424 / 670
weighting was used for both analytes. Confirmation was achieved through analysis of ion ratios
between samples and reference standards for at least one additional transition listed in the table.
The concentration of each analyte in a sample was calculated by the following equation:
ܴ ൌ ൈ ൈ ܵ
ൈ ͳͲͲ ͳͲ
Ȁ
where R = results expressed in mg/100g, C = concentration of the analyte in the injected solution
in ng/mL, V = volume of the initial extract in mL, S = sample weight in g, and D = dilution
factor, the inverse of any dilution made. All results were calculated on a ready to feed or
reconstituted basis of 25 g diluted to 225 g with water, except for SRM 1849a that was calculated
back to powder basis.
Results and Discussion
Linearity was assessed from the percent deviation from theoretical concentration across
the working standard range (Table 3). The results from the validation showed an overall
repeatability for free and total choline of 1.9 and 2.3 %RSD
r
, while the overall intermediate
precision obtained for free and total choline was 2.4 and 2.7 %RSD
INT
, respectively (Table 4).
Free and total carnitine had an overall repeatability of 2.9 and 2.7 %RSDr, while the overall
intermediate precision obtained for free and total carnitine was 3.3 and 3.1 %RSD
INT
,
respectively (Table 5). The average results obtained from the analysis of NIST SRM 1849a were
within the certified ranges or close to the information mass fraction value. Only an information
mass fraction value is given for free choline, while there is currently nothing provided for total
carnitine from NIST. The average total recoveries (endogenous + added) shown in Tables 6 and
7 for all matrices tested from LOQ to the upper ranges required in the SMPR were 95.9
–
103.6%. Analysis of bound sources of carnitine and choline analyzed in duplicate over three
days gave average recoveries of 104.6% for acetylcholine, 96.7% for phosphatidylcholine, and
Chol-08/Carn-07 (February 2016)
FOR ERP USE ONLY
DO NOT DISTRIBUTE