Abstract
<p>Attachment in children with autism spectrum disorder (ASD) has typically been assessed retrospectively after diagnosis (> 30 months), primarily using non-standardized protocols, making it difficult to interpret and compare across studies in normative and atypical samples. We assessed attachment security prospectively at 15 months in high-risk infants with later (3 year) ASD (High-Risk/ASD, n=16), and high- (High-Risk/No-ASD, n=40) and low-risk (Low-Risk/No-ASD, n=39) infants without later ASD- using the standard Strange Situation Procedure. High-Risk/ASD infants were disproportionately more likely to be classified as insecure (vs. secure, 2-way) and insecure-resistant (vs. secure vs. avoidant, 3-way) than High-Risk/No-ASD and Low-Risk/No-ASD (2-way, 56.3%, 15.0%, 20.5% respectively; 3-way, 37.5%, 7.5%, 10.3%, respectively). Compared to high- and low-risk infants without later ASD, high-risk infants with later ASD did not show less intense attachment behaviors with their parents, but were more likely to display insecure-resistant attachment. High-risk infants with insecure attachments were 7.28 times more likely to receive an ASD diagnosis than high-risk infants with secure attachments. As an index of early social-emotional functioning, attachment security in high-risk infants may serve as a potential behavioral marker and target for intervention.</p>