RESEARCH
Review Food Composition D ata: The Foundation of Dietetic Practice and Research
JEAN A. T, PENNINGTON, PhD, RD; PHYLLIS J. STUMBO, PhD, RD; SUZANNE P, MURPHY, PhD, RD; SUZANNE W, McNUTT, MS, RD; ALISON L, ELDRIDGE, PhD, RD; BEVERLY J. McCABE-SELLERS, PhD, RD; CATHERINE A. CHENARD, MS, RD
benefit from becorning more informed about food compo- ABSTRACT sition data, exploring new ways to educate themselves Food composition databases and dietary assessnent sys- about databases and database systems, and advocating tems are impo::tant tools for food and nutrition profes- for what is most needed in dietetic practice. sionals. The availability and accesslbility of data have J Ant Diet Assoc. 2047;fi7:2105-2LL3. improved over time along with the technoiogy to convert the information into useful formats for planning diets, writing educational materials, counseling patients, and ood composition databases are used by food and nu- conducting research. Prirnary sources of food composition trition professionals to assess the d.ietary status of data include government, academic, and other iastitu- patients, ilients, and consumers and to plan and tional databaJes; the food industry; and scientific litera- evaluate the dietary adequacy of meals and diets. Accu- ture. Changes in the rnarketplace affect food availability rate assessment ofthe intake ofnutrients and other food and composition and affect the aciuracy and adequacy of components depends on the accuracy of the databases. food composition databases. Improvements in both food The databases are never complete due to the dynamic composition data and in dietary assessment metirods nature of the food supply; for example, foods are con- have worked synergistically to irnprove .estimates of di- stantly introduced to and removed from the market, and etary intake. The development of databases for food fre- also because of newly recognized food components that quency assessment systems requires special consider- are identified as being associated with health and/or dis- ations for data aggregation for each food or food grouping ease, The purpose ofthis review is to provide information in the questionnaires. Considerations for selecting ? gi- and guidance to food and. nutrition professionals with etary aisessment system itrciude appropriateness of the reepect to primary sources of food composition data and data for the intended audience or purpose, efficiency of their use, advances in food composition data and data- the search strategy for letrieving data, content and for- bases, the effects of the changing marketplace on data mat of summary infoi:nation, and cost, Needs for food and databases, the synerry between food composition composition data vary depending on dietetic practice data and dietary assessment, the incorporation of data atea; howe.rer, most food and nutrition professionals will into dietary assessment systems, and the selection of a dietary assessment system. An overriding theme is meet- ing food and nutrition professionals'need for food compo- J. L T. Pennington is a researclt' nutritionist' Dlvision sition data. of Research Coordination, National Insti-tutes of Health, Bethesd.a, MD. P. J. Sturnbo and C. A' Ch'enard are re- search dietitians, General Clinical Research Center, PRIMARY SOURCES OF FOOD COMPOSITION DATA (Jniuersity of lowa, Iowa City. S. P' Murphy is a nutri' The primary source of food composition data for the tion researcher, Cancer Research' Center, University of United States is the Nutrient Database for Standard Hawoii, Honolulu. S. W' McNutt is a senior research Reference (NDSR) (1), which is maintained by the Agri- nutritionist, Westat Inc, Salt Lake City, UT' A' L. El' culturai Research Serr.ice (ARS) of the US Department of dridge is a manager, Nutrition Scieneg Diuision, Gen' Agliculture ruSDA) and freely available on the Internet. erat MiIIs BelI Institute of Health and Nutrition, Minne- NDSR, Release 19, contains ?,293 foods and up to 140 apolis, MN. B. J. McCabe'Sellers is a research. nutrients for each food; the data are derived from ARS- ci o r d.inat or, lJ S D ep artrn e nt of $g r i cultur e Ag ricultur al contracted food analysis, the food industry, and the sci- Research Seruice Lower Mississippi Delta Nutrition In- entific literature as wel] as some values that are esti- teruention Research Initiatiue, Little Rock, AR. mated or ca.lculated based on recipe ingredients or Address correspond.ence to: Jean A' T' Pennington, similar foods. There are misbing values in the database PhD, RD, Research Nutritioizist, Diuision of Nutrition because some foods have not been analyzed for all the Rese:arch C oor dina.tio n, N ational Institute s of He altlt, food components and because some missing values have 6707 Democracy Blud, Roont 629, Bethesda MD 20892' not been estimated. One feature of the on-line NDSR is 546 1. E-mail: JP 157 [email protected] the ability to aciess food lists based on their content of a Copyright @ 2007 by thq American Dietetic food component per 100 g or Daeasures provided by ARS. Association. Another ARS database is the Food and Nutrient Data- 0002-8223 I 07 I 107 1,2-0007 $32. 00 / 0 base for Dietary Studies (FNDDS) (2), which is used to doi: 1 0. 1 0 1 6 lj j ada'2007. 09' 004 assess dietary intakes for the National Health and Nu-
@ 2007 by the Americon Dietetic Association Journal of the AMERICAN DIETFIIC ASS0CIATI0N 2105 Feature NDSR (1) FNDDS (2)
