Mental health symptomology is more common in CYP with ASD compared to the general population (8–10). In the United States, nearly 78% of children with ASD have a comorbid mental health condition [Kerns et al. (11)]; anxiety is one of the most common of these, with ~50% of individuals with ASD experiencing clinically significant symptoms of anxiety that impact on daily life (12). Further, it is estimated that 84% of CYP with ASD experience sub-clinical symptoms of anxiety (10), which typically warrant intervention and are predictors of increasing severity (13).

There are several possible explanations for the high rates of anxiety observed among CYP with ASD. First is the interaction between deficits related to ASD, such as social communication difficulties and environmental factors (e.g., parenting and peer relationships), resulting in increased vulnerability to anxiety (14). Second is emotional regulation (ER) which is described as a person’s ability to effectively manage and respond to an emotional experience (15). Third, the cognitive mechanism of Intolerance of Uncertainty (IU), characterized by the inability to cope with unfamiliar or uncertain situations (16), is commonly observed among CYP with ASD. The presence of ER and IU offer an explanation for anxiety in the younger ASD population, as a multitude of changes and unpredictability occur during childhood and adolescent years which are often beyond their coping abilities (10, 17).

Childhood and adolescence are a crucial time for the development and acquisition of social skills and behavioral regulation (18). These skills – including social communication, reciprocity, engagement, willingness to interact with others, and ability to form meaningful relationships – are particularly challenging for CYP with ASD (2). This contributes to feelings of social isolation and loneliness, both of which have been identified as common experiences of CYP with ASD (19), who often struggle with the unpredictability of everyday social situations, resulting in maladaptive behavioral responses to social interactions (20). As such, adapting to the changes experienced throughout development, such as the transition from childhood to adolescence, can be especially challenging for CYP with ASD (21). Yet, while CYP with ASD struggle to develop and maintain interpersonal relationships, their desire for these relationships is similar to the general population (22).

CYP with ASD often have difficulty effectively regulating or expressing their emotions (23) which manifests in aggressive behaviors (24). Among individuals with developmental disorders, aggressive behaviors are one of the strongest predictors of the need for crisis intervention, admission to residential care facilities, and the prescription of psychotropic medications (24). These behaviors become particularly challenging as children age and the potential severity of risk of harm to self and others increases.

Mindfulness – the practice of maintaining awareness of the present moment, while being conscious and attentive to this awareness, in a non-judgemental way (25) – is moderately effective in improving symptoms of anxiety and depression (26) and reducing emotional dysregulation by increasing emotional awareness and acceptance (27) for adults without ASD. Mindfulness-based interventions also improve stress, anxiety, depression, and cognitive performance among CYP without ASD (28, 29). Recently, a meta-analysis of mindfulness-based interventions for individuals with ASD and their caregivers found significant improvements in subjective wellbeing for all participants (30). Yet, only four systematic reviews exist of mindfulness-based interventions for CYP with ASD (31–34) and all have limitations that we aim to address in the current systematic review, as outlined below (35).

First, a systematic review conducted by Cachia et al. (32) examined the efficacy of mindfulness-based interventions on psychological, social, and behavioral outcomes in CYP and adults with ASD. This review included six studies published prior to February 2016 (32). Findings suggested the interventions were efficacious in improving a range of psychological, social, and behavioral problems for individuals with ASD (32). Only six children and 29 adolescents were included in the total 161 participants in the review conducted by Cachia et al. (32). As most of the participants included were adults (N = 126), only two of the interventions reviewed by Cachia et al.(32)were tailored to the specific needs of CYP. Further, this review had restrictive definitions of mindfulness-based interventions. Specifically, authors omitted “yoga” and “mind–body therapy,” which are holistic interventions that include mindfulness. These movement-based mindfulness interventions may be more efficacious for CYP with ASD due to a lack of dependence on cognitive ability, a limitation for this population in other interventions, such as Cognitive Behavior Therapy (35).

Second, Hourston and Atchley (33) conducted an exploratory systematic review to investigate the types of mind–body therapies used for individuals with ASD and the outcomes which are targeted (33). Consistent with findings reported by Cachia et al. (32), of the 16 studies included in the review, improvements in mental health were reported in multiple studies (33). While many of the studies reported a decrease in aggressive behaviors, resulting from a lack of controlled trials and lengthy follow-up periods, authors were unable to draw any conclusions on the efficacy of the interventions for behavioral outcomes (33). There were no criteria for the outcomes included in the systematic review, and as such, there was considerable variance in the outcomes of the studies. Further, participants of all ages were included, meaning intervention studies targeting adults with ASD and the parents and caregivers of CYP with ASD were also included in this review.

