Predicting social competence in ASD from age 2-age 8

Sally J. Rogers, PhD1, Susan Hepburn, Ph.D.2, Amy Philofsky, Ph.D.3, David Most, Ph.D.4, and Deborah Fidler, Ph.D.4. (1) The M.I.N.D. Institute, UC Davis, 2825 50th St., Sacramento, CA 95817, (2) School of Medicine/Psychiatry, University of Colorado at Denver, 13121 E. 17th Avenue, Anshutz Medical Center, Box C-234, 5th Floor, Aurora, CO 80045, (3) University of Colorado Health Sciences Center, Denver, CO 80262, (4) Colorado State University, Fort Collins, CO

Background: Current developmental explanatory theories of autism include the imitation/mirror neuron theory (Rogers/Pennington,Williams), executive function theory (Russell), triadic joint attention/intersubjectivity theory (Mundy, Tomasello), sensory overarousal theory, and a dyadic social-emotional theory (Hobson and colleagues).

Objectives: to compare the predictive power of variables involving imitation, joint attention, emotional , executive function, and imitation at age 2 and social competence at age 8.

Methods: We used a longitudinal design, from age 2-3 to ages 8-10  in two groups of children, Autistic Disorder (n = 31), and other developmental difficulties (n = 33) matched on MA and CA. The test battery at age 2  included measures of intellectual function, adaptive behavior, and autism severity, as well as measures of  executive function (spatial reversal), imitation (manual, oral, and object), initiating and responding to joint attention,  emotional communication (directed smiles), and sensory abnormality (via parent report).

Results: At age 2-3, only three measures: emotion (directed smiles), initiating joint attention, and imitation, significantly differentiated the groups (p<.05).  At age 8, the ASD group was significantly superior to the DD group in overall IQ (20 point advantage; p=.02) and nonverbal IQ (25 point advantage, p=.002).  Regression analyses revealed that, for ASD, imitation and executive function at age 2 each accounted for a significant amount of the variance in age 8 Vineland language and social skills and Verbal IQ. IJA was significant only for communication outcomes. In the DD group, almost no variables predicted significantly to age 8 outcomes.

Conclusions: Age 2 neuropsych variables had much more influence on age 8-10 outcomes for ASD than DD group, and were more predictive than age 2 IQ measures. The imitation variable alone carried significant weight in all outcome measures. The nature of the IQ deficits appear different in ASD, more linked to the social deficits, than in DD.