Type of database-- - . Reference database Survey database piijor. Natjonal US food composition database; standard DBveloped for use with the National Health and Ior food composition information in the United Nutrition Examination Survey States primary uses Dietetic reference, patient care, studenVconsumer Assessment of food component intakes for US education; product development by industry; nutrition surveys and studies used by some other countries to create national databases No. of foods 7,293 -7,000 No. o{ food components 140 62 Basis ofvalues' Per 100 g and usually several serving potlions Linked file for serving size portions; -30,000 serving pottions total Missing values? Yes No; values missing from the NDSR lor the 62 food components are estimated Source of data Laboratory analyses, food companies, literature, NDSR, estimation estimation Data provided Mean/median varues "$yJHlJ'l,Lii?l;:::li,::.1',,ouffiJ'JiJ:.i.o median values for all others Special items Provides documentation about data sources and o lncludes default food items for use when suruey derivation of non-analytical values .cannot be specific about foods :rrrfr.#ents o Contains recipe items calculated from NDSR foods . iri i roiiiuie unJ trt adjustments file and a . nutrient retention file that may be used to calculate the composition ol recipe items
Figure 1, Features of the US Department of Agriculture Nutrient Database lor Standard Reference (NDSR) and Food and Nutrient Database {or Dietary Studies (FNDDS). tr.ition Examination Survey (NIIANES) (3)' The FNDDS components that are required to be on the label. Proce- includes the foods reporbed as being consumed by NIIANES dures for attempting to determine the composition o{ participants and contains 62 food components for approx-- products based on the label data and the ingredient list- imateiy 7,000 foods. The data in FNDDS are derived ing are provided by Rand and colieagues (12). primarily from the NDSR, and all missing values are Journal articles are timeiy sources of data gmerging imputed so that the database is corhplete for making from anall'bicai laboratories and can be accessed througb diitary intake assessments. The FNDDS is also freely literature searches. One primary source is the Journal o1 available on the Internet. Figure 1 summarizes the fea- Food Composition snd Analysis (13). Literature data on tures of the NDSR and FNDDS. the levels of bioactive food components (eg, carotenoids. A dietary supplement database was developed by the flavonoids, and polyphenols) and the antioxidant poten' National Center for Health Statistics for use with the tial of foods are of increasing interest because of their NIIANES dietary supplement intake data (4,5)' The Of- association with health and disease prevention (14,15). fice of Dietary Supplements at the National Institutes of Many food composition databases have been developec Health is working on a label-based dietary supplement in academic, clinical, and private institutions using tht database and is collaborating with ARS to develop an primary sources of data mentioned above. Some of thesr analysis-based dietary supplement database (6-8). An- are listed in Lhe International Nutrient Databaitk Direc othei government database is that for the Food and Drug tory (16), which is maintained by the National Nutrien Administration's Total Diet study (9)' This database Databank Conference. The 2006 edition of this directorl Iists about 280 core foods of the US food supply and lists 33 organizations that provide databases and/or soft contains annual analytical data for the levels ofdietary ware systems for dietary analysis, recipe calculations minerals, folic acid, heavy metal,s, radionuclides, pes- and./or other functions. The directory provides detaile< ticide residues, industrial chemicals, and other chemi- information about the nurnber of foods, food components cal contaminants (9-11). data sources, sofbware features, avaiiability, and cost o Information on the composition of packaged foods is each item. available from the Nutrition Facts and Supplement Facts The International Food Cornposition Tables Directoq labels and from food companies, trade associations, and (17) maintained by the United Nations International Net their Web sites. The Nutrition Facts label data have been work of Food Data Systems is an extensive archival col roundeil and adjusted to be in compliance with govern- lection fiat lists the names of hardcopy -and electroni ment regulations and are usually lir-aited to the food databases from around the world. Food and nutritior
2106 December 2007 Volume 1 07 Number 12 professionals may find this directory useful for iocating more than one t11pe of food (eg, a truffle is a chocoLate data on ethnic and inported foods that are not present in candy and a mushroom), and food name synonlms (eg, US databases. broccoli raab is a synonym for rapini) may make it diffl- cult to find some foods. The quality of food composition data is usually deter- ADVANCES IN FOOD COMPOSITION DATA mined from the sampling design, the number of samples Information on nutrient aud non-nutrient components of analyzed, the analltical methods used, and the reproduc- foods has been rapidly expanding. The NDSR database ibility of the data, However, because this type of informa- increases in size with eacir new release, and the FNDDS tion is usually available only with original data (ie, in is routinely updated for consistency with foods reported scientific papers and reports) and not carried over to to be consumed in NHANES. The USDA National Food compiled databases, the task of ensuring data quality and Nutrient Analysis Program provides funds to update usually falls on the database compiler, The quality of food aad expand the NDSR and help assure the quality of composition data has been addressed in great detail in analytical data through use of standard reference nrate- Greenfie1d and Southgate (36). rials and sound sampling arrd statistical handling of data Food composition variability is a reflection of inheient (13-20). As information on food componentb that may be (eg, genetic), environmental (eg, climate and tempera- related to health outcomes becomes availabie in the sci- ture), processing (eg, cooking and preservation methods), entific iiterature, the data may be incorporated into the and analltical factors (eg, sampling designs and analyt- NDSR and into academic and other institutional food ical methods). Some food components are more variable composition databases. Examples of such components in- than others becauie they may be susceptible to loss by (eg, clude flavonoids (21), iignans (22), giycemic index (23,24), heat, light, and processing methods cereal grain re- fluoride (25), tlnamine (26), biotin (27), hqterocyclic f,nement). Food component variability should be consid- amines (28), and glucosinolates (29). ered when making practical use of the data f,or patients New forms and units for some food components have and clients. Food and nutrition professionals should not been added to the NDSR and other databases that offer recommend one food over another as a better or lower data users choices of ways to express the levels of food source of a food component unless the differences be- components. For dxample, the Institute of Medicine pro' tween the foods will be of practical importance for the posed several new ways of expressing nutrient activity for patient or client, the Dietary Reference Intakes, such as folate as Dietary For national, regibnal, or local dietary intake surveys Folate Equivalents, vitamin A as Retinoi Activity Equiv- in the United States, the FNDDS or similar survey data- alents, and vitamin E as a-tocopherol only (30,31). base is appropriate. Survey databases should include The MyPpamid Web site (32) allows for the assess- foods commonly consumed by the target'population and rnent ofdietary intakes based on food groups; that is, the have no missing values. For a reference in the office or amount of fruits, vegetables, and grain products con- classroom, the NDSR or a similar database derived from sumed may be compared to MyPyramid recommenda- the NDSR is a good choice and should be used with the tions. This type of food group intake measure may also be understanding that there are missing values for some incorporated into analyses relating intakes to disease food components. The NDSR or similar database is also outc&nes such as cancer and heart disease. Food compo- appropriate for patient care such as dietary management sition databases may be used to assign food group serv- for weight loss or'disease. For patients who are following ings to each food item, in much the same way that nutri- therapeutic diets like severe sodium restriction, food and ents are assigned to foods; for example, the number of nutrition professionais will need databases with specific servings of vegetables per 100 g of the food (33). Some brand name i.nformation. For research or clinical studies food composition databases now have the ability to accu- to test the eflects of food components on biomarkers, mulate such food group intake variables (34), researchers should use either an isolated form ofthe food The increasing availability of data on the composition component, or, if a food is to be used, it should be pur- of dietary supplements is essential for accurate estimates chased in bulk and a representative sample should be of total food component intakes. The use of dietary sup- analyzed for the food component(s) of concern to the plements is increasingly common in the United States study. Reference and survey databases may not be appro- Lnd worldwide, and accurate estimates of totai intakes priate for some types of clinical research. The knowledge are not possible if supplements are not included. The offood and nutrition professionals can be especially use- accuracy and completeness of supBlement composition fui in guiding researchers to the most recent food compo- databases have traditionall.y lagged behind those of food sition data and pointing out the limitations of using ref. cornposition databases, but efforts to improve data on the erence databases in ciinical researih protocols. conaposition of dietary.supplements are underway (8). MAINTAINING DATABASES WITH A CHANGING FOOD SUPPLY The constant changes in the food marketplace preseni GONSIDERATIONS FOR USE OF DATA SOURCES issues and challenges for those who maintain databases When searching for food composition data, food and nu- The process of updating databases r'equires timeiy inputs trition professionals should consider food names and de- a detailed understanding offoods in the food supply, ani scriptions, data quality, and data variability (35). Accu- difficult decisions about priorities. One issue facing data. rate food names and descriptors help ensure that the base developers is globalization ofthe food supply. Foods associated data correspond to the food for which informa- may be iocaliy grown, regionally produced, or importei tion is sought. Sometimes a food name may pertain to from almost anywhere in the world. Exotic varieties oi
December 2007 . Journal of the AMERICAN DIEIETIC ASSoCIATI0N 210i fruits and vegetables are readilv available, and seasonal- Data aggregation is a challenge for database compilers ity is no longer a f'actor in accessibility. Because of this, when brands try to distinguish themselves from compet- year-round sampling and analysis may be necessary to itors with different added nutrients and functional ingre- accurately refiect the components in some foods as the dients. Database developers need to di.fferentiate a*orrg values may differ significantly due Lo dift'erences in vari- similar producls that may have very different nutrient ety, growing conditioirs, storage, and transporlation. In content. For example, the food name "granola bars" fails contiasi, spLcialty vegetables may be glown in only one to capture the hundreds of products on the market de- location and be picked and packaged during one har- signed for athletes, for