Hyperlinks to non-FAO Internet sites do not imply any official endorsement of or responsibility for the opinions, ideas, data or products presented at these locations, or guarantee the validity of the information provided. The sole purpose of links to non-FAO sites is to indicate further information available on related topics.
This electronic document has been scanned using optical character recognition (OCR) software. FAO declines all responsibility for any discrepancies that may exist between the present document and its original printed version.
© 1991 Academic Press
Reproduced with the permission of Academic Press
INFOODS Guidelines for Describing Foods: A Systematic
Approach to Describing Foods to Facilitate International
Exchange of Food Composition Data
A. Stewart Truswell, *,†,1 Deborah J. Bateson, †
Kathryn C. Madafiglio, † Jean A. T. Pennington, *,‡ William M. Rand, *,§
AND John C. Klensin *,||;
* Nomenclature and Terminology Committee, INFOODS;
† Human Nutrition Unit, University of Sydney. New South Wales 2006 Australia;
‡ Division of Nutrition, Center for Food Safety and Applied Nutrition, Food and Drug Administration, Washington, D.C., U.S.A.;
§ formerly with the INFOODS Secretariat, Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A., and currently with the Department of Community Health, Tufts University School of Medicine, Boston, Massachusetts, U.S.A.; and
|| INFOODS Secretariat, Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A.
Received October 28, 1988, and in revised form October 26, 1990
This report presents the INFOODS Guidelines for Describing Foods, which were designed with the intent of facilitating interchange of food composition data between nations and cultures by compilers of nutrient databases. Familiarity with the system should also be useful in other areas of nutrition, e.g., in recording food intakes. The system is a broad, multifaceted, and open-ended description mechanism using a string of descriptors. Criteria are proposed for deciding whether a food is “single” or “mixed” (multi-ingredient), and different sets of descriptive facets are provided for these two classes of foods. The INFOODS system is the result of extensive international consultation and is intended to be culture independent. © 1991 Academic Press, Inc.
The science and practice of nutrition depend substantially and unavoidably on good food composition data. Knowledge of the quantities of nutrients and other components in foods is essential for estimating nutrient intakes, for dietetic advice of all kinds, for all judgments about the relative healthfulness of foods (Truswell, 1985), and for epidemiological research that relates indices of disease to intakes of food components. Yet no country has food composition tables (FCTs) or nutrient databases (NDBs) which have kept up with the concepts of modern nutritional science and the proliferation of foods and processing techniques. The best provided countries such as the United States and the United Kingdom have missing foods and missing components in their NDBs, while many countries have no NDB of their own. Where there is a value for a particular nutrient and food, this single value is often misleading (Rand, 1985): NDBs do not usually indicate which country each value comes from, whether from modern or obsolete analyses, how many samples it represents, or any estimate of the sampling and analytical variance.
1 To whom reprint requests should be addressed.
The only solution is international sharing as in other fields of science, including animal feed information (Robards and Packham, 1983). Where a local analytical value is not available, it is better for compilers of FCTs to find and use results from other countries than to indicate “-” or “0” in the data field. The International Network of Food Data Systems (INFOODS) was set up to promote international participation and cooperation in the acquisition and interchange of quality data on the nutrient composition of foods, beverages, and their ingredients in forms appropriate to meet the needs of government agencies; nutrition scientists; health and agriculture professionals; policy makers and planners; food producers, processors, and retailers; and consumers (Rand and Young, 1984).
INFOODS GUIDELINES FOR DESCRIBING FOODS
Equivalent Food Terms in the United States and the United Kingdoma
United States term
|United Kingdom term|
whole wheat flour
|whole meal flour|
a Additional examples can be found in Appendix 1 of Davidson et al. (1979).
Food composition data cannot be usefully exchanged between countries until we can standardize the nomenclature and description of foods (Rand and Young, 1984). The major task of the INFOODS Nomenclature and Terminology Committee was to develop proposals for this process. Different scripts and many languages may be smaller obstacles than inconsistent terminology by English speakers.2 English-speaking people may not recognize some food names used by English-speaking people in other parts of the world (Table 1) or even the same country. In the same city in the United States, trained interviewers who independently collected 24-h dietary recalls from the same children disagreed on 24% of the food names (Frank et al., 1984). Polacchi (1986) has published striking examples of inconsistencies in meaning and classification of foods with the same name among food composition tables of different countries.
2 These Guidelines have been developed in English with the hope that hey can later be translated into other international languages.
Most compilers of food tables are not interested in detail about food names; all they want is a representative figure for (say) “apples.” But if compilers are to derive this, they need detailed information about the different apples that have been analyzed so that they can aggregate the figures in a logical way, e.g., in proportion to the market share of different varieties or into broad groups like fresh or cooked—dessert or cooking apples (Paul and Southgate, 1978). The INFOODS Guidelines were designed primarily to help communication between producers of analytical data and compilers of NDBs and between compilers in different countries. For applied nutritionists in the field, simpler systems of food names applicable in the particular country can and should be used.
To assist data users in locating foods in food composition databases, most database compilers list food items alphabetically within food groups. Code numbers may be assigned to foods based on the food groupings and subgroupings. Thus, some food grouping systems may also be considered food coding systems. Simple food grouping systems are usually based on food source (e.g., vegetables, fruit, grain, meat, fish, poultry), but as the food grouping systems become more complex, the groups or subgroups may also be based on product use or type (e.g., desserts, entrees, snacks, dairy products, breakfast cereals). Although food source is generally well understood internationally, product use/type is often culture dependent.
The INFOODS Nomenclature and Terminology Committee recognized this culture-dependent concept of food groupings at the onset of its deliberations and sought a faceted approach to the description of foods which would take the emphasis off “grouping” foods while allowing the food source, cultural food use, and other characteristics of the food to be identified independently.
Many different structured food description systems have been proposed starting from different perspectives and with different goals and objectives. Examples that have appeared in the published literature include modifications of the INFIC (International Network for Feed Information Centres) system for describing animal feeds (Harris et al., 1980a,b) such as that of Butrum (1985), a system for epidemiological use in Europe called the Eurocode 2 (Arab et al., 1987), and the system developed by the U.S. Food and Drug Administration (FDA) that was originally called the Factored Food Vocabulary (McCann et al., 1988). That system, now renamed Langual, was developed originally to meet the regulatory responsibilities of the FDA and has been expanded for other applications.
