Skip to Main content

Nutrient Composition of Food
Related terms:
Nutrient Database
Food Composition
Food and Nutrition Program
Metabolizable Energy
Nutrition Education
Human Nutrition
Extruded Food
View all Topics
Microbial Spoilage of Foods
Leonardo Petruzzi, ... Antonio Bevilacqua, in The Microbiological Quality of Food, 2017
1.3.2 Food Composition
The nutrient composition of food also influences the growth of the most suitable species of microorganisms. Protein foods such as meat, fish, and eggs are liable to be attacked by proteolytic organisms; "carbohydrate" foods such as bread, flour, pasta, syrups, and jams are more liable to attack by fermentative organisms; fats are liable to be attacked by lipolytic organisms (Modi, 2009).
Lipids can be degraded enzymatically to produce free fatty acids that have rancid and unpleasant off-aromas. Breakdown of lipids is important in meat, dairy, and olive oil systems (Howell, 2016).
Some foods also have inhibitory substances or naturally occurring antimicrobials present (Davidson and Critzer, 2012). Such substances can inhibit or slow the growth of some microorganisms (Kreyenschmidt and Ibald, 2012). For example, egg albumin contains the enzyme lysozyme, which disrupts the cell walls of Gram-positive bacteria by hydrolyzing the glycosidic bonds of N-acetylmuramic acid and N-acetylglucosamine in the peptidoglycan layer, avidin, which ties up the vitamin biotin, conalbumin, which ties up iron, and protease inhibitors, which inhibit protein degradation. All of these compounds act in concert to inhibit microbial growth. Raw milk also contains lysozyme and the lactoperoxidase system, which requires the interaction of the enzyme lactoperoxidase, thiocyanate, and hydrogen peroxide to produce the antimicrobial hypothiocyanate (Davidson and Critzer, 2012).
Lactoferrin is also present, which sequesters the iron necessary for microbial growth similar to conalbumin found in eggs (Davidson and Critzer, 2012).
In general, Gram-positive bacteria are sensitive to many molecules such as citrate, nisin, butylated hydroxyanisol or butylated hydroxytoluene, as well as molecules known for their antifungal activity, such as sorbates or benzoates. Gram-negative bacteria are more resistant than Gram-positive bacteria, but are still susceptible to a broad spectrum of additives as well as SO2 (Baron and Gautier, 2016).
There are some yeast species that are tolerant to ethanol at high concentrations, although their metabolism may be affected, and these include Zygosaccharomyces, Dekkera, Pichia, and Saccharomycodes. Many yeast species have evolved resistance to weak organic acids and include Zygosaccharomyces, along with some strains of Candida krusei and Pichia membranifaciens (Howell, 2016).
Foods with low sugar content, such as juices, jellies, and jams, contain acids, sweeteners, and hydrocolloids. These additives are needed for sensory purposes. Aspartame is among the most used sweeteners, though the use of the natural sweetener stevia is increasing in recent years. The presence of mentioned additives affects the growth of Zygosaccharomyces bailii (Campos et al., 2015).
Some yeasts are able to grow in media or substrates with high salt or sugar concentrations. For example, Debaryomyces hansenii is very salt tolerant, and some strains can tolerate up to 24% (w/v) NaCl. Saccharomyces cerevisiae also exhibits significant salt tolerance, but this is strain-dependent and is also influenced by pH of the food or beverage in question (Howell, 2016). Some species such as Hemimysis anomala and Candida pseudotropicalis may grow at NaCl up to 11% (Campos et al., 2015). Zygosaccharomyces rouxii is exceptionally tolerant to high sugar concentrations (up to 70% w/v sucrose) (Howell, 2016). Similarly, molds are tolerant to high concentrations of sugar or salt (Modi, 2009).
