Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A computerized system for instruction in food selection practice Prince, Peter Robert 1979

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1979_A6_7 P75.pdf [ 19.89MB ]
Metadata
JSON: 831-1.0094593.json
JSON-LD: 831-1.0094593-ld.json
RDF/XML (Pretty): 831-1.0094593-rdf.xml
RDF/JSON: 831-1.0094593-rdf.json
Turtle: 831-1.0094593-turtle.txt
N-Triples: 831-1.0094593-rdf-ntriples.txt
Original Record: 831-1.0094593-source.json
Full Text
831-1.0094593-fulltext.txt
Citation
831-1.0094593.ris

Full Text

A COMPUTERIZED SYSTEM FOR INSTRUCTION IN FOOD SELECTION PRACTICE by PETER ROBERT PRINCE B. Sc., University of British Columbia, 1969 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES SCHOOL OF HOME ECONOMICS DIVISION OF HUMAN NUTRITION  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA July 1979 (c)  Peter Robert Prince, 1979  In p r e s e n t i n g  this  thesis  an advanced degree at the L i b r a r y s h a l l I  f u r t h e r agree  for scholarly by h i s of  this  written  fulfilment of  the U n i v e r s i t y of B r i t i s h  make i t  freely available  that permission  for  the requirements  Columbia,  I agree  reference and  f o r e x t e n s i v e copying o f  this  It  for financial  is understood that gain s h a l l  permission.  Depa rtment The U n i v e r s i t y of B r i t i s h  2075 Wesbrook Place Vancouver, Canada V6T 1W5  Columbia  not  copying or  for  that  study. thesis  purposes may be granted by the Head of my Department  representatives. thesis  in p a r t i a l  or  publication  be allowed without my  ii  ABSTRACT  This thesis has developed arprototypical system which provides information on dietary practices for those individuals interested in applying nutritional principles to their eating habits.  The system has the poten-  tion to provide information which both accurately reflects nutritional guidelines and facilitates adoption of recommendations, by providing a self-explanatory statement of foods to consume and by 1imiting suggested changes in present food pattern. The prototypical computerized system developed has two major functions: ( i ) , diet-assessment to appraise the acceptability of individual's dietary practices; and ( i i ) , diet-planning to recommend modifications in the diets of those individuals not meeting specified limits.  The focus of the  system is a constrained-optimization algorithm that generates a revised food plan which both satisfies nutrient constraints, and minimizes the deviation of food items rand item groups from the original amount consumed by the client. Testing has been restricted to a descriptive evaluation of some of the algorithm's characteristics - - specifically, the design assumptions which define the acceptability of deviating from an original inventory, and the revised diets developed when these assumptions are modified.  The results  illustrate that altering these design assumptions produces marked variations in the revised diets with respect to observed parameters.  Further  modifications in the algorithm have been suggested. The explorative evaluation provides'ia foundation for more systematic evaluation of the validity of the algorithm.  Recommendations for f a c i l i -  tating the further development and testing of the system are^outlined.  1 ii This thesis has shown that mathematical modeling provides an effective means of collating the vast amount of data required to develop cogent dietary recommendations which are nutritionally accurate, straightforward, and acceptable to the client.  iv  TABLE OF CONTENTS 1 1  ABSTRACT TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ACKNOWLEDGEMENTS 1.  1 V v1 x  1 Background and Need for the Study  1  1.1.1  Fundamental Tasks of Nutrition Education Programs...  1  1.1.2  Criticisms of Nutrition Education Programs  2  1.1.2.1  Developing a Scientific-Nutritional Code  4  1.1.2.2  Communicating a Nutritional Code to the Public  5  1.1.2.2.1  Communication With a Variety of Audiences  1.1.2.2.2  Complex Dynamics of Communication  6  With the  Public  7  1.1.2.2.2.1  Changing Food Market  7  1.1.2.2.2.2  Mass Media and Advertising  9  1.1.3 1.2  2.  1  x 1  INTRODUCTION 1.1  . . "" .  Summary and Conclusions Thesis Goal :and Objectives  11 13  1.2.1  Thesis Goal  13  1.2.2  Thesis Objectives  13  1.2.2.1  The First Objective of the Thesis  13  1.2.2.2  The Second Objective of the Thesis  14  REVIEW OF LITERATURE  15  2.1  15  2.2.1  Definition of an Adequate Diet Influences of Diet in Human Populations  15  2.1.1.1  Influences on the Individual  15  2.1.1.2  Socio-Economic Influences of Diet  17  2.1.2  Criteria of an Adequate Diet  17  2.1.3  Nutrient Requirements  19  2.2 2.2.1  Relevant Systems  21  Food Guides  21  2.2.1.1  F a l l i b i l i t y in Guiding Food Selection Practices...  22  2.2.1.2  Inapplicability to the Present Food Supply  24  2.2.1.3  Relevance to Contemporary Nutrition Problems  25  2.2.1.4  Ineffectiveness as a Teaching Tool  26  1  V  2.2.2  Other Relevant Systems  27  2.2.2.1  Pennington's Dietary Nutrient Guide  27  2.2.2.2  Nutrition Labelling  28  2.2.2.3  Computer Applications in Nutrition and Dietetics..  29  2.2.3 2.3 2.3.1  Critique and Conclusions  32  Dietary Assessment  35  Data Collection  39  2.3.1.1  Data Collection Methods  39  2.3.1.2  Evaluation of Data Collection Methods  42  2.3.1.2.1  Validity and Repeatability of Methods...  44  2.3.1.2.2  Time Frame of Methods  46  2.3.1.2.3  Randomness and Size of Population-Samples  48  2.3.1.2.4  Other Metholodigical Considerations  49  2.3.1.3  Summary and Conclusions  50  2.3.2  Data Analysis  52  2.3.3  Data Evaluation  62  2.3.3.1  Dietary Standards  62  2.3.3.1.1  Formulation of Dietary Standards  64  2.3.3.1.2  Use and Interpretation of Dietary Standards  66  2.3.3.1.3  Limitations of Dietary Standards  72  2.3.3.2 2.3.3.2.1 2.3.3.2.2 2.3.3.3 2.4 2.4.1 2.4.1.1 2.4.1.1.1 2.4.1.1.2 2.4.1.2 2.4.1.2.1 2.4.1.2.2 2.4.1.2.3.  Further Dietary Recommendations  74  Recommendations for Nutrients Not in the Dietary Standard.;  74  Recommendations for Maximum Intakes  76  Conclusions  77  Dietary Prescription  80  Diet Planning Food Planning Models  80 82  Food Planning Without Palatability : ,.: i Considerations Food Planning Models with Palatability  82  Considerations  87  Menu Planning Models Menu Planning Models - The Random Approach Mathematically-Programmed, Multistage, MenuPlanning Mathematically-Programmed, Single-Stage, MenuPlanning  93 93 94 97  vi  2.4.2 2.4.2.1  Information and Behavior Change  98  Development of Food Behavior and Factors in Food Selection  3.  99  2.4.2.2  The KAP Gap  100  2.4.2.3  Increasing the Effectiveness of Communication  102  DEVELOPMENT OF THE PROTOTYPICAL SYSTEM  105  3.1  105  Introduction  3.1.1  Developing Nutritional Guidelines  105  3.1.2  Communicating Information to Individuals  106  3.1.2.1  Comprehensibility of the Information  106  3.1.2.2  Acceptability of Suggested Changes  107  3.2 3.2.1  System Design and Characteristics  109  Data-Collection  110  3.2.1.1  Client's Initial Diet  110  3.2.1.2  Client's Demographic Data for Defining Nutrient Limits  113  3.2.2  Data-Analysis  113  3.2.3  Data-Evaluation  117  3.2.4  Diet-Planning  121  4. TESTING OF THE PROTOTYPICAL SYSTEM  128  4.1  Introduction  128  4.2  Diet-Planning Model's Premises and Assumptions  136  4.2.1  The First Premise and Related Assumptions  136  4.2.2 4.3  The Second Premise and Related Assumptions Observations of the Objective Function's Characteristics  139 140  4.3.1  Unconstrained Objective Function  140  4.3.2  Constrained Objective Function  141  4.3.2.1  First Term:  Shape of the Curve  4.3.2.2  First Term:  Penalty Coefficient w  4.3.2.2.1 4.3.2.2.2 4.3.2.2.3 4.3.2.3 4.3.2.4  141 i  Penalty Coefficient, w., Based on Amount Consumed Penalty Coefficient, , Based on Initial Consumption Penalty Coefficient, First Term: Algorithm  , Further Comments  145 146 158 175  Further Modifications of the  Second Term of the Objective Function  176 177  vii  .4.3.2.5  Second Term;  Shape of the Curve  4.3.2.6  Second Term:  Penalty Coefficients, w.. and P j  4.3.2.6  Second Term':  Further Modifications of the  Algorithm 4.4 5.  Summary and Comments on Testing the Algorithm  180 k  182 182 196  SUMMARY, RECOMMENDATIONS, AND CONCLUSIONS  200  5.1  200  Summary  5.1.1  The Thesis Goal  200  5.1.2  First Objective of the Thesis  200  5.1.2.1  Diet-Assessment Function  200  5.1.2.2  Diet-Planning Function....  201  Second Objective of the Thesis  202  5.1.3 5.2  Recommendations  203  5.2.1  First Recommendation  203  5.2.2  Second Recommendation  205  5.3  Conclusions  LIST OF REFERENCES  205 207  APPENDICES. A  Food-Item File  224  B  Abridged Food-Item File  236  C  Food-Composition File  238  D  Abridged Food-Composition File  284  E  Nutrient-Limits File  286  F  Abridged Nutrient-Limits File  301  G  Attribute Group Matrix  303  yiii  LIST OF TABLES 2.1 3.T 3.2  Annual subsistence diet for a moderately active adult male, calculated using linear programming methods  85  Excerpt from prototypical intake questionnaire - - dietary intake format  Ill  Excerpt from prototypical intake questionnaire - - client demographic data  114  3.3  Excerpt from the evaluation output used for system testing  3.4  Excerpt from a proposed evaluation output format  3.5  Excerpt from the diet-planning output format used for  .  117 118  system testing  121  4.1  Upper and lower nutrient constraints for a standard subject .  129  4.2  Standard Initial Diet 1 (SID-1) and Standard Initial Diet 2 2 (SID-2) Nutrient composition of SID-1 and SID-2, and the upper and lower nutrient constraints for a standard male subject  131  4.3 4.4 4.5  135  Revised diets developed from SID-1 and SID-2 using quadratic objective functions (Eqns. 4.8, 4.9)  148  Nutrient composition of SID-1 and SID-2, and the revised diets developed using quadratic objective functions (Eqns. 4.8, 4.9)  152  4.6  Average absolute deviation from i n i t i a l consumption levels for items in revised diets developed from SID-1 and SID-2 using quadratic objective functions (Eqns. 4.8, 4.9) 155  4.7  Revised diets developed from SID-1 using quadratic objective functions (Eqns. 4.9, 4.12)  160  Revised diets developed from SID-2 using quadratic functions (Eqns. 4.8 - 4.12)  164  4.8 4.9  4.10  objective  Nutrient composition of SID-1, and the revised diets developed using quadratic objective functions (Eqns. 4.9.,4.12)  168  Nutrient composition of SID-2, and the revised diets developed using quadratic objective functions (Eqns. 4.8 4.12)  169  4.11  Average absolute deviation from i n i t i a l consumption levels for items in revised diets developed from SID-1 using quadratic objective functions (Eqns.; 4.9, 4.12) 172  4.12  Average absolute deviation from i n i t i a l consumption levels for items revised diets developed from SID-2 using quadratic objective functions (Eqns. 4.8 - 4.12)  173  Revised diets developed from SID-2 using quadratic objective functions (Eqns. 4.8, 4.26 - 4.28)  184  4.13 4.14 4.15  4.16  Nutrient composition of the SID-2, and the revised diets developed using quadratic functions (Eqns. 4.26 - 4.28)  ...  188  Average absolute deviation of item clusters'and of attribute groups in each hierarchical level from i n i t i a l levels, and the . penalty coefficients assigned for each hierarchical " level  190  Effect of penalty assignment on the number of attribute groups containing consumed items  192  LIST OF FIGURES 2.1  Individual variability in nutrient requirements  2.2  Proportion of the population having actual requirements above nutrient intake and the probability of deficiency in individuals ingesting a particular level of nutrients  2.3 3.1 4.1 4.2  4.3  68  ..  69  Theoretical model of the relationship of the nutrient intake of a population to the prevalence of deficiency  70  Overview flowchart of prototypical system for assessment and planning of individual's diets  109  Graph of quadratic function reflecting penalties for deviation from i n i t i a l consumption of food i  142  Graph of quadratic and linear functions reflecting penalties for deviation from i n i t i a l consumption of food i  144  Graph of quadratic function for deviation from i n i t i a l consumption of attribute group G -  180  xi  ACKNOWLEDGEMENTS  For those who have borne with me in my discovery that the shortest distance between two points is not a thesis, my sincerest thanks.  John H. Mil sum William T. Ziemba Donald B. Hausch Nancy E. Schwartz Melvin Lee Joseph Leichter Thomas J . Abernathy  CHAPTER 1 INTRODUCTION  1.1  Background and Need for the Study  1.1.1  Fundamental Tasks of Nutrition Education Programs Nutrition education, as a process to promote the public welfare,  is universally needed to ensure healthful food selection.  Since "there  is no instinct that guides man to select those foods which meet the nutritional needs of the body . . . each new generation must be taught what foods to select . . . " (Tddhunter 1969, p. 9).  Furthermore, i t is  needed because societal forces, which influence food availability, must parallel'iknowledge of nutritional well-being to ensure that appropriate products will be available for selection. The American Dietetic Association (1973), in "its "Position Paper on Nutrition Education for the Public", defines nutrition education as " . . . the process by which beliefs, attitudes, environmental influences, and understandings about food lead to practices that are scientifically sound, practical, and consistent  with,.individual  needs and available  food resources" (p. 429). As the above definition implies, the fundamental tasks of nutrition education personnel are two-fold.  First, nutrition educators provide a  code of nutritional practice through the judicious interpretation of scientific studies, and strive to reduce the time lag between the discovery of nutrition knowledge and its application to food practice (Gifft et a l . 1972; Leverton 1974).  Second, nutrition educators  2  communicate available knowledge to the public sector with awareness of the variety of audiences - individual citizen, "activated" consumer, governmental bodies, and industry - that may be approached to ultimately effect changes in food practices by their influence on either available food supply or food selection behaviours (American Dietetic Association 1973; Leverton 1974).  Also, in communicating with the public sector, the  nutrition educator must be aware of the complex dynamics of communication with target audiences that potentiate or deter change in food habits.  1.1.2  Criticisms of Nutrition Education Programs Man's knowledge of food and its implication for well-being has  evolved through the course of centuries.  Traditional knowledge of food,  previously based on " . . . a long series of trials and errors - - sometimes mortal errors - - . . . " (Mayer 1973, p. xxi) has been greatly extended by the systematic investigation of the present day.  The new scientific-  technologic tradition has produced an explosion in knowledge of dietary consequences for health, and with this a growing suspicion that this knowledge is not being applied to its full  potential.  In 1941 nutrition-related medical problems were shown to be affecting much of the potential American military manpower.  In that year  ...John R. Murlin, participating in a White House Conference on Nutrition and Defense, stated: "Knowledge is like a rock set upon a shelf. It does no harm, and it does no good, so long as i t rests there, but let somebody jar the shelf and let the rock fall off, and then something happens. That is the state of things today. We know more than we are doing." (Hill 1969, p. 14) Similar sentiments have been echoed to the present day.  Briggs (1969)  suggests that "advances in nutrition-related research are only useful when extensively applied . . . but the link between this knowledge and its  3  application is extremely weak" (p. 8).  Concerning the state of knowledge  in the science of nutrition, White (1976) states that "obviously much more is known about nutrition and human needs than is manifest by the current practices of our population" (p. 54). Current criticisms on the gap between nutritional knowledge and its apparent application, stem from western population studies (U. S. Department of Agriculture 1968; U. S. Department of Health, Education, and Welfare 1972; Canada 1973) identifying significant'incidences of nutritionrelated problems, and in some, a trend for decline in dietary quality as compared to previous decade's.  Also, there have been great increases in  mortality from diseases with dietary involvement, which appear to be largely preventable.  While health and medical expenditures have continued  to rise, no strong evidence exists to indicate that any corresponding benefits have occurred in the major health indices.  For example in the  United States, l i f e expectancy of those over twenty has not improved in the last twenty-five years, a period when expenditures have increased eightfold (Cornely 1974).  Mayer (1975) aptly comments, "like Alice in  Wonderland, we are running faster and faster to stay in the same place" (p. I D Briggs (1969) suggests that " . . . we do have sufficient  scientific  knowledge to conduct sound programs of nutrition education . . . " (p. 8). If so, then what are the reasons for the discrepancy between this knowledge and population food practices?  The reasons for this discrepancy may be  categorized in two primary areas which coincide with nutrition education's aforementioned responsibilities, namely, the problem of developing a scientific-nutritional code, and the complexity of translating this code to the public sector.  4  1.1.2.1  Developing a Scientific - Nutritional Code  As indicated in "A New Perspective on the Health of Canadians" (Lalonde 1974): The spirit of enquiry and skepticism, and particularly the Scientific Method, so essential to research, are, however, a problem in health promotion. The reason for this is that science is full of "ifs',' "buts", and maybes" while messages designed to influence the public must be loud, clear and unequivocal (p. 57). Indeed, uncertainty in scientific circles is often reflected in ambiguous statements to the public on which course of action to follow.  As Mayer  (1975) indicates, "we have been telling people what they should eat but avoiding strong statements about what they ought not to eat or to eat less of.  We 'stay away from controversial subjects'  . . . " (p. 8).  Unresolved  scientific issues would not matter provided they stayed in academic circles.  However, the environment in which sound nutrition practices are  promoted is competitive and often hostile.  Misinformation, often with a  seed in scientific debate, is part of this environment - - a fact which vacillation on issues by the scientific community does nothing to alleviate. In fact, vacillation may ultimately breed distrust, by a significant proportion of the public, in science and technology generally, and specifically in the quality of the food supply and in the potential of nutritionists to reflect meaningfully on these issues (Leverton 1974). When information becomes public knowledge, regardless of whether scientists believe the evidence is sufficient for a conclusion, the public begins to act through the voices and opinions of media.  Mayer (1972)  recommends "given this situation, i t is better to act on the conjecture of scientists than on the guesses of newspapermen" (p. 240).  Otherwise,  "as the nutritionally trained people contemplate their disagreements on approaches to nutrition, many persons not trained in nutrition are making  5  decisions on what foods are available in stores and in what form" (Ullrich 1973a, p. 184). As the complexity and scope of nutrition knowledge increases, and therefore the task of interpretation appears more d i f f i c u l t , it becomes particularly important that comprehensive and coherent recommendations on dietary practices be provided.  The failure of nutrition education to meet  fully its obligations to provide a code of appropriate practice is suggested by criticisms (Robinson 1976) indicating the absence of clear recommendations for action on controversial aspects of diet, and on newer findings in dietary relationships for health.  1.1.2.2  Communicating a Nutritional Code to the Public  The ultimate recipient of nutrition information is the public and i t is their dietary well-being which indicate the success of nutrition education activities.  The delivery of nutrition information to the public  is not the exclusive domain of the nutrition educator, nor is the medium of exchange restricted to "bookish ministrations".  In order to  effectively  communicate nutrition information "nutrition and health educators must be concerned with human behaviour and therefore must compete with all internal and external forces that define and control how an individual behaves" (White 1976, p. 54).  Criticisms, discussed below, suggest that nutrition  educational programs have failed to keep pace with the changing communication mileau of nutrition information in western societies, both in terms of the variety of audiences that must be approached to ultimately influence the dietary practices of the public, and also the complex dynamics of communication with the public in an informationally-competitive society.  6  1.1.2.2.1  Communication with a Variety of Audiences  Food habits of individuals are determined by a number of factors including the foods available in the marketplace, and information disseminated:.in the media - - both factors which have been influenced massively by food technologists, manufacturers, advertisers, and legislators.  Thus,  in addition to communicating a nutritional code directly to the public through schools, hospitals, and media, nutrition educators must recognize audiences in both industry and government through which the public is influenced, and through which nutrition information can indirectly reach the public (Anon 1972).  For example, dietary lifestyle can be influenced  through food processing and manufacturing regulations, media policies and advertising guidelines. As Ullrich (1974a) indicates, "objective nutrition education should not be the exclusive responsibility of any one group but a balance among government agencies, food industry, and educational institutions" (p. 84). However, in efforts to promote sound nutrition, as Gussow (1972) observes, " . . . most professionals were much more worried about the excesses of the 'health food' stores than the excesses of what at least one observer has called the 'unhealth food stores'  [and] until very recently, this  misplaced concern extended even to advertising" (p. 48 and 49).  In attempts  to combat misinformation, the colourful, but perhaps insignificant  antics  of the charlatan have attracted the greatest attention of nutritionists. For example, recent popular food movements, the "health foodists" as coined by Wolff (1973), may be allies and not enemies of nutrition education. Professional emphasis on such groups may have resulted in nutritionists ignoring other, perhaps more significant, forces which influence food habits and which may be important factors promoting misinformation.  Further,  7  Hall (1975) suggests that nutritional science is outmoded in its ability to deal with the modern world of nutrition.  Nutritionists " . . . practice  a science, outmoded by technological reality, [thus] they can not influence ...  [the] political process" (Hall 1974 p. 9).  1.1.2.2.2  Complex Dynamics of Communication with the Public  Beyond the influence nutrition educators can have on the public's dietary practices via intermediary bodies, such as government and industry, the nutrition educator can affect public consumption patterns by directly providing information to individuals. The food choices of individuals are determined by a complex of internal and external forces that define and control behaviour (Gifft et a l . 1972).  Among the external or environmental forces influencing food  behaviour are the foods available, mass media, cultural tradition, and governmental policy.  Thus, in providing information directly to indivi-  duals, consideration must be given to the many competing forces that influence food practices - - forces which define the complexity of the informational environment in which nutrition educators must function.  With respect to  two environmental influences - - foods available in the marketplace and diet-related discussions generated by mass media and advertising - criticisms suggest that nutrition education programs have been inadequate competitors.  1.1.2.2.2.1  Changing Food Market  Food technology, industrial development, and rapid transportation have greatly increased the number, kinds, and availability of food products (Todhunter 1969).  Whereas in 1928 the average supermarket  8  contained around 900 items, large supermarkets now carry in excess of 10,000 items (Gussow 1972).  Thus, "today, more than ever, there are more  opportunities to make poor food choices because of the broad array of new foods available" (Anon 1972, p. 34).  " . . . The shopper has the difficult  problem in properly selecting the best buys in both nutritional and monetary value" (Todhunter 1969, p. 9).  As Ullrich (1975) states, "the  technology of providing a large variety of foodstuffs  in the marketplace  has far outstripped the knowledge of the consumer to make wise choices" (p. 48). The nature of the food supply has been virtually transformed as a result of technological development.  "In 1941, only 10 percent of our  foods were highly processed; today, that amount has risen to 50 percent" (Mayer 1972, p. 239).  In addition to the alterations that processing  may cause in nutrient characteristics and distribution of nutrients in foods, approximately 1,830 additives are available for routine use in foods (Hall 1973).  Many of these substances are either new in the diet  or are present in greater proportion than previously. With the change in food supply has come a change in food habits as illustrated in the following statements by Ullrich (1974b): A look at the present state of our national food consumption compared with 25 or 50 years ago shows a decline in the consumption of foods of high nutrient quality in relation to consumption of foods of low nutrient quality. In part, the decline may be due to the overzealous development and heavy advertising of "fabricated" food products that are not equal to the conventional foods they replace (p. 4). There is concern in nutritional circles that consumption of products of low nutrient quality coupled with current patterns of low energy expenditure may lead to nutritional problems (Harper 1974).  Mertz (1972)  suggests that increased use of textured vegetable proteins and refined  9  grains may result in depletion of trace elements.  However, the Ten State  Survey (U. S. Department of Health, Education, and Welfare 11972) indicates in its "Highlights", " . . . inadequate information is available on the distribution of nutrients in today's food supply . . . " (p. 12).  The long  term effect of fabricated diet consumption is not known. Additionally, the consequence of long term consumption of many additives and particularly r  of combinations of additives is not known. To be effective,  nutrition education programs must keep pace with  the changing food supply, and with the consequent influence of food supply on behaviour and nutritional well-being.  Regardless of the cause  of the present transformation in foods available, or of the difficulty of interpreting the consequences of these changes, i t is the responsibility of the nutrition education profession to provide rational and useful guidelines for the public, industry, and government - - guidelines that are relevant to the present type and variety of foods available.  1.1.2.2.2.2  Mass Media and Advertising  As Gussow (1972) points out, " . . . between 1928 and 1968 people had learned to eat thousands of new food items" (p. 48).  If people, as  nutritionists say, have food habits which once established are difficult to change, then what change agent has been effective?  Her implication is  that advertising is a significant factor, and particularly advertising on television.  Manoff (1973), who has also publicized the importance of  media and advertising as a nutrition education (or miseducation) force, indicates that, "of the 6,000 to 8,000 items on sale in American food stores, 50 percent of them did not exist 20 years ago.  This would not  have been possible without commercial television, which began roughly  10  20 years ago" (p. 126).  Presumably the Canadian situation is similar.  These factors are of particular significance to the nutritionist since the massive influence of advertising is motivated by marketing forces and not by a concern for nutritional value.  As Manoff (1973)  suggests, "food manufacturers produce anything they can sell at a profit. This is the elementary law of marketing" (p. 128).  Although the food  industry claims to act in the consumer's interest by supplying consumer demands, millions of dollars are spent each year on advertising to influence consumer purchase of foods hav.ing high profit, but often low nutritive value (Gussow 1972; Manoff 1973 ).  Mayer (1975) further questions  the motives of the food industry, stating, " . . . the opposition to the proposed Consumer Protection Agency in [the United States] is being led by the large food companies. (He adds,]]  I leave you to draw your own con-  clusion on what this may represent" (p. 11).  Manoff (1975), paraphrasing  and quoting Mayer, suggests: . . . the reasons that food manufacturers appear reluctant to concentrate on foods of higher nutritional value . . . is that there is no insistent demand for such foods. "Until the American mentality changes", he said, "food manufacturers will feel no strong injunction to provide such products", (p. 139). Although mass media are a significant force in molding population food practices, " . . . the mass media has been virtually abandoned educators  by nutrition  to the commercial food marketer and his nutrition education . . . "  (Manoff 1973, p. 125). If consumer mentality is to change, "legitimate" nutrition education, in addition to the traditional nutrition education vehicles of school, c l i n i c , and hospital, must make effective use of media and advertising techniques to match the efforts of the food industry (Anon 1972; Manoff 1973; Manoff 1975).  Additionally, the formidable force of advertising in  11  molding present attitudes and practices, establishing relevant public nutrition issues, and influencing the nature of the food supply, must be recognized when establishing nutrition education programs.  