energy, or for various health is- vest season, thus requiring a local and seasonal sam- sues. If database developers do not distinguish among pling protocol. Globaiization of the food supply also these products, data users will not be able to accurately affects manufaitured products, For exarnple, tomato- estimate nutrient intakes. based soups may be prodr.iced ali year round, but the source of.the tomatoes could be from California, Mexico, or Israel depending on the season. SYNERGY BETWEEN FOOD COMPOSITION DATA AND DIETARY l)atabases generally contain composite values for com- ASSESSMENT monly eaten fruits and vegetables that are averaged for Dietary intake information is coilected as foods con- seasonal variables and market share of major varieties. sumed, and increasingly, also as dietary supplements When new varieties that differ greatly in nutrient content consumed, These food and supplement intake data are are developed, new database entries may become neces- then converted to food component intakes using food com- sary. For example, commercial.ly grown heritage potato position databases (Figure 2). The uitimate goal of di- varieties are purpie and yellow. If these differ sigtrifi- etary assessment is usually to compare reported intakes cantly in carotenoid or flavonoid content frorn regular to intake stand.ards such as the Dietary Reference In. potatoes, new database entries wiII be needed. Ethnic takes (42), or to food guidelines such as MyPyramid (32) diversity in the population and preferences for ethnic (or, frequently, to both). foods require expansion of databases to account for new Improvements in food composition databases comple- foods and recipe ingred.ients. Many foods that may have ment many ongoing advances in methods for collecting been thought of as exotic in years past are now part of and evaluating dietary intake data, For example, the American cuisine. USDA Automated Multiple-Pass Method appears to cap- The number of manufactured producLs continues to ture food intakes from the previous day without the un- expand. Yearly output of new products from the top 25 derreporting bias that has been problematic with tradi- food companies increased by more than 70Vo since 1999 tional 24-hour recall methods (43). A:rother important (37). Since 2004 there have been more than 1,900 new new method for collecting survey data involves the use of product introductions for just the top 25 food manufac' a propensity questionnaire for measuring intakes of in- lureis. The biggest contributors were Nestl6 (Nestl6 frequently consumed foods (44,45). At the 2006 Interna- USA, Glendale, CA), Kraft (Kraft Foods, Inc, Northfleld, tional Conference on Dietary Assessment Methods, these IL), and Unilever (Unilever PLC, London, UK), with 209, and numerous. other advances in assessment methods 207, and 190 products, respectively, in 2005' When new were described (46). As computerg have become smaller foods are introduced into the food supply, those who and less expensive and access to the Internet has ex- maintain nutrient databases need to decide if these foods panded, dietary assessment has become more available to should be added; that is, ifthey are likely to be short-lived the public. For example, the USDA MyPyramid Tracker or may have staying power in the marketplace. is a user-friendly Internet tool that can be used by con- Food manufacturers are quick to respond to consumer sumers to evaluate their food and nutrient intakes (47). trends. For example, in response to the popularity of Better estimates of food component intakes are now low-carbohydrate diets, which began in 2003 and peaked possible because the accuracy of the dietary assessment in 2004, the food industry introduced more than 2,000 methods and the accuracy of food composition databases low-carbohydrate products by August 2004 (38). Ten per- are improving. The credibility of the final intake evalua- cent ofthe total adult population was on a low-carbohy- tion is increased synergistically if both the food composi- drate diet that year (39). Popularity ofthe diets waned, tion data and the dietary intake methods are of the high. and by 2005 only 2.8Vo of consumers reported interest est quality. in jow-carbohydrate dieting. Introductions of new low- carbohydrate products continue to decline, and many products have been removed from the market. Another INCORPORATION OF DATA INTO DIETARY ASSESSMENT ixample of a rapid change in the marketpiace is that of SYSTEMS whole-grain products in response to the reiease of the Dietary assessrnent instruments such as 24.hour dietar5 Dietary Giidelines for Americans 2005 (40), which in- recalls, food records, and food frequency questionnairer cluded a recommendation to consume at least three (FFQs) are linked to food composition databases thar servings of whole-grain products per day. In anticipa- contain food names and levels of food components. Thr tion of consumer demand, food companies increased database associated with each instrument should be ac production of whole-grain products, and in the 8 weeks curate, current, and represent foods and components o. after the release of the Guidelines, consumers pur- interest; be specific, precise, and uniform; and have nr chased 12Vo more whole-grain breads, IgVo more whole- missing values. Researchers use 24-hour recalls and fooc grain rice, and 16Vo more whole-grain ready-to-eat record.s to obtain snapshots of individual diets in cross breakfast cereals (41). sectional studies and use FFQs to collect usual diets
21OB December 2007 Volume 107 Number 12 Dietary Food Re?erenc€ Connposition lnta kes Dietary Datahasss 1 Evaluatisn I tta 1Ftor.Lat ciraat5gr tqirh.r.d.F.t I ' tu:{r4run*/rjy'daIreI r*,*-! *.!.*fg'*s- I *1 VI ---.- t
1
Foo,d G,Foup Serving,s Compositio,n
