DSM-IV-TR Criteria That Best Differentiate Intellectual Disability from Autism Spectrum Disorders

Background:  A high proportion of children with Autism Spectrum Disorder (ASD) also have intellectual disability (ID), complicating their symptom presentation. While ID is characterized by global delays in cognition and adaptive skills, children with both ID and ASD additionally show deficits in social interaction, communication, and restricted interests and repetitive behaviors. The somewhat similar characteristics of the two disorders pose a challenge for accurate differential diagnosis. 

Objectives:  In order to minimize diagnostic confusion, the current study sought to identify the DSM-IV-TR criteria that best differentiate between ID and ASD among children with features of both disorders.

Methods:  Population-based surveillance of ASD was conducted in 14 U.S. sites on children who turned 8 years old in 2006 and 2008. Evaluation of clinical and educational records identified 2,816 children who met case definition for ID, with or without ASD. Sensitivity and specificity, logistic regression, and an Area Under the Curve (AUC) Receiver Operating Characteristic analysis were conducted to determine which DSM-IV-TR criteria for ASD best differentiated children with ID and some ASD features (who did not meet criteria for ASD; N = 937) from those with co-occurring ID+ASD (N = 1,879). 

Results:  DSM-IV-TR Criteria related to social skills deficits, restricted interests and repetitive behaviors showed high sensitivity and specificity in differentiating the two diagnostic groups. AUC analyses indicated that, when considered singly, adherence to routines (AUC = 0.821) and stereotyped and repetitive behaviors (AUC = 0.823) showed the greatest power to discriminate between the two diagnostic groups. Criteria related to communication deficits, such as delay in spoken language (AUC = 0.660) and impairment in conversation ability (AUC = 0.659) showed the least discriminatory power. Results of the logistic regression indicated that a subset of 6 criteria in the areas of social skills deficits, restricted interests and repetitive behaviors was the most effective at differentiating the two diagnostic groups (AUC = 0.923). The specific subset included impaired use of nonverbal behaviors, lack of social reciprocity, stereotyped patterns of interest, adherence to routines, stereotyped and repetitive behaviors, and preoccupation with parts of objects. 

Conclusions:  Restricted interests and repetitive behaviors appear to be critical in differentiating individuals with ID and some ASD features from co-occurring ID+ASD. This is in contrast to prior assertions that, due to high rates of repetitive behaviors in the ID population, this is not a meaningful method of differentiating global cognitive delays from the specific constellation of behaviors found in ASD. Social skills deficits appear to have less discriminatory power than previously reported. Whereas deficits in social and communication skills may initially trigger concern regarding the presence of ASD, the absence of such skills does appear to as clearly differentiate ID from co-occurring ID+ASD as the presence of more unusual behaviors (restricted interests and repetitive behaviors).  An understanding of the discriminatory power of different DSM-IV-TR criteria may help clinicians reduce diagnostic confusion when patients have features of both ID and ASD, and may lead to improvements in diagnosis.