As an off-shoot of a study examining the reliability and validity of an adapted version of the Pre-Linguistic Autism Diagnostic Observation Schedule (A-PL-ADOS), 13 individuals with Down syndrome with IQs ranging between 24 and 48 were administered the Autism Diagnostic Interview-Revised (ADI-R) and the A-PL-ADOS, which are well-validated interview and observational diagnostic measures. Three out of 13 met lifetime criteria on the ADI-R, but none of these three showed behavior that met the criterion for autism on the A-PL-ADOS (although two nearly did so). However, two individuals did meet the A-PL-ADOS criterion and showed behavior that fell only just short of meeting lifetime criteria on the ADI-R. Altogether, 5 individuals with Down syndrome may be considered to show an autism spectrum disorder. Of the remaining 8, some showed a few autistic features, and some showed none. The findings raise both methodological and conceptual issues.

The Autism Diagnostic Observational Schedule (ADOS) is a semi-structured observational scale developed to assess social interaction, communication and play in individuals who are suspected to have autism. Since the ADOS is not suitable to be used with severely or profoundly mentally retarded adolescents and adults with very limited language skills, materials and some of the tasks of the PL-ADOS and the original ADOS (the former versions of the current ADOS) were adapted. Results indicated that almost all of the overall ratings showed good reliability and discriminative diagnostic validity. Furthermore, the combination of codings into an overall algorithm score on social/communicative behavior resulted in a sensitivity of .82 and a specificity of .85 when using a cut-off score of 15.

We sought to determine if the family loading for either the broader autism phenotype or for cognitive impairment differed according to whether or not autism was accompanied by severe mental retardation. The sample comprised 47 probands with autism meeting ICD-10 criteria, as assessed by the Autism Diagnostic Interview and the Autism Diagnostic Observation Schedule. Family history interview and findings were compared with those for the higher IQ autism and Down syndrome samples in the Bolton et al. (1994) study. The familial loading for autism and for the broader phenotype was closely comparable to that in the study of higher IQ autism, and different from that for Down syndrome. The family loading for scholastic achievement difficulties was slightly, but significantly, higher when autism was accompanied by severe retardation.