Nonverbal IQ in Young Adults with Autism Spectrum Disorder: Correspondence with Scores from Early Childhood

Background: IQ has been studied as an important factor in predicting both current and future functioning in children with autism spectrum disorder (ASD).  However, relatively little is known about longitudinal measurement and stability of IQ, particularly in children who span the full range of cognitive ability.  

Objectives: To assess the stability of nonverbal IQ scores in a cognitively heterogeneous sample of individuals with ASD from toddlerhood to young adulthood.  

Methods: A sample of 84 children diagnosed with ASD at age 2 years was tested again at age 3 and then followed through age 18.  At each assessment, participants were administered developmentally-appropriate tests to obtain estimates of cognitive functioning either through standard IQ scores, if possible, or ratio-derived developmental quotients (DQs).  

Results: Nonverbal IQ/DQ scores generally declined from 2 to 18 for the majority of individuals with ASD in the sample.  While age 2 scores were predictive (R²=.44), age 3 scores were slightly more predictive (R²=.55).  However, at both ages, relationships were generally stronger for individuals with lower nonverbal IQ/DQ.   At age 2, R² = 0.35 for children with NVIQ/DQ < 70, while R² = 0.08 for children with NVIQ/DQ ≥70.  At age 3, R² = 0.26 for children with NVIQ/DQ < 70, and R² = 0.17 for children with NVIQ/DQ ≥70.  Broad score divisions (e.g., IQ/DQ over vs. under 70) showed relatively high stability, whereas narrower DSM-IV-based ability ranges (e.g., Borderline, Mild MR, etc.) were much less stable over time. At age 2, children were distributed as follows: Average (NVIQ/DQ ≥ 85, n = 17), Borderline (NVIQ/DQ ≥ 70, n = 26), and the remainder (n = 41) fell into the ID range. Twenty-one were in the moderate to profound category (NVIQ/DQ< 50). At age 18, when 67% of the sample required use of a test outside of its standardized age range, the number of participants with NVIQ/DQ < 50 jumped to 45, while the Average (n = 21) and Borderline (n = 9) groups declined.  Use of Vineland-II Daily Living Skills scores in place of nonverbal IQ/DQ scores at the age 18 time point did not improve the predictive validity of age 2 scores (R²=.41).

Conclusions: These findings indicate that most children with ASD who exhibit significant nonverbal cognitive “delays” as young children score in the range of intellectual disability as adults.  On the other hand, the majority of individuals with nonverbal IQ scores in the average range or above by age 3 tended to continue to score in the average range or above at age 18.  Particularly within the lower ranges of ability, this study indicates that “absolute” nonverbal IQ scores of individuals with ASD decline from age 2 to 18.  However, limitations of available and appropriate tests may be significant factors in this decline, as well as growth in verbal skills during this time period.  Clinicians and researchers should use caution when making comparisons based on absolute scores or specific ability ranges within or across individuals with ASD of different ages.