Are Restricted and Repetitive Behaviours Related to Executive Function Performance? a Correlational Meta-Analysis.

Background: A growing number of correlational studies have been undertaken to evaluate the proposed relationship between executive function (EF) skills and restricted and repetitive behaviours (RRBs) in young children (e.g. Tregay, 2009) and individuals with autistic spectrum disorder (ASD) (e.g. South et al., 2007). The EF theory is a strong contender for explaining RRBs across both populations, as it proposes that an inability to control behaviours does not facilitate the inhibition of RRBs. For example, this theory is capable of accounting for the age-related changes occurring in typical development around the age of four/five, when EF skills undergo a rapid development (e.g. Carlson, 2005) and RRBs decrease (e.g. Evans et al, 1997). Despite this, the evidence regarding the relationship between EF skills and RRB levels is mixed, as some studies have found associations with several EF components, such as set-shifting and inhibitory control (e.g. Lopez et al, 2005), whereas others have found that only set-shifting performance predicted RRBs (Van Eylen et al, 2014). The current meta-analysis will, therefore attempt to evaluate the relationship with a focus on set-shifting and inhibitory control measures, but also parental-reports. The parental-report measures will be included as a separate analysis, as although both types of tasks are commonly used to measure EF skills, Toplak et al (2013) found that they measured different levels of cognition, namely, cognitive abilities and goal pursuit achievement. There is now a need for a systematic attempt to evaluate the relationship across studies and between task batteries.

Objectives: This meta-analysis (following recommendations from Quintana, 2015) will allow us to examine the general EF hypothesis, but also identify whether the link between RRBs and EF skills is a general one, or if it is caused by one specific mechanism.

Methods: Following a comprehensive search in three databases, three random-effects analyses were run. The set-shifting analysis consisted of 20 studies (with 26 correlations) and a sample size of 869. The inhibitory control analysis comprised of 14 studies (with 17 correlations) and a sample size of 664. Finally, the parental-report analysis included 8 studies (with 12 correlations) and a sample size of 1362. Moderators such as age, diagnosis, EF task and type of RRB measure were investigated.

Results: The analyses found significant but moderate associations between RRBs and set-shifting performance (.27, 95% CI 0.17-0.41, p< 0.0001), inhibitory control performance (.20, 95% CI 0.02- 0.37, p 0.02) and scores on parental-report measures (.31, 95% CI 0.0736-0.5514, p <0.001). We found that for set-shifting tasks and parental-scores; the association remained stable with age, between diagnoses and across different types of EF and RRB measures. In contrast, the relationship between RRBs and inhibitory control performance was moderated by age and diagnosis.

Conclusions: This meta-analysis offers support for the EF hypothesis, and suggests that at least two cognitive mechanisms may underpin the high RRB levels in individuals with ASD, as well as in young children. The relationship between cognitive skills and behaviours may have clinical training implications so directions for future intervention research will be proposed.