Location: Methods and Application of Food Composition Laboratory
2022 Annual Report
Objectives
Objective 1. Determine the impact of industrial packaging methods (canning, freezing and drying) on the nutrients and bioactive compounds in fresh fruits and vegetables.
Objective 2. Validate a software program based on mathematical optimization techniques for estimating nutrient contents of commercial multi-ingredient foods.
Objective 3. Determine the impact of dietary fiber methodology on fiber composition and intake estimates.
Approach
Objective 1. Industrial processing alters nutrients/bioactive compounds in fruits and vegetables. A 2-step, 2-year study for sample collection will be conducted. Consulting USDA plant scientists/other collaborators, multiple same-cultivar ripe samples will be collected simultaneously from one. Samples will be analyzed for vitamins, fiber, minerals, polyphenols, and metabolomics (baseline). Portions will be transported and stored to emulate typical commercial storage conditions; nutrients/polyphenols will be conducted in stored raw samples every 3 days until they decay. Shipping practices for samples will be simulated through collaborations with processing plants near harvest locations, emulating agricultural and industrial practices. Analyses at 0, 14, 35, 70, 120, 180 and 360 days post processing, and analysis in validated commercial laboratories, using AOAC methods, polyphenolic compounds analysis by ARS/academic collaborators will address the impact of processing.
Objective 2. Linear programming software for estimating missing nutrient values in commercially processed foods (using label values, ingredient lists) requires improved automation and analytical ingredient data. Food types (e.g., baked products) and nutrients will be identified, program functionality improvements completed, and tests, where ingredient proportions and nutrient values are known, will allow determination of estimation accuracy. The Virginia Tech (VT) Food Analysis Laboratory Control Center will prepare foods and QC materials for analysis; food manufacturers will be consulted on ingredient proportions, and an equivalence study of the estimates will be conducted to determine classes of nutrients and food types where the estimated and analytical values are similar, i.e. within ± 20%. Program validation will ensue to assess which nutrients to include.
Objective 3. The McCleary method (MCF) is a more complete determination of dietary fiber (DF) content in foods compared to the enzymatic-gravimetric (EGF) method, enabling better intake estimates. Select high-fiber foods and frequently consumed, fiber-containing foods will be analyzed by EGF (985.29) and MCF (2011.25) methods at a USDA-qualified commercial analytical lab. Foods with isolated or synthetic non-digestible carbohydrates may be analyzed. EGF (AOAC 985.29; total DF) and EGF (AOAC 991.43; soluble and insoluble DF), and MCF (AOAC 2009.01 and 2011.25 (fractions) will be studied and summed. Sumswill determine the food types where the fiber method used does not make a difference for measuring total DF. This allowsbetter understanding of the effect of fiber methodologies for selecting the appropriate analytical method for specific foods.
Progress Report
Objective 1. Determine the impact of industrial packaging methods (canning, freezing, and drying) on the nutrients and bioactive compounds in fresh fruits and vegetables. The effects of post-harvesting, cooking, and industrial processing on nutrients and bioactive compounds in sweet corn was investigated; 7 types of samples (3 raw, 2 cooked samples and 2 industrially processed, 39 total) were analyzed for carotenoids, carbohydrates, minerals, and phenolic compounds. Prior to the analysis of free and bound carotenoids, an analytical method including extraction, High Performance Liquid Chromatography (HPLC) separation/quantification was evaluated and optimized. Different fractions of carbohydrates (sugar, starch, dietary fiber) and minerals were measured by commercial labs; statistical analyses were carried out for carotenoids, carbohydrates, and minerals to support 3 manuscripts. The analytical method for phenolic compounds was optimized for use in this study. Select samples of raw and processed/cooked sweet corn are analyzed by University of California, Davis to examine if and how processing/cooking affect the structures of complex carbohydrates. A separate study specifically looked at the effects of cooking (steaming, boiling) on the carbohydrates of sweet corn, and compared between raw and cooked samples; these results were published in ACS Food Science and Technology. A review article “Are processed tomato products as nutritious as fresh tomatoes? - effects of industrial processing on nutrients and bioactive compounds in tomatoes” was published in Advances in Nutrition, will help design experiments for tomato analyses. A proposal on the systematic investigation of cooking and canning on the content and structure of dietary fiber in sweet corn and the effects on modulation of gut microbiome was submitted by Beltsville Human Nutrition Center (BHNRC) scientists to USDA ARS Research Associate Program, Class of 2023.
