First Year Inventory Attention Constructs in Infants at Risk for Autism Spectrum Disorders

Background: Atypical patterns of attention are thought to be early predictors of autism spectrum disorders (ASD; Chawarska et al., 2016; Keehn et al., 2013). Parent report of attentional patterns could provide crucial information in early identification of ASD. One screener, the First Year Inventory (FYI; Baranek et al., 2003) shows promise in identifying 12-month-olds at risk for later ASD diagnosis (Reznick et al., 2007; Rowberry et al., 2014). The FYI may also be valuable in evaluating early attentional behaviors. Stephens et al. (2017) developed three new scores for the FYI: Responding to Social Attention (RSA), Initiating Social Attention (ISA), and Nonsocial Sensory Attention (NSA). These attention constructs predicted general social responsiveness at 3 years, but it is unclear how well the constructs predict diagnostic status or other aspects of development.

Objectives: To investigate how the attention constructs of the FYI relate to developmental outcomes at 24 months in infants at high and low risk for ASD.

Methods: Parents of 102 infants with older siblings diagnosed with ASD (HR; 70 male) and 63 infants with no family history of ASD (LR; 32 male) completed the FYI around their child’s first birthday. At 24 months, children completed the Mullen Scales of Early Learning (MSEL), ADOS-T, and Vineland-II. Based on outcome diagnoses at 24/36 months of age, participants were categorized as: ASD (n=18), LR typically-developing (LR-TYP; n=50), HR typically-developing (HR-TYP; n=37), and children with other developmental delays (ATYP; n=60).

Results: One-way ANOVAs revealed significant effects of diagnostic group in RSA, F(3,161)=5.92, p<.01, and ISA, F(3,161)=5.79, p<.01, but not in NSA scores. Post-hoc Tukey HSD tests revealed that for RSA, LR-TYP children scored lower than children with ASD or ATYP, both p<.05 (Figure 1). For ISA, LR-TYP children scored lower than ASD, ATYP, and HR-TYP children, all p<.05. Multiple regressions showed that attention constructs from the FYI at 12 months predicted the following at 24 months: MSEL Verbal Developmental Quotient (VDQ), F(3,161)=6.46, p<.01, R²=.11; MSEL Nonverbal Developmental Quotient (NVDQ), F(3,161)=2.77, p=.04, R²=.05; Vineland Communication SS, F(3,151)=6.60, p<.01, R²=.12; Vineland Daily Living SS, F(3,151)=3.67, p=.01, R²=.07; and Vineland Social SS, F(3,151)=6.09, p<.01, R²=.11. RSA was a significant predictor in the Vineland-II Communication model; ISA was a significant predictor in the MSEL VDQ and Vineland-II Daily Living and Social models; and NSA was a significant predictor in the MSEL VDQ model, p<.05 and a marginally significant predictor in the MSEL NVDQ model, p=.06. The attention constructs did not predict ADOS-T total scores or Vineland-II Motor SS, p>.28.

Conclusions: Children later diagnosed with ASD and other developmental delays scored more atypically on two of the three FYI attentional constructs (RSA and ISA) than LR typically-developing children. Furthermore, the constructs differentially predicted the cognitive ability and adaptive behavior of the children one year later. This suggests that parent-reported information via the new attentional constructs of the FYI show promise not only in identifying children who are at risk for developing ASD, but also predicting later vulnerabilities in verbal and nonverbal development and adaptive behavior.