Third, unlike the reviews conducted by Cachia et al. (32) and Hourston and Atchley (33), Semple (34) focused specifically on interventions targeting CYP with ASD. Semple (34) conducted a systematic review of the effects of mindfulness-based interventions on core symptoms of ASD. Social skills and aggressive behaviors are captured within diagnostic criteria for ASD, meaning these outcomes were included in the review. Improvements were observed across all eight studies included in the review, providing promising evidence that mindfulness-based interventions may be efficacious in improving prosocial behaviors among CYP with ASD (34). The impacts of mindfulness-based interventions upon psychological wellbeing, particularly levels of anxiety, were not considered in this review (34).

Last, (31) conducted a systematic review assessing the effect of Yoga for musculoskeletal function, cardiovascular function, neurological function, and behaviors of CYP with ASD. Only two of the 18 studies in this review included outcome measures of communication and social interaction. While positive results were found in both studies, authors state there is insufficient research with relevant outcome measures to determine the effects of Yoga for improving behaviors of CYP with ASD (31).

In summary, only three systematic reviews of mindfulness-based interventions for CYP with ASD exist. None captures the effects of mindfulness-based interventions on anxiety, social behaviors, and aggressive behaviors in CYP with ASD. Additional limitations exist surrounding the sample population age range and/or the definition of mindfulness-based intervention. All authors emphasized the limited and often poor-quality research with individuals with ASD and the importance of further research into mindfulness-based interventions for this population (31–34). Furthermore, since the publication of these reviews, mindfulness-based interventions have grown in popularity and are more commonly used in psychotherapy (36). Together, this highlights the need for an updated review.

The purpose of the current systematic review is to answer the question: Does mindfulness-based therapy – including yoga, mind–body, and mindful-movement based interventions – improve anxiety, social behaviors, and aggressive behaviors in behavior CYP aged 6–25 years with ASD? Doing so will provide a novel understanding of the efficacy of mindfulness-based interventions specifically for CYP with ASD, focusing upon outcome measures related to anxiety, social skills, and aggressive behaviors. A risk of bias analysis will inform whether the methodological quality of the research has improved since the reviews conducted by Cachia et al. (32), Hourston and Atchley (33), Artchoudane (31), and Semple (34). Further, the review will summarize the results across clinic, home, and school contexts, and evaluate the quality of the studies for clinical practice. Assessing these differences provides an opportunity to draw general conclusions about the efficacy of mindfulness-based interventions for CYP with ASD across a range of different settings.

For the purpose of the current review, “young people” are defined as those aged between 6 and 25 years based upon the Australian Department of Health definition of “young adults” (Support for Young Adults (Ages 18–25), n.d.). Participants under the age of six were not included in the review as children younger than this are unlikely to participate effectively in a mindfulness-based intervention. In addition, research on mindfulness interventions for children commonly focuses on ages six and above [Weare (37)]. CYP with a diagnosis of ASD, Pervasive Developmental Disorder (PDD) or Asperger Syndrome were included, in accordance with the DSM-IV and DSM-5 diagnostic criteria (2, 38) and/or ICD-10 (39).

Mindfulness-based interventions in this review included both mindfulness-only interventions as well as mindful-movement/mind–body therapies such as yoga that contain significant mindfulness components. Interventions targeting the parent or caregiver were not included.,

Studies were not required to have a control or comparison group in order to be included in the review.

The three outcomes of interest were anxiety, social skills, and aggressive behaviors. Studies were included if they had a measure of anxiety, social skills, such as increased social responsiveness, or aggressive behaviors, such as physical outbursts. Outcomes can be measured through interviews and a range of standardized questionnaires, including the Social Responsiveness Scale [SRS; (40)] and Child Behavior Checklist [CBCL; (41)], among other measures. Studies with an outcome not relevant to anxiety, social skills or aggressive behaviors were excluded.

Both qualitative and quantitative research which implemented a mindfulness-based intervention across any context (i.e., home, clinic, school) were included in the review. Relevant reviews and meta-analyses were consulted as sources; however, they were not included in the analysis. Grey literature, including non-peer reviewed articles and unpublished studies, were excluded from the current review as this data is more likely to be inaccurate or incomplete [Benzies et al. (42)].

A total of 334 articles were screened and 56 full-text articles were retrieved for further review. Of these, 23 articles were included in the larger systematic review, to evaluate the efficacy of mindfulness-based interventions for improving psychological well-being, anxiety, social skills, and aggressive behaviors in CYP with ASD. All 23 studies were included in the current review, examining the effects on anxiety, social skills, and aggressive behaviors. The articles were peer reviewed and published between 2010 and 2021. The study selection and screening process is available to review in the PRISMA flow chart (Figure 1).