The system needed by INFOODS was one that did not necessarily group or code foods or allow for computerized retrieval, but one which would consider the great cultural and commercial complexity of human foods and allow these foods to be described at the level of analysis in sufficient detail so that subsequent users of the analytical data (especially users from other countries) could clearly recognize a food and distinguish it from other foods. The INFOODS food description system which evolved provides for open-ended descriptive information relative to specific characteristics. The INFOODS system also distinguishes between single and mixed (multi-ingredient) foods because the Committee recognized that not all facets were applicable for single and mixed foods.
An international coding or registration system, as for telephone numbers, cars, or books, was not considered workable for foods. First, it is not possible to be sure that two foods with the same short name in different countries and cultures (or even in different parts of the same country) are so similar that they can be assumed to have the same chemical composition. Second, an international coding system for foods would require an expensive registry. Foods could only be given a code number if a letter or telex was sent to this international registry. Some scientists and some countries would not know about the registry, while others would be too busy to communicate with it or not accept its decisions.
Monohierarchical systems for classifying foods are not suitable for international exchange because in different cultures people see the relationships between foods in different ways. Even in countries where Western science is well developed, there is disagreement about grouping and classification of foods. Should they be grouped on the basis of similarity of nutrient content or on their culinary use or following biological taxonomy? However, for other purposes, e.g., cancer epidemiology, it may be more informative to group foods in other ways (type of wrapping, food additives present, etc.).
Food composition data are being used today in many more ways than they were when the first food tables were originally compiled for use in therapeutic dietetics. When a monohierarchical classification is constructed, decisions are made about what is and is not important forever. For international interchange, none of the existing monohierarchical classifications of foods is sufficiently flexible. This does not mean that such classifications may not be very useful in national and even regional food tables.
A MULTIFACETED DESCRIPTION SYSTEM
By using several words to describe a food, we can improve the probability of predicting its chemical composition; we can make it easier for someone else to locate a similar food; and we can make it more practical to determine when two foods are not similar. For example, vitamin C content varies with cultivar of a fruit (Paul and Southgate, 1978; Breene, 1983; Wenkam and Miller, 1965); iodine content depends on where a food was grown (Iodine Content of Foods, 1952). If the variety and location of origin are supplied with the name, we improve the chances of predicting vitamin C and iodine content.
The most practical approach for international sharing is a multifaceted description system. The facets are independent. As in a diamond or a mountain, there are different ways of looking at the object. For each facet there are several possible descriptors. The most useful extended description of a food for a nutrient database, therefore, is a string of descriptors. The following examples could well be used at present in compiling a food intake record:
Food 1. Fried calf liver
Food 2. Roast New Zealand lamb
Food 3. Heinz baked beans
Food 4. Sugared stewed cooking apples
These four foods, so briefly named, require at least seven facets among them: origin (calf, lamb, bean, apple), part (liver), country of origin (New Zealand), manufacturer (Heinz), added ingredient (sugar), preparation or cooking (fried, roasted, baked, stewed), and cultivar (cooking apple).
However, none of the names made by these brief strings of descriptors is adequate for compilers of FCTs. Because it can affect composition, we still would like to know for Food 1, what fat or oil was used for frying (added ingredient)? For Food 2, was this leg of lamb (part)? For Food 3, was the scientific name of the beans Phaseolus vulgaris and were they canned in tomato sauce (added ingredient)? For Food 4, were the apples heavily or lightly sugared (added ingredient) and were they peeled (preparation)?
SINGLE VERSUS MIXED (MULTI-INGREDIENT) FOODS
In developing the INFOODS guidelines for food description, one of the two major problems has been that (unlike most animal feeds) many human foods are compounded from several single foods. In peasant communities most foods are obtained as single items and only mixed with other foods late in preparation of meals. Single foods are generally straightforward to name and describe. In industrial countries, however, an important proportion of foods is bought compounded with other foods and ingredients. Examples include mixed grain breads, cakes, and pastries, puddings, blended oils, margarines, sausages, pizzas, fish fried in batter, fish fingers, canned soups, fruit jams and jellies, mixed fruit juice drinks, salad dressings, bottled sauces, liqueurs, and cocktails.
Mixed foods presented three difficulties. The first was where to set the dividing line between single and mixed foods. Does a tiny amount of an added ingredient (e.g., salt or sugar) make a single food mixed? If not, how much of other ingredients may be added before a food is classed “mixed” and how can this be estimated by NDB compilers who did not produce the food? Second, can mixed foods be described using the same system as for single foods without the multiplication of facets becoming long and tedious? Third, are different facets needed for mixed processed foods made in a factory and for recipes using a mixture of foods prepared in a home kitchen? Many discussions, trials, and new beginnings were needed to produce the system outlined below.
BREVITY VERSUS COMPREHENSIVENESS
This is the other major problem. From reviewing the names of foods in the major food tables and during discussions at INFOODS terminology meetings, the list of facets grew. As well as the obvious facets for a food description, such as scientific name and method of processing or preparation and proprietary name, other facets can sometimes affect the composition of a food and might be useful if one wants to match one food with another. These include local or dialect name used by the people who usually eat the food, country (or area of the ocean) where the food originates, degree of cooking, agricultural production conditions, maturity, storage conditions, grade, physical state, and color. For a little-known food, it is important also to know if there is a photograph or line drawing on file or published and where this can be found.
When all these facets for single foods are put together in a guide, with brief explanations for each, and a different arrangement of facets for mixed foods, the description scheme looks discouragingly long. Will people use it?
The work involved in providing all the relevant descriptors for the full description of a food is not as much as it seems on first reading Table 2 for these reasons:
The INFOODS system is a complete listing of the information that might possibly be needed. It is obvious that few, if any, of the providers of information will be able to supply all that is called for. It is thus implicit in the INFOODS description scheme that every user, either as supplier or receiver of data, will need to have a specifically tailored individual protocol, program, or form for entry and retrieval. Such protocols will not request information that is unavailable or unnecessary.
INFOODS SYSTEM FOR DESCRIBING FOODS IN NDBs
The system shown in Table 2 is a sequence, program, or flow diagram for working systematically through facets for describing foods that may be needed by compilers of NDBs. Table 2 is presented here as an example of a form which a supplier of data would complete, e.g., at a computer keyboard, when preparing food composition data for transmission to an operator at the NDB in another country. Successive prototypes of the scheme were evaluated by members of the Nomenclature and Terminology Group between April 1985 and December 1987.
Table 2 is a summary of headings. Each of the facets is described briefly in the following text. The letters and numbers refer to the headings in Table 2. Thesauri with preferred terms, definitions, and alternative terms may eventually be developed for some of the facets, and the Multilingual Food thesaurus of the Commission of the European Communities (1979) may be used to translate these thesauri into other languages.
