African American Toddlers with ASD Demonstrate More Social-Communication Symptoms Than Caucasian Toddlers

Background:   Understanding differences in symptom presentation across racial groups may be helpful in reducing disparities in the early identification of autism spectrum disorder (ASD). Given that African American (AA) children are typically misdiagnosed or diagnosed later than their Caucasian (CA) peers, it is necessary to explore differences in the direct observation of autism-specific symptoms. 

Objectives:  This exploratory study investigated differences in direct observations of social-communication symptoms in toddlers at risk for ASD. Racial differences will be explored while controlling for socioeconomic status, as measured by maternal education.   

Methods:  Parents completed the M-CHAT(-R) at pediatric 18- or 24-month well-child visits. Children at-risk based on the M-CHAT(-R)+Follow-up or a play-based screening measure, or whose parents/pediatrician expressed concerns about ASD were invited to complete a diagnostic evaluation (N=195,Mage=26.56 months, SD=5.13). Evaluations included measures of ASD symptomatology (Autism Diagnostic Observation Schedule Module 1-ADOS), developmental level, and adaptive abilities. For the purpose of this investigation, children who received a diagnosis of developmental delay or developmental language disorder (non-ASD group), and ASD were included in the analyses. Typically developing children and those missing race and income data were excluded.  

Results:   Caucasian (CA; N=102; M=15.51 years, SD=2.67) mothers had significantly greater years of education than AA parents (AA; N=74; M=13.60 years, SD=2.45), t=5.19, p<.001. A series of 2x2 between-subjects ANOVAs examined race and diagnosis (N_ASD=106, N_non-ASD=70), controlling for maternal education. Results indicated a significant main effect of race on ADOS Communication, F(1, 195)=7.41, p=.007 and Play, F(1, 195)=9.55, p=.002. AA toddlers demonstrated greater communication (M_AA=4.09, SD=1.92 > M_CA=3.42, SD=1.87) and play deficits than CA toddlers (M_AA=2.93, SD=1.17 > M_CA=2.56, SD=1.12). There was also a significant race x diagnosis interaction for ADOS Social Interaction, F (1, 195)=4.20, p=.04. Within non-ASD, AA toddlers displayed fewer social symptoms than their CA peers (M_AA=2.97, SD=2.59 < M_CA= 3.57, SD=2.54). In contrast, within ASD, AA toddlers displayed more social symptoms than CA toddlers (M_AA=9.91, SD=2.92 > M_CA=8.90, SD=2.34).  There were no significant racial differences in the restricted/repetitive behaviors or ADOS comparison scores. 

Conclusions:  These findings suggest that under direct testing conditions AA toddlers may display weaknesses in their communication skills and pretend play. Additionally, AA toddlers with ASD may display more social deficits (e.g., poor eye contact, reduced giving and showing) than CA toddlers with ASD.  It is possible that AA toddlers may need to demonstrate more severe social deficits to receive a diagnosis from evaluators. Although maternal education was controlled for in these analyses, unmeasured factors related to SES, such as toys available in the home, exposure to other children, and experience in childcare were not investigated. For example, some children may be encouraged to sit and play quietly as opposed to being encouraged to play pretend. These factors may also have influenced toddlers’ play and social skills during the observation. Future studies would benefit from exploring how cultural differences in the emphasis of pretend play and other aspects of social communication play a role in the assessment of early ASD symptomatology.