability (ID) to ensure that these individuals are not under-represented and that the reporting of these characteristics should not be neglected. The importance of such participant characteristics has been examined previously: it was established that the proportion of individuals on the spectrum with ID is close to 50%, yet there tends to be a selection bias in all fields of autism research that leads to the under-representation of the autistic population with ID (Russell et al., 2019).

Limitations

The findings of the systematic review should be considered in the context of several limitations. Primarily, even though this review aimed to summarise the most relevant studies on temporal synchrony in autism, the study selection criteria may reflect the author’s personal perspectives, approaches, expertise and training. The quality of the review conclusions is also dependent on the quality of the included studies. Some of those included had significant limitations, such as having very small sample sizes. The review also focused on peer-reviewed studies to help ensure a minimum standard; however, this meant that it was difficult to assess the extent to which publication bias might have been present. Finally, because the studies were judged too heterogeneous in terms of design with only small numbers of studies of each design, no quantitative synthesis was attempted. As this literature reaches a greater level of maturity and more primary studies are published with similar designs, a meta-analytic investigation with moderator analyses will be helpful for further investigating the extent of temporal asynchrony tendency in autistic individuals compared to neurotypical individuals and the person, task, and broader contextual influences on this.

Future Directions

The findings of the systematic review raise the possibility that studies on temporal synchrony in autism could provide new insights into understanding the social communication and sensory difficulties in autism. Temporal synchrony difficulties experienced by autistic individuals in everyday tasks and in more naturalistic settings such as speaking back-and-forth on the telephone without visual cues, engaging in "flowing" one-to-one in-person conversation, and taking turns in social interactions appear to be examples of novel areas for further study. Further, it could be investigated whether temporal synchrony tendencies mediate the association between meeting diagnostic criteria for autism (or showing high levels of symptoms) and everyday differences and challenges experienced relative to typically developing individuals.

Intervention studies targeting improvement on sensory integration and functioning in autism might have effects on the social-communication experiences of autistic individuals: this is an area of potentially significant innovation for impactful interventions and should be investigated further. Likewise, a better understanding of temporal synchrony differences could help communication between typically developing and autistic individuals and inform the adaptation of work and social environments to better accommodate the latter.

Finally, temporally extended body representation in autism in visuo-tactile synchrony studies needs to be investigated further by administering other experimental paradigms aside from rubber hand illusions. This might allow for a better understanding of temporally asynchronous behaviours in autism overall.

Conclusions

This review provided evidence that temporally asynchronous behaviour and atypical temporal processing are evident in autism when autistic individuals are expected to integrate audio-visual, audio-motor, visuo-tactile, visuo-motor, social motor, and conversational information. Results from the included intervention studies also highlight the possibility of improving socially and temporally synchronous behaviours in autistic individuals.

While the cause of temporal asynchrony has not yet been investigated, it may be appropriate to hypothesise that temporal processing and multisensory integration difficulties might cause temporally asynchronous behaviours in autistic individuals, and therefore may contribute to social-communication and sensory difficulties in autism.

It is evident that the clinical representation of autism and the experiences of autistic people is more diverse than the concepts of ‘social-communication and interaction, and restricted and fixated behaviours, interests and activities’ (APA, 2013). Therefore, it is essential to move away from the ‘Core-Deficit Hypothesis’ in research and clinical practice, and document alternative profiles of the condition. Further understanding of the mechanisms and outcomes of temporal synchrony may help improve clinical practice and develop new biomarkers and successful interventions to improve possible temporal asynchrony related sensory and social-communication difficulties of autistic individuals.