Is the Prevalence of Autism Associated with Maternal Ethnicity and Nativity?

Friday, May 12, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
R. Bruno1, M. Khalil2 and M. Elsabbagh2, (1)Research Institute of the McGill University Health Centre, Montreal, QC, CANADA, (2)McGill University, Montreal, QC, Canada

The prevalence of autism is highly variable within and across geographic regions (Elsabbagh et al., 2012 Aut Res). Several factors have been proposed to explain this variability. Among these factors are differences in level of awareness and recognition as well as the availability of services in different communities. Previous research has suggested that socio-demographic factors such as maternal ethnicity and nativity are associated with prevalence, which may in turn explain some of the variability in estimates.


We conducted a systematic review to establish the reliability of association between sociodemographic factors and autism prevalence.


We searched Medline and PsychInfo for manuscripts where association between autism prevalence and sociodemographic factors was investigated. We included research articles exploring ethnicity and nativity or immigration in relation to autism regardless of their design. By following PRISMA guidelines, this approach allowed for inclusion of large population studies from which quantitative data can be extracted to compare relative risk between groups as defined by these sociodemographic variables. Non-original articles and those not reporting prevalence estimates were excluded.


Our search identified 130 records, from which we retained 44 full text articles after screening and application of inclusion and exclusion criteria. Of these 16 original articles reported sufficient data on the association between prevalence and sociodemographic risk factors to be included in the final analysis. A meta-analysis was possible to conduct only for immigration status, where data extraction provided clearly differentiated risk estimates for immigrants vs. non-immigrants. The total population included in the analysis was 76.7 Million (321,188 individuals with ASD). Relative risk was 0.947 in immigrant group versus non-immigrant (CI- 0.66- 1.35, P<0.001). However, we also found significant heterogeneity in the estimates (Q = 1518.04, p < .0001), limiting confidence in the findings of the meta-analysis.


Several strong claims have been made in the literature regarding the association between socio-demographic factors and autism prevalence. Our finding, using a systematic review methodology, suggests that the number of studies and their quality limits the validity of these claims. Our review also identified sampling biases where risk factors like ethnicity and immigration are more often than not confounded by other socio-demographic factors. Future studies need to also shift from post-hoc accounts and toward specific and biologically plausible hypotheses explaining how sociodemographic factors might impact autism prevalence.

See more of: Epidemiology
See more of: Epidemiology