Hietary Guide,lines for Arnet Figure 2, The dietary assessment process, in which food and supplement intake data are converted to food component intakes using food composition databases then compared to intake standards such as the Dietary Reference lntakes (42), or to food luidelines such as the Dietary Guidelines for Americans (40) or MyPyramid (32) (or, frequently, to both), resulting in a dietary evaluation. generally in retrospective assessments, to evaluate past provide intake estimates). The next step is to determine diets to determine relationships with current health mea- the source of the data used to build the instrument; the sures. data should represent the foods and components com. Researchers use a database such as the FNNDS to code monly consumed in the population of interest. foods reported by participants in 24-hour recalls and food The food list is constructed by identifying what the records; however, FFQs, which are the most commoniy target population consumes or by using stepwise muj.tiplr used instruments in large dietary epidemiology studies, regression techniques to identify foods that are the besl require special considerations for food composition data- predictors of dietary factors of interest. In the first ap bases. FFQs do not obtain as much detail as 24-hour proach, frequency data are evaluated to determine whicl recalls and food records and, therefore, require the con- foods are most often consumed by the target population struction of a database of eornposite f6od codes with as- The multiple regression approach is favored when thr sociated food components that represents each foqd line goal is to identify foods that are most predietive of certair (48). FFQs consist of a list of foods with very little detail, dietary compbnents, such as fiber intake, percent energJ and dach food on the list represents many variations of from fat, or vitamin C. A list of foods is compiled from fo6i that food or even a group of related foods, For example; compoSition tables or Surveys that identify importan. pizzais a typical FFQ line,item, and the food components nutrient sources, pilot tests are conducted to test thr associated with pizza must represent the m1r'iad variet- predictive value of the foods in the list, and foods with lov ies ofpizza that are availabie. Because the development predictive values are discarded. After multiple iterationr of a FFQ food composition database is complicated and of analysis a modest number of foods are identified tha requires much consideration, existing FFQs (and associ- are most predictive of the nutrients of interest (49). ated. food composition databases) that are used in nutri- __\Yith regard to portion sizes for the foods, options fo: tion epidemiology research could be consulted and con- IIQI *"V be to collect no portion informaiion (simpl, sidered before attempting to develop a new one. FFQ), specify a portion size as part of the food tine itin The frst step in the development of a food composition (semiquantitative FFQ), or include a discrete question fo database for an FFQ is to identify the population of in- portion (quantitative FFQ), fne Block Questionnaire fo terest, foods and food components of interest, and the Kids-Ages 2-7 (50) is an example of the first option. Air objective of the data collection (eg, rank individuals or age-specifi.c portion size is applied to each food. Iine iten
December 2007 o Jeg1n21 of the AMERI0AN DIETET|0 ASSoctATtoN 210 that represents the average amount consumed by that composition of foods eaten over severa.I days, If printed population based on national survey data. The Harvard materiai is needed for education purposes, the printed Adult Questionnaire (51) is sr
2110 December 2007 Volume 107 Number 12 research, institutional and restaurant foodservice, food sibility of data have improved along with the technology Iabeling, and education of practitioners (16,64). Novel to convert the information into useful formats for plan- software applications, including smart cards, persgnal ning diets, writing education materials, counseling pa- digital assistants, and artificial intelligence (65-72) are tients, and conducting research. Despite the avaiiability belng deveioped for a variety ofpractice settings. Despite of current food composition information, data for some this potential avallability of data and software, not ev- food components are still iimited. Because new products eryone has access to or uses the available technoJ.ogy continually appear in the food supply and scientific inter- (73-75). Food and nutrition professionals in all practice ests in new food cornponents emerge through research, settings need easy access to the hardware and software gaps will always exist between what databases contain required to use nutrient data, and software appiications and what food and nutrition professionals need from should perform required caiculations such as evaluations them. Needs for food composition data vary depending on based on intake standards at speeds needed for daily dietetic practice area; however, most food and nutrition practice and with output in formats suitable for the in- professionais will benefil from becoming more informed tended audience. about food cornposition data, expioring new ways to edu- Busy professionals need assistance to keep up'with the cate themselves about databases and d.atabase systems, profusion of new foods, exotic foods, modifrcation of cur- and advocating for what is most needed in dietetic prac- ient foods, and new food components. Information can be tice. found on the Internet, colleagues can be queried via prac- Improved features of food composition databases work tice group listservs, and food composition experE from synergistically with improved dietary assessment soft- aroottd the world can be quickly contacted via E-mail. wane to provide a greatly