24958
Are Longitudinal Associations Between Joint Attention and Language Attributable to Nonsocial Attention in Infancy? an Infant Sibling Study
Longitudinal associations between joint attention (JA) and language commonly documented among children with and without autism (Brooks & Meltzoff, 2005; Sigman and Ruskin, 1999) are often interpreted as evidence that both JA and language reflect social understanding (Charman et al., 2000). Keehn and colleagues (2013) hypothesized that atypical non-social attention may underlie impaired JA and contribute to language difficulties among autistic children. Positive associations between the duration of non-social attention in infancy and later language have been demonstrated among typically developing (TD) children (Kannass & Oakes, 2008). Longitudinal associations between social attention, JA, and language have also been documented among TD children (Salley et al., 2013).
Objectives:
To investigate if 12-month nonsocial or social attention is associated with 18-month JA and examine if nonsocial attention accounts for associations between JA and language.
Methods:
Participants with (n = 65) or without (n = 41) an autistic sibling were classified as autistic (n = 9), other concerns (n =17), or TD (n = 80) by a clinician at 36 months or later. At 12 and 18 months, RJA and IJA were assessed with the ESCS. Examiners also caught participants’ attention and shook two rattles for approximately 15 seconds. The rate of gaze shifts between rattles (nonsocial attention) and the duration of attention to the examiner (social attention) were coded. The MSEL were administered at 18, 24, and 36 months.
Results:
Nonsocial and social attention at 12 and 18 months were not correlated with 18-month IJA or RJA (ps > .07). We ran repeated measures general linear models with raw expressive and receptive language at 18, 24 and 36 months as dependent variables and outcome, 18-month RJA and IJA as predictors. We then added nonsocial and social attention at 12 and then 18-months to each model. IJA, RJA and outcome were associated with overall expressive language (ps< .05). Outcome predicted changes in expressive language (p< .001). When 12-month attention was entered, nonsocial attention (fewer gaze shifts), IJA and outcome were associated with overall expressive language (ps< .05) while RJA and social attention were not (ps> .12). When 18-month attention replaced 12-month attention, only nonsocial attention and outcome were associated with overall expressive language (ps < .01); RJA, IJA, and social attention were unrelated (ps> .12). RJA, IJA and outcome were initially associated with overall receptive language (ps< .02). When 12-month attention was entered, nonsocial attention, IJA, and RJA (ps< .02) were associated with overall receptive language; social attention and outcome were unrelated (p> .06). When 18-month attention was entered, RJA, nonsocial attention and outcome were associated with overall receptive language (ps < .05) while IJA and social attention were not (ps > .06).
Conclusions:
These findings demonstrate that reduced distractibility contributes to commonly observed associations between JA and language among children with and without autism. However, nonsocial attention was unrelated to JA and some associations between JA and language remained apparent when nonsocial attention was accounted for. Nonsocial attention is not a clear precursor for JA but both contribute to language development.