Section A. Source of Food Name(s) and Descriptive Terms
The source of the information should be documented as fully as possible in a standard format.
Aggregate foods should be described as such, e.g., “apple, raw; values represent weighted market shares of all apples consumed in the country in a particular year.”
Section B. Name and Identification of the Food
B1 is used for the food's name in a national language of the country in which it was obtained for analysis. This is usually the country in which it is intended for consumption. This facet is to be used for every food. The name should be narrow rather than broad. Two or three words may be needed here.
If the food has a proprietary (trade) name, this should be included. Proprietary names like Coca Cola, Rice Krispies, Marmite, or Drambuie increase the precision of identification of a food or drink. Even if it is policy not to use proprietary names in food tables (e.g., because of legal problems if specific foods deviate from an average value), these names are needed at the level of the original NDB. Formulation of most proprietary foods is maintained within fairly narrow limits by the manufacturer and hence the composition is more reproducible than for some other foods. The recipe is, however, occasionally changed, e.g., if there is a large movement in price of an ingredient. Levels of specific components may change in response to public health issues, e.g., in some proprietary foods, the sodium content has been reduced in recent years. If the food has a proprietary name, its generic name could also be provided, e.g., Post Toasties, corn flakes; Big Mac, hamburger sandwich.
The name in the national language is not required to use Roman letters. For example, in the Japanese food tables, the food name is given in Japanese characters after the code number (international numbers) and then the corresponding name in English and a brief description follow in the next column (Standard Tables of Food Composition in Japan, 1982). If necessary, the name can be transliterated from Japanese, Cyrillic, Arabic, etc., characters into Roman letters. In FCTs prepared in Beirut (before the civil war), up to four names are given for each food, common (generic) name in English, proprietary name (where appropriate) also in English, scientific name in Latin, and the Arabic name transliterated into Roman print (Pellet and Shadarevian, 1970). In several countries there is more than one official national language, eg, Belgium, Canada, Ireland, Hong Kong, India, Kenya, South Africa, and Switzerland. Where the language is not English, French, or Spanish, the name of the language should follow (in parentheses) the name of the food.
The local name for the food (B2) is not usually needed, but gives greater specificity in a large country like China or in South America where Spanish is the national language everywhere except Brazil. Even in a smaller country, the local name may be useful for a regional recipe or dish, a local variety of a food plant, or a wild food not known in the major cities of the country. Local names are indispensable for reporting on wild, “bush,” or traditional tribal foods (Brand et al., 1982; Kuhnlein et al., 1979). Local names are used in the Malaysian FCTs (Siong, 1982), for example, for a number of vegetables. The FCT for use in Latin America has English, scientific (Latin), and “principal” Spanish names in the main table (Leung and Flores, 1961). At the end there is a glossary of local names. The name of the local language should be given where it is known, but in some cases the word has been adopted into the national language; localization can then be provided by supplying the locality. If written in characters other than Roman, it may be convenient to transliterate the local name.
INFOODS System for Describing Foods
Use English, French, or Spanish, except for B1 and B2.
|E. Customary uses of food||Optional for single or mixed foods|
The next facet is the nearest equivalent name of the food in English, French, or Spanish (B3), whichever of these is the language for the rest of the description of the food. Where B1 (and B2) are already in one of these languages, this facet is not needed. Well-known food names like “lasagne” or “macaroni” that have become internationalized need not be translated into English, French, or Spanish, but “rigatoni” or “fusilli” in B1 could be referred to as “type of pasta” in B3. As with all translations from one culture to another, this name cannot usually be as precise as those in B1 and B2, but it can at least put the food in a small food group such as wild native leafy vegetable or small stone fruit.
B4 is the country or area where the food is obtained (e.g., bought) for analysis. Usually this coincides with the country where the food is intended for consumption. If a native food obtained in Ethiopia is sent to Sweden for analysis (Ågren and Gibson, 1968), the country for B4 is Ethiopia (Sweden goes in Section A). If an infant milk formula is imported from Europe to Nigeria and bought for analysis in Nigeria, the country where the food was obtained is Nigeria. If a fruit grown in New Zealand for export is analyzed in New Zealand (Visser and Burrows, 1983), the country where the food was obtained is New Zealand. In large countries, the state (in U.S.A., India, Australia, etc.), province (e.g., in Canada), republic (e.g., in the USSR), the general area (“in the north”), or a specific location matching the local name under B2 may need to be indicated.
B5 is for the food group and code in the NDB used in the country (in B4). For apple in Britain, the group is fruit and the codes are 675 to 681 depending on type and preparation (Paul and Southgate, 1978). If the apple is in the Federal Republic of Germany, its food group is fruits with cores including their products, and its code is 290201 (Scherz and Senser, 1989). These are the codes for aggregated data, but for individual food composition data used in NDBs, a more complex numbering system may be used. Many smaller countries do not have comprehensive FCTs of their own. At present they may borrow US or UK NDBs or use regional tables.
B6 is for the food group and code in the regional NDB. For apples in either the UK or West Germany, the Eurocode group is fruits, -products, -dishes and the code number is 10.303 (Arab et al., 1987). Countries which do not have a regional NDB may have regional food tables. For apples in an African country the group and code would be fruits, code number 860 (Leung et al., 1968).
B7 is for the food group at the international level. A system of relatively narrow food groups which can cover all possible foods without a “miscellaneous” group is needed for searching and retrieval of food composition data. There appear to be two alternatives. First, the Codex Alimentarius Classification of Foods and Animal Feed-stuffs (Joint FAO/WHO Food Standards Programme, 1984) has a classification of single, unprocessed foods, but is at present inadequate for classifying processed foods. Second, an INFOODS set of keywords, at present in draft, could be developed.
Sections C and D. Descriptions of Single and Mixed Foods
Here we consider and give rules to help decide whether a food is “single” (go to Section C) or “mixed” (go to Section D). A string of descriptors works very well for identification of single foods. The majority of food entries in tables designed for use in those developing countries which do not have their own NDB (Chatfield, 1954; Platt, 1957) are single foods—mostly raw, but in some cases cooked (unmixed) or milled. In the industrial countries, many of the foods that people buy are packaged and contain multiple ingredients. Likewise, many of the cooked dishes that people eat (and report to nutritionists) are mixed dishes. If ingredients are added in small amounts, e.g., a preservative, a food can be described as for a single food; if there is a mixture of several major ingredients, it becomes impossible and unnecessary to have a long string of descriptors for each. In mixed foods, minute specification of the ingredients becomes less essential in predicting nutrient composition. The most important information is the relative amounts of the ingredients and how they are processed and/or prepared.