Five studies had specific outcomes relating to anxiety, with varying results in improvement following the intervention. There was one qualitative report on a reduction in anxiety post-intervention (49). Conner et al. (51) found a significant reduction in anxiety in parent-reported measures, but no significant changes in self-reports of anxiety. Tanksale et al. (66) found a significant reduction in self-reported performance anxiety, however there were no significant reductions in other sub-types of anxiety. Two studies (52, 67) report no improvements in anxiety symptoms post-intervention. Two of the five studies measuring anxiety included a follow-up analysis ranging from 6 to 9-weeks post intervention (66, 67). The post-intervention improvements found by Tanksale et al. (66) were retained at the time of follow-up.

For outcomes relating to anxiety symptoms an effect size of 0.59, indicating a medium effect size (68), was reported by Conner et al. (51). De Bruin et al. (67) and Tanksale et al. (66) reported effect sizes on the relevant outcome measures to be <0.40. Hartmann et al. (52) did not report effect sizes for anxiety outcomes. Due to the qualitative design of the Black and Rosenthal (49) study, effect sizes were not applicable.

Out of 14 studies, half found an improvement in social skills (17, 50, 52, 56, 59, 65, 67). Scroggins et al. (60) found improvements in social skills; however, the significance values were not reported as the research was a single subject case study. Black and Rosenthal (49) and Ridderinkhof et al. (58) conducted qualitative research, with improvements in social skills observed in both studies. Black and Rosenthal (49) interviewed young people who had undergone a transcendental meditation intervention. Three out of the six participants reported improved social skills (49). Of the remaining four studies focussing on social skills, two found no significant improvement in the treatment group (46, 55). The third found comparable improvement in social skills, measured using the SRS-2, in both the treatment and control group (52). Radhakrishna (57) also found an improvement in social skills, however not at a significant level. Five of the studies measuring social skills included a follow-up analysis, ranging from one to 4 months post-intervention (17, 52, 56, 59, 67). The improvements in social skills observed at the conclusion of the intervention were retained, or increased, at the time of follow-up.

For social skills outcomes Cohen’s d effect sizes of 0.68 and 0.40 respectively, indicating a medium to large effect size (69), for the relevant outcome measures were reported by Chan et al. (50) and de Bruin et al. (67). Pahnke et al. (56) expressed effect sizes as partial eta-square with 0.01 reflecting a small effect, 0.06 moderate, and 0.14 a large effect size (69). Pahnke et al. (56) reported a moderate effect size for improvements in social skills (ɲp² = 0.06). Ho et al. (53), Ridderinkhof et al. (17) and Salem-Guirgis et al. (59) reported effect sizes on the relevant outcome measures to be <0.40. Effect sizes were not applicable to three of the studies as two were qualitative studies (49, 58) and one was a single subject case study with no standardized outcome measures (60). The remaining 11 studies did not report the effect sizes for social skills.

Outcomes measuring aggressive behaviors were examined in 14 of the 23 studies in this review (17, 46–54, 61–64). Of these, seven studies reported a significant decrease in aggressive behaviors following a mindfulness-based intervention (17, 47, 48, 50, 51, 54, 64). In Black and Rosenthal’s (49) qualitative study, two of the six participants reported a reduction in behaviors of aggression. Singh et al. (61–63) also found a reduction in aggressive behaviors, however the significance level was not reported in the results of these studies. Four studies measuring aggressive behaviors conducted a follow-up analysis ranging from one to 4 years post-intervention (48, 61–63). All four studies found that the effect of the intervention, a reduction in aggressive behaviors, had been retained or improved at follow-up.

Effect sizes >0.40 for outcome measures related to aggressive behaviors were found for the following two studies: 0.66 (50), and 0.76 (51). Ahemaitijiang et al. (48) reported effect size (−0.96) using the Tau-U statistic, where 0.80 or above indicates a very large effect (70). Ridderinkhof et al. (17) reported effect sizes <0.40 on relevant outcome measures. Effect sizes were not applicable to Black and Rosenthal (49) as this was a qualitative study. The remaining nine articles did not report effect sizes for aggressive behaviors.

The risk of bias tool was used to determine the methodological quality of the studies included in this review. An overall rating of a study’s methodological quality, either strong, adequate, or weak, was then awarded to each study (43, 44). The results of the quality analysis are further detailed below.

The overall methodological quality of a research article is assessed by the criteria outlined by Reichow et al. (43) and Reichow (44), including both the primary and secondary quality indicators. As reflected in Table 6, of the 23 articles included in the current review, four were found to be of strong methodological quality, five to be adequate quality, and 14 to be weak quality.

This review included studies which incorporated mindfulness-based CBT interventions like Acceptance Commitment Therapy (ACT), Dialectical Behavior Therapy (DBT), and Emotion Awareness and Skills Enhancement Program (EASE), however, their use obscures which intervention component(s) led to significant effects. There was also substantial variance in the mindfulness techniques employed in the studies, precluding recommendations for a specific type of mindfulness for CYP with ASD and making it difficult to ascertain which aspects of the mindfulness-based interventions were most successful.