An analysis of nearly 100 million medical records suggests that local rates of autism increase with rates of certain birth defects. The researchers suggest that this is further evidence - though indirect - that prenatal exposure to certain toxic chemicals can increase risk of autism.
Their report appears online today in PLOS Computational Biology.
The study found that, county by county, autism rates jumped by 283 percent for every 1 percent increase in the frequency of birth defects affecting male genitals. The study looked specifically at these types of birth defects because of other research linking them to prenatal exposure to chemicals such as lead and endocrine disruptors.
“This study is of interest because it may add further evidence that environmental factors can increase or decrease the risk of autism,” says Alycia Halladay, Autism Speaks senior director for environmental and clinical research. “However, caution is needed in interpreting its results.” Other studies have looked for and not found the evidence presented in this report, Dr. Halladay notes. Particular care must be taken in looking for a particular suspected link in an information source as huge as 100 million medical records, she says.
The research team – from the University of Chicago – analyzed health insurance claims that covered nearly one third of the U.S. population. Based on these records, they established a baseline frequency of autism across the country. Then they looked for differences in rates between counties. In their analysis, they corrected for other county-by-county differences such as average parent age, ethnicity and local wealth.
Focus on birth defects in boys
The researchers said they focused on birth defects of the male reproductive system based on their interpretation of research on how chemical exposures affect the system’s prenatal development. In particular, they noted research on the toxic effects of lead and chemicals that mimic hormones. In addition, they found that boys with autism were almost six times more likely than other boys to have genital malformations.
“We interpret the results of this study as a strong environmental signal," Dr. Rzhetsky said. He also noted that the geographic clustering of autism is evidence that, if vaccines play any role, it's an extremely weak one – because vaccinations are given uniformly across the country.
The researchers also found a weak link between autism and reproductive birth defects in girls. But the association was small enough that it could have resulted in chance. They found a similar, weak association between these birth defects and intellectual disability.
The researchers acknowledge limitations to their study and conclusions. For instance, they were able to access medical records more easily in certain counties over others. Also, they noted that autism rates appear to differ between states based on mandates on how children are evaluated for the disorder.
Finally, the study was not designed to find a direct link between autism and exposures to toxic chemicals. Rather the exposures were indirectly inferred by the presence of certain birth defects.
“These findings definitely warrant follow-up research,” Dr. Halladay says. Autism Speaks is already funding a broad range of studies directly investigating the role of environmental exposures and gene-environment interactions in autism, she notes.