Characteristics of Toddlers with Autism Spectrum Disorder Are Associated with the Severity of Anxiety Symptoms at Adolescence: A Long-Term Follow-up Study

Background: Anxiety is among the most common mental health problems in autism spectrum disorder (ASD). Up to 40% of individuals with ASD are diagnosed with at least one anxiety disorder at some point in their lives. Previous studies reported that anxiety in ASD negatively impacts functioning. However, research has not focused on which early signs in toddlers with ASD may predict later development of anxiety.

Objectives:

1. To examine the prevalence of anxiety subtypes in adolescents who were diagnosed as toddlers with ASD.
2. To compare parental evaluation of the severity of anxiety subtype symptoms with the self-perception of adolescents with ASD.
3. To search for predictors of social and separation anxiety severity in adolescence at the time of ASD diagnosis.

Methods: The study included 65 participants, 60 males and 5 females (mean age=13:8y), diagnosed with ASD at toddlerhood. Participants underwent a comprehensive assessment of cognitive ability, adaptive skills and autism severity at toddlerhood and adolescence. The severity of anxiety was assessed by the Screen for Child Anxiety Related Disorders (SCARED) completed by parents (n=61) and study participants (n=37).

Results: According to parental SCARED reports, 39.3% of the study population had significant (≥5) separation anxiety symptoms, 27.7%, social anxiety (≥8), 18% generalized anxiety (≥9), 6.6% panic/somatic complaints (≥7) and 3.3% school avoidance (≥3). Comparing parents' and adolescents' SCARED reports (n=37) revealed that the participants noted more severe symptoms than their parents for panic/somatic complaints, separation anxiety, and school avoidance. We then examined the correlations of SCARED subdomain scores at follow-up with cognitive ability, autism severity and adaptive skills at toddlerhood. The SCARED social anxiety subdomain correlated negatively and significantly with VABS communication scores (r=-.41, p<.01) at diagnosis; SCARED separation anxiety correlated negatively and significantly with DQ (r=-.27, p<.05) scores, and positively and significantly with ADOS-RRB-CSS (r=.26, p<.05); SCARED generalized anxiety correlated positively and significantly with VABS socialization (r=.28, p<.05) and DQ (r=.30, p<.01) scores and negatively and significantly with ADI-R social interaction and communication subdomain (r=-.24,p<.05) scores. Next, we looked for predictors at toddlerhood for the two most frequent anxiety subdomains, social and separation anxiety at adolescence, using two hierarchical regression analyses. Age, sex, DQ, VABS composite and ADI-R subdomain scores at toddlerhood were used as the independent variables. For SCARED separation anxiety symptoms, the model explained 25.6% of the variance, but only the DQ scores at baseline correlated negatively and significantly with separation anxiety scores (β=-.42, p<.01). For SCARED social anxiety symptoms, the model explained 19.7% of the variance, but only the VABS composite scores at baseline correlated negatively and significantly with social anxiety scores (β=-.36, p<.05).

Conclusions: In adolescents with ASD, separation, social, and generalized anxiety symptoms are highly prevalent. Adolescents with ASD have a more severe perception of their anxiety symptoms than their parents’ evaluation. Toddlers with higher cognitive ability or poorer functioning are at increased risk for anxiety at adolescence. In light of these findings, it is important to assess anxiety symptoms in adolescents with ASD, including self-reports when available, in order to provide timely interventions.