We propose that foods described as single foods should fit into one of the following categories:
Mixed or multi-ingredient foods, on the other hand, are those which have more than one major ingredient, such as mixed grain breads, cakes, pastries, puddings, luncheon meats, hamburger sandwiches, pizzas, soups, salad dressings, and cocktails. Some foods are usually, but not always, made from a mixture of similar ingredients, e.g., margarine from several vegetable oils, fish sticks/fingers from several species and parts of fish, and frankfurters from several species or parts of animals. Their ingredients can vary with market price and availability.
Fried foods are best treated as mixed foods. French fried potatoes (British chips) contain about 16 g of fat compared with 0.1 g/100 g in boiled potatoes and their energy content is 315 kcal compared with 86 kcal/100 g in boiled potatoes (U.S. Department of Agriculture, 1984). From the nutritional viewpoint, information about the fat or oil is more important than details about the potatoes.
Section C. Systematic Description of Single Foods
C1(a), the food source in English, French, or Spanish must be expressed as the whole animal or plant. For some foods, the food source is the same as the food name (B1–B3), e.g., cabbage, onion, oyster. For many common foods, however, the food name is different from the food source:
|B1. Scrambled egg.....................||C1(a) chicken or hen|
|B1. Cheddar cheese...................||C1(a) cow|
|B1. White wine............................||C1(a) grape (vine)|
|B1. Brown bread.........................||C1(a) wheat|
The scientific name of food source (C1 (b)) is usually the zoological or botanical name in Latin and usually consists of two words, one for the genus and one for the species. Corresponding names are
|C1(a) Chicken............................||C1(b) Gallus domesticus|
|C1(a) Cow..................................||C1(b) Bos taurus|
|C1(a) Grape vine........................||C1(b) Vitis vinifera|
|C1(a) Wheat...............................||C1(b) Triticum aestivum (or vulgare)|
The scientific name is indispensable for little-known plants and animals, for wild bush plants, e.g., Terminalia ferdinandiana (Brand et al., 1982), and for fish, whose popular names can vary. The scientific name is least useful for major classical food plants and animals because these have been extensively hybridized, and the exact species name for a type of wheat or brassica would require time-consuming research in botanical taxonomy.
The variety of the source (C1(c)) is a subdivision of the species, usually in English, French, or Spanish or sometimes in Latin. Examples are different varieties of potato (new, maincrop), different varieties of apple (Granny Smith, Golden Delicious), milk from British cows in general or from Channel Island cows (i.e., Jersey, Guernsey) (Paul and Southgate, 1978).
Part of plant or animal (C2) is the part of the food source used as food (and analyzed for composition). For plants, this may be the root, tuber, stem, leaf, flower, fruit, nut, seed, juice, or oil. For animals it maybe a particular cut of mostly muscle (e.g., rump steak), an organ (e.g., liver), or blood. This facet also includes products from animals and plants, such as milk, egg, roe, gum, or syrup. In a few cases, the whole of the food source is eaten, e.g., whitebait, sardines.
Country or area of origin (C3) is the country or area where the plant was grown and harvested or where the animal was raised and/or slaughtered. For example, USDA Agriculture Handbook No. 8–9 (1982) gives two sets of data for avocados (Persea americana) depending on area of origin. The representative lipid value for California avocados is 17 g/100 g; the lipid value given for Florida avocados is 9 g/100 g. For fish and other seafood, the corresponding locality is the area of ocean or name of the sea or lake where they were caught or harvested. This can be standardized using the FAO Fishery Statistics.
C4 is the manufacturer's name and address. If a multinational company is named, the factory in which the analyzed food was produced should be noted because recipes and ingredients can vary among countries for products with the same brand name. They can also change over time so that a batch or lot number should be recorded with the primary data in a NDB.
Other minor food ingredients or additives (C5) include nutrient enrichment (e.g., added vitamin C). The ingredients should be listed in decreasing order by weight as is commonly used for ingredient labeling in several countries. Included as other ingredients are packing mediums such as water or vegetable oil in canned tuna, syrup in canned fruit, tomato sauce in canned sardines, or brine with olives. Most of the chemical additives (except flavors) appear as E code numbers on the labels of packaged foods in Scandinavia, the European Community, and Australia (e.g., E102 is tartrazine, E621 is monosodium glutamate). These international code numbers may be conveniently used to record food additives. If all the ingredients and additives or their relative amounts are not known, these uncertainties should be noted. Some foods are specifically sold as containing “no additives” or “no artificial colors” and such statements also belong here.
Food processing and/or preparation (C6) is one of the major facets. Food tables which omit this information are very difficult to use. It is a weakness of the otherwise excellent Souci et al. (1981) German food tables that, although they give the name of each food in German, English, and French, they do not make it clear whether the food has been cooked or by which method. The new 4th ed. by Scherz and Senser (1989) now gives data for vegetables raw and cooked. A sequential list of processing/ preparation treatments applied to the food should be listed, beginning with the first step of processing and ending with the last. Methods of cooking (frying, baking, boiling, etc) should be recorded.
Food processing can be defined as large-scale treatment of food in a factory or plant at a different place from where the food is eaten. Food preparation can be defined as treatment(s) applied at the same place where the food is consumed or purchased as a take-away food. There is considerable overlap in the treatments of foods (and technical terms) for processing and preparation. The place where each processing or preparation step occurred should also be sequentially indicated. This could be factory, institution, restaurant, home, laboratory simulation of home kitchen, street vendor stand, or picnic/ barbeque, etc. As an example, the processing/preparation of apple to make applesauce could be written: peeled (home), cored (home), stewed (home), sweetened with sucrose (home), or peeled, cored, stewed, sweetened with sucrose (home). The number of processing and preparation steps is generally fewer for single foods than for mixed foods.
The preservation method (C7) used to prevent microbial spoilage should be indicated. Some foods have two or more preservation methods. The most common methods include canning, freezing, freeze-drying, drying, and the use of chemicals. This facet overlaps to some extent C6 (processing) and C5 (additives). It is assumed that most of these methods are applied in an industry setting. If they are done at home or elsewhere, this should be noted.
Degree of cooking (C8) refers to whether the food is raw, partially cooked, or fully cooked. Such information is important when one considers the biological safety of the food, both the likelihood of contamination by pathogenic microorganisms with undercooked food and the possibility of mutagen formation when foods are cooked at high temperatures. This description may be used, for example, to indicate rare versus well-done steaks or hard-boiled versus soft-cooked (boiled) eggs. Length of cooking should be noted too for vegetables where losses of vitamin C and folate have wide ranges depending on how long they are boiled.
