Malleability of Comorbid Anxiety and ADHD Symptoms and Moderating Effects across the PEERS® Intervention for Autistic Adolescents

Background: Comorbid anxiety and ADHD symptoms are common among autistic adolescents, with rates of roughly 40% (van Steensel, Bögels, & Perrin, 2011) and 30-80% (Mattila et al., 2010), respectively. Additionally, each of these comorbid symptoms have been found to exacerbate social difficulties (Ames & White, 2011; Bellini, 2004). Social skills interventions may have secondary effects on anxiety (Corbett, Blain, Ioannou, & Balser, 2017; Lei, Sukhodolsky, Abdullahi, Braconnier, & Ventola, 2017; McVey et al., 2016; Schohl et al., 2014); secondary effects on ADHD are unknown. In one study of moderating effects, anxiety did not influence intervention response, but ADHD was associated with nonresponse (Antshel, Polacek, Dygert, Miller, & Faisal, 2011). More work is needed to better understand secondary effects on anxiety and ADHD symptoms and the potential moderating role of anxiety and ADHD symptoms on improvements via social skills intervention.

Objectives: The present study examined the malleability and moderating effects of anxiety and ADHD across the Program for the Education and Enrichment of Relational Skills (PEERS^®).

Methods: One-hundred and thirteen adolescents aged 11-16 with IQ ≥ 70 were included in this post hoc study. Autistic adolescents were randomly assigned to an experimental or a waitlist control group as part of a larger randomized controlled trial. Parents reported upon social behavior via the Social Responsiveness Scale (SRS; Constantino & Gruber, 2002). Anxiety and ADHD symptoms were measured via parent-report on the Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2001).

Results: Gender and IQ were included as covariates in the following analyses, given that groups differed on IQ (Table 1) and gender was related to the CBCL ADHD Problems subscale (r(99)=0.20, p<.05). Four mixed model MANCOVAs were conducted. Results indicated improvements in some SRS subscales for the experimental, but not the waitlist control group (Table 2). For anxiety, results revealed nonsignificant main and interaction effects of Group, Time, and Group*Time. Exploratory paired samples t tests revealed that both groups showed declines across time (EXP: t(41)=1.96, p=.056); WL: t(48)=2.91, p=.005; Table 2). Results of ADHD also revealed nonsignificant main and interaction effects of Group, Time, and Group*Time. Exploratory paired samples t tests similarly revealed that both groups declined across time (EXP: t(41)=2.29, p=.027); WL: t(48)=4.97, p<.001; Table 2). Results examining the possible moderating effects of anxiety, ADHD, both, or neither, revealed a nonsignificant Group*Comorbidity*Time interaction (p>.05).

Conclusions: Some changes in social behavior were evident for the experimental but not the waitlist control group, consistent with prior literature (Laugeson, Frankel, Gantman, Dillon, & Mogil, 2012; Schohl et al., 2014). Declines in anxiety and ADHD were seen across the intervention for the experimental group, yet this pattern was also observed in the waitlist control group. Contrary to expectations, the presence of clinically-significant anxiety and/or ADHD symptoms did not moderate improvements in social behavior. Importantly, this evidence may suggest that, while it is unclear if PEERS^® affects anxiety and ADHD symptoms, the integrity of the intervention is upheld for autistic adolescents with clinical levels of anxiety, ADHD, or both.