improved ability to evaluate The American Dietetic Association (ADA) also helps prac- diets at both the individual and group levels. Food and titioners keep current by including food composition in- nutrition professionals can play an important role in en- formation in their Daily News from ADA's Knowledge suring that food composition databases meet the highest Center postings (76) and lhe Journal and by evaluating standards and that they are used appropriately by con- nutrition and health outcomes via the ADA Evidence sumers who are seeking better nutrition information and Analysis Library (77). The National Nutrient Databank by health professionals in their research and practice. Conference is devoted to issues of food composition and Better dietary assessment methods and better food com- fosters communication among database generators and position data ultimately result in greatiy improved esti- users. The first meeting in 19?6 was jointly sponsored by mates of dietary intakes, AIA and the Academy of Pediatrics (78), and the 30th National Nutrient Databank Conference in 2006 was a satellite to the A-DA Food and Nutrition Conference and References Exposition. Food and Nutrition Professionals are encour- 1. US Department of Agriculture, Agricultural Research Senice Nutri- attend this meeting for yearly updates with re- ent Data Laboratory I{ome Page. Search the USDA National Nutrient ag.d to Database for Standard Reference, Reiease 19. Anailable. at: http:/ con- gata to advances in food composition and to-visit the www,nal.usda.gov/Eoic/foodcomp/search./. Accessed April 11, 2006. Lret ." Web site 6ttp://www'na1.usda. gov/fnie/foodcomp/ 2. US Department ofAgriculture, Agricultural Research Service. USDA conf/) for information about conference proceedings. Food and Nutrient Database for Dietary Studies. Available at: http:/, Food and nutrition professionals need a basic under- www.ars.usda.gov/Seruices/docs.htm?docid:7.673. Accessed October generated, 26, 2006. standing ofhow food composition data are the 3. National Center for Health.Statistics. National Health and Nutri. levei of precision needed for the practice area, how- to tion Exarhination Survey. Avail ible at: http:/iww'w.cdc,gov/nchs, apply data to individuals and group-s, and knowledge nhanes.htm, Accessed October 26, 2006. features for associated software. Despite 4. Ervin RB, Wright JD, Kennedy-Stephenson J. Use of dietary supple aLout important 1988-94. Vitol Health Stat 11. L999i244 database soft- ments ijn the United States, available information about how to select i-iii, 1-14. ware, food and nutritjon professionals may be uncomfgrt- 5. Dwyer JT, Picciano MF, Raiten DJ. Food and dietary supplemen able making appropriate choices (79). With the increasing databaies for What We Eat ia America-NIIANES. J Nutr. 2003 databases and sofbware, food ald 133(suppl):624S-63 45, variety of available 6. National Institutes of l{ealth, Office of Dietary Suppleinents. Di nutrition professionals must be prepared to reconcile dif- etary Supplement Ingredient and Labeling Databases. Availabl, ferences between different programs or databases a t: httpy'/dietary-supplements.info.nih.gov/Ilealth-tnformationlDietary (80,81) and guide clients in selecting and using software . Supplembnt-Ingredient-and-Labeling-Databases.aspx. Accessed OcLc (82), Although graduates ofdietetics education programs ber 26, 2006. (83), ?. Dwyer JT, Piccia::o MF, Betz JM; Coates PM. Mission and activitie should be a6le lo interprdt the composition of foods of the NIH O{fce of Dietary Supplements. J Food. Comp Anol. 200c . most college and university dietetics prograrns include 17:493-5 00. few coursei and iittle training in food composition data. 8. Dwyer JT, Picciano MF, Betz JM, Fisher XD, Saldanha LG, Yetle EA, Coates PM, Radimer I! Bindewald B, Sharpless KX, Holden . As appear for food and nutrition profession- opportunities Andrews K, Zhao C, Harniy J, WoIf WR, Perry CR. Progress i: als to-use nutrient data (84-86), it is important that they development ofan integrated dietary supplemeit ingredient databas develop competency in this specialized practice area, Per- at the NIH Office ofDietary Supplements. J Food Comp AnoL 2o0( haps a new ADA special interest or Dietetic Practice 19(suppl):S 108-S 114. of competencies for food and nutrition 9. Food and Drug Administration. Total Diet Study. Available at: http: Group or a list www.cfsan.fd.a.gov/-comm/tds-toc.html. Accessed October 26, 2006. profeisionals who specialize in this area is needed. 10. Penniagton JAT, Gundersoa EL. I{istory of the Food and Dng Ac mi.nistration's Total Diet Study-1961-1987. JAssoc Offic Anal Chen 1987:70;772-7 82. 11. Pennirgton JAT, Capar SG, Parfitt CH, Edwards CW. Hlistory of tl Food composition databases are importdnt tools for im.- Food and Drug Administration's Total Diet Study (Part II), 198' proving nitional health (87). Both availability and acces- 1993. J:{ssoc Offic Anal Chem Intl. 1996;79:163-170,
