Convergent Validity with Externalizing Behavior
Externalizing behavior was negatively correlated with RSA-R recorded during the social condition (r = -.673, p = .008). As RSA-R decreased, indicating withdrawal, externalizing symptomatology increased (see Figure 1). Externalizing behavior was not significantly related to RSA-R measured during the cognitive or emotional conditions (r s < 0.468, p s > .05). Descriptive statistics of RSA-R across domains and correlations with externalizing behavior can be found in Table 3.
Discussion
This study is the first to evaluate the feasibility of the MRP and examine domain-specific RSA-R in relation to externalizing behaviors in young autistic children. Physiological data collection was feasible for young children, with adequate completion and good to excellent stability of the cognitive, emotional, and social tasks, and lower completion rate and stability of the sensory task (Aim 1). Increased RSA withdrawal during a social challenge correlated with higher levels of externalizing behavior (Aim 2). Taken together, our findings highlight the feasibility of eliciting a reliable RSA response and underscore a potential relationship between PNS reactivity in social contexts and externalizing behaviors.
Our results support the use of the MRP. Acquisition rates suggestthat the social, cognitive, and emotional tasks are feasible in young autistic children with phrase speech or fluent language. Notably, fewer children wore sensors for the full sensory condition (Visit 1: 68%; Visit 2: 63%) compared with the other three tasks (79-84%). Given heightened sensory sensitivities in this population (Chistol et al., 2018), gustatory sensory tasks may not be sufficiently feasible to elicit a stable RSA signal. Nonetheless, additional task refinement to increase acquisition rates may provide essential information about sensory reactivity for a subgroup of young autistic children.
The negative relationship between externalizing behaviors and socially elicited RSA-R highlights the value of measuring domain-specific reactivity. Decreased RSA-R, or increased withdrawal, may reflect physiological dysregulation and limited coping responses to challenging situations (Graziano & Derefinko, 2013; Mezzacappa et al., 1997; Beauchaine, 2001). However, both RSA-withdrawal and reactivity may be adaptive depending on the context (Fenning et al. 2019). Indeed, 3.5-year-olds with elevated internalizing behaviors and moderate baseline RSA levels had fewer longitudinal externalizing problems than those with high or low baseline RSA (Ugarte et al. 2021). Characterizing physiological reactivity across domains could determine optimal domain-specific or domain-independent responses.
The specificity of the relation between externalizing behavior and RSA-R in the social domain replicates work in neurotypical individuals and may highlight the battery’s sensitivity to autism-specific regulatory demands. In a meta-analysis, only tasks using negative emotional reactivity RSA for non-autistic adults (Beauchaine et al., 2019). Although cognitive and emotional differences may be impacted by the heterogeneity of ASD, reduced social functioning is a core feature of autism. Consequently, engaging in face-to-face interactions may place the greatest self-regulatory demands on autistic children and best probe reactivity. Indeed, difficulty adapting to social norms and reading social cues has been linked to externalizing behaviors among autistic individuals (Shea et al., 2018). Therefore, socially induced RSA may reflect mechanisms linked with externalizing behaviors or better elicit negative emotional reactivity amongst autistic children.