No Link Found Between Epidural Use and Autism

Tara Haelle

April 20, 2021

Exposure to epidural analgesia during labor did not show a link to a later diagnosis of autism spectrum disorder (ASD) in a population-based cohort study published April 19 in JAMA Pediatrics.

Though the initial analysis showed an association, adjustment for a wide range of demographic, medical, and birth factors eliminated the link. The authors note that their findings contrast with those of a cohort study in California published in the same journal last year.

"It is possible that residual confounding explains this positive association because key perinatal variables, including induction of labor, labor dystocia, and fetal distress, were not included as confounders in that study," write Elizabeth Wall-Wieler, PhD, of the University of Manitoba in Winnipeg and her colleagues. "To limit potential bias from unmeasured confounders, we included the aforementioned variables within a wide set of potential confounders."

The researchers analyzed linked datasets from all singleton infants born in a hospital from 2005 to 2016 in Manitoba, Canada, to compare use of epidurals during birth with diagnoses of ASD before 18 months of age. The four data sources included the Statistics Canada, Manitoba Education, Manitoba Families, and Manitoba Health, Seniors and Active Living, which includes the Manitoba Health Insurance Registry, Medical Services, Hospital Abstracts, and Drug Program Information Network. The researchers excluded women with cesarean deliveries because it was not possible to differentiate between scheduled and unscheduled cesarean deliveries.

Among 123,175 children born to mothers with an average age of 28 years, 38.2% had been exposed to epidural analgesia during their labor. Autism diagnoses occurred among 2.1% of those exposed to epidurals and 1.7% of those not exposed to epidurals. After the researchers controlled for a range of potential confounders, the difference became nonsignificant (hazard ratio, 1.08).

The adjusted analysis accounted for mother's age; high school degree; marital status; neighborhood socioeconomic status; receipt of public assistance during pregnancy; and presence of diabetes, hypertension, anxiety, or depression in the year before the birth. Other covariates included in the adjustment included the following pregnancy factors: "parity, gestational diabetes, gestational hypertension or preeclampsia, self-reported and diagnosed drug use, smoking, alcohol use, premature rupture of membranes, antepartum hemorrhage, infection of the amniotic sac and membrane, urogenital infection, antenatal mental health hospitalization, hypothyroidism, benzodiazepine use, antidepressant use, and antiepileptic use." The researchers also included birth year, labor induction or augmentation, labor dystocia, fetal distress or macrosomia, gestational age at birth, the infant's sex, and hospital type.

"There were substantial differences in maternal sociodemographic, preexisting, pregnancy-related, and birth-specific covariates between births who were exposed vs. nonexposed to epidural labor analgesia," the authors report. "For example, births exposed to epidural labor analgesia were more likely to be nulliparous, have premature rupture of membranes, antepartum hemorrhage, induction of labor, augmentation of labor, and fetal distress."

To take family history of ASD into account, the researchers also compared siblings who were and were not exposed to an epidural during labor: 80,459 children in the cohort had at least one sibling in it as well. The researchers still found no association between use of an epidural and a subsequent autism diagnosis (HR, 0.97). The authors conducted several sensitivity analyses for first-born children, those with two or more diagnostic codes for ASD on different days, and women with missing data on high school completion or marital status who delivered at 37 weeks of gestation or later; these results consistently showed no association between epidurals and ASD.

The findings are important but unsurprising, said Scott M. Myers, MD, a neurodevelopmental pediatrician and associate professor at the Geisinger Commonwealth School of Medicine's Autism & Developmental Medicine Institute in Scranton, Pa. Myers, who was not involved in the study, said it was strengthened by the inclusion of a wide range of covariates and multiple sensitivity analyses.

"It confirms the suspicion of many experts who were skeptical of the association reported previously, that the small increase in ASD in offspring of mothers who had epidural labor analgesia was likely attributable to other factors that differed substantially between the exposed and unexposed groups," Myers said in an interview. "The plausibility of exposure to epidural analgesia in labor having a large effect on ASD risk and accounting for changes in ASD prevalence over time is low."

It's possible to hypothesize about subgroups that are genetically susceptible to certain environmental risk factors, including epidurals, but such an association should show up in epidemiological research if the subgroup is large enough.

