AOAC SPIFAN Nutrients ERP-Final Review (Jan. 2019)

2015.10 (Jan. 2019) Carn-07 MLT Report FOR ERP USE ONLY DO NOT DISTRIBUTE

Laboratory Qualification Two qualification samples and the NIST SRM 1849a were provided to be run for the total carnitine/choline method. These qualification samples were used to optimize each participant’s instrument performance and chromatography. Before proceeding with the MLT, each participant’s qualification results were sent to and approved by the Study Director. After preparation of the stock and working solutions, the performance of the instruments was determined. Full curves for carnitine and choline were injected and linearity was assessed. Coefficients of determination (r 2 ) were required to be 0.995 or greater and back-calculated concentrations of the calibration standards did not exceed ±10% of theoretical. Signal to noise for the low calibration standard was at least 5:1 for each analyte. Solvent blanks and extracted blanks were used for carryover demonstration that was acceptable at <20% peak area counts of the low calibration standard level. After the above criteria were met, the analyst could analyze the qualification samples and extracted method blanks. The qualification samples had expected amounts of choline or carnitine. After analysis of the qualification samples, the laboratories sent the qualification result form to the Study Director to review the data. After successful qualification sample analysis, the data was approved by the Study Director and the analyst proceeded with the analysis of the test samples. Data Reporting Participants were asked to record all observations and any potential method deviations. In addition they were to investigate any potential aberrant results, for example incorrect calculations, use of wrong units, transposition errors, incorrect standard preparation or contamination. It was also a requirement to have all the results and calculations reviewed by a peer, laboratory supervisor or manager. Data Analysis The Study Director reviewed and compiled all the data submitted by the participants. Statistical analysis was conducted to determine outliers using Grubbs tests for removal of laboratories with extreme results. The spreadsheet was used to calculate the following: • standard deviation (S r ) and relative standard deviation (RSD r ) for repeatability • standard deviation (S R ) and relative standard deviation (RSD R ) for reproducibility • number of valid data points • HorRat value (RSD R /predicted RSD R ) Discussion Ten samples were analyzed as blind duplicates by 9 independent laboratories for both free and total carnitine/choline. These laboratories were from industry and contract research organizations. Laboratories were located in North America, Europe, and Asia. Sample results are presented in concentrations of mg/100g, indicating different concentrations in many cases between liquids and powders. Results for carnitine and

- 4 -