Trajectory of Early Development: ASD, Broader Phenotype, Typical Development

Background: Prospective and retrospective studies indicate that language development is an area of vulnerability in children with autism, with trajectories characterized by regression in 30-50% of cases. Also, there is evidence for an intermediate phenotype associated with autism, where trajectory of language and social development appears slower than in typical development, but more robust than in children with autism (Landa et al., 2006). 

Objectives: To examine latent trajectories across multiple developmental systems in order to identify patterns of growth in children with and without autism.

Methods: Participants were assessed with the Mullen Scales of Early Learning at 6, 14, 18, 24, 30, and 36 months of age.  Dependent variables: Visual Reception, Fine and Gross Motor, Receptive and Expressive Language standard scores (Gross Motor norms not available at 36 months).  Outcome classifications at 36 months of age included: Autism Spectrum Disorder (ASD; n=59); Intermediate Phenotype (IP, involving language or social delay; n=36); Non-IP (n=121 siblings of children with autism; n=49 low risk controls). 

Latent growth curve models were used to study child-specific patterns of performance.   A 3-class model was favored.  Latent class membership was then related to clinical outcomes through regressions of latent classes on indicators for clinical diagnoses. The most likely latent class membership for an individual is based on posterior probabilities of being in a certain latent class, which suggested good separation between the classes.

Results: Class 1 consisted of an expected proportion of 60% of the sample, and was characterized by stable and normal development in all domains until 18 months of age, after which there was substantial acceleration in all areas except motor (which decreased by .5 standard deviations).  This class was comprised of 60% of the non-IP group, 75% of the IP group, and 52% of the ASD group. 

Class 2 consisted of another 28% of children; it was characterized by rapid development in all domains through 18 months, followed by stabilization of growth rate relative to gain in age (except motor, which showed disproportionate slowing in relative to age gain). This class consisted of 40% of the non-IP group, 17% of the IP group, and one child with ASD.

Class 3 was characterized by slowed development in all domains, and consisted of 0 children from the Non-IP group, 8% of the IP group, and 46% of the ASD group). 

ASD diagnosis increased the odds of assignment to class 3 relative to class 1 (OR=40, p<0.001) or class 2 (OR=43, p<0.001).  Outcome classification of ‘Intermediate Phenotype’ did not influence odds of assignment to class 2 versus class 1 (OR=-0.76, p=0.33).

Conclusions: About half of children with ASD are likely to have slowing in development beginning during infancy and continuing through the third birthday.  Children with an Intermediate Phenotype involving language and/or social delays are likely to show a burst in language and cognitive development beginning at 24 months.  The common practice of combining Visual Reception and Fine Motor scores to estimate Nonverbal Developmental Quotient is not supported by the trajectories identified herein.