Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Uses of nutrient data bases for identifying nutritional relationships to public health and nutrition education in the united states

Carol T. Windham, R. Gaurth Hansen, and Bonita W. Wyse
Utah State University, Logan, Utah, USA

Ann W. Sorenson
Johns Hopkins University, Baltimore, Maryland, USA

The release of several reports in the United States that imply a relationship between diet and degenerative diseases has renewed public and professional interest in nutrition, especially as it relates to public health. Coincidentally, increased leisure time and interest in physical fitness have resulted in diet becoming an all pervasive consideration for some of the population. Many people in the United States have turned to promotional schemes expecting health benefits to be derived from food supplements and/or unusual combinations of foods. The need has never been more critical for reliable food composition information that can readily accommodate new research interests and ever-changing food consumption patterns.


The availability of food composition data together with food consumption data makes the application of computer technology feasible to the analysis of human dietaries and thus serves as a basis for advancing understanding of human nutrition. For purposes of evaluating food and nutrient consumption practices, defining the relationships between diet and disease, and developing relevant and credible nutrition education materials, three steps are critical:

1. More Complete Food Composition Data

For foods that comprise the majority of calories being consumed by the population, nutrient composition data are necessary. It is no longer adequate to have only the traditional nutrients, i.e., water, kilocalories, protein, total fat, carbohydrate, calcium, phosphorus, iron, vitamin A, thiamin, riboflavin, niacin and ascorbic acid. Increasingly, the food industry, educators, and consumers need food composition information for such nutrients as magnesium, potassium, sodium, zinc, copper, manganese, selenium, pantothenic acid, total vitamin B6, folacin, and vitamin B12, as well as for specific fatty acids, cholesterol, specific amino acids and dietary fibre. Many of these represent emerging interests, and for this reason the demand for reliable data exceeds availability

2. Detailed and Continuous Knowledge of Food Consumption Patterns

In the United States, national food consumption surveys have been conducted periodically since the late 1930s. Because of rapid changes in and expansion of the food supply, it has become increasingly important to know what people are eating as a basis for understanding the relationship between diet and disease and developing realistic nutrition education education programmes. Any reasonable definition of the association between diet and health or disease must rely on estimates of present and previous food intake. Food composition data bases provide the fundamental information necessary for these purposes.

3. Generally Accepted Standards of Nutrient Intake

To be useful for public health and nutrition education purposes, food consumption data must be evaluated with respect to standards of optimum nutrient needs. In the United States, recommended dietary allowances (RDA) have been developed, and for some nutrients, estimated safe and adequate dietary intakes have been suggested for purposes of evaluating quantitatively diets of population groups. RDA values need to be refined and expanded as better food composition data become available to serve as a basis for validating decisions about the nutrient qualify of the diet.


Food composition data in conjunction with dietary standards provide the means for evaluating dietary survey data such as food disappearance data, household food consumption data, and individual dietary intakes. Each of these measures food consumption at different levels, and thus the needs of researchers are varied with respect to which foods, nutrients, non-nutrients, processing factors, and other information must be accessible in a data base.

Food Disappearance Data

Researchers concerned with food disappearance data generally need information on the average levels of nutrients available from commodities in national civilian per capita food supplies. It is, nevertheless, important to these users to have specific information on the variability in nutrient levels of foods produced in different regions within a country. Improvements in food distribution and storage in the United States have resulted in wide dispersement of food commodities across the nation. It is still possible however, for local food supplies to contain less than adequate (or sometimes toxic levels) of specific nutrients. An example is selenium content in fruits, vegetables, and grains produced and consumed in local areas where selenium levels in the soil are very low {or very high). The issue of nutrient variability in foods is also of international concern, especially for foods that comprise the basic commodities of international commerce.

Household Food Consumption

Users of household food consumption data are concerned with nutrient values of edible portions of foods as they are purchased and enter the household. These include a broad spectrum of foods in various states: raw, partially cooked, ready-to-eat, fresh, canned, frozen, etc. The researcher, in this case, not only needs food composition data on basic commodities, but also on the nutrient levels of these commodities as they are modified during processing, pre-preparation, and packaging procedures prior to purchase.

Food and Nutrient Intake of Individuals

Researchers studying individual food intake practices must have detailed data on nutrient values of foods as consumed. There is a growing need for information on new and processed foods on the market, as well as for additional nutrients, such as zinc, sodium, and fatty acids that are emerging as important nutritional issues. These data benefit both the researcher to advance understanding of diet, health, and disease and the consumer to his nutritional and informational needs.