Objective 2. Validate a software program based on mathematical optimization techniques for estimating nutrient contents of multi-ingredient foods. Work on development and expansion of IngID, a system for parsing and systematic reporting of ingredients used in commercial packaged foods continued. IngID was tested to characterize top-selling U.S. baked products by determining the distribution and co-occurrence of ingredients used and its potential to answer research questions such as types of flours, fats, sweeteners, additives, etc. A manuscript was submitted to the Journal of Food Composition and Analysis. The thesaurus file, a component of IngID was further expanded to over 30,000 ingredients. Ingredient statements from USDA’s Global Branded Food Products Database were used to predict food categories and nutrients using machine learning techniques. In collaboration with University of Maryland, several techniques were developed and tested, achieving up to 99% accuracy for food classification and 0.98 R2 for calcium estimation (0.93~0.97 for calories, protein, sodium, total carbohydrate, total lipids, etc.).
Objective 3. Determine the impact of dietary fiber methodology on fiber composition and intake estimates. High carbohydrate foods were identified to support research on the impact of dietary fiber methods and changes in the distribution of carbohydrate fractions during storage, senescence, and cooking e.g., corn, potatoes. Work includes measurement of dietary fiber (enzymatic gravimetric AOAC 991.43), McCleary fiber (AOAC 2009.01, 2011.25), starch (digestible, residual, and retrograde, after cooking/cooling/reheating) and oligosaccharides; data are included in FoodData Central (FDC) Foundation Foods. Alpha- and beta-glyosidic linkages were analyzed by collaborators at the UCDavis for potato starch and oat flour. We analyzed fresh harvested corn for soluble and insoluble dietary fiber. Sampling of potatoes (multiple cultivars (fresh and cooked amylose and amylopectin), additional pulses/legumes, and corn (fresh and cooked) are underway. Discussions with Methods and Application of Food Composition Laboratory (MAFCL) chemists and the Institute for the Advancement of Food and Nutrition are involved. Review of external carbohydrate data/datasets are being reviewed for linking and/or inclusion in FDC.
Human Breast Milk Composition. MAFCL co-leads the Human Milk Composition Initiative (HMCI), a joint undertaking by federal U.S. and Canada agencies to articulate human milk-related data needs relevant to federal programs, policies, and regulations. It provides a framework for collection and reporting of human milk composition data and metadata. A scoping review on current literation on human milk composition and volume and listening sessions with stakeholders has been initiated. With contributions from 19 federal scientists, it fulfils an important public health need and gap in literature to inform future public health monitoring efforts.
Iodine: Low iodine status is a public health issue due to iodine’s crucial role in fetal development; iodine intake is inadequate for about 20% of U.S. women of childbearing age. Iodine data for foods has been scarce so MAFCL, Food and Drug Administration (FDA) and Office of Dietary Supplements, ODS-NIH scientists are collaborating to provide iodine data for foods and dietary supplements. Data for 430 foods (2020), were released in amounts per serving in 2022 and linked to food intakes from survey data (NHANES WWEIA 2013-14 through 2020) for estimating iodine intake for the U.S. population, based on participants’ urinary iodine measurements. Bi-monthly sampling for a national study of iodine variability in retail milk over one year in twelve locations was completed. Samples of other foods were collected for laboratory analysis and inclusion in the database. Data support health policy and nutrition guidance for a vulnerable population and 2 manuscripts were published. The Iodine Special Interest Database serves as a model for SIDs for Nitrates/Nitrites, Purines and Glucosinolates, and ancillary datasets for Iodine Uptake Inhibitors perchlorate and thiocyanate, in progress.
