Who Are We Missing: Why Would a Community-Derived Sample of Children with ASD Have Higher CBCL Symptoms Than Clinic-Based Samples?

Background:   Children with Autism Spectrum Disorders (ASD) have increased rates of mood and anxiety symptoms and suicidal ideation and attempts (SI/SA).  The available literature indicates demographic characteristics (age 10+, male, African American/Black, Hispanic/Latino, low SES) and psychological characteristics (depression, teasing/bullying, aggression, impulsivity and anxiety) as risk factors for SI/SA in ASD.  In order to probe levels of disparity of care, a school-based program that did not require previous diagnosis of ASD was compared to a group of youths who received clinical evaluations in a specialty clinic or were referred from a clinical setting for research testing. 

Objectives:   To compare Child Behavior Checklist (CBCL) syndrome scales, DSM scales, and SI/SA item endorsement between clinical and community groups and among different levels of maternal education. 

Methods:   Parents of 147 youth with ASD (female=19), age 5.92-19.17 (M=10.29, SD=2.5), FSIQ 62-149 (M=103.79, SD=19.43), who met CPEA criteria for ‘broad ASD’ on the ADI-R and/or ADOS, completed CBCL on their children.  123 families were evaluated in a clinical setting, and 24 families were evaluated in a community setting.  Maternal education in 7 categories was used as a proxy for SES; 74.8% of mothers in the clinic sample and 56.5% of mothers in the community sample have a college or graduate education.  Clinical and community groups did not differ significantly on sex or IQ; 11.6% of the clinical sample and 21.7% of the community sample is African American/Black.  Group differences were explored with Chi-square and independent sample t-tests. 

Results:   15.6% of individuals had parent-reported SI/SA. 83.7% of individuals showed clinical elevation on CBCL syndrome scales overall, 73.6% showed clinical elevation on CBCL DSM scales overall, and 43.1% showed clinical elevation on the affective problems scale.  Endorsement of SI/SA items was associated with significantly higher CBCL anxiety (t(26.42) = -2.36, p<.05) and depression t-scores (t(145) = -2.81 ,p<.01). 

There was no significant difference between clinical and community groups on endorsement of suicide items and overall CBCL syndrome scale elevation.  Among maternal education levels, there were no significant differences on CBCL syndrome and DSM scales or SI/SA endorsement, possibly because such a large percent of the group was highly educated. 

There is a significant difference (p<.01) on overall CBCL DSM scale elevation between community and clinical groups (chi square).  Examination of DSM scale categories indicated that significant differences exist between clinical and community groups on CBCL clinical level affective problems (t-score ≥65, p<.05) (chi square).  This holds true when considering the continuous CBCL affective problem data with an independent sample t-test (t(142) = 2.49,p<.01). 

Conclusions:   Although significant differences based on SES were not found, there were differences between clinical and community samples (see Figure 1).  Based on a relatively small sample of community data, significantly more DSM symptomatology was reported by parents.  Future directions include expanding the sample size of the community group and understanding what services this group is and is not receiving.