Quick Study Analysis: Do Omega-6 Fats Cause Autism?

tldr; Seed oils are associated with Autism, but there is no definitive mechanism in this paper.

Background.

A number of folks on X have drawn my attention to this recently-released paper looking at metabolites of Ω-6 fats and Autism-Spectrum Disorder (ASD, “autism”), some have suggested that this paper is evidence of a link between seed oils and autism.

“Arachidonic Acid-Derived Dihydroxy Fatty Acids in Neonatal Cord Blood Relate Symptoms of Autism Spectrum Disorders and Social Adaptive Functioning: Hamamatsu Birth Cohort for Mothers and Children (HBC Study)” (Hirai, 2024)

(Hirai, 2024)

It’s an interesting paper, it builds upon the work done by Bruce Hammock (who discovered the soluble epoxy hydrolase (sEH) enzyme which is discussed in this paper. I interviewed him here:

“Omega 6 Linoleic Acid Research with Prof. Bruce Hammock, Prof. Bruce German, and Tucker Goodrich”

In a nutshell, avoiding a lot of complexity, the Cytochrome P-450 enzyme converts polyunsaturated fats (PUFA) into signalling/mediator molecules, known as epoxy fatty acids, which are then further converted by sEH into dihydroxy fatty acids (AKA diols). The source fats (substrates) can be either Ω-3 or Ω-6 PUFAs.

If you’re dying to learn more about this, I discussed it a bit more in depth in this post, concerning leukotoxin (an epoxy fatty acid derived from linoleic acid (LA)), and leukotoxin diol (derived from leukotoxin via sEH):

But it’s not necessary to understand that post to read this one.

The epoxy fatty acids and the diols can be both beneficial or harmful. Leukotoxin diol, for instance, is a normal part of our metabolism, and is helpful in promoting the growth of brown fat, but in excess will kill you via Acute Respiratory Distress Syndrome (ARDS).

New paper on Autism.

This paper looks at the association between autism and “cord blood” with diols produced via CYP450 and sEH.

This study examined the association between PUFA metabolites in the cord blood and ASD symptoms and adaptive functioning in children. (All quotes from Hirai, 2024, unless noted otherwise.)

Cord blood is blood taken from the umbilical cord when the baby is born, so it represents the nutrients and, possibly, contaminants or toxins, in the baby’s body before it is born.

They discuss two diols, 8,9-diHETrE and 11,12-diHETrE, which are both produced from arachidonic acid (AA), an Ω-6 fat which is either consumed in animal fat or produced from LA in seed oils. The body cannot produce this itself (it’s an ‘essential’ fat, AKA Vitamin F), so we know that somehow this fat is related to the diet of the mother.

They found that these two diols are related to two different aspects of autism. I say related because they took a single sample of blood at birth, and then followed these kids until they did or didn’t develop autism, measured via the ADOS-2 diagnostic questionnaire.

“We found a significant correlation between the ADOS-2 CSS [calibrated severity score] and AA-derived diol, and no other significant correlations were found.”

They have some interesting discussions on why this might be meaningful:

“Studies on mice have demonstrated that exposure to inflammatory cytokines during pregnancy leads to behavioral impairments reminiscent of ASD.”

Such cytokines, typically a response to infection or injury, can be produced via the CYP-450 enzyme and sEH.

“Polyunsaturated fatty acids (PUFAs), represented by arachidonic acid (AA) and metabolites, are key mediators of immune modulatory processes.”

So it’s plausible that PUFA could lead to autism. The most common PUFA in the human diet is linoleic acid from seed oils, which is converted to arachidonic acid, and consumption of LA has increased as autism prevalence has increased.

“The last decades have shown a spectacular and partially unexplained rise in the prevalence of autism spectrum disorders (ASD). This rise in ASD seems to parallel changes in the dietary composition of fatty acids. This change is marked by the replacement of cholesterol by omega-6 (n-6) fatty acids in many of our food products, resulting in a drastically increased ratio of omega-6/omega-3 (n-6/n-3).” (van Elst, 2014)

As noted by the authors of Harai:

“The total ratio of omega-6 to omega-3 in PUFAs was observed to be elevated in children with ASD.”

Unfortunately, this too is a correlation. So they have an interesting idea.

“A study also reported that … the gene encoding sEH, is high in the postmortem brains of patients with ASD… In addition, exposure to high concentrations of glyphosate during pregnancy and lactation induces ASD-like behavior in the offspring of pregnant mice, along with significant decreases in 8,9-EET levels and increases in sEH levels in the plasma, prefrontal cortex, hippocampus, and striatum. The administration of sEH inhibitors to pregnant mice exposed to glyphosate also improved ASD-like behavior, such as increased grooming time and social interaction deficits in the offspring. These findings suggest that the presence of CYP-PUFA metabolites during the fetal period plays an important role in generating a biological background that leads to the development of ASD-like behavior.”

Do they prove it?

The shortcomings of this paper.

So I’m certainly of the opinion that the massive increase in autism is related to the massive increase in seed oils. There are nearly a thousand papers listed in PubMed as a response to the search “autism oxidative stress”.

Search of PubMed: 985 papers.

The most common markers used to identify oxidative stress are the oxidized metabolites of the fats contained in seed oils.

But, does this paper prove, or even support, that case?