Dietary Supplements (DS). Nutrient content and release from prescription prenatal Multivitamins (RxP MVM) solid oral dosage forms (tablets, caplets and soft gels). Pregnant women in the U.S. are at risk of dietary deficiencies of calcium, vitamin D, iron and folate. Folic acid and iron supplementation is recommended in the periconceptional period, during pregnancy and lactation to reduce risks of neural tube defects in infants, iron deficiency anemia in pregnant women and related adverse consequences in their infants. The goals of this Dietary Supplement Ingredient Database study were to measure micronutrients in RxP MVMs and to assess dosage form adherence to the USP disintegration/rupture and dissolution standards set for MVM dosage forms in the United States Pharmacopeial general chapter <2040>. Multiple lots of 24 RxP MVMs, representing 61% of the market were tested by commercial laboratories. The USP dissolution testing for folic acid, iron, riboflavin and vitamin A/retinol was applied - all 12 soft gel RxP MVM passed the rupture test and most of 19 tablets did pass the disintegration test and dissolution test for at least one or two index nutrients, the entire sample of nationally representative (and high consumption) RxP MVMs failed to meet dissolution standards for all three-four index nutrients, as required by USP for compliance. Conclusion: The analyzed sample of RxP MVMs contained several micronutrients which significantly exceeding the RDA; this could close nutrient gaps but failure to release critical micronutrients raises concerns for the health of periconceptional, pregnant or lactating women, and newborns because of unavailability for absorption in the GI tract.
Cranberry DS: proanthocyanidin (PAC) content and dosage form disintegration assessed with USP standards. Several clinical trials have shown that consumption of cranberry juice or taking cranberry DS significantly reduced the recurrence of urinary tract infections (UTI) compared to a placebo. A-type PACs appear to be critical for changes in human metabolomes that could mediate the anti-UTI effects of cranberries; intake of cranberry juice/DS may overcome the bacterial resistance and reduce antibiotic use. Cranberries, ranked 6th in herbal ingredients sold in the U.S. (Nutrition Business Journal, 2020) are allowed in FDA-approved health claims but based on the weight of botanical material; this doesn’t adequately inform researchers and consumers about the precise but variable content of bioactives. The range of type A and type B PACs in cranberry DS and ability by dosage form to disintegrate and release bioactives in USP tests are largely unknown. This multi-laboratory study: 1. Evaluates popular cranberry DS for their total PAC content (by 4-di-methylaminocinnamaldehyde (DMAC) colorimetric assay used in industry label claims); 2. Evaluates new developments in testing of PAC content e.g., separating/quantifying A-type and B-type PAC content (cranberries are predominantly A-type PAC so all products are tested for their ratio of A-type and B-type PACs to determine identity and exclude adulteration with non-cranberry PACs ); and 3. Tests the cranberry DS against the USP disintegration standards. Popular cranberry DS were identified using the National Health and Nutrition Examination Survey, retail sales data from SPINS, the Dietary Supplement Label Database and web-based DS sellers; label categories were identified as those with a trademarked cranberry ingredient, those without a trademarked cranberry ingredient, and those with a numeric PAC claim (e.g., DS containing branded ingredients that were used in clinical trials) and a variety of dosage forms and DS forms.
Accomplishments
1. Ingredients in commercial packaged foods. There is general lack of information in scientific literature on the ingredients used in commercial packaged foods. Researchers at Beltsville, Maryland, in collaboration with University of Maryland, recently developed and tested a prototype of IngID, a framework for parsing and systematic reporting of ingredients used in commercially packaged foods in the U.S., using ingredient statements of baked products mainly from USDA’s Global Branded Food Products Database. IngID enables characterization of what is in the foods we eat in a systematic manner, in dimensions other than the traditional nutrient profiles. The ingredient statements were also used to predict food categories and nutrients, using Artificial intelligence techniques. The researchers achieved up to ninty-nine percent accuracy for food classification for nutrient estimation. Such approaches can potentially be used for other such challenging and resource-intensive applications.
