Is the slight risk of autism with an epidural important?


Two large studies have hinted at small increases in the risk of autism spectrum disorder (ASD) in infants born to mothers who received epidural anesthesia during childbirth, but their authors said it was not clear whether ‘there was a real association.

And an editorial in JAMA, where the studies were published, argued that while it’s real, it’s too small to matter.

One of the studies, conducted in the Canadian province of British Columbia, found a crude risk ratio of 1.32 (95% CI 1.24-1.40) in an analysis of nearly 400,000 births, which fell to 1.09 (95% CI 1.00-1.15) after adjusting for a variety of covariates.

The other, from denmark, gave a statistically significant association of similar magnitude before adjustments (HR 1.29, 95% CI 1.21-1.37), but became non-significant when other factors were considered (HR 1 05, 95% CI 0.98-1.11), despite including nearly 500,000 births.

Their respective authors both cast doubt on the reality of the association.

Gillian E. Hanley, PhD, of Vancouver General Hospital in British Columbia, and colleagues rated the apparent increase in ASD risk as “low” and said that although it had reached statistical significance ( barely), “the likelihood of residual confusion” was large enough to make the strength of the evidence weak.

And the Danish study authors, led by Anders Pretzmann Mikkelsen, MD, of Copenhagen University-Rigshospitalet Hospital, grasped the statistical insignificance of their analysis adjusted to oppose true association.

In both studies, the absolute difference in ASD rates was small: 1.53% vs. 1.26% in Canadian data, and 1.52% vs. 1.30% in Denmark, both before adjustments. As the study’s authors and editorial writers pointed out, this left a lot of room for doubt of an effect.

But what if it was for real? The 95% confidence intervals for the adjusted analyzes indicated that the risk of ASD increased from 10% to 15% could not be excluded. Some previous studies (although the literature is generally mixed) had also identified a potential risk of ASD with epidurals. Isn’t that another reason to avoid epidurals, especially given the other risks associated with them?

Not really, said Cynthia A. Wong, MD, and Hanna Stevens, MD, PhD, both from the University of Iowa in Iowa City, in the JAMA editorial.

“For individual patients and their delivery professionals, the benefits and risks of labor neuraxial analgesia, compared to alternatives, must be carefully weighed,” they wrote. The couple explained that epidurals are “unmatched by any other pain reliever option” and that their safety in terms of most results is well established. That said, they added, epidurals come with documented risks, including intra-lab fever, abnormal fetal heart rhythms, and prolonged labor for some women.

Due to recent studies, including those now published in JAMA, “patients and clinicians will legitimately ask whether an ASD risk should be included in this benefit-risk analysis,” wrote Wong and Stevens. The answer, they said, is no.

“Based on these studies and current knowledge about ASDs, it appears that concern for ASDs should not carry weight in this decision,” they said. “The current evidence does not support taking the risk of ASD into account when deciding whether to use (patients) or recommend (to healthcare professionals) occupational neuraxial analgesia.”

Study details

The two studies were similar in that they were retrospective reviews of medical records, but not identical. Hanley and her colleagues drew on data for virtually all women in British Columbia who gave birth to single children from 2000 to 2014, for a total of 388,254 births after excluding cesarean sections, premature births and lost children. of view before the age of 2 years. In this cohort, about 30% of infants were exposed to the epidural.

In the Danish study, Mikkelsen’s group analyzed the results of 479,178 infants born between 2006 and 2013. Less than a quarter of these deliveries involved epidurals. The exclusions were narrower than in the Canadian study, including emigration, infant mortality, and postnatal diagnoses like Down syndrome, totaling about 6,000 infants (compared to over 200,000 in British Columbia).

Adjustments in the Canadian study included year of birth, age of parents, neighborhood income, community size, maternal complications of pregnancy, parity, smoking during pregnancy, BMI maternal, onset of labor, gestational age, infant sex, infant size relative to norm, and presence of birth defects.

Because Mikkelsen and his colleagues included a wider range of births, some factors that excluded many infants from the Canadian study ended up being adjusted in the Danish study. These included year of birth, age of parents, gestational age, sex of infant, firstborn or not, elective cesarean delivery, small (but not large) for gestational age , induction of labor, maternal complications of pregnancy (but less than in the Canadian study), obesity, smoking status (not limited to pregnancy), family history of ASD, history of family psychiatric diagnosis, maternal education, employment maternal, region of Denmark and previous medical research behavior of the mother.

Another potentially important difference was that follow-up was somewhat longer in the Canadian study, around 9 years compared to 7 on average.

The Danish study found that firstborn status, induction of labor, region, year of delivery and infant gender were factors, in addition to epidural analgesia, most strongly correlated with subsequent diagnosis of ASD. Hanley and colleagues found that most of the statistical risk mitigation, relative to unadjusted data, occurred when they took into account pregnancy complications, parity, smoking during pregnancy, and pregnancy. Maternal BMI.

Both studies had the usual limitations of retrospective chart reviews. Additionally, while British Columbia and Denmark both have criteria for ASD diagnoses, there was no system to ensure clinicians were following them in every case. And while both studies attempted to fit known and suspected ASD covariates, the possibility of residual confusion remained substantial.

  • John Gever was editor from 2014 to 2021; he is now a regular contributor.


Both studies were funded by grants from the government and foundations.

All of the study’s authors and columnists said they had no relevant relationship with any business entity.


Comments are closed.