Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BHNRC) » Beltsville Human Nutrition Research Center » Nutrient Data Laboratory » Research » Research Project #426694

Research Project: USDA Dietary Supplement Ingredient Database

Location: Nutrient Data Laboratory

2016 Annual Report


1a. Objectives (from AD-416):
The Dietary Supplement Ingredient Database (DSID) (http://dsid.usda.nih.gov; (U.S. Department of Agriculture, 2012) currently provides nationally representative supplement composition data for adult and children’s MVM, based on chemical analysis. All other dietary supplement databases contain label information only. In addition to the evaluation of vitamin and mineral levels in dietary supplements, the DSID will be expanded to include supplements with other ingredients of interest, including omega-3 (n-3) fatty acids in fish and plant oil supplements, and catechins, caffeine, and flavonoids in botanical supplements. Objective 1: Determine, evaluate, and expand knowledge about the composition of frequently consumed dietary supplements in the U.S. by collecting and compiling data and information on essential nutrients and emerging components of public health importance. This includes developing methods for sampling new dietary supplement product types, statistically evaluating analytical results to portray representative estimates, and improving the value of the Dietary Supplement Ingredient Database by utilizing new technologies to generate novel datasets.


1b. Approach (from AD-416):
Several types of studies are planned, including ongoing national and monitoring studies and new pilot and national studies. NDL will design and carry out research to analytically determine the ingredient composition of representative dietary supplements, including essential nutrients and bioactive components. NDL will identify laboratories for sample analysis and evaluate data quality. Results will be statistically evaluated and made publicly available through the DSID. On-going DSID Studies: 1) Omega-3 fatty acid national study: scientifically evaluate the relationships between labeled and analytical levels of omega-3 fatty acids in representative dietary supplements, including fish oil, plant oil and fish/plant oil blends. Release mean results and variability information from this study in DSID-3. 2) Over-the-Counter (OTC) prenatal MVM national study: scientifically evaluate the relationship between label and analytical values and assess the variability for up to 22 vitamins and minerals in OTC prenatal MVM. Release mean results and variability information from this study in DSID-3. 3) Adult MVM monitoring study: systematically evaluate the statistical relationships between label and analytical data since the first adult MVM study in a monitoring study of adult MVM, and release results in DSID-4. The information from this study will be used to plan the frequency and scope of updates to DSID for other MVM and possibly other supplement categories. Pilot Studies: 1) Green tea study AND flavonoid-containing dietary supplements (DS) study: methods of analysis will be evaluated by testing representative and top-selling products for prioritized ingredients of interest. 2) Calcium in DS and OTC antacids study: DS and OTC antacids, a major source of supplemental calcium intake in the U.S., will be assessed for their calcium and vitamin D content. National Studies: 1) Green tea DS and flavonoid-containing DS studies: obtain mean estimates for and measures of variability for analyzed catechins, caffeine, and for analyzed flavonoids (such as flavonols, flavones, flavanones, catechins, and anthocyanidins) in nationally representative botanical DS. 2) Calcium in DS and OTC antacids study: determine national estimates for ingredients in calcium-containing DS, a commonly reported supplement


