Cesarean Birth and Autism Spectrum Disorder

Background: Some studies have suggested a higher risk for autism spectrum disorder (ASD) after cesarean birth.  Such an association could stem from physical consequences of the operation itself or from the indications which led to the decision to perform the cesarean delivery.

Objectives: We hypothesize that children with autism spectrum disorder are more likely to have been the product of aberrant labor courses that result in non-elective cesarean delivery.

Methods: The CHARGE (Childhood Autism Risk from Genetics and the Environment) Study is an ongoing case-control study of the etiology of autism.  Data from maternal self-report and medical records documenting the course of labor and delivery were available for the mothers of 477 children with a diagnosis of autism spectrum disorder (ASD) and 272 population-based frequency-matched controls.  We collected demographic data and information about the pregnancy, delivery, and child’s early life in the Environmental Exposure Questionnaire, a telephone-administered interview.  Covariates related to medical conditions of pregnancy as well as the labor and delivery process were abstracted in a systematic fashion.  We performed the Autism Diagnostic Interview–Revised and the Autism Diagnostic Observation Schedule on study cases to confirm the diagnosis of autism.  Logistic regression was used to examine the relationships between delivery mode and autism status.  Because the risk of intraamniotic infection (IAI), use of antimicrobials, and cesarean rates and indications vary according to gestational age, we generated a Cox proportional hazards regression model with time-dependent covariates as a mechanism to evaluate the combined effects of IAI risk and antimicrobial use on mode of delivery.

Results: After adjustment for preeclampsia and / or diabetes and IAI risk, non-elective cesarean delivery following labor and / or ruptured membranes was more common among mothers whose children developed ASD compared with those who delivered vaginally (OR 1.71, 95% CI 1.04, 2.81).  There was no association between elective cesarean delivery and ASD.  A Cox proportional regression analysis examining the effect of IAI risk and antimicrobial use on the outcome of non-elective cesarean delivery and controlling for labor duration identified that women with both IAI risk and antimicrobial usage were twice as likely to have had a non-elective cesarean delivery compared to women who did not have IAI risk and did not take antimicrobials (OR 2.14, 95% CI 1.12, 4.07).

Conclusions: Our analysis suggests that it is not birth by cesarean itself that is associated with ASD.  Rather, it appears that factors associated with difficult labor courses, including prolonged labor and membrane rupture, as well as occult and overt infection within the amniotic cavity may drive the relationship between non-elective cesarean delivery and ASD.