News The latest developments in autism research.

Large study links autism to autoimmune disease in mothers

by  /  22 August 2013

Brain-bound: Immune molecules in mothers who have children with autism target neurons in the frontal cortex, cerebellum and hippocampus of mouse brains.

About one in ten women who have a child with autism have immune molecules in their bloodstream that react with proteins in the brain, according to a study published 20 August in Molecular Psychiatry1.

Several research groups have found these immune molecules, called antibodies, in mothers of children with autism, and have shown that prenatal exposure to the antibodies alters social behavior in mice and monkeys.

The new study, which includes more than 2,700 mothers of children with autism, is the largest survey yet on the prevalence of these anti-brain antibodies.

“It’s a very large sample size,” says study leader Betty Diamond, head of the Center for Autoimmune and Musculoskeletal Disorders at The Feinstein Institute for Medical Research in Long Island, New York. The scale gives a clearer impression of the prevalence of these antibodies, she says.

Antibodies help the body’s immune system recognize and fight off disease-causing microorganisms such as bacteria and viruses, but sometimes the body mistakenly produces antibodies to its own proteins. In some people, this results in autoimmune diseases such as rheumatoid arthritis and lupus, in which the body attacks its own tissues.

Researchers say anti-brain antibodies do not harm the brains of the women who produce them because of the blood-brain barrier, a filter that prevents most molecules from entering the brain. But the immature blood-brain barrier of a developing fetus may let them through, allowing them to damage the brain and perhaps cause autism.

Diamond’s team also found that women who have autism-linked antibodies are more likely to have other markers of autoimmunity compared with those who don’t carry these antibodies. Studies have shown that women with an autoimmune disease also have an increased risk of having a child with autism2.

“This ties together the epidemiological finding that women who have autoimmune disease are more likely to have kids with autism, with the idea that there are actually antibodies against fetal brain in their serum,” says Paul Patterson, professor of biology at the California Institute of Technology, who was not involved in the work.

Sample size:

In the new study, Diamond’s team screened blood plasma samples from 2,431 mothers enrolled in the Simons Simplex Collection (SSC). The SSC is a registry of families with one child affected by autism and unaffected parents and siblings, and is funded by’s parent organization.

The researchers found that plasma from 260 of the women, or 10.5 percent, reacts strongly with mouse brain tissue, a signal that the blood contains anti-brain antibodies.

They also screened samples from 318 mothers enrolled in a different autism registry, the Autism Genetic Resource Exchange, and found that 28, or 8.8 percent, of them also have anti-brain antibodies.

In contrast, among a group of 653 controls drawn from the general population of women of childbearing age in New York City, only 17, or 2.6 percent, carry the autism-linked antibodies.

This means that the prevalence of anti-brain antibodies is about four times greater among mothers of children with autism than among the controls.

However, “The control population could have had mothers of children with autism in it,” notes Judy Van de Water, professor of clinical immunology at the University of California, Davis, who was not involved in the work. In fact, some of the women in the control group may not have children at all.

Still, Van de Water says, the study “corroborates, from an independent investigator, that women in this population have a higher incidence” of the antibodies.

Van de Water has found anti-brain antibodies in autism mothers who are part of the Childhood Autism Risk from Genetics and the Environment study using a different methodology3, known as Western blotting, and says that the differences in study design make the similar results more persuasive. “It’s further support for this potential pathway to autism,” she says.

However, she cautions that measuring reactivity to mouse brain tissue also has drawbacks, because of the action of chemicals used in preparing the brain tissue for analysis. “The perfusion and fixation process that you have to go through is going to alter some of the antigens,” the proteins to which the antibodies bind, she says. This may ‘mask’ some of the binding patterns that occur in live brains.

Using the mouse brain enabled Diamond’s team to pinpoint which regions of the brain the antibodies affect. They found that the antibodies bind primarily to neurons in the frontal cortex, hippocampus and cerebellum, all areas that have been implicated in autism.

Fetal findings:

Patterson notes that the researchers tested binding of the antibodies to adult mouse brain, but protein expression in the adult brain is often very different from that in the fetal brain. “Optimally, one would like to know what the binding is to fetal human brain or fetal monkey brain,” he says.

According to Diamond, the team did find that the antibodies also bind to fetal human brain extracts and to whole fetal mouse brain. However, she says, it wasn’t possible to see which regions of fetal mouse brain they bind to because the brains are too small.

Her team performed several other analyses that uncovered tantalizing links between autism-linked antibodies and autoimmunity more generally.