View chapterPurchase book
Nutrition and Food Access
Ruth MacDonald, Cheryll Reitmeier, in Understanding Food Systems, 2017
7.1.1 Nutrient Composition of Foods
The study of nutrient composition of foods in the United States began in the mid-1800s with S.W. Johnson who had studied under Liebig and became the first professor of biochemistry at Yale University. Johnson's student, Wilbur O. Atwater, received the first appropriation from Congress to conduct research in human nutrition. Atwater eventually led the Agricultural Experiment Stations within the USDA and was intrigued with the concept of the energy value of foods. One of Atwater's accomplishments was the construction of a respiration calorimeter similar to that developed in Germany. His work led to the understanding that dietary fats and carbohydrates, not only proteins, could be used for mechanical work by the body. The Atwater factors of 4 kcal/g of protein or carbohydrate, 7 kcal/g of alcohol, and 9 kcal/g of fat are still used today to estimate the amount of energy generated from foods. In 1896, Atwater created the first proximate analysis database (nitrogen, fiber, ash, ether extract, moisture, and carbohydrates) of foods to teach poor people how to obtain their protein requirements at the lowest cost. The USDA publication from his work was the first to define food within five categories of macronutrients: protein, carbohydrate, fat, energy, and water.
By 1950, the USDA had compiled a substantial database on the composition of foods and produced Agricultural Handbook No. 8, Composition of Foods—Raw, Processed, Prepared. This document listed the proximate analysis (protein, carbohydrate, fat, and water), five vitamins (vitamin A, thiamine, riboflavin, niacin, and vitamin C), and three minerals (calcium, phosphorus, iron) of 750 foods. In 1963, a revision of Handbook No. 8 with added data for cholesterol, fatty acids, sodium, potassium, and magnesium, was published. Analytical methods to measure nutrients in foods continued to be developed during this time, and the demand for information about food composition was high as the role of food in health and disease was being recognized. An electronic version of Handbook No. 8, now called the National Nutrient Database for Standard Reference (NNDSR), was released by the USDA in 1980. The need for more data, especially for quantification of the wide range of bioactive compounds of foods, continued to grow through the 1990s. Bioactive compounds are naturally occurring chemicals in foods that may have health benefits, but are not considered essential nutrients. These bioactive compounds include antioxidants found in plants such as the flavonoid family of anthocyanidins, isoflavones, and catechins. In 1997, the USDA created the National Nutrient and Food Analysis Program to address the complexity of food composition and to coordinate and ensure the accuracy of data being collected from government labs, the food industry, and academic research. The USDA Nutrient Data Laboratory is responsible for overseeing the publicly available food composition database for the United States, which is freely accessible through their website: www.ndb.nal.usda.gov.
View chapterPurchase book
Sources of the Vitamins
Gerald F. CombsJr, in The Vitamins (Fourth Edition), 2012
Vitamin Content Data
Collation of best estimates of the nutrient composition of foods and feedstuffs has been an ongoing activity by several groups in the United States since the turn of the century.1 Nutrient composition data for foods and feedstuffs are now available in many forms (e.g., books, wall charts, tables, and appendices of books, computer tapes, and diskettes). However, most compilations derive from relatively few sources. This is particularly true for the nutrient composition data for foods. For US foods, almost all current versions are renditions of the USDA National Nutrient Database,2 developed through an ongoing program of the US Department of Agriculture. Similar, but less extensive, databases have been developed for other countries,3 and efforts are being made to standardize the collection, compilation, and reporting of food nutrient composition data on a worldwide basis.4
The nutrient composition of feedstuffs has, with few exceptions,5 been developed less systematically and extensively. Data sets presently in the public domain have been compiled largely from original reports in the scientific literature. Therefore, effects of uncontrolled sampling, multiple and often old analytical methods, multiple analysts, unreported analytical precision, unreported sample variance, etc., are likely to be far greater for the nutrient databases for feedstuffs than for the corresponding databases for foods.
Use of any database for estimating vitamin intake is limited for reasons concerning the accuracy and completeness of the data. Although the United States Department of Agriculture (USDA) National Nutrient Database (Table 19.1) is much more complete than most feed tables with respect to data for vitamins, it is only reasonably complete with respect to thiamin, riboflavin, and niacin; data for vitamins D and E are relatively sparse. In addition, accurate and robust6 analytical methods are available only for vitamin E, thiamin, riboflavin, niacin, and pyridoxine. For the other vitamins, this means that the quality of available analytical data can render them unacceptable for inclusion in the database, or be low enough to raise serious questions concerning reliability.
