Nonshared environmental factors in the aetiology of autism and other neurodevelopmental conditions: a monozygotic co-twin control study.

Exposure information was collected retrospectively from parents and medical records, so recall or confirmation bias is possible. Birthweight agreement with medical records was excellent, but agreement was lower for some other risk factors, such as breech position and cumulative early medical events. Prenatal factors were not analysed because identical twins share the prenatal environment, so maternal medication and teratogen exposure could not be studied. The design was observational, so it cannot prove that any risk factor caused autism, ADHD, intellectual disability or lower IQ. Reverse causality could not be excluded, because the condition or related factors may have increased the chance of reporting some early medical events. The sample was selected to include many twins discordant for neurodevelopmental conditions, so it is not a representative population sample. Some specific exposures had very few discordant twin pairs, making those analyses underpowered. Some exposures were indirect indicators of medical problems, such as oxygen therapy for hypoxia or phototherapy for jaundice. Many comparisons were made, and some associations that were significant at P < 0.05 did not remain significant after correction for multiple comparisons.

The work was funded by Vetenskapsrådet, VINNOVA, Svenska Forskningsrådet Formas, Hjärnfonden, Forskningsrådet om Hälsa, Arbetsliv och Välfärd, Stockholm Brain Institute, Autism and Asperger Association Stockholm, Queen Silvia Jubilee Fund, Stiftelsen Solstickan, PRIMA Child and Adult Psychiatry, the Pediatric Research Foundation at Astrid Lindgren Children’s Hospital, Stiftelsen för Strategisk Forskning, Stiftelsen Sven Jerrings Fond, Svenska Frimurarorden, Stiftelsen Kempe-Carlgrenska Fonden, Stiftelsen Sunnerdahls Handikappfond, Jeanssons Stiftelser, EU-AIMS, the Innovative Medicines Initiative Joint Undertaking, the European Union’s Seventh Framework Programme, and Karolinska Institute. The supplied text lists funders but does not state author competing interests.

Do not read these twin-pair associations as proof that birth weight discordance, infant medication, or early illness caused autism, ADHD, or lower IQ. The study was observational, prenatal factors could not be analysed, reverse causality could not be excluded, and many specific exposures had too few discordant pairs for firm conclusions.