Changes in Autistic Social Impairment with Treatment: Exploration of SRS Treatment Scales

Background:  Social deficits are a defining feature of Autism Spectrum Disorders (ASD), but research on the efficacy of interventions targeting improved social competence has yielded inconsistent findings. Improvement is variable across individuals, and long-term gain is often not realized (e.g., Bellini et al., 2007; Rao et al., 2008). Anxiety, common in ASD (e.g., White et al., 2009), is one individual variable that may limit treatment response if unaddressed.  For instance, anxiety may lessen ability to focus and inhibit learning, diminish opportunities to practice newly learned skills, or decrease the fluency of mastered skills in social situations.  Reducing anxiety while treating social skill problems might improve treatment response.

Objectives:  This study evaluated change in social competence across specific ASD-related domains in an intervention targeting both anxiety reduction and social skill improvement.

Methods:  Adolescents (n = 30) with a confirmed ASD and concomitant anxiety disorders participated in a randomized controlled trial (RCT) comparing a structured individual and group therapy program compared to waitlist. The purpose of the trial was to evaluate feasibility and preliminary efficacy of the combined psychosocial treatment program.  This report presents results on the Social Responsiveness Scale (SRS; Constantino & Gruber, 2005) over the course of treatment and at three-month follow-up.

Results:  The sample was predominantly Caucasian (87%) and male (77%), with estimated Verbal IQ ranging from 73 to 126 (M = 97.07). The SRS demonstrated excellent internal consistency (alpha = .93) in this sample. Baseline Total SRS scores were correlated, non-significantly, in the expected direction with theoretically similar constructs: r = .29  (p = .12) with ADOS algorithm scores, r = -.31 (p = .09) with Vineland Socialization domain score. Participants assigned to the combined psychosocial intervention (n = 15) had significant improvement in all SRS subscales compared to no change in the wait-list comparison group (n = 15).  Within the active treatment group effect sizes ranged from 0.73 (Social Motivation domain) to 1.18 (Autistic Mannerisms).  Follow-up data are available on nine of the 15 participants assigned to active treatment. No significant deterioration is apparent across the SRS subscales or total score between endpoint and three-month follow-up, indicating that treatment gains largely persisted after the intervention ended.

Conclusions:  The SRS appears to be a useful measure of social impairment in adolescents with ASD and these results indicate that it is sensitive to change with treatment. Addressing individual factors, such as anxiety, may be a useful consideration in treatments targeting improved social competence in adolescents with ASD, though such remains to be clearly demonstrated. Estimated effect sizes, on the SRS, in this trial are larger than the effects reported in recent clinical trials of interventions that did not also target anxiety reduction (e.g., Lopata et al., 2010; White et al., 2010). The next step will be to directly compare a ‘pure’ social skill development intervention to a combined approach.