Review Publications
Pehrsson, P.R., Roseland, J., Patterson, K., Phillips, K., Spungen, J., Andrews, K. 2022. Iodine in foods and dietary supplements: A collaborative tool by ODS-NIH, FDA and USDA. Journal of Food Composition and Analysis. https://doi.org/10.1016/j.jfca.2021.104369.
Fukagawa, N.K., Mckillop, K.A., Pehrsson, P.R., Moshfegh, A.J., Harnly, J.M., Finley, J.W. 2021. USDA’s FoodData Central: What is it? and why is it needed today? American Journal of Clinical Nutrition. https://doi.org/10.1093/ajcn/nqab397.
Durrazo, A., Sorkin, B.C., Lucarini, M., Gusev, P.A., Kuszak, A.J., Crawford, C., Boyd, C., Deuster, P.A., Saldanha, L., Gurley, B.J., Pehrsson, P.R., Harnly, J.M., Turrini, A., Andrews, K.W., Lindsey, A. 2022. Analytical challenges and metrological approaches to ensuring dietary supplement quality: International perspectives. Frontiers in Pharmacology. 12:1-23. https://doi.org/10.3389/fphar.2021.714434.
Ma, P., Wang, Q., Yu, N., Li, Y., Zhang, Z., Sheng, J., Ahuja, J.K., Mcginty, H. 2022. Deep learning accurately predicts food categories and nutrients based on ingredient statements. Food Chemistry. 391:133243. https://doi.org/10.1016/j.foodchem.2022.133243.
Wu, X., Pehrsson, P.R. 2021. Current knowledge and challenges on the development of a dietary glucosinolate database in the U.S. Current Developments in Nutrition. 5:nzab102. https://doi.org/10.1093/cdn/nzab102.
Wu, X., Yu, L., Pehrsson, P.R. 2021. Are processed tomato products as nutritious as fresh tomatoes? Scoping review on the effects of industrial processing on nutrients and bioactive compounds in tomatoes. Advances in Nutrition. 13(1):138-151. https://doi.org/10.1093/advances/nmab109.
Zhang, W., Zhu, B., Childs, H., Yu, L., Whent, M.M., Pehrsson, P.R., Zhao, J., Wu, X., Li, S. 2022. Effects of boiling and steaming on the carbohydrates of sweet corn. ACS Food Science and Technology. 2:951-960. https://doi.org/10.1021/acsfoodscitech.2c00103.
Ershow, A., Haggans, C., Roseland, J., Patterson, K., Spungen, J., Gahche, J., Merkel, J., Pehrsson, P.R. 2022. Databases of iodine content of foods and dietary supplements–Availability of new and updated resources. Journal of the Academy of Nutrition and Dietetics. https://doi.org/10.1016/j.jand.2022.03.017.
Roseland, J.R., Bahadur, R., Pehrsson, P.R. 2022. Iodine and vitamin D content and variability in U.S. shell eggs and processed eggs. Journal of Food Composition and Analysis. https://doi.org/10.1016/j.jfca.2021.104166.
Dwyer, J., Saldanha, L., Bailen, R., Durazzo, A., Le Donne, C., Piccinelli, R., Andrews, K., Pehrsson, P.R., Gusev, P., Calvillo, A., Connor, E., Goshorn, J., Sette, S., Lucarini, M., D’Addezio, L., Camilli, E., Marletta, L., Turrini, A. 2021. An impossible dream? Integrating dietary supplement label databases: Needs, challenges, and next steps. Journal of Food Composition and Analysis. https://doi.org/10.1016/j.jfca.2021.103882.
Dwyer, J., Saldanha, L., Bailen, R., Gahche, J., Potischman, N., Bailey, R., Jun, S., Long, Y., Connor, E., Andrews, K., Pehrsson, P.R., Gusev, P. 2022. Do multivitamin/mineral dietary supplements for young children fill critical nutrient gaps? Journal of the Academy of Nutrition and Dietetics. https://doi.org/10.1016/j.jand.2021.10.019.