Some specific experiences with existing food composition data bases and individual food consumption surveys are highlighted in this issue. Historically, nutrient intake analysis in the United States has attempted (i) to determine average amounts of energy and nutrients consumed by various sex-age groups in the population, and (ii) to arrive at diet quality assessments by comparing group averages with the appropriate recommended dietary allowances (RDA) in the United States are expressed separately for males and females of different ages because of differences in growth rate or body weight and body composition. However, conversion of the individual RDA values into recommended nutrient levels per unit or energy illustrates that the variability in the recommendations per calorie is not very great, especially when considering the reliability of estimated need. Current recommended dietary allowances for many nutrients, when converted to allowances per 1,000 kcal, are constant, or approximately constant across all sex-age categories. Single-value nutrient allowances per 1,000 kcal (table 1) have been derived and are designed to meet the needs of all groups in the population when the energy needs of each group are met. For those nutrients whose RDA values per 1,000 kcal are not constant for all sexage classifications, the single-value allowances suggested as a standard are based upon the needs of those individuals whose calorie requirements are least, because they find it most difficult to meet all dietary allowances. Standard values so derived, when used in conjunction with food composition data, become immediately useful in evaluating dietary patterns across population groups.

TABLE 1. Single-Value Nutrient Allowances per 1,000 Kcal

Nutrient per 1,000 Kcal
Protein 24 g
Calcium 450 m
Iron 8 mg
Magnesium 150 mg
Phosphorus 450 mg
Vitamin A 2,000 IU
Thiamin 0.5 mg
Riboflavin 0.6 mg
Niacin 7 mg
Vitamin B6 1.0 mg
Vitamin B12 1.5 g
Vitamin C 30 mg

Condensed from R.G. Hansen and B.W. Wyse, "Expression of Nutrient Allowances per 1,000 Kilocalories." J. Amer. Dietet Assn 76: 223 (1980).

To illustrate this concept and its utility, current food consumption information has been analyzed to determine the nutritional quality of food consumed in the United States, with the nutritional adequacy of diets being measured relative to single-value nutrient standards derived from the RDA. An additional objective has been to identify food consumption patterns by examining the nutrient-toc-alorie ratio of foods consumed by various sex-age groups in the population.

Data used for the analysis were taken from the spring quarter of the Nationwide Food Consumption Survey, 19771978, conducted by the US Department of Agriculture (USDA). This survey included 8,661 individuals from a statistically selected sample of households in the contiguous states. The 24-hour recall and two day food intake records were used to collect data on individual food consumption. Nutrient values were computed by the USDA using the food composition data base developed for the survey by the Consumer Nutrition Center of the USDA. Agriculture Handbook No. 8 was the primary data source, with more recent data included when available. Nutrient data were included for over 5,000 foods and the following nutrients: kilocalories, protein, fat, carbohydrate, calcium, iron, magnesium, phosphorus, vitamin A, thiamin, riboflavin, preformed niacin, vitamin B6, vitamin B12, and vitamin C. The data base for this survey was developed in the early 1970s, when specific fatty acids, cholesterol, added calories as carbohydrate, sodium, zinc, and other trace minerals were not critical nutrition issues.

The findings of the study illustrated a consistency of nutrient consumption patterns across the US population (figure 1). Males and females surveyed were found, statistically, to be consuming diets whose average nutrient contents per unit of energy were almost identical, except for differences in intake of vitamins A and C. The average concentration of nutrients in diets was also relatively constant for all individual four years of age or older, with the most notable exceptions again being vitamins A and C. A highlight of this study is the indication that, on the average, the nutritional quality of food consumed in the United States is the same for almost everyone, regardless of sex or age. The results not only increase our understanding of food as consumed in the United States, but they also lend direction to food and nutrition policy development and provide additional support for the more universal standard for evaluating nutrient intake.

FIG. 1. Nutrient Density Profile of Average Daily Dietary Intake for Male (Ml and Female IF) Individuals, Based on Standard Allowances per 1,000 kcal

Limitations of Composition Data for Survey Uses

Nutrient data. A major limitation in conducting this study was the lack of food composition data for certain nutrients of interest, including sodium, folacin, zinc, cholesterol, specific fatty acids, and others. As mentioned earlier, these were emerging interests at the time the survey was conducted, but the lack of data for these nutrients illustrates the need for flexible data bases that can accommodate ever-changing nutritional concerns. Efforts are being made, within the limits of current assaying procedures, to update the USDA food data bases with respect to many of these nutrients.