Information about conditions of agricultural production or animal husbandry (C9) may be available and relevant. For plant foods, conditions that could affect nutrient composition include type of fertilizer used (or the opposite—organically grown), hydroponically grown, hand-picked versus mechanically harvested, late harvest, early season, etc. For animals such conditions include type of feed, free range versus stall fed or intensively housed, wild versus farmed (e.g., duck), method of slaughter, etc.
The maturity or ripeness (C10) of a food is one of the six INFIC facets used to describe animal feeds, but a minor one for human food. It can be considered at two states, at the time of harvest/slaughter or at the time of consumption. Polacchi (1986) has shown how legumes are misclassified in some food tables which do not make clear whether beans are green (i.e., immature, including pods) or dried seed only (i.e., mature). They have sometimes been classified as vegetables although their water content indicates that they must be dried seed. Nutrients in fruits can vary with degree of ripeness. Examples of immature animal foods are veal and young turkey. Information about maturity at the time of consumption (or preparation for consumption) may be desirable for meats, cheese, and wines, which are aged to develop properties such as flavor, tenderness, and bouquet over a period of storage during which physiochemical changes occur in the food.
Length of storage or product freshness or storage conditions (C11) may be available and relevant for some foods. Foods may be stored at several stages between harvesting or slaughter and the time of consumption. Examples of storage between harvesting/ slaughter and processing include meat aged under controlled conditions of temperature and humidity and fruit ripened by holding in a controlled atmosphere. Examples of storage between processing and (home) preparation include concentrated orange juice stored frozen or canned baked beans stored between the factory and opening the can in the kitchen. An example of storage between processing and consumption (or analysis) is bread which loses water as it goes stale. Potatoes could be freshly dug from the ground or stored for 3 months; the latter contain only one-quarter the vitamin C of the former (Paul and Southgate, 1978). In many countries, “use by” dates are stamped onto factory-processed foods and may be the only information available. In recording primary data for NDBs, the “use by” date should be noted (along with the date of analysis).
The grade (C12) of food is one of the six facets used by INFIC for animal feeds, but it is a minor facet for human foods. For some foods like apples and eggs, size is often used for grading; for cream it is the percentage fat. For meat, characteristics which correlate with fat content have changed over time because of concern about excessive consumption of animal fat. The measurements and characteristics of food grade differ according to food regulations in each country, but some international standardizations are being developed by Codex Alimentarius and the EEC. Alternatively, an informal, subjective grading could be used, e.g., “fruit of good quality.” Most of the fruit analyzed for databases probably is of better quality than that which most poor (and nutritionally vulnerable) people usually eat. It would therefore be important to identify poor quality staple foods if they are analyzed for nutrient content.
The container and food contact surface (C13) are the subject of special studies which investigate the possibility of uptake of metals (e.g., tin, iron, aluminum, or lead) or of plastic monomers (e.g., vinyl chloride). Another common surface is glass. Some contact surfaces can add small amounts of contaminants, but they are unlikely to have much effect on the nutrient content, except for iron.
C14 is for physical state, shape, or form. This is sometimes described by a word in the primary name (B1), e.g., beef mince, fish fingers, corn flakes, or it may be described under food processing or preparation (C6). This facet is necessary in case the physical state, shape, or form is not mentioned earlier; it helps to identify the food and might affect its nutrient content. Some examples include chopped ham, potato chips, split peas, compressed yeast, shredded suet, dried egg powder, and tomato puree.
C15 is for color of the food, if not stated earlier. There are a few foods for which the color affects nutrient content, notably the carotene in yellow (as against white) maize, orange (as against white) carrots, and orange (as against white) sweet potatoes. Description of color improves identification of a number of other plant foods, e.g., red, green, or white cabbage, red or green peppers, black or white dried ground pepper, purple or yellow passion fruit, pink or white grapefruit, black or green grapes. With most of these there are some differences in nutrient composition.
C16 is reserve space for additional descriptors that analysts or NDB compilers may wish to include if they feel that such information would be helpful to others.
A good photograph or drawing (C17) can save many words. Before an unfamiliar single food is analyzed, it should be photographed in color. Examples can be seen in a series of articles on analysis of Australian Aboriginal bush foods (Brand et al., 1983, 1985; Cherikoff et al., 1985). For little-known plant foods and for fish, photographs are essential, and diagrams are very helpful for describing cuts of meat. Pictures from food labels may be kept on file. This pictorial material may at present have to be stored and transmitted separately from written descriptions if the latter are in computer files. The record should indicate that a photograph, picture, or drawing of the food is available and where it is located. Some articles reporting analysis of unfamiliar plants are accompanied by a map of where they were collected (Kuhnlein et al., 1982).
Section D. Description of Mixed (Multi-ingredient) Foods
Mixed or composite foods contain at least two substantial ingredients (each single or mixed foods) which are incorporated into the mixed food in a factory or kitchen. A list of ingredients and a recipe should be available. These constitute the two most important facets for describing mixed foods.
The ingredients (D1) of multi-ingredient foods should be listed in descending order by weight. The ingredients may be obtained from the label of a commercial food or from a recipe. The source of the ingredient information should be indicated (e.g., food label, cookbook citation, or other recipe source). Quantities or relative proportions of ingredients, if known, should also be listed. Water should be included if it is an ingredient. The ingredients can be described with a few essential words using the characteristics of Section C. The amount of description appropriate for each ingredient should be sufficient to reproduce the recipe. The most useful of these characteristics are C1(a) food source, C2 part of plant or animal, C5 other ingredients, C6 food processing/preparation and where processed/prepared, and C7 preservation method. Table 3 provides an example using homemade meat sauce to be served with pasta. To use the full set of characteristics for single foods, as in Section C, for each of the ingredients of a multi-ingredient food would in most cases be both impractical and unnecessary. The individuality of multi-ingredient foods reflects the combinations and proportions of the ingredients and the way they are processed or prepared together. In a mixed food, each component food has less influence on the nutrient content than in a single food. If the individual food components of multi-ingredient foods have been previously described using these INFOODS guidelines, the descriptors in Section C for these foods will be on file and can be cross-referenced to the multi-ingredient food.
For recipe procedure (D2), the intent is to provide a brief, relevant summary of the preparation involved in making the multi-ingredient food, emphasizing those methods which could affect the nutritional composition (e.g., parts discarded, heating methods). For commercial products, little information may be available. Table 3 provides an example of a homemade food item. A flow diagram or algorithm of the recipe may provide useful additional information if available. D2 deals only with treatments that occur during combination of ingredients into the multi-ingredient foods, not with previous processing/preparation of individual ingredients. Some FCTs (Paul and Southgate, 1978; Food Composition Tables for the Near East, 1982) and their supplements (Wiles et al., 1980; Tan et al., 1985) give ingredients and simplified recipe methods in a style similar to that of Table 3.
