The activity of α-amylases is frequently determined with an assay consisting of a single-point measurement at 20°C (this protocol). Work within INFOGEST “Working Group 5 - Starch digestion and amylases” found significant interlaboratory variation using this protocol.

The current study aimed to evaluate the repeatability and reproducibility, measured as intra- and interlaboratory coefficients of variation (CVs), of a newly optimized version of that protocol based on four time-point measurements at 37°C. Human saliva (pool from ten healthy adults) and three porcine enzyme preparations (two pancreatic α-amylases and pancreatin) were tested in 13 laboratories (12 countries, 3 continents). Mean repeatability CVs of the α-amylase activity values were between 7% and 13% for all products at both incubation temperatures. Assay reproducibility was greatly improved: reproducibility CVs ranged from 9% to 21%, i.e. up to four times lower than previous observations following the original method.