Researchers have been chasing the elusive links between gluten and major mental illness for decades. Despite some hyperbolic coverage in Wheat Belly and slightly more convincing coverage in Grain Brain, there is, so far, quite a bit of smoke, but no fire outside a few case studies. Curt Dohan had quite a few papers back in the day (including this one), and much more recently Faith Dickerson, now armed with antibody titres, could be more precise (including in this paper).
In the last couple years the rather stunning data from the CATIE trial (a very large multi-center study of schizophrenia treatment run by NIMH in the last decade) that schizophrenics were 5X as likely to have anti-tTG antibodies as healthy controls and over 7X the likelihood of having high AGA (antibodies to gliadin) compared to normal controls has made more researchers take notice. Yet on face, all we could really say is, wow, a certain subset of people with schizophrenia sure do have some suspicious antibodies to different wheat proteins, and it is pretty clear that devastating neurological illness can be caused by gluten (dystonias in some people, for example) without the classic celiac gut findings, but is the issue in schizophrenia a leaky gut (thus higher antibody titres to certain food moieties) or the wheat itself, or both? I covered these questions in a bit more detail here.
One major issue with the theory that wheat causes schizophrenia is that schizophrenia seems to have a similar prevalence in gluten and non-gluten eating areas, but since "schizophrenia"is pretty clearly recognized as a final common pathway for a number of different genetic and environmental pathologies, it wouldn't necessarily torpedo the gluten theory. Now, finally, we can test whether gluten-free diets help symptoms in the subset of schizophrenics who have suspicious wheat antibodies. The newest round of researchers, led by Jessica Jackson (along with Alessio Fasano) at the University of Maryland and Emily Severance at Johns Hopkins, are following these leads.
First off, we have "A gluten-free diet in people with schizophrenia and anti-tissue transglutaminase or anti-gliadin antibodies." This paper starts off with discussing the mixed results of previous trials (7 in all) of gluten-free diets in schizophrenia, showing a subset with real improvement (and some with remission, which is an astonishing finding), but many without improvement whatsoever. None of these studies tried to differentiate schizophrenics with or without anti-tTG and AGA, mostly because they were done before these titres were available. The paper makes the distinction that anti-tTG antibodies are more likely to signify celiac disease, whereas AGA is more likely to signify non-celiac gluten sensitivity. In the current paper, exactly two patients with schizophrenia (one woman symptomatic since 1976 and a man symptomatic for the past 8 years) and positive antibody titres (one for anti-tTG and one for AGA) who were stable on medicines but still symptomatic (pretty common) were put in an inpatient unit and observed on a gluten-free diet for two weeks.
The woman had improved concentration and attention (critical, because psychotic symptoms often respond relatively well to medication, but poor executive functioning, attention and concentration are not so responsive, and those deficits keep many people with schizophrenia from being able to function independently). The man had some reduction in psychotic symptoms and increased insight into his condition. Since schizophrenia is a progressive neurodegenerative disease, seeing improvement, particularly in the woman who had been sick since 1976 from a non-medicine intervention in two weeks' time is at the least interesting.
The limitations of this study are profound. Open label, about as tiny as you can get, and obviously taking someone and sticking him or her in an inpatient unit with structure and observation is an intervention all on its own. But the level of improvement was enough that Schizophrenia Research (not the topmost tier of psychiatry journals, but certainly no Medical Hypothesis) published the paper, and it is available free full text on pubmed if you care to click the link above.
The second paper was sent to me by the amazing Victoria Prince (who just finished her last rotations in medical school. Woo hoo!) I love this paper, and I want to give Emily Severance a hug just for the ideas it brings together. She already deserves a hug for the previous paper I discussed in this article: Schizophrenia and the Gut. We know schizophrenia is multitudes, it's complex, it's genetic and environmental and immune-mediated. Ergo: Anti-Gluten Immune Response following Toxoplasma gondii Infection in Mice. (I know, mice.) It's also available free full text over at PLOSone.