Ingredients and Recipe for a Mixed Food
Homemade Italian style meat sauce for pasta
onion, chopped, 2 medium size
lean ground beef muscle, 500 g
green pepper, chopped, 1 medium
celery, chopped, 2 medium stalks
carrots, chopped, 2 medium
garlic, chopped, 1 clove
tomatoes, 1 can (no additives), 425 g
tomato paste, canned, 100 g
water, 120 ml
basil, fresh, chopped, 1 tablespoon (or equivalent dried)
salt, 5 g
black pepper, to taste
parsley, chopped, to taste
Saute onion and ground beef in nonstick pan over moderate heat until the meat browns. Add green pepper, celery, carrots, and garlic and stir over heat for 15 min. Add tomatoes, tomato paste, hot water, basil, and pepper. Simmer covered for 15 min. Add parsley and heat an additional 5 min. Serve with boiled pasta (400 g before cooking).
The place where the multi-ingredient food was processed or prepared (D3) may be a factory, a fast food outlet, a large- or small-scale kitchen, or a laboratory simulation of kitchen preparation. Sometimes processed mixed foods are stored after purchase and then undergo a final stage of preparation where the food is consumed (home, restaurant, or institutional kitchen). This last preparation step can then be described under D9. If the food is consumed (as analyzed) directly after its preparation from ingredients, facets D6, D7, D8, and D9 do not need to be used.
If a photograph or picture (D4) of the multi-ingredient food is available, its location can be indicated here. Pictures from food labels may be useful to keep on file. If the multi-ingredient food was prepared by a factory, restaurant, or other institution, the name and address of the manufacturer (D5) should be indicated. For homemade dishes, this space would remain blank.
The container and food contact surface (D6) might be a can, box, or aluminum tray. The preservation method (D7) should be indicated if not already covered by the information above. Storage conditions (D8) may include time on grocery or kitchen shelves at room temperature or time in commercial or domestic freezers. The “use by” date may be the only information available on storage for some foods.
The final preparation (D9) of a mixed food may be heating a canned spaghetti bolognese, heating a frozen dinner, or cooking sausages. Some stored mixed foods, e.g., canned fruit salad or ice cream, do not need further preparation. Reserve space (D10) is provided as for C16 above.
Section E. Optional Supplementary Information
Information on customary uses of food is of value to nutritionists in other countries who cannot visualize the food and wish to know who eats it, how frequently, and how much at a time. The weight of a typical portion also has many uses inside the country. If there is a health claim or special dietetic use for a food, this is an essential facet for its complete description. This section applies equally to single and mixed foods.
Some food tables (Adams, 1975; Corden and Thomas, 1971) give the weight of typical portions (E1). There are three possible types of typical portion weight:
Portion sizes are essential data for food frequency determinations in epidemiological studies (Willett et al., 1985; Thompson et al., 1987). Typical portion weights should be the denominator in all tables for nutrition education that show good (and poor) sources of particular nutrients. They can be used as the basis for indices of nutritional quality (Hansen et al., 1979). The corresponding household measure or size [e.g., 1 stalk (celery), 1 medium slice (bread), 1 standard can] must be given as well as the gram weight.
E2 includes availability, frequency, and season of consumption. Frequency of consumption depends on both availability and social custom. The food describer should indicate whether the food is a staple or core food (eaten at least once a day) or eaten sometimes, occasionally, rarely, or only on ceremonial occasions (e.g., Christmas cake, Easter egg). Many plant foods exhibit seasonal variation in their availability, although advances in preservation and storage technology have extended these seasons in affluent communities. If a food is not commercially available or is difficult to obtain, this should be noted.
Usual place of the food in the diet can be recorded in E3. Is this a breakfast food, a main course food, a dessert food, a snack, a spread, a condiment, a beverage, etc?
E4 identifies the users of the food. Some foods are primarily prepared or marketed for a particular section of the community, e.g., for infants, children, a particular ethnic group, or perhaps the food is a luxury.
Specific purpose(s) of the food (E5) include ceremonies (e.g., wedding cake), religious observances (Passover food), and a variety of special dietetic purposes (e.g., lowfat yogurt as part of a weight-reducing regimen; fructose-sweetened jam for diabetics; gluten-free products for celiac disease). Any health claim should be noted here, e.g., high fiber. It should be possible to differentiate between claims which fit legal criteria and those which are the manufacturer's opinion.
Section F. Sampling and Laboratory Handling
Information on sampling and laboratory handling is the bridge between describing the food and recording its chemical composition. In the INFOODS structure, this area is common ground between the Nomenclature and Terminology Committee and the Data Quality Committee. The headings F1 to F10 in Table 2 are given as an example of a description system. They have been accepted by Dr. DAT Southgate for this purpose. There will be more detail on sampling and laboratory handling in the INFOODS manual Guidelines to the Production, Management and Use of Food Composition Data Systems (Greenfield and Southgate, in press).
All the questions in this section are essential when transmitting food composition data from one country to another. Lack of information about sampling and laboratory handling is as common an obstacle in using data from another country as are vague names for the foods in the NDB. Much of the presently published food composition data have limited value because they do not have attached to them information about when the samples were collected and analyzed (F1); whether adequate size samples were available (F2); where samples were collected and whether samples were representative of the market, specially made by the manufacturer, or quite casual (from the corner shop nearest to the laboratory) (F5); whether the laboratory analyzed multiple samples separately (and derived a weighted average from the results) or pooled multiple samples and homogenized them before a single analysis (F9); and whether in handling the food before and in the laboratory the investigators took appropriate precautions to preserve sensitive nutrients like vitamin C and folate (F6 and F7).
Most FCTs, however, are more informative about F3, the percentage edible portion (usually of the food as purchased) and its description, and F4 the percentage refuse (inedible or not usually eaten portion) and its nature. When food information is shared among countries, one cannot assume that inedible parts of a food in one country are not eaten in another country.
The food nomenclature system stops before laboratory chemical methods. This is the area dealt with by Dr. Southgate's INFOODS Data Quality Committee. But the method of analysis for each nutrient is seen by INFOODS as part of the name of the nutrient. This is incorporated in the report Identification of Food Components for INFOODS Data Interchange (Klensin et al., 1989).
