The Autism Inpatient Collection (AIC): Problem Behaviors and Communication Difficulties

Background: Problem behaviors are linked to the level of language/communication skills in children with autism spectrum disorders (ASD). Previous studies, however, have primarily included school-enrolled children with limited representation of children with a range of skills. The Autism Inpatient Collection (AIC) provides rigorous phenotypic and biological data on a large cohort of children and adolescents with the full spectrum of ASD characteristics, allowing a more full examination of the relationship between problem behaviors and communication skills in ASD.

Objectives: The language/communication profile and problem behavior rates of inpatients with ASD were examined based on verbal level.

Methods: Participants were 105 children and adolescents aged 4-20 years, admitted to six specialized inpatient psychiatry units and prospectively enrolled in a study performing phenotypic assessment and collection of biological samples.  All subjects scored ≥12 on the Social Communication Questionnaire (SCQ) before study enrollment.  ASD diagnosis was confirmed by ADOS-2 administration by a research-reliable examiner and extensive inpatient observation by expert clinicians.  Other measures included the Aberrant Behavior Checklist (ABC), Repetitive Behavior Scale - Revised, Leiter-3, and Vineland-2. 

Results: Participants were categorized based on the administered ADOS module [Modules 1/2 = Low Verbal (LV), n=58; Modules 3/4 = Verbal (V), n=47]. The groups did not differ significantly in age [LV Mean age 12.58 years (SD 3.61), range 5.77-20.08; V Mean age 12.77 (SD 3.37), range 4.58-17.92, t(103)=-.272, p=.787]. They were similar in gender (LV=25.9%; V=23.4% female), race (LV=77.6%; V = 78.7% Caucasian), and ethnicity (LV=92.2%; V=90.0% non-Hispanic/non-Latino). The groups were significantly different in mean non-verbal IQ [LV Mean 49.47 (SD 17.65), range 30-99; V Mean 94.35 (SD 22.52), range 35-135; t(81)=-10.15, p<.0001]; mean SCQ total score [LV Mean 26.36 (SD 5.34); V Mean 21.52 (SD 5.89); t(90)=4.13, p<.0001]; and, mean Vineland-2 receptive communication score [LV Mean 6.29 (SD 2.89); V Mean 8.13 (SD 1.85), t(69)=-3.09, p=.003] and expressive communication score [LV Mean 4.10 (SD 2.78); V Mean 9.50 (SD 3.07); t(67)=-.759, p>.0001]. Initial analyses of measures of problem behavior indicated that the groups did not differ on mean ABC-Irritability scores [LV Mean 28.13 (SD 10.87); V Mean 24.93 (SD 13.17), t(99)=1.34, p=1.85] but did differ significantly on mean ABC-Stereotypic behavior scores [LV Mean 10.15 (SD 6.36); V Mean 6.04 (SD 5.87), t(99)=3.34, p=.001] and mean for self-injurious behavior subscale from the RBS-R [LV Mean 9.75 (SD 6.02); V Mean 7.00 (SD 5.41), t(88)=2.24, p=.028]. Sixty-three percent of the children who were Low Verbal were reported to use a communication system; 55.2% of these used the Picture Exchange Communication System (PECS); 24.1% used sign language; and, 37.3% used a voice output device.

Conclusions: This cohort of children and adolescents with ASD have a high proportion of significant cognitive and spoken language impairment. Children with more limited verbal skills had more stereotypical and self-injurious behaviors as compared to children with fluent speech. In addition, only about half of the participants with limited verbal skills were reported to use a method of augmented communication, a variable that should be further examined with respect to problem behaviors.