Objectives: This study used the recently published severity metric to: 1) examine the distribution of ADOS severity scores across diagnostic groups; 2) further investigate the relationship between severity and participant characteristics (IQ, adaptive functioning, verbal impairment); and 3) examine ADOS severity scores over time in a subset of the sample.
Methods: To date, 256 children completed diagnostic assessments, including administration of the ADOS (168 module 1, 88 module 2), measures of nonverbal and verbal IQ (Mullen Scales of Early Learning, Differential Ability Scales-II), and adaptive functioning (Vineland II). At the time of the assessment, 141 children were diagnosed with autism (mean age=54.0 months, SD=21.9), 30 with PDD-NOS (mean age=44.9 months, SD=12.5), 53 non-ASD developmental delays (mean age=47.1 months, SD=14.9), and 32 with typical development (mean age=33.0, SD=6.6). A subset of children with autism (n=64) had a second ADOS between 4 and 15 months (mean=10.3 months, SD=3.4) following the first assessment.
Results: Preliminary findings revealed a Time 1 mean ADOS severity score of 7.54 (SD=1.4) for children with autism, 5.80 (SD=1.7) for PDD-NOS, 2.45 (SD=1.61) for non-ASD developmental delays, and 1.25 (SD=.44) for TD, consistent with the original paper in which ADOS classification anchored scores of 6-10 for autism, 4-5 for PDD-NOS, and 1-3 for nonspectrum.
For children with ASD (autism n=141, PDD-NOS n=30) at Time1, ADOS raw total scores were highly correlated with developmental/cognitive scores: nonverbal (r=-.49) and verbal (r=-.67) IQ. These correlations were lower with ADOS severity scores: nonverbal (r=-.21) and verbal (r=-.28) IQ, and found to be significantly lower when statistically compared (verbal difference: z=4.77, p<.001; nonverbal difference: z=2.93, p=.003). In addition, ADOS raw total scores were significantly correlated with adaptive functioning (r=-.38), while ADOS severity scores were not (r=-.11) (correlation difference: z=2.62, p=.009). For children with autism with two ADOS time points (n=64), results of a one-way within subjects ANOVA revealed no significant difference between Time 1 (mean= 7.4) and Time2 (mean= 7.27) ADOS severity, F(1, 63)=.46, p=.50. The majority of scores at time 2 remained within 1 point of Time 1 scores.
Conclusions: Results from this study indicate that ADOS (module 1 or 2) severity scores in a sample of young children with autism, PDD-NOS, and nonspectrum are similar to the mean severity scores for these groups in the original standardization sample. ADOS severity scores also appear more independent from cognitive and adaptive functioning. In addition, severity scores remained relatively stable in this sample with varying treatments. Further analysis of larger samples that contain multiple predictor variables will be important in determining when and how severity scores change in individuals with ASD.