Examining Early Developmental Trajectories for Children with and without Parent-Reported Skill Regression

Background: Roughly 20 - 30% of parents of children with ASD report that their child experienced a loss of previously acquired language skills (Jones & Campbell, 2010; Meilleur & Fombonne, 2009). Findings have been somewhat inconsistent as to whether developmental regression in ASD is associated with different autism symptomatology or skill profiles as compared to children with non-regressive ASD (Jones & Campbell, 2010; Meilleur & Fombonne, 2009; Rogers, 2004). Much research in this area has relied on parent reports made after early childhood, which can lead to parents reporting developmental milestones being achieved at later ages than they actually occurred (Lord et al., 2004; Luyster et al., 2005). Therefore, there is a need for further investigation of developmental regression in ASD utilizing parent reports during early childhood and examining longitudinal associations between different types of skill loss and developmental outcomes.

Objectives: The objective of this study was to examine parent-reported developmental regression on the ADI-R Toddler version and its association with language skills, nonverbal cognition, and autism severity at ages two and five.

Methods: Participants were 114 children (mean age at Time 1 = 31 months) with ASD who are part of a larger study of early language development. Best estimate diagnoses were determined at Time 1 using comprehensive diagnostic evaluations that included the ADOS or ADOS-Toddler and a Toddler version of the ADI-R. Measures used in the current study (ADOS; Mullen; Preschool Language Scale-4^th Edition) were drawn from Time 1 for all participants and, for a subset of children, a subsequent visit three years later (mean age at Time 2 = 68 months).

Results: 27% of participants had a parent-reported language loss of 3 or more words for at least 1 month. Approximately 20% had a clear loss of other skills (e.g., social engagement, play skills). Children with language loss spoke their first words at a significantly earlier age (mean = 15 months) than children with no language loss (mean = 21 months; t = 4.83, p <.001). Among children with language loss, the majority also lost vocal imitation skills and/or communicative intent. Among those who lost other skills, most children lost social engagement/responsiveness. Language or other skill regression did not predict performance on the Mullen, PLS-4 or ADOS at Time 1. Preliminary analyses indicate loss of communicative intent was associated with higher ADOS severity scores at Time 2.

Conclusions: Consistent with previous research, children with ASD who experience language regression begin talking earlier than their peers with ASD who do not have a loss. Although differences between the two groups in language skills, nonverbal cognition, and autism severity were not present at age 2, some differences may emerge by age 5. This research can improve our understanding of the significance of early regression for developmental trajectories in ASD.