Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
1 Executive Summary Since 1936 the U. S. Department of Agriculture (USDA) has been responsible for conducting periodic surveys of food consumption. Currently, the agency's Nationwide Food Consumption Survey (NFCS), a large study of the food consumption patterns in the United States, includes information on individual dietary intake, which serves as a basis for determining the magnitude of inadequate nutrition in the general population. To ensure that the estimates of inadequacy are based on scientifically valid parameters, the USDA asked the National Research Council to develop criteria for the use of survey data in this effort. As a result of this request, the Subcommittee on Criteria for Dietary Evaluation was formed within the Food and Nutrition Board of the Research Council's Commission on Life Sciences. The subcommittee was charged by the USDA with estab- lishing criteria reflecting the degree of risk associated with intakes of the following nutrients: ascorbic acid; vitamins A, B6, and B12; calcite; folacin; iron; magne- si~m; riboflavin; niacin; phosphorus; thiamin; zinc; food energy; and protein. The agency also requested that crite- ria be established for the evaluation of the proportion of dietary intake derived from protein, fats, and carbohy- drates as well as from total energy (caloric) intake. Dur- ing the course of its study, the study group examined efforts of the USDA and others in the scientific community to assess the nutrient adequacy of diets in the U.S. population and considered the analytic methods used in the past. Data from the most recent NFCS survey were provided to the subcom- mittee to permit empirical testing of different approaches for dietary evaluation. 1

OCR for page 1
2 USDA food assistance programs and other nutrition- related projects are based in part on estimates of inade- quate nutrient intake derived from the NFCS. These esti- mates are also of interest to nutrition policymakers out- side the USDA, scientists, and others who wish to identify population groups at risk of developing nutrient defi- ciency and to learn the determinants of unsatisfactory dietary intake for the country as a whole and for specific population groups. Although estimates based on dietary intake data are useful for examining adequacy of nutrient intake in a population or subpopulation, and may be useful in identifying individuals at relative risk of developing nutrient deficiency, they cannot be used alone to deter- mine the nutritional status of individuals or population groups. For these purposes, biomedical and clinical measures are necessary. The Recommended Dietary Allowances (RDAs) are often used as the basis for determining whether nutrient intake is adequate. They are standards for nutrient intake designed to meet the nutrient needs of virtually all healthy individuals in the United States. Because there is variation in nutrient needs among people despite simi- lar physiological characteristics, margins of safety are built into the RDAs for many nutrients. Therefore, most people who receive less than the RDA for a specific nutri- ent will nevertheless meet their own nutrient require- ment. For a number of years, a fixed cutoff point, such as two-thirds or three-fourths of the RDA, has been used by analysts to estimate the prevalence of inadequate intake for specific nutrients. The subcommittee considered the merits of this type of analysis and concluded that it may lead to imprecise estimates, partly because it does not consider fully the variability in requirements among individuals. Conse- quently, some persons who are meeting their nutrient requirement may be judged by this method to have inade- quate intake while some with inadequate intake will not be identified. A different approach based on the probability that a specific intake is inadequate to meet an individ- ual's requirement was identified by the subcommittee, and guidelines were developed for interpreting the resulting estimates. Although the new approach is not difficult, it requires some familiarity with basic statistical theory. In this probability approach, estimates of average require- ments and variability (i.e., the standard deviation) for

OCR for page 1
the nutrient are used along with the shape of the distribu- tion (e.g., normal or skewed) as the criteria for judging adequacy of dietary intake. The approach also requires information on the distribution of usual intakes among individuals examined in the survey. Dietary data from the NFCS are derived from interviews to determine the foods respondents have eaten for 3 days. Because the intake of an individual varies over time, it is necessary to adjust the distribution of dietary data in order to estimate the distribution of usual dietary intakes. The subcommittee also recognized that the analysis of nutrient intake ade- quacy may be constrained by systematic errors such as underreporting or overreporting of food intake and lack of information on the mean and shape of the requirement dis- tribution for many nutrients. The subcommittee believes that the prevalence of inade- quate intake can be estimated for many nutrients and food components by using the probability approach. Empirical tests of the approach were made using intake data for iron, protein, vitamin A, and vitamin C in men and women and for thiamin in men. These data, from the 1977-1978 NFCS, were provided to the subcommittee by the USDA. The probability approach is not indicated for some nutrients, especially energy, as will be discussed below. The overriding constraints in the application and interpretation of the probability approach are the limita- tions, validity, and reliability of estimates of mean nutrient requirements and survey data on dietary intake. At present, direct estimates of mean nutrient requirements are not available for most nutrients. Thus, the proposal to undertake probability analysis of dietary intake calls for the assignment of a higher priority to the development of the knowledge base on mean nutrient requirements and to improvement of the data on dietary intake by the collec- tion of least two independent (i.e., nonconsecutive) observations for the same individuals. The subcommittee suggests that priority be assigned to the development of improved estimates of mean nutrient requirement for nutri- ents that a substantial proportion of the population is consuming at levels less than the RDA. In the meantime, the subcommittee believes that the use of the probability approach will both stimulate and guide efforts to improve the validity and reliability of nutri- ent requirement estimates by permitting examination of the