The INFOODS food description system is open-ended (i.e., without standardized terms) allowing data generators to describe foods in their own words from different perspectives. The intention of the INFOODS system is to maximize the exchange and sharing of food composition data by allowing data users to clearly discern the nature of the foods associated with the data. The INFOODS description system will not supersede or replace systems used at present in regional, national, or specialized collections of food composition data. It should support and be compatible with all of them.
The INFOODS food description system is intended for use by chemists, nutritionists, dietitians, and food scientists who work in government agencies, academia, industry, and private agencies where foods are sampled, collected, and analyzed for nutrient content or where food composition data are gathered and compiled for databases. Although the system was originally designed for food composition databases, it is applicable to other food-related databases (e.g., food sales, inventory, imports, exports). Lastly, it is our belief that exposure of trainee nutritionists and dietitians to the food description system will help to improve clarity and reproducibility in food consumption measurement.
This INFOODS food description system is the product of international meetings of interested specialists in Madrid, Spain (July 1984); Cambridge, Massachusetts, United States (February 1985); Heidelberg, Germany (April 1985); Brighton, United Kingdom (August 1985); Norwich, United Kingdom (August 1985); Oxford, United Kingdom (April 1987); and Copenhagen, Denmark (July 1987). Contributors to the development of concepts are listed under Acknowledgments. The next stage of testing in the field and implementation will require international or multinational funding. We intend for various groups to begin to use it and to modify and improve it for their own specific needs. Food analysts and/or database compilers may also develop their own listings of preferred terms to describe foods within each facet. The use of preferred terms will help ensure consistency in the vocabulary of each laboratory or database.
The authors are especially grateful to the following for advice, discussions, and constructive criticisms during development of this INFOODS food description system: Jennie Brand (Australia), Ritva Butrum (U.S.A.), Lena Bergstrom (Sweden), Marianne Sylvada da Soza (Australia), Dianne Feskanich (U.S.A.), Heather Greenfield (Australia), Terry Leche (Australia), Anders Møller (Denmark), Wanda Polacchi (Italy), Roselyn Romeberg (U.S.A.), Dave Roberts (Australia), Nevin Scrimshaw (U.S.A.), David A. T. Southgate (UK), Jean Stewart (U.S.A.), Clive West (The Netherlands), and Vernon Young (U.S.A.). The authors are also grateful to other participants at INFOODS meetings on Nomenclature and Terminology held from 1984 to 1987: Lenora Arab (Germany), Clive Balch (UK), David Buss (UK), Max Feinberg (France), Harold Haendler (Germany), Georg Karg (Germany), Julien Perissé (Italy), J. Christopher Rigg (The Netherlands), Loll Rolling (Luxembourg), Derek Singer (UK), Christine Borrmann (Germany), Hans Pfannendorfer (Germany), Wolfgang Sichert (Germany), Ian Unwin (UK), Aree Valyasevi (Thailand), and Marion Wittler (Germany).
Adams, C.F.(1975). Nutritive Value of American Foods in Common Units. U.S. Department of Agriculture, Agriculture Handbook No. 456. U.S. Government Printing Office, Washington, DC.
Agren, G., AND Gibson, R. (1968). Food Composition Table for Use in Ethiopia. Institute of Medical Chemistry, University of Uppsala, Sweden.
Arab, L., Wittler, M., AND Schettler, G. (1987). Eurocode 2. In European Food Tables in Translation. Springer, Heidelberg.
Brand, J.C., Cherifoff, V., Lee, A., AND Truswell, A.S. (1982). An outstanding food source of vitamin C. Letter to the Editor. Lancet 2, 873.
Brand, J.C., Cherikoff, V., AND Truswell, A.S. (1985). The nutritional composition of Australian Aboriginal bushfoods. 3. Seeds and nuts. Food Technol. Aust. 37,275–279.
Brand, J. C., Rae, C., McDonnell, J., Lee, A., Cherikoff, V., AND Truswell, A. S. (1983). The nutritional composition of Australian Aboriginal bushfoods, 1. Food Technol. Aust. 35,293–298.
Breene, W. M. (1983). Raw product selection: Factors to be considered. Proceedings of a meeting of the Food Processors Institute February 1983, Los Angeles, CA, pp. 1–13. Washington, DC.
Butrum, R. R. (1985). Considerations in designing a system for description of foods and flexible retrieval of relevant data. Proceedings of a CoDATA Meeting on the Role of Data in Scientific Progress (P.S. Glaiser, Ed). Elsevier Science, Amsterdam.
Chatfield, C. (1954). Food composition tables—Minerals and vitamins for international use. FAO Nutritional Studies No. 11. FAO, Rome, Italy.
Cherikoff, V., Brand, J. C., AND Truswell, A. S. (1985). The nutritional composition of Australian Aboriginal bushfoods. 2. Animal foods. Food Technol. Aust. 37,208–211.
Commission of the European Communities (1979). Food: Multi-lingual Thesaurus. Vol.5. Saur, Munich/ New York/London/Paris.
Corden, M., AND Thomas, S. (1971). Simplified Food Composition Tables. Composition of Selected Raw and Processed Foods Expressed in Common Household Portions. Commonwealth Department of Health, Nutrition Section. Australian Government Publishing Service, Canberra, Australia.
Davidson, S., PASSMORE, R., Brock, J. F., AND Truswell, A. S. (1979). Human Nutrition and Dietetics, 7th ed., pp. 543–547. Churchill Livingstone, Edinburgh/London.
Food Composition Tables for the Near East. (1982). FAO Food and Nutrition Paper 26. FAO, Rome, Italy.
Frank, G. C., Hollatz, A. T., Webber, L. S., AND Berenson, G. S. (1984). Effect of interviewer recording practices on nutrient intake—Bogalusa Heart Study. J. Amer. Diet. Assoc. 84, 1432–1439.
Greenfield, H., AND Southgate, D.A.T. (in press). Guidelines to the Production, Management and Use of Food Composition Data Systems. An INFOODS Manual.
Hansen, R. G., Wyse, B. W., AND Sorenson, A. W. (1979). Nutritional Quality Index of Foods. AVI Press, Westport, CT.
Harris, L. E., Butrum, R. R., AND Sorenson, A. W. (1984). International food component data system. J. Amer. Diet. Assoc. 84, 1210–1215.
