Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 11/5/2008
Publication Date: 11/8/2008
Citation: Andrews, K., Roseland, J.M., Holden, J.M. Status of Vitamin Methodology based on SRM 3280 Results for the DSID Project. Dietary Supplement Vitamin Methodology Workshop, Gaithersburg, MD, July 7-8, 2008.
Technical Abstract: The Nutrient Data Laboratory (NDL), Beltsville Human Nutrition Research Center, U.S. Department of Agriculture is working with the Office of Dietary Supplements and other federal agencies to plan and develop a Dietary Supplement Ingredient Database (DSID) to provide levels of ingredients in dietary supplement products. The major product of this research will be a publicly available ingredient database with reliable estimates of nutrients and other bioactive components in dietary supplements. Initial analytical work for the Dietary Supplement Ingredient Database (DSID) focused on multivitamin/minerals (MVMs) due to their high prevalence of use. Pilot studies were initiated to review sample handling procedures and methods of analysis and to qualify laboratories for the analysis of nutrients and other dietary components in supplement products. Preliminary assessments of precision and accuracy for methods and laboratories were established in 2005. These assessments were made based on the results of several laboratories’ analyses of 11 vitamins in the NIST SRM 3280 MVM material, then under development. These vitamins were measured 5 or 6 times over a period of several weeks or months. After acceptable methods and labs were identified by a federal expert panel, two MVM studies were undertaken using the NIST MVM product as a control material. This presentation summarizes the ongoing assessments of laboratory method accuracy and variability using the NIST MVM as a reference material. Mean values and % relative standard deviations (% RSD)were generated for NDL by contracted laboratories over the past 2 years. In general, NDL’s analytical measurements of NIST SRM 3280 MVM were closer to the 2008 preliminary NIST values determined by NIST for the water-soluble vitamins than for the fat-soluble vitamins. NDL results for 7 vitamins (niacin, riboflavin, thiamin, vitamin B6, vitamin C, folic acid, and alpha-tocopherol) showed % RSDs < 10% and mean values (n = 20 to 23) < +10% of the 2008 preliminary NIST means. For vitamin B12 analyzed using a microbiological method, the mean NDL value was within 1% of the 2008 preliminary NIST mean and the % RSD (<15) was much lower than was initially determined as acceptable (<30%). Beta-carotene was measured in two laboratories and mean results for both were < 10% of the 2008 preliminary NIST mean with % RSDs of 15 and 22. Laboratory analysis of two of the fat-soluble vitamins (vitamin D and retinol) is currently being re-evaluated because NDL results for the NIST SRM 3280 MVM are outside of acceptable ranges. These results indicate that for some nutrients, especially water-soluble vitamins, the methods of analysis used in this study are generally acceptable for the analysis of SRM 3280. The fat-soluble vitamins, except for alpha-tocopherol, appear to be more difficult to obtain consistent and accurate analyses. Some of the research questions that remain concern the best methods for grinding and homogenizing dietary supplements and for extracting encapsulated nutrients.