1.1.3  Summary and Conclusions Since the public is the ultimate user of nutritional information, i t  is the task of the nutrition education profession to provide the consumer with comprehensive food selection guidelines; and to influence which foods and information are available through activity in.government, the food :  industry, educational services, and the marketplace.  Nutrition education  programs in western countries have been criticized for not keeping pace with the complex changes in food available in the marketplace, and the complex discussion within the community on dietary issues influencing health, whether of legitimate or artifactual origin. If the above difficulties plague the nutrition education profession in its efforts to guide nutritional practice within society, then they also plague individual citizens when nutrition education programs are not effective.  That i s , when these programs cannot properly provide the neces-  sary information to the public, the individual must become the sole arbiter of nutrition information.  Thus, the individual has the confusing tasks of  resolving the complex dietary issues; of interpreting the worth of conflicting recommendations and admonitions from nutritional, medical, media, and marketing sources; of coordinating food selection practice in the face of a vast number of dietary provisos; and of selecting items from an overwhelmingly complex food supply - - in short, an information overload.  As  stated by Mayer, and quoted by Ullrich (1973b), concerning the nutritional literacy of the American people:  " . . . we have a very tired mind subjected  12  day in and day out to a tremendous amount of information which is mostly misinformation by people who have something to sell" (p. 224). Programs providing nutrition education services directly to the public must consider the complex forces influencing food decisions of individuals and families - - forces which nutrition education can not presently buffer through means other than public education.  Unfortunately,  the tendency has been to believe that individuals cannot handle complex information on foods, and therefore will benefit more with simplified guidelines on food selection practice.  In so doing, nutrition educators  leave the consumer as prey to the whims of industry and advertising. Oversimplified nutrition information does not properly equip the consumer for the present world of nutrition information. Although i t is not necessary that each individual be a nutrition specialist, she or he should be offered the opportunity to confidently eat as one.  It is the responsibility of the nutrition education profession  to provide the consumer with the resources necessary to collate available information on foods and nutrition, so that rational choices can be made. This includes useful guidelines which are comprehensive in coverage of recognized and proposed dietary issues, and relevant to the present variety and types of food available in the marketplace.  Proposals by leading  authors in nutrition indicate a trend to more comprehensive and detailed information.  For example, Mayer (1975) states, "in order to teach nutrition  to a broader audience, we obviously have to embrace a much broader concept of nutrition than talking about nutrients and foods which existed 25 years ago" (p. 8).  13  1.2  Thesis Goal and Objectives  1.2.1  Thesis Goal The overall goal of the thesis was to develop a prototypical system  which provides information on dietary practices for individuals who: (i) (ii)  want to apply nutritional principles to their eating habits; and have sufficient resources (eg. time, energy, education, money) to use the information which defines healthful dietary practices for them.  This development was undertaken bearing in mind: (i)  The complexity of developing a scientific-nutritional code due to the many unresolved dietary issues under investigations, and the consequent problem of translating available knowledge for health promotion at the community level,  (ii)  The complex factors influencing communication with individuals which the nutrition educator must consider - - that i s , an environment with an  overwhelming array: of foods to choose from,  for an equally overwhelming number of reasons - - and the unsatisfactory resolution of this condition by individual exposure to contradictory information provided by media sources.  1.2.2  Thesis Objectives  1.2.2.1  The First Objective of the Thesis  The first objective, arising from the thesis goal, was to develop a prototypical computerized system with two major functions, namely:  14  (i)  diet-assessment in order to appraise the dietary intake of individuals, and  (ii)  diet-planning in order to recommend modifications in food intake for those individuals with diets which do not meet specified nutrient limits.  Specific characteristics of the system are defined in "System Design and Characteristics" (p.109").  1.2.2.2  The Second  Objective of the Thesis  The second objective, arising from the thesis goal, was to test the diet-planning component of the prototypical system.  This testing explored  some of the conceptual assumptions of a model designed for modifying diets which do not meet specified nutrient limits.  The results of this work is  reported in "Testing of the Prototypical System" (p.128).  15  CHAPTER 2 REVIEW OF LITERATURE  2.1  Definition of an Adequate Diet The criteria used for appraising diets are ultimately derived from  a knowledge of the dynamics of these diets in human populations.  Presently,  the criteria for appraising diets of normal individuals are defined in the dietary standard (Passmore e_t aVi 1974; United States 1974; Canada 1975) - - in a sense, a compendium of known nutrient requirements used in the evaluation and design of diets. Any discussion about providing information on' food selection practice should consider the theoretical foundation of the dietary standard, and more broadly, the basis for defining an adequate diet.  Thus, this  discussion begins with a cursory examination of the influence of diet on the human individual and on the society in which the individual lives.  2.2'il  Influences of Diet in Human Populations  2.1.1.1  Influences on the Individual  Diet has many influences on the human organism.  Both human physiology  and perception are influenced by dieUs nutrient and non-nutrient constituents, and its characteristics conferred by human imagination and cultural beliefs. With respect to nutrient constituents, more than forty are presently recognized as essential  for normal body functions of growth, maintenance,  and repair (United States 1975).  Deficiencies of one or more of these  nutrients results in a plethora of physiologic and psychologic symptoma-  16  tology, touching all systems of the body (Goodhart and Shils 1973; Pike and Brown 1975). symptomatology.  Similarly, overnutrition is also associated with disease For example, a nutritional component has been suggested  for many of the degenerative diseases (Canada 1976a), and for such ubiquitous problems as dental disease (McBean and Speckmann 1974).  In -  addition to nutrient involvement in the causation of disease, nutrients may alter physical and mental potentials within the vaguely defined limits of normal well-being.  For example, a special athletic dietary regimen,  called the glycogen loading diet, maximizes available energy for performance (Astrand and Rodahl 1970; Astrand 1973).  As a further example,  Davis and Williams (1976) suggest that diet may influence such factors as sleeping pattern and healing time. Diet can also be a vehicle for an immense array of potentially harmful or beneficial non-nutritive substances.  Toxic food substances are found o  in the environment naturally (United States 1973) - - for example, l.athyrism and aflatoxin - - and as a biproduct of modern technology which has introduced many new chemicals into the environment and into foods (Hall 1973; Hall 1977; W. H. 0. 1978).  Beneficial non-nutritive substances found in  food include the active anti-infective property called lactobifidus factor found in human milk (JeTTiffe and J e l l i f f e 1971), and dietary fiber (Burkitt and Painter 1974; Klevay 1974; Spiller and Amen 1975).  Further,  food includes substances which provide taste, color, texture, smell, and other sensations.  These influence the desirability of items for consumption,  and the nature of the eating experience (Kinder 1973).. Diet in some cultures is believed to have properties beyond the physical.  For example, the idea that the "heat" or flesh of a brave enemy  or animal conferred courage has existed in cultures from as early as the  17  Stone Age (Lowenberg et_ al_. 1974).  Additionally, beliefs based on hygenic  observations, and taboos not so empirically founded, exist inrmost cultures (Lowenberg ejt a]_. 1974).  Even i f these beliefs have no basis in fact,  their influence may be sufficient to confer either benefits or detriments; in any case they have consequences for the population even i f a physiological effect is absent.  2.1.1.2  Socio-Economic Influences of Diet  Diet has implications beyond its physiologic and psychologic  influences  on the individual, extending to those of a social, economic, and political nature.  Authors who have addressed these issues include:  Berg (1973),  who examines the role of nutrition in national development; Correa (1975), who discusses the influence of nutrition on socio-economic development; and Mitchell (1975) with a consideration of the dynamics of food production and supply in context of Canadian economic policy.  2.1.2  Criteria of an Adequate Diet As indicated, nutritive and non-nutritive elements of the diet have  many influences on the individual and on the society in which the individual lives.  Thus, to properly identify an optimal or ideal diet potentially  requires consideration of this multitude of dietary agents and their consequences.  In turn, to identify desirable consequences may require con-  sideration of the objectives and aspirations of individuals and their society. "A major objective of the Food and Nutrition Board of the NRC continues to be to encourage the development of food use practices by the population of the United States that will allow for maximum dividends in  18  the maintenance and promotion of health" (United States 1974, p. 1 ); where "health is defined according to the World Health Organization, as 'a state of complete physical, mental and social well-being and not merely 1  the absence of disease or infirmity" (United States 1974, p. 1 ). However, even when collective recognition of the desirability of healthful food practices exists, as exemplified by the objectives of a recognized body such as the Food and Nurtition Board, defining the optimal diet may s t i l l be difficult - - quite apart from the problems of defining health alone.  For example, i t may be that not all apparently desirable  dietary practices are mutually compatible.  "It is not at all certain . . .  that the nutritional requirements for maximum size, early maturity, active sex l i f e , and maximum muscular development are identical with those for maximum longevity" (Goodhart 1973, p 403 ), resistance to chronic diseases, and maximum physical and mental performance in old age (Harper 1974). Furthermore, "diets designed to protect the individual against bacterial infections, such as Tuberculosis, may lower resistance to certain viral infections and predispose the individual to obesity and coronary heart disease in later years" (Goodhart 1973, p.403 ).  Thus, as Goodhart (1973)  summarizes, " . . . statements such as 'An adequate diet is one which meets in full all the nutritional needs of the person have l i t t l e meaning unless they can be interpreted in terms of either the person's ambitions for himself or the community's designs for or on him" (p.403 ). Another factor that may make definition of the ideal diet difficult is conflict between dietary objectives and other objectives of an individual or a society.  For example, bottle feeding of infants is becoming socially  fashionable in third world societies, but economic and technologic constraints often make this practice nutritionally unsound (Jelliffe 1973).  19  However, increased well-being may conceivably result in some instances when apparently less than ideal diets are used in favour of meeting other personal or social priorities, as for example, that moderate alcohol consumption increases well-being, with respect to coronary risk (Yano et a l . 1977). In summary, the optimal diet - - the best dietary option described by environmental, biological, societal and personal circumstances - - is the diet which most effectively  contributes to the achievement of the goals and  aspirations of the society and of the individuals in the society.  Its  definition involves, in addition to identification of the many relevant consequences diet has for societal  and individual realization, choosing  among a variety of apparently equivalent but mutually incompatible dietary objectives, which in turn conflict with other personal and societal aspirations.  2.1.3  Nutrient Requirements Although complex c r i t e r i a , such as those discussed above, are relevant  for defining an optimal diet, at present a more modest formulation of desirable dietary practices has been defined.  This is the "adequate diet",  as described by Goodhart (1973), which is based on human metabolic requirements for essential nutrients. Nutrient dynamics in the body can be envisioned as a contiuum from the extremes of depletion to repletion, presumably with a circumscribed optimum between (Arroyave 1971).  Ideally, to monitor these changes,  criteria are required which differentiate the metabolic continuum from these extremes, and which identify the optimum for an individual in any particular situation for any nutrient.  In the absence of suitable methods  20  and criteria to identify a "nutrient optimum", the establishment of a requirement for a nutrient, defined by the Food and Nutrition Board (United States 1974) as " . . . the minimum intake that will maintain normal function and health" (p. 8 ) , " . . . rests on the production of a deficiency? and on the definition of that daily intake which prevents or cures the deficiency state" (Mertz 1972, p .19).  Criteria to determine these requirements  are based on sensitive measures of biological response to nutrient depletion, such as, gross clinical symptomatology, and metabolic, biochemical, or other biophysical-functional changes (Arroyave 1971). The known nutrient requirements are largely outlined in the dietary standards of various countries and international agencies, such as:  the  "Dietary Standards for Canada" (Canada 1975), the "Recommended Dietary Allowances" (United States 1974) in America, and the "Handbook on Human Nutritional Requirements" (Passmore et_ al_. 1974) by the World Health Organization.  The dietary standards, will be discussed more fully in  "Data Evaluation" (p. 62).  As discussed later (p. 74), other dietary  goals beyond those presented in the dietary standard have been proposed.  0.  Deficiency is defined as "...an habitual intake by the individual below his own true requirement. The manifestations of such 'deficiency' will depend upon the criteria by which the requirement has been defined . . . " (Beaton 1972, p.357).  21  2.2  Relevant Systems  2.2.1  Food Guides  Although there are many different approaches to teaching nutrition (Ahlstromand Rasanen 1973), food guides have been the standard model for outlining the adequate diet in nutrition education (Chandler and Perloff 1975; Winarski 1976).  Examples of food guides include:  Canada's Food Guide  (Canada 1977) the Basic Four and Basic Seven (U. S. D. A. 1976, U. S. D. A. 1971), and the Type A Pattern for school lunches (Head ejt al_. 1973).  Food  exchange l i s t s (Caso 1950)are examples of guides developed for use in planning therapeutic diets. The food guide is intended as a simple and reliable nutrition education device for teaching the principles of healthful food selection.  The food  guide is a translation of nutrient requirements into a guideline of suggested serving of items from a few basic groups of foods with roughly equivalent nutrient composition (Ahlstrom and Rasanen 1973; Hertzler and Anderson 1974).  Foods are classified into basic groups according to both  their major nutrient contributions to the diet, and  other criteria which  assure the guides' applicability for the population and its u t i l i t y as a teaching tool.  These other criteria include:  the food's function in the  meal system of the population; foods available; educational status, income, and lifestyle of the population; and local nutritional needs.  (Hayes et a l .  1955; Lachance 1972; Ahlstrom and Rasanen 1973; Hertzler and Anderson 1974; United States 1974). The problem in the design of food guides is to strike the balance between simplicity - - so that the information can be understood and remembered - - and validity - - so that the guide accurately reflects the dietary standards.  To this end adaptations and modifications have been  22  made to meet the needs of different populations and for different purposes, but the basic model of a limited number of natural food groups has remained (Hertzler and Anderson 1974). Even though the u t i l i t y of the food guide is recognized, i t has been criticized as a teaching tool for a number of reasons, including:  its  f a l l i b l i t y in guiding proper food selection practices; its inapplicability to the present food supply; its lack of relevance for contemporary nutrition problems of the propulation; and its ineffectiveness as a teaching tool in an informationally-competitive society.  Each of these criticisms will be  dealt:; with below.  2.2.1.1  F a l l i b i l i t y in Guiding Food Selection Practices  The validity of the simple plan, based on four or five groups, which is used i.hothe United States and Canada for directing food selection practice has been questioned (Anon 1972).  Even for populations whose preferred  style of eating is reflected by the food guide, it is recognized that the guide is f a l l i b l e .  For example, i t is known that the food guide system  can be invalidated by consistently making poorer choices in a food group, skimping on serving sizes, or using improper cooking and preparation practices (Bogert et al_. 1973).  Also, the required allowance of nutrients  can be obtained from a wide variety of food combinations and patterns, besides those of the food guide (United States 1974). A simplified eating plan, as in the food guide, is based on two basic assumptions:  f i r s t , only a few key nutrients need to be monitored out of  forty-plus possible known and unknown nutrients; and second, the nutrients  23  not being monitored inevitably occur in the variety of foods selected for the key or index nutrient (Bogert et al_. 1973).  To test these assumptions  Pennington (1976) evaluated two diets selected by Page and Phipart (1957) in accordance with the "Basic Four" plan, against the  1973 Recommended  Dietary Allowances of the Food and Nutrition Board (United States 1974). Although the four index nutrients (protein, calcium, vitamin A, and vitamin C) were acceptable, other nutrients (thiamin, riboflavin, niacin and iron) and energy were below the recommended levels.  Pennington (1976)  comments as follows: Whether or not the remaining essential this plan) are met If "empty calorie" adequate nutrients  deficit for these 4 nutrients or any of the nutrients (which total 45 and are ignored with depends on foods chosen to round out energy needs. foods are selected . . . the chance of getting is lessened (p. 4 ).  Further Pennington (1976) states that "the major problem with the Basic Four seems to be the nonuniformity of major and coincidental nutrients . . . within the groups" (p. 6 ).  Thus, the Basic Four food group concept inten-  tionally provides for only 4 nutrients out of a possible 50 or so.  The  observed nonuniformity of the index and other coincident nutrients in the Basic Four plan would indeed indicate potential f a l l i b i l i t y of the plan. In a more recent evaluation, King e_t al_. (1978) compared the nutrient content of 20 published menus based on the Basic Four Food Guide with the 1974 Recommended Dietary Allowance (RDA) for an adult reference male.  The  Basic Four Foods met or exceeded the RDA's for only 8 of the 17 evaluated nutrients.  For 5 of these nutrients, menus supplied 60 percent or less of  the standard.  A modification of the Basic Four Food Guide for adults was  suggested, in order to improve the adequacy of menus developed from this guide. Two studies of food guides were also undertaken in Canada.  In the  first Canadian study, Milne e_t al_. (1963) compared nutrient intakes to  24  patterns of food usage for adolescents.  They found teenagers may consume  nutrients in recommended amounts without ingesting foods in the amounts suggested in Canada's Food Guide.  Similarly, McClinton et al.''(1971),  in the second Canadian food use study, indicated a wide divergence between patterns of foods recommended in the food guide, and those patterns of foods selected by the study population.  Out of 1,418 people who met the recom-  mended allowance for nutrient intake, only one person included foods as recommended in the Canada Food Guide. the findings^of Pennington (1976).  These results agree in concept with  On the other hand, these findings do  not necessarily evaluate the educational effect of the food guide.  The  fact that the population sampled had adequate nutrient intakes may reflect favourably on the educational efficacy of the food guide, i f in fact the food guide had served as a basis of instruction.  2.2.1.2  Inapplicability to the Present Food Supply  Present food guides have also been criticized as inapplicable to present food supplies because of the trend, in western society, towards the use of highly refined fabricated foods, and of nutrient supplements (Bogert et a_]_. ,1973; Hertzler and Anderson 1974; Fremes and Sabry 1976). The development of an effective food guide requires that the food supply, and consequent pattern of food usage, have a reasonably consistent nutrient distribution which allows translation of the foods into a limited number of food groups.  Many of the synthetic and fabricated foods may be concentrated  sources of one or more nutrients which reflect those in a food group but do not supply sufficient amounts of other nutrients.  Thus, although the  food guide is presumed to give reasonably sound information on a diet selected from a limited number of natural food groups, many of the manu-  25  factured items presently available which resemble natural foods and may in fact replace items in the food plan, do not compare favourable in nutrient content (Bogert e_t al_.1973; Hertzler and Anderson 1974; Fremes and Sabry 1976).  2.2.1.3  Relevance to Contemporary Nutrition Problems  Food guides have been criticized for their lack of relevance to the present nutrition problems of the population.  A food guide is a tool to  improve eating habits and consequently i t should indicate where improvement is needed.  Food guides, although revised occasionally, were developed  at a time when nutrition knowledge was not as extensive as today, and when emphasis on dietary needs was different (Hertzler and Anderson 1973).  For  example, Canada's Food Guide was originally established in 1942, at a time when the needs of national defense dictated a policy of food allocation and rationing practices.  However, as suggested by Fremes and Sabry (1976),  the nutritional problems of today are not ones of shortage, as they were during thai war years, but problems of abundance. Bogert et a]_. (1973) states that in developing a food guide the goal is to " . . . add to the scientific basis of the Four Food Groups, making finer distinctions and discriminations among the group alternatives based on newer knowledge of nutrients and the foods in which they occur" (p. 442). Examples of this evolution exist (Hertzler and Anderson 1974).  For example,  embellishments to meet particular needs have been incorporated in food guides, such as indications of yitamin D requirements, and provisos for the use of iodized salts.  Similarly, subclassification of the fruit and vege-  table group has been used to discriminate foods rich in vitamin A and C. Modern examples of food guides for use by vegetarians (MacMillam.'and Smith  26  1975; Smith 1975) and for coronary prevention (Conner 1967; Jansen et a l . 1975) are available on a limited basis.  However, these guides must  necessarily incorporate additional food groups or qualifying remarks to direct food choices.  2.2.1.4  Ineffectiveness as a Teaching Tool  To clarify the concept of healthful food selection and to facilitate learning, the need for maximum simplicity in food guide design has been stressed.  Simplicity has been obtained by limiting the number of food  groups and by using familiar names of foods (Hertzler and Anderson 1974). However, the trend in North America to attain maximum simplicity may not be as practical today as in earlier years, consequently, some authors have questioned the present food guide approach as the exclusive model for teaching food selection.  (Hayes et_ aj_. 1955; Anon 1972; Lachance 1972;  Poolton 1972; Manoff 1975). The media and advertising in the current era of food fads and fad diets have made the public aware of the basic nutrients, and a host of other food-associated  claims (Ullrich 1971).  Simple food-grouping systems may not  be effective as teaching tools because the increased impact of media on nutritional awareness has established issues to which nutritional instruction by food guides does not adequately respond (Anon 1972; Lachance 1972; Manoff 1975). Another criticism of the food guide as a teaching tool has been the effect on interest in nutrition observed when i t has been used extensively and repeatedly through several years of nutrition education (Poolton 1972; Bogert et_ al_. 1973; Leverton 1974).  Food guides do not adequately hold  Peoples' interest and consequently their ability to learn suffers.  As  Lachance (1972) stresses, " . . . teaching nutrition by food groups is like  27  teaching mathematics by astronomy.  It can be done, but it's not exactly  the optimal method" (p. 44)  2.2.2  Other Relevant Systems The food guide is not the only method for presenting the principles  of appropriate food selection.  As indicated there are reasons that a more  elaborate teaching tool should be considered to replace or supplement the-present food guide.  The sacrifice of the food guide's simplicity may  be substantially offset by benefits in both teaching effectiveness and accuracy of nutritional information.  Such tools include:  food labelling  (Anon 1973), the Dietary Nutrient Guide (Pennington 1976), and a variety of computerized dietary assessment programs (Hanson 1969; Hansen 1973; Johnson et al_. 1974; Eddison 1975; Action B. C. 1976; Kugler 1976).  These tools  provide a .nutrient-emphasis rather than the more traditional food-emphasis.  2.2.2.1  Pennington's Dietary Nutrient Guide  Pennington (1976) has developed a "Dietary Nutrient Guide" for evaluating diets which utilizes a limited number of "index" nutrients to monitor the total nutrient contribution of diets. index nutrients ensures adequacy of other essential few other guidelines are followed.  Adequacy of the seven nutrients provided a  The concurrence of index and other  nutrients in foods has been determined by using extensive statistical analysis and computer procedures. This approach does circumvent the inadequacies of the traditional food guide by ensuring adequacy for 45 nutrients rather than four.  Pennington  has been able to avoid the usual pitfalls encountered when the dietary standard is translated into food guides or other patterns for desirable  28  eating - - that i s , to deliberately or inadvertently ignore many essential nutrients in order to simplify the computation required in order to group foods. It is interesting to note that the index nutrients identified by Pennington, as indicative of dietary quality in conventional foods, are not those customarily recognized as key nutrients in food guides and which presently appear as key nutrients in nutrition labelling.  Pennington's  seven index nutrients are vitamin B-6, magnesium, pantothenic acid, vitamin A, folacin, iron, and calcium.  2.2.2.2  Nutrition Labelling  Nutrition labelling, as defined in the 1973- U. S. Code of Federal Regulations for food label information panels (Anon 1973), outlines conditions for voluntary and mandatory declaration of nutrient composition information on food products.  While nutrition labelling is not solely  and specifically designed as a nutrition education device, i t can be "exploited" for nutrition education purposes (Moore and Wendt 1973).  Thus,  nutrition labelling provides a reference standard for directly comparing the nutritive value of foods, and illustrates the major nutrient contributions of products whether or not they correspond to the food group concept. It also acts as a planning guide for balancing meals by providing clues to selecting food combinations which contain adequate amounts of key nutrients. Moore and Wendt 1973 suggest that nutrition labelling uses " . . . a technically sound vocabulary for describing the nutritive quality of foods . . . " (p. 123) - - that i s , objective nutrient information on specific food products - - rather than a " . . . contrived index of nutritive quality that sacrifices accuracy-for oversimplification" (p. 122). Although Moore and Wendt (1973) support the use of percentage of the  29:  U. S. RDA per serving or portion as a more practical and meaningful method of providing nutrient content information than either nutrients per calorie or nutrients per unit weight, other authors (Wittwer et al_. 1977) have promoted the use of nutrients per calorie or nutrient density.  Nutrient  density is expressed as an "Index of Nutritional Quality" which is the ratio of a food's nutrient contribution as a percentage of the nutrient allowance, to its caloric contribution as a percentage of the energy requirement.  2.2.2.3  An index value of 1.0 for a nutrient is the basic goal.  Computer Applications in Nutrition and Dietetics  During the previous decade, electronic data processing and system design have been extensively applied in food service systems management, hospital dietetic departments, dietetic education, and nutritional research centers to perform a variety of functions (Hoover 1976). Computers have been variously utilized:  in hospital dietetic  departments and nutrition-research centers to perform nutrient analysis on a variety of dietary data (Brisbane 1964; Hjortland et_ a}_. 1966; Schaum 1973); in diet- and menu-planning procedures which allow simultaneous satisfaction of nutritive, production, economic, and palatibility constraints (Smith 1963; Gelpi et al_. 