Variables at the Time of ASD Diagnosis in Toddler Years Predict Outcome in Adaptive Functioning and ADHD Co-Morbidity in Adolescence: A Long-Term Follow-up Study

Background:

Some children with autism spectrum disorder (ASD), show limited progress, while others make rapid, remarkable gains. Short-term follow-up studies in ASD reported that higher cognitive ability, less severe autism symptoms, better adaptive functioning, and younger diagnosis age predicted better outcome in different domains. Adolescents with ASD have shown high rates of attention deficit/hyperactivity disorder (ADHD) co-morbidity.

Objectives:

To examine how measures of age of diagnosis, cognitive ability, autism severity and adaptive skills at diagnosis can predict long-term outcomes in adaptive skills and ADHD symptoms.

Methods:

The study included 61 participants, 55 males and 6 females, with an age range of 10:5 – 17:10 years (M=13:6±1:10y). All the participants were diagnosed with ASD (T1) in the age range of 1:3-3:1 years (M=2:2±0:5y) and were reassessed (T2) after 8:7-15:7 years (M=11:5±1:10y). ASD diagnosis at baseline employed two standardized tests, the Autism Diagnosis Interview-Revised (ADI) and the Autism Diagnosis Observation Schedule (ADOS), including the calibrated severity scale (CSS). Outcome measures included: Vineland adaptive behavior scales (VABS) composite score for adaptive skills; ADOS social-affect calibrated severity scale (SA-CSS) and restrictive and repetitive behaviors CSS (RRB-CSS) (≥4-ASD) at T1 and T2 to measure autism severity; Inattention and Hyperactivity/Impulsivity scores from the Conners' Rating Scales–Revised (CRS-R) (score>60 is considered significant) at T2.

Results:

At T2, four participants (6.7%) received a score of <3 for the ADOS-SCC. Seventy percent received a score >60 for the Conners Inattention scale and 58.3% a score >60 for the Conners Hyperactivity/Impulsivity scale. To explore which variables at T1 predicted outcome in functioning (VABS composite scores) and ADHD symptoms (Inattention and Hyperactivity/Impulsivity scores) at T2, we performed three hierarchical regression models for these dependent variables.

Independent variables in each model included: T1 DQ in the first step, T1 VABS score in the second step, ADOS-SA-CSS and ADOS-RRB-CSS scores in the third step and age of diagnosis in the fourth step.

The first model, where the VABS score served as the dependent variable, explained 19.7% of the variance. ADOS-SA-SCC score (β=-.36, p<.01) and age of diagnosis (β=-.23, p<.05) correlated negatively and significantly with VABS score at T2. The second model, where the Conners Hyperactivity/Impulsivity score served as the dependent variable, explained 27.5% of the variance. DQ score at T1 (β=.33, p<.05) correlated significantly and positively and VABS score at T1 (β=-.33, p<.05) correlated significantly and negatively with the Conners Hyperactivity/Impulsivity score. The third model, where the Conners Inattention score served as the dependent variable, explained 17.7% of the variance. DQ score at T1 (β=.27, p<.05) and ADOS-RRB-CSS at T1 (β=.36, p<.05) correlated significantly and positively with the Conners inattention score.

Conclusions:

Less severe social-communication symptoms and younger ASD diagnosis age predicted better functioning in adolescence. Baseline lower adaptive skills and higher cognitive abilities predicted more severe symptoms of hyperactivity/impulsivity in adolescence. More severe RRB symptoms and higher cognitive abilities at the time of ASD diagnosis predicted more severe inattention symptoms in adolescence. Autism severity and level of functioning in early diagnosis of ASD predicted adaptive skills and ADHD co-morbidity.