Exploring Relationships Between Negative Cognitions and Anxiety Symptoms in Youth with Autism Spectrum Disorder

Background:  Negative cognitions (NC) are central to the conceptualization and treatment of anxiety disorders in youth without ASD (Beck & Rush, 1985; Davis et al., 2011) and demonstrate relationships with anxiety symptoms, intolerance of uncertainty (IU), and worry (McEvoy et al., 2014). In ASD, meta- and socio-cognitive impairments raise questions about the role of NC in the phenotype of anxiety, and modified cognitive-behavioral interventions (MCBT) often place less emphasis on addressing NC. Yet, two recent studies found that youth with ASD reported more NC than anxious and non-anxious youth without ASD, and NC predicted severity of anxiety, depression, and disruptive behaviors (Farrugia et al., 2006; Ozsivadjian et al., 2013). These findings, coupled with data showing that about one third of youth with ASD do not respond to MCBT (Vasa et al., 2014), suggest that further research on the relationship between NC and anxiety is needed to guide interventions.

Objectives:  The objectives of this study are: 1) To examine relationships between negative cognitions (NC) and anxiety symptoms, worry, and intolerance of uncertainty (IU) (Aim 1), and 2) To assess predictors of negative cognitions (NC) in youth with anxiety and ASD (Aim 2).

Methods:  Forty-one children and adolescents with ASD, ages 8 to 14 years participated in the Face Your Fears MCBT group therapy program as part of a multi-site treatment study (Reaven et al., 2011). The diagnoses of ASD and anxiety were established using standardized instruments. NC were assessed using the Children’s Automatic Thoughts Scale (CATS; Schniering & Rapee, 2002), a self-report measure comprised of four sub-categories: physical threat, social threat, personal failure, and hostile intent. Children also completed measures of IU (IUS-C), anxiety (SCARED), and worry (PSWQ-C) preceding the intervention. Spearman Rho correlations were conducted to examine Aim 1. Multiple linear regressions were run to examine Aim 2.

Results:  Preliminary analyses indicated moderate to strong relationships between NC and anxiety symptoms (Total NC: r = 0.56, p < 0.001; NC sub-categories: r = 0.33 - 0.64, p < 0.05) and IU (Total NC, r = 0.52, p < 0.001; NC sub-categories r = 0.34 - 0.54, p < 0.05). All NC sub-categories (r = 0.44 - 0.51, p < 0.05) except hostile intent were associated with worry (Table 1) (Aim 1). Results of the regression analyses showed that anxiety and IU were each independently associated with NC (Table 2) (Aim 2). Age, gender, and IQ were not significant predictors of NC and were excluded from the model.

Conclusions:  Results support previous literature indicating a relationship between NC and anxiety symptoms in ASD youth. New findings from this study show that IU and worry are associated with NC, implicating NC as a significant component of anxiety in youth with ASD. IU may be key in the development of NC and may be a treatment target. Further analysis of interaction effects to differentiate relationships between NC and anxiety associated variables will be pursued. Relationships between specific NC (i.e., item-level analysis) and anxiety disorders, IU, and worry will be investigated.