AOAC SPIFAN Review Team Meeting Book-July 22, 2015

Accuracy:

In order to determine method accuracy, a two step approach was taken. The first

step was to spike a placebo product with the free non-phosphorylated forms of the

vitamins to probe accuracy without the additional variable of enzyme efficiency. Spiking

was performed at two levels: the first being at the middle standard concentration and the

second at 5 times below the middle standard concentration. The second step was to spike

the placebo and SRM 1849 with the phosphorylated forms of the vitamins at the mid

standard level. In the case of thiamin and riboflavin, both phosphorylated forms were

spiked. Pyridoxine phosphate, however, was not spiked because it does not occur

naturally. A wider acceptance criteria than normal was necessary because of complexity

of enzymatic digestion. The results for each compound are summarized below, and in

Tables 1 and 2.

Total

Pyridoxine:

Overall, the average spike results for pyridoxine met the expected criteria of 85-

115%; although, the overall variability of the high spike data was slightly higher than

expected (Table 4). There was no assignable method based cause to the variation, and

the triplicate within day data is very tight. Further, data from the low-spike level

exhibited a variability similar to the precision data above. Given these reasons, the

variability in the high-spike data is most likely assigned to the limited amount of data.

Nevertheless, the results are still within the expected criteria and therefore are considered

acceptable.

Pyridoxal:

The pyridoxal spike data was excellent with only one day’s average falling

slightly outside 105% (Table 4). The method performance for PAL is well within the

expectations. The pyridoxal 5’-phosphate recoveries were also well within expectations

of 85 – 115% (Table 5). Overall, the method is considered accurate at determining the

total pyridoxal content of milk-based infant formulas.

Pyridoxamine:

The spike recoveries for non-phosphorylated pyridoxamine are very good

beacause of the method’s ability to accurately determine the free levels of pyridoxamine.

The recovery of pyridoxamine 5’-phosphate, while consistent, is at the 80% level. This is

potentially due to stability of the stock spike solution or a problem with the enzyme

activity. Considering that the inherent pyridoxamine concentration is a small fraction of

Total

and the phophorylated contribution is approximately 50%

of the free

BVIT-02

FOR ERP/WORKING GROUP USE ONLY

DO NOT DISTRIBUTE