December 2007 . Journal of the AMEHICAN DIFIETIC ASSOCIATION 211 12. Rand WM, Penningbon JAT, \Iurphy SP, Klensin JC. Compiling Data Group, Ltd. Available at: http/iwww.gnpd.com/sinatra/gnpd/frontpagd. lbr Food. Composition I)ata Bases. Hong Kong: Urrited Nations Uni- Accessed September 19, 2006. versily Press; 1991. 39. Balizer H, the NPD Group. 19th Annual Repart on Eating Patterns in 13. Joumal of Food Composition and Anaiysis. Available at: http://www. America. PortWashington, NY: NPD Group; 2004. elsevier,com/w?s/find,/joumaldescription.cws-home/622878/des cription. 40. US Departments of llealth and Human Serviees and Agriculture. Accessed October 26, 2006. Dietary Guidelines for Ameiicans, 2005. Available at: irttp://www. 14. Peonington J.{T, Food composition databases for bioactive food com- healthierus.gov/dietaryguidelines. Accessed November ZZ, 2006. ponents. J Food Comp AnaL 2002;15:419-434. 41. Mancino,L, Going with the grain: Consumers responding to new 15. GrafBA, Milbury PE, Blumberg JB. Flavonols, ff,avones, flavanones, dietary guidelines. Econ Res Ser Amber Waues. 2005;3:4. and human health: Epidemiologicai evidence. J tuted Faod. 2005;8: 42. Institute of Medicine, Food and Nutrition Board. Dietary Reference De1_ton Intakes, the Essential Guide to Nutrient Requirements. Washiagton ro. i"tuiiir"it. E, Burlingame B, Chenard C, Selley B, Stumbo P. lnter- DC: National Academies Press; 2006. national Nutrient Databank Directory, 2006. Available at:. htlp:// 43. Blanton CA, Moshfegh AJ, Baer DJ, Kretsch MJ. The USDA multiple- www.nal.usda.gov/fnidfoodcomp/confl. AccesSed February 1, 2007. pass method aecurately estimates group total energy and nutrient 1.7. Internationai Network of Food Data Systems. lnternational Food intahe. J Nutr. 2006;136:259a-2599. Composition Tables Directory Available at: http://ww.fao.orglinfoods/ 44. Subar AF, Dodd KW, Guenther PM, Kipnis V, Midthune D, McDowell directory-en.stm. Accessed October 26, 2006. M, Tooze JA, Freedman LS, Krebs-Smith SM, The food propensity 18. Phillips KM, Patterson KY, Rasor AS, Exler J, Haytowitz DB, Holden questionnaire: Concept, development, and validation for use as a JM, Pehrsson PR. Quality-control materials in the USDA Natioaal covariate in a model to estimate usual intakes. J Am Diet Assoc. Food and Nutrient Analysis Program (NFNAP). Anal Bioanal Chem. 2006;106:1 55 6- 1563. 2006;384:134 1-55. 45. Tooze JA, Midthune D, Dodd KW, Freedman LS, Krebs-Smith SM, 19. Pehrsson PR, Haltowitz DB, Holde JM, Perry CR, Beckler DG. USDA's Subar AF, Guenther PM, Carroll RJ, Kipnis V. A new statistical National Food and Nutrient Analysis Program: Food sampling. JFood method for estimating the usuai intake of episodicaily consumed foods Comp Anol. 2000;13:379-389. with application to their distribufior,. J Am Diet Assoc. 2006;106: 20. Hayiowitz DB, Pehrsson PR,.I{olden JM. The identification of irey 15 75-1587. 46. foods for food coinposition researeh. J Food Comp AnoI. 2002;15:183- Heitmm BL, Lissner L, Winlivist A. Abstracts for the 6th Interna- 194. tional Conference on Dietary Assessment Methods. Available at: 21. Franke AA, Custer LJ, Arakaki C, Murphy SP. Vitamir C and fla- http://w.1CDAM6.dk. Accessed November 13, 2006. 47. vonoid Ievels of lruits ald vegetable foods consumed in Hawaii. J US Department of Agricultue. MyPpamid tracker. Available at: http:/iwww.MyPyramidTracker.gov. Accessed FooQ Comp Anal. 2004;17 :7-35. November 13, 2006. 48. McNutt S. lncorporation of data 22. BLitz CL, Murphy.SP, Au DL. Adding lignan values to a food compo- into dietary assessment systems. J Food Comp AnaL 2O08;21(supp1 1):520-526. sition table. J Food Comp Anol. 2007;20:99-705. 49, Wiileti W. Epidemiok)gy 23. Foster:Powell K, Holt SH, Brand-Mil]er JC. International table of Nutritional .2nd ed. Nev York, NY: Oxford University Press; 1998:?4-100. glycemic index and glycemic lbad values. Arr J Clin Nutr. 2002;76: 50. Block G. Block kids questionnaire-ages 2-?. Available at: http:// 5-56. www.nutritionquest.com/products/questionnaires_screeners.htm. Ac- 24. Schauer RR, Sehakel SF, Harnack IJ. Workirg with a glycemic index . cessed November 17, 2006. . database: Recognizing its uses and limitations in dietetic practice and 51, Willett W. Harvard Food Frequency Available at:httpll nutrition research. J Am Diet Assoc. 2006;106:A-58. Questionnaire. www.channing.harv ard:edr:/nhs/questionnaires/pdfsA{HSIl199 8.PDF. 25. Dep artment of Agriculture, Agricultural Research Service. USDA US Accessed October 15, 2007 National Fluoride Database of Selected Beverages and Foods. Avail- 52. Block G. Block 2005 Food Frequency Questionnaire. Available at: abie at: httpy'/www.nal.usda.gov/fnidfoodcomplData,rFluoride/Fluoride. hltp://wil'w.nutritionquest.com/productVquestionnaires_screeners.htm December 11,2006. htrnl. Accessed Accessed November l7, 2006, BJ, Staggs CG, Bogle ML. Tyramine in foods md 26. McOabe-Sellers 53. National Cancer Institute. NCI Diet History Available oxidase inhibitor drugs: A crossroad where medicine, Questionnaire. monoamine at: http:iiriskfactor.cancer,govllHQ)/. Accessed November 17, 2006. nutrition, pharmacy, and food industry converge. J Food Comp Anol. 54. Stumbo P. Selection of.dietary assessment systems. J Food Comp 2006;19(suppl 1):S58-S65. AnoI, 2008;21(suppl 1):S13-S19. McCabe Teague AM, Mock DM. Deter- 2?. Staggs CG, Sealey WM, BJ, 55. Buzzard IM, Price KS, Warren RA. Considerations for selecting nu. content of seiect foods using accurate and mination of the biotin trient-calculation software: Evaluation of the nutrient database, sensitive HPlC/acidia binding. J Fr)od Comp Anal.2O04:77:767-176' Am J CIin Nutr. !99I;54:7-9. Ibanez Garcia:Closas R, Pera G, Amiano P, 28. Jalszyn P, Agudo A, R, 56. Probst YC, Tapsell LC. Oveniew ofcomputerized dietary assessmenl food database of nitrosamines, het- Goizalez CA. Development of a programs for research and practice in nutrition education. J Nztr Eo . erocyclic amires, and polycyclic aromatic hydrocarbons. JNulr. 2004; Behau. 2005;37:20.26. 57. Frank GC, Pelicm S. Guidelines for selecting a dietary analysil 29. MeNaughton SA, Marks GC. Development oi a food crrmposition da- systdm. J.Am Diet Assoc. 1986;86:72-75. glucosinolates, the tabase for the estim?rtion of dietary iqtakes of 58. Hoover LW, Pelican S. Nutrient Data Bases-Considerations for ed. biologically active constituents of crtrciferous vegetables. Br J Nutr. ucators. J Nufr Edue 1984;16:58-64. 2003;90:687-697. 59. Vozeniiik GP, Choosi-ng fhe best nutrient analysis software for you; 30. Murphy SP. Changes in dietary guidance: Implications for food and needs. J An Diet Assoc. 1999;99:1356-1357. nutrient, databases. J Food Comp Anal.200L;14:269-278. 