Thursday, 1 May 2014

Casomorphins and autism reignited?

For those who know me and some of my particular autism research interests, they'll understand that I was always going to be intrigued by the results reported by Oleg Sokolov and colleagues [1] on elevated levels of bovine casomorphin-7 immunoreactivity being reported in a small group of children with autism. Indeed, as per my last post plugging my new book (apologies for the very blatant plug!), casomorphins, the peptide breakdown products from the protein casein complete with opioid activity, are something that I've known about for quite a few years.

BC1-7 @ Wikipedia 
From the outset I have to come clean and say that I did peer-review a version of the Sokolov paper when submitted to another journal last year (2013). Without hopefully creating too many waves, my comments whilst supportive did highlight a few areas of contention with that previous submission. Not that this should influence yours or my interpretation of the current findings; neither should it seem unusual that papers are reviewed in one journal and end up finding a home at another journal (some of my manuscripts have endured a similar journey).

Anyhow, with thanks to Natasa for passing the full-text version of the published paper to me, a few details:

  • This was a small trial comprising 10 children (4-8 years old) diagnosed with an autism spectrum condition (including autism and Asperger syndrome) and 10 control children.
  • First production spot urine samples were collected and some initial preparation carried out (SPE) prior to the main event which was analysis by ELISA looking at immunoreactivity to bovine (cow) casomorphin-7 (CM-7). Just in case you didn't know CM-7 is the same as beta-casomorphin 1-7 (BC1-7) which as I've indicated previously, is a peptide fragment derived from cow milk protein (or at least some types of cow) which has some history in autism research (see here).
  • Looking at the current method and the comments I made on the previous incarnation of this manuscript, quite a few of the issues I had highlighted originally did seem to have been answered in this paper including a much more detailed explanation of the analytical methods employed on samples such as things like limits of detection (LOD). Indeed, the authors report: "The minimal detectable limit of CM-7 immunoreactive material was 25 fmol" which is pretty low: a billionth of a millionth (10-15).
  • Results: some important trends emerged from the data not totally out-of-line with other findings [2]. So: "autistic children have significantly higher levels of urine CM-7 than control children". Bearing in mind the already discussed small participant numbers included, further analysis of CM-7 results for those children with autism (N=5) and those with Asperger syndrome (n=5) seemed to suggest a link with the "severity of autistic symptoms" too, confirmed in a plot of CM-7 levels and CARS scores achieving a correlation coefficent of 0.85. Ergo, cows milk or at least some of the by-products from cows milk might show involvement with some cases of autism [3].

OK, to reiterate, this was a study of 10 children with autism/Asperger syndrome vs. 10 asymptomatic children. There is still very much more to do on this topic before one might entertain discussions about trends and facts in this area including the question of what effect dietary intervention such as the use of a casein-free diet might do to the CM-7 results obtained. You may well be saying 'dietary intervention should see CM-7 levels drop' but I'd personally like to see the cold data showing this effect before any assumptions are made.

Glancing through some of my review notes made about the original incarnation of this paper and it's current format, I do note that at least one important point that I originally raised does not seem to be answered in the most recent version. The authors note: "The identity of bovine CM-7 detected by ELISA was confirmed by reverse phase and gel-filtration HPLC". The point I made was that whilst I've published on the use of HPLC (high-performance liquid chromatography) for the analysis of urine samples, this technique alone and without any reference to the type of detection method used is pretty meaningless. Certainly unless detection is via mass spectrometry (MS) or even nuclear magnetic resonance (NMR) there will always be questions about accuracy and precision.

In amongst some of the discussions included in the Sokolov paper, I noted a few points of added interest which should be mentioned. There is for example, some chatter about dietary intervention and autism, and as I suggested a few sentences back, how removal of casein (and gluten) from the diet should affect the source materials of those peptides as per what is done in coeliac disease with gluten peptides (see here). I note also that the authors make mention of different types of milk as potentially being useful: "casomorphin-free casein" although incorrectly sourcing it back to A1 cows when it should be A2 cows (see here). Indeed, it is timely that the paper by Barnett and colleagues [4] has also entered the peer-reviewed publication domain but I'm not going to say too much more on that for now aside from saying that cows are not the only makers of milk.

I've probably been a little harsher than usual on the Sokolov paper given the history in this area of food and autism and finding data to fit models and hypotheses around the use of dietary intervention with a specific focus on peptides [5] and on more than one occasion [6]. By no means do I wish to belittle the potentially important observations made in the current paper; merely stating however that technology can help in this area and we should be using it where and whenever possible to cast further scientific light on the unusual connection between food and its potential effects on at least some cases of autism.

Oh, and whilst we're on the topic of gluten and casein peptides and autism, I'd like to introduce some rather novel findings from Dick Deth and colleagues [7] on a potentially important link to cysteine and glutathione which might also be relevant to some autism...

Something a little bit different to finish today. For those who know a little about kata and karate, how about a little synchronised kata? (my brood have already started their training...)


[1] Sokolov O. et al. Autistic children display elevated urine levels of bovine casomorphin-7 immunoreactivity. Peptides. 2014. March 20.

[2] Reichelt KL. et al. Peptides’ role in autism with emphasis on exorphins. Microbial Ecology in Health & Disease. 2012; 23: 18958.

[3] Kost NV. et al. Beta-casomorphins-7 in infants on different type of feeding and different levels of psychomotor development. Peptides. 2009 Oct;30(10):1854-60.

[4] Barnett MP, et al. Dietary A1 β-casein affects gastrointestinal transit time, dipeptidyl peptidase-4 activity, and inflammatory status relative to A2 β-casein in Wistar rats. Int J Food Sci Nutr. 2014. March 20.

[5] Hunter LC. et al. Opioid peptides and dipeptidyl peptidase in autism. Dev Med Child Neurol. 2003 Feb;45(2):121-8.

[6] Cass H. et al. Absence of urinary opioid peptides in children with autism. Arch Dis Child. 2008 Sep;93(9):745-50.

[7] Shah JJ. et al. Casein and gluten-derived opiate peptides affect cysteine uptake and redox status. FASEB J. 2013; 27: 1075.1.

---------- Sokolov, O., Kost, N., Andreeva, O., Korneeva, E., Meshavkin, V., Tarakanova, Y., Dadayan, A., Zolotarev, Y., Grachev, S., Mikheeva, I., Varlamov, O., & Zozulya, A. (2014). Autistic children display elevated urine levels of bovine casomorphin-7 immunoreactivity Peptides DOI: 10.1016/j.peptides.2014.03.007