Obstetric complications in autism: consequences or causes of the condition?

OBJECTIVE To determine whether and why obstetric complications are associated with autism. METHOD Obstetric histories, obtained at maternal interview and coded as an optimality score (OS), were compared in two groups: 78 families containing an autistic proband (ICD-10 criteria) and 27 families containing a down syndrome (DS) proband. The OS was examined in relation to offspring diagnosis, proband characteristics, and familial loading for autism and its phenotypic variants. RESULTS Autistic and DS probands had a significantly elevated OS compared with unaffected siblings, regardless of birth order position. The elevation was mainly due to an increase in mild as opposed to severe obstetric adversities. In autistic probands, the OS was best predicted by familial loading for autism and its phenotypic variants, but in the absence of this measure by the number of autistic symptoms. Among siblings of autistic probands affected with autism or its variants, the OS was best predicted by the probands' OS, and in its absence, by the measure of familial loading. In DS probands and siblings the OS was associated with increased maternal age, although this did not account for the OS elevation in DS probands. CONCLUSIONS Rather than playing any principal etiological role, the obstetric adversities associated with autism either represent an epiphenomenon of the condition or derive from some shared risk factor(s).