Table 19.1. Adequacy of the USDA National Nutrient Database Vitamin Contents of Foods
FoodADEKCThiaminRiboflavinNiacinB6Pantothenic AcidFolateB12Baby foods■•□■■■■•□••Baked foods: Breads□•□■■■■••□ Sweet goods••■■■•••• Cookies/crackers••■■■■•••Beverages••••••••Breakfast cereals•□•□••••••••Candies•□•□•••••□Cereal grains: Whole•••■■■■■■ Flour•□■■■■■■ Pasta□□■■■■■■Dairy products■■••■■■■■■•■Eggs, egg products••••□■■■■■•■Fast foods■□•□•■■■••□•Fats and oils••••Fish, shellfish: Raw•••□••••••• Cooked□••□□□□□••□Fruits: Raw■□•■■■■■•• Cooked•□□••••••• Frozen, canned■□□•■□■•••Legumes: Raw•■•■■■•■■ Cooked□■□■■■•■■Meat: Beef■•••■■■■••■ Lamb•••□■■■•••■ Pork•••□■■■■••■ Sausage••□□■■■■■••■ Veal•••□■■■■•■■ Poultry•••□■■■••••Nuts, seeds•••□••••••□Snack foods•••□□•■■•••□Soups■□□■■■■•□□•Vegetables: Raw■••■■■■•□• Cooked••□•••••□• Frozen■□□■■■■□•• Canned■□□■■■■□••
Key: □, few or no data; •, inadequate data; ■, substantial data.
Source: Beecher, G. and Matthews, R. (1990). Nutrient composition of foods. In Present Knowledge in Nutrition (Brown, M., ed.), 6th edn. International Life Science Institute – Nutritional Foundation, Washington, DC, pp. 430–439.
View chapterPurchase book
Sources of the Vitamins
Gerald F. CombsJr. Ph.D., James P. McClung Ph.D., in The Vitamins (Fifth Edition), 2017
Vitamin Content Data
Collation of best estimates of the nutrient composition of foods and feedstuffs has been an ongoing activity by several groups in the United States since the early 1900's.2 Nutrient composition data for foods and feedstuffs are now available in many forms and from many sources. However, most compilations derive from relatively few primary sources. This is particularly true for the nutrient composition data for foods. For US foods, almost all current versions are renditions of the USDA National Nutrient Database3 developed through an ongoing program of the U.S. Department of Agriculture. A similar database, the Canadian Nutrient File has been developed by that country.4 Less extensive databases have been developed for other countries,5 and efforts are being made to standardize the collection, compilation, and reporting of food nutrient composition data on a global basis.6 Variances are to be expected in national food composition databases due to differences in analytical methodologies (a particular problem for folate) and data presentation, e.g., whether tocotrienols are included in the calculation of α-tocopherol equivalents.7
The nutrient composition of feedstuffs has, with few exceptions,8 been developed less systematically and extensively. Data sets presently in the public domain have been compiled largely from original reports in the scientific literature. Therefore, effects of uncontrolled sampling, multiple and often old analytical methods, multiple analysts, unreported analytical precision, unreported sample variance, etc., are likely to be far greater for the nutrient databases for feedstuffs than for the corresponding databases for foods.
Use of any database for estimating vitamin intake is limited for reasons of accuracy and completeness of the data. Although the USDA National Nutrient Database is much more complete than most feed tables with respect to data for vitamins, it is least complete with respect to vitamins D, E, and K, as well as pantothenic acid.
View chapterPurchase book
Dietary intake measurement: Methodology
A.A. Welch, in Encyclopedia of Human Nutrition (Third Edition), 2013
Quantitative Analysis
The data collected by dietary methods are converted into food and nutrient consumption by calculating the amount of food eaten and linking this to a database with values for the nutrient composition of foods.
The databases of nutrient composition of foods are provided by the governments of many countries. They consist of nutrient composition data for the average composition of commonly consumed foodstuffs and are usually available as printed publications, computerized databases, or as part of software packages. Values in nutrient composition databases are expressed as either per 100 g of food or per common household measure. Nutrient databases vary in the coverage and comprehensiveness of the foods and nutrients. They are revised periodically to cover newer foods of different nutrient compositions or to modify or extend the nutrient coverage. Some issues concerning the choice of nutrient databases are shown in Table 4. It is important to read the information distributed with the printed or electronic versions of databases to determine the uses and limitations of the data.