First, they screened the blood of study participants for anti-nuclear antibodies (ANAs), which bind to molecules found in the cell nucleus and are commonly found in people with a variety of different autoimmune disorders. They found that 52 percent of autism mothers who carry anti-brain antibodies also have ANAs. In contrast, only 13.4 percent of autism mothers without anti-brain antibodies have ANAs, as do 15 percent of controls.

Mothers with anti-brain antibodies are also more likely to have autoimmune diseases, especially rheumatoid arthritis and lupus, compared with mothers who don’t have these antibodies.

Finally, the team screened a group of 363 women with rheumatoid arthritis for autism-linked antibodies. They found that 13.5 percent of these women have anti-brain antibodies, similar to the prevalence among mothers of children with autism and much higher than that of controls.

That’s a rather puzzling finding, Patterson says. “If they’re just as likely to be found in rheumatoid arthritis women, that means they’re not especially specific to autism,” he says.

But a link between rheumatoid arthritis and autism isn’t unheard of, Diamond says. In a separate series of experiments, she has found that antibodies that cause kidney damage in women with lupus also cause cognitive impairment in mice exposed to the antibodies in utero4.

Diamond says she has unpublished data indicating that some of the brain proteins that bind the antibodies are already implicated in autism or other neurodevelopmental disorders — but declined to reveal details. Are any of the targets she has identified the same as those reported by Van de Water and her colleagues in July? “It’s sort of interesting,” Diamond says. “They’re not.”


1: Brimberg L. et al. Mol. Psychiatry Epub ahead of print (2013) PubMed

2: Atladóttir H.O. et al. Pediatrics 124, 687-694 (2009) PubMed

3: Braunschweig D. et al. Transl. Psychiatry 3, e277 (2013) PubMed

4: Lee J.Y. et al. Nat. Med. 15, 91-96 (2009) PubMed

36 responses to “Large study links autism to autoimmune disease in mothers”

  1. ASDDad says:

    I thought the very last paragraph intriguing.

    “Diamond says she has unpublished data indicating that some of the brain proteins that bind the antibodies are already implicated in autism or other neurodevelopmental disorders — but declined to reveal details. Are any of the targets she has identified the same as those reported by Van de Water and her colleagues in July? “It’s sort of interesting,” Diamond says. “They’re not.”

    That raises the possibility that the percentage of mothers with fetal antibodies could go higher than the reported 23% UC David MIND – Van de Water

  2. Steve White says:

    But on the other hand, it seems like Dr. Diamond is finding less maternal antibody related autism overall and different antigens, and not so tight a link between the antigens and autism. Did Dr. Van de Water miss those Diamond has found or vice versa? I feel confused right now.

  3. Ashley says:

    Interesting, My son has autism and I have an autoimmune disease. How about them apples.

    • poo poo says:

      Have you been tested for those specific antibodies which cause autism? I doubt it. You should get checked.

    • Steve White 2 Ashley says:

      According to a presentation by Dr. Van de Water, if I understood it, there is a genetic variant most of the mothers with antibodies to brain have. In fact, nearly all of them have, which also ties with the tendency for people to get autoimmune diseases.
      I think it makes sense to try to get the antibody test if you can, if there is any chance of more children. I am not sure when they estimate it will be commercially available but if you do have the antibodies, pregnancy would be very risky according to the folks at UC Davis Med School. Maybe they have info on how to get the test in some other way.

      • Ashley says:

        I’m done having kids, but I’d do it for research. I already am part of a research program for my autoimmune disease, what’s one more 🙂


        • Jewls says:

          Hi Ashley,

          Is there a specific test for autoimmune disorders; I have been through almost every test possible and doctors can not seem to find where my severe pain is coming from (severe right and left flank pain; shooting pain going into legs and arms, sore joints, I have a very hard time moving my head all the way back; digestive problems). I am working with a specialist right now but everything comes back normal.

          • Ashley says:

            I’m not sure on specific tests. But it’s my understanding that if it were an auto immune disorder it might be more uniform? It affect your right side as well? I don’t know for sure though. That’s frustrating they can’t figure it out. It’s not nerve related?

          • Sandy Ryers says:

            There are several tests for specific autoimmune disorders but you can also test false negative and false positive. You should look into getting a CT of your kidneys, seeing a geneticist, having a full spine MRI, and looking into Ehlers-Danlos Syndrome, a connective tissue disease.

  4. Parent of a recovered child says:

    It is high time someone ran a treatment trial on IVIG alongside plasmapheresis in autism!

