Food and Behaviour Research

Donate Log In

31 January 2019 - Questioning Answers - Milk opioid peptides and dipeptidyl peptidase-4 (DPPIV) linked to autism

Paul Whiteley


The aim of the this study was to determine "BCM7 [β-casomorphin-7] influence on DPPIV functioning in children with ASD in comparison to healthy children."


This study adds some new experimental evidence to the longstanding question of whether opioid-like peptides derived from A1 beta-casein (found in standard cows' milk, but not in human breastmilk or other animal milks) may play a role in some symptoms of autistic spectrum disorders (ASD) - and related conditions.

Previous studies - like this one - have found some evidence of abnormal peptide profiles in at least a subset of children with ASD.  However, ASD is not a unitary condition, so differences in the populations studied and methodologies used have made any clear conclusions difficult.

As usual, 'more research is needed'. 

Meanwhile, the A1 form of beta-casein is found in standard cows' milk in most developed countries - and at least some individuals with autism and related conditions appear to react badly to cows' milk, and to benefit from 'casein-free' diets.

Many others do not benefit from excluding milk and dairy products. And as milk is a highly nutritious food for children, excluding it may require careful planning of the diet to ensure adequate intakes of all essential nutrients.   

Importantly, however, A1 beta-casein is not found in goats' milk, sheeps' milk, or the milk of any other mammals - including human breastmilk. These milks all contain A2-type beta-casein (the 'original' form in evolutionary terms).  Only standard cows' milk contains the A1 form of beta-casein (which is digested slightly differently, producing opioid peptides to which some individuals may be sensitive).

Anyone suspecting that cows' milk might be linked with 'intolerance' symptoms could therefore try exploring whether other forms of milk - containing only A2 beta-casein - might be tolerable, as an alternative to excluding all milk and dairy products from the diet (which can increase risks for nutritional deficiencies unless the diet is carefully planned to compensate for this).   

Very importantly, however, A2-type milks are NOT recommended if classic cows' milk protein allergy is known or suspected. In this case, all animal milks should be avoided.

For more information on the differences between A1 and A2 milk, see

Read the abstract of the underlying research:

See here  for other articles relating to autism and diet.
So: "we have concluded that milk-derived opioid peptides and DPPIV [dipeptidyl peptidase-4 (DPPIV)are potentially factors in determining the pathogenesis of autism."

That was the quite sweeping statement made in the paper published by Beata Jarmołowska and colleagues [1]. It continues a quite a long running research topic in autism circles on whether the chemical arrangement of certain dietary components *might* have an important biological effect on at least some people diagnosed as being on the autism spectrum.

The Jarmołowska paper is open-access so there is no real need for me to go over the suggested hows-and-whys of some diets being potentially related to (some) autism. If you need some further reading on the topic, I'll direct you to other posts on this blog where I've discussed this 'gluten and casein' issue and onward, my professional interest in it for quite a few years.

The aim of the Jarmołowska study was to determine "BCM7 [β-casomorphin-7influence on DPPIV functioning in children with ASD in comparison to healthy children." 'Healthy children' is the term for the control group used by the authors by the way, not me. They "examined content and activity of serum DPPIV, content of BCM7 in serum and urine, and studied the effect of hydrolysed bovine milk, as a source of opioid peptides, on DPPIV gene expression in peripheral blood mononuclear cells (PBMC) in both groups."

Results: "We found that the content of BCM7 in serum was significantly higher (p < 0.0001) in ASD than in the control group." Urine concentrations of BCM7 were not significantly different among those with autism compared with controls. Also: "Concentration of DPPIV was found to also be significantly higher in serum from ASD children compared to the control group (p < 0.01)."That was about the sum of the differences noted by researchers.

Caveats? Well some. So: "ELISA test enabled identification of BCM7 contents in the serum and urine from patients, as well as in tested peptide extract obtained from hydrolyzed bovine milk." Authors do mention how the testing was carried out "in triplicate" following a previously published protocol. I don't dispute the results they got but am not exactly enamoured with the ELISA method used. I'd much rather see the analysis undertaken using something like mass spectrometry or similar technology, given the precision that comes with such methods based for example, on the use of internal standards. Perhaps if the authors still have their samples, they might consider further analyses if available to them?

Although researchers provide quite a bit of information about their participant groups, I also noted one important detail to be missing: were any of their participants - diagnosed with autism or not - following any special dietary regime? Y'know, they talk about casein (milk) free diets and how such diets are supported by "numerous scientific reports." So I guess in a cohort of 86 children diagnosed with autism, at least a few of them might be following such a dietary intervention? I've searched their paper but couldn't find anything to say that they were or weren't.

Putting such issues to one side, I don't want to take anything away from the Jarmołowska findings. As they end their paper: "this issue requires further investigation." I wouldn't disagree.