Schizophrenia and milk

Only two posts ago I was extolling the virtues of milk and dairy produce and yet here I am again plunging the reputation of the white stuff into disrepute. Michael MacIntyre highlighted such a paradox best in one of his comedy sketches about his children: you love them when they are sleeping ('the little darlings') but seem to find other emotions and words when you want to get them to go to sleep.

The paper in question is this one* from David Niebuhr and colleagues published in Schizophrenia Research. I admit that I am a little late getting to this paper. The main finding from the study is that from looking through quite a few hundred, nay thousand, serum samples for the presence of IgG antibodies to cow milk derived casein, there was a reported association between the risk of schizophrenia and elevated serum levels of casein antibodies before symptoms onset.

I will come back to the findings in more detail momentarily but first there are a few other important things to bring into this. Those who follow this blog will know that schizophrenia in relation to gluten and casein has been discussed before. The original work of the late Curt Dohan on schizophrenia and grains has been talked about (here) alongside the more recent work from people like Faith Dickerson and colleagues detailing findings in relation to gluten antibodies (here). Dohan is perhaps the key person here, following his suggestions over 40 years ago, that some cases of schizophrenia might be tied into diets which contain gluten and casein, and using such diets might actually be quite beneficial to some people with schizophrenia in terms of ameliorating some of their symptoms.

Back to the Niebuhr study, reading through their paper I get the impression that whilst there was great scientific interest in looking at casein antibodies in relation to schizophrenia risk, there was a more practical perspective also. The study was carried out on soldiers, and in particular soldiers who were later diagnosed with schizophrenia and hence medically-discharged from the military as a result. Aside from having a large bank of 'volunteers' and all the accompanying controls in terms of accurate diagnosis and tests for military discharge, I assume that screening serum samples provided by new recruits might eventually provide some further information about those soldiers who might be prone to developing schizophrenia (assuming that the casein hypothesis holds up)?

Anyway, a summary of the methods and findings:

• 6106 serum samples from 855 cases and 1165 controls were analysed alongside a few thousand samples from cases of bipolar disorder (some participants provided more than one sample).
• Three time periods were examined: 2-4 years before diagnosis, 2 years before diagnosis and 2 years after diagnosis.
• There was a significant differences between schizophrenia cases and controls for casein antibody levels for the 2-4 years before diagnosis period.
• There was a significant difference between schizophrenia cases and bipolar cases for both the 2-4 years period before and 2 year period before diagnosis.
• Those cases with a high initial level of casein antibodies at first sampling who also showed an increase in antibody levels on subsequent samples were faced with an 18% increased risk of developing schizophrenia per standard value increase in casein IgG level (this trend not seen in bipolar disorder cases).

With the number of participants included in this study, I don't think anyone can say that the study authors did not recruit enough people to conduct an adequately powered study. Indeed most researchers could only wish of having such a large and well-defined patient group as this. As the authors point out, the military treat their personnel well. That means that with some confidence, they knew that their patient group were not initially presenting with schizophrenia or bipolar disorder at recruitment alongside the comprehensive investigations carried out post-symptom onset prior to and as part of their medical discharge from the military.

As to why such results were found is more of a mystery. This study was based on levels of circulating IgG antibodies to casein. I have tried to go through the various immunoglobulin isotypes in a previous post. IgG does not seem to imply 'allergy' in the way that IgE (type 1 hypersensitivity) implies. Rather IgG is perhaps better described as demonstrating some degree of immunologic sensitisation, in this case to casein. As per the authors' comments on the opioid-excess hypothesis, I do wonder if such findings are potentially present as a result of an increase in circulating neuroactive casein peptide species. Speculation at least.

All in all, a well-designed, well-performed study. I should perhaps end with an apology to milk, or rather casein, for this post. In future posts I will try and say something more positive about milk and casein again, this time without the sting in the tail. To end a song about apology and regret from the very Beautiful South.

* Niebuhr D. et al. Association between bovine casein antibody and and new onset schizophrenia among US military personnel. Schizophrenia Research. May 2011.