Table 4. Factors to consider when choosing a nutrient database to calculate nutrient intakes
Comprehensiveness of food item and beverage coverage?Does the database contain entries for important foods consumed by the population to be studied?How comprehensive is the coverage of nutrients?Does the database contain data for mixed or multiple ingredient recipes or dishes?What analytical techniques were used to derive nutrients in the database? (There can be differences in nutrients measured by different techniques.)Are the data officially evaluated?What compilation methods were used to construct the database?Which conversion factors are used to calculate metabolizable energy content of foods for protein, fat, carbohydrate, and alcohol?What proportion of missing values exists within the database? (Missing values are counted as zero in calculations and so result in systematic underestimates of intake.)For international studies or comparisons how do the analytical methods for determining nutrient composition and compilation techniques affect the resulting data?
Several steps are involved in calculating nutrient intake (also known as coding or processing). The first is to choose an item in the database, which corresponds most closely with the food consumed. If the food consumed is not in the database a suitable alternative can be chosen by considering food type, general characteristics, and likely nutrient profile. Once the food has been chosen the nutrient composition of the food quoted in the database is multiplied by the amount of food eaten, e.g., for 60 g food the nutrients would be multiplied by 0.6 (where nutrients are expressed per 100 g of food).
To calculate daily intake for an individual the contribution of each food is calculated and all the foods for a day summated. If more than one day's data have been collected it is usual to calculate the average of the number of days recorded. Data from FFQs are usually computed to consumption per day but can also be computed per week.
Although it is possible to compute intake by hand, using a calculator and a printed copy of a nutrient database, this is very labor intensive and in practice for most purposes has been superseded by computerization.
View chapterPurchase book
Taste disorders in disease
Laurent Brondel, ... Luc Pénicaud, in Flavor, 2016
15.3.3 Taste disorders in metabolic syndrome
Several different dietary patterns, including Mediterranean-style diets (Jones et al., 2012), low-fat diets (Neuhouser et al., 2012), and vegetarian and vegan diets (Rizzo et al., 2011) have been shown to be protective against metabolic syndrome. However, different tasting profiles may moderate a person's ability to adhere to a healthy diet. The tasting profile, such as sweet likers (SL) or supertasters (ST), may be influenced by genetics, and may influence dietary intake. ST are thought to have heightened taste perceptions, and are able to discriminate small changes in the nutrient composition of foods, more so than non-ST (Hayes and Keast, 2011).
Results concerning ST and BMI are contradictory; some show that non-ST have a higher BMI (Tepper et al., 2008), others show no differences in BMI between ST and non-ST (Drewnowski et al., 2007; Yackinous and Guinard, 2002), and no effect of the ST genotype on BMI or dietary intake (Hayes, 2010). Concerning SL, a preference for sweet solutions has been associated with increased preference for sugar-sweetened desserts (Drewnowski et al., 1999), but other studies have reported contradictory results. Turner-McGrievy et al. (2013) tested around 200 patients, and investigated the relationships and interactions between both SL and ST status and metabolic syndrome. This study found that the pattern of dietary intake was consistent with the pattern of metabolic syndrome occurrence, and revealed lower intakes of fiber and higher intakes of caloric beverages among individuals with a combined SL (high consumption of added sugars) + ST (avoidance of vegetables) status, followed by those with a non-SL + non-ST status, as compared to ST or SL only. Despite several limitations in this study, the authors concluded that counseling individuals who are both SL and ST or are non-SL and non-ST on ways to increase fiber and decrease caloric beverage intake may be needed to modify metabolic syndrome risk.
Future direction: Assessing genetic differences in taster profiles may be useful in the development of specific dietary interventions to prevent and treat metabolic diseases. Better understanding of the role of taste thresholds in overweight and obesity might help us manage these diseases. Better understanding of the physiology of modified eating behavior after RYGB is a current objective.
View chapterPurchase book
Dietary Assessment Methodology
FRANCES E. THOMPSON, AMY F. SUBAR, in Nutrition in the Prevention and Treatment of Disease, 2001
D. Choice of Nutrient Database
It is necessary to use a nutrient composition database when dietary data are to be converted to nutrient intake data. Typically, such a database includes the description of the food, a food code, and the nutrient composition per 100 g of the food. The number of foods and nutrients included varies with the database.