Harris, L. E., Haendler, H., Riviere, R., AND Rechaussat, L. (1980a). International Feed Databank System: An Introduction into the System with Instructions for Describing Feeds and Recording Data. International Network of Food Information Centres Publ. No. 2. International Feedstuffs Institute, Utah State University, Logan, UT
Harris, L. E., Jager, F., Leche, T. F., Mayr, H., Neese, U., AND Kearl., L. C. (1980b). International Feed Descriptions, International Feed Names, and Country Feed Names. International Network of Feed Information Centres Publ. No. 5. International Feedstuffs Institute, Utah State University, Logan, UT.
International Committee of Medical Journal Editors. (1982). Uniform requirements for manuscripts submitted to biomedical journals. Brit. Med. J. 284, 1766–1770.
Iodine Content of Foods. Annotated Bibliography 1925–1951. (1952). Chilean Iodine Educational Bureau, London.
Joint FAO/WHO Food Standards Programme. Codex Alimentarious Commission. (1984). Codex Alimentarious Part 4. Codex Classification of Foods and Animal Feedstuffs. CAC/PR4. FAO, Rome, Italy, and WHO, Geneva, Switzerland.
Klensin, J. C., Feskanich, D., Liu, V., Truswell, A. S., AND Southgate, D. A. T. (1989). Identification of Food Components for INFOODS Data Interchange. WHTR-14/UNUP-734. United Nations University, Tokyo.
Kuhnlein, H. V., Calloway, D. H., AND Harland, B. F. (1979). Composition of traditional Hopi foods. J. Amer. Diet. Assoc. 75, 37–41.
Kuhnlein, H. V., Turner, N. J., AND Kluckner, P.D. (1982). Nutritional significance of two important root foods (springbank clover and Pacific silverweed) used by native people on the coast of British Columbia. Ecol. Food Nutr. 12, 89–95.
Leung, W. T., Busson, F., AND Jardin, C. Food Composition Table for Use in Africa. (1968). FAO, Rome, Italy, and U.S. Department of Health, Education, and Welfare, Bethesda, MD.
Leung, W. T., AND Flores, M. Food Composition Table for Use in Latin America. (1961). Institute of Nutrition of Central America and Panama, Guatemala City, Guatemala.
McCann, A., Pennington, J. A. T., Smith, E. C., Holden, J. M., Soergel, D., AND Wiley, R. C. (1988). FDA's Factored Food Vocabulary for food product description. J. Amer. Diet. Assoc. 88, 336–341.
Pao, E. M., Fleming, K. H., Guenther, P. M., AND Mickle, S. J. (1982). Foods Commonly Eaten by Individuals: Amount per Day and per Eating Occasion. U.S. Department of Agriculture, Human Nutrition Information Service, Home Economics Research Report No. 44. U.S. Government Printing Office, Washington, DC.
Paul., A. A., AND Southgate, D. A. T. (1978). McCance and Widdowson's The Composition of Foods, 4th ed., H.M. Stationery Office and Elsevier/North-Holland, London.
Pellet, P. L., AND Shadarevian, S. (1970). Food Composition Tables for Use in the Middle East, 2nd ed. American University of Beirut, Lebanon.
Platt, B. S. (1957). Tables of Representative Values of Foods Commonly Used in Tropical Countries. Medical Research Council Special Report Series No. 253. H.M. Stationery Office, London.
Polacchi, W. (1986). Standardized food terminology: An essential element for preparing and using food composition data on an international basis. UN Univ. Food Nutr. Bull. 8, 66–68.
Rand, W. M. (1985). Food composition data: Problems and plans. J. Amer. Diet. Assoc. 85, 1081–1083.
Rand, W. M., AND Young, V. R. (1984). Report of a planning conference concerning an International Network of Food Data Systems (INFOODS). Amer. J. Clin. Nutr. 39, 144–151.
Robards, G. E., AND Packham, R. G. (1983). Feed information and animal production. In Proceedings, 2nd Symposium of the International Network of Feed Information Centres. Commonwealth Agricultural Bureau, Farnham Royal, Slough, UK.
Scherz, H., AND Senser, F. (1989). Food Composition and Nutrition Tables 1989/90, 4th ed. Wissenschaftliche Vergesellschaft mbH, Stuttgart.
Siong, T. E. Nutrient Composition of Malaysian Foods—A Preliminary Table. (1982). Division of Human Nutrition, Institute for Medical Research, Kuala Lumpur, Malaysia.
Souci, S. W., Fachmann, W., AND Kraut, H. (1981). Food Composition and Nutrition Tables 1981/82. 2nd ed. Wissenschaftliche Vergesellschaft mbH, Stuttgart.
Standard Tables of Food Composition in Japan, 4th ed. (1982). Resources Council, Science and Technology Agency, Tokyo.
Tan, S. P., Wenlock, R. W., AND Buss, D. H. (1985). Immigrant Foods. Second Supplement to McCance and Widdowson's The Composition of Foods. H.M. Stationery Office and Elsevier/North-Holland, London.
Thompson, F. E., Lamphiear, D. E., Metzner, H. L., Hawthorne, V. M., AND Oh, M. S. (1987). Reproducibility of reports of frequency of food use in the Tecumseh diet methodology study. Amer. J. Epidemiol. 125, 658–671.
Truswell, A. S. (1985). ABC of nutrition: Other nutritional deficiencies in affluent communities. Brit. Med. J. 291, 1333–1337.
U.S. Department of Agriculture, Human Nutrition Information Service. (1982). Composition of Foods. Fruits and Fruit Juices—Raw, Processed, Prepared. Agriculture Handbook No. 8–9. U.S. Government Printing Office, Washington, DC.
U.S. Department of Agriculture, Human Nutrition Information Service. (1984). Composition of Foods: Vegetables and Vegetable Products—Raw, Processed, Prepared. Agriculture Handbook No. 8–11. U.S. Government Printing Office, Washington, DC.
Visser, F. R., AND Burrows, J. K. (1983). Composition of New Zealand Foods I. Characteristic Fruits and Vegetables. Department of Scientific and Industrial Research, Wellington, New Zealand.
Wenkam, N. S., AND Miller, C. D. (1965). Composition of Hawaii Fruits. Hawaii Agricultural Experiment Station Bull. No. 135. University of Hawaii, College of Tropical Agriculture, Honolulu, Hawaii.
Wiles, S. J., Nettleton, P. A., Black, A. E., AND Paul, A. A. (1980). The nutrient composition of some cooked dishes eaten in Britain: A supplementary food composition table. J. Hum. Nutr. 34, 189–223.
Willett, W. C., Sampson, L., Stampfer, M. J., Rosner, B., Bain, C., Witschi, J., Hennekens, C. H., AND Speizer, F. E. (1985). Reproducibility and validity of a semiquantitative food frequency questionnaire. Amer. J. Epidemiol. 122, 51–65.