Anyway, we already know that folks with schizophrenia have higher levels of gut inflammation (measured by checking antibodies to known infections that get into the system when there is gut inflammation or infections that actively cause gut inflammation, such as our old friend Toxoplasma gondii), and the newer the onset of illness, the more likely you are to find gut inflammation, AND the more antibodies to gluten and casein you have, the more likely you are to have these signs of gut inflammation. So Dr. Severance sought to answer some of the questions raised by this finding. Did the infection cause a gut pathology that allowed neurotoxic food fragments to attack the brain of the genetically susceptible? Were the infections themselves the problem in the brain, and the food antibodies just secondary to the infections? Well, it is difficult (not to say unethical) to do the sorts of experiments you need to answer these questions in humans, but mice can be housed and infected and their little immune systems examined in greater numbers over several generations more readily.
So the researchers took mice and gave them delicious T gondii infected rodent chow (via infected ground up mouse brains!!). They infected some adult mice and a subset of female mice who were then knocked up so they could check the pups for gut inflammation as well…there are a lot of mini-experiments in this paper and I won't explain them all to death here, as the paper is freely available. Anyway, after infection with T gondii, serum antibodies to wheat proteins and complement activation (not a sign of well-bred mice but rather a measure of inflammation) increased in the infected groups but not in the mock-infected or uninfected groups. The anti-wheat antibodies in mouse pups born to the infected moms were also significantly higher than in those born to uninfected mouse moms.
So here we have proof, in mice, that infection with Toxoplasma, a known risk factor for schizophrenia in humans, leads to the generation of anti-gluten antibodies, presumably via a gut inflammatory mechanism. Most importantly, in the mouse pups, the anti-gluten antibodies and infection happen at a time of critical neurodevelopment. Thus the combination of infection and, perhaps, a dietary enhancer (such as, possibly, gluten) could be working in concert to make someone vulnerable to developing schizophrenia later on. The "gut inflammatory" mechanism is vague at this point. In celiac disease in humans (more associated with the anti-tTG antibodies), there is definitely gut damage and permeability. In non-celiac gluten sensitivity (more associated with AGA), there doesn't seem to be frank leakage, but apparently large gluten peptides can cross the border via transcytosis and this may happen more readily if the gut is infected and the immune system is on the case and things…frankly the exact details of gluten and the gut continue to elude us. Check out the last paragraph of this paper (BIG HUGS):
In summary, the models described in this paper provide appropriate experimental tools to examine the impacts of gluten peptides, T. gondii and associated immune activation on brain physiology. As we accumulate more information from analyses of clinical biomarkers, we can adapt these animal models to test the effects of dietary modifications and other types of infections on behavioral endpoints, the pharmacological outcomes of specific antipsychotics on immune system parameters, and the autoimmune response responses triggered by T. gondii infection. Ultimately, we envision a translational system by which we can fully evaluate the interface of environmental perturbation and genetic predisposition as it relates to serious neurodevelopmental disorders such as schizophrenia, bipolar disorder, and autism.
I've never been a very linear person; I tend to absorb and think about things all at once. That's part of what I like about my so-called Evolutionary Psychiatry. We can think about lots of things at once as they impact physiology, immune activation, and genetics. The researchers who also seem to think this way, but can also break down these questions and not leave gaping holes (Severance's previous experiment where she took the trouble to go across the ocean to study gut and immune activation in medication naiive and medicated schizophrenics, taking out a major confounder in most schizophrenia research in the US) are the kinds of thinkers we need who can do good science to work out these big complex tangles. I can't wait for the next papers to come out. In the mean time, there is no clinical guidance. Is it worth checking your schizophrenics for anti-tTG and AGA? What are the risks of recommending a gluten-free diet and what is the likelihood it will be strictly followed in an outpatient setting?
Always, more questions than answers.