Nutrient variability. Another concern in the study was the inability to explain adequately the variability in the reported data. For example, standard errors about the mean values were relatively small for protein intake (figure 2) when compared to those for vitamin B12 intake(Figure 3). This may have resulted from error in food composition data, error in food consumption data, from the inherent distribution of different nutrients in the food supply, or, most likely, from some combination of these and other factors. It does highlight, however, the need for more precise food composition data, along with indices of variability or quality of the reported values. It is essential, when documenting similarities or differences in consumption practices of various populations, to know that these derive from actual food habits as reported and not from inconsistencies in the basic food composition data.

FIG. 2. Mean Protein Intake per 1,000 Kcal for Individuals by Sex and Age. Index of nutritional quality of 1 (left vertical axis} is equivalent to the standard protein allowance of 25 9 per 1,000 kcal designated at right vertical axis. Confidence intervals computed at the 0.01 level.

Dietary standards. To determine the nutritional adequacy of diets consumed by populations, food consumption data must be evaluated with respect to standards of optimum nutrient needs. Current food consumption practices and food composition data are among the many factors used in the development of dietary standards such as the RDA. Examination of the RDA tables reveals some problems areas and raises the question of reasonableness of dietary standards for some nutrients

FIG. 3. Mean Consumption of Vitamin B12 per 1,000 Kcal for Individuals by Sex and Age. Index of Nutritional Quality of 1 (left vertical axis) is equivalent to the standard vitamin B12 allowance of 1.5 g per 1,000 kcal designated at right vertical axis. Confidence intervals computed at the 0.01 level.

For example, the RDAs for calcium are given in increments of 400 mg for different age groups, implying considerable uncertainty in the recommendations. Yet, the average American is being defined as being at risk with respect to calcium consumption. Vitamin A allowances, when expressed as a proportion of recommended calorie allowances, are higher for adults than for the developing child, although there appears to be no scientific basis for this. recommendation. As additional data become avail able, dietary standards need to be reviewed, revised, and expanded. Improvements in food composition data are essential for these processes, particularly since standards are fundamental to validating decisions about the nutritional quality of the food supply, as produced and consumed.


Nutrition education programmes and dietary guidance materials, when credibly developed and implemented, facilitate the selection of a nutritionally adequate diet. Food composition data are used in developing food and nutrient guides and in evaluating their impact on the nutritional quality of foods consumed by target populations.

Food Guides

Prior to 1970, categorical food guides were used almost exclusively as the tools of nutrition education. Even today, relatively unadaptable systems for categorising acceptable types and amounts of foods are the most common approaches to nutrition education. The assumption behind many of these food grouping systems is that consumption of a designated number of servings from specified food groups will provide the foundation of an adequate diet. As improvements in analytical procedures lead to more precise compositional data for a wider variety of foods and nutrients, this assumption may be invalidated.

Another concern with respect to most food selection guides is that they fail to consider dietary constituents such as sodium, sugar, fat, and other nutrients of current concern to the public for health-related reasons.

FIG.4. Nutrient Profile of Dairy Sub-group Cluster Consisting of Skim Milk, 1% Milk, 2% Milk, Buttermilk, and Lowfat Plain Yogurt

The rationale is that food guides must be so simple, convenient, and attractive that everyone can use them for choosing a total daily diet. There are certainly advantages to this approach, but increasing numbers of consumers are becoming more sophisticated in their understanding of nutrition and are requesting more specific nutrition and dietary information.

Researchers in the field of nutrition education are attempting to deal with these consumer demands by developing both expanded food guides and nutrient guides. A recent innovation has been the application of sophisticated statistical and mathematical classification procedures for objectively grouping foods on the basis of similar nutrient content. Using these procedures, the researcher is able to define food groups based upon the contained amount or concentration of whatever nutrients are of concern or interest. For example, dairy commodities have been classified into sub-groups which are identified in terms of the levels of fat, cholesterol, sugar and other nutrients contained in the foods comprising the sub-group (figures 4 and 5). The only limitation to this analysis is the requirement for precise compositional data in order to obtain accurate results, as the clustering algorithms are sensitive to small variations in nutrient values.

