Subtypes of Toddlers with Autism Spectrum Disorders

Background: Identification of subtypes of toddlers with autism spectrum disorders (ASDs) can improve knowledge on early manifestations of the disorders. Moreover, identification of subtypes of ASDs may inform early identification efforts by examining how characteristics that most distinguish subtypes affect performance on common diagnostic instruments. Past research suggests that social relations, verbal abilities, nonverbal abilities, and the presence of certain stereotyped interests and behaviors (SIB) may be important factors in delineating subtypes of ASDs in toddlers. Yet there is no published study that examines empirically derived subtypes in a sample of such young children.

Objectives: To determine whether distinct subtypes can be derived from a sample of toddlers who fail an autism screen and are subsequently diagnosed with an ASD.

Methods: Children were identified by two ongoing screening studies. Families were given the Modified Checklist for Autism in Toddlers (M-CHAT) during a visit to their pediatrician or early intervention provider. Families that indicated risk for developmental delay were offered a comprehensive clinical evaluation. The total sample evaluated (n = 288) was 80% male and 20% female. One hundred and seventy two children had race data; 143 of these children were White, 11 were Black, seven were Hispanic (including Puerto Rican), four were Asian, two were bi-racial, one was Hawaiian, and four chose “other” to describe their race. The average age at time of evaluation was 26 months (range = 13-37 months; SD = 5 months). One hundred eighty six children were diagnosed with an ASD after the clinical evaluation. Of these 186 children, 113 were diagnosed with Autistic Disorder, 72 were diagnosed with PDD-NOS, and one was diagnosed with Asperger’s Disorder. Items from the Childhood Autism Rating Scale were used to identify subtypes of children subsequently diagnosed with an ASD.

Results: Cluster analysis found three subtypes of toddlers with ASDs. Discriminant function analyses revealed two functions that accounted for group differences: a social-communication function and a SIB function. Group differences were found on social, communication, and cognitive skills and the rate and intensity of certain SIB (e.g., sensorimotor behaviors). Therefore, groups were labeled low social-communication with high SIB, low social-communication with low SIB, and high social-communication with low SIB. Lower-order or sensorimotor SIB that distinguished subtypes were repetitive use of objects and unusual sensory response. Items that assessed higher-order cognitive rigidity did not differentiate subtypes in this sample.

Conclusions: We found three subtypes of toddlers with ASDs delineated by social-communicative maturity and the rate and intensity of lower-order SIB. Consequently, defining toddler subtypes by these factors may improve knowledge of early manifestations of the disorders and early identification efforts. Repetitive use of objects and unusual sensory response should be given utmost priority when classifying toddlers with ASDs since these lower-order SIB distinguished subtypes in our sample. Higher-order SIB that focus on cognitive rigidity have less utility and, therefore, may be less useful in distinguishing ASD subtypes in younger cohorts.