Location: Methods and Application of Food Composition Laboratory
2021 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
Determination of the impact of industrial packaging methods (canning, freezing and drying) on the nutrients and bioactive compounds in fresh fruits and vegetables. Some of the work proposed for FY21 were completed in FY20, including the nutrient analysis of the samples from the pilot study and the statistical analysis. In FY21, additional statistical analysis was performed to determine the sampling plan. Detailed research plan for sweet corn was developed. Due to COVID-19 pandemic, no additional samples were picked at the farm during 2020 harvest season, and no lab work was carried out. Thorough literature reviews were conducted for another two vegetables – tomato and spinach, to develop the research plans. A review article entitled “Are processed tomato products as nutritious as fresh tomatoes? - effects of industrial processing on nutrients and bioactive compounds in tomatoes” was submitted to Advances in Nutrition for publication. In addition, thorough literature review on an in vitro digestion model (INFOGEST) was conducted. This method is included in the current research plan as an addition to assess the effects of processing/cooking on the digestibility of nutrients and bioactive compounds in foods. The physiological effects of dietary compounds are related to their fate during processing and in the gastrointestinal tract, the data obtained from the proposed studies will help us better elucidate the relationships between the selected foods and health.
Validation of a software program based on mathematical optimization techniques for estimating nutrient contents of multi-ingredient foods. Ingredients were parsed from ingredient statements of top selling commercial 15 food categories (~14,500 food products) - baked products, beverages and mixed dishes, using an in-house custom program. Ingredient statements were very inconsistent and several major challenges were identified. The identified parsed ingredients were reviewed to identify equivalent ingredients, such as synonyms, spelling and usage variants, common names, possible errors, etc. The parsed list of ingredients was used to develop a thesaurus of ~ 16,000 ingredients, using standardized formats and procedures. These form the basis for IngID, a framework for parsing and systematically reporting ingredients used in commercially packaged foods. Manuscripts to characterize baked products and mixed dishes sold in the U.S., as an application of IngID is under development. A list of most frequently used flours were identified to prioritize for inclusion in Food Data Central (FDC) Foundation Foods. These data are also being used for a project to categorize foods based on their description and ingredients using machine learning; a manuscript is being is being drafted.
Determination of 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: Garlic, onions, apples, flours, and corn. A publication for changes in slightly, ripe, and overripe bananas was accepted and presented with data for how the ripeness stage affects changes in carbohydrate composition. Objective 3 was expanded to include 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 in highly consumed, high carbohydrate commodity foods. All data were included in FDC Foundation Foods. Alpha- and beta-glyosidic linkages were analyzed by collaborators at the University of California, Davis. Sampling frames for potatoes (multiple cultivars (fresh and cooked amylose and amylopectin), additional pulses,/legumes, and corn (fresh and cooked) are underway. Discussions with the Carbohydrate Committee of IAFNS set the stage for collaboration and includes scientists at BHNRC chemists/scientists in this area. Review of external carbohydrate data/datasets are being reviewed for linking and/or inclusion in FDC.
Human Breast Milk Composition: Methods and Application of Food Composition Laboratory (MAFCL) co-leads the Human Milk Composition Initiative (HMCI), a joint undertaking by federal U.S. and Canada agencies to articulate human milk (HM)-related data needs relevant to federal programs, policies, and regulations. A manuscript series exemplifying potential public health opportunities in HM research is underway with perspectives from 50 scientists regarding HM composition data and metadata, public health relevance of such data to US and Canadian populations, and potential uses of data to support federal programs and policies.
Macro- and micro-nutrients in raw plant foods. A tool for categorizing raw plant foods based on similarity of nutrients was developed. Correlations among 25 macro- and micronutrients of 268 raw plant foods across five food categories (fruits, vegetables, legumes, grains, and nuts) were analyzed. Distribution of nutrients among plant foods using 2D and hoverable 3D principal component analyses (PCA) was visualized to guide selection of balanced diets. Literature search was conducted for glucosinolate (GSL) contents in commonly consumed foods for the period 1980-2020. The data were reanalyzed to summarize current knowledge and discuss challenges for development of a dietary GSL database for US foods. Results suggested that currently available US data are insufficient to develop a valid GSL database. A book chapter summarizing common dietary antioxidant measurements and discussing the controversies of dietary antioxidants was prepared along with 2 publications on glucosinilates.
Dietary Supplements: National study of calcium and vitamin D dietary supplements (DS) for adults. In the U.S., the current under-consumption of vitamin D and calcium is of public health concern. To evaluate if consumption of adult Ca DS may help to close the gaps in Ca and vitamin D intakes, we compared the nutrient labeled and measured content with the highest Recommended Dietary Allowances (RDA) and Tolerable Upper Intake Levels (UL) (for ages >70 years) in nationally representative DS. DS with calcium as the primary ingredient at = 80 mg/serving with or without vitamin D and other minerals were analyzed. The overall mean percentage differences from labels for individual DS were: for calcium, 7.9%, ranging from -5 to 28%; n=102; for vitamin D, 24.9%, ranging from -88% to 88%; n=83. For calcium, 30.4% of DS were labeled at or above its RDA but for vitamin D, zinc, manganese and copper, 53-67% of DS were labeled at or above their RDAs (but below the ULs). Manufacturers added vitamin D overages to amounts exceeding the vitamin D RDA or UL. Thus, Ca DS may provide significant amounts of calcium and vitamin D, as well as magnesium, phosphorus, zinc, manganese or copper in some DS. For more accurate calculation of total nutrient intake from foods and DS in a population, it is important to account for the nutrient amounts above or below labels.
