The first and only time I ate Marmite was in high school. I’ve never had the desire to eat it again. But many people love Marmite. If you are one of these people, and you have celiac disease, you should know that this product may not be gluten free (as we define gluten-free in the United States).
What exactly is Marmite?
Marmite is a yeast extract spread that is wildly popular in the United Kingdom. According to the Marmite website (marmite.com), this spread is “made from brewer’s yeast that’s been used to ferment sugars into alcohol.” The ingredients list of Marmite reads, “Yeast Extract, Salt, Vegetable Extract, Niacin, Thiamin, Spice Extracts (Contains Celery), Riboflavin, Folic Acid, Vitamin B12.”
Why assess Marmite for gluten?
Marmite was assessed for gluten contamination as part of preliminary testing for a planned larger study on yeast extract/autolyzed yeast extract derived from brewer’s yeast. Brewer’s yeast may be a by-product of the beer brewing process and as such may be contaminated with malt and grain. In the US, individuals with celiac disease are advised to avoid food products containing brewer’s yeast as an ingredient.
An unopened 125 g container of Marmite manufactured by Unilever in the UK was sent to the Food Allergen Research and Resource Program at the University of Nebraska (farrp.org). The product was tested in duplicate using both the standard R5 sandwich ELISA and the competitive R5 ELISA. The results are as follows:
Sandwich R5 ELISA
Extraction One: 28 ppm gluten
Extraction Two: 31 ppm gluten
Lower limit of quantification for this assay is 5 ppm gluten
Competitive R5 ELISA
Extraction One: 3,700 ppm gluten peptide
Extraction Two: 3,400 ppm gluten peptide
Lower limit of quantification for this assay is 1,250 ppm gluten peptide
Please note that 1 ppm gluten peptide is NOT equivalent to 1 ppm gluten. It is not easy to evaluate peptide concentration in terms of ppm gluten. According to R-Biopharm, the manufacturer of the competitive R5 ELISA,”The degree of hydrolisation is variable in the analysed food matrices and therefore smaller and bigger fragments in different amounts and forms are occuring.” What we can conclude from the results of the competitive assay, is that Marmite contains peptide fragments that are potentially problematic for individuals with celiac disease.
Very important caveat: Only 1 container of Marmite was tested. This container may not be representative of the general gluten content of Marmite.
Marmite in the United Kingdom
Marmite is manufactured by Unilever in the UK. Marmite is not labeled gluten free (at least the jars I have). Unilever includes Marmite in a listing of products with the heading, “The products on the below list DO NOT contain Gluten or products thereof (wheat, rye, barley oats) in flavour or as ingredient.”
Historically, Marmite has been included in the Food and Drink Directory of Coeliac UK (http://www.coeliac.org.uk/node/156). In email correspondence with representatives from Coeliac UK, I was advised that for the 2009 version of the directory, manufacturers had to sign a declaration that products contained less than 100 ppm gluten. Prior to 2009, products had to contain less than 200 ppm gluten.
The 2010 directory has two sections. To be listed in the first section, foods must meet the gluten-free labeling law introduced in the UK in 2009. Products in this section contain less than 20 ppm gluten as verified through testing.
To be listed in the second section, foods must meet 2005 UK Allergen Labeling laws. Products in this section do not include any ingredients that contain gluten. Manufacturers are not required to submit gluten test results for products to be included in this section. Marmite is listed in the second section.
According to the Website of Coeliac UK, “All foods listed in the Food and Drink Directory are suitable for people with coeliac disease. Even though the products in Section 2 are not specialist products and may not be labelled as ‘gluten-free’, they are still suitable for people with coeliac disease.” For more information, see the webpage of the Food and Drink Directory.
Results of the testing done on Marmite were provided to both Coeliac UK and Unilever. Unilever has prepared a statement regarding the test findings for their “Careline” for concerned consumers (see the contact page of marmite.com).
To eat or not to eat Marmite
Until more research is done on the gluten content of Marmite, my recommendation is to not eat this product.
Statement from Unilever
Marmite Yeast Extract Spread – Gluten Content
Unilever UK Statement (1st March 2010)
Unilever is aware of the interesting research being carried out in the US, investigating gluten peptide levels in yeast containing products. We will be looking into this ourselves to understand the possible implications for coeliacs and the validity of the tests in relation to Marmite.
We would like to reassure our customers that Marmite has been available for over 100 years, currently selling approximately 10 million jars per year, and that there have been no recorded instances of allergic or intolerant reactions to gluten.
Marmite has always been produced from ‘spent’ yeast arising as a by-product from beer-making and, as such, we recognise the possibility of trace cross contamination from gluten-containing cereals used in the brewing processes. However, the spent yeast undergoes extensive purification steps – including washing, centrifugation and filtration – before being processed into Marmite. We would therefore expect the gluten content to be very low (although not completely absent).
Although we do not currently make any claims for the product in terms of its gluten content, we do conduct analytical checks for gluten from time to time in order to assist consumers via our careline. These tests have confirmed that the gluten content in Marmite is low and we would regard the results reported in the US of gluten contents between 28-31ppm as fairly typical for our product.
These results would meet the criteria for the old international Codex standard for “gluten free” (below 200ppm) and the recently introduced EU regulatory limit for “very low gluten” (less than 100ppm), but not the recently introduced EU regulatory limit for “gluten-free” (less than 20ppm). We do not however use these terms in product labelling – they are voluntary claims.
Regarding the US finding of much higher levels of ‘gluten peptides’ in our product, we would advise that we do not carry out sophisticated analysis of this type as a matter of routine, but have now commissioned such testing as part of our investigation. It is important to recognise that results from this kind of testing can suffer from interference arising from the high salt content of the product and also require careful interpretation.
The methodology used for such testing (known as competitive ELISA testing) gives results for gluten peptides: ie fragments of gluten protein molecules which in theory also correspond to gluten content, but with an uncertain relationship. It is not fair to say that 1mg of gluten peptide has an equivalence of 1mg of gluten. Although we would stress that it is inappropriate for us to give medical advice, our understanding is that people with coeliac disease respond adversely only to those peptides still containing the ‘repeating units’ to which coeliacs are intolerant. In simple terms, what matters is not how many fragments are present in total, but the size of the fragments and how they were split from their original parent proteins.
Another point arising from this issue is that coeliacs react adversely to the actual amount of gluten consumed, rather than the concentration in the product. Since Marmite is consumed in very small quantities (we recommend a portion of just 4g), it would seem unlikely that low concentrations of gluten in the product would present a problem for coeliacs under normal circumstances.
Copyright © by Tricia Thompson, MS, RD
Also available at