Nutrient Guides

A nutrient approach, with the educational emphasis placed on nutrients most difficult to obtain, has been recommended as a way to teach nutrition effectively. Teaching about foods as vehicles for nutrients obviously requires detailed composition information. The advantage to the individual is in having a more rational basis for making decisions about food. This decision-making ability is the ultimate goal of nutrition education.

A nutrient density nutrition education programme has been developed at Utah State University and evaluated for kindergarten through sixth grade students (ages 5 to 12). The programme allows nutrition education to be integrated into existing classroom programmes. Nutrient density, which compares the nutrients in a food with its caloric content, served as the conceptual framework. Materials were developed for student and teacher use and for teacher training.

The programme was evaluated in nine public elementary schools in Utah and Idaho with 806 student participants. Pre- and post-tests were conducted in addition to classroom evaluation. Students, teachers, and parents reacted favorably to the programme. The results indicate that elementary school-aged children possess sufficient academic sophistication to work effectively with a programme based on nutrient composition of foods. Further, with proper training and well-designed teaching materials, elementary school teachers with only a limited nutrition background can effectively use the nutrient density approach to teaching nutrition.

FIG. 5. Nutrient Profile of Dairy Sub-group Cluster Consisting of Whole Milk, Plain Whole Yogurt, Natural Cheese, Processed Cheese, and Cheese Spread

Limitations of Composition Data for Educational Uses

Nutrients. Nutrition educators have assumed the task of providing the public with information about foods and how to use them to obtain a nutritionally adequate diet. Advertising claims and food editors have sensitized the public to nutrient composition of the diet with respect to such food constituents as fibre and various trace elements for which good data are currently limited. Regardless of whether the educator chooses to teach about foods, nutrients, or some combination of the two, he or she is constrained by the lack of availability and precision in current food composition data.

Foods. There has always been a need for dietitians and physicians to prescribe diets of special nutrient composition. Now, with computer-based data processing techniques and the availability of small computers in both the school and the home, requests are being made for assistance in performing quality analyses of individual diets, presumably as an aid to food selection and preparation. An appreciable degree of sophistication is required to accumulate and keep current a data base that includes the important nutrients in foods. Nutrition educators are faced with the decision-making task of which foods and nutrients should be included in condensed data bases for clinical, educational, or home use on small computers. One approach would be to include those foods most commonly consumed by the population. Researchers at Utah State University, using data from the Nationwide Food Consumption Survey, have identified 85 foods that comprise 50 per cent of the calories consumed by the US population. In the ongoing Total Diet Study conducted by the US Food and Drug Administration (FDA), 240 basic foods have been identified and are being used to track levels of essential elements and some contaminants in foods and total diets. The availability of comprehensive, precise composition data for such basic food lists would greatly aid nutrition educators in their tasks

Dietary standards. For most nutrition education purposes, dietary standards such as the RDA are used as the bases for development of food guidance materials and as a measure for their effectiveness. There is a need for greater precision in these standards. Improvements in food composition data are essential for the task of determining human nutrient needs and developing appropriate standards of intake.


Food composition information is also important to food producers and processors. This information is a basic requirement for the development of methods to optimize the uses of food to provide adequate nutrition in an economical manner. This need is universal, hence an international system to meet the needs for food data would be useful world-wide. Such information can be used for assessing nutrient intakes and nutritional status, and for developing adequate diets for healthy and sick individuals. This need for food composition data is especially critical where food availability is marginal.

Food composition data serve as the basis for establishing the daily requirements for nutrients, establishing maximum acceptable intake of nutrients when appropriate, designing supplemental foods for vulnerable poblation groups, and designing national nutrition policies. Food composition data can thus be directed to agricultural production policies to better meet the nutritional needs of populations. Information on food production and availability, together with knowledge of food composition, contribute to the overall data base on which sound agricultural policies can be based.


1. Develop dietary standards of reasonable and adequate nutrient intake.

2. Identify food and nutrient consumption patterns of population groups and evaluate the adequacy of consumption

3. Relate food consumption habits to degenerative disease, including coronary heart disease and cancers.

4. Develop dietary guidance tools and education programmes and plans for individuals and groups across the population.

5. Perform nutritional assessments, diet evaluation, menu planning, and evaluation.

6. Facilitate the continuing efforts of the food industry to provide a more nutritious, safe, and saleable food supply

7. Develop national plans that project the demand for food.

8. Develop projects, including food assistance programmes, designed for ordinary and special needs, and to monitor and evaluate programme effectiveness.

Contents - Previous - Next