    Presence of autoantibodies (both in mothers and their affected children) may well turn out to be a good predictor of responsiveness, a ‘subtyping’ biomarker, for such treatment/s. There are at the moment so many medical interventions that seem to be very helpful for a small percentage of affected individuals, including IVIG or steroids. However the number of ‘responders’ never reaches statistical significance in studies, precisely because no treatment trials so far have managed (or even attempted) to subgroup their patients into biologically-based subtypes.

    As long as treatment trials continue to recruit ASD subjects according only to their superficial diagnosis of autism, parents and doctors will be left in the dark as to what treatments might be best suited for their individual child.

    The result of this study, and the preceding ones by MIND institute, will be truly ground-breaking only once they are used to inform treatment (and prevention) in real life situations. Otherwise they could easily join the dust-gathering and ever-growing pile of ‘fascinating’ autism findings.

    • Payman says:

      You are so right in every word. I have the same philosophy. We are being misled by this diagnosis, during any clinical studies related to autism, hardly every the autism group is subgrouped itself for any meaningfull data analysis.

      Can we please keep in touch? Did you ever try IVIG?


  5. RA Jensen says:

    Celiac disease is an autoimmune disorder. Premature deliveries and low birth weight is associated with autism risk. Celiac disease, especially if untreated, appears to increase the risk of repeated miscarriages and premature deliveries, and impaired fetal growth with reduced birthweight The recent Danish study that found a 7% recurrance risk risk for autim is interesting since untreated Celiac disease in mothers is associated with increased recurrance of low birth weight and preterm deliveries.

  6. Steve White says:

    Some wild speculations by a layman parent:

    Recurrence rate is of great interest I think.

    It seems to me any large pool of school children can give you a very good recurrence rate, (though you have to do genetic testing to be assured kids who are said to be full siblings actually are). Just look at all the kids who are in special ed who have a sibling in special ed.

    It also seems to me compiling a very good database for this is potentially of great value, because you can run some numbers and test out various theories of causation.

    Or maybe I should say, set limits on the maximum possible frequency of certain causes.

    Although these are not good examples, 7% recurrence would tell you there are no more than 14% of cases of autism caused by dominant single genes. And no more than 28% of cases could be caused by two independent genes interacting dominantly, and so forth.

    Of course no one thinks it’s that simple, but more complex analysis can be done to try to narrow it down.

    The frustrating thing is the numbers do not seem to line up enough to do this kind of analysis with any confidence.

    • Payman says:

      Please elaborate the confidence analysis you are refering to? what numbers you are trying to line up here. In any research, if you have any science background…you dont try to line up anything…in general the data analysis talks for itself and not the other way around. Thanks for the post

  7. Steve White says:

    One more quick thought. I saw a study on PubMed showing fraternal twin autism concordance was much higher than sibling concordance. If I remember right, the concordance was higher than would have been explained by maternal antibodies. The latest numbers on maternal antibodies might have changed that but it was quite high concordance.

  8. misty says:

    I have a son on the Spectrum—it has been found over the years that I have elevated numbers on certain autoimmune tests, but no diagnoses outside of Raynaud’s Syndrome. My son, however, has EXCESSIVE autoimmunological responses to little 24 hour bugs. his body attacks itself (and usually puts itself in remission as well.) The last episode attacked his kidneys and spiked his BP. I am wondering if other children have this in common on the spectrum.

    • SAM says:

      I too have a kid on autism, non verbal, more of a regressive autism.

      The allergies or allergy like symptoms seem to be the trigger for bigger inner body reaction which leads to a regression state where my son not only feels well but his behavior and state of mind degrades. He feels this cycle, since he feels it and at its initial state he will go for medication cabinet (he is non verbal) and reaches for medication so that we can administer him. But we dont know why this is happening and why these regressions pick up at night. But when he is in an progression mode, he seems as a normal boy.

      Why this two states of mind? How do you make other researchers know of this facet of autism?

  9. Lynn says:

    Lightbulb moment! I have 2 children on the spectrum-both high functioning PDD-NOS and I have Hashimoto’s it too is an autoimmune disorder that has plagued me for 18 yrs first revealing itself after the birth of my first child . Then a “Dr” that treated me for two years and pronounced me “cured” and took me off the meds to treat the high levels of TSH in my system. This awesome disease once again revealed its ugly face yes you guessed with the other child. Both of who are the ones on the spectrum. We have one unaffected child -my thyroid function remained normal during this pregnancy. I believe that it is my genetics as the oldest has a different father than the youngest.I know it wont change anything but if they were to find a direct connection it would give alot of answer and maybe treatment options . So glad i came across this article.

  10. Terry says:

    Does anyone know where one can get tested for these antibodies?

