A new statistical framework developed by researchers at the Johns Hopkins Bloomberg School of Public Health, Johns Hopkins University School of Medicine, and Kaiser Permanente Northern California offers improved understanding of how genetics and environment contribute to autism risk.
Large-scale genetic studies have led to the development of genetic risk scores that estimate a person's predisposition to diseases and health conditions based on their DNA profiles. The new framework allows researchers and clinicians to analyze these scores using family data and characterize the risk of conditions such as autism and other developmental conditions in children based on their own DNA, parental factors, and environmental influences such as maternal diet and lifestyle.
For their study, published June 2 in Nature Genetics, the researchers analyzed more than 18,000 case-parent trios—autistic children and their parents—across diverse ancestral populations in the Simons Foundation Powering Autism Research for Knowledge consortium and the Genes and Environment Autism Research Study.
Key Takeaways
- A previously published genetic risk score predicts autism risk strongly in individuals of European ancestry, but is less accurate in other populations, especially African-ancestry populations.
- Maternal genetic susceptibility for certain traits, including obesity and certain neurocognitive characteristics, may increase the risk of autism in children.
- Known maternal risk factors, such as pregnancy complications, multiply the risk of autism, on top of the risk due to genetic susceptibility.
"This new framework allowed us to gain novel insights into the complex interplay between genes and environment in developmental conditions such as autism and has important future implications for the discovery of risk factors and biomarkers using data collected from children, parents, and families," says senior author Nilanjan Chatterjee, Bloomberg Distinguished Professor of Biostatistics and Genetic Epidemiology at Johns Hopkins University, with appointments in the Bloomberg School's Department of Biostatistics and the School of Medicine's Department of Oncology.
The researchers expect their framework will provide the foundation for new types of studies and analyses, using family data, to better understand the genetic basis and gene–environment interactions for developmental conditions like autism. They plan to build out these methods to analyze data from broader family structures, including extended sets of relatives, and unify the results from such family studies with larger population-based studies of unrelated individuals.
The low accuracy of the existing genetic score for predicting autism risk in non-European populations highlights the need for more data collection and curation from diverse populations, especially individuals of African ancestry, to develop updated genetic scores for autism that can be applied more universally. The researchers are initiating efforts to compile existing data from diverse populations to develop genetic scores for children and adults separately, to predict various health conditions, lifestyle factors, and biomarkers. These scores can be applied to very large genetic studies of autism and other developmental conditions to test novel hypotheses regarding the roles of genes and environments.
The researchers note that although the findings linking maternal genetic susceptibility for certain non-autism traits to autism risk are compelling, the study design may introduce biases that could influence the results, and these results need to be further validated using alternative study designs.
This study was a collaboration led by lead author Ziqiao Wang, a postdoc at the Bloomberg School at the time, across the departments of Biostatistics, Epidemiology, and Mental Health. This work was supported by the National Institutes of Health (R00HG013674, R01HG010480, U01CA249866, R01ES034554, R35GM150836, and R01DE031855).
"Estimation of Direct and Indirect Polygenic Effects and Gene-Environment Interactions using Polygenic Scores in Case-Parent Trio Studies" was co-authored by Ziqiao Wang, Luke Grosvenor, Debashree Ray, Tianyuan Cheng, Ingo Ruczinski, Terri H Beaty, Heather Volk, Christine Ladd-Acosta, and Nilanjan Chatterjee.