1972; Balintfy 1976); in patientand hospital-information systems to support patient care activities such as menu ordering and production (Schaum and Sharp 1973); in obtaining dietary histories by automated-interactive  interviewing (Evans and  Gormican 1973); and in various other areas of nutrition research such as the study of eating patterns (Pao and Burke 1974).  Computer-assisted  ::  systems have also been used extensively in food services to provide automated control of resources (Andrews et_ al_. 1967; Tuthill and Moore 1974),  30 and in food production (Fleetwood et al.1974; Sager 1974). In addition to the industrial, c l i n i c a l , and research applications of computers in dietetics and nutrition mentioned above, systems utilizing some of the same functions are available to support nutrition education activities.  These systems include Dietronics (Hanson 1969), Nutrimetrics  07 (Eddison 1975), the Nutrient Adequacy Reporting System (NARS) (Johnson et a l . 1974), the Nutrition, Health and Activity Profile (Kugler 1976), the Nutrient Quality Index (Hansen 1973; Sorenson and Hansen 1975; Sorenson et al_. 1976; Wyse et al_. 1976; Wittwer et al.. 1977), and the Action B. C. nutrition evaluation program (Action B. C. 1976).  In addition to their  potential nutrition education function, some systems, in particular Dietronics and NARS, were designed as dietary-screening devices for clinical use. All of these programs are designed primarily to evaluate individual's diets, by providing information on the nutrient characteristics of the diet in relation to the estimated needs of the individual.  Nutrient  analysis and evaluation are performed on information from self-administered dietary questionnaires.  Food composition data is recorded as one (Action  B. C. 1976) or two-day recalls (Eddison 1975), as records of one (Johnson et a l . 1974) to seven days (Hanson 1969; Action B. C. 1976), or as usual consumption pattern (Hanson 1969; Kugler 1976).  Additionally, client  demographic data are collected on the questionnaire and used for estimating the client's nutrient needs. The various systems employ computer-generated outputs which provide information on nutrient consumption with a comparison of nutrient intake to anindividualized dietary standard for between one (Eddison 1975) and 25 nutrients (Hanson 1969; Kugler 1976).  This information is usually  31  displayed as actual nutrient intake and/or percentage of the nutrient allowance.  Additionally some programs include information on nutrient  ratios and percentages (Hanson 1969; Kugler 1976); the number of servings in the diet from each of the food groups (Hanson 1969; Johnson et_ a]J 1974) and other dietary measures such as sucrose consumption (Hanson 1969; Kugler 1976) and plaque frequency exposure (Hanson 1969).  NARS also provides an  overall measure of nutrient adequacy for the client, by averaging the percentage of the recommended allowance obtained over 12 nutrients. Although most systems provide tabular displays for the results of analysis and evaluation, the Nutrient Quality Index utilizes a graphical display. In some systems instructional material is appended to, or forms an integral part of the output, so that the results of the nutrient analysis and evaluation can be retranslated to an eating pattern.  Nutrimetrics 07,  which analyses the diet solely for caloric intake, prescribes caloric restriction where required to achieve ideal body weight by identifying and suggesting limitation of low-nutri nt density foods, such as alcoholic beverages and sweets.  The Action B. C. program output provides summary  statements on desirable items to include in the diet when a nutrient is limiting.  Information on nutrient functions and a recommended activity  program for weight loss is also available. The most elaborate output from the systems available is the Nutrition, Health and Activity Profile which proves a 9 to 13 page computer-generated response incorporating extensive discussion on diet and other aspects of lifestyle. output.  Another form of prescription is utilized in the Dietronics  Brand-name nutrient supplements are recommended to supply those  nutrients falling below the nutrient standards.  32  These examples illustrate some of the present uses of computer systems in dietetics and food services, and particularly in nutrition education. As Hoover (1976) indicates, in her review of computers in dietetics,  the  next decade may bring more extensive use of computers in present areas, and the exploration of further computer applications in dietetics and nutrition.  Present computer technology offers more capabilities than have  been presently exploited in nutrition.  While the computer is the tool  used to perform the operations required in the above systems, the operations have nothing to do with the computer per se.  However, the solution  of problems, and performance of tasks of realistic size cannot be practically attempted without the aid of high-speed computers. is a tool which can be effectively free processing is required.  The computer  utilized in situations where rapid error-  For example, the numerical manipulations,  report preparation, and other routine decision-making functions found in many areas of dietetics and food services make effective use of computer methods.  Within these limits the computer has demonstrated substantial  advantages in cost and time-saving over conventional methods.  2.2.3  Critique and Conclusions To develop a teaching tool or guide which is generally useful is not  an easy task'(UIlrich 1971).  As indicated in Section 1.1, the complexity  of present food markets, of dietary issues influencing health, of sources of nutrition and food related information, and of communicating to a public with a diversity of needs and levels of knowledge, must be considered in designing any tool providing information on food selection practice. This information provides a guideline for both the assessment of present practices and a goal for directing dietary modifications.  The acceptability  33  of any system for food selection instruction depends on the worth of these two functions - - assessment and prescription. The food guide has attempted, as Bogert et_ al_. (1973) suggests, to " . . . combine scientific knowledge with Western cultural wisdom and, faulted though it i s , there is not now an equally simple, equally workable alternative plan for Western food patterns" (p. 422).  On the other hand,  Pennington (1976) comments; A workable and usable food guide should be reliable and understandable but i t need not be overly simplified. The Basic Four, as presently used is oversimplified; and is certainly not fool proof' (p. 7). Unlike the food guide the other systems discussed have potential capabilities of monitoring most of the known nutrients in the diet, and in this sense are theoretically less f a l l i b l e guides for assessment of food selection practices.  The use of nutrients as the basis of assessment  provides an objective foundation on which to effectively  coordinate  nutritional information effectively with available food supplies and consumer habits.  But these systems have been criticized for this detailed  information allegedly exceeding the average consumer's comprehension (Moore and Wendt 1973).  Unfortunately, after elaborate analysis and evaluation of  the client's diet, these systems must then utilize less accurate methods to generate a food plan, such as the food guide or a listing of primary sources of nutrients. Neither of these concepts - - the food guide with its traditional food-emphasis nor nutrition labelling with its nutrient-emphasis - - are designed to stand alone (Moore and Wendt 1973; Winarski 1976).  They both  require the informational skills of a nutritionist/dietician to interpret nutrient information and to develop a dietary plan.  Presumably the  computer systems discussed also require some supplementary instruction  34  to ensure the development of a viable food plan.  However, unless the  professional has specific auxiliary tools available, the effectiveness of assessment and prescription will be limited by the capacities of the food guide and nutrition labelling. Although the nutrition educator should play an integral part in programs for instruction in food selection practices, the actual interpretation and translation of food and nutrient information to human diets may be a mathematical and technologic problem (Balintfy 1973; Balintfy 1976).  The logic of mathematical modelling, and the data handling  capability of computers may be required to resolve the conflict between the requirements for accuracy and for simplicity in the design of systems for food selection instruction. As indicated, several computerized systems for dietary analysis and evaluation are available for nutrition education.  These systems provide  useful feedback on the adequacy of present practices through accurate and comprehensive  assessment of individual diets.  Although no computerized  nutrition education tools which provide individualized diet or menu planning are presently available, illustrative models do exist in hospital food services and other industrial applications.  These systems allow for  simultaneous satisfaction of a large number of variables including nutritive, economic, and palatibility.  These may be adapted to nutrition  education purposes to provide recommendations on needed alterations in individual diets, and married to present evaluative programs.  35  2.3  Dietary Assessment The objective of dietary assessment is to provide an estimate of a  population's or individual's nutritional risk, by accurately estimating food and nutrient intake parameters of a population, group, or individual, and evaluating these parameters against appropriate standards (ICNND 1963; Mongeau 1974; Pike and Brown 1975).  Dietary assessment has three phases:  f i r s t , the collection of data on food intake, second, analysis of dietary data to determine the nutrient intake, and third, evaluation by comparing estimated nutrient intake with appropriate standards. Dietary assessment is only one of the procedures used in the assessment of nutritional status.  Present methods for assessing the nutritional  status of individuals includes dietary studies, clinical studies and laboratory investigations  (Jelliffe 1966; Christakis 1973).  Each of these  techniques is important for investigation of nutritional status, and for defining desirable dietary practices.  However, only in comprehensive  nutrition surveys would all techniques be simultaneously invoked.  Other-  wise different techniques would be selectively employed depending on the objectives of the study or requirements of the situation. Nutritional status according to Christakis (1973) is "the health condition of an individual as influenced by his intake and utilization of nutrients, determined from the correlation of information obtained from physical, biochemical, c l i n i c a l , and dietary studies" (p. 80).  Consequently,  nutritional status by definition cannot be judged from dietary intake data alone for two primary reasons. (i)  Since the evaluative standard for nutrient intake - - the dietary  standard - - is built on a normal curve which has inherent variability and consequently makes the individual's exact requirement  36  unknown, a certain diagnosis of nutrient inadequacy cannot be made from a knowledge of nutrient intake alone (United States 1974; Hegsted 1974). Although it is highly unlikely that intake is inadequate when the dietary allowance or standard is met or exceeded, there s t i l l exists a risk that deficiency will occur.  Similarly, an intake below the allowance is not,  in i t s e l f , evidence of nutritional inadequacy.  This criticism applied to  other methods of nutritional status assessment: . . . data, like dietary data, have been traditionally interpreted by selecting some arbitrary cut off point, tabulating those who fall below that point, and using this value to indicate the number of :'>. "deficient" individuals in the population. Obviously, the significance <: of any measure of nutritional status - whether based on dietary, biochemical, c l i n i c a l , or anthropometric data = depends on how it relates to the conditions one is attempting to diagnose or the specificity of the measure and the variability of the measure itself (Hegsted 1975, p. 18). Any measure of nutritional status can only indicate a degree of risk which may be attached to that measurement.  Thus, to assess the nutritional status  of an individual, records of nutrient intake should be considered in conjunction with the results of c l i n i c a l , biochemical, and anthropometric measures.  This procedure simply increases the probability of making a  correct diagnosis. (ii)  Different methods of assessment are not exactly equivalent.  Nutritional assessment tools have sensitivity to different aspects of the deficiency continuum as predictors of risk, and|.,provide different types of information about nutritional status. J e l l i f f e (1966) classifies the assessment tools as:  direct methods,  including c l i n i c a l , biochemical, anthropometric, and biophysical techniques; indirect methods, such as information on vital statistics; and ecological methods, wich consider food consumption, cultural influences, economic factors, and infectious diseases.  Jelliffe's  socio-  classification  illustrates a basic difference in the type of information provided by the  37  different methods.  Whereas, the direct and indirect methods of assessment  "ideally" identify the consequence of nutrient influence, ecological measures such as diet assessment are used to establish causation.  Diet is only one  of many factors that may result in poor nutritional status and consequently diet adequacy does not ensure nutritional health. However, the different u t i l i t y of dietary assessment methods confers some distinct advantages to its use.  Dietary assessment monitors a differ-  ent aspect of the deficiency risk progression and therefore is an important component of nutritional assessment.  The progression from sufficiency to  deficiency is illustrated in a model of the development of deficiency diseases as follows.  Five stages of interference in metabolism, give rise  to progressively more overt clinical symptomatology:  f i r s t , the preliminary  stage with depletion of body stores for a nutrient subsequent to a poor diet, disease, or other conditioning factor; second, the biochemical stage where biochemical defects become apparent due to enzymatic depletion of coenzymes and the like; third, the physiologic stage which presents unspecific clinical findings such as general malaise, and i r r i t a b i l i t y ; fourth, the clinical stage where overt clinical signs are evident but tissue pathology exhibits nonspecific syndromes such as skin lesions and anemia; and f i f t h , the pathologic stage where specific syndromes exist and pathology has progressed to more vital functions (Krehl, W. A. 1964). Although there is actually considerable overlap in the specificity of the methods for monitoring the risk continuum, i t would appear that " . . . dietary information should be the most sensitive method of anticipating risk of deficiency" (Hegsted 1975, p. 18) from dietary origin.  Also, dietary  assessment may provide cues to nutritional problems that may not be evident from other methods.  38  In summary, dietary assessment procedures aid in the interpretation of other survey findings by, f i r s t , contributing to the diagnostic sensit i v i t y of nutritional status measurements, and second, by providing etiologic clues to morphologic, biochemical, or functional abnormalities. Assessment by itself is of limited value without knowledge of causation, since this is important for defining corrective measures.  Additionally,  dietary assessment acts as an indicator of early nutritional risk, and provides mechanisms to detect risk of nutritional problems where clinical and biochemical procedures are either not sensitive or are presently unavailable.  Consequently, as a public health tool, dietary data have  considerable merit in that this serves as the earliest  preventive  nutritional health monitor. Thus, while recognizing the limitations of defining nutritional status by dietary status alone, dietary assessment can be an effective public health tool to provide evaluation of food selection practices of individuals - - and in fact may be the only practical method to do so. The following three sections deal more specifically with the three phases of dietary assessment - - data collection, analysis, and evaluation - - and how they can be best incorporated into a system to provide instruction on food selection practice.  39  2.3.1  Data Collection  2.3.1.1  Data Collection Methods  "The aim of all dietary surveys, whether made on individuals or on groups, is to discover what the persons under investigation are in the habit of eating. Their diets must be those to which they are accustomed and which they freely choose" (Marr 1971, p. 108, quoting Widdowson). Methods are available to collect data on the intake of populations (ICNND 1963), several hundreds of individuals (Paul et aJL 1963), families or households (ICNND 1963), and individuals (Chappell 1955; Taggart 1962).  The present concern is with data collection on indivi-  duals. Food consumption data on individuals may be collected, according to the schema presented by Marr (1971), in the following ways: (i)  Present intakes of food can be recorded by weighing methods.  The weighing methods can be divided into two types, the precise weighing method and the weighed inventory method (Marr 1971).  Both are applicable  to the free-living individual, unlike metabloic balance studies which require controlled laboratory conditions (Wilson e_t al_. 1964).  With the  precise weighing method, all ingredients used in the preparation of dishes, the inedible wastage, cooked portions of food, and the plate waste are recorded.  The weighed inventory method only requires that the prepared  food be weighed immediately before consumption and the plate waste be weighed at the end of the meal.  In  either method, the actual measurements  of foods eaten may be under the investigator's control (Pekkarinen and Roine 1964).  Analysis of the diet may be chemically determined from  aliquot samples or may be calculated using tables of food composition.  40  (ii)  Present intakes of food can be recorded in household measures.  This method, as described by Marr (1971), requires that the record of the food eaten, over a period of usually three to seven days, be described in terms of household measurements or be compared in size with food models. These descriptive terms are then converted to weights which are used for assessing the nutrient and caloric intake by using tables of food conposition.  Although direct supervision is not necessary, where cultural  patterns do not predispose to systematization, or where literacy is low supervision may be required. (iii)  Present intakes can be recorded as a menu.  The problem of  collecting dietary data on large groups of individuals for epidemiological study, has prompted the development of methods which require only a minimal amount of data from the subjects (Wiehl and Reed 1960; Marr 1971).  This  method measures the foods consumed over the study period as a menu without quantities.  The assumption, when using unquantified intake instead of a  measured or weighed intake, is that the size of helping does not vary to any great extent between individuals, or at least not to such an extent that course measures of dietary quality cannot be determined.  In some  cases factors corresponding to the amount of food in an average portion, or derived from multiple regression analysis, are applied to the frequencies to obtain quantitative approximations of foods or nutrients consumed (Mongeau 1974). (iv)  Past intakes of foods actually consumed can be recalled.  This  method measures actual past intakes as remembered at an interview (Adelson 1960; Guggenheim et_ aj_. 1960; Huenemann et_ al_. 1961; Bransby et al_. 1964) or on self-completion questionnaires (Keen and Rose 1958). The subject is usually asked to recall all food consumed during the  41  previous day and to estimate quantities in ordinary measures or servings. To facilitate quantitication estimation aids such as food models and measuring utensils may be utilized in interviews, and have also been used in questionnaire methods (Johnson ejt al_. 1974).  Customarily, foods  eaten in the previous 24 hours are recalled - - and for this reason the technique is usually termed the 24-hour recall - - , although recalls covering up to the previous three days (Guggenheim ejt al_. 1960) and for periods of one week (Huenemann e_t al_. 1961) have been reported. (v)  Past intakes of foods usually consumed can be recalled.  The  diet history method obtains data on general dietary pattern, as opposed to current diet or past foods actually eaten (Pekkarinen 1970).  The method,  developed by Burke (1947), makes use of several different approaches to obtain information from an individual about average food intake over a period of time.  The three phases of the diet history are, f i r s t , a deter-  mination of the past pattern of eating by using the 24-hour recall coupled with inquiry about the usual eating pattern.  Second, using a detailed  predetermined l i s t of foods, the pattern of eating reported in the f i r s t section is cross-checked to determine inconsistencies  and inaccuracies.  Finally, a three day food record of present intake in the form of a nonquantified menu is kept by the subject.  The results of the history are  recorded in household measures, and are converted by use of food composition tables to their nutrient values.  The r e l i a b i l i t y of the data on usual  consumption, obtained by the f i r s t phase of questioning, is verified by the later two phases, for which reason the method is called the crosscheck dietary history (Hartog e_t al_. 1965).  Burke (1947) indicates the  history method is suitable only when a well defined dietary pattern exists, vi)  Past intakes of foods can be recalled as an unquantified menu.  42  In addition to short-cut methods of data collection by recording, abbreviated methods for collection of recalled dietary data have been developed.  These are radically shortened versions of the Burke method  which measure usual frequency of consumption data.  Both interview  (Stefanik and Trulson 1962) and questionnaire formats (Hankin and Huenemann 1967; Hankin et al_.1967; Balogh et al_. 1968):have been tested.  2.3.1.2  Evaluation of Data Collection Methods  Although a variety of techniques for determining dietary intake for free-living individuals are available, none has received general acceptance. Indeed as Marr (1971) points out, controversy has raged over what constitutes the best method.  She indicates that perhaps the "extremes of con-  sidered opinion" (p. 109) are respectively illustrated in the views of Widdowson and of Burke.  The former emphasizes "the necessity for [accurate]  measurement of current intake and the other the need for assessment of [representative] 109).  food habits over a considerable period" (Marr 1971, p.  Thus, there are conflicting views on the value of accurate measure-  ment which arise primarily because procedures to maintain validity of measurement may interfere with the representativeness  of the sample popu-  lation, or with the eating patterns of that population (Thomson 1958; Mann et_ al_.  1962; Marr 1971).  Mann et al_. (1962) complains in response  to this confusion, that " . . . a superficial examination of the technical problems experienced in measuring dietary intake meets such a morass of conflicting opinions that the f i r s t inclination is apt to be a decision for abandonment" (p. 212). Pekkarinen (1970) and Marr (1971) indicate, in part, the problem of appraising dietary intake methodologies is that no independent device,  43  or absolutely accurate method of dietary data collection, exists against which the validity of the methods used in dietary assessment can be measured.  The validity of each method can only be tested comparatively  (Marr 1971). Although no single ideal method exists, each method has its advantages and disadvantages.  Consequently, the choice of method, its potential and  validity, must be assessed against the objectives and purposes tbf the study; and against the circumstances in which the technique is to be used, for example, funds and personnel available for the  study, and the study  population's characteristics. In terms of validity of measurement alone the precise weighing technique, with investigator supervision and analysis of aliquot samples, theoretically is the method providing the most valid estimate of actual food and nutrient intake (Marr e_t al_. 1959; Whiting and Leverton 1960; Marr 1971).  The  weighed inventory method with investigator supervision and analysis of aliquot samples would be expected to be a close second.  However, the  expense required for performing weighing methods, (Pekkarinen and Roine 1964) and the intrusiive nature of the collection procedure (Marr 1971), limit the u t i l i t y of such studies to small, often non-randomized populations of cooperative individuals (Pekkarinen and Roine 1964; Hartog e_t al_. 1965), and to relatively short periods of reference - - one or two weeks is generally considered to be the maximum duration (Pekkarinen 1970).  Further, the  intrusive character of the weighing techniques may interfere with normal behavior and thereby limit the validity of the measurements (Pekkarimen 1970; Marr 1971).  "To what extent this [intrusion] alters the food intake  is d i f f i c u l t , i f not impossible, to determine" (Marr 1971, p. 110). Inconsidering other methods, i t must then be determined to what  44  extent measurement validity is influenced by the various departures which are necessary in practice to reduce costs and improve subject cooperation, such as:  the use of food composition tables instead of  chemical analysis for nutrient analysis (Marr 1971); the use of subject versus investigator measurement for data collection (Paul et al_. 1963; Marr 1971); the use of descriptive measures or no measurement instead of direct weighing (Bransby et_ al_. 1948; Young et_ al_. 1952a; Thomson 1958; Marr 1971); and the reliance on memory in recall methods rather than on direct observation (Huenemann et_ al_. 1961; Marr 1971; Mongeau 1974).  The loss in accuracy of measurement can in turn be judged against  the increased usefulness of data derived from representative samples of a population living their normal lives, for which the weighing methods are not a practical technique (Marr 1971).  The less intrusive methods encour-  age greater subject cooperation resulting in opportunities for longer term studies of larger sample size with randomized samples; and for reduced interference in eating pattern (Pekkarinen 1970; Marr 1971).  Data  collection at relatively lower cost also permits increased sample size and longer time frame (Marr 1971).  2.3.1.2.1  Validity and Repeatability of Methods  Although absolute validity for the weighing methods cannot be established; relative validity has been demonstrated by comparing weighted surveys with the results of other dietary data collection methods (Marr 1971).  The weighed inventory method has also been indirectly validated,  Durnin (1961) found that caloric expenditure, as measured by indirect calorimetry, and energy intake determined by the weighed inventory method were balanced over seven days.  Tests of repeatability show reasonably  consistent results with both the precise weighing method (Marr et_ a]_. 1959) and the weighed inventory method (Heunemann and Turner 1942; Adelson 1960). Compared to data recorded by weighed methods, data recorded in household measures and descriptive estimates are considered less precise largely because of errors in estimation of portion size (Young et al."1952a; Thomson 1958; Marr 1971).  Interestingly,  in a study conducted by Bransby  et al_. (1948), comparisons of weighed and measured intakes for both individuals and groups did not show marked discrepancies, providing the measuring equipment was standardized. Dietary records collected as menu items without an indication of quantity are considered less precise than either weighed methods or those using household measures.  However, investigators (Marr 1971) testing the  u t i l i t y of frequency-of-consumption information indicate i t was accurate enough to classify populations into broad categories with respect to nutrient intake. According to many investigators, grouped mean intakes from recalls of actual intakes and those from weighed records (Morrison ejt a]_. 1949; Adelson 1960; Combs and Wolfe 1960), estimated records (Payton et al_. 1960), or diet histories (Stevens et al_. 1963) give comparable results and thus can replace each other in group surveys.  This is so particularly  if  samples are large and i f daily and seasonal variations in intake are small (Pekkarinen 1970).  However, agreement of grouped data between recalls and  other methods is not found in every case (Bransby et al_.  1948; Young et a l .  1952c; Trulson 1954; Thomson 1958; Pekkarinen et al_. 1967).  When individual  consumption is sampled, variation in intake data are observed between dietary recalls and other methods by some investigators (Morrison et a l . 1949; Young et al_. 1952c; Trulson 1954).  This finding is not universal.  46  Some authors find that when individual consumption is sampled over the same time frame (Flores e_t al_. 1965), or even different time frames (Adelson 1960), results of recalls and weighed records are in good agreement. Considering the dietary history method, Balogh et_ al_. (1968) have found comparable results from the history method and a weighed record. However, in most cases (Huenemann and Turner 1942; Young et aJL 1952b; Trulson 1954; Paul et al_. 1963; Hartog et al_. 1965; Hart and Cox 1967), attempts to validate dietary histories against weighed or measured food records have demonstrated discrepancies between the methods for group and individual values, particularly when surveying children (Trulson 1954; Beal 1967).  Repeatability of the dietary history method had been demonstrated  by repeat histories (McCann et_ al_. 1962; Paul e_t al_. 1963).  Burke et a l .  (1943) provides some evidence that indirectly validates the dietary history method.  In this study the results of a diet history was validated against  an independently measured variable - - the incidence of preeclampsia. Results of validation studies indicate favourable comparisons between the shortened recall methods and research history interviews, weighed, or measured weekly records for group data (Marr 1971).  