An Exploration of the Relationship Between the Child Behavior Checklist and ADOS Comparison Scores Including Possible Mediating Factors

Background: The Withdrawn syndrome scale and the Pervasive Developmental Problems (PDP) DSM-oriented scale on the Child Behavior Checklist (CBCL) have received attention in the literature as one way to identify children with symptoms consistent with an Autism Spectrum Disorder (ASD) diagnosis.  Although previous research has supported a relationship between the CBCL Withdrawn scale and an ASD diagnosis, inconsistent results have emerged regarding the relationship between the PDP scale and an ASD diagnosis.  However, Gotham et al.’s work (2009) in developing calibrated ADOS Comparison Scores now allows for an examination of these scores as they relate to both the Withdrawn and PDP scales.  ADOS Comparison Scores range from 1-10 and are divided into four interpretive categories that correspond to the level of autism spectrum-related symptomatology that is observed during the ADOS or ADOS-2.  

Objectives: The aim of the present analysis was to further explore the relationship between the Withdrawn and PDP CBCL scales and ADOS Comparison Scores while taking into account possible confounding effects of child race or ethnicity, sex, and cognitive ability.  

Methods: Data were obtained from 2161 children between the ages of 2 and 5 years who participated in the Autism Treatment Network.  Subjects were included if they had a calculated ADOS Comparison Score, a complete parent-report CBCL, and had completed a standardized measure of cognitive or developmental ability (overall IQ).        

Results: Correlational analyses indicated a significant relationship (p<.01) between ADOS Comparison Scores and T-scores of multiple CBCL scales, including the Withdrawn scale (p<.0001).  The relationship between ADOS Comparison Scores and the PDP Scales approached significance at the .01 level (p=.0190).  However, when T-scores on both the Withdrawn and PDP scales were greater than 65, the relationship between Comparison Scores and PDP T-scores was no longer significant.  Gender, race, and ethnicity did not change the relationship between Withdrawn and PDP T-scores and Comparison Scores.  After controlling for overall IQ, the relationship between Comparison Scores and PDP T-scores was no longer significant while the relationship between Comparison Scores and Withdrawn T-scores remained significant.  

Conclusions: Although CBCL Withdrawn and PDP scale T-scores were significantly correlated with ADOS Comparison Scores in the present study, the relationship was complicated by the impact of overall IQ.  It is possible that the items unique to the PDP scale capture behaviors of children with lower cognitive abilities and artificially inflate the relationship between this scale and ADOS Comparison Scores.  Our results suggest that caution should be exercised when CBCL PDP scale scores are used to indicate the presence of ASD as designated by higher ADOS Comparison Scores as a greater number of false positives are likely unless cognitive level is accounted for. 

This research was conducted as part of the Autism Speaks Autism Treatment Network. Further support came from a cooperative agreement (UA3 MC 11054) from the U.S. Department of Health and Human Services, Health Resources and Services Administration, Maternal and Child Health Research Program, to the Massachusetts General Hospital. The views expressed in this publication do not necessarily reflect the views of Autism Speaks, Inc.