Environmental risk factors for autism: an evidence-based review of systematic reviews and meta-analyses.

The review only included previously published systematic reviews and meta-analyses, so studies published after those reviews or not already reviewed were excluded. Many meta-analyses had estimates that were not precise, inconsistent, or indirect, and some had publication bias. Most evidence came from observational studies, so an association may reflect other factors, recall bias, or shared genes and environment rather than cause. Nutritional studies often assessed deficiency or supplementation after ASD was already present, which makes causal interpretation difficult. Chelation studies were uncontrolled and lacked placebo arms, so their reported symptom improvement cannot be taken as evidence that chelation works. ASD was defined broadly and includes multiple subtypes, so risk factors may not apply equally to all patients.

Do not conclude that environmental factors cause autism. The included reviews report associations, and many estimates were not precise, inconsistent, indirect, or confounded by other factors.