The applicability of  this method to individual dietary analysis is doubtful.(Marr 1971).  2.3.1.2.2  Time Frame of Methods  Information on long-term food consumption is the objective of most dietary surveys.  The efficacy of extrapolating short-term data to  habitual intake is questionable (Cantoni ejt a]_. 1961; Pekkarinen 1970; Marr 1971; Mongeau 1974). Studies using weighed records to explore weekly variations in individual's diets have provided conflicting long term intake.  Some authors  (Thomson 1958; Morris et_ al_. 1963) believe that seven-day weighed records  47  give a sufficiently accurate estimate of usual intake, since in their study populations considerable individual stability of food intake was found over widely separated weeks.  However, others (Yudkin 1951; Keys et  al. 1966) indicated that a one week survey was not predictive of long term intake.  Adelson (1960) reports that for some individuals weekly diet  patterns are relatively stable, while this is not the case for other subjects.  Authors using weighing methods (Yudkin 1951; Chalmers et a l .  1952; Chappell 1955; Fidanza et al_. 1964; Hartog et al_. 1965; Hankin and Huenemann 1967) have indicated that daily, weekly, seasonal, and yearly variations in dietary intake are evident.  The extent of the temporal  variations is influenced by factors such as age, gender, and occupation (Chalmers et_ al_. 1952; Adelson 1960).  Also the different nutrients show  different degrees of variability (Chalmers e_t al_. 1952). A general statement about the appropriate length of time to be surveyed using weighing methods, or in fact any method of measurement, cannot be given (Pekkarinen 1970; Marr 1971).  The length of time required  to establish a representative picture of intake is obviously dependent on the individual's or population's food intake patterns.  If extreme varia-  tions exist in food intake pattern, no method short of extensive sampling is likely to accurately categorize long term dietary intake.  As the variation  in the diet decreases the time frame required for-accurate estimates of usual consumption can be reduced accordingly.  Providing the duration of  sampling corresponds to a consistent dietary cycle or rhythm, the sampling frame should be indicative of usual intake.  To define the time frame  required to record dietary intakes of usual consumption for an individual or group the precision required must be stated, and intake variation per time must be determined for each nutrient and for each sample of the  48  population to be appraised (Marr 1971). Thus, the efficacy of extrapolating from the sample period depends on the consistency of the dietary pattern (Pekkarinen 1970).  Even the diet  history method, which is intended to measure long-term food pattern, is suitable only when a well defined dietary pattern exists (Burke 1947). Sampling period can be more effectively lengthened by sampling in a number of separate one week periods over the year (Chappell 1955) or as Balogh et al_. (1971) have shown by repeated 24-hour recalls.  2.3.1.2.3  Randomness and Size of Population Samples  A major disadvantage of the weighing methods is that they demand a high degree of cooperation from the individual, consequently response rates in population studies may be low.  For this reason, weighing methods  may not be suitable in studies carried out on random samples (Pekkarinen and Roine 1964), as has been found in certain countries (Hartog et a l . 1965).  However, in some other countries (Buzina et_ a]_. 1964; Fidanza et_  al. 1964) the use of weighing methods has not prevented randomization of the survey population.  Since records using household measures are less  demanding than weighirigmethods a higher degree of cooperation may be expected.  However, Marr (1971) indicates that higher cooperation rates  are not necessarily achieved by descriptive as compared to weighed studies. For all its faults, the 24-hour recall is the only suitable method for use in large scale surveys of heterogeneous populations.  The method  should allow randomized surveying of large representative sample populations, since cooperation rates would be expected to be high (Pekkarinen 1970; Marr 1971).  However, Marr (1971) indicates that l i t t l e information  on cooperation rates in recalled surveys of actual food consumption is  49  available in the literature.  The diet history method is also reasonably  acceptable for random surveys of large population samples, particularly in its abbreviated forms (Marr 1971).  2.3.1.2.4  Other Methodological Considerations  In addition to the generic features of the dietary-data collection methods, which have been presented above and which contribute to discussions of the relative validity of these methods, several other methodological considerations which effect the accuracy of dietary intake data have been identified.  These considerations include:  the professional skills of  the interviewer or technician doing the data collection (Adelson 1960); the use of auxiliary procedures, such as visual aids and models to assist in quantification (Balogh et al_.  1968); the characteristics of the diet  under study - - its complexity or monotony (Chalmers e_t al_. 1952); the circumstances of the study, such as season undertaken (Mongeau 1974); the study populations characteristics including their skills (Young et a l . 1952c; Stevens e_t al_. 1963), age (Marr 1971), attention span, memory capabilities, education, literacy, and intelligence  (Pekkarinen 1970); and  perhaps a variety of procedural nuances used to reduce errors of portion size estimation and of omission, for example, explicit inquiry about postprandial consumption patterns when doing recalls (Balogh e_t al_. 1968), and a detailed interview following completion of client records to enable the quantities and size of helpings to be checked (Kitchin et_ al_. 1949). The accuracy of data collection methods may be greatly influenced by situational factors, as those above.  For example, in quantitative studies  using estimation methods, the ability of the subject to accurately estimate quantities is important for achieving accurate results.  Stevens et a l .  50  (1963) and Young et_ al_. (1952c) found good agreement between methods when using informed subjects, such as Home Economics graduates, whereas the results from other groups of subjects provided inconsistent answers. Presumably had all the subjects in the experiment been able to estimate quantities accurately, the conclusions reached by these authors on the comparative accuracy of the methods would have been different.  Present  comparative studies of data collection procedures have not explicitly addressed the impact of these factors on the accuracy of methods or their possible importance for defining the exact method appropriate for every field situation.  In this regard, new methods are being proposed and  developed (Mongeau 1974).  Explorations into short schedule epidemiological  methods (Hankin and Huenemann 1967) are being made.  2.3.1.3  Summary and Conclusions  Evaluation has identified some of the limitations and potentials of each method.  The basic problem is that i t is d i f f i c u l t , and perhaps  impossible, to measure accurately the dietary intake of a large random sample of free-living individuals for long periods of time (Marr 1971), particularly i f extreme''daily variation in intake exists (Christakis 1973). No method presently available ensures, simultaneously, validity of measurement and relatively unbiased sampling and experimental methodology.  That  i s , attempts to ensure absolutely accurate measurement in most cases do not permit data collection from large samples of people, from random z samples of those populations, for long periods of time, and from freeliving individuals. As indicated by Marr (1971) and Pekkarinen (1970) the choice of any method - - its acceptability - - is dependent on the characteristics of the population considered, the purposes of therstudy, and the circumstances of  51  the study.  Each situation and study purpose has its own best data  collection method. utility.  Similarly, each of the methods has its own specific  Any consideration of the acceptability of a method must take  into account the purpose and situation for which the method was designed. In the present evolving field of collection methodology for dietary data, definition of an appropriate method for any given situation may be dependent on more than simply selection from the available genera of study procedures.  Available methods provide guidelines from which to  define some aspects of procedure; beyond this, in order to ensure an accurate record of the types of foods eaten and the quantities consumed, the standard protocol must be modified and elaborated as dictated by circumstance. Thus, both the application as well as the choice of method are significant in/determining the accuracy and acceptability of any method.  52  2.3.2  Data Analysis Beyond choosing how to collect dietary data representative of an  individual's intake, and choosing how a particular intake is to be judged, analysis of the dietary intake data is required to provide indices for evaluation.  These indices may take the form of nutrient values (Marr 1971),  information on foods and food groups (McClinton ejt al_. 1971), or alternatively, dietary scoring patterns (Trenholme and Milne 1963).  For the  purposes of this thesis, analysis which provides nutrient composition information reflective of the actual nutrient intake is required. Two methods for determining the nutrient content of diets are available. These are calculation of nutrient values from food tables and nutrient analysis by chemical methods.  Whiting and Leverton (1960) indicate that,  laboratory analysis of individual foods, or of composites, provides the most reliable estimate of the nutrient values of foods actually eaten. However, the laboratory procedures and data collection practices needed to secure food samples for analysis, require careful and time consuming handling.  Consequently, available resources generally limit use of these  methods to special  studies.  More commonly, nutrient values are calculated from food tables - - a practice for which the merits have been debated (Harris 1962; Marr 1971). Reflective of the dichotomy of opinions, Widdowson and McCance (1943), in their discussion of the scope and limitations of food tables, wrote: There are two schools of thought about food tables. One tends to regard the figures in them as having the accuracy of atomic weight determinations; the other dismisses them as valueless on the ground that a foodstuff may be so modified by the s o i l , the season or its rate of growth that no figure can be a reliable guide to its composition. The truth, of course, lies somewhere between these points of view (p. 230).  53  •Comparisons of nutrient determinations using food tables with those using chemical analysis, on either group mean or individual data, show somewhat contradictory results.  Some early studies (Bransby et a]_. 1948;  Groover ejt al_. 1967) conclude that substantial differences exist between the two methods.  They suggest that calculated values do not represent  actual nutrient consumption.  Interestingly, a reexamination of the study  of Bransby et al_. (1948) by Marr (1971) suggests that although absolute agreement between analyzed and calculated values for every individual was not achieved, the conclusions reached by Bransby et_ al_. were overly harsh. Marr found that the calculated and analyzed values were essentially the same for individual and group means.  Other authors (Buzina et a]_. 1966;  Pekkarinen 1967) indicate relatively good agreement between calculated and analyzed results.  Whiting and Leverton (1960) find that whereas the group  mean values for some nutrients agree using the two methods, others do not. Differences between the results of the methods depend primarily on how closely the food composition table value corresponds to the nutrient values of the food consumed.  Obviously, i f the values assigned by food  composition tables are the same as the laboratory results, the calculated and analyzed results for a given portion of food will agree.  This is  illustrated in studies where food tables composed largely of analytic data of local foods are being used.  In such cases the results of calculation  and analysis have been found to agree (Pekkarinen 1967).  Furthermore,  closer correspondence of analyzed and calculated values is achieved with nutrient values that are more stable, such as those for energy and protein (Whiting and Leverton 1960).  Where the food compositional values may vary  widely, for example with vat values, the differences between calculated and analyzed values may be larger (Whiting and Leverton 1960).  It is important,  .54  therefore, that the food tables used should be appropriately matched to the foods which are eaten and analyzed, i f exact correspondence between laboratory and computational methods is desired. In comparing food values obtained by analytic methods with those by calculation from food tables, it should not be expected they would agree exactly.  The two methods measure somewhat different information about  foods consumed.  The analytic value determines the nutrient intake of  foods eaten on those days of the study, the calculated value determines the nutrient intake based on food composition which represents the average of food samples taken over an extended time period; perhaps a year or more. Although small special purpose food tables and simple tables for local studies exist, in a complex marketing economy such as North America, i t is not useful to disaggregate compositional values for all the local factors that effect nutrient values.  Thus, food tables in common use in North  America (Watt and Merrill 1963; Church and Church 1975; Adams 1976) contain food compositional information for the country or area which is averaged for factors such as:  genetic differences - - variety or breed; environmental  effects, for example soil f e r t i l i t y , f e r t i l i z e r s , diet for animals, light, temperature, precipitation, and other climatic elements influencing conditions of growth; seasonal and geographic differences; methods of harvesting, handling, and storage; and manufacturing and processing procedures (Asenjo 1962; Watt and Merrill 1963). Averaging across factors which create variability in nutritive values is done in order to develop a final figure as representative of the product available the year around for the area considered (Watt and Merrill 1963). Additionally, the statistical  procedures used may incorporate averages  weighted according to pertinent factors, such as availability or use; other  55  adjustments in the figures, such as excluding widely discrepant results from the final calculations, may be used so that representative figures are developed (Whiting and Leverton 1960; Watt and Merrill 1963).  Food values  in the 1945 U. S. Department of Agriculture publication, "Tables of Food Composition in Terms of Eleven Nutrients (United States 1945), the 1950 and 1963 editions of Agriculture Handbook No. 8 (Watt and Merrill 1950; Watt and Merrill 1963) are treated in this way. Although averaging the compositional values in this way adds potential variability to nutrient values calculated from food tables when compared to analyzed figures, a certain advantage is attached to the use of average values.  When information on usual consumption is desired, then, providing  that the average values are reflective of the actual food consumed, the use of food composition tables with average values provides values which are unbiased by season and other variations in food supply throughout the year. In addition to the questions about the suitability of calculations of nutrient values from food tables raised due to the variation between analyzed and calculated values, there are other limitations when using nutrient analysis from food tables.  Some of these problems are common to  chemical methods. (i)  Standard food composition tables (Watt and Merrill 1963) have  not provided information on .all the nutrient values in dietary standards, and the standard i t s e l f does not provide requirements for all known essential  nutrients (United States 1974), Hertzler and Hoover (1977)  suggest that:  56.  Although more analyses are needed for every nutrient, new tabulations are especially needed for: vitamin E, molybdenum, phosphorus, iodine, chromium, manganese, selenium, amino acids, and individual carbohydrates. Analyses of dietary fiber are needed because crude fiber represents only a fraction of total fiber in foods (p. 22). Recent progress in the development of larger composition tables may largely rectify this difficulty (Schaum ejt al_. 1973; Hertzler and Hoover 1977). (ii)  Food tables do not include values on compounds known to inter-  fere with nutrient availability, nor do the chemical analysis' on which the tables are based estimate the true availability of all nutrients. Hence, calculations from food tables may overestimate the amounts actually absorbed (Harris 1962; United States 1974).  However, in the 1963 Agriculture  Handbook No. 8 (Watt and Merrill 1963) foods with high oxalic acid concentrations are footnoted with respect to availability of calcium.  Also,  although the total amounts of iron, sodium, potassium, and magnesium in foods are listed, the amounts available to the body are recognized, as variable by this publication. (iii)  Although until 1940 food tables produced by the U. S. Department  of Agriculture listed maximum, minimum, and average values for nutrients when more than one analysis was available (Herzler and Hoover 1977), present tables of nutrient composition to not indicate the variability of nutrient values used to determine the average values (Whiting and Leverton 1960; United States 1974).  Thus, while ranges of values would provide  some guidance for r e l i a b i l i t y of calculations, present calculations can only be done from the mean and not the range. (iv)  Nutrient analysis is needed for many items, such as:  ethnic  foods; restored or supplemented items; products with new formulations; and dehydrated, frozen, and ready-prepared products (Hertzler and Hoover 1977). Regradless of the extent of food table development, difficulties will  57  probably s t i l l exist in:  mbnitoring constantly changing manufacturing  practices; describing a l l the variations of preparation practice such as cooking time;,".and providing food names which cover all local idioms. Regardless of the problems encountered with the use of food tables, resource limitations and circumstances necessitate that nutrient analysis be performed using food composition table values.  This applies especially  when the intakes of large numbers of people are monitored for many nutrients. Numerous tables of food composition have been developed in North America since the 1890's. Hoover (1977).  This history has been recently reviewed by Hertzler and  In one of the f i r s t comprehensive tables of nutrient compos-  ition for American foods, Atwater (1895) included values of calories, protein, fat, and carbohydrate for approximately 235 items.  Since that time  advances in nutrition knowledge, refinements in laboratory procedures, expansion in the number of foods available, and a trend towards increased specificity for most nutrient classifications  have resulted in food com-  position tables of considerably larger size.  A projected development of a  data bank by the U. S. Department of Agriculture will consider up to 215 food constituents for an undefined but potentially enormous number of foods (Hertzler and Hoover 1977). Tables of food composition used in North America have as their basis tables developed primarily by the U. S. Department of Agriculture. The most comprehensive of these tables, among the bibliography of the USDA food composition publications provided by Chandler and Perloff (1975), is contained in Agriculture Handbook No. 8.  This volume was first published  in 1950 (Watt and Merrill 1950) and revised in 1963 (Watt and Merrill 1963).  Table 1 of the revised publication provides calories and 16 nutrients  for 2,483 foods in 100 gram portions.  Additional nutrient values for some  58  foods are also included in supplementary tables in the same volume. Publications, such as Church and Church's (1975) "Food Values of Portions Commonly Used", contain food composition tables with items listed according to household portions.  Other publications, including the USDA  publications Home and Garden Bulletin No. 72, f i r s t published in 1960 (United States 1960) and most recently revised in 1971 (United States 1971); and Agriculture Handbook No. 456 (Adams 1975), provide nutrient data for foods in terms of common household measures and market units. Since the late 1950's computer stored data bases have been compiled by the USDA (Hertzler and Hoover 1977).  Presently, USDA food composition  data sets areoavailable to the public in a variety of machine readable issues.  These include the data contained in the 1963 edition and 1972  revised edition of Handbook No. 8 (United States 1977a-b); the 1971 edition of Home and Garden Bulletin No. 72 (United States 1977c); and Handbook No. 456 (United States 1977d). Additionally the USDA began compiling a Nutrient Data Bank in 1972 to serve as an international repository of nutrient data (Watt et al_. Chandler and Perloff 1975; Hertzler and Hoover 1977)).  1974;  This f a c i l i t y will  integrate data submitted from a variety of sources including industrial, governmental, public, and private laboratories, as- well as that already stored in the USDA's Nutrient Data Research Center.  In order to increase  handling capacity, a computer based system has been instituted to facilitate storage, processing, and retrieval of data. will be processed in three ways.  The data stored in the bank  Data from individual analyses will  comprise Data Base I; those from the averaged values of identical items create Data Base II; and the averaged values of similar items will form the present revised equivalent of Agriculture Handbook No. 8, Data Base III (Chandler and Perloff 1975).  Provision has been made for storage of  59 up to 215 nutrients, related constituents, and analytic values (Hertzler and Hoover 1977). computer use.  Data Base II will be made available on tape for  Presently, two issues of Data Base III - - those for dairy  and egg products, and for spices and herbs - - are available in loose leaf format or on tape (United States 1977e). Schaum et al_. (1973) have reported on another extensive nutrient data bank.  The storage space presently allotted permits expansion to a  maximum of 10,000 food items with 63 nutrients each.  I n i t i a l l y , the f i l e  contained 3,600 items, and subsequent work has increased the food item f i l e to approximately 4,800 items (Hertzler and Hoover 1977). In addition to the extensive compilations mentioned above, compressed tables of food composition have been developed and used by a number of authors (Leichsenring and Wilson 1951; Dawber et al_. 1962; Mann et a l . 1962; Browe et al_. 1966) for shorthand determinations of nutrient intake. These investigators compress food item listings by foods of similar nutritive value into food groups, and establishing mean nutrient values for each group by arithmetic averages or by a weighted average adjusted to food consumption or preference practices of the population.  However, these  methods are better for group determinations than for individual determinations of nutrient intake, since neither the population-averaged nor arithmetically-averaged nutrient value for a food group will accurately reflect the individual's actual intake of foods in that food group (Whiting and Leverton 1960). "The extent to which one may group food in a table depends entirely on the methodology, the nutrients one is interested i n , and the amount of detail one is willing to forego for the sake of simplicity and time" (Browe et al_. 1966, p. 107).  In nutrition-education programs providing dietary  assessments a shortened l i s t of food items is desirable.  However,  "60  compressed tables have the disadvantage that they may necessitate severe reductions in the number of nutrients assessed and the number of foods considered.  Inaccurate assessment may result both from ignoring some  nutrient values in order to effectively group foods, and from substituting nonequivalent food items in the condensed food l i s t . Pennington (1976) has produced a "Miniature Food List" and tables of nutrient composition which circumvent some of the problems usual in compressed food composition tables, by appraising the coexistence of 45 nutrients in a large number of foods.  The miniature food l i s t is composed  of 202 index foods, from among those most commonly consumed in the United States, with values of 45 nutrients for each food.  The index-food items  can be substituted with a variety of other designated items which have been found to be similar in nutrient characteristics.  The assumption used  in defining substitutable items is " . . . that the nutrient variation of any one nutrient in any one food (index item) is greater than the variation of all the means of that same nutrient in all the foods included in the group" (Pennington 1976, p. 9).  The validity of this tool is based on the  demonstrated fact that errors in dietary nutrient value due to the use of food:;substitution and to established serving sizes are much less than errors due to simple variation of nutrients in foods (Pennington 1976). The l i s t does not cover all foods that are potentially available in the marketplace, but does cover a considerable number relative to the index items presented. Pennington's miniature food l i s t may provide an acceptable compressed table of food composition for a nutrition education system performing assessment of individual's diets.  It provides a comprehensive table of  food items and their nutrient values, without compromising the accuracy of  61 the compositional values.  The table provides a balance which maximizes  validity of assessment but minimizes computational and data-collection difficulties. Beyond defining the indices for dietary analysis, in this case nutrient composition, the computational method for processing dietary data should be considered.  Traditionally, nutrient analysis was performed by hand cal-  culation, however, the advent of electronic data processing offers methods for rapid and accurate nutrient analysis (Marr 1971). Computers have been used to perform nutrient analysis on a variety of dietary data, for example, on patient food consumption records (Thomson and Tucker 1962; Brisbane 1964; Eagles et al_. 1966; Beal 1967; Schaum 1973; Tuthill 1974), metabolic diets (Hjortland et_ al_. 1966), epidemiological data (Goodloe et_ al_. 1963; Hayes et aJL 1964), and nutrition survey data (Tie et_ al_. 1967).  Calculation of menu nutrients by computer demonstrates  advantages in cost (Flook and Alford 1974), accuracy, and response time over manual computation (Hoover 1976). Computer systems offer the advantage of comprehensive analysis of nutrients and the option of complex data manipulation, for example, statistical interpretation of relationships among nutrients.  Mathematical  accuracy, of course, will not remove inherent errors in food composition tables.  However, the computer will carry out routine calculations with  speed and complete accuracy.  It therefore seems advantageous to use com-  puters when both ease and accuracy of nutrient assessment procedures would be improved.  62  2.3.3  Data Evaluation The final phase of dietary assessment - - data evaluation - - requires  that estimated nutrient intake be judged against appropriate standards.  The  dietary standard provides the present formal basis on which nutritional science and the nutrition educator define acceptable nutrient intake. Other recommendations which may supplement the values contained in the dietary standard have been proposed by recognized nutrition agencies.  The  following section discusses the characteristics of the standards available for appraising dietary well-being, and their suitability for evaluating individual dietary practices.  2.3.3.1  Dietary Standards  Available information on human nutrient requirements has been incorporated into dietary standards, which are, in a sense, compendia on the known nutrient requirements of man.  The most recent revision of the Canadian  dietary standard (Canada 1975) is defined as "a statement of the daily amounts of energy and essential  nutrients considered adequate, on the  basis of scientific data, to meet the physiologic needs of practically all healthy persons in a population" (p. 5).  The Recommended Dietary  Allowances of the United States Food and Nutrition Board (United States 1974) embodies the same principles. Although the dietary standard is based on actual or extrapolated human requirements, the figures given are "recommended daily intakes" or "acceptable daily intakes", or as the Food and Agriculture Organization of the World Health Organization states "safe level of intakes", and not a statement of the actual nutrient requirements for each individual in the population (Passmore et al_. 1974; United States 1,974; Canada 1975).  This  63  distinction is very important in understanding both the design and use of the dietary standard. Individual differences in nutrient requirements, as a result of normal biologic variability, have been considered in establishing a standard which should meet the nutrient requirements of most healthy people in the population (Beaton 1972; United States 1974; Canada 1975).  Con-  sequently, an individual whose average daily nutrient intake meets the standard for all nutrients, would have l i t t l e likelihood of nutritional inadequacy.  In fact, the requirement of most individuals should be less  than the standard. different.  The formulation of standards for energy intake is  Unlike the standard for other nutrients, the recommended  caloric intake approximates the predicted average requirement of the population members, instead of lying substantially above the average requirement.  For this reason, there is relatively l i t t l e relationship to indivi-  dual requirements which may be above or below the recommended intake. Some qualifications for the use of dietary standards are recognized. First, the dietary standards are not formulated to cover additional requirements of persons depleted by infection or injury, other traumatic stresses, prior dietary inadequacies, genetic and metabolic disorders, and the use of pharmaceutical preparations (eg. oral contraceptives); nor do they consider losses of nutrients that occur during processing and preparation of foods (United States 1974; Canada 1975).  