Reviews Exploring the intersection of autism and the arts.
Opinion / Reviews

Book review: Autism’s twisted immune links, untangled

by  /  2 January 2012

In the summer of 1917, an Austrian doctor named Julius Wagner-Jauregg gave a man with advanced symptoms of syphilis, including psychotic episodes, a very peculiar treatment: malaria.

The parasitic infection causes fever, and Wagner-Jauregg had a theory that fever treatment, or ‘pyrotherapy,’ could cure psychosis. The doctor injected the man with blood drawn from a soldier with malaria. Within weeks, as expected, this brought on episodes of fever. After the sixth fever, the man’s psychotic fits ended, and he eventually recovered completely from both syphilis and malaria. Wagner-Jauregg repeated the experiment on nine more men with psychosis, and six of them improved.

This marked the first time that scientists had identified a physical treatment for a mental disorder — a feat that earned Wagner-Jauregg a Nobel Prize. To this day, scientists don’t know the mechanisms underlying his success.

Despite a century-long history, we rarely hear about the link between the immune system and mental illness. But the science behind it is rich and varied — from massive epidemiological studies of twins and pregnant women, to the screening of immune molecules in amniotic fluid and postmortem studies of brain inflammation.

For the non-expert, this field can be more intimidating than a box of jumbled Christmas decorations. In his new book, Infectious Behavior: Brain-Immune Connections in Autism, Schizophrenia, and Depression, biologist Paul Patterson nimbly untangles the strings of lights.

For the past decade, Patterson, professor of biology at the California Institute of Technology, has conducted animal studies to tease out the molecular mechanisms underlying immune links to autism.

For example, he found that when pregnant mice receive a mock infection, their pups later show an array of abnormal behaviors. The book sometimes explores his own work, but mostly focuses on the large body of indirect evidence supporting the idea of an immune component to autism, schizophrenia and other psychiatric disorders.

The most famous data for this link come from some six decades of epidemiological studies showing that when a pregnant woman gets a viral infection, it substantially increases her risk of having a child with schizophrenia or autism.

There are dozens of other intriguing associations. Genetic studies, for example, have tied genes in the major histocompatibility complex — a molecule that guides immune reactions — to both schizophrenia and autism. Some individuals with autism carry abnormal profiles of immune molecules in their blood. Others, according to postmortem studies, show activated immune cells in the brain.

Patterson also highlights a rarely discussed aspect of autism twin studies. When an identical twin has autism, the likelihood that the other twin also has it is high, around 80 percent. The so-called concordance rate of fraternal twins, who share only half of their genes, is around 30 percent. Because both kinds of twins share the same womb, most scientists see this difference as evidence that autism is a strongly genetic disorder.

But the fetal environment is actually not the same for fraternal twins and identical twins: Fraternal twins have separate placentas, but identical twins usually share one. Patterson also points out that whereas the fraternal twin concordance rate is 30 percent, for non-twin siblings — who also share half their DNA, but do not share a womb — it’s only about 15 percent.

Patterson lays out this complicated work clearly and concisely, with little editorializing. But anyone who has heard him give a talk knows his primary motivation for writing the book: He’s tired of all the attention in the autism community going toward genetics.

As he notes briefly in the introduction: “I believe that [immune research] is not yet fully appreciated by the biomedical community, overshadowed as it is, in both publicity and funding, by genetic approaches to these diseases.”

He’s not wrong. According to the most recent government report, some $64 million in public and private research funds went toward autism genetics in 2009, compared with $27 million for research on environmental causes and gene-environment interactions.

I wish Patterson had discussed the question of why, with such a long history and so many interesting scientific leads, immune research in autism has been largely ignored.

It could be because genomics technologies, which have evolved rapidly over the past decade, draw research dollars away from other subfields. Or it could be that the anti-vaccine ruckus has made scientists gun-shy of anything even remotely related to immunity.

