Determination of aflatoxin M1 in powdered formula: an inter-laboratory study and the surveillance in Japan

A method to determine aflatoxin M1 (AFM1) levels by using an immunoaffinity column-based clean-up procedure and HPLC with fluorescence detection was validated by an inter-laboratory study among ten laboratories in Japan. Using the validated method, we surveyed AFM1 contamination in powdered formula. Samples for validation included a blank, three levels (blind pairs) of AFM1 spiked into liquid milk, naturally contaminated liquid milk, and naturally contaminated powdered formula. All samples were frozen and sent to the ten participating laboratories. For the liquid milk spiked at 1.0, 0.5, and 0.05 μg/kg levels, recoveries were 89.9, 91.6, and 88.2%, respectively. The repeatability relative standard deviation (RSDr) and reproducibility relative standard deviation (RSDR) were less than 7.4 and 8.1%, respectively. The recovery, RSDr, and RSDR of the powdered formula were 94.5, 8.9, and 11.9%, respectively. The RSDr and RSDR of the naturally contaminated milk were 13.3 and 20.9%, respectively. The Horwitz ratio (HorRat) values of all six samples were less than 1.0. For surveillance, 108 commercial powdered formulae were obtained in Japan. The average value of AFM1 in the powdered formulae was 0 .002 μg/L, as ready-for-infant liquid milk (14 g powdered formula in 100 mL water). The highest contamination was 0.025 μg/L.