Second, although the standards  are expressed as a daily average, the ability of the body to adapt to daily fluctuations in nutrient intake implies that recommended intakes do not have to be met on a daily basis (United States 1974; Canada 1975).  Therefore,  it is considered acceptable when estimating dietary adequacy to average weekly intake of nutrients.  64  Third, when using dietary standards, even accurate knowledge of actual nutrient intakes is not synonymous with evaluation of the nutritional status of eigher the individual or the population surveyed.  2.3.3.1.1  Formulation of Dietary Standards  The ideal method, rarely i f ever achieved, to develop an allowance would be to (1) determine the average requirement of a healthy and representative segment of each age group for the nutrient under consideration; (2) assess statistically the variability among the individuals within the group; and (3) calculate from this the amount by which the average requirement must be increased to meet the needs of nearly all healthy individuals (United States 1975, p. 5). Insofar as possible, the above methodology is followed for developing a nutrient allowance.  Population groupings are selected for many of the  major identifiable factors influencing requirements - - that i s , age, sex, body size, physiological state, and physical activity (Harper 1974; United States 1974).  Utilizing available estimates of average requirements and  variability within the populations studies, allowances for nutrients other than energy (which is an estimate of average needs of the group), are developed by increasing the average nutrient requirement two standard deviations (Lorstad 1971; Beaton 1972; Harper 1974; United States 1974). Providing or assuming that individual requirements f i t a statistically normal distribution, an allowance set two standard deviations above average should be sufficient to meet or exceed the needs of 97.5% of the individuals in the population (Lorstad 1971; Beaton 1972).  Unfortunately, such  allowances cannot be established for most nutrients, with the exception of thiamin, riboflavin, niacin, iron, and protein, because there is inadequate information about the variability of individual requirements (Beaton 1972).  65  When information on the requirements of large groups of individuals is not available, average requirement and estimated variability are derived from the following data:  one or two controlled feeding trials  on a limited number of subjects; c r i t i c a l metabolic studies on animals with values extrapolated to man (the translation is fraught with undertainty); or dietary surveys on the minimum amount of a nutrient known to be consumed by apparently healthy people (Harper 1974; United States 1974; Canada 1975).  Unfortunately, the basic requirement and figures for assumed  or actual population variability from which the allowances have been derived are not stated in the dietary standards (Lorstad 1971). Factors that influence the efficiency of nutrient utilization are considered in setting allowances.  These include precursor conversion ratios,  efficiency of absorption, digestibility, assimilability, and utilization of complex nutrients like protein (Young 1964; J e l l i f f e 1973; Harper 1974; United States 1974).  The importance of each of these factors differs  from nutrient to nutrient, so the extent to which the allowance for different nutrients must exceed requirements varies accordingly. Although the methodology for establishing allowances is generally accepted, estimates of nutrient requirements arrived at by committees represent the results of accommodation and judgement.  Consequently,  recommendations from different committees may differ (Goodhart 1973; Harper 1974; Canada 1975).  66  2.3.3.1.2  Use and Interpretation of Dietary Standards  Dietary standards have come to serve as guides in an expanding number of areas.  However, "there are two primary uses of dietary standards — to  serve as guides for the planning of diets and food supplies and to evaluate nutritional adequacy from food consumption data" (Hegsted 1975, p. 13). With respect to both uses, publications of dietary standards have regularly indicated that the values are intended for use as guides applicable to populations and large groups, and that they are not intended for application to the individual.  Additionally, as mentioned, dietary standards cannot  be used by themselves for the assessment of nutritional status of either individuals or population groups (Goodhart 1973). Although the limitations on the use of dietary standards have been published repeatedly, Beaton (1972) indicates, " . . . the meaning and interpretation of the figures remains a matter of doubt, confusion, and often argument" (p. 356).  For example, there is a " . . . dichotomy represented by  the Food and Nutrition Board's insistence that they were intended for use only when dealing with groups of people, but at the same time, the Board's willingness to give, in the 1974 edition, individual figures for twentyfour age-sex groups plus those for pregnancy and lactation" (Leverton 1975, p. 9).  It appears as suggested by Beaton (1972) " . . . the recommendations  are based upon a consideration of the individual, not population requirements - - or rather, upon a consideration of individual requirements within a population" (p. 356).  Referring to the Recommended Dietary Allowance of  the Food and Nutrition Board, Hegsted (1975) has further criticized the u t i l i t y of one standard for two purposes - - evaluation and planning of diets - - and suggests that neither/purpose is f u l f i l l e d with the present standard.  67  The dispute over whether the standards are appropriate for use with populations and not for individuals is of particular importance to nutrition educators.  The warnings provided in the dietary standard, about their use  for evaluating or planning individual's diets, although instructive, fail to address the needs of nutritional practitioners who must evaluate and plan individual's diets (Hegsted 1975).  Beaton (1972; 1975) and Lorstad  (1971) have been instrumental in the development of a logical approach to the interpretation of nutrient intake data for both populations and individuals.  Their approach, discussed below, recognizes that population  data defining dietary standards are in fact data on a population of individuals.  The dietary standards so derived can be used rightfully to evaluate  and plan the diets of individuals. The objective of a dietary study is to determine whether deficiency exists in the individual, or in the case of a population, the proportion of the population that is deficient.  The problem posed then is:  what inter-  pretation should be placed on those individuals with intakes below the recommended level, arid for the population as a whole, what prevalence of deficiency might be expected (Beaton 1972; Beaton 1975)?  The Canadian  (Canada 1975) and American dietary standards (United States 1974) have clearly indicated that the intake of a person habitually consuming nutrients below the recommended levels cannot be interpreted as an inadequate intake for those nutrients. The method proposed by Beaton (1972; 1975) i s , in short that, providing data are available to describe the distribution of individual requirements in a population, then, by the application of probability statistics i t is possible to determine the risk or probability of deficiency to an individual, or the prevalence of deficiency in a population.  As Beaton (1975) suggests,  68 " . . . assessment of the adequacy of nutrient intake should be based upon a judgement of the probability or risk of deficiency rather than on an 'adequate' or 'inadequate' basis . . . " (p. 31). diet is adequate will depend on where  Whether the individual's  the individual's true requirement  lies in relation to the assigned requirement, and this is not known. Similarly, population averages of nutrient intakes can be compared to the recommended allowance, but a population that fails to meet the standard does not necessarily have inadequate diets (Hegsted 1975).  Thus, i t is only  within the framework of probability that the dietary standard can be used legitimately and meaningfully to interpret the relationship of nutrient requirement and the individual's nutrient intake (Beaton 1972; United States 1974; Beaton 1975) or the population's nutrient intake (Beaton 1972; Beaton 1975). The concept outlined by Beaton (1972) will be summarized below.  In  order to determine individual risk of deficiency, a knowledge of the d i s t r i bution of the  .individual requirements in the population is needed.  This  is illustrated below in Figure 2.1 with the assumption of a normal distribution about the mean requirement.  The recommended intake lies at two  standard deviations above the mean, a level which should meet or exceed the requirements of 97.5% of the individuals in the population.  mean  . _ u + 26 requirement Individual variability in nutrient requirements (Beaton 1975). -  Figure 2.1  u  69  Utilizing the data on the variability of nutrient requirements in the population (Figure 2.1), a cumulative distribution can be developed to (Figure 2.2) describe the proportion of the population with requirements above a certain intake level.  This curve corresponds to a distribution  of probability or risk of deficiency for an individual at any given intake level, and provides a basis for interpretation of individual nutrient intakes.  Probability of deficiency in individual s  ~* Figure 2.2  nutrient intake  u  Proportion of the population having actual requirements above nutrient intake and the probability of deficiency in individuals ingesting a particular level of nutrients (Beaton 1975).  Beaton (1972) derives a f i r s t principle for use in interpreting individual and population nutrient intake data from the above graph (Figure 2.2).  The principle states " . . . an individual habitually consuming the  recommended intake (or more) of a nutrient must be considered to be at low risk of deficiency.  As . . . intake f a l l s , the risk of deficiency  increases" (Beaton 1972, p. 358), in a manner predictable from a knowledge of the distribution of requirements.  Thus, although i t is not known  whether an individual intake is inadequate, the individual's dietary intake can be interpreted in terms of the risk of deficiency associated with that  70  intake. This concept can also be applied to populations to predict the prevalence of deficiency in a population. derives a second principle.  For this purpose Beaton (1972)  The second principle states, " . . . when  population data are considered, i t is necessary to consider both the variability of nutrient requirements among individuals and the variability of habitual  intake among individuals" (Beaton 1972, p. 358), before the  prevalence of deficiency can be meaningfully determined.  The variability  of both requirement and intake must be considered since the ratios of these variabilities will influence the prevalence of predicted deficiency as illustrated in F„igure 2.3 below.  A consideration of the average intake  of a population alone will not provide a measure of prevalence of deficiency. In this respect Beaton (1972) derives a third principle which states, " . . . an observation that the average intake of a population group is at or above the recommended intake does not mean that all individuals are well nourished" (p. 359).  Figure 2.3  Theoretical model of the relationship of the nutrient intake of a population to the prevalence of deficiency (Beaton 1972).  71  Thus, the dietary intake of an individual cannot rightfully be deemed adequate or inadequate on the basis of a comparison with the dietary standard, since i t in not known where the individuals true requirement lies in relation::to the standard. such assignments.  However, i t has been customary to make  Similarly, nutritional adequacy is not assured i f the  average intake of a population meets the standards.  The prevalence of  deficiency will depend on the range of individual requirements and intakes within the population.  Therefore statements about population groups such  as, " . . . 'The average intake meets RDA standards; therefore, there is no problem of nutritional inadequacy.'are . . . invalid" (United States 1974, p. 14).  Also statements such as the " . . . 'RDA includes a large safely  factor; therefore a diet that meets two-thirds of the RDA standard should be adequate', have no validity" (United States 1974, p. 14).  A nutrient  intake of two-thirds of the dietary standard will be adequate for some but f  inadequate for others; there is no way of knowing who falls into which category.  However, providing the distribution of requirements is known the  risk of deficiency at this level can be computed. Unfortunately, the distribution of requirements is not available for most nutrients for which requirements are established.  In these cases,  requirements are established by committee judgement to include an estimate of population variability.  Whether true or estimated variability is used,  it would s t i l l seem desirable to utilize Beaton's concept of risk so that an individual habitually consuming the recommended intake, or more, must be considered to be at low risk of deficiency. of deficiency increases.  As intake falls,."the risk  Since individual requirements are not known, the  individual's intake can be interpreted in terms of the risk of deficiency associated with a particular intake.  Depending on the available data on  population variability, this interpretation can be performed either with  72  high numerical accuracy, or alternatively, simply as an indication of a trend towards greater or lesser risk.  2.3.3.1.3  Limitations of Dietary Standards  Even i f the nutrient content of the daily food ration meets the dietary standard, there are several reasons why this is  insufficient  information to formulate an ideal, or adequate, or low risk diet.  Dietary  values should be applied with recognition of the following limitations. Present knowledge of nutritional needs is incomplete. as yet undiscovered may s t i l l exist.  Nutrient needs  For this reason the Bureau of  Nutritional Sciences in their publication "Dietary Standards for Canada " (Canada 1975) suggests, " . . . recommended intakes should be achieved by eating a variety of foods because unknown nutrients may be present which are essential  for the maintenance of health' (p. 7).  are forwarded in the American standards.  Similar qualifications  However, this is not completely  consistent with the indication in the same volumes, that standards are acceptable for use as guidelines in processing and fabrication of foods (Hegsted 1975; Leverton 1975). Requirements for many nutrients recognized as essential  have not been  established (Harper 1974; United States 1974), and for those that have, the requirements are based on a criteria of deficiency rather than a nutrient optimum (Mertz 1972).  Since requirements are based primarily on knowledge  of deficiency states, they do not differentiate long-term nutrient consequences other than some conditions of deficiency.  Further, safe maximal  levels of nutrients have not been established. Many possible factors which may significantly influence requirements in individuals have not been established.  For example, standards do not  73  explicitly consider:  drug-nutrient interactions, such as oral contracep-  tives; nutrient-nutrient interactions, such as the effect of vitamin-D on calcium retention, protein intake and the vitamin-B6 requirement, and amino acid balance with protein requirement; the effect with noxious chemicals; individual sensitivity to nutrients; diurnal rhythms; diet pattern; nutrient history; climatic conditions; and the effect of a potent i a l l y large number of environmental factors.  As " . . . a source of  nutrients, food has psychological and social values that are difficult to quantify" (United States 1975, p. 2), these psychological vectors of nutrient requirements have not been considered explicitly in defining the present standard.  Other factors important in determining the dietary  requirements of individuals include:  the organism's ability to adapt to its  environment, and the aspirations of the individual and the expectations that society has for the individual.  Factors, such as those above, must be  considered in defining human nutrient requirements, however, there is s t i l l lack of knowledge concerning the factors that effect nutrient requirements in individuals.  (Goodhart and Shi 1 s 1973; United States 1974; Canada 1975).  Additionally, for many nutrients for which requirements have been identified, there exists considerable conflict in their formulation (Harper 1974).  For example, Hegsted (1967) points out that,  although most of the people of the world do not consume enough calcium to meet the dietary recommendations, and thus are often said to be calcium deficient, there is no convincing evidence that this is true. We do not even know what calcium deficiency looks like in man" (p. 107). It isluncertain.whether criteria to determine requirements in adults should be set to maintain body weight, or prevent nutrient depletion as indicated by balance studies, tissue concentrations, specific functions, or specific deficiency signs (Harper 1974).  Note that "for some nutrients there is  74 a considerable difference between the amount that will prevent the development of specific signs of deficiency and the amount required to maintain maximum body stores" (Canada 1975, p. 6).  Additional criteria that may  be useful in determining requirements, such as sleeping pattern and wound healing time, have been proposed (Davis and Williams 1976).  Further, data  used in estimates of requirements are often fragmentary and based on limited experimental data on humans (Canada 1975), or in fact may be extrapolated from animal studies (Hegsted 1975).  2.3.3.2  Further Dietary Recommendations  2.3.3.2.1  Recommendations for Nutrients Not in the Dietary Standard  The recent Canadian dietary survey (Canada 1973) indicates that some nutritional problems exist in the Canadian population with respect to nutrients for which the "Dietary Standard for Canada" (Canada 1975) has established requirements.  However, vital statistics (Canada 1976b) report  negligible mortality in the Canadian population from diseases associated with these nutrients. Although the "classic" nutritional deficiency diseases appear to be of minimal public health significance in Canada, dietary components have been implicated as etiologic factors in diseases which are among the major causes of death in Canada.  These diseases include, most notably, cardio-  vascular disease (ischaemic heart disease and cerebrovascular disease) and cancer of the gastro-intestinal  tract (Lalonde 1974).  Diabetes mellitus  also ranks as a significant cause of mortality (Canada 1973). Additionally, these aforementioned diseases have importance for population morbidity figures, as do the diet related problems of dental caries, periodontal disease, obesity, and alcoholism.  75. Whereas the dietary standard provides figures for the nutrients associated with the diet-related diseases of low mortality, appropriate standards are not included for the dietary components and nutrients associated with the high-mortality diseases.  However, recommendations and  guidelines for these nutrients and food components, which may be used to supplement the dietary standards, are available from recognized nutritional agencies. The American Heart Association (1973), Canadian Committee on Diet and Cardiovascular Disease (Canada 1976a), and Senate Select Committee on Nutrition and Human Needs (United States 1977f) provide recommendations for the dietary intake of fats and salt.  These are proposed primarily because  of their implication in the etiology of cardiovascular disease.  To reiterate  briefly, the suggestion by these agencies is for a reduction in consumption of fat.  Reduction in saturated fat with a proportionate increase in poly-  and mono-unsaturated fat intake is also recommended.  The American recom-  mendations (American Heart Association 1973; United States 1977f) also suggest limiting cholesterol intake.  Lower intake of salt is recommended  by these agencies since excess dietary sodium is considered an adverse factor in some people prone to hypertension (Men eel y and Battarbee 1976). However, in accepting these recommendations, i t should be appreciated that the diet-heart hypothesis has not stood uncriticized (Werko 1976; Mann 1977). Also, the goals of the Senate Select Committee on Nutrition and Human Needs have been both supported (Latham and Stephenson 1977) and criticized (Harper 1977) in recent publications. The Senate Select Committee on Nutrition and Human Needs also presents guidelines for total and simple carbohydrate composition of the diet. Additionally, the Bureau of Nutritional Science of Canada (Cheney 1976)  is  76  developing guidelines for meal replacements which provide values for dietary fibre.  Decreased dietary fibre intake has been implicated with increased  incidence of gastro-intestinal  cancer and other gastro-intestinal  disorders  (Burkitt and Painter 1974; Spiller and Amen 1975); and with coronary heart disease (Klevay 1974). Although dietary recommendations which are based primarily on epidemiological data, such as those from the American Heart Association (1973), may be less precise than those derived from metabolic studies, the potential public health significance of these recommendations may justify, or in fact necessitate, their inclusion in the formulation of prudent nutrition education programs.  The implication is that the contribution to the total  risk of mortality and morbidity from the secondary nutritional component of the high incidence  diseases in North America is as substantial, i f not  greater, than the mortality and morbidity due to nutritional diseases where diet is the primary agent.  Unfortunately, recommendations in this area  are highly controvertial and therefore difficult to interpret.  2.3.3.2.2  Recommendations for Maximum Intakes  The Canadian dietary standard (1975) suggests that " . . . a consideration of possible effects from intakes far in excess of estimated nutrient requirements is outside the scope of present standards" (p. 7).  Similarly the  American standards (United States 1974) and those of the World Health Organization (Passmore nutrient intakes.  et_ a]_. 1974) do not explicitly consider upper levels on  Whereas the recommended values presented by the dietary  standards are minimum suggested levels of intake, with the exception of calories, many of those discussed directly above correspond to maximum recommended levels of intake.  77  Although quantities in excess of requirements for individual nutrients have been proposed as beneficial  (Pauling 1976), the Food and Nutrition  Board (United States 1974) suggests that: "... we are aware of no convincing evidence of unique health benefits accruing from consumption of a large excess of any one nutrient. Large doses of individual nutrients may have some pharmacological action, but such effects are unrelated to nutritional function. Claims that large intakes of individual nutrients will cure nonnutritive diseases should be viewed with skepticism . . . (p.3 ). No specific benefits or disadvantages are recognized from the ingestion of excessive quantities of many nutrients.  However, toxic ceilings have  been identified for some nutrients (Goodhart and Shils 1973). In a program designed to monitor nutrient intakes, i f possible, assessment should explicitly articulate an acceptable range of intake - that i s , both maximum and minimum nutrient limits should be stated.  When  using standards i t is suggested (United States 1974) that intakes substant i a l l y above requirements are not harmful. specified.  However, these amounts are not  Ignoring values above the requirements established in dietary  standards, in light of the availability of high-potency nutrient supplements and extensive food fortification, may be unwise.  The necessity of setting  minimum and maximum values for nutrients in new products is advised in the American Recommended Dietary Allowance (United States 1974).  2.3.3.3  Conclusions  Standards against which individual nutrient intake can be judged are available both from publications of dietary standards and from the recommendations proposed by recognized nutrition agencies.  The standards are  not absolute limits which separate good from poor diets, but rather are guidelines or goals which indicate the likelihood or probability that a diet is adequate for an individual.  The standards may be used meaningfully  78  to provide guidelines for individual and population dietary practices. When dealing with total populations, a particular prevalance of deficiency can be set as the public health objective - - that i s , a certain risk to the population.  When intake variability in the population is high,  and therefore per-capita nutrient increases needed to reduce the prevalance of deficiency would.be unreasonable, then the public health effort would be to selectively identify those persons with low intakes and raise their intakes.  Since actual requirement and consequently true risk are not known,  the usual approach in counselling individuals is to urge consumption•of sufficient nutrients to achieve the dietary standard, which corresponds to a low level of risk.  Both the individual counselling and the national  planning approach can be defended as proper public health objectives.  In  using the dietary standard, a public health or statistical concept is used rather than a clinical or therapeutic one (Clements 1975: Latham and Stephenson 1977). In many cases these standards or goals are not well accepted.  However,  the importance of accepting some uncertainty in developing standards is emphasized by Latham and Stephenson (1977), who, in_discussing the nutritioneducation profession's responsibility for establishing goals, state: Worrying to us is the use by the opposition of the argument that we lack information to set goals at a l l . That surely is an abrogation of responsibility, a "cop-out". There will always be uncertainties and more research that needs to be done. But advice has to be given, Goals need to be set . . . (p. 152). Further the results of evaluation should not be erroneously extrapolated into a judgement about the nutritional status of the individual.  Although  there is l i t t l e virtue in assigning interpretations to dietary data that cannot possibly be valid, one cannot deprecate the value of dietary data as an indication of potential nutritional status.  Evaluation of dietary  79  data can be utilized as an effective public health tool to provide assessment of food selection practices of individuals - - and in fact may be the only practical method presently available to do so.  80  2.4  Dietary Prescription  2.4.1  Diet Planning Conventional methods (Ohlson 1972) of planning human diets for out-  patient diet counselling, or for modifying pre-assessed hospital routines, include:  generalized and simplified models for assessing dietary adequacy,  such as food guides, concepts of dietary variety, the prudent diet, and checks on a limited number of nutrient values; computational aides, such as exchange l i s t s ; and a variety of rules of thumb and trial-and-error procedures.  The above methods help to simplify the inordinately complex  procedure of dietary assessment and prescription.  However, these methods  do not make explicit use of all the information necessary to provide precise conclusions in the planning of human diets (Gelpi et_ al_. 1972; Balintfy 1973a; Head et_ al_. 1973).  This problem becomes particularly evident as the  complexity of a diet problem increases, or alternatively, as the constraints on an acceptable solution become more severe. Human diet problems are amenable to mathematical definition, formulation, and solution (Balintfy 1973b; Balintfy 1976a), and in fact may require mathematical modelling for effective resolution (Balintfy 1973a).  Balintfy  (1973a) indicates that " . . . the problem of diet planning is not a nutritional, but a mathematical one . . . [which] . . . defies definition, both conceptually and operationally, unless the problem is cast into some mathematical model . . . " (p. 581).  Therefore, it would seem important that  realistic mathematical formulations be available for human diet problems. Without mathematical techniques, nutritionists have limited use for the potentially extensive information required to ensure that nutrient adequacy, preference, production requirements, and budgetary restraints  81  are met when planning diets - - " . . . more data means only that more information will have to be ignored" (Balintfy 1973a, p. 581).  Consequently,  conventional procedures exhibit suboptimal decision processes as compared to mathematical programs (Balintfy 1976b).  For example, menu planning  models, utilizing mathematical programming techniques designed to optimize costs, have demonstrated raw food cost savings of 5 to 34 percent and improved nutritive control without sacrificing patient satisfaction, as compared to conventional planning techniques (Gelpi e_t al_. 1972; Balintfy 1975). Additionally, increased technical demands on dietitians necessitate the use of tools which improve efficiency.  Computerized mathematical  programs for diet planning provide dietitians with data, and techniques for data manipulation, which would otherwise be impracticable.  Also, there is  some indication that the demand for nutrition services may outstrip available manpower committments unless supplemented by technological means (Witschi et al_. 1976). Food selection or diet planning problems can be modelled mathematically as constrained optimization problems (Balintfy 1976a).  As such, the planning  of human diets is a decision process to assure satisfaction of a number of simultaneous requirements.  These requirements are certain attributes of  food and food intake, such as food cost, nutritive value, palatability, and production characteristics (Gelpi ejt al_. 1972).  Although any suitable set  of optimization objectives and constraints may be defined for diet planning, typically coefficients  of food cost have been optimized within constraints  established by nutritive specifications and palatability (Smith 1963; Balintfy 1976a).  Recently, considerable exploration has been undertaken of  models which optimize measures of food preference (Balintfy 1976a).  82  Human diet problems, according to Balintfy (1973b), fall into two major categories; food planning and meal planning.  Food planning is  concerned with decisions of food allotment, over an undifferentiated but specified time period, which characteristically must meet given budgetary, nutritional, and acceptability requirements.  