Abstract: Autism spectrum disorders encompass well-known behavioral and cognitive disturbances. Heritable and environmental factors may influence normal development or alter regulatory processes, with environmental factors perhaps exerting their effect during the prenatal and early postnatal periods. This study focused on the association between autism and labor induction or augmentation or both using a population-based data set of linked child educational and birth records from North Carolina to investigate whether birth induction/augmentation is associated with increased odds of being diagnosed with autism.
Demographic and obstetric data were obtained from all recorded live births occurring in North Carolina. Information on autism diagnosis was extracted from statewide educational records. Birth records from 1990 to 1998 were linked with educational data from the 1997 to 1998 to 2007 to 2008 academic years for the entire state (match rate, 74.4%). An individual met special needs eligibility criteria by having deficits in 3 of the 4 areas: communication, social reciprocity, behavior, or sensory. During delivery, a mother was either not induced or augmented (reference), induced only, or augmented only, or both induced and augmented. The association between induction/augmentation was examined in a range of models with such specifications as (1) only induction/augmentation and infant sex; (2) adding standard maternal-level characteristics; (3) adding maternal medical conditions and health behaviors affecting the pregnancy; (4) expanding model 3 to control for events of labor and delivery; (5) adding to 4 an ordinal variable for child birth year; and (6) adding an interaction between infant sex and induction/augmentation.
About 1.3% and 0.4% of male and female children had autism, respectively. For both sexes, the percentage of induced or augmented mothers was higher among children with autism compared with noncases. In model 1, a child whose mother was induced and augmented during delivery had 23% higher odds of being diagnosed as having autism than a child whose mother was neither induced nor augmented (odds ratio [OR], 1.23; 95% CI, 1.02–1.47). The ORs for the induced-only and augmented-only categories were 1.10 (95% CI, 1.01–1.19) and 1.15 (95% CI, 1.07–1.24), respectively. Compared with girls, boys had increased odds of autism diagnosis (OR, 3.04; 95% CI, 2.86–3.24). In models 2 to 4, including potential confounders related to socioeconomic status, maternal health, and pregnancy-related events/conditions did not significantly change the ORs for induction/augmentation from those in model 1. In model 5, a child whose mother was both induced and augmented at delivery had 27% higher odds of being diagnosed as having autism (OR, 1.27; 95% CI, 1.05–1.52). The ORs for the induced-only and augmented-only categories indicated a 13% to 16% increase in the odds of autism, respectively. Fetal distress and meconium were associated with autism with ORs of 1.25 and 1.22, respectively (95% CIs, 1.15–1.36 and 1.11–1.34, respectively).
Children born to mothers who were either induced, augmented, or both had increased odds of having autism. Further research should target the agents used during induction/augmentation and acute medical and obstetric events during labor. The results do not suggest altering the standards of care for induction or augmentation but do indicate that additional research is warranted.
Center for Human Genetics, Department of Medicine, and Duke Institute of Molecular Physiology, Duke University Medical Center, Durham, NC; Children’s Environmental Health Initiative, School of Natural Resources and Environment, University of Michigan, Ann Arbor, MI; and Children’s Environmental Health Initiative, Duke University; and Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, NC