AOAC SPIFAN Nutrients ERP Review (March 14, 2019)

2016.13 (Jan. 2019) Carot-02 MLT Report FOR ERP USE ONLY DO NOT DISTRIBUTE

trans-lycopene and any peaks between the first cis peak and all-trans-lycopene are also cis peaks. The method will be updated to give better instructions for identifying lycopene peaks. 4. Broadened peaks observed for some of the lutein samples. Suspect this may be due to unhydrolyzed lipids still present in the sample. Study director’s response: This is likely the cause of the problem. Because carotenoids readily isomerize and the method aims to measure cis isomers, a short KOH treatment at room temperature is used instead of full saponification. Consequently, not all lipids are hydrolyzed. The residual lipids necessitate the use of 2-propanol as the sample solvent instead of mobile phase, and this also makes the separation of lutein isomers challenging. Given these constraints, it is especially important to use a small injection volume and to plumb the LC system to minimize extra-column volume. 5. The analytical column was a YMC C30 Carotenoid, 3 µm, 2.0 × 150 mm. A Phenomenex SecurityGuard cartridge, C18 Gemini, 4 × 3 mm ID was used in the guard column. The HPLC program has been changed, and the resolution is now slightly better than it was when I analysed the practice sample. I have used the following flow rate: 0 min, 0.25 mL/min; 25 min, 0.25 mL/min; 26 min, 0.35 mL/min; 33 min, 0.35 mL/min; 34 min, 0.25 mL/min. The gradient program was as follows: 0 min, 3% B; 1 min, 10% B; 8 min, 15% B; 25 min, 100% B, 25.5 min, 3% B, 34 min, 3% B. The changed program ended after 34 min. Study director’s response: Your values are in the expected ranges and it appears your modifications work well. Thank you for sharing what you learned. 6. We presumably found two further cis isomers of β-carotene in some samples between all trans and 9 cis β-carotene which are not included into the reported results. Study director’s response: In some samples, there are minor α- and β-carotene peaks which are not included in the calculations. Refer to the example chromatograms for peak identification. Conclusions Ten laboratories representing 7 countries completed this multi-laboratory study of carotenoids in infant formula and adult nutritionals. The method showed acceptable accuracy with NIST SRM 1869. Precision data for all-trans-β-carotene, total β-carotene, and lycopene showed that the method meets the SMPR for these nutrients in all of the valid matrices. Although the method does not meet the SMPR for total lutein, HorRat values indicate that lutein precision is better than expected based on historical data. Recommendations It is recommended that AOAC 2016.13 be endorsed as a Final Action method for determination of β-carotene and lycopene in infant formula and adult nutritionals. The ERP should review the precision data for lutein and determine if it is acceptable.

27