select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

air pollution and autism

A study in Archives of General Psychiatry suggested exposure to traffic-related air pollution, particulate matter and nitrogen dioxide during pregnancy and during the first year of a child’s life appears to be associated with an increased risk of autism. This analysis accompanied a roundup which can be viewed here.



Title, Date of Publication & Journal

‘Traffic-Related Air Pollution, Particulate Matter, and Autism’ by Heather Volk et al., 27 November 2012, Archives of General Psychiatry



Claim supported by evidence?

  • This paper does show an association between estimated air pollution and autism.
  • This study does not show that air pollution causes autism for two reasons:
    1. The study cannot completely rule out the impact of other factors that could cause the apparent association.
    2. The study cannot describe how pollutants would interact with other factors to cause autism
  • The authors say that considerable further research must be done before a causal link could be inferred.


  • The paper observed an association between air pollution and autism using a case-control design based in California.
  • This is a detailed study. The paper does account for some factors such as age, sex and ethnicity and socioeconomic factors. However, we do not know from this research whether other factors are causing this association

Study Conclusions

The study found that local estimates of traffic-related air pollution and regional measures of particulate matter (PM) 2.5, PM10, and nitrogen dioxide were higher in children with autism.

The authors conclude that additional research is needed to replicate these findings, and also to investigate potential biological pathways, before we could say that air pollution causes autism.




  • Rigorous methods were used to identify children with autism and children without autism.
  • The study has taken account of a number of factors such as:
        • socioeconomic and demographic factors
        • child’s age, sex and ethnicity
        • parental education
        • mother’s age and smoking status during pregnancy
        • a rough measure of urban/rural living
  • Two independent methods were used to measure air pollution and each showed an association
  • The first method was measured air quality data taken from regional measurement stations.
  • The second type was estimated/modelled local air-traffic pollution, based on annual traffic volumes, speeds  and emissions, prevailing weather conditions and distance from each road.


  • They authors point out that the apparent association could be caused by unmeasured risk factors including lifestyle, nutritional or other residential exposures. Family history is another such known and unmeasured risk factor.
  • The authors also point to other potential subtle explanations, such as geographic clustering in areas where autism is more likely to be diagnosed.
  • The study is not powerful enough to show whether there is a particular time period in foetal or childhood development where the association is strongest.


A case-control design finds cases (children with autism) and controls (children without autism), but who are similar in other respects (such as sex and age) and examines which group has seen greater exposure (in this case, has been exposed to more polluted air).


‘Before the headlines’ is a service provided to the SMC by volunteer statisticians: members of the Royal Statistical Society (RSS), Statisticians in the Pharmaceutical Industry (PSI) and experienced statisticians in academia and research.  A list of contributors, including affiliations, is available here.

in this section

filter Headlines by year

search by tag