Association between early androgens and autistic traits: A systematic review and meta-analysis

Background: We systematically reviewed evidence for the androgen theory which proposes exposure to elevated levels of androgens in early development predisposes to autistic behaviour. Method: MEDLINE, EMBASE and Pubmed were searched for studies measuring androgens in mother or child during pregnancy or the first year of life and examined autistic behaviours (including social ability and repetitive behaviour) and language measured up to age 24 years. Results: Twenty-five of 3,041 publications met inclusion criteria, exploring 11 unique cohorts. Overall quality of evidence was very low as studies were non-experimental and most had high risk of bias. Only one research group found significant associations between autistic behaviour and androgens in amniotic fluid. There were mixed findings across the studies reviewed. Metaanalysis indicated a small significant pooled association between autistic behaviour and androgens in amniotic fluid (males and females combined; 3 studies), 0.28 [95 % CI 0.14, 0.41], also significant in males and females separately. Conclusions: Despite interest in this topic, of studies exploring direct measures of early androgens and later autistic traits, there is only a small amount of low-quality evidence from independent cohorts. The androgen theory of autism is neither confirmed nor refuted by the existing association studies included in this review.