Labor Induction or Augmentation and Autism

Autism spectrum disorder (ASD) is characterized by social and communicative developmental deficits and repetitive, restrictive, or unusual behaviors 1. Autism spectrum disorder includes the diagnosis of autism, Asperger syndrome, and pervasive developmental disorder not otherwise specified 1 2. Approximately 1 in 88 American children is affected by ASD, although the reported prevalence varies widely, exhibiting a 4:1 male predominance and racial and ethnic variation 3. Although the cause of ASD is unclear, it demonstrates a strong genetic predisposition and multifactorial influences. A wide variety of exposures, including many perinatal factors, have been linked to ASD but the suggested associations in many cases are weak, inconsistent, or both among studies, and cannot be equated with a cause and effect relationship 4 5 6 7 8 9.

Recent research suggests a role for endogenous oxytocin in normal human social and cognitive behavioral development 10. Functional oxytocin deficiency and a faulty oxytocin signaling pathway have been observed in conjunction with ASD 11. Because exogenous synthetic oxytocin commonly is administered for labor induction and augmentation, some have hypothesized that synthetic oxytocin used for these purposes may alter fetal oxytocin receptors and predispose exposed offspring to ASD 12.

Studies to date that have investigated a potential link between oxytocin and ASD have a number of limitations, such as small size, retrospective data collection, and limited control for possible confounding variables. Such characteristics reduce the value of these studies and suggest the need for more research. Among nine studies summarized by Guinchat et al in a 2012 review 13, three studies 7 8 9 demonstrated a weak but significant association between labor induction and autism in univariate analysis, and six studies 6 14 15 16 17 18 found no such association. Only one study found an association that persisted after adjustment for potentially confounding variables 7. A 2011 meta-analysis found insufficient evidence to suggest an association between labor induction or augmentation and an increased risk of autism (odds ratio [OR], 1.21; 95% confidence interval [CI], 0.9–1.62) 19.

A larger 2013 analysis reported an increased odds ratio of autism among children born following induced or augmented labor 20. Data for this study were obtained from the North Carolina Detailed Birth Record and Education Research databases linked with children’s school records and included 5,500 children with autism (as indicated in their educational records) from among 625,042 live births. Among males, multivariate logistic regression showed a weak association between ASD diagnosis and labor induction and augmentation (induced and augmented [OR, 1.35; 95% CI, 1.1–1.66]; induction only [OR, 1.18; 95% CI, 1.08–1.3]; and augmentation only [OR, 1.15; 95% CI, 1.05–1.25]); in contrast, among females, ASD diagnosis was modestly associated with labor augmentation only (OR, 1.18; 95% CI, 1.02–1.36); but not with induction or both 20. If risks like these were substantiated in future work, it would represent an attributable risk of approximately 2 per 1,000 for male offspring (95% CI, 1–3 per 1,000) and 3 per 10,000 for female offspring (95% CI, -2–8 per 10,000) for induction or augmentation of labor 20.

Although the Gregory study suggested an association between ASD and labor induction or augmentation, the study design could not determine if such findings were truly a result of cause and effect. This was recognized by the authors, who noted that interpretation of their findings was limited by missing data regarding important potential confounders, the use of education as a proxy for socioeconomic status, and a lack of data regarding induction indications and methods. They concluded that the “results are not sufficient to suggest altering the standard of care regarding induction or augmentation…though additional research is warranted” 20.

Subsequent to its publication, the Gregory study has been criticized because of limitations in defining the exposure and the outcome of interest 21. Critics note that investigators did not know the specific individual or combination of agents that were used for labor induction or augmentation. The critics also note that the American Psychiatric Association reported an editorial error in the criteria listed for the diagnosis of pervasive developmental disorder not otherwise specified in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV); an error potentially leading to overdiagnosis during much of the time covered by the Gregory study 21.

In addition to recognizing limitations of available evidence, it is important to view any concerns about an association between labor induction or augmentation and ASD in the context of obstetric practice. Labor induction and augmentation are common in contemporary obstetric practice. Reducing indicated oxytocin use for labor augmentation or induction (eg, when delivery is deemed necessary or when labor is not progressing appropriately) would increase the cesarean delivery rate and could adversely affect maternal or neonatal health.