60. Orta J, Reinarts CL. Comparison ofhandheld computers for nuf,ritior 31. Murphy SP. Dietary Reference Intakes for the US and Canada: Up- assessment and support. J Am Diet Assoc. 1994;94:1409-1414, date on implications for nutrient databases. J Food Comp AnaL 61. Viocare. ProNutra. Available at: ):ttp;,/www.viocare.com,/Products 2002;I5:4II-4I7 ProNutra.aspx. Accessed October 31, 2006. 32. US Department of Agriculhrre. MyPparnid. Available ah http//Mtrw. 62. University of Texirs Health Science Center at l{ouston. Food Intak' MyPyamid.gov. Accessed November 13, 2006. and Analysis System. Availabie at: http://www.sph.uth.tr::rc.edu 33. US Department of Agriculture. Plramid servings database for USDA DellHealthyliving/default.asp?id=4008. Accessed April 2, 2007. survey food codes,. version 2. http://us+a.usda.gov/cnrg/services/ 63. University of Minnesota Nutrition Coordinatiag Center. Nutrien foodlinl<. Accessed November L3, 2006. Dlta System for Research. Availab1e at: http://www.ncc.umn.edu 34. Sharma S; Murphy S, Wilkens L, Au D, Shen L, Kolonel L. Extending products/ndsr.html. Accessed October 3 1, 2006. a multiethnic food compositioa table to include standardized food 64. McCabe-Sellers BJ, Chenard CA. Meeting the needs of US dietitian group servings. J Food Comp AnaJ. 2003;16:485-495. for food composition data. J Food Comp AnqL 2008;2(suppi 1):S2? 35. Penningion JAT. Applications offood composition data. Data sources s34. md considerations fot use. J Food Comp Anal, 2008;2L(suppl 1):S3- 65. Stumbo PJ, Novel software applications using nutrient databases, , s12. Food Comp AnaL 2003;16:293-297. 36. Greenfield H, Southgate DHT. Food Composirton Dota: Production, 66. Lambert N, Plumb J, Looise B, Johnson IT, Ilarvey I, Wheeler C Monagement, and Use. London, UK: Elsevier Applied Scieace; 1992. Robirson M, Roife P. Using smart card technology to monitor th 37. Aron. New Product Activity at 25 Top Companies. !'oo d Industry Rep. eating habits of children in a school cafeteria: 1, Developing an validatiag the methodology. J Hum Nutr Diet.2005;18:243-54. sr. a?:f;itol;f Products Database. chicago, IL: Mii'rtel International 67. Yon BA, Johnson RK, Harvey-Berino J, Gold BC. The use of a per
2112 December 2007 Volume 107 Number 12 so:ral digitai assistant ibr dietary self-monitoriag does irnprove not at: http://www.adaevidenceiibrary.con/default.cfm?auth = 1. Accessed the validity of self-repori,s of energy intake. J Art Diet Assoc. December 11, 2006. 2006:106;1 25 6- 125 9. 78. Hoover LW. Computerized nutrient data bases: 1. Comparison of 68, Wang D, Kogashiwa M, Kira S. Developnrent of a n.ew irstrument for nutrient analysis systems. J Am Dibt Assoc. 1983;82:501-bbb. evaluating indir.idualsr dietary iutakes. J Am Diet Assoc. 2006;106: 79. Johnson TE, Borja ME, Loftus MK, McClintock NK, Factors inffuenc- 15 88-1 593, ing entrepreneurial dietitians in the use ofthe computer as a profes- 69, Raper N, PerloffB, Ingwersen L, Steitlfeldt L, Anand J. An oven-iew sional tool. J Am Diet Assoc. 7999;99:!440-1442. of USDA's dietary intake data system. J Foad Comp AnaL. 2004;!7: 80. Lee RD, Nieman DC, Raiuwater M. Comparison of eight microcom- puter dietary analysis programs with the USDA Nutrient Data Base 70. Petot GJ, Marling C, Sterling L. An artificial iatellgence system for for Standard Reference. J Am Diet Assoc. 1995;95:858-86?, computer-a-ssisted menu plamring. J Ant Diet Assoc. 1998;98:1009-1014. 81. Snetselaar LG, Chenard CA, Hunsicker LG, Stumbo PJ. Protein 71. Gaal B, Vassaayi I, Kozmann G. A novel artificial iltelligence method calculation from food diaries of adult humans underestimates values ibr weekly dietary menu plannil;g. Methods Inf Med. 2005;44:655-64, determined using a biological marker. J Nutr. Lgg1;l2b:2833-2340. 72. Noah SA, Abdullah SN, Shahar S, Abdul-Ha:uid H, lCrairudirr N, 82. Conlors P. Nutrient databases: When the end-user is not a scientist. Yusoff M, Gl:azali R. Mobd-Yusoff N, Shafii NS, Abdul-tr{anaf Z J Am Diet Assoc. 2006;106:4,-80. DietPal: A Web-based dietary menu-generating* and management sys- 83. Commission on Accreditation for Dietetics Education, American Die- lent. J ]vled Internet Ra.s. 2004;6:e4. tetic Association. 2002 Eligibility requirdments and accreditation 73. Connors PI-, Simpson DF. influence of menu plaming strategies on standards. Available at; http://www.eatright.orglada,/files/2002_ER-A,S- the nutrient conposition ofTexas schooi luncbes. J Food Comls AnaL Web-updated_8-06.pdf. Accessed April 5, 2007. 2004;Ii:a59-468. 84. Aimanza BA, NeIson.D, Chai S. Obstacles to nutrition labeli:rg in 74. Fox tr&! Cripinsek MI( Conuor P, Battaglia M. Sdrool Nutrition Dietary restaurants. J Am Diet Assoc. 7997:97:157-167. Assessnrent Study-II summary of fiading:. AVail.airle at: h@//www. 85. Wootan MG, Osborn M. Availabiliiy of nutrition iaformation from fos.rrsda. govloandmenu,€ublisheCl/CNP/F ILES/SI'lDAlmnd.pdf. Accessed cbain resta'.rrants in the United States. Az J Preu Med.2006;30:266- April 5,2007. 268. 75. Turner PR, Burgin CE, Funderburg KM, Van Grevenhof J, Knehans 86. American Dietetic Association. At meal assembly stores, registered AW, Personal digital assistants: Do dietitians use them? J Am Diet dietitians help shoppers create fast, affordable and heal.ihy dinners Assoc. 2005;105(supp] 2):A-3?. [press release]. Available at: http://www.eatright.orgicps/rde/xchg/ 76. Graham L. How do I keep up with research in food and nutrition? ada.&rs.*sVmedia_10 109_ENU_HTML.htm. Aceessed December 8, 2006. J Am Diet Assoc. 2005;105:855. 87. Champagne CM. Food composition databases: Important tools for ?7. Ameiican Dietetic Assbciation. Evidence aaalysis library. Available improving national health. J Food Comp AnaL 2003;16:251-253.
December 2007 . Journal ot the AMEBICAN DtffEnC ASSoCIATI0N 211