    • gregboustead says:

      Hi Terry,

      I spoke with the researchers of the main study above, and they say that there isn’t a screen available yet to test for the discussed antibodies. This is in the discovery stage, and the assay has not been validated for clinical use. They further warn that an assay like this would, at best, determine a probability and not a certainty, in any case.

      community manager,

  11. Sarah says:

    The fact that many autism mothers have an auto immune problem is a valuable clue that perhaps their children also have a fragile immune systems. So why are we vaccinating these potentially vulnerable children from day one afterall vaccines are designed to provoke the immune system. Perhaps these children cannot tolerate the load. Perhaps all these vaccines are tipping the child’s immune sysem over the edge into dysfunction. Perhaps there should be a protocol in place to pre-screen these infants based on family history of
    autoimmune, neurological and neuropsychiatric disorders prior to vaccinating.
    Why is noone addressing this?

    • Doceorcecat says:

      What part of the MOTHER’S immune system causing the disorder didn’t you get?

      • Brooke says:

        Oh gosh,cmon,the apple doesnt fall far from the tree.Auto immune disorders run rampid in our family and i vaccinated my vulnerable child and boom overload.What dont you get? IMO its the innability to handle the toxins thats hereditary. Funnier mom gets the recommended flu shot to protect herself and her baby right? With a mouth full of amalglams.There’s a clue.

    • Kip says:

      Your logic is bad, and you should feel bad.

  12. Sarah says:

    BTW, always looking to cover themselves, the CDC commissioned study below which only models outcomes using a stastical modeling method called conditional logistic regression. The researchers never actually took blood samples from the children with autism (compared to controls) to see how their immune systems (antibodies) were reacting to each shot.

    “Increasing exposure to antibody-stimulating proteins and polysaccharides in vaccines is not associated with risk of autism.”

  13. Vicky says:

    I’m a bit skeptical about this just because it is hard to tease apart potential influences on brain development. I am currently 15 weeks pregnant with hashimotos and know that t4 levels are central to brain development. If these we’re not properly managed then brain development is affected. So just because a woman has autoimmunity does not mean it is the antibodies that are causing the issues. The same with type 1 diabetes – glucose levels can influence brain development. Plus, there is the placenta which I have heard is a barrier, therefore applying blood directly to brain tissue is perhaps not representative of real life.

    • Peaches says:

      Vicky, Moms Antibodies cross the placenta to the baby. So, you don’t have to apply blood directly to the baby’s brain. Yeah, the placenta can keep some molecules out, but IgG can cross and does. 14

  14. Anonymous says:

    I just don’t buy it. Every female in my family has an autoimmune disorder. We have no family members with autism,and I have a big family. Sure it may play a small part, but publishing the results of this like it is fact is going to cause a lot of pregnant women unnecessary worry.

    • Kip says:

      It doesn’t say that all autoimmune disorders result in autism. It only says that some antibodies present from autoimmune diseases occur commonly in mothers of those children with autism.

  15. Brooke says:

    While I dont discount an immune dysfunction issue. Ive seen it in my 5 yr old with ASD,it seems its always something such as overweight,birth control,maternal age etc.Every system in these kids are screwed up.I know im going to get knocked for this but metals seem to be the only likely scenerio. When you really study what these toxins do to the body it’s clear.Im not a Dr and i certainly dont know it all but you don’t have a genetic epidemic.

  16. Candace L says:

    Our son is going through experimental treatment for a rare syndrome called ROHHAD. After chemo and IVIG his autism improved drastically! In fact, almost as if he was a different child. Testing to show the progress and improving of behavior and cognitive ability was done by Kennedy Kriegers top doctors.

    John’s Hopkins believes ROHHAD may be caused by an autoimmune response and chemo resets that response. And so far… They are right. There have been huge improvements in many areas but the greatest improvement is in the autistic behavior area.

  17. Lauren says:

    I think it’s odd that women can make antibodies for their own brain without having brain cells come into contact with their immune system. The brain is an immunopriviledged site people! You can’t make antibodies for your child’s brain without making them for your own brain!

  18. j franklin says:

    yes they are connected, i have hashi`s my mum had it and RA and add celiac to that
    and what do you get a mess. both my sons are autitic one has aspergers and i have dyslexia. so their you have it .i have auto-antibodies to the thyroid, but have not be tested for the RA and celiac. undiagnosed hypothyroidism in pregnancy causes the auto-antibodies to attack the thyroid and stop vital hormones reaching the fetus. and other things may be in play like the RA and celiac and other autoimmune condisions.

  19. Secret Name says:

    Change the link to the thing saying autism and R. Arthritis are connected. I have heard of the link before but the study you quoted does not support the association

Leave a Reply

Your email address will not be published. Required fields are marked *