Genetics Significantly Influences Variability of Autism Occurrence

Boy holding up a dns helix model in Daycare center, Pre-school or Kindergarten, Cologne, NRW, Germany
Variation in autism spectrum disorder occurrence may be due to genetic influences, rather than maternal or environmental effects.

Variation in autism spectrum disorder (ASD) occurrence may be due largely to genetic influences rather than maternal or environmental effects, according to multinational cohort study results published in JAMA Psychiatry.

Investigators conducted a population-based cohort study of children born between January 1, 1998, and December 31, 2011, in Denmark, Finland, Sweden, Israel, and Western Australia. Patients were identified from the birth registers of their respective countries and followed for up to 16 years. Outcome and covariate data were abstracted from government-maintained national health registries.

Related Articles

A multigenerational family design was employed for analysis; variance in risk for ASD was decomposed into additive genetic (A), maternal (M), shared environmental (C), and nonshared environmental (E) components. Three nested models were then fitted based on these components: the AE model; the ACE model; and the ACME model. Using each model, country-specific point estimates for the contribution of each component were calculated. Pooled analyses were also conducted for the Nordic countries, given demographic similarities between Denmark, Finland, and Sweden.

The total analytic sample comprised 2,001,631 children (51.3% boys); 22,156 children (1.11%) were diagnosed with ASD. The median (95% CI) ASD heritability was 80.8% (73.2% to 85.5%) for country-specific point estimates, ranging from 50.9% (25.1% to 75.6%) in Finland to 86.8% (69.8% to 100.0%) in Israel. In the Nordic pooled sample, heritability estimates were 82.7% (79.1% to 86.0%) for the AE model, 82.2% (77.2% to 85.9%) for the ACE model, and 81.2% (73.9% to 85.3%) for the ACME model. Per the ACME model, country-specific point estimates of maternal effects ranged from 0.4% to 1.6%. Point estimates for shared environmental effects were also small in the ACME model, ranging from 0.0% to 14.5%. Estimates for nonshared environmental effects ranged between 16.0% and 19.3% for the AE model, 15.1% and 32.8% in the ACE model, and 17.5% and 33.6% in the ACME model. Overall, genetic and non-shared environmental factors contributed most consistently to variability ASD occurrence; in the three Nordic countries, genetic factors explained at least 73.9% of variability in risk, while non-shared environment explained at most 26.5%.

The investigators noted, “[The] effective sample size for individual countries was limited by the low prevalence of ASD.” Owing to sample size limitations, only the ACE model could be fitted for Israel and Western Australia.

These data identify genetics as the most significant influence on variability in ASD risk, with no support for the contribution of maternal effects. Modest differences in ASD risk factors may exist between countries. The impact of gene-environment interactions on ASD variability warrants further research.


Bai D, Yip BHK, Windham GC, et al. Association of genetic and environmental factors with autism in a 5-country cohort [published online July 17, 2019]. JAMA Psychiatry. doi:10.1001/jamapsychiatry.2019.1411