Some values in nutrient databases are obtained from laboratory analysis; however, because of the high cost of laboratory analyses, many values are estimated based on conversion factors or other knowledge about the food [302]. In addition, accepted analytical methods are not yet available for some nutrients of interest [303], the analytical quality of the information varies with nutrient [223, 303], and the variances or ranges of nutrient composition of individual foods are in most cases unknown [304]. Rapid growth in the food processing sector and the global nature of the food supply add further challenges to estimating the mean and variability in the nutrient composition of foods.
One of the U.S. Department of Agriculture's (USDA) primary missions is to provide nutrient composition data for foods in the U.S. food supply, accounting for various types of preparation [305]. USDA produces and maintains the Nutrient Database for Standard Reference, which includes information on up to 82 nutrients for 6200 foods. Interest in nutrients and food components potentially associated with diseases has led to development of databases for limited numbers of foods. These include databases for isoflavones, carotenoids, selected trans-fatty acids, sugar, and vitamin K. Values for selenium and vitamin D have been incorporated into the current Release 13 of the Nutrient Database for Standard Reference. Information regarding all of USDA's nutrient databases is available at the USDA's Nutrient Data Laboratory home page: http://www.nal.usda.gov/fnic/foodcomp. The International Network of Food Data Systems (INFOODS) [306] maintains an international directory of nutrient composition data, which is available at http://www.fao.org/infoods.
Research on nutrients (or other dietary constituents) and foods to improve current estimates is ongoing, and there is constant interest in updating current values and providing new values for a variety of dietary constituents of interest. Methods for converting dietary intake information into number of servings recommended in the U.S. Food Guide Pyramid [307] have been developed using the Minnesota Nutrition Data System [308] and the U.S. Department of Agriculture pyramid servings database [309]. One limitation in all nutrient databases, particularly for fatty acids, is the variability in the nutrient content of foods within a food category [310, 311]. For example, one study found that among 17 brands of crackers, trans-fatty acid content varied from 1 to 13 g per 100 g of crackers [23]. Depending on the level of detail queried on the dietary assessment instrument, the respondent's knowledge of specific brand names, and the specificity of a particular nutrient database, estimating accurate fatty acid intake can therefore be problematic. For FFQs, this problem is compounded by the collapsing of foods into categories (such as crackers) that might have highly variable nutrient content.
Many other databases are available in the United States for use in analyzing records, recalls, or FFQs, but most are based fundamentally on the USDA database, often with added foods and specific brand names. Estimates of nutrient intake from dietary recalls and records are often affected by the nutrient composition database that is used to process the data [312–314]. Differences are due to the number of food items in the database, the recency of nutrient data, and the number of missing or imputed nutrient composition values. Therefore, before choosing a nutrient composition database, a prime factor to consider is the completeness and accuracy of the data for the nutrients of interest. For some purposes, it may be useful to choose a database in which each nutrient value for each food also contains a code for the quality of the data: e.g., analytical value, calculated value, imputed value, or missing. Investigators need to be aware that a value of zero is assigned to missing values in some databases. The nutrient database should also include weight/volume equivalency information for each food item. Many foods are reported in volumetric measures (e.g., 1 cup) and must be converted to weight in grams. The number of common mixtures (e.g., spaghetti with sauce) available in the database is another important factor. If the study requires precision of nutrient estimates, then procedures for calculating the nutrients in various mixtures must be developed and incorporated into nutrient composition calculations. Another key consideration is how the database is maintained and supported.
Research or guidance on how to best compile a nutrient database for a FFQ or diet history is limited [56, 75, 315]. However, it is clear that an approach that is primarily data driven, using national or other carefully collected dietary data connected to a trustworthy nutrient database, is critical.
