Medical history of discordant twins and environmental etiologies of autism.

The environmental contributions to autism spectrum disorder (ASD) and their informative content for diagnosing the condition are still largely unknown. The objective of this study was to investigate associations between early medical events and ASD, as well as autistic traits, in twins, to test the hypothesis of a cumulative environmental effect on ASD risk. A total of 80 monozygotic (MZ) twin pairs (including a rare sample of 13 twin pairs discordant for clinical ASD) and 46 dizygotic (DZ) twin pairs with varying autistic traits, were examined for intra-pair differences in early medical events (for example, obstetric and neonatal factors, first year infections). First, differences in early medical events were investigated using multisource medical records in pairs qualitatively discordant for ASD. The significant intra-pair differences identified were then tested in relation to autistic traits in the remaining sample of 100 pairs, applying generalized estimating equations analyses. Significant association of the intra-pair differences in the MZ pairs were found for the cumulative load of early medical events and clinical ASD (Z=-2.85, P=0.004) and autistic traits (β=78.18, P=0.002), as well as infant dysregulation (feeding, sleeping abnormalities, excessive crying and worriedness), when controlling for intelligence quotient and attention deficit hyperactivity disorder comorbidity. The cumulative load of early medical events in general, and infant dysregulation in particular, may index children at risk of ASD owing to non-shared environmental contributions. In clinical practice, these findings may facilitate screening and early detection of ASD.