SPIFAN Nutrients ERP Book_9-29-15

BVit-02 - MEF FORM FOR ERP USE ONLY DO NOT DISTRIBUTE

Overall the method performs well, and meets SMPR's for each of the four vitamins.

SRM statistical analysis was done even though the SRM values are not "totals" for vitamins B1, B2, and B3. However, there is very little of the phosphated forms of thiamine and very little nicotinic acid in the SRM, so the results are in good agreement. There is a high bias for riboflavin, but that is expected due to the release of the phosphated forms by the enzymatic digestion. For B6, only the pyridoxine results were used for statistical analysis, and they compare well. Both repeatability (RSDr) and intermediate precision (RSD iR) are given. There are no SMPR criteria for intermediate precision, but they are less than the RSDR requirement which is promising for this method meeting RSDR if it proceeds to multilab evaluation (Table 3). Intermediate precision was collected by three different analysts, on two different instruments of the same make and model, and on six separate days over the course of about a month. Linearity criteria are not contained in the SMPR, but linearity is a daily method suitability criteria. Table 4 shows the 5-day averages of the standard solutions calculated as samples against the curve. Typically, 95-105% recovery is desired, but larger deviations can be acceptable depending on circumstances. Pyridoxal recovery range is a little larger (90-110%) because the signal is more than 10x lower than the other forms. This is by method design because pyridoxal is a minor component. Therefore, the curve must be of low concentration to keep product levels within the curve. The larger deviation does not negatively impact Total B6 quantitaion becuase pyridoxal is a minor component. If the levels were higher, the precision would be similar the other compounds. Table 5 contains the practical LOQ (PLOQ) data derived from the Linearity study. PLOQ's is set as the low standard of the curve after appropriate unit conversion and application of the dilution factor of 500. The PLOQs meet the SMPR's except for B6 by a narrow margin. However, the LOD for B6 is nearly 200x lower than the PLOQ, so a slight adjustment to the working standard concentration range would allow the B6 LOQ to be easily met. Spike recoveries for all 4 vitamins are in the 90-110% range as required except for B3 recovery of 110.7% for one sample. All five placeboes and three of the fortified matrixes were studied. The remaining five matrixes were not studied because they are not different enough from the eight matrixes tested to warrant additional scrutiny. The relative amount of each vitamin form in the spiking solution is given below: B1: Thiamine (~80%), thiamine monophosphate (~12%, and thiamine pyrophosphate (~8%). B2: Riboflavin (~73%), riboflavin phosphate (~18%), and FAD (~9%). B3: Nicotinamide (~93%), nicotinic acid (~7%). B6: Pyridoxine (~80%), pyridoxal (~5%), pyridoxal-5'-phosphate (~4%), pyridoxamine (~6%), pyridoxamine-5'-phosphate (~5%). Spikes were done at approximately 50 and 200% of fortification levels.

Table 1

Table 2a-2d

Table 3

Table 4

Table 5