Is the calculation of the gluten content by multiplying the prolamin content by a factor of 2 valid?

Presently, the only essential therapy of celiac disease (CD) is the permanent strict withdrawal of gluten from the diet. With respect to gluten containing foods from wheat, rye, barley, and oats, CD patients have to consume surrogates that must be gluten-free according to the Codex Alimentarius Standard for Gluten-Free Foods. The recent Draft Revised Standard proposes a gluten threshold of 20 mg per kg gluten-free product. For gluten quantitation, the alcohol-soluble prolamins should be extracted and analyzed by an immunochemical method; the amount of gluten is calculated by multiplying the prolamin content by the factor of 2. To investigate, whether this calculation is valid in any case of contamination of gluten-free products by wheat, rye, barley, and oats, wholemeal or white flours from common wheat, spelt, durum wheat, kamut, emmer, einkorn, rye, barley and oats were analyzed for the ratio of prolamins to glutenins (PROL/GLUT) by a combination of extraction and reversed-phase HPLC procedures. Additionally, different industrial wheat starches were analyzed for their prolamin and total gluten content using different extraction and concentration steps followed by gel permeation HPLC. The results for the cereal flours revealed that the ratio PROL/GLUT was generally higher than 1.0 as proposed by the Draft Revised Standard and strongly influenced by cereal species and variety. Common wheat showed the lowest ratio (1.5-3.1), followed by oats and spelt (1.7-3.3), barley (1.4-5.0), durum wheat and kamut (3.1-3.4), emmer (3.5-7.6), rye (6.3-8.2), and einkorn (4.0-13.9). In any case, the gluten content of gluten-free products contaminated with CD activating cereals was generally overestimated, when the prolamin content was multiplied by the factor of 2. In extreme cases, e.g., contamination with rye, the overestimation amounted to 72-79%. Completely different PROL/GLUT ratios were found in ten commercial wheat starches ranging from 0.2 to 4.9. Obviously, the quality of wheat cultivars used for starch production and/or different process parameters, e.g., washing steps, influenced the composition of gluten proteins adherent to starch granules. For wheat starch, the calculation of the gluten content by 2 x PROL may either lead to underestimation (-71% at most) or overestimation (+66% at most). In conclusion, this calculation is invalid; therefore, a future task will be the development of immunoassays with antibodies against all types of storage proteins from wheat, rye, barley, and oats.