"For example, epidural labor analgesia can prolong labor, and if it were a significant risk factor for ASD, one might expect that longer labor would have been demonstrated to be associated with increased ASD risk, but this has been examined and is not the case," he said. He also noted that other perinatal factors previously linked to ASD, such as cesarean delivery, may result from a shared factor that affects risk of both ASD and cesarean delivery.

"Although there haven't been enough systematic postmortem brain studies to be certain that the findings are generalizable, the most consistent neuropathological findings associated with ASD clearly arise long before birth,"Myers said. "The information I would provide to a concerned pregnant mother is that the current weight of the evidence does not suggest an association between epidural analgesia in labor and increased likelihood of ASD in offspring, much less a causal association."

Clay Jones, MD, a hospitalist specializing in neonatal-perinatal medicine at Newton (Mass.)–Wellesley Hospital, was not involved in the research and offered a similar assessment of the findings.

"Our understanding of autism is that it is more of a genetic condition which interferes with the organization of brain architecture, so the evidence for any environmental cause would need to be robust for it to change medical practice or our recommendations to the general public," Jones said in an interview. Compared to the previous California study, "this new research is larger and better accounts for confounding variables that might increase the risk of a child eventually being diagnosed with autism," he said.

While recognizing the value in conducting studies to uncover any potential environmental factors contributing to autism diagnoses, Jones also addressed science communication challenges related to this research.

"While many of these studies are valid early efforts at honing in on potential risk factors, they can be overhyped and lead to increased patient anxiety and potentially harmful changes in behavior," Jones said. "There is already a significant amount of pressure for many women to avoid certain safe and effective pain reduction strategies during labor, such as epidural labor analgesia. This pressure is often based on misunderstandings of the risks, pseudoscientific beliefs regarding the benefits of so-called 'natural childbirth,' and blatant misogyny. I hope that this new study helps to reassure women that it is okay to request to be more comfortable during their labor experience with the help of epidural labor analgesia."

The authors of the study also noted the benefits of epidural use during labor.

"It is recognized as the most effective method of providing labor analgesia," they write. In addition, "the presence of an indwelling epidural catheter allows epidural anesthesia to be administered for an unplanned (intrapartum) cesarean delivery, thus secondarily avoiding any maternal complications or fetal exposure from general anesthesia."

JAMA Pediatrics editor Dimitri A. Christakis, MD, MPH, wrote his second-ever Editor's Note about this topic after the journal published two similar studies with different conclusions.

"Because there will never be experimental studies of environmental exposures, we are left with imperfect observational studies that are always at risk for residual confounding, especially when observed effect sizes are small," Christakis writes. "Science is an imperfect and iterative process, and our responsibility as journal editors is to manage the process as best we can. Publishing two conflicting studies in such a short time frame serves as testament that we recognize the process for what it is." His personal opinion is that any association has yet to be definitively established but that the journal will publish the study if a more definitive one is done.

In considering potentially contributing environmental risk factors to ASD, Gillian E. Hanley, PhD, of the University of British Columbia in Vancouver and two colleagues write that "meta-analyses have been unable to identify a single perinatal and neonatal factor associated with ASD risk, although some evidence suggested that exposure to a broad class of conditions such as fetal presentation, umbilical-cord complications, fetal distress, or multiple births, reflecting compromised neonatal health, may increase risk."

Yet, they add, these studies are inconsistent in their effect size, likely because of differences in study methodology, comparison groups, sample size, diagnostic criteria, and exposure assessment.

"Thus, we continue to ask questions about whether biologically plausible associations exist or whether associations reflect residual confounding related to yet-to-be-determined maternal genetic or environmental factors," Hanley and her colleagues write. They discuss precise differences between the California and Manitoba studies and the inevitability of selection bias since people who choose an epidural will differ in other ways from those who don't.

"Epidural labor analgesia is an extremely effective approach to obstetric analgesia, and we have a collective responsibility to understand whether it is a safe option that sets a healthy developmental pathway well into childhood," Hanley and her colleagues conclude. "Women have the right to make a truly informed choice about their pain relief during labor."

The research was funded by the Canadian Institutes of Health. One author reported receiving personal fees or grants from Aetion, Alosa Foundation, Lilly, GSK, Pacira, and Takeda. No other authors had disclosures. Jones, Myers, and the editorial authors had no disclosures.

This article originally appeared on, part of the Medscape Professional Network.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: