Characterization of Final Action Official MethodSM 2011.19 and First Action Official Method 2015.06 Performance at Analyte Levels Corresponding to CODEX STAN 72 (1981) Minimum Levels.

Research paper by Joseph J Thompson, Lawrence L H Pacquette

Indexed on: 25 Jan '17Published on: 25 Jan '17Published in: Journal of AOAC International


A limited single-laboratory validation (SLV) was conducted in the authors' laboratory to investigate the performance ofAOAC Official Methods 2011.19 Determination of Chromium (Cr), Selenium (Se), and Molybdenum (Mo) in Infant Formula and Adult Nutritional Products by Inductively Coupled Plasma/Mass Spectrometry and 2015.06 Determination of Minerals and Trace Elements in Infant Formula and Adult/Pediatric Nutritional Formula by Inductively Coupled Plasma/Mass Spectrometry at analyte levels below the practical LOQs (PLOQs) already published for these Final Action Official Methods. This work was needed to verify that the actual LOQs were below the minimum requirements for minerals in infant formula as given in CODEX STAN 72 (1981). Linearity studies at low levels were conducted as well as the analysis of blanks over multiple days to establish the LOQs (as opposed to PLOQs) for these nutrients. Several placebo matrixes from the AOAC Stakeholder Panel on Infant Formula and Adult Nutritionals (SPIFAN) program were tested over multiple days at two different sample sizes to quantitate the effect of doubling the sample size given in the original publications. The SLV results indicate that both methods can meet the Codex minimum requirements as-is, without modification of the methods, albeit with a relaxation of the stringent precision criteria originally established for these methods by SPIFAN. Precision can be improved by doubling the sample size, but this step is not necessary to use the method for its intended purpose. A concurrent collaborative study of Method 2015.06 showed that the RSDR obtained across eight laboratories for several infant formula placebos containing mineral concentrations between the PLOQ and LOQ were indeed worse than SPIFAN expectations, but reasonable Horwitz ratios (HorRat) were nonetheless obtained for these analytes.