But it could also be that the bulk of research on immune links to autism is a jumble of intriguing anecdotes and statistical correlations. To gain a better handle on the mechanisms at work, the field needs more scientists like Patterson who test these associations on animal models.

If for no other reason, it’s worth studying because unlike genetic influences, environmentally influenced risk factors can potentially be avoided.


Virginia Hughes is a freelance journalist based in Brooklyn, New York. She earned an M.A. in science writing from Johns Hopkins University in 2006, and has written for Nature, New Scientist, Popular Science and Scientific American.

8 responses to “Book review: Autism’s twisted immune links, untangled”

  1. Jon Brock says:

    Thanks Virginia and Happy New Year! I will definitely look out for this book.

    Just a point of clarification. You note that fraternal twins do not share a placenta, but you didn’t explain why that was important.

    If viral infection (or a mother’s immune response to the viral infection) plays a role in causing autism, then both twins would be exposed regardless of whether they were identical or not. This might explain why concordance is higher for fraternal twins than for non-twin sibs (although there are lots of other potential explanations).

    Is the idea that the placenta acts as an imperfect buffer against this, so sharing the same placenta means that you’ll definitely get the same exposure, whereas twins with their own placentas might be exposed differently?

    As you can tell, I have absolutely no idea what I’m talking about here, but it seems to be a crucial missing link in the argument. Perhaps I should just read the book!

  2. says:

    Hey Jon,

    The idea is just that the difference between identical concordance rates (80%) and fraternal concordance (30%) is not only due to genetics. Fraternal twins grow in different placentas, and each placenta, according to Patterson, “is somewhat different in its ability to transport nutrients and hormones to the fetus.” In terms of the maternal immune response, Patterson says that each placenta “could filter microbes or cytokines somewhat differently,” potentially leading to one with autism and the other without the disorder.

    Amazingly, even identical twins who do share a placenta can have different exposures, depending on the arrangement and size of blood vessels in the placenta.

    Hope that helps clarify things! Thanks for reading.

  3. usethebrainsgodgiveyou says:

    Do great minds think alike?
    Not for you to read, but just as “Proof”, my blog link.

    From there I’ll give an outstanding link about the “trigger” that Persico proposes. I have a very good feeling about this stuff.

    Thank you for your work. What a great way to start the New Year!!!!

  4. Jon Brock says:

    Cool. That was kind of what I assumed. I’ll still check out the book though!

  5. RAJensen says:

    Not all identical twins share the same placenta. The East Flanders Prospective Twin Study (EFPTS) regisrty is one of the largest and oldest twin registries. The EFPTS is the only twin registry that records placentation status at delivery. In over two thousands twin pairs entered into the registry they reported that 70% of monozygotic twins are monochorionic (single placenta) and about 30% of monozygotic twins are dichorionic (seperate placenta). Every twin study in autism has reported a 60% concordance rate for narrowly defined autism, It would be interesting to learn if concordance rates for narrowly defined autism is consistent with ratio of monochorionic v dichorionic in monozygotic twin pairs.

    Gareri et al found that the levels of exposure to alcohol in multiple placenta births varies significantly.

  6. usethebrainsgodgiveyou says:

    I bought the book kindle edition.

    Now if I can only FOCUS long enough to read it…

  7. Paul Whiteley says:

    Interesting read. Paul Patterson also keeps an interesting blog related to his work and book:
    Without trying to complicate things too much, viral infections are perhaps only one piece of the puzzle when it comes to the immune system and psychiatry. There is for example a growing body of research suggestive of a possible association between schizophrenia and cases of parasitic infection as per the work on protozoa like Toxoplasma gondii. Of course association is not causation but looking back at the work of the late Reed Warren on the C4B null allele, perhaps this is an area which requires more exploration for any overlap. Many thanks.

  8. Paris call girls says:

    I envy your capability to publish wonderful article on simply wanted to say I like this !

Leave a Reply

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