AOAC OMB Final Action Recommendation (December 2019)-2016.14

2016.14 (October 2019) – FOS-03 MANUSCRIPT

FOR ERP USE ONLY DO NOT DISTRIBUTE

have been available, AOAC 997.08 (6) and AOAC 999.03 (7). Both of those methods have their 1 limitations, which have previously been discussed (4, 5, 8), and this lead to a call for new methods, 2 initiated by the Stakeholder Panel for Infant Formula and Adult Nutritionals (SPIFAN) after 3 defining the standard method performance requirements (SMPR) for fructan analysis (9). Of the 4 methods submitted to the AOAC in answer to the call for methods, two methods had sufficient 5 single laboratory performance to be promoted to first action status; the method submitted by 6 Haselberger and Jacobs (8), first action 2016.06 and the method submitted by Brunt, et.al. (5), first 7 action 2016.14. This paper describes the collaborative study carried out to collect repeatability and 8 reproducibility data of first action method AOAC 2016.14. 9 Collaborative Study Description 10 Twenty-nine laboratories were invited to participate in this collaborative study and 15 11 laboratories located in 12 different countries expressed interest in participating. One lab had to 12 resign because of difficulties in importation of the samples and two labs dropped out because of 13 instrumentation problems and resource availability. Finally, 12 laboratories from 9 countries 14 managed to provide results on time for this study. The study was designed in two parts (1) Method 15 set up and qualification of participants and (2) Multi-laboratory test participation. 16 Five fructan-containing infant formula available in the SPIFAN sample kit were used in the 17 study and an additional two adult nutritional products and one ready-to-feed (RTF) infant formula 18 from manufacturing sites in Europe were added. A Standard Reference Material (SRM 1869 19 Infant/Adult Nutritional Formula) from the National Institute of Standards and Technology 20 (Gaithersburg, MD) was also included. The additional samples were added to the SPIFAN kit 21 because only five of the products in the kit contained fructans. All samples were blinded and coded 22 before being dispatched to the participants. Homogeneity of the fructan contents in the extra 23 samples was assessed by duplicate analyses of test samples from separate bottles. Out of the 5 boxes 24 of 24 bottles each received for the RTF infant formula (120 bottles), one bottle was randomly 25 selected in each box and analysed in duplicate for its fructan content ( n =5). The two adult 26

3