Symptom dimension in ASD. A study based on ADI-R and ADOS-G factor analyses

Background:

Many studies on ADI-R algorithm items pointed out a two-factor structure composed of a socio-communicative and a stereotyped behaviours factor, whereas factor analysis of the ADOS-G identified only the socio-communicative factor (Robertson, 1999; Lord, 1999, 2000; Constantino, 2004). Besides, recent findings suggest that the restricted, repetitive behaviour area may contribute to diagnostic stability (Lord, 2006). These findings suggest that the current ADI-R and ADOS-G structures need to be revised so that they can reflect more accurately the relationships between sub-scales and create algorithms composed of items that best differentiate between clinical diagnoses.

Objectives:

to individuate the factors underlining autistic symptoms as identified through ADI-R and ADOS-G in a sample of carefully diagnosed individuals with ASD.

Methods:

ADI-R and ADOS-G were administered to 70 ASD referred for diagnosis to a second level hospital. Participants included 56 individuals with Autism (mean age: 8.5years; sd: 3.5years; mean total IQ: 68,6; sd: 26,3) and 14 with PDD-NOS (mean age: 9.5years; sd: 3years; mean total IQ: 69,4; sd: 26,3). All subjects met the DSM-IV clinical criteria for the two diagnoses. Exploratory factor analyses were conducted on ADI-R and ADOS-G.

Results:

ADI-R shows two main factors grouping most of socio-communicative (22% of accounted variance) and restricted, repetitive sub-scales (19% of accounted variance). ADOS-G shows one main factor (% of accounted variance ranges from 18% to 38%) relating to socio-communicative area.

Conclusions:

This study confirms the studies of Robertson (1999) and Constantino (2004) and suggests that two factors underline two different functional systems in carefully diagnosed individuals with ASD. This study can help to organise the ADI and ADOS items into a new algorithm and could contribute to reorient future research in genetic of ASD.