A new study published in the sci-glam journal Science purports to point to a way to prevent or “treat” autism in utero by way of oxytocin-related therapeutics. But where the findings really point, if one takes them literally, is away from scheduled c-sections.
During delivery, rodent infants undergo a shift in chloride ion signaling in the brain, thanks to the action of oxytocin, which also drives uterine contractions. This inhibition, these study authors say, goes wonky in the brains of baby rats that have been exposed to valproic acid, which is used to treat epilepsy but also has been associated with increased autism risk following in utero exposure. The inhibition also fails in the brains of baby mice with the rodent version of human Fragile X.
Both valproic acid and Fragile X influence the activity of the GABA-based system in neurons, a signaling pathway that regulates the movement of chloride ions in and out of cells. A heart failure drug and diuretic, bumetanide, also affects GABA signaling, possibly serving as a corrective where oxytocin presumably is unavailable or unable to act. In this new study, Tyzio et al. report that giving rodent mothers bumetanide around the time they deliver valproic acid-exposed or Fragile X pups improves the ability of their pups to vocalize. They also report that blocking oxytocin in the mothers worsens this vocalization ability and the pups’ ability to find their home nest.
Their interpretation is that these findings
validate the clinical actions of bumetanide and oxytocin and emphasize the importance of investigating how and when developmental sequences are disrupted in animal models of autism in order to develop novel therapeutic avenues.
In an article at the Indepdendent that hilariously asks in its headline if autism “could be caused at birth by salt in the nerves?,” study lead author Yehezkel Ben-Ari, takes the paper conclusions even further, claiming that
Chloride levels during delivery are determinants of the occurrence of autism spectrum disorder.
Well, that’s a new one.
Ben-Ari’s eagerness could arise from an interest in confirming some mild findings from his own small clinical trial of bumetanide in autistic children ages 3 to 11 (well after birth, obviously), results that one researcher dismissed as “very problematic” in terms of outcome measures.
In an editorial accompanying the publication, however, Andrew Zimmerman and Susan Connors aver that “the findings raise the possibility of preventing the autistic phenotype in offspring by predelivery pharmacological treatment” because in examining “two seemingly unrelated rodent models of ASD” the authors have found a common, GABA-based pathway between them. Zimmerman and Connors write:
With this new insight into a convergent pathogenic mechanism downstream from different etiologies, we may now begin to understand the variability, as well as sameness, among people with ASD and related disorders.
Except that …
Fragile X is associated with about 5% of autism cases. Valproate exposure in utero is linked to about a 4% increased risk of autism. That’s a lot of autism remaining that’s not linked to either of these factors, both of which were not “seemingly unrelated” when it came to involvement in GABA signaling. That was alreadyknown. It’s not shocking or amazing that pharmaceutical effects can mimic genetic effects–that is, indeed, the basis for much therapeutic intervention, but this work doesn’t give us much insight into any “unifying theory of autism,” which likely doesn’t exist. It also ignores a large body of evidence suggesting early gestational in utero influences and, well, a lot of non-Fragile X genetics. Nevertheless, watch as the news media breathlessly takes up the take-home being pushed here and cites a “switch” at birth for autism, “salt” involvement in autism, “cure” in the offing for autism, etc. In an odd twist, prenatal exposure to valproic acid can also enhance social behavior in rodents, which takes me to …
Rodents are not people, and the ability to vocalize or to do so at an appropriate frequency is a mouse problem, not an autistic person problem. Autistic people can make sounds. The issue with many nonspeaking autistic people isn’t whether or not they can make sounds, and certainly many appear to have perfect pitch. The issue is with their motor coordination for forming words. That’s very different from simple vocalizing. Autism is a sociocultural construct based on human communication differences, one that shows sex-based differences and is subject to cultural interpretation. If we’re going to model a complex social behavior that involves complex species-specific communication and culture, perhaps we should turn to animal models with those features to make some real headway. Dolphins? Corvids? Dunno.
The alleged implications of this work are that without the action of oxytocin during birth, autism is a risk because the oxytocin doesn’t exert its necessary, protective effects on the GABA system. This ‘brain-protective’ action of oxytocin has been established in rodents. It has not been established in humans. If it were relevant, one would expect a strong autism signal among children born via planned cesarean section, as that generally doesn’t involve oxytocin influence. The relationship, if any, is ratherweak, and so many other factors might influence the decision to schedule a c-section that teasing them away as involved in autism hasn’t been done.
Just as scheduled c-sections should be a big correlate with autism if delivery-fueled oxytocin influence is important, the use of oxytocin during labor should be associated with reduced autism risk. It’s not. Indeed, as the authors of the editorial note: “The use of exogenous oxytocin for the initiation or augmentation of labor in humans has been the focus of much speculation as a possible cause of ASD. For example, oxytocin has been associated with increased odds of ASD, especially in male children.” They go on to try to efface this link by noting that “it is possible that pregnancy conditions that lead to the administration of oxytocin may predetermine abnormal development of GABA-associated physiology. It is also possible that improved obstetric and neonatal care allow survival of infants with preexisting brain damage.” That’s a bit of contorting to get these conflicting data to wrap around the oxytocin-is-good-for-autism concept. Perhaps this might have helped them. Overall, studies seem to suggest that oxytocin administration during labor doesn’t influence autism risk one way or the other.
We have no way of identifying autism in utero, claims of accurate tests notwithstanding. As Zimmerman and Connors note in their editorial, using butenamide in this way would require figuring out who’s autistic before they’re born, which they say, would be difficult because autism symptoms “often do not appear until the second year of life.” I’d say rather that they’re often not noticed until then. At any rate, in the absence of confirming an autism diagnosis before birth, would the expectation be that all women should just be administered butenamide at delivery, in spite of the huge variation in autism manifestations and likely etiology and the fact that some attributes of autism can, indeed, be positive? Based on these findings in rodents, should women ask for induction and augmentation for delivery just in case? Should women avoid c-sections at all costs if the prospect of having an autistic child frightens them? That last is the real take-home of this study if you take “autistic-like behaviors in rodents” as a serious measure of autism itself. I’d suggest leaving that one in the lab, along with the rodents.
Forbes Forbes is among the most trusted resources for the world's business and investment leaders, providing them the uncompromising commentary, concise analysis, relevant tools and real-time reporting they need to succeed at work, profit from investing and have fun with the rewards of winning.