Further Study on Dietary Supplements: Single- and multi-ingredient green tea (GT) DS studies: label assessment for precision and consumer safety. Epidemiological studies and DS consumers rely on product labels for information on DS content and safety and address “Does the Food and Drug Administration (FDA) required label format provide reliable estimates for intakes of phytochemical content for researchers and consumers?” and “Do highly concentrated DS provide sufficient instructions and warnings to consumers?”. Two lots of commonly sold single- (GT-1, n=32) and multi-ingredient (GT-2, n=36) GT DS were tested for their catechin and caffeine content. Epigallocatechin gallate (EGCG) is the primary catechin in GT leaves and extracts. Regression analyses indicated that, in the GT-1 DS study, increased weight of GT material (FDA required information) was not significantly associated with an increase in EGCG measured content. For both studies, labels that voluntarily listed an amount of EGCG were better predictors of measured EGCG content than label information for the weight of GT material. Thus, the FDA-required label format does not provide reliable estimates for catechin content in GT DS for intake calculation by researchers and for consumers seeking health benefits. In line with the current state of knowledge about GT DS safety and potential toxicity, most of the studied DS listed instructions and/or warnings about use (e.g., related to pregnancy, taking with food). However, some GT DS containing high amounts of catechins or caffeine did not display such warnings and instructions. Since 2019, manufacturers who choose to comply with United States Pharmacopeia recommendations must add ”Do not take on an empty stomach. Take with food” instructions and a liver toxicity warning to labels of DS with GT extracts, which may improve consumer safety further.
Iodine in Foods: 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. Thus, MAFCL, FDA, and ODS-NIH scientists are collaborating to provide and maintain iodine data for foods and dietary supplements. Our data for iodine content of 430 foods, released in 2020, were linked to food intakes from survey data (NHANES WWEIA 2013-14) for estimating iodine intake for the U.S. population, based on participants’ urinary iodine measurements. A national study of iodine variability in retail milk over a one-year period in twelve locations began, expanding upon MAFCL’s foundational study conducted in 2018. Samples of other foods were collected for laboratory analysis and inclusion in the database. These data support health policy and nutrition guidance for a vulnerable population.
Accomplishments
1. Ingredients in commercial packaged foods. Commercially packaged foods are an integral part of the U.S. diet and are important contributors to the food security and nutrient intake. There is a lack of information in the scientific literature on the ingredients used in packaged foods. USDA’s Global Branded Food Products Database makes publicly available a compiled dataset of ingredient lists for over 350,000 commercial food products. Researchers at Beltsville, Maryland, have developed a framework for parsing and systematic reporting ingredients used in commercially packaged foods which can help improve the understanding of the ingredients enhacing the traditional nutrient profiles using new tools.
2. Nutrient similarities and dietary diversification. One in three people in the world suffers from some form of malnutrition. A possible solution to malnutrition is dietary diversity, especially diversifying plant food consumption. However, there has not been a suitable method to guide the acquisition of balanced nutrients from diverse plant sources. Researchers at Beltsville, Maryland, have developed a tool for categorizing raw plant foods based on the similarity of nutrients. This is the first report of a nutrient-based food categorization system. Importantly, this process can be used to guide the development of other food categorization systems for nutrient diversification.
Review Publications
Ma, P., Li, A., Yu, N., Bahadur, R., Qin, W., Li, Y., Ahuja, J.K. 2021. Application of machine learning for predicting label nutrients using USDA Global Branded Food Products Database (BFPD). Computers and Electronics in Agriculture. https://doi.org/10.1016/j.jfca.2021.103857.
Ahuja, J.K., Li, Y., Bahadur, R., Nguyen, Q., Haile, E., Pehrsson, P.R. 2021. IngID: a framework for parsing and systematic reporting of ingredients used in commercially packaged foods. Journal of Food Composition and Analysis. 100. http://doi.org/10.1016/j.jfca.2021.103920.
Li, Y., Bahadur, R., Ahuja, J.K., Pehrsson, P.R., Harnly, J.M. 2021. Macro-and micronutrients in raw plant foods: the similarities and implication for dietary diversity. Nature Food. https://doi.org/10.1016/j.jfca.2021.103993.
Woodruff, R.C., Zhao, L., Ahuja, J.K., Gillespie, C., Goldman, J.D., Harris, D.M., Jackson, S., Moshfegh, A.J., Rhodes, D.G., Sebastian, R.S., Terry, A., Cogswell, M. 2020. Top Food Category Contributors to Sodium and Potassium Intake - United States, 2015-2016. Morbidity and Mortality Weekly Reports. 69(32):1064-1069. https://doi.org/10.15585/mmwr.mm6932a3.
Fukagawa, N.K., Pehrsson, P.R., Mckillop, K.A., Phillips, K.M., Mcginty, R.C., Couture, G. 2021. Dietary fiber, starch, and sugars in bananas at different ripeness in the retail market. PLoS ONE. https://doi.org/10.1371/journal.pone.0253366.