Alternatively, meal or menu  planning is concerned with a temporally-defined food-allotment decision. Meal planning defines " . . . an optimum sequence of meals consisting of combinations of prepared foods, called menu items, . . . such that the required structure of meals and given budgetary, nutritional, and food production specifications are met" (Balintfy 1973b, p. 1).  As discussed  below, a number of approaches to defining, formulating, and solving human diet problems by mathematical means have been undertaken.  2.4.1.1  Food Planning Models  2.4.1.1.1  Food Planning Without Palatability Considerations  Early food planning models were designed to find "minimum-cost" combinations of foods which met specified nutritional standards.  These  purely nutritional models were formulated to find the lowest cost of physiological subsistence and not to consider elements of dietary palatab i l i t y or acceptability such as:  variety, prestige, or other cultural  and personal concerns (Smith 1963). In order to optimize food costs within a prescribed nutrient allowance three types of data are required:  a l i s t of nutritional requirements, the  nutrient composition of available foods, and price coefficients food item.  for each  A formal statement, in algebraic notation, of the essential  relationships between the optimization objective and the three kinds of  83  information used in the construction of least-cost diet models is as follows: Let z be the total expenditure, n the number of foods, p- the unit J  price of commodity j , x- the quantity of commodity j to be consumed, J  m the number of nutritional requirements (restraints), b. the quanth titative requirement set by the i  restraint, and a ^ the quantity  of nutrient i contained in one unit of commodity j .  In this  notation, the problem is to n E p.x. j=l subject to (1) x- ^ 0 minimize z =  J  J  (j = 1, 2, 3, . . . , n)  0  n E a, .x. » b. (i = 1, 2, 3 m) j=l That i s , z, the total expenditure on foods purchased, is the sum of the expenditures on each food, where p.x. is the expenditure on the th  (2)  1 J  j  food.  J  1  The quantity z is to be made as small as possible,  subject  to the two requirements that (1) no negative quantities of foods may be purchased and (2) the total quantity of the i  nutrient (the sum  of the quantities, a-j-x., provided by each food) shall equal or exceed the required amount, b^, for each of the m nutrients (Smith 1963, pp. 6-7). As indicated by Smith (1963), the earliest formulations of the minimumcost diet problem were by Cornfield (1951) in 1941 and by Stigler (1945) in 1945.  Although both authors developed solutions, Cornfield's was  proposed for the case of two foods with any number of nutritional restraints (Dorfman ejt al_. 1958; Smith 1963).  Stigler provided the f i r s t solution  for a general problem with nine equations of nutrient requirements and seventy-seven unknowns for the amounts of food items in the diet (Danzig 1963; Smith 1963).  Stigler's solution was not obtained by linear programming but  by a systematic procedure of trial and error.  Consequently, he could not be  84  certain that an optimal solution had been obtained, although he believed that his result was close to the optimum (Smith 1963). Linear programming of human diets, and in fact linear programming in general, was not available until Dantzig and Laderman offered a general method of solution to the least cost diet problem, called the Simplex Method, in an unpublished paper written in 1947 (Dorfman et_ al_. 1958; Dantzig 1963).  Thus, Dantzig's discovery of the Simplex method of linear programming  allowed for realistic solutions of an econometric problem previously only solvable in principle. The solutions obtained by Stigler (1945) and in 1947 by Dantzig and Laderman (Dantzig 1963) appear in Table 1.  The recalculated solution to  Stigler's problem by Dantzig and Laderman, used Stigler's August 1939 price index data, his food composition data for seventy-seven items, and his nine nutrient allowances.  This linear programming solution produced an  annual subsistence diet with beef liver instead of evaporated milk, and required that the proportions of other items change.  However, just as  Stigler believed, his solution was close to the true least-cost diet - within 25 cents per annum.  In fact, Stigler's answer was only one iteration  from the optimum (Vajda 1958).  85  Table 2.1  Annual subsistence diet for a moderately active adult male, calculated using linear programming methods.*  Commodity  Wheat flour (enriched) Evaporated milk Cabbage Spinach Dried navy beans Beef liver  Stigler solution  Dantzig-Laderman.solution  amount (lb.)  amount (lb.)  cost ($)  370 57 cans 111 23 285  13.33 3.84 4.11 1.85 16.80  total annual expense...  39.93  cost ($)  299  10.77  111 23 380 2.4  4.11 1.85 22.28 .69 39.68  * Calculated by Stigler (1945) and recalculated by Dantzig and Laderman (Dantzig 1963) + Cost determined using Stigler's August 1939 price index data.  Food planning models using linear programming have been developed by other authors including Vajda (1958), Beckman (1960), and Smith (1963). These models provide solutions of minimum cost diets which meet from 3 to 13 nutrient restrictions, and consider a commodity l i s t of from 8 to 73 food items.  Since linear models usually provide solutions with fewer foods  than restrictions, the diets obtained from these models contain a predictably small number of food items.  Beckman's solution contains 4 foods,  Vajda's solution 3 foods, and Smith's midget model solution 6 foods.  Smith  (1963) indicates that neither Vajda's model nor his own model can be considered, atrue least-cost subsistence model since the food l i s t used was chosen with consideration of each item's palatability.  Lower-cost, unpalatable  items may not have been included in the calculations.  The food l i s t used  obviously effects the character of the solution obtained.  86  A recent reflection on the low-cost diet, called the "three-consideration diet" (for the three statements of nutrient allowance, food composition, and food prices), has been introduced by Lewis and Peng (1977). Daily diets were computed for each member of a four-person reference family using the 1974 U. S. Recommended Dietary Allowances, current food composition data, and Bureau of Labour retail price statistics for Atlanta, October 1975.  The authors indicate that the diets produced are less than  satisfactory with respect to variety and palatability, although they are somewhat more varied than earlier models.  The additional nutrient constraints,  17 altogether, have resulted in diets of 7 or 8 items which includes vegetable oils and grain products in addition to other food groups.  In com-  1  parison Stigler s solution contained no animal fat or vegetable o i l , and the Beckman diet did not contain grain products. Some authors have focused on the application of linear programmed least-cost diet problems for field conditions rather than for illustrating the mathematics of linear programming.  Florencio and Smith (1969; 1970)  have utilized least-cost diet methods to measure the efficiency of food purchasing among working-class families in Columbia.  Efficiency was  determined by an index which compared the cost of actual foods consumed with a mathematically optimized least-cost diet which was developed from commonly consumed foods.  Least-cost models have also been used by  Chamberlain and Stickney (1973) and Kansra ejt al_. (1974) for development of least-cost nutritionally balanced  multimixes suitable for children in  developing countries. All of the models presented above, and other diet models to be presented below, utilize linear nutrient constraints which are specified in absolute amounts, or where nutrient interdependences exist, as fixed  87  ratios or exact constants.  Smith (1974) has developed a food planning model  which modifies the standard formulation of dietary constraints by the introduction of nonlinear constraints for determining protein allowances.  The  model considers simultaneously the effect of both protein concentration and protein composition on protein allowance in determining the most economical diet which satisfies protein needs and other nutrient requirements.  Utilizing the nonlinear protein constraint provides an opportunity  to economize by consuming smaller quantities of higher quality protein or larger quantities of lower quality protein while meeting protein requirements.  The problem is solved using separable programming to obtain linear  approximations to nonlinear restraints. Smith's (1974) model was designed for national food planning problems where efficient use of protein resources is required, for example in the developing countries, Carmel (1976) has developed a model of similar nature and application using the concept of NDpCal percent to account for the relationships between energy content, protein quantity, protein quality, and protein value of diets.  Diet planning with variable nutrient coefficients  has been discussed by Armstrong and Balintfy (1975).  2.4.1.1.2  Food Planning Models with Palatability Considerations  Comparison of the results of purely nutritional models with low-cost diets prepared by nutritionists illustrates discrepancies in dietary cost and composition due to factors beyond those explicitly considered in the modelled diets (Smith 1963).  As Calavan (1976) states:  The solution to this prescriptive food selection model typically consists of an unpalatable, boring combination of three to seven food items. Except for populations on the verge of starvation, such solutions are optimum only in an arbitrary mathematical sense (pp. 65-66).  88  Consequently, many later diet models have included explicit statements of palatability in their formulations, in attempts to produce diets which are more congruent with actual consumption patterns.  Both linear and non-  linear models have been developed. In an early example of formulations considering palatability, Brown s  (1954) attempted to develop a descriptive programming model which would produce a diet for one person similar to the actual diet of the British working class.  As Calavan (1976) indicates, this objective was different  from that of prescriptive models which attempt to define nutritionally adequate diets. Three i n i t i a l models tested by Brown were similar to the nutritional models which have been previously discussed.  In these models prediction  of consumer behaviour was based on the hypothesis that the consumer's objective was to select the least-cost diet which met specified minima of twelve food composition factors.  These factors corresponded to nutrient  restraints established by population practices and not to physiological requirements identified in dietary standards.  The computed diets were  derived from a basic l i s t of 15 food groups averaged for price and nutrient composition.  The solutions obtained consisted of from five to eight groups  for the fifteen food groups.  Not unexpectedly, Brown found that as more  nutrient restraints are included, the diet becomes more varied and more expensive. In Brown's fourth model an additional objective of consumer behaviour was included.  This was an explicit restraint on the levels of bread and  potatoes consumed.  Brown's most sophisticated model provides a weekly diet  for one person which includes eight of the fifteen groups consumed by the population.  He judged the fourth model as adequate because it selected  89  foods in all the major groups used by the population, except f r u i t , and because the calculated weekly expenditure is only about twelve percent lower than actual population expenditures. The major difference between the formulation of Brown's models and earlier nutrition models is the use of quantity restraints on some items. This concept has been further extended by Smith (1963).  Smith has developed  three food planning models beyond the midget model which was previously discussed (p. 85).  These are; the midget model with cooking aids, the  small model, and the large model.  These models utilize an objective function  of cost minimization, and explicit palatability constraints designed to provide solutions which conform to conventional consumption patterns. Smith's most sophisticated model, the large model, provides an inexpensive, nutritious, and reasonable palatable diet of 62 items for a family of three over a four week period.  Unlike the other Smith models  (and those of Stigler, Beckman, Vajda, and Brown) which have a limited number of food classes, the large model utilizes an expanded commodity l i s t of 572 widely consumed individual food items and narrow commodity classes. In the large model, item acceptability and dietary variety depends primarily on the use of a large number of constraints on the amounts and associations of foods included in the diet, rather than on a restricted commodity l i s t as in earlier models.  In fact many of the items in the  expanded commodity l i s t are unpalatable raw materials or ingredients which become acceptable only by their association with other ingredients in the diet. The 85 commodity constraints used in the large model include ten complimentary constraints which define proportionalities among different items, 28 maximum quantity and 41 minimum quantity limits on the consumption  90  of certain foods, and six requirements for specified amounts of some commodities. traints.  Only the large model uses all four types of commodity res-  Additionally, thirteen nutritional restraints are defined.  Twelve identify minimum acceptable intakes while the thirteenth provides a maximum total caloric intake for the diet.  Thus, a total of 98 restraints  have been defined in the large model. The commodity constraints protect against excessive amounts of some items, and against exclusion of important items which are common in the usual diet or which are required to make some ingredients palatable.  The  constraints are in most cases based on exact or adjusted figures derived from the consumption patterns of populations, although in some cases arbitrary or experimental values are incorporated. Calavan (1976) used a linear programming approach to develop a descriptive model of food selection practice which would be appropriate for research on the epidemiology of malnutrition.  Data on the socio-economic  variation of food-use practice in a northern Thai village was used to identify indices or goals around which the residents optimized dietary behavior.  The identified goals of the population were to satisfy energy  requirements, maximize dietary variety, maximize intake of animal foods, and maximize fat intake.  A tentative model was constructed utilizing a  series of linear programs which, while staying within budgetary limits, selected foods to satisfy a specified sequency of the above householder dietary goals. Comparison of model-generated and household-generated food l i s t s was the intended evaluative format, however, insufficient data were available for field testing.  Instead, less conclusive tests of comparison indicated  that the goals adopted in the original model were consistent with  91  population data on socio-economic variations in dietary behaviour, provided the sequency of these goals was slightly modified. Various nonlinear food models have been developed.  The Consumer and  Food Economics Division of the U. S. Department of Agriculture has developed a nonlinear programming model to aid in adjusting their family food plans to coordinate with food price fluctuations, changes in established nutrient allowances, and changes in eating habits of the population (Balintfy 1976a). The model provides an adjusted food plan which is nutritionally adequate and which approximates food-group consumption patterns for each of 22 sex-age groups .and income levels established by the 1965-66 Household Food Consumption Survey.  The,precise formulation is as follows;  Let q.j denote the past consumption of food quantity i and x^ the corresponding quantity in the new food plan.  The optimal food  plan is thus formulated as the following quadratic programming model: n 2 £ w.(q. - x.) , i=l 2  minimize  1  1  1  subject to Ax > b, Rx > d, where w. are weights to equalize the relative contributions of deviations, and where A is the matrix of food cost and nutrient composition data for up to 17 food groups and 18 nutrients per group.  The R matrix represents a set or upper and lower bounds  as well as proportionality constraints imposed on the components of the solution vector to assure strictly positive and acceptable food quantities Balintfy 1976a, p. 328). In 1958 Wolfe developed a nonlinear food planning model which reduces dietary monotony while ensuring economy (Smith 1963).  The model incor-  porates a quadratic index of d i s u t i l i t y or "fatigue" in the ojbective  92  function which is based on the assumption that excessive consumption of any food would cause disutility proportional to the square of the quantity consumed.  The quadratic expression ensures a disproportionately large  penalty at higher levels of consumption for any item and therefore limits the total intake of any item. Wolfe!smodel is formulated as follows: Let n be the number of foods, p. the unit price of commodity j , x- the quantity of commodity j to be consumed, m the number of restraints in the model, b. the quantitative requirement set by the i  restraint,  a., the quantity of nutrient i contained in one unit of commodity j , J +h and f. the "fatigue" or disutility function for the j food. Minimize 1  z for all of the arbitrary numbers P between 0 and 6.8077, where n n z = P Z-PiX,- + z f.x, (j = 1, 2, 3, n) j=l j=l 2  J  J  3  subject to (1)  3  x. > 0  n (2) E a. .x. * b. j=1  (i = 1, 2, 3, . . . , m) (Smith 1963, p. 30)  In the above equation, P.is an arbitrary weighting factor which determines the emphasis to be given to dietary economy.  At P = 0 there is no cost  consideration, whereas at P = 6.8077 cost considerations are dominant as in the classical cost minimization linear solution, weighted cost  -  z is the total of the  of the diet plus the index of d i s u t i l i t y .  To test the model Wolfe used data from the Stigler model for the nutrient requirements and for cost and nutrient coefficients. items from Stigler food l i s t were used.  Only twenty  Wolfe found that as the emphasis  on economy was decreased, ;the solution developed from the basic Stigler solution of five items to incorporate all twenty items available in the abridged food l i s t .  The cost of the diet rose accordingly.  93  In a further nonlinear example, Balintfy (1976a) discusses a model, called the "weightwatcher.'s quadratic diet model", which contains a nonlinear objective function in its formulation.  The model functions to  maximize the total dietary u t i l i t y or "preference" while meeting cost and caloric constraints.  The program uses data on the estimated quadratic  u t i l i t y functions for sixteen food groups, and on the cost and caloric content of each food group.  Nutrient considerations are not extended  beyond the caloric content of the diet since, it is assumed, protein needs will be satisfied by the strong u t i l i t y of meat and milk products in the objective function, and other nutrient needs can be satisfied by supplementation.  Presumably the model is not ultimately restricted to this  limited nutrient domain.  2.4.1.2  Menu Planning Models  Menu or meal planning is a decision process of defining the serving sequence and/or serving frequency of prepared foods, called menu items, such that required nutritive, production, economic, and palatability constraints are satisfied  (Balintfy 1973b; Balintfy 1976a).  Two approaches  to menu planning by computer have been reported - - a non-mathematical method called the "random approach" (Eckstein 1967) and a mathematical modelling approach which has evolved to include a variety of linear and nonlinear programming models (Balintfy 1976a).  2.4.1.2.1  Menu Planning Models - The Random Approach  The random approach to menu planning (Eckstein 1967) attempts to simulate the routine decision making processes used by dieticians in planning menus. The menu is compiled by randomly selecting items from a variety of meal component classes and sequentially evaluating these items against predeter-  .94  mined acceptability criterion used by dieticians.  These criteria include  colour, texture, shape, flavour, caloric content and cost of the food item, as well as other acceptability factors.  Items from each meal com-  ponent that satisfy the criteria are included in the solution. The random approach is based on the concept of bounded rationality - a decision process where acceptable solutions are reached without considering all the possible alternatives.  This approach to menu planning has  been largely supplanted by mathematical programming methods which, by contrast, produce optimal solutions based on considering all the possible alternatives (Balintfy and Nebel 1966).  2.4.1.2.2  Mathematically Programmed, Multistage, Menu Planning  Menu planning as a mathematical programming problem was f i r s t identified and solved by Balintfy (1963; 1964).  This first approach used a multistage  decision rule which planned an optimal combination of menu items for a sequence of meals by considering the problem of scheduling on a sequential meal-by-meal, day-by-day basis.  This model has been formulated in both a  linear programming version (Balintfy 1966; Neter and Wasserman 1970) and as an integer programming version which approximates the theoretical solution to the problem (Calintfy 1964; Gue and Liggett 1966). This f i r s t model utilized menu-item quantities of fixed-portion size as decision variables to produce a nonselective menu.  Such a nonselective  menu is obtained by assigning one menu item to each of several menu-item classes or courses (for example, appetizer, entree, cereal, bread, and beverage for breakfast courses) over a defined period of time.  Gue and  Liggett (1966) have extended the concept of menu-items as decision variables to fixed-choice groups of two or more items, and thereby developed  95  selective menu planning model.  Selective menu planning allows a choice  of items from each menu item class.  This concept has been further extended  to selective menu planning from variable-choice groups by Balintfy (1971). A refined model based on the above approach was converted into a selfcontained food services information processing and menu planning computer package, called "System/360 Computer Assisted Menu Panning" or "CAMP" (Balintfy 1969), for use in institutional food planning in the public domain.  The system includes a systemmatic approach to data collection and  data management, as well as to menu planning (Balintfy 1975). The menu-planning objectives, which define optimality in the CAMP system, are to determine the least-cost combination of menu items for a sequence of days which meet nutritional and acceptability requirements Balintfy (1975).  Item acceptability and dietary variety are provided by structural,  separation, and attribute constraints which ensure compatitiTity of items between meals and within meals.  Compatibility within meals is provided by  use of attribute codes which restrict the entry of items of similar attributes from appearing more than desired in any meal. ments on the structure of the menu are imposed.  Further, formal requireThe structural require-  ments partition the menu into a customary array of menu components for each meal, to which only appropriate menu items can be assigned.  Compatibility  between meals is achieved by separation constraints which indicate the desired serving frequency.  The separation constraint defines a minimally  elapsed number of days between consecutive rescheduling of the same or similar items.  Other constraints on proportionality of items, and on  production requirements are also included.  The constraints used in CAMP  are based upon the expressed preferences of patients and upon institutional policy.  96  CAMP has provisions for selective and nonselective menu planning (Balintfy 1975).  Selective menu planning is intended to further enhance  menu acceptability.  Although structural, attribute, and separation  constraints are operative in selective menu planning, nutritional constraints are not.  It is reasoned that is nutrient-assured meals are required the  nonselective f i r s t choice must be accepted. Multistage menu planning has evolved to include procedures for variable portion-size and chance-constrained modelling.  Armstrong and Sinha (1974)  have developed a quasi-integer programming algorithm to plan nonselective menus in which the portion size of the menu items can vary over a specified positive range.  This is an advance over earlier systems in which fixed  portion size was a technical necessity.  Another advance is demonstrated by  Balintfy (1976a) with the introduction of random variables as constraints. This particular application considers the nutrient content of any item in probabilistic terms and the constrained solution as meeting nutrient requirements individually or collectively with a specified probability. Multistage menu planning can also incorporate recently developed nonlinear preference maximization objectives which utilize some measure of consumer satsifaction  (Balintfy 1976a).  Typically cost-minimization  objectives have been used since adequate quantitative measures for consumer preference were absent, and since earlier modelling objectives were chosen to demonstrate the economic impact of mathematical optimization as opposed to conventional methods (Balintfy 1973b).  The use of preference maximization  objectives has been demonstrated in single-stage menu-planning models to be di scussed. Multistage menu planning has been implemented in a variety of institutional food services including hospitals, colleges, penal and mental  97  institutions.  Applications of the CAMP system (Bowman and Brennan 1969;  McNabb 1971; Gel pi et al_. 1972; Balintfy 1975) have demonstrated savings of 5 to 34 percent in raw food costs, improvements in nutritive control, and equivalent acceptability and variety standards as compared to conventional techniques. and Nebel 1966)  Prototypes of the CAMP system (Balintfy 1964; Balintfy  and other multistage systems (Gue and Liggett 1966)  demonstrate similar advantages.  2.4.1.2.3  Mathematically-Programmed, Single-Stage, Menu Planning  Menu planning can be done in a single stage when the frequency of occurrence of menu items over a finite time period is to be determined and not the specific sequence of items in the menu is to be determined (Balintfy 1973b; Balintfy 1976a).  Unlike multistage menu planning which provides a  consecutive sequency of integer solutions for each of the smallest periods within the planning horizon, single-stage menu planning offers only one mathematical solution for the entire cycle (Balintfy 1966).  The common  shortcoming of single-stage solutions is that computer scheduling on a meal-by-meal basis is s t i l l required to provide a sequence of daily meals (Balintfy 1976a).  The difficulty of scheduling for a fixed-time horizon  in a single stage is discussed by Balintfy (1974a). The f i r s t models incorporating single-stage menu planning (Balintfy 1966) inherited the cost-minimization objective and linear constraints on nutrients, structure, and attributes.  The concept of minimum separation  of items used in multistage scheduling, as a safeguard for variety, was used in single-stage models to establish upper bounds:onrjthe frequency of occurrence of items.  Although the use of these upper-bound constraints  was a crude method of maintaining preference levels, studies in a number of situations indicated single-stage menu planning models produced diets  98  which met nutrition and acceptability levels while reducing costs (Gelpi et a l . 1972; Balintfy 1976a). With the discovery of time-related food-preference functions (Balintfy et al.197 4) and time-related food-preference and quantity functions (Balintfy 1973b) i t became possible to optimize the quantity and frequency of food intake based on an empirical measure of food preference or u t i l i t y (Balintfy 1976a).  Recent single-stage programming models have incorporated nonlinear  objective functions which maximize measures of total food preference subject to given nutrient, cost, structural, and assorted attribute, proportionality, production, and other constraints - - constraints similar to those in CAMP (Balintfy 1974b; Balintfy 1976a) An alternative non-linear programming formulation with the preference quantity functions has been proposed (Balintfy 1976a).  This model maintains  the linear cost-minimization objective normally used in institutional menuplanning models, and incorporates a nonlinear preference constraint which maintains a given food-preference  2.4.2  level.  Information and Behavior Change  As indicated at the outset, the task of nutrition educators is both to provide a code of nutritional practice and to communicate this available knowledge to the public sector.  One view is that this process is one of  planned change, where planned change is " . . . a conscious effort to alter food-related practices or attitudes when the need exists" (Gifft et a l . 1972, p. 255).  Thus, the ultimate objective of nutrition education  programs, according to this approach, is to modify food behaviour through deliberate intervention.  99  2.4.2.1  Development of Food Behavior and Factors in Food Selection  Food selection behavior, as with other human behaviors, is the product of a complex interaction of situational and developmental variables, and of individual and environmental variables. is dependent of two factors:  Situationally, food selection  food availability - - the environmental  variable - - and food acceptability - - the individual-related variable (Gifft et_ al_. 1972).  The availability of food in the marketplace and home  is of paramount importance in food selection behavior.  Availability is  governed by climatic and geographic factors, by economic, political and technologic factors, and by public policy and individual decisions (Gifft et al_. 1972).  A food's acceptability, in turn, determines which of the  available items will be selected and eaten.  Item acceptability for the  individual is determined psychologically through motivations or needs, such as biogenic (sensory) needs, psychogenic (emotional) needs, and sociogenic (goal) needs; and through cognitions such as ideas, attitudes, and beliefs (Lund and Burke 1969). Developmentally, the psychologic elements of an individual are rooted in the unique interaction of his or her biological nature and socio-cultural factors (Gifft ejt aj_. 