3. Progress Report:
Objective 1: Determine, evaluate, and expand knowledge about the composition of frequently consumed dietary supplements in the U.S. by collecting and compiling data and information on essential nutrients and emerging components of public health importance. Adult MVM 2 study. Release 3 of the DSID (http://dsid.usda.nih.gov) reports analytical estimates from the adult MVM-1 study, and children’s and non-prescription prenatal MVM studies. The DSID-3 predicted estimates for adult MVMs purchased in 2006-07 are linked to NHANES 2003-2008 and could be used instead of labels for more precise nutrient intake assessment in surveys. Currently, the DSID estimates are linked to the NHANES cycles preceding, coinciding with and following the products’ year of purchase. The adult MVM-2 study was initiated to answer questions about how the adult MVM label information, sales markets and quality change over time. In addition, new ingredients were added to the DSID for this product category (vitamins A and D, iodine, and chromium). Representative adult MVM-2 products were identified and sent for chemical analysis to qualified laboratories and results were evaluated for quality. Final lab results were statistically evaluated. In both studies, overages in mean ingredient content, at the product level, were found for all vitamins (with thiamin a consistent exception) and all minerals. Improvements in analytical methodology for iodine, chromium, vitamins A and D provide reliable data for these ingredients in adult MVMs for the first time. We have found 20-40% mean overages for these ingredients which are largely within the USP requirements for these nutrients. For some ingredients the mean overages have changed significantly over the 5-year period. Among ingredients that may require updates in the DSID are iodine, magnesium, folic acid, and vitamins B-12, B-6 and E. For all other ingredients (copper, iron, manganese, phosphorus, selenium, zinc, niacin, riboflavin, thiamin) there was no difference in the overall mean percent differences from label and product distributions between the two studies. There were also changes in the type of regression equations describing the relationships between the labeled and analytical ingredient content over the 5 years: 12 out of 18 (66.7%) equations demonstrated changes in predicted content dependence from label level. Also, for some ingredients, the most common labeled levels shifted (for example, for calcium, magnesium, phosphorus, and selenium). Final statistical evaluation of the regression models and product variability in the two studies is pending. Findings above UL and below label claims. Products labeled at or above the Upper Tolerable Intake Level (UL) were among those analyzed in one or more of the MVM DSID studies for seven ingredients (vitamins A and B-6, folic acid, niacin, iron, magnesium and zinc). However, even for the label levels at the UL, overages were still measured for some ingredients. DSID findings indicate that nutrients in some products are below their label claims within the shelf life. Thiamin has a mean predicted level below label for the non-prescription prenatal MVMs (at -9.2% of label for the entire regression range) and for adult MVMs (at -6.4% of label at the most commonly labeled level). A prescription prenatal MVM pilot study is underway and is designed to determine the analytical ingredient content of representative products. The results of this study will be compared to the information for non-prescription prenatal MVMs reported above. Prescription prenatal MVMs are not categorized as approved drugs by the FDA, but as dietary supplements, which have the same manufacturing regulations as foods. Thus, the manufacturing specifications for both the non-prescription and prescription products are the same. Some of these products are sold at a much higher price than the non-prescription products. The labeled ingredient distributions, analytical content and product variability will be determined. Botanical DSID initiative. The goal of the Botanical Initiative for the DSID is to evaluate the levels of ingredients and ingredient constituents in botanical DSs, estimated to be taken by 7.5% of the US population. Labeling requirements for botanicals are based on the amount of plant material or extract, without any requirement for concentration information. The unlabeled amounts of botanical constituents may be significant in highly concentrated extracts, therefore the DSID analytical data will be highly beneficial for better phytochemical intake assessments from botanical DSs, especially since many ingredients in botanicals are also in foods or beverages (tea, flavonoids, spices, fatty acids, etc). Commonly consumed and frequently studied, green tea (GT) DSs were chosen as the first botanical DSs for the DSID botanical initiative. The most abundant GT constituents are the