The Utility of Psychophysiological Approaches in Assessing the Validity of Co-Morbid Anxiety in Autism Spectrum Disorders

Background: According to the literature, individuals with autism spectrum disorders (ASD) experience high rates of comorbid psychopathology, particularly anxiety disorders. No validated measures of associated psychiatric symptoms exist for the ASD population. Psychophysiological approaches offer promise for differentiating symptoms of anxiety from core autism symptoms among individuals with ASD.

Objectives: To describe potentially useful psychophysiological approaches for detecting anxiety in youth with ASD. Data from two studies piloting reliable physiological measures of anxiety with youth with ASD will be presented. One approach, startle response, will be discussed given its relation to amygdala function and implications for the pathogenesis of both autism and anxiety. 

Methods: In Study 1, 20 high-functioning adolescents with ASD and 19 typically developing (TD) adolescents participated in a fear potentiated startle (FPS) paradigm. Measures of social (Social Skills Rating Scale; Social Responsiveness Scale) and psychiatric (Revised Children’s Manifest Anxiety Scale; RCMAS and Child Behavior Checklist; CBCL) symptoms were administered to teens and their parents; eyeblink magnitude and latency (electromyographic activity; EMG) was measured during the FPS paradigm. In Study 2, twenty participants with ASD, ages 7-14, have completed a startle paradigm, during which EMG, skin conductance, and heart rate is collected. Participants also complete measures of anxiety (e.g., Pediatric Anxiety Rating Scale; PARS and Multidimensional Anxiety Scale for Children; MASC).

Results: In Study 1, parents of teens with ASD reported significantly higher rates of associated psychopathology compared to TD peers (CBCL Internalizing: t(35) = 8.19, p < .001; CBCL Anxious/Depressed: t(35) = 4.17, p < .001). Both groups demonstrated normal potentiated startle response, evidenced by larger response to the Threat versus Safe condition (t(893) = 5.80, p < .001). Social impairment was unrelated to startle response. In Study 2, convergent validity was examined through testing the association between physiological arousal and manifest anxiety symptoms; preliminary findings indicate that baseline skin conductance is significantly related to a total score on the PARS (Overall Severity of Avoidance of Anxiety-Provoking situations; r(7) = .688, p < .05), and marginally correlated with the MASC Somatic/Autonomic parent rating (r(10) = .536, p = .072). Convergent validity between heart rate, EMG, and behavioral measures of anxiety will also be examined in a larger group of children expected to be enrolled within the next 5 months (N = 40).

Conclusions: Results indicate that individuals with ASD demonstrate normal startle response, suggesting this aspect of amygdala function is intact (and unrelated to social deficits associated with ASD). Further, physiological reactivity may be related to manifest anxiety symptoms, which mirrors patterns observed in the TD population. These results support the notion that anxiety symptoms in youth with ASD are mediated by analogous physiological mechanisms as those observed in TD children; as such, a subset of individuals on the autism spectrum may experience anxiety as symptoms that are distinct from core ASD symptoms. Additional approaches, including examination of cortisol levels for participants in Study 2, will be presented as potential approaches for further differentiating anxiety from core symptoms in ASD.