1972).  The intricate process of this social and  emotional acculturation provides the food-related experiences which contribute to the development of food patterns, whereas the biological heritage determines the physiological needs and capacities, and the potential psychologic and sensory structures which will interact with the environment to create a person's food pattern. Consequently, each individual has a characteristic pattern of eating which has developed over a lifetime as a result of many complex processes and influences, and which has an integral part of that individual's total  100  behavior (Gifft et al_. 1972).  This pattern is resistent to change, espec-  ially complex change, unless immediate benefits are evident, or a change is forced by circumstance (Gifft et_ al_. 1972).  Much is vested in a par-  ticular pattern and the tendency is to move towards familiarity or to reinforce what is already known (Thompson 1969; Gifft et_ al_. 1972). Future gains are obscured by the immediate rewards of not changing.  2.4.2.2  1  The KAP Gap  Food practices do not change just because people have accurate facts about nutrition (Leverton 1974).  Numerous studies in nutrition education  (Hampton et al_. 1967; Baker 1972; Bell and Lamb 1973)and other fields of applied education (Young 1967) have indicated the incongruity of different aspects of behavior.  Material taught may not be learned, once learned  i t need not be believed, and even i f a change in attitude did occur '" practices would not necessarily be altered.  The reciprocal is also true.  Behaviors are not based only on particular types of knowledge.  In fact  a person may not be able to justify his or her actions and beliefs to the satisfaction of others.  For example, individuals may not be able to  provide valid nutritional reasons for their nutrition practices (Emmons and Hayes 1973). Although evidence of the effect of nutrition knowledge on food practice is limited and conflicting, some evidence does suggest that acquisition of formal nutrition knowledge is positively related to behavior change in  1.  For further discussion of the knowledge - - attitude-practice (KAP) gap see Travers (1963)  101  individuals (Young et a]_. 1956 a+b; Hinton et_ al_. 1963; McKenzie and Mumford 1965; George 1971; Boysen and Ahrens 1972). not a linear relationship (Gifft et al_.  Admittedly this is  1972) indicate, studies comparing  ;theinfluence of formal nutrition education on food practices are difficult to perform and interpret due, in part, to the conflict of other variables which may simultaneously effect behavior.  These include factors such as  emotional stability and maturation age (Hinton et_ al_. 1963), informational sources besides the nutrition-intervention technique (Rosenstock et a l . 1966) , and early learning and food patterns (Litman et_ al_. 1964; Brown 1967) . That nutrition education is able to influence the eating habits of populations, is illustrated most notably by the successful campaigns of various food companies (Tyler 1962; Gussow 1972; Manoff 1973) and other mass media campaigns (Rosenstock et_ al_. 1966).  In these cases, most  campaigns were aimed at influencing product choice for reasons other than nutritional value.  Therefore, there is some reason to believe, provided  similar resources are available, that eating habits can be influenced for nutritional reasons (Manoff 1973). In any case, accurate facts are essential  for rational decisions, even  though information on proper diet may not prevent, and may even result in contrary, incorrect, or exacerbating behavior.  Information or messages are  a source of guidance in food pattern development and redevelopment, and thus form a tool through which the nutrition educator can promote planned change (Gifft e t e H .  1972)  102  2.4.2.3  Increasing the Effectiveness of Communication  Although human behavior is complex and difficult to analyze, predict or manipulate, research has provided some principles and guidelines to increase the efficacy of the educational process, and thereby increase the chance of individual compliance with recommendations.  However, precise  predictions of the efficacy of using these principles are not possible (Gifft et aJL 1972). There are many approaches to and l i s t s of principles for enhancing the effectiveness of educational programs (Gifft et_ al_. -1972).  The underlying  theme of these principles is to enhance the learner's receptivity to the message by: (i)  increasing the learner's interest through: (a)  incentives such as teacher and program credibility, individualized involvement, and advertized benefits; and  (b)  appropriate exposure to the message, for example, by enlisting the learner's active involvement in the learning situation, changing process demands, keeping messages short to ensure active attention, and emphasizing thinking versus recall activities  (ii)  (Gifft et a]_. 1972).  increasing the learner's receptivity by constructing a message of maximum potential meaning for the learner (Gifft e_t a]_. 1972)  t  In eithertcase the focus is on the learner's relationship to the information rather than on the message alone.  The present concern is with the construc-  tion of a message with maximum potential meaning for the learner, and not with the development of other aspects of the teaching-learning process> used to increase the learner's receptivity.  103  Change occurs when the information provided is significant enough to the learner to motivate action (Gifft ejt al_. 1972).  It is axiomatic, in  the field of communication theory (Berlo 1960), that response to bids for change are governed by the ratio between the anticipated benefit and the energy required to respond. put -out,  The more reward, the more effort which will be  and the less reward the less the effort.  Therefore, the  potential response to a message can be increased by increasing the reward or decreasing the effort required, or both. Expected benefits can be increased by choosing messages relevant to the learner's interests, perceived needs, and concerns (Gifft et al_. 1972; Leverton 1974). Effort to respond to recommended alterations in food practices can be decreased by gearing information to the physical and mental s k i l l s , cognitive sets, attitudes, resources, and emotional readiness of the receiver.  i  For example, information can be adjusted to the individual 's:  physiological needs; psychological barriers such as values,  attitudes,  beliefs, likes and dislikes, and pursuits such as, family responsibilities, professional demands and leisure-time activities  (Gifft et_ al_. 1972).  An important consideration in decreasing the effort necessary to respond is complexity of change.  Since people's food patterns remain  relatively stable and are resistant to change, especially complex change, recommendations for change should avoid unnecessarily complex demands. Complexity of change has been suggested as perhaps " . . . the strongest determinant of the speed and extent of adoption" p. 265).  (Gifft e_t al_.  1972,  Thus, nutrition education programs are more effective when  emphasis is placed on the maintenance of present desirable habits and the improvement of current food patterns rather than on radical alterations  104  in diet (Todhunter 1969).  For example, persuading a person to eat more  of a food he or she already eats may be less  complex and therefore easier  to accomplish than inducing him or her to use a food which has neven been tasted (Gifft ejt aj_. 1972).  Similarly, the individual should not be  overloaded with information (Leverton 1974).  105  CHAPTER 3 DEVELOPMENT OF THE PROTOTYPICAL SYSTEM  3.1  Introduction The project goal (p.13) was to develop a prototypical system to provide  information on dietary practices for those adults who want to apply nutritional principles to their eating habits, and have sufficient  resources  (eg. time, energy, education, money) to make use of the information which defines healthful dietary practices for them.  This development was under-  taken bearing in mind two primary difficulties, coincident with nutrition education's aforementioned tasks (p. 1), namely:  the problem of developing  nutritional guidelines suitable for health promotion in the public sector, and the problem of communicating this information to individuals.  3.1.1  Developing Nutritional Guidelines With respect to the issue of developing a message which incorporates  accurate nutritional guidelines, standard dietary assessment procedure of data collection, analysis, and evaluation have been discussed (Section These appear to be suitable for the prototypical system's design.  2.3).  Guide-  lines for evaluating diets can be derived from the dietary standards of various countries and international agencies, and the dietary goals proposed by recognized scientific agencies. considered despite certain problems, namely:  This procedure has been considerable conflict exists  about the formulation and application of dietary standards; dietary assessment alone does not provide the means to establish nutritional status; and "optimal" nutritional status cannot presently be defined.  106  3.1.2  Communicating Information to Individuals With respect to the second problem, that of communicating this  information to individuals with the ultimate objective of changing food practices, the use of educational principles was considered relevant to the system's development.  The focus has been the development of a message  for motivated individuals - - in this instance a statement about what to consume  which facilitates adoption of recommendations.  two features of the message are significant, namely:  In this context,  the comprehensibility  of the information, and the acceptability of the suggested changes.  3.1.2.1  Comprehensibility of the Information  Education principles indicate that the comprehensibility of information recommending change in eating behavior can be increased by not overloading the client with unnecessary information.  In particular, recommendations  for change in eating habits should not require extensive alteration in the person's perception of his food environment and eating habits.  The type  of information typically presented in nutrition education programs is either an explicit statement about foods to consume, for example as a daily menu, or alternatively an implicit statement, for example as a daily nutrient allowance.  Although the presentation of nutrient information  may provide a clear rationale for food selection, nutrient information becomes functional only when translated into foods and meals.  Thus, a  statement based on foods to consume should be easier to comprehend than one based on nutrients. In developing such a food-based statement, the emerging problem is to accurately translate nutrient requirement data into a viable food plan. The problem arises because the nutrient contribution of each food with its  107  characteristic nutrient pattern must be considered in developing the diet plan.  As the number of nutrients under consideration increases and the  restrictions on acceptable solutions become more severe., the accounting problem becomes formidable.  Mathematical-programming techniques, based on  experimental menu-planning procedures for food service applications and on a variety of programs with industrial applications, are useful in overcoming these problems.  They provide for simultaneous satisfaction of many variables,  including nutritive, production, economic, and palatability constraints. This method may be used for nutrition education purposes to resolve the problem of accurately translating data on nutrient requirements and food composition into a viable food plan.  3.1.2.2  Acceptability of Suggested Changes  The acceptability of suggested changes can be increased by limiting the complexity of change.  Complexity of change has been suggested as  perhaps the strongest determinant of the speed and extent of adoption. Thus, nutrition education programs are more effective when emphasis is placed on the maintenance of presently desirable habits and the improvement of current food patterns rather than on radical alterations in diet. Factors which may contribute to the perceived complexity of these changes include:  budgetary considerations, socio-cultural patterns, foods  available, food habits, taste preference, colour preferences, likes and dislikes, needs and interests, and beliefs about foods.  In short, any  deviation from the characteristics of a usual or desired food pattern may contribute to the perceived complexity of change. Hence the best approach for compliance with nutritional recommendations should be a diet which deviates as l i t t l e as possible from a dietary inventory  108  identified as desirable either by past consumption  or stated preference.  Presumably, this would be similar to the diet which the individual would choose i f she or he understood, accepted, and used nutrition knowledge. Using the mathematical-programming techniques previously mentioned, a constrained-optimization algorithm can be formulated to find the combination of foods which are similar to an individual's actual or desired food plan while simultaneously considering nutritional, budgetary, and palatab i l i t y requirements, or for that matter, any other measurable vector of food or food-selection  2.  behavior.  Usual food pattern may provide a better monitor of actual preference, being the diet chosen under present budgetary restrictions and long term cultural and personal habits.  109  3.2  System Design and Characteristics The prototypical system, as illustrated below in Figure 3.1,  is  designed both to assess and plan diets by systematizing the procedures used in dietary analysis and counselling by nutritionists and dietitians. Input data from both a client questionnaire and computer files are processed within the two phases of analysis-evaluation and planning.  Within this  framework, the nutrient characteristics of the client's i n i t i a l diet can be analyzed and evaluated, and a diet plan can be produced which should remain close to an individual's actual or desired food plan while simultaneously meeting specified nutrient limits. is an output statement of:  The outcome of this procedure  the client's i n i t i a l diet; a recommendation of  altered food intakes - - the revised diet;and an analytic and evaluative statement of the original nutrient intake.  INPUT DATA OUTPUT STATEMENTS Client Questionnaire client's i n i t i a l diet client's demographic data Computer Files food-item f i l e food-composition f i l e  DATA PROCESSING Analysis and Evaluation  Planning  client's i n i t i a l diet analysis and evaluation of client's i n i t i a l diet client's revised diet  nutrient-limits f i l e attribute-group . matrix Figure 3.1  Overview fl owchart of prototypical system for assessment and planning of individual's diets.  no  3.2.1  Data-Collection  3.2.1.1  Client's Initial Diet  The client's i n i t i a l diet includes those food items and their quantities, selected from the system food-item f i l e (Appendix A), which the client 3 consumes habitually.  This information is used both in diet-assessment and  in diet-planning, since i t provides the basic data with which food-composition analysis and evaluation can be performed in diet-assessment and with which the client's diet and dietary-structure can be defined in the planning phase.  The client's i n i t i a l diet is stored in the system during processing  and compared with the revised diet on the output presentation. The client's i n i t i a l diet is determined by using a multiple-purpose intake questionnaire (Table 3.1) which permits use as a one-day r e c a l l , weekly record, long-term food frequency history, or other variants.;  Items  can be quantified by weighing, by household measures using standard portionsizes or measured portions, or by estimation.  The questionnaire can be  self-administered by the client, or used in an interview format for supervised inquiry.  The individual and/or institution can select the most  appropriate data-col lection^procedure for their purposes.  Flexible data  collection procedures may be more useful for meeting the variety of client requirements, program capabilities, and other situational factors.  3.  Alternatively, the diet can represent the desired regime or a diet previously designed in collaboration with a nutritionist.  -i  Ill  Table 3.1  Excerpt from prototypical intake questionnaire - - dietary intake format:*  FOOD ITEMS AND DESCRIPTION  ENTREE ITEMS -- DAIRY AND EGGS AMERICAN PROCESSED CHEESE. BLUE o r ROQUEFORT CHEESE. CHEDDAR, JACK o r SWISS CHEESE COTTAGE CHEESE, creamed o r uncreamed, any c u r d . CREAM CHEESE o r CHEESE SPREADS. SOUR CREAM o r CHIP DIP. YOGHURT, whole m i l k base. YOGHURT, skim m i l k base. YOGHURT, p a r t - s k i m m i l k base. EGG: raw, b o i l e d , poached, f r i e d (add f a t ) . EGG: scrambled, omlet, s o u f f l e , spoon bread. ENTREE ITEMS — CEREALS CORN CEREAL, e n r i c h e d , r e a d y - t o - e a t . WHEAT CEREAL, e n r i c h e d , r e a d y - t o - e a t . WHEAT CEREALS, more r e f i n e d , e n r i c h e d : cooked. OATMEAL, a l l t y p e s : cooked. WHEAT CEREALS, l e s s r e f i n e d : cooked. PANCAKES, WAFFLES o r FRITTERS: made w i t h m i l k and eggs NOODLES, egg-type, e n r i c h e d : cooked. SPAGHETTI, MACARONI o r NON-EGG PASTAS, e n r i c h e d : cooked RICE, brown: cooked. RICE, w h i t e , e n r i c h e d , u n e n r i c h e d o r p a r b o i l e d : cooked CORNMEAL o r CORN GRITS, e n r i c h e d : cooked. WHEAT GERM. FRENCH o r SOURDOUGH BREAD, e n r i c h e d : f r e s h o r t o a s t e d . RAISIN o r RAISIN-NUT BREAD, e n r i c h e d : f r e s h o r t o a s t e d . ETC '.. ETC  *  AMOUNT PER .UNIT SERVING  CONSUMPTION  STANDARD PORTION _SK£  DAY  ADJUSTED PORTION SIZE  FREQUENCY  SERVINGS PER WEEK MONTH  V ' - V ' - I V (1 oz.)  r-r-iv d oz.) r-r-iv (i oz.)  H cup (4 oz.) 2 1 1 1 1 1 2  t b s p . (1 oz.) tbsp. cup cup cup egg eggs  1 cup (1 oz.) 1 cup (1 oz.) H cup h cup k cup 2 a t 4" d i a . h cup h cup H cup H cup H cup 3 t b s p . (1 oz.) 1 slice 1 slice  A complete l i s t of foods and portion sizes is contained in Appendix A.  112  The questionnaire food items, contained in the system food-item f i l e (Appendix A), were derived from the "Mini Food List with Food Substitutions" developed by Pennington (1976).  Pennington's Mini Food List contains a  table of 202 commonly-consumed American food-items, called index items, and a comprehensive l i s t of 49 compositional values for each of these index items.  Each index item represents one-or-more substitutable items  classified according to the coexistence of the 49 nutrient values. In order to develop the system's food-item f i l e , 200 of Pennington's 202 index items were selected.  This i n i t i a l l i s t was elaborated to  incorporate most of the substitution items within their pre-assigned groups.in order to f i t the system's attribute-group matrix.  The food item  f i l e (Appendix A) developed contains 221 item clusters, where an item cluster is composed of either a single item or two or more exchangable items of similar nutrient characteristics. An abridged food-item f i l e (Appendix B) of 127 item clusters was developed for testing the system. The item clusters have been categorized within a hierarchical structure of groups and subgroups used for planning purposes, called the attributegroup matrix (Appendix G).  Each of the item clusters was assigned an  attribute-group code number and an item cluster code number which appears with the item in Appendix A and B.  The makeup of this hierarchy is discussed  further in the material on diet-planning (p. 121). To aid in quantification of questionnaire items, the description of portion sizes in the system food-item f i l e has been elaborated beyond that available from Pennington's text.  Food Portions Commonly Used (Church and  Church 1975) and Agriculture Handbook No. 456 (Adams 1975) were used in this extension.  This procedure was not in every case free of difficulty as noted  113  in Appendix A, since the descriptions .of standard portions per gram quantity, used by the three authors, were considerably different in some instances. Manual computer input of the questionnaire data was used.  Alternative  input methods such as interactive terminal interviewing or computer read cards, although perhaps preferable in on-going use required for present purposes.  3.2.1.2  Client's Demographic Data for Defining Nutrient Limits  Data on the individual's sex, age, size, activity pattern, and pregnancy status are collected on the questionnaire (Table 3.2).  This  information is used in the assessment and planning phases of the program to define the nutrient limits appropriate to the individual.  n.4  Table 3.2  Excerpt from prototypical intake questionnaire - - client demographic data.*  AGE (in years)  IDEAL BODY WEIGHT (in pounds)  SEX and PREGNANCY STATUS (check one) male  female not pregnant female pregnant 1st trimester female pregnant 2nd or 3rd trimester female lactating  ACTIVITY PATTERN (select pattern appropriate for sex from table below and check one) A B C D Types of Activity 1) Resting metabolism 2) Sitting or standing s t i l l 3) Walking slowly Light domestic work (eg. ironing, sweeping floor, cooking, dishwashing, dusting) Light office of industrial work (eg. typing, lab. work, sewing, printing, garage mechanics, machine-tool operation) Sports involving light activity (eg. bowling, golf, sailing) 4) Walking at moderate speed Moderate domestic work (eg. scrubbing floor, window cleaning, furniture polishing) Moderate industrial work (eg. painting, plastering, brick-laying, modern farming) Hobbies with moderate activity (eg. gardening, woodwork, dancing) Active sports (eg. tennis, cycling, skiing, gymnastics, swimming) Total *  Act. Pat.(men ) Act.Pat. (women;) A • B C D A B C D ' No. hours/day No. hours/day 8 8 8 9 8 8 8 9 10 10 12 13 11. 11 12 13  4  5  4 2  4  5  4 2  2  1  0 0  1  0  0 0  24 24 24 24  24 24 24 24  Adapted from the Dietary Standards for Canada (Canada 1975).  115  3.2.2  Data-Analysis  Following collection of dietary data, the client's i n i t i a l diet is analyzed for nutrient values.  Nutrient data used for this analysis and  for generating the new diet plan are contained in the system's foodcomposition f i l e (Appendix C). The food-composition f i l e is stored in the system for retrieval during operation.  It is derived from Pennington's Mini Food List (Pennington  1976), and includes 41 composition values in nutrients per 100 grams of edible food portions for each item cluster in the food-item f i l e .  The  nutrient values contained are for total calories; protein and eleven amino acids (tryptophan, threonine, isoleucine, leucine, lysine, methionine, cystine, phenylalanine, tyrosine, valine, histidine); total, saturated, and polyunsaturated fatty acids; cholesterol; total carbohydrate; sucrose; fiber; twelve vitamins (thiamin, riboflavin, niacin, pyridoxine, folate, cobalamin, ascorbate, pantothenate, biotin, retinol, cholecalciferol, tocopherol); and nineminerals (calcium, phosphorus, magnesium, iron, iodine, zinc, sodium, potassium, copper).  It was assumed that the values  obtained from Pennington's publication are applicable to the Canadian marketplace. The composition values used are specific for processing and preparation effects outlines with the item description.  Consequently, recipes  using raw items cannot necessarily be calculated from baked items due to changes in weight with cooking.  Further Pennington (1976) suggests that:  when possible, fresh cooked items should be substituted by fresh cooked; canned by canned; and frozen cooked by frozen cooked. This will prevent large errors in water-soluble vitamins, sodium, and vitamin-E (p. 16). The sodium content of items is without either salt added at the table or  116  salt added in preparation.  Therefore, any added salt must be included as  salt listed in the food-item f i l e . An abridged food-composition f i l e (Appendix D) of 22 nutrients was developed to coordinate with the abridged food-item l i s t of 127 item clusters contained in Appendix B, and to provide a reduced composition format for system testing. follows:  The nutrients selected for the f i l e are as  total calories; protein; total, saturated, and polyunsaturated  fatty acids; total carbohydrate; sucrose; fiber; nine vitamins (thiamin, riboflavin, niacin, pyridoxine, folate, ascorbate, retinol, cholecalciferol, tocopherol); and five minerals (calcium, phosphorus, magnesium, iron, potassium). For the nutrient t a l l y , the aggregate available quantity of any nutrient is assumed to be the sum of the quantities contained in each of the foods consumed.  No allowance has been made for factors that reduce  the availability of nutrients, either in a particular item or in other foods eaten simultaneously, with the exception of cashews, spinach, and spinach substitutes where oxalate concentration has been considered. Therefore, the total nutrient composition of the diet is the summation across all foods for each nutrient, namely: n (3.1)  E  i=l  a . x? q i  (q = 1, 2,  m)  1  where: x°  . is the amount of item cluster i per period time in the original diet.  An item cluster is a single item, or two or more substitutable  items of a similar nutrient composition. n  ;  is the total number of item .clustersMn the diet(Appendix A) or 127 (Appendix B).  =n is 22>  117  a :  is the amount of nutrient q in a unit of item cluster i . The values for a . are contained in the m x n matrices of food composition (Appendix C or D).  m  ' is the total number of nutrients considered in dietary analysis - either 41 (Appendix C) or 22 (Appendix D).  3.2.3  Data-Evaluation  In the data-evaluation phase the system evaluates the client's  initial  diet by comparing calculated nutrient-intake with nutrient limits which are individualized for the client.  The evaluation output comprises a  graphical and/or tabular display of the quality of the diet. format utilized is shown in Table 3.3.  The output  This provides a functional  tabular output for system testing and development.  An example of a pro- .  posed output format (Table 3.4), which is modelled on the "Nutrient Quality Index" (Wittwer et^ al_. 1977), incorporates a graphical presentation of the evaluation output.  Table 3.3  Excerpt from the evaluation output format used for system testing  NUTRIENT  MINIMUM (/week)  MAXIMUM (/week)  PROT (gm) CHO-T (gm) T-FAT (gm) KCAL CHO-F (gm) SFA (gm) SUCR (gm) PUFA (gm) VIT-A (iii) VIT-D (iu) VIT-E (mg)  392.00 2735.4 44.209 18899. 79.576 .0 .0 44.209 35000. 700.00 63.000  558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 .14000E+06 4200.0 700.00  INITIAL DIET (/week) 818.46 1572.2 825.87 18366. 21.596 363.74 396.53 445.79 44903. 1623.6 48.232  REVISED DIET (/week) 558.60 2735.4 566.47 19301. 79.576 208.08 746.02 317.61 72916. 1244.7 63.000  118  Table 3.4  Excerpt from a proposed evaluation output format.*  The table below contains an estimate of your average daily nutrient intake over a period of (1 day, 1 week, 1 month or longer), and a comparison of your estimated intake with minimum and maximum intake standards for an individua-1 with the following characteristics: (activity, sex, age, size, pregnancy status). The table also indicates the average daily amount of nutrients provided by a diet recommended for you, i f your old diet does not meet the standards. Nutrient  Recommended Intake /day Min. Max.  Estimated Diet Intake /day Old New  Energy (kcal) Protein (gm) Carbohydrate (gm) Fiber (gm) Fat (gm) 6tC•••«*••••••  2700 56 391 11 6  2624 117 225 3 118  2984 80 639 23 95 ••••  2757 80 391 11 80  Estimated old dietary intake as % of minimum recommended intake 100% xxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxx xxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx  Estimated new dietary intake as % of minimum recommended intake 100% xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxx  Nutrient  Estimated old dietary intake as % of maximum recommended intake  Energy (kcal) Protein (gm) Carbohydrate (gm) Fiber (gm) Fat (gm) etc  xxxxxxxxxxxxxxxxxx^^ xxxxxxxxxxxxxxxxxxxxx xxxxxxxx xxxx xxxxxxxxxxxxxxxxxxxxx  Estimated new dietary intake as % of maximum recommended intake 100% xxxxxxxxxxxxxxxxxx xxxxxxx xxxx xxxxxx xxxxxxxxxxxx  Nutrient  Energy (kcal) Protein (gm) Carbohydrate (gm) Fiber (gm) Fat (gm) etc  The following nutrients, although essential for the maintenance of health, have not been included in assessing your diet: (water, chromium, e t c . . . . ) . The following factors are not considered to be nutrients and consequently are not included in assessing your diet: (nucleic acid, popsicle magic factor, etc ). Nutrient values presented in the evaluation output could correspond to the total evaluation f i l e , to selected values as requested by the client or as needed by the counsellor for illustrating diet problems, or to composites of groups of nutrients  119  The nutrient-limits f i l e (Appendix E) provides the basic data, used in both the assessment and planning phases of the program, to define the range of appropriate nutrient intake for a given period of time.  This  f i l e is stored in the system for retrieval during operation. Nutrient limits for the client are generated from" the values contained in the nutrient-limits f i l e by applying the rules outlined in Appendix E. These rules are used to translate values from the nutrient-limits f i l e to measurement units which are common to those of the food-composition f i l e , and to individualize these for the client based on the client's demographic data collected on the questionnaire (p.TT.3).  Nutrient limits expressed as  ratios of two nutrients do not require this modification. The nutrient-limits f i l e contains maximum and minimum nutrient limits expressed as nutrient ratios, as quantities per day, and as quantities per kilogram per day.  These limits are disaggregated by age, sex, activity  level and pregnancy status.  Standards from Health and Welfare Canada,  and recommendations from other established nutrition sources, have been used to designate minimum nutrient limits.  As indicated, these values  with their respective sources have been outlined in Appendix E.  It should  be noted that the philosophy of the "Dietary Standard" has not been represented in the development of every lower intake limit used in the system.  This is due, in part, to the inclusion of "nonessential" nutrients  in the evaluation phase - - for example sucrose, saturated fatty acids, and fiber.  For these components a lower limit of zero was used, unless benefits  from some intake were documented.  Where dietary standards were not avail-  able for essential nutrients included in the system, such as polyunsaturated fatty acids, sodium, and potassium, a measure of minimum maintenance was used as outlined in Appendix E.  