View chapterPurchase book
Fibre in extruded products
N. Sozer, K. Poutanen, in Fibre-Rich and Wholegrain Foods, 2013
12.1 Introduction
There is a remarkable demand worldwide to design foods with high nutritional profile. Research evidence on the health benefits of dietary fibre (DF) has brought with it the need to increase DF content in food products. The challenge is to combine high fibre and wholegrain content with low saturated fat, sugar and salt contents and nevertheless produce acceptable sensory appeal. Extruded food products are one of the cereal product categories in which an increased fibre level would offer a nutritional benefit. According to the Nutrition Labeling and Education Act of 1990 USA, extruded snacks can be labelled as 'good source' or 'high fibre source' if the DF amount per serving is at least 2.5 or 5 g, respectively (Eastman et al., 2001). Implementing the data from Healthy Eating Index 2005 and MyPyramid, the US developed a classification system based on the nutrient composition of foods (http://www.health.gov/DietaryGuidelines/dga2005/document/). In the Nutrition and Health Claims directive of the European Union (EU), foods can be claimed to be a source of fibre at the level of 3 g DF/100 g and a good source at 6 g/100 g, provided they fulfil the other nutritional profile requirements (Annex I, EC No 1924/2006).
Extrusion technology is one of the processing techniques to design expanded food matrices. Food extrusion applications can be found as early as the 1930s for processing of cereals, snack foods and texturized foods (Ilo et al., 2000). Extrusion technology enables formulation and design of a large variety of food products. It can be used either to form ready-to-eat foods (snacks, cereals, pasta, confectionery) or to modify food ingredients which can form solid, semi-solid and liquid products after preparation steps for consumption. Structure–texture properties of the expanded food products strictly depend on the operational parameters.
The structural elements of an extruded food product are formed via the physicochemical changes occurring in the raw materials. It is known that extrusion processing results in irreversible changes in starch granules and polymers, denaturation of proteins and formation of starch–lipid, protein–lipid and protein–protein complexes (Lai and Kokini, 1991; Cremer and Kaletunç, 2003, Hagenimana et al., 2007). Cereal extrudates consist of a continuous starch matrix and a discontinuous protein phase (Hermansson, 1988). It has been suggested that structural segregation within the extruded material occurs due to formation of protein fibrils (Noguchi, 1989). It is a food engineering challenge to design palatable extruded foods containing a large amount of DF, since polymer matrices with high levels of DF have low expansion capability. The majority of high DF products exhibit poor textural (high hardness and low crispiness) and morphological (small pores, high density) properties. In this chapter the aim is to review the effects of DF and outer grain parts on extrusion processing of expanded foods, and summarize the ongoing research into development of high DF foods with enhanced palatability.
View chapterPurchase book
Dietary Assessment Methodology
Frances E. Thompson, Amy F. Subar, in Nutrition in the Prevention and Treatment of Disease (Third Edition), 2013
E Choice of Nutrient and Food Database
It is necessary to use a nutrient composition database when dietary data are to be converted to nutrient intake data. Typically, such a database includes the description of the food, a food code, and the nutrient composition per 100 grams of the food. The number of foods and nutrients included varies with the database. Research on nutrients, other dietary components, and foods is ongoing, and there is constant interest in updating current values and providing new values for a variety of dietary components of interest.
Some values in nutrient databases are obtained from laboratory analysis; however, because of the high cost of laboratory analyses, many values are estimated based on conversion factors or other knowledge about the food [538]. In addition, accepted analytical methods are not yet available for some nutrients of interest [539], analytical quality of the information varies with the nutrient [539,540], and the variances or ranges of nutrient composition of individual foods are in most cases unknown but are known to be large for some nutrients [541]. Rapid growth in the food processing sector and the global nature of the food supply add further challenges to estimating the mean and variability in the nutrient composition of foods eaten in a specific locale.
One of the USDA's primary missions is to provide nutrient composition data for foods in the U.S. food supply, accounting for various types of preparation [542]. Information about the USDA's nutrient composition databases is available at the USDA's Nutrient Data Laboratory home page [543]. The USDA produces and maintains the Nutrient Database for Standard Reference (SR). New releases are issued yearly; these include information on new foods and revised information on already included foods, and they identify foods deleted from the previous version of the database. The most recent release, SR24, includes information on up to 146 food components for more than 7900 foods [544], and it is available online.
Interest in nutrients and food components potentially associated with diseases has led the USDA to develop specialized databases for a smaller number of foods, such as flavonoids, proanthocyanidins, choline, and fluoride [545]. A separate database developed by the USDA Food Surveys Research Group—the Food and Nutrient Database for Dietary Studies (FNDDS)—is used by many investigators in analyses of foods reported in NHANES' What We Eat in America dietary recalls and is based on nutrient values in the USDA SR database [89].
