AOACSPIFANMethods-2017Awards

182 / 258

230 McMahon: Journal of AOAC I nternational Vol. 99, N o. 1, 2016
Acetate and both fortified with α-tocopherol acetate) were used
by the participating laboratories so that the laboratories could
become familiar with the analysis procedure. The results were
submitted to the Study Director for approval prior to commencing
the collaborative study. Results within a range of expected levels
indicated that the laboratory was capable of successfully running
the analysis. The same practice samples were used as QC samples
during the analysis of the study samples.
Most of the laboratories submitted practice sample results that
met the acceptance criteria. A couple of laboratories submitted
results just outside the acceptance limits but were accepted to
participate in the study as they met the acceptance limits for at
least three of the four analytes.
Milk-Based Products
For retinyl palmitate, RSD r was calculated in a range of
1.06–4.72% and RSDR in a range of 6.51–10.52%.
For retinyl acetate, RSDr was 16.60% and RSDR was
22.61%.
For α-tocopherol acetate, RSDr was calculated in a range of
0.60–3.84% and RSDR in a range of 4.15–8.67%.
For α-tocopherol, RSDr was calculated in a range of
1.57–5.78% and RSDR in a range of 5.68–12.47%.
For total vitamin E, RSDr as tocopherol equivalents (TEs)
was calculated in a range of 0.81–3.74% and RSD R in a range
of 3.84–7.17%.
Note: One milligram of α-tocopherol acetate is equal to
0.671 mg TEs. One milligram of α-tocopherol is equal to
1 mg TEs.
Low Fat Product
For total vitamin A palmitate, RSD r was 15.78% and RSDR
was 21.73%.
For α-tocopherol acetate, RSDr was 2.11% and RSDR was
8.50%.
For α-tocopherol, RSDr was 8.90% and RSDR was 43.56%.
For total vitamin E, RSDr as TEs was 2.71% and RSDR was
10.78%.
The concentration of α-tocopherol was relatively low in the
product: 0.14 mg/100 g ready to feed (RTF).
Hydrolyzed Protein and Elemental Products
For retinyl acetate, the partially hydrolyzed soy-based
product gave an RSDr of 2.30% and an RSDR of 11.93%.
For α-tocopherol acetate, RSDr was calculated as 3.65% and
RSDR was 11.25%.
For α-tocopherol, RSDr was calculated as 1.67% and RSDR
was 11.94%.
For total vitamin E as TEs, RSDr was calculated as 5.46%
and RSDR was 10.15%.
For retinyl palmitate, the elemental product gave an RSDr of
15.13% and an RSDR of 16.25%.
For α-tocopherol acetate, RSDr was calculated as 3.38% and
RSDR was 6.66%.
For α-tocopherol, RSDr was calculated as 15.48% and RSDR
was 17.44%.
For total vitamin E as TEs, RSDr was calculated as 4.90%
and RSDR was 5.68%.
Soy Product
For retinyl palmitate, RSDr was 6.84% and RSDR was 9.66%.
For α-tocopherol acetate, RSDr was calculated as 1.67% and
RSDR was 6.47%.
For α-tocopherol, RSDr was calculated as 7.89% and RSDR
was 8.74%.
For total vitamin E as TEs, RSDr was calculated as 2.07%
and RSDR was 4.22%.
For two blind duplicate samples of the milk-based
hydrolyzed protein formula, where the repeatability was
observed as being very high (up to 50% RSDs), repeat
analysis was performed by a selection of laboratories. All
initial results were confirmed. Repeatability obtained at the
SLV stage (duplicates over 6 separate days) was very good for
this matrix, 5.4 and 5.78% for cis and trans retinyl palmitate
and 5.5 and 10.2% for α-tocopherol acetate and α-tocopherol.
Results obtained for this sample were deemed invalid due
to the material being expired at time of testing following
an investigation and confirmation that other study directors
encountered similar difficulties and following a discussion at
the September 2014 ERP meeting. The results for this sample
have been removed from the AOAC Final Action Method
document.
AOAC INTERNATIONAL Interlaboratory Study
Workbook Revision 2.1 (6) was used to perform the statistical
evaluation of the data. Outliers were detected and reviewed
prior to exclusion from the data set. Where possible, detected
outliers were reviewed by the participating laboratories for
possible transcription or calculation errors. Across the four
anaytes, 13-cis and all-trans Vitamin A, and α-tocopherol
acetate and α-tocopherol, laboratories 4–8 and 10–12 had no
statistical outliers. Laboratory 13 had one outlier, laboratories
1 and 15 had 2 outliers, laboratory 2 had 3 outliers, laboratory
3 had 4 outliers, laboratories 9 and 14 had 14 outliers. The
statistical summary (Tables 1-20) lists the outliers and the
reasons for removal (Cochran test outlier or Single Grubbs
test outlier), as well as Cochran test or single Grubbs test
outliers.
An invitation was sent to all participating laboratories
to comment on the performance of the method in their
laboratories. In general, the comments were very positive.
Laboratory 3 indicated that sample preparation was easy.
According to the laboratory, it is not necessary to prepare the
working standard (6 point curve) fresh daily, as the reagent
used to prepare the stock and the working standards is the
same. Laboratory 3 also indicated that the papain solution
should be sonicated to ensure it is fully dissolved. Laboratory 4
changed the dilution of the standards to be within the working
range of the spectrophotometric measurements and HPLC
calibration curves. It used an injection volume of 35 µL
instead of 50. An injection volume between 20 and 100 μL
can be used with the method, depending on sensitivity. The
laboratory indicated that 35 µL presented areas (or heights)
to give optimum quantification and optimization. Laboratory
5 commented that the sample preparation procedure is
182