Abstract
This study evaluated the effectiveness of the Japanese food system in preventing unintended exposure to milk and wheat protein allergens in processed foods. We analyzed protein contamination using a Bayesian approach with left-censored models to estimate concentration distributions, utilizing published literature and direct measurements of food products with precautionary labeling. Two consumption scenarios were developed: a high-intake model, and a low-intake model. These scenarios estimated unexpected protein intake across age groups. In the high-intake scenario, the age-specific mean intake of milk protein was 0.040–0.046 mg, whereas wheat protein averaged 0.53–0.61 mg. When the low-intake scenario was applied, means declined to 0.0051–0.0062 mg for milk and 0.106–0.129 mg for wheat. At the 97.5th percentile, predicted intakes were 0.223 mg for milk and 3.21 mg for wheat, both sitting below the provisional FAO/WHO reference doses of 2 mg and 5 mg, respectively. It is important to note that this assessment focuses on potential contamination under typical processing conditions and does not model rare, acute incidents of high-level accidental allergen contamination. Overall, the results indicate a low likelihood of allergic reactions from inadvertent milk or wheat protein intake under Japan’s current labelling regulations.
•Hidden milk proteins found in 8.1 % of unlabeled processed foods.•Hidden wheat proteins detected in 11.2 % of unlabeled processed foods.•Mean exposures to milk and wheat proteins were below reference doses.•Bayesian methods estimated minimal risk of unintended allergen exposure.•Japanese labeling system effectively reduces unintended allergen exposure.