An Exploratory Analysis of Autism and Reactive Attachment Symptoms in in-Care and at-Risk Children

Background: This study examines the factor structure of symptoms of Autism Spectrum Disorders (ASD) and Reactive Attachment Disorder (RAD) in a combined sample of foster care and community preschoolers. The etiology of ASD and RAD are notably different, with ASD being largely organic in nature, and RAD being rooted in negative early experiences with caregivers. Despite these differences, ASD and RAD symptoms are markedly similar, with both diagnoses characterized by difficulties with social interactions and communication with caregivers. Despite recognition of this overlap in clinical settings, (Moran, 2010), there has been little empirical research on the overlap of ASD and RAD symptoms. This research aims to identify how symptoms of ASD and RAD cluster, and to identify different profiles of children with ASD and RAD symptoms.

Methods: Participants were the primary caregivers of 177 children (ages 3-6; 117 in foster care and 60 SES-matched community children). Caregivers completed the Early Childhood Inventory-4, a questionnaire that measures the behavioral symptoms of DSM diagnoses relevant to preschool children including diagnostic symptoms of RAD and ASD.

Results: A principal components analysis with varimax rotation was conducted to determine the factor structure of all symptoms (i.e. 12 ASD and 2 RAD) when analyzed together. Four components with eigenvalues greater than 1 were extracted:  one representing social difficulties (33.12% variance), one representing “Aspergers” symptoms (9.62% variance), one representing communication difficulties (9.19% variance), and one representing RAD (7.93% variance).

Item loadings onto the four components were generally consistent with established ASD symptom clusters, however cross-loadings indicated symptom overlap. More specifically, two of the four ASD social difficulty items loaded strongly and exclusively onto the social difficulties component (.83, .77), while another loaded most highly onto the social difficulties component (.60), but also loaded onto the “Aspergers” (.34) and RAD components (-.35). The fourth social item loaded more evenly onto the social difficulty (.42), communication (.44), and RAD (.30) components.

The first RAD item, “afraid, distrustful with familiar people,” loaded exclusively onto the associated RAD component (.78), and the second, “seeking affection from people, even strangers,” loaded strongly on both the RAD (-.56) and communication (.38) components.

Results suggest that, while some ASD symptoms are specific to the disorder, others are shared by ASD and RAD. Likewise, while the RAD symptom of distrust of familiar people is unique to RAD, indiscriminant friendliness is also related to the ASD communication difficulty symptom cluster. While this variable-centered analytic approach provides useful preliminary information on the extent to which these symptoms relate, we will use an complementary approach, Latent Profile Analysis, to identify how individuals cluster according to their symptom presentation, thereby further clarifying the relationship between ASD and RAD symptoms.  Results of this pending analysis will also be reported.

Implications: Understanding the ways that ASD and RAD symptoms cluster has implications for correctly diagnosing children and identifying most effective treatments. This research is particularly relevant to children who have experienced multiple early life stressors and display a diagnostically-challenging array of social and behavioral challenges.