Nutrient composition data are also compiled by a number of other countries, and the International Network of Food Data Systems maintains an international directory of nutrient composition tables [546]. Combining different food composition databases across countries poses comparability challenges, however. The European Food Information Resource [547] was formed to support the harmonization of food composition data among the European nations. The International Nutrient Databank Directory, an online compendium developed by the National Nutrient Databank Conference, provides information about the data included in a variety of databases, national reference databases, and specialized databases developed for software applications, such as the date the database was most recently updated, the number of nutrients provided for each food, and the completeness of the nutrient data for all foods listed [548].
In addition to nutrient databases, databases that can relate dietary intake to dietary guidance have been developed in the United States [549,550]. The USDA Food Patterns provide quantities of foods to consume from specific food groups in order to attain a diet consistent with the guidelines at a variety of calorie levels [551]. Just as FNDDS provides a nutrient profile for each food, the Food Patterns Equivalents Database (FPED) provides food group data for each food in FNDDS in order to allow assessment of the intake in terms of these Food Patterns. The FPED contains 32 food group components (e.g., dairy, fruits, and vegetables) and provides the amount of each food group per 100 grams of each food [552].
Other databases are available in the United States for use in analyzing dietary records and 24-hour recalls, but most are based fundamentally on the USDA SR database, often with added foods and specific brand names. One prominent such database is the University of Minnesota's Nutrition Coordinating Center's (NCC) Food and Nutrient Database [553]. This database includes information on 162 nutrients, nutrient ratios, and other food components for more than 18,000 foods, including 8000 brand-name products. The NCC is constantly updating its database to reflect values in the latest release of the USDA SR.
One limitation in all nutrient databases is the variability in the nutrient content of foods within a food category and the volatility of nutrient composition in manufactured foods. Recent changes in the sodium and fatty acid composition of manufactured foods, for example, illustrate the difficulty in maintaining accurate nutrient composition databases [554]. Obviously, a key consideration is how the database is maintained and supported.
Estimates of nutrient intake from 24-hour recalls and dietary records are often affected by the nutrient composition database that is used to process the data [555–557]. Inherent differences in the database used for analysis include factors such as the number of food items included in the database, how recently nutrient data were updated, and the number of missing or imputed nutrient composition values. Therefore, before choosing a nutrient composition database, a prime factor to consider is the completeness and accuracy of the data for the nutrients of interest. For some purposes, it may be useful to choose a database in which each nutrient value for each food also contains a code for the quality of the data (e.g., analytical value, calculated value, imputed value, or missing). Investigators need to be aware that a value of zero is assigned to missing values in some databases, whereas for other databases, the number of nutrients provided for each food may fluctuate depending on whether or not a value is missing, and for others all unknown values may be imputed.
The nutrient database should also include weight/volume equivalency information for each food item. Many foods are reported in volumetric measures (e.g., 1 cup) and must be converted to weight in grams in order to apply nutrient values. The number of common mixtures (e.g., spaghetti with sauce) available in the database is another important factor. If the study requires the precision of nutrient estimates, then procedures for calculating the nutrients in various mixtures must be developed and incorporated into nutrient composition calculations.
Developing a nutrient database for an FFQ presents additional challenges [558] because each item on the FFQ represents a food grouping rather than an individual food item. Various approaches that rely on 24-hour recall data, either from a national population sample or from a sample similar to the target population, have been used [114,138,559]. Generally, individual foods reported on 24-hour recalls are grouped into FFQ food groupings, and a composite nutrient profile for each food grouping is estimated based on the individual foods' relative consumption in the population. For this approach to be effective, the 24-hour recall data must be representative of the population for whom the FFQ is designed and connected to a trustworthy nutrient database.
View chapterPurchase book
Recommended publications:
Journal of Food Composition and Analysis
Journal
Appetite
Journal
Journal of Nutrition Education and Behavior
Journal
Nutrition in the Prevention and Treatment of Disease (Fourth Edition)
Book • 2017
Browse Journals & Books

About ScienceDirect
Remote access
Shopping cart
Advertise
Contact and support
Terms and conditions
Privacy policy
We use cookies to help provide and enhance our service and tailor content and ads. By continuing you agree to the use of cookies.