1. Arachidonic acid is a problematic causative pathway.
   Non-industrial societies with low intake of seed oils tend to have higher, not lower amounts of arachidonic acid in breast milk, while they do have lower linoleic acid, as expected. Additionally, arachidonic acid levels seems to be pretty tightly controlled in the body:
   “The percentage of AA was not significantly correlated with its precursor LA in either Tsimane or Cincinnati milks, suggesting AA synthesis from LA is not a limiting factor for milk AA content.” (Martin, 2012)
   AA doesn’t vary much in breast milk samples from around the industrialized countries (Brenna, 2007).
   So it’s hard to see how a dietary intake of AA or LA has driven AA levels to a problematic level.

2. These authors don’t find any relationship to LA.
   It’s not that they didn’t look; they found other connections between LA and poor outcomes in child development, just not with autism (Umeda, 2022). That doesn’t mean there might not be some pathway, but it’s not in the current paper.

3. It’s possible that some pathway other than dietary Ω-6 is upregulated which is driving excess conversion of AA to diols in autism.
   They describe a possible pathway involving glyphosate, which is plausible. Bruce Hammock, in fact, has explored that connection (Hashimoto, 2021). If that connection is a valid one, then it suggests that it is not seed oils that is driving this pathway. However, increasing prevalence of autism was noted long before glyphosate was introduced to the market in 1974 (Lotter, 1966), so it’s unlikely that’s the only factor.

4. Other papers have found conflicting results.
   “However, in humans, reduced cord blood diHETrE has been reported in boys with ASD, contradicting the present study’s results.”
   Ideally one wants to see fairly consistent results, with the understanding that variation in lab methods and human error can always introduce some variance.
   There just hasn’t been a lot of research done yet in this specific area.

5. There’s no longitudinal tracking of the biomarkers.
   It’s a single point-in-time data point. It’s interesting, but this is weak evidence from an epidemiological perspective. As the paper itself notes.
   “Importantly, we did not quantify CYP-PUFA metabolites at age 6, making it impossible to describe the association between the inflammatory state caused by postnatal diols and ASD symptoms and adaptive functioning. Verification of changes along the developmental trajectory is needed to understand the impact of the prenatal and postnatal inflammatory milieu.”

Conclusion

Autism used to be seen at a rate of 1 in 10,000, now it’s more like 1 in 44 (this paper) to 1 in 36, I have heard.

Clearly some environmental change is driving this rapid alteration. Glyphosate (or some other component of the pesticide Roundup) may well be be a partial driver.

It’s an interesting paper, but I don’t think it makes a case that excess consumption of seed oils is driving increased rates of autism.

I avoid foods containing seed oils and glyphosate, which seems like a wise course in general.

---

References

Brenna, J. T., Varamini, B., Jensen, R. G., Diersen-Schade, D. A., Boettcher, J. A., & Arterburn, L. M. (2007). Docosahexaenoic and Arachidonic Acid Concentrations in Human Breast Milk Worldwide. The American Journal of Clinical Nutrition, 85(6), 1457–1464. https://doi.org/10.1093/ajcn/85.6.1457

Gornoski, D. (2021, May 24). Omega 6 Linoleic Acid Research with Prof. Bruce Hammock, Prof. Bruce German, and Tucker Goodrich (No. 5/24/2021) [Mp3]. https://aneighborschoice.com/omega-6-linoleic-acid-research-with-prof-bruce-hammock-prof-bruce-german-and-tucker-goodrich/

Hashimoto, K., & Hammock, B. D. (2021). Reply to Reeves and Dunn: Risk for autism in offspring after maternal glyphosate exposure. Proceedings of the National Academy of Sciences, 118(2), e2016496118. https://doi.org/10.1073/pnas.2016496118

Hirai, T., Umeda, N., Harada, T., Okumura, A., Nakayasu, C., Ohto-Nakanishi, T., Tsuchiya, K. J., Nishimura, T., & Matsuzaki, H. (2024). Arachidonic Acid-Derived Dihydroxy Fatty Acids in Neonatal Cord Blood Relate Symptoms of Autism Spectrum Disorders and Social Adaptive Functioning: Hamamatsu Birth Cohort for Mothers and Children (HBC Study). Psychiatry and Clinical Neurosciences, n/a(n/a). https://doi.org/10.1111/pcn.13710

Lotter, V. (1966). Epidemiology of Autistic Conditions in Young Children. Social Psychiatry, 1(3), 124–135. https://doi.org/10.1007/BF00584048

Martin, M. A., Lassek, W. D., Gaulin, S. J. C., Evans, R. W., Woo, J. G., Geraghty, S. R., Davidson, B. S., Morrow, A. L., Kaplan, H. S., & Gurven, M. D. (2012). Fatty Acid Composition in the Mature Milk of Bolivian Forager-Horticulturalists: Controlled Comparisons with a Us Sample. Maternal & Child Nutrition, 8(3), 404. https://doi.org/10.1111/j.1740-8709.2012.00412.x

Umeda, N., Hirai, T., Ohto-Nakanishi, T., Tsuchiya, K. J., & Matsuzaki, H. (2022). Linoleic Acid and Linoleate Diols in Neonatal Cord Blood Influence Birth Weight. Frontiers in Endocrinology, 13. https://doi.org/10.3389/fendo.2022.986650

van Elst, K., Bruining, H., Birtoli, B., Terreaux, C., Buitelaar, J. K., & Kas, M. J. (2014). Food for Thought: Dietary Changes in Essential Fatty Acid Ratios and the Increase in Autism Spectrum Disorders. Neuroscience & Biobehavioral Reviews, 45, 369–378. https://doi.org/10.1016/j.neubiorev.2014.07.004