catechins, especially epigallocatechin gallate (EGCG), and caffeine. EGCG accounts for about 50% of the total catechins in green tea and is the catechin most studied for its health effects in clinical trials and animal studies. GT DS pilot study 1: NDL finalized results for the first GT pilot study, which was limited to products containing GT as the only or the primary botanical in order to minimize interferences from sources of the same phytochemicals in non-GT ingredients. Analytically measured EGCG amount ranged from 0.5-533 mg/serving. It was significantly different between the products with and without label claims for EGCG content (299.8 ± 206.3, n=18; 114.2 ± 95.25, n=14, respectively; mean ± SD mg/day, t-test, p=0.0024). Importantly, total green tea amount required to be present on the labels by FDA regulation did not predict analytically measured amount of EGCG: at the most commonly labeled level 500 mg/serving of green tea, the percent of the green tea amount that was EGCG ranged from 1 to over 60% (0.1 – over 300 mg per serving, n=9). Amounts of EGCG in 9 of 12 products that had label claims for EGCG were within ± 20% of such claims. Analytically measured levels of caffeine for the products not decaffeinated ranged from 0 to 158 mg/day (n=25). The mean analytical caffeine content in products without caffeine label claims was lower than in products with such claims (27.3 vs. 63.9 mg/day), similar to our 2007 caffeine study. The % difference from label for the 7 products labeled for caffeine content ranged from -84% to +71%. Analytical results for the decaffeinated products averaged 2 mg/day. GT DS pilot study-2: The goal of this study is to purchase GT products in multiple matrices and analyze them for catechins and caffeine from all botanical sources present. This study targets commonly purchased multi-ingredient DS that are estimated to contain at least 20 mg/serving green tea, regardless of the other ingredients. Epidemiological studies suggest that even a small increase in flavonoid intake may significantly decrease risk of CVD disorders. Thus, it is important to evaluate how well analytical methods capture and measure small amounts of flavonoids, especially in complex botanical DSs. Currently, 2 lots of 37 DS products and 3 in-house control materials have been purchased and these were sent for analysis of catechins and caffeine to two of the laboratories used in GT pilot study-1. Additionally, 5 NIST SRMs were purchased and these were sent for analysis with the samples and in-house control materials to evaluate laboratory accuracy and precision. Approximately ½ of the products contain GT in blends and ½ products contain GT identified separately on the label. Products were sold in a variety of forms (capsules, caplets, tablets, soft gels, gummies, liquid-filled capsules, powder) and for a variety of purposes (antioxidant blends, MVMs + botanicals, weight loss and energy products, sports supplements). Disintegration and dissolution studies on DS containing green tea: The efficacy of a dietary supplement is determined not by only by the active ingredient amount but also by the formulation design. Formulations and capsule materials can greatly influence the fraction of the ingested amount that is absorbed and reaches the target site. Therefore, in vitro disintegration and dissolution methods are being used in DSID research to assess the performance and quality of dietary supplements. In collaboration with FDA and ODS scientists, we are evaluating whether GT DSs met the general USP chapter requirements for both disintegration and dissolution. Disintegration is the ability of a tablet or capsule to break apart within a specified amount of time after being ingested, so that the ingredients become available to the body. The GT-1 and GT-2 supplements were tested for disintegration and the pass rate was 55.2% and 67%, respectively, using the USP general guidelines. Dissolution measures the amount of marker compound dissolved in one hour in a solution mimicking stomach conditions. A product passes if, after 1 hour, > 75% of the amount of the marker compound (EGCG) is dissolved in the solution. Measurement of parent vitamin D and 25(OH)D in foods and DSs. Health effects of the vitamin D metabolite, 25(OH)D, may be more potent than non-metabolized vitamin D. However, reliable sources for the analytically estimated 25(OH)D content in foods and DSs are very limited or nonexistent. An 18-month Interagency Agreement between ARS and ODS set up a pilot study to measure vitamin D and 25(OH)D in foods and DSs. The goal was to estimate the variability among both domestic and international labs in measuring 25(OH)D content reported for varied foods and matrices. In addition, the outcome of the trial may help National Institute of Standards to formulate Standard Reference Materials necessary for the harmonization of 25(OH)D methods and reporting protocols.


