March 20, 2019

expert reaction to study on early life pesticide exposure and risk of autism spectrum disorder

Research published in The BMJ demonstrates that children’s risk of autism spectrum disorder increases following prenatal exposure to pesticides within 2000 m of their mother’s residence during pregnancy.

Dr Oliver Jones, Associate Professor of Analytical Chemistry, RMIT University, said: 

“This is an interesting study that appears to have been well thought through and carried out. The authors have used databases of historical data and statistical models to predict pesticide exposure and link this to possible health effects. They suggest a possible small increase in the likelihood of autism in children exposed to a number of different pesticides either while their mother was pregnant, or just after their birth. As the accompanying press release points out, the increase in risk identified is very small and there are a number of possible cofounders detailed in the study, such as the fact that the authors relied on historical records and didn’t measure pesticide levels directly. Neither did they suggest a way in which the effects they saw could be caused.”

Prof Stephen Evans, Professor of Pharmacoepidemiology, London School of Hygiene & Tropical Medicine (LSHTM), said:

“This paper describes a carefully conducted case-control study of autism spectrum disorder and exposure to pesticides.

“The associations the authors found are very weak and possibly beyond the limit of reliability of observational studies of this type. It was conducted in an area where the exposure to pesticides seems extremely high (75% of the controls, and 77% of the cases with autism were exposed to the most common pesticides- compounds containing glyphosate).  Adjustment for risk factors did not change the results very much it appears (not clearly stated in the paper) and the contribution of unmeasured confounding factors is unknown. It is now well-known that genetic factors are very important in development of autism. These are acknowledged but not measured in this study.

“Consequently, the results presented should be treated with a great deal of caution rather than assuming there is a definite causal effect and raising another “scare” for pregnant women or parents with children who have autism. Following the precautionary principle, one could avoid unnecessary exposure to pesticides during pregnancy, however this study can’t tell us if doing so would reduce the risk of autism.”

Dr David Menassa, Research Fellow at the University of Southampton, Stipendiary Lecturer in Neurophysiology at the University of Oxford and Neuroscience Theme Lead at The Physiological Society, said:

“Autism spectrum disorder (ASD) is largely genetic as demonstrated by family and twin studies. A small percentage of cases may have an environmental aetiology and this is what this current case-control study attempts to address.  A number of previous studies have emerged examining whether the baby’s exposure to pollutants during pregnancy is associated with a higher risk for the baby having an ASD. This was from a Finnish study which demonstrated an association between maternal exposure and getting ASD. This is not the first study of this kind although it has a higher number of cases against the controls examined. 

“Case control studies only give us odds ratios that have no relationship to risk. There is no causation here, only a weak association. However, one cannot conduct randomised control trials in this type of cohort because of the associated risks. Therefore the effort here can be commended and observational studies can be useful. 

“The paper reports a background rate of ASD of 8% in this cohort against a national USA CDC background rate of 1.5 percent in 2012/2013. It is curious that they have such a high rate in this special population. The study is conducted in a niche area that the authors describe as agricultural and there may be a higher risk of ASD more specifically there. This is unlikely to apply elsewhere. More mechanistic work is needed to demonstrate how something like glyphosate works and whether it would have an impact on the brain (for example in an animal model). One has to be careful in interpreting these findings here and the main message is that there is a weak association and no definitive causation in already a very small percentage of cases where the ASD can be attributed to an environmental cause. The authors are clear to highlight the strengths and weaknesses in their discussion.”

‘Prenatal and infant exposure to ambient pesticides and autism spectrum disorder in children: population based case-control study’ by Ondine von Ehrenstei et al. will be published in BMJ at 23:30 UK time on Wednesday 20th March.

Declared interests

Dr Oliver Jones: No conflicts of interest

Prof Stephen Evans: No conflicts of interest

Dr David Menassa: No conflicts of interest