The Autism Inpatient Collection (AIC): Methods and Sample Desription

Background:  Individuals severely affected by autism spectrum disorder (ASD), including those with intellectual disability, significant expressive language impairment, and/or self-injurious behavior (SIB) are underrepresented in the autism literature and in current collections of phenotypic and biological data. 

Objectives:  The Autism Inpatient Collection (AIC) was developed to provide a resource for the study of the full autism spectrum.  Rigorous phenotypic and biological data are being collected on a large cohort of children and adolescents with ASD, hospitalized in six specialized inpatient psychiatry units.

Methods:  Children and adolescents aged 4-20 years, admitted to specialized inpatient psychiatry units, are prospectively enrolled in a six-site study.  Subjects scoring ≥12 on the Social Communication Questionnaire are offered participation in the study.  After informed consent, the Aberrant Behavior Checklist (ABC), Repetitive Behavior Scale - Revised, Leiter-3, Vineland-2, Parent Stress Index – Short Form-4, Difficult Behavior Self-Efficacy Scale, Functional Assessment Screening Tool (FAST), Emotion Dysregulation Inventory and Child and Adolescent Symptom Inventory-5 are completed.  ASD diagnosis is confirmed by ADOS-2 administration by a research-reliable examiner and extensive inpatient observation by expert clinicians. Additional data, including psychotropic medication usage, co-morbid psychiatric and medical diagnoses and sleep observations are collected.  Biological samples from probands and their biological parents are sent to the Rutgers University Cell and DNA Repository, where they are processed for DNA extraction, and creation of induced pluripotent cell lines from the proband.

Results:  Mean age of the first 108 subjects is 12.70 years (SD 3.50, range 4.58-20.08); 24% are female; 77.8% self-identified as Caucasian, and 91.4% non-hispanic/non-latino.  Mean non-verbal IQ = 70.69 (SD 30.21, range 30-135) with a tri-modal distribution (see Figure 1) and a mean adaptive behavior composite score of 56.9 (SD 21.70, range 27-196).  The majority of subjects (55.7%) were administered Module 1 (single words or less, 47.2%) or Module 2 (phrase speech, 8.5%) of the ADOS-2, indicating very limited verbal ability.  Children administered Module 1/2 had a significantly lower mean non-verbal IQ score (M=49.18, SD=17.55) than Module 3/4 (M=94.35, SD=22.52; t=-10.30, p<.001). The mean Vineland-2 expressive communication sub-scale score was 6.36 (SD 3.94, range 1-17), three standard deviations below the population mean.  The mean ABC-Irritability sub-scale score was 26.67 (SD 11.91, range 0-45) and 26% of the sample were observed to have SIB by a psychologist or board certified behavior analyst.

Conclusions:  Preliminary data on the first 108 subjects enrolled in the AIC indicate that this is a cohort of children and adolescents with ASD with a high level of cognitive and spoken language impairment.  More than half the sample has less than fluent speech, the majority have intellectual disability, over one-quarter have self-injurious behavior and their mean adaptive functioning score is three standard deviations below the population mean.  The Autism Inpatient Collection is a substantially new resource to advance study of the full autism spectrum, which will augment the wealth of existing data on higher-functioning cohorts.  Although it is too early to tell, we hypothesize that the genetic diversity of the AIC cohort may differ substantially from extant patient data collections.