Is There Any Early Developmental Factor That Verbal Acquisition Will Appear in Children with Autism?

Background: Autism (MIM 209850) is a complex neurodevelopmental disorder defined by impaired social interaction, communication and restricted, repetitive and stereotyped behaviour.
Whether there are differences in the early developmental abilities of children with autism spectrum disorder (ASD) and whether those differences in a specific period of time might predict verbal development, are questions of continued debate in autism research and that emerged from our clinical experience.

Objectives: Understand if there is any marker in the developmental/cognitive profile that can predict later verbal development.

Methods: The sample includes 34 subjects with autism based in Autism Diagnostic Interview Revised (ADI-R) and Childhood Autism Rating Scale (CARS) positive results. Inclusion criteria were two separated assessments with Griffiths Scale of Mental Development, the first at pre-school age and the second one at school age. All children benefited from speech-language therapy and proper teaching model based in the Treatment and Education of Autistic and related Communication Handicapped Children (TEACCH). To find a marker in the developmental profile which can predict verbal development, Global Developmental Quotient (GDQ), Language Developmental Quotient (LDQ) and Performance Developmental Quotient (PDQ) were assessed at both periods. Later (at school age) we stratified the sample as nonverbal (NV) and verbal (V) according to the acquisition of language (spontaneous, flexible use of at least two words in combination, one of which must be a verb-ADI-R definition).
Statistic analysis (SPSS 15) was performed comparing the developmental profile of both subgroups with Mann-Whitney test for unpaired samples and Wilcoxon test in paired samples. Significance level (σ)=0.05.

Results: 34 subjects (26M/8F: 3/1) ADI-R positive and a median score CARS=35 (P₅ =30; P₉₅=50). At pre-school assessment (median age 49M) all children were NV and had a median GDQ=60. The group median LDQ was very low (40) and PDQ median value was higher (67), g<0.001. At the school period assessment (median age 81M) the median GDQ-65, and median PDQ-70 were similar to the first assessment, but the language development had increased significantly LDQ (40/52; g=0.001), as expected, since 17 of 34 children had become verbal.
We compared the developmental profile at preschool age between NV/V children. At the first evaluation both subgroups (NV/V) had low LDQ as expected (N/V=28; V=45), but they were very different in PDQ (NV=45; V=83; g<0.001). At the time of the second evaluation the verbal subgroup had a marked improvement in the LDQ=85 (first time=45; g<0.001), but the PDQ=95 (first time PDQ=83; g=0.522) remained equal. At the non verbal subgroup the results of LDQ (19) and PDQ (45) did not experience any change.

Conclusions: These findings demonstrate that in non verbal preschool children with autism, PDQ may be an index that verbal acquisition will appear, regardless of educational intervention. These preliminary findings should be replicated in a larger sample.