OCR for page 1
4 implication" of different requirement estimates for a population, given current levels of dietary intake. There is now a need for further empirical testing of the pro- posed approach to determine the applicability of the method and to establish directions for further research. MAJOR CONCLUSIONS - The prevalence of inadequate intake can be estimated for many nutrients by the probability approach described in this report. For others, however, the method cannot be applied until research leads to a better understanding of both the average nutrient requirement and its variability, which are needed in probability analysis, and an ~mprove- ment in the reliability of food composition data. These estimates are Important in identifying determinants of inadequate intake, identifying possible interventions, and designing them for maximal efficiency. They are depen- dent on more comprehensive surveys to validate dietary anal- yses through biochemical and clinical measurements, such as are currently done in the National Health and Nutrition Examination Survey (NHANES) of the National Center for Health Statistics (NCHS). A basic statistical assumption of independence between requirement and intake is necessary for the probability approach. Thus, this method cannot be used meaningfully when the level of dietary intake and the required intake are cor- related, as for dietary energy (calories), which most people in prosperous nations with low requirements consume at low levels. After reviewing the work of Lorstad (1971), however, the subcommittee concluded that this is not a problem. There is a need for continuing studies to improve research methods and thus data on dietary intake, which are essential for analysis regardless of the approach used. There is also a need for continuing attention to the validity of food composition data and research to improve such data. After examining the methods with which dietary intake data and reference data on the nutrient composition of foods are collected and conducting several types of analysis to determine the impact of random error, error due to the sampling technique, and systematic biases on the estimates of adequacy, the subcommittee concluded that such errors diminish the accuracy but do not necessarily destroy the utility of

OCR for page 1
5 estimates of the prevalence of inadequate intake. The subcommittee believes that sensitivity testing is needed to assist in determining which factors have the greatest effect on prevalence estimates and hence should be given priority for research to improve the approach. MAJOR RECOMMENDATIONS . Nutrient requirements based on multiple criteria of adequacy should be developed and applied. For a given nutrient, one might focus on the intake adequate to pre- vent clinical deficiency, to maintain functional integrity of metabolic systems, and to maintain tissue stores. This would permit multitiered population assessments. If the probability approach is adopted, the following suggestions should be considered when planning for future NFCS surveys: Changes may be advisable in the design of food intake data collection. For example, the nether of 1-day food intake observations per subject might be reduced; it would be preferable to use the same data collection meth- ods for each day of intake data; and it might be desirable to avoid sampling on adjacent days and to continue to sample on representative days of the week. These changes should be made only after full consideration of all the uses of the data and of the integration of survey planning for all these purposes. Methods to reduce, or take into account, respondent or interviewer bias should be developed to improve the accuracy of food intake data. Continuing research on food intake methods and the design of sampling strategies is recommended. Research should be conducted to determine the magni tude of any correlation between dietary intake and nutri- ent requirement. - The subcommittee also recommends a n~mber of actions that should be considered in order to improve the reference tables on nutrient composition of foods. These recommenda- tions, which are presented in Chapter 9, relate to documen- tation and analysis when data are missing, increases in sam-

OCR for page 1
6 ple size for nutrient composition analyses, and improvements in sampling methods. ~ . The design selected for future surveys should take into account all important uses of the survey data. The subc~mmittee's attention has been directed to one partic- ular type of use. Other purposes may impose additional design demands on data collection approaches. The subcom- mittee believes that agencies responsible for the design and conduct of national or regional surveys would benefit from conducting analyses analogous to those discussed in this report, including full statistical consideration of the implication of design decisions on the precision and reliability of data analyses. ~ It is imperative that future surveys include ques- tions on intake of dietary supplements as well as of foods.