4. Accomplishments
1. Monitoring measured nutrient content in nationally representative samples of adult MVM products for DSID. Average content of many ingredients in adult MVMs was close to label levels, but the actual content of some vitamins and minerals may differ significantly from label claims. It is unknown how these discrepancies (that may cause errors in the assessment of the nutritional status of the US population), persist over time due to DS product reformulation. Research by ARS scientists in Beltsville, Maryland, indicates that for some ingredients the mean overages have changed significantly over a 5-year period. Among ingredients that may require updates in the DSID are iodine, magnesium, folic acid, and vitamins B-12, B-6 and E. For these ingredients, scientists conducting calculations of nutrient intake from adult MVMs can apply updated label adjustments.

2. Findings above Upper Tolerable Intake Level (UL) label claims. The concept of overages consistent with Current Good Manufacturing Practices (CGMP) is recognized in the food and DS industry as a means to ensure that claimed ingredient amounts are valid at the end of product shelf life. However, the ingredient amounts added above those claimed by the labels are unknown. The analytical nutrient content of DSs is also important for evaluating risk of exposure to doses in excess of labeled amounts and Recommended Daily Allowances (RDAs). ARS scientists in Beltsville, Maryland, demonstrated that products labeled at or above the Upper Tolerable Intake Level (UL) were among those analyzed in one or more of the MVM DSID studies for seven ingredients (vitamins A and B-6, folic acid, niacin, iron, magnesium and zinc). DSID findings indicate that manufacturers apply overages in MVM products even for high and potentially harmful levels.

3. Botanical DSID initiative. Botanical dietary supplements (DSs) are taken by 7.5% of the U.S. population. The unlabeled amounts of botanical constituents may be significant in highly concentrated extracts and should be included in nutritional surveys of phytochemical intake in addition to the intake from foods. However, labeling requirements for botanicals by FDA are based on the total amount of plant material or extract, without any requirement for concentration information. Based on an analytical pilot study of green tea (GT-1) DSs, ARS scientists in Beltsville, Maryland, concluded that labels that provide only the label information currently required by FDA do not predict actual content of GT phytochemical constituents. Information about the actual content is important for consumers who make decisions for DS use and for accurate intake assessments in epidemiologic studies.

4. Disintegration tests on DSs containing green tea. The efficacy of a DS is determined not only by the active ingredient amount but also by the formulation design. Formulations and capsule materials can greatly influence the fraction of the ingested amount that is absorbed and reaches the target site. Therefore, in vitro disintegration methods are being used in DSID research to assess the performance and quality of dietary supplements. In collaboration with FDA and ODS scientists, ARS scientists in Beltsville, Maryland, showed that significant number of DSs containing green tea extracts do not meet specifications set by the United State Pharmacopeia (USP) for in-vitro disintegration tests. Our findings may indicate the necessity of stricter quality control of DS dosage forms marketed in the US.

5. Measurement of vitamin D and 25(OH)D in foods and DSs. Assessment of total vitamin D intake from foods and DSs may be incomplete if 25-hydroxyvitamin D [25(OH)D] intake is not included. However, 25(OH)D data for such intake assessments are lacking and no food or DS reference materials (SRMs) are available. ARS scientists in Beltsville, Maryland, selected food and DS samples and analyzed data from five experienced laboratories from the United States and other countries, all using liquid chromatography tandem-mass spectrometry but no common analytical protocol for determining vitamin D3 in the DSs. The results demonstrated that vitamin D3 and 25(OH)D3 concentrations in food and DS materials could be measured with acceptable reproducibility. Thus, it is possible to obtain consistent results among experienced laboratories for vitamin D3 and 25(OH)D3 in foods and a DSs and to collect reliable data for food composition databases.


5. Significant Activities that Support Special Target Populations:
Dietary supplements are consumed by various age, ethnic and socioeconomic populations. The DSID-3 has published data for children’s MVM supplements and prenatal MVM supplements. Data acquired through DSID studies will be used to monitor the total dietary intake and health status of children and pregnant women who are considered a high-risk population sub-groups. The national sampling program was designed to select sample units from all marketing channels (retail, multi-level, Internet) and from a large selection of known or popular brands to assure that products purchased by diverse populations are included.


Review Publications
Roseland, J.M., Patterson, K., Andrews, K.W., Phillips, K.M., Phillips, M.M., Pehrsson, P.R., Dufresne, G.L., Jakobsen, J., Makowski, A.J., Scheuerell, C.R., Larouche, G.P., Gusev, P.A., Savarala, S., Nguyen, Q.V., Taylor, C.L., Wise, S.A., Harnly, J.M., Williams, J.R., Betz, J.M. 2016. Interlaboratory trial for measurement of vitamin D and 25(OH)D in foods and a dietary supplement using liquid chromatography-mass spectrometry. Journal of Agricultural and Food Chemistry. 64:3167-3175.