Comparison of allergen ELISA kits for soy detection in multiple food matrices

Soybeans are widely used as a protein source in food, but even trace amounts can trigger harmful reactions in allergic individuals. To minimize safety risks, food producers can use commercially available ELISA kits to detect unintended soy presence. However, selecting the appropriate ELISA kit is challenging due to limited information on assay performance across different matrices. Therefore, this study compared seven commercial soy ELISA kits across nine different food matrices: cookie, chocolate, beverages, snacks, plant-based drinks and products, dairy products, ice cream, meat replacers and herbs/spices. Soy content was quantified as milligram of total soy protein per kilogram of the food product, in both soy-spiked and incurred food products. Recovery, intra-assay, and inter-assay variability were evaluated for the spiked matrices. All kits could qualitatively detect soy at the eliciting dose (ED01) of 0.5 mg total soy protein based on relevant consumption size for each matrix. Quantitatively, detected soy concentrations increased with higher soy levels, though recovery varied by kit and matrix. The number of ELISA kits meeting recovery criteria was as follows: dairy products (7), cookies (6), ice cream (6), plant-based drinks and products (6), snacks (5), beverages (5), herbs/spices (5), chocolate (4), and meat replacers (3). Meat replacers proved the most challenging matrix, with only three kits meeting the criteria. These findings provide comparative performance data for commercially available soy ELISA kits and can assist the food industry in selecting appropriate methods for reliable soy detection in specific matrices, contributing to improved allergen risk management and consumer safety.