Other suggested indices of dietary well-  120  being not specifically included in the Dietary Standard, such as polyunsaturate and saturate ratios, calcium/phosphorus were also included in the nutrient-limits f i l e .  Similarly, maximum limits have been  empirically based where possible, and otherwise are arbitrarily set at twice;"! the minimum value for present test purposes.  Maximum and minimum  energy intakes have been arbitrarily set at plus or minus five percent of the intake for any age, sex, and activity category. In total, 82 nutrient limits - - 41 maximum and 41 minimum - - have been assigned.  These values correspond to the 41 food compositional values  in Appendix C.with a few exceptions.  First, in addition to specific  limits on calcium, phosphorus, polyunsaturated and saturated fatty acid intake, restrictions have been included on the ratios of polyunsaturated to saturated fatty acids and of calcium to phosphorus.. Second, instead of maximum and minimum limits for each of the eleven amino acids, only 18 limits have been imposed to cover 7 single amino acids and 2 pairs of amino acids - - phenylalanine and tyrosine; methionine and cystine. An abridged nutrient-limits f i l e (Appendix F) has been designed for system testing.  This table has 24 minimum and 24 maximum standards to  coordinate with the abridged food-composition f i l e (Appendix D) and fooditem l i s t (Appendix B).  The nutrients considered are:  total calories;  protein; total, saturated, and polyunsaturated fatty acids; polyunsaturate to saturate ratio; total carbohydrate; sucrose, fiber; nine vitamins (thiamin, riboflavin, niacin, pryidoxine, folate, ascorbate, retinol, cholcalciferol, tocopherol); five minerals (calcium, phosphorus, magnesium, iron, potassium); and calcium to phosphorus ratio.  121  3.2.4  Diet-Planning  The system described thus far, has analyzed the nutrient composition of the client's diet and evaluated the client's diet by comparing the diet to nutrient limits.  If the client's i n i t i a l diet does not meet these  limits, then the system generates a revised diet for the client.  This  computer prints out this revised diet for comparison with the i n i t i a l diet.  Table 3.5 illustrates the numerical listing of the item clusters  with quantities in grams per week for the i n i t i a l and revised diets.  Table3.5  Excerpt from the diet-planning output format .used for system testing. Item Cluster Code Number 003 004 005 006 008 010 012 013 015 016 017 018 019 020 021 023 024 027 028 031  Initial Diet (100 gram/week) 2.2400 0.0 0.0 0.0 1.1000 0.0 0.0 7.2000 3.6000 0.0 0.0 0.0 3.0000 0.0 0.0 0.6900 1.3800 0.0 1.3800 0.0  Revised Diet (100 grams/week) 3.9887 0.0 0.0 0.0 0.0 3.5409 1:5189 4.4315 0.0 0.0 0.0 0.4904 3.0426 0.0 0.0 0.0 3.0740 0.0 0.6316 0.0  The revised diet is developed by a constrained-optimization formulated to find the combination of item clusters that minimizes the sum of the squared differences between the amount of specific item clusters and of  122  g attribute groups  in the i n i t i a l and revised diet, while satisfying  nutrient constraints.  The formal mathematical statement of the model is: I 2_. K. j 2 minimize I w.(x. - x?) t E P.. (E (x- - x?)) i=l k=lj=l UG i.  _t_  (3.2)  1  1  1  J  k  1  1  j k  (3.3)  subject to  I m >, E a . x- > n q ^ _i q q n  1  (q = 1, 2, . . . , Q)  1  I E  (3.4)  r  a  ui  where:  X i  i  ^ ifl I 1-! ^  (3.5)  x  >0  > t  (u,v = any specified set of nutrient pairs)  ^  (i = 1, 2,  I)  x^ is the amount, in grams, of item cluster i per time period in the revised diet.  The values of x^ are assumed to be additive  and independent x*? is the amount, in grams, of item cluster i per time period, in the client's i n i t i a l , desired, or nutritionist-prescribed diet, as determined from the client questionnaire. I  is the total number of item clusters in the diet.  The I vector  equals 221 in the complete food-item f i l e (Appendix A), or 127 when the abridged food-item f i l e (Appendix B) is considered.  5.  As indicated, an item cluster is one item, or two or more food items, varieties of an item, or prepared variations of an item which are considered nutritionally equivalent.  6.  An attribute group is a group of item clusters with similar characteristics or attributes (see Appendix G).  123  K is the total number of hierarchical levels in the model, namely seven.  Each hierarchical level corresponds roughly to a  criterion for classifying item clusters into attribute groups on the basis of item similarities.  The designated attribute  groups form a hierarchical pyramid of groups and subgroups. The matrix of the hierarchical levels and attribute-groups is given in Appendix G. The matrix developed is based, in part, on the examples of food classification schemes by other authors (FAO/WHO 1 9 4 9 ;  Davenport 1 9 6 4 ; Gue and Liggett 1 9 6 6 ; Chandler and Perloff 1 9 7 5 ; Canada 1 9 7 7 ) .  It represents an attempt to direct appropriate  food substitutions among items of the food l i s t by, f i r s t , defining general food and diet attributes that may important for indicating an item's or a diet's similarity to another item or diet.  Second, foods are classified into groups according  to these attributes.  By attempting to provide the nearest  acceptable substitute for any change in the quantity of an item or class, the compatibility relationships among items and the general characteristics of the altered diet should be most successfully maintained.  For further discussion refer to  Section 4 . 2 ( p . 1 3 6 ) . J  is the total number of attribute groups j in the diet.  J equals  2 7 8 in the large model and 1 7 8 in the abridged model, is the number of attribute groups j in the k level.  In the large model J-j = 6 7 , ^  = 3 5 , J g = 2 7 , and  J-, =50, J  9  =39,  = 4.  =  hierarchical  55,  =  52,  = 38,  In the abridged model matrix  =34, J. =23, Jr =20, J  f i  =12,  and J  7  = 4.  124  Gj^ is the set of item clusters i within the j of the  level.  attribute group  The total amount of food within any attribute  group is the sum of the quantities of item clusters contained in that group.  The item quantities are assumed to be additive  and independent; that i s , i t is assumed that the common characteristic of items within any group can be measured by the aggregated quantity of the contained item clusters, and that an item cluster can hold simultaneous membership in any number of groups, w-  is the weighted penalty associated with deviations of item cluster i from the desired amount.  Values of w^ have been  assumed in the absence of empirical data.  These values are  discussed in Chapter 4 (p.145). Pj^  is the weighted penalty associated with deviations of the j^* attribute group in the k Values of data.  a •  1  level from the desired amount.  have been assumed in the absence of empirical  These values are discussed in Chapter 4 (p.182)-  is the amount of nutrient q in a unit portion of item cluster i.  The values of a ^ are obtained from the I x Q matrix of  food composition (Appendix C or D) which contain nutrient coefficients for edible 100 gram portions of foods.  The  aggregate available quantity of any nutrient is assumed to be the sum of the quantities contained in each of the foods consumed.  The elements of the food composition table are  assumed to be additive and independent, and to be constants. Allowance has not been made for factors that reduce the availab i l i t y of nutrients in either a particular item or in other  125  foods eaten simultaneously, with the exception of cashews, spinach, and spinach substitutes where the effect of oxalate concentration has been considered. Q is the total number of nutrients q considered in the diet. For the large model, Q is 41; for the abridged model, Q is 22. The  nutritional adequacy of solutions for mathematical diet-  models can only be assumed for those nutrients specifically included in the model.  For this reason, the large-model formu-  lation incorporated as many nutrient values as was reasonably possible. m and n are Q vectors of respectively, the maximum and minimum amount of nutrient q allowed in the individual's diet over a defined time period according to the client's age, sex, size, activity and pregnancy status.  Values for m^ and-TV  are derived from  the nutrient-limits f i l e (Appendix E or F). a - and a - are the amounts, respectively, of specified nutrients u and ul  v1  v in a unit of item cluster i .  u corresponds to either calcium  or polyunsaturated fatty acids,  v corresponds to either  phosphorus or saturated fatty acids.  Values of a - and a^ ul  are contained in Appendix C and D. r and t uv  are, respectively, the maximum and minimum allowed ratio for the nutrient pair (u,v).  The nutrient pairs considered are  calcium and phosphorus, and saturated and polyunsaturated fatty acids. or D.  Values of r  u v  and t  are derived from Appendix C  126  The objective function (Eqn. 3.2) defines a l i s t of item clusters by minimizing the aggregate squared difference between specified characterist i c s ^ the i n i t i a l and revised diets.  The f i r s t term of the objective  function sums the weighted squares of the difference between the amounts of item clusters in the i n i t i a l and revised diets. The  The second term sums  weighted squares of the difference between the i n i t i a l and revised  amounts of a hierarchical sequence of item cluster groups, called a t t r i bute groups.  The quadratic term introduces disproportionately larger  penalties as the revised diet deviates more widely from the i n i t i a l  diet.  This characteristic of the objective function tends to spread deviations uniformly over all item clusters and all attribute' groups. The modelling constraints comprise:  nutrient constraints (Eqns. 3.3,  3.4) which designate maximum and minimum limits for the nutrients provided by the diet, and a non-negativity constraint (Eqn. 3.5) which prevents entry of non-negative quantities of variables in the solution.  With  respect to the nutrient constraints, only those which can be stated either as a fixed amount (Eqn. 3.3), or as linear ratios of two nutrients (Eqn. 3.4) are incorporated in the model.  As written the second constraint  (Eqn. 3.4) is a nonlinear equation.  However, i t is easily transformed  into linear form.  Apart from Equation 3.4, nonlinear constraints were not  considered in the model. The model was formulated into matrices suitable for input into a preprogrammed quadratic package based on Lemke's Complementary Slackness Algorithm (Cottle and Dantzig 1968; Lemke 1968). formulation can be briefly outlines as follows: 1  Minimize:  c' x + % x  D x  Subject to:  A x >, b  Where:  x ^ o is the vector of foods.  The matrix algebra  127  c' is a vector containing the coefficients  for linear  terms in the objective function. D is a matrix containing the coefficients of quadratic terms in the objective function. A  is the food composition matrix,  b  is the vector of nutrient constraints.  128  CHAPTER 4 TESTING OF THE PROTOTYPICAL SYSTEM  4.1  Introduction The second project objective (p.14) involved testing of the diet-planning  phase of the system, since this phase was considered to be the most significant obstacle to overall system development.  This testing was restricted  to a descriptive evaluation of some of the objective function's characteristics.  Specifically, assumptions defining the concept of minimum deviation  between diets, which are implicit in the objective function, were a r t i culated and then explored by altering some of these assumptions and observing the consequences for revised diets developed for hypothetical individuals.  The impact of altering nutrient constraints was not considered.  This evaluation was undertaken to explore the conceptual and technical feasibility of using a mathematical model to provide an effective dietary recommendation - - the individual's nutrient-constrained food-choide - - from a dietary inventory identified as desirable for the individual by his or her past consumption or stated preference.  Also, it provides a basis for  more definitive evaluation and development of the model.  Explorative  evaluation does not constitute the means to validate^ the diet-planning model's design. In order to reduce computational costs and clerical work required for evaluation of the diet-planning phase, all testing was done using the abridged data bases outlined in; Appendices B (abridged food-item f i l e ) ,  7.  Validation tests that the model is a reasonable representation of reality.  129  D, (abridged food-composition f i l e ) , F (abridged nutrient-limits f i l e ) , and G (attribute-group matrix).  The nutrient constraints for test runs  were determined from Appendix F for a standard male subject of age 19 to 35, activity level code B, and 70 kilograms body weight.  The nutrient  assigned for this standard subject for a weekly period are outlined below in Table 4.1.  The daily intake equivalent is also provided for  comparison.  Table 4.1  4  Upper and lower nutrient constraints* for a standard subject : expressed as weekly amounts (and daily equivalents) for 24 nutrients.  Nutrient  Lower Limit (/wk)  Upper Limit (/wk)  Lower Limit (/day)  Upper Limit (/day)  Nutrient  Lower Limit (/wk)  Upper Limit (/wk)  Lower Upper Limit Limit (/day) (/day)  ENERGY (kcal) PROT (gm) FAT-T (gm) SFA (gm) PUFA (gm) P/S  18894  20889  2699  2984  28000  2000  4000  392.0  558.6  56.0  79.8  2800  200  400  44.21  663.13  6.32  94.73  3500  30  500  0.00  221.04  0.00  31.58  44.21  663.13  6.32  94.73  1  2  1  2  2735.4  4476.1  390.8  639.4  0.0  746.02  0.0  106.57  79.58  159.15  11.37  22.74  PYR 14000 (ug) F0L 1400 (ug) VIT-C 210 (mg) VIT-A 35000 (1u) VIT-D 700 (iu) VIT-E 63 (mg) CAL 5600 (gm) PH0SP 5600 (mg) CA/P 0.8  9.94  19.89  1.42  2.84  131.30  262.60  18.76  37.51  11.94  23.87  1.71  3.41  MAG (mg) IRON (mg) POT (mg)  CHO-T (gm) SUCR (gm) CHO-F (gm) THIA (mg) NIAC (mg) RIBO (mg)  140000 5000  20000  4200  100  600  700  9  100  11200  800  1600  11200  800  1600  1.2  0.8  1.2  2205  4410  315  630  70  140  10  20  9800  19600  1400  2800  * Values for these constraints are derived from Appendix F. + Standard subject was age 19 to 35, activity level code B, and 70 kg. body weight.  130  Two seven-day food-intake records were defined to represent the i n i t i a l diet obtained from the standard male subject (Table 4.2).  One record,  called Standard Initial Diet 1 (SID-1), contained 3 items; the other, called Standard Initial Diet 2 (SID-2), contained 83 items.  Table 4.3  compares the amounts of nutrients supplied by each of these diets with the upper and lower nutrient constraints.  131  Table 4.2  GROUP-ITEM CODE #  Standard Initial Diet 1 (SID-1) atid'.Standard Initial Diet 2 (SID-2): Two seven-day food-intake records for a standard male subject, age 19 to 35, activity level code B, and 70 kilograms body weight. FOOD ITEM  SID-1 (grams/week)  SID-2 (grams/week)  ENTREE-DAIRY 001-003 001-004 001-005  CHEDDER CHEESE... COTTAGE CHEESE... CREAM CHEESE  002-006 002-008  SOUR CREAM YOGHURT  003-010  EGG  182  495  ENTREE-CEREALS 004-012 004-013  CORN CEREAL WHEAT CEREAL  28  005-015 005-016  OATMEAL WHEAT CEREAL  270  006-017  PANCAKES  135  007-018 007-019  NOODLES SPAGHETTI  008-020 008-021 008-023  RICE, brown RICE, white WHEAT GERM  009-024 009-027 009-028  FRENCH BREAD WHITE BREAD WHOLE WHEAT BREAD  40 414  010-031 010-032 010-033 010-035  BISCUITS HAMBURGER BUN MUFFIN ENGLISH MUFFIN...  70 46 40 46  011-037 011-040  SALTINES RYE KRISP  24  9375 280  ENTREE-MEATS 012-041 012-042 012-043 012-046 012-047 012-048  BEEF, BEEF, BEEF, PORK, PORK, BACON  30% f a t . . . . 20% f a t . . . . 15% f a t . . . . lean cuts.. a l l hams...  013-049 013-051  CHICKEN, steamed. CHICKEN, f r i e d . . .  170 85 170 85 43 64 340 JJ  132 Table 4.2  (Continued)  014-055 014-056 014-058 01.4-062 014-063 014-065  FRIED FISH BROILED FISH OYSTERS SARDINES SHRIMP. .• TUNA  015-067  LIVER  017-069 017-070 017-071  FRANKFURTERS FRESH SAUSAGES LIVERWURST  018-073 018-075  BEANS, dried SOYBEANS  019-076 019-077 019-079 019-080 019-081  ALMONDS CASHEW NUTS PEANUT BUTTER PEANUTS PECANS  1 •  43 60  40  15 "28 120  ENTREE-VEGETABLES 020-082 020-083 020-084 020-085  POTATOE, baked.... POTATOE, f r i e d . . . . POTATOE, mashed... SWEET POTATOE  100 100 200 180  021-089 021-091 021-092 021-095 021-098 021-099 021-101  BEANS, green BROCOLLI CABBAGE LETTUCE PEAS PEPPERS SPINACH  65 63 65 180 170  022-103 022-104 022-105 022-106 022-111  BEETS CARROTS, cooked... CARROTS, raw CORN TOMATOE  023-113 023-114 023-115  CUCUMBER MUSHROOMS ONIONS  024-116  SUCCOTASH  025-117 025-118  OLIVES PICKLES, sweet  025- 119  PICKLES, sour  85 1  80 76 50 518 275 8  20 34  ENTREE-FATS 026- 121 026-124  LARD SOYBEAN OIL  027-125  BUTTER  5 125  Table 4.2  (Continued)  028-127 028-128  CHEESE SAUCE GRAVY  029-132 029-133  MAYONNAISE SALAD DRESSING....  105 90  BEVERAGES-DAIRY 030-136 030-137  WHOLE MILK SKIM MILK  031-140 031-141  TABLE CREAM... WHIPPED CREAM  732 5658 195 8  BEVERAGES-FRUIT 032-143 032-144 032-145 032-148  APPLE JUICE GRAPEFRUIT JUICE.. LEMON JUICE ORANGE JUICE  033-150  TOMATOE JUICE  120 5 360  BEVERAGES-MISC. 034-151  COLA-TYPE  035-153 035-154  COFFEE TEA  2800 1600  036-155 036-156 036-158  BEER DISTILLED SPIRITS. DRY WINES  3240 43 400  339  SOUPS 037-160 037-161 037-163  CREAMED SOUPS PEA SOUPS MEAT + VEGIE SOUPS  198 200  DESSERTS-CEREALS 038-166 038-168 038-170  COFFEE CAKE FRUITCAKE ICED CAKES  75  039-172 039-172  FRUIT PIES PUMPKIN PIES  160 150  040-173 040-174  COOKIES FRUIT COOKIES  60  041-175 041-177  CAKE DOUGHNUTS.... DANISH PASTRY  60 38  60  DESSERTS-DAIRY 042-178 042-179  ICE-CREAM SHERBERT  270  043-181  PUDDINGS  185  1,34 Table 4.2  (Continued)'' /:.: DESSERTS-FRUIT  044-183 044-184 044-185 044-186 044-187 044-188 044-192  APPLE APPLESAUCE BANANA CANTALOUPE GRAPEFRUIT ORANGE PINEAPPLE  045-193  DRIED FRUIT  300 185 100 100 200 150  DESSERTS-SWEETS 047-195 047-197  HONEY SUGAR  048-198 048-199  JAMS. SYRUP  049-201 049-202  CHOCOLATE CANDY... MARSHMALLOW  lUb 100 40  MISCELLANEOUS 051-205  POT PIES  227  063-217  SPAGHETTI + MEAT..  330  067-221  COCOA MIX Grams/Week..  15313  1 18871  Total # of Items..  3  83  Total  135  Table 4.3  Nutrient composition of SID-1 and SID-2 and the upper and lower nutrient constraints for a standard male subject.  Nutrient  PROTEIN (gm) CHO-T (gm) T-FAT (gm.) KCAL CHO-F (gm) SFA (gm) SUCR (gm) PUFA (gm) VIT-A (iu) VIT-D (iu) VIT-E (mg) VIT-C (mg) RIBO (mg) NIAC (mg) VIT-B6 (ug) FOLATE (mg) POTAS (mg) CAL (mg) PHOSP (mg) IRON (mg) MAG (mg) CA/P P/S  Nutrient Constraints Minimum Maximum (/week) (/week) 392.00 2735.4 44.209 188899. 79.576 .0 .0 44.209 35000. 700.00 63.000 9.9470 11.936 131.30 15000. 1400.0 9800.0 5600.0 5600.0 70.000 2205.0 .8 1.0  713.82 1872.4 981.39 20195. 32.250 367.46 549.60 543.42 69952. 1111.3 75.273 10.292 13.803 178.25 12942. 1460.9 22490. 6137.3 11258. 120.08 2386.5 .54514 1.4789  SID-1 Initial (/week) 512.54 2809.9 92.428 14210. 35.125 5.6000 30.925 22.400 0.0 2319.8 56.550 16.329 13.963 148.67 20777. 2019.5 17082. 8172.8 15349. 73.195 4471.8 .53245 4.0000  ,  SID-2 Initial (/week) 558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 .140E+06 4200.0 700.00 19.894 23.873 262.60 28000L 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  136  4,2  Diet-Planning Model's Premises and Assumptions A satisfactory procedure for dietary evaluation and suitable data on  food composition and nutrient requirements were available for developing constraints to define dietary adequacy.  However, no obvious procedure  was available to define the relative acceptability of modifications in an initial dietary inventory.  As previously mentioned (p.80), an extensive  tradition does exist for hospital menu planning and Food Guide generated diets, but precise guidelines are not available for systematically developing individualized dietary prescriptions which are both nutritionally adequate and maximally acceptable to the client.  Presumably the nature  of this process is not fully understood. In the absence of a detailed precedent which defines the relationship between an individual's nutrient-constrained food-choice and his or her i n i t i a l dietary inventory, two major premises and related assumptions were used to define and operationalize the objective function.  4.2.1  The First Premise and Related Assumptions The first major premise, for developing the model's objective  function is that the client's most acceptable diet can be defined either by past consumption patterns - - namely, previous choice tends to define future choice - - or by a food plan identified as desirable by the client. In practice, past consumption may provide the better monitor of actual preference since i t is the diet chosen under presently operating budgetary restrictions and long-term cultural and personal habits. It is assumed that an individual identifies a diet's character by a complex of attributes.  These attributes provide cues to distinguish one  dietary pattern from another and an index to estimate the relative  137  significance of dietary change.  In this context, attributes are defined  as measureable aspects of food or of behavior applied to foods, such as food taste, colour, ethnocultural pattern of food intake, nutrient content, menu function, food group, cost, serving frequency, preparation procedure, and a variety of food item relationships. These attributes partition or categorize food substance into an array of parts and subparts; for example, into such typical categories as ingredients, food items, food groups, menu items, as well as less formal or personal dietary distinctions such as an individual's preferred items.  Thus for each individual, an attribute map of considerable complex-  ity can be envisioned.  Depending on the attributes utilized by the indivi-  dual, the map may correspond, more or less, to ethno-cultural patterns, familial patterns, and so on.  Further, i t should be noted that depending  on the attributes considered, two diets may differ by one set of criteria but be alike by another. The attribute map developed for the prototypical model is based, in part, on the examples of other authors (FAO/WHO 1949; Davenport 1964; Gue and Liggett 1966; Chandler and Perloff 1975; Canada 1977) who developed food classification schemes for various purposes.  Only a limited number  of illustrative attributes is considered for the prototypical model - attributes presumably used by the average Canadian for identifying the similarities between food substances.  This provisional assignment of  attributes includes a definition of a standard item.  These are mono-  ingredient menu items which are low in complementarity requirements - that i s , items which do not require the coexistence of other items to ensure their palatability and acceptability - - and which can be assigned nutrient values from available literature.  The other attributes  13a  considered are: -generic terms of common usage such as cheeses, poultry, bread, and cultured milks; -meal classification such as breakfast or lunch; food group assignments such as dairy products or fruit products; and -some general physical characteristics including taste, colour, and physical state. On this basis dietary items were partitioned into a hierarchical, attributegroup matrix based on the apparent similarities between defined items.  The  matrix elements created are discussed in Section 3.2.4 and described in Appendix G. Attributes not explicitly considered in developing the attribute map, such as pricing, are considered independent parameters for present purposes, even though their influence could affect the diets acceptability. It is assumed that under conditions of change the map elements maintain their integrity.  However, in some extreme situations, such as where  radical dietary alterations are required, new attribute specifications may become apparent as food substance shifts from i n i t i a l levels. A vast number of attribute maps can be defined, each of which has a particular item l i s t with its inherent interactional properties.  The  attribute map chosen, and the items and item groups thereby defined, are central to the eventual u t i l i t y of such a model.  The items and item groups  selected will effect the accuracy and simplicity of the diet assessment procedure, the acceptability of the diet produced by the planning phase, and the outcome of communicating dietary modifications.  For example, using  a commodity l i s t of palatable items for diet planning largely guarantees palatibility of the diet developed.  139  4.2.2  The Second Premise and Related Assumptions The second major premise is that acceptable dietary modifications - -  namely, the best approximation for compliance to nutritional prescriptions - - can be provided by determining the least-altered diet which meets nutrient constraints.  The least-altered diet is presumably the best  alternative, as perceived by the individual, to his or her presently chosen, or preferred diet.  Thus, the mathematical equation for determining the  best alternative diet should reflect this concept. The individual is assumed to perceive dietary change as a shift of food substance in his or her personal attribute map.  The acceptability or  perceived extent of this change then depends on the significance of changing each element of the map from or to other elements.  its i n i t i a l condition with respect to itself  For the diet-planning model,.the acceptability of  this change is assumed to decrease as the square of the deviation from i n i t i a l amounts for any element.  Where unbounded, this produces a  symmetrical quadratic curve centered around the i n i t i a l amount of the element.  This deviation is weighted by a penalty coefficient assigned to  represent the relative significance of each attribute element deviating from i n i t i a l amounts.  The value assigned for this coefficient can be  adjusted for a variety of hypotheses including client acceptability or preference, i n i t i a l consumption levels, average serving size of items, or some other provisional pattern such as Canadian average consumption or nutritionist-recommended consumption levels.  In the diet-planning model  penalties are assigned to reflect i n i t i a l consumption levels, consumed versus non-consumed status, and for hierarchical membership.  Thus, the  acceptability of a diet deviating from i n i t i a l levels is defined as the summation of weighted squares of the difference between the amounts of the attribute elements in the i n i t i a l and revised diets.  140  4.3  Observations of the Objective Function's Characteristics  4.3.1  Unconstrained Objective Function The unconstrained objective function of the mathematical programming  model (viz. no constraints on the nutrients) corresponds to the solution of a problem with an input diet which already satisfies nutrient constraints in all respects.  The unconstrained solution allows the sum of weighted  squared differences between items and item groups in the i n i t i a l and revised diets to go to zero.  This should result in a recommended diet identical to  the i n i t i a l diet - - the presumed ideal. To verify this, a number of solutions for i n i t i a l diets were obtained using the diet-planning model.  These solutions (Table 4.4, 4.13) were  developed from the SID-2 as the input diet and with a variety of objective functions (Eqns. 4.8, 4.9, 4.28) which incorporated different penalty coefficients,  as described later in this text.  The solutions for these  diets were rerun using the same programs that generated them.  As expected,  in each case the original solution was identical to the rerun solution, allowing for rounding errors. These observations illustrate, f i r s t , that each solution provided by the model is an optimal one, for the given input diet and nutrient constraints.  Second, providing the diet satisfied the nutrient constraints,  altering the penalty coefficients  in the objective function for positive  values of w- and P.. will not affect the outcome. I  The same phenomena would  JK  be expected with other objective functions incorporating the same theme of minimizing the deviation from the original diet, such as a linear objective function.  141  4.3.2  Constrained Objective Function If the most acceptable diet for the client occurs when the sum of  differences is allowed to go to zero, as with the unconstrained solutions, less acceptable diets would be those where, because of nutrient constraints, minimization towards zero is less successful.  Presumably the larger the  deviation the less acceptable the recommended diet will be for the client. The nature of this deviation and perhaps the relative acceptability of the outcome can be altered by modifying the characteristics of the objective function.  The following material explores alterations of the objective  function which produce different nutrient-constrained output profiles.  4.3.2.1  First Term;  Shape of the Curve  The mathematical expression used in the model's objective function to minimize deviations of any item cluster from i n i t i a l dietary levels, irrespective of specific concurrent deviations in other items, is the summed weighted square of differences between the amounts of item clusters in the i n i t i a l and recommended diet, as follows: I (4.1) minimize . E w. (x- - x.) 2  i=l  1  1  1  This expression, corresponding to the f i r s t term of the objective function (Eqn. 3.2), describes a symmetrical quadratic curve centered at the i n i t i a l consumption level, as shown in figure 4.1.  As the difference between the  initial and recommended values increases the quadratic objective the presumed unacceptability of this divergence.  intensifies  The choice of a quadratic  curve is arbitrary, but it is the simplest nonlinear function which captures the essence of this phenomenon.  142  Figure 4.1  Graph of quadratic function reflecting penalties for deviation from i n i t i a l consumption of food i .  A simpler alternative, but perhaps less realistic formulation, uses an objective that minimizes the absolute linear difference between the i n i t i a l and recommended diet, as follows: (4.2)  minimize  E w. (|x• i=l 1  - x?  1  1  In order to program this linear objective function (Eqn. 4.2),  the  formulation of the algorithm previously described (Eons. 3.2-3.5) can be rewritten in a piece-wise linear form, as follows: 1  (4.3)  minimize  , £ (w- x- + w- x-) i=l +  +  1  1  1  1  subject to (4.4)  x  (4.5)  m  + i " i x  q *  o =  x  (i = 1, 2,  x  i " i  a  qi  X  i *' q  a  ui  x  i  i=l  ,n  ( q  =  l  s  2  '  I)  Q)  I (4.6)  E r  ,,w ^ i l I =  ^t  u v  E'  1=1  u v  a . x. V 1  1  (u,v = any specified set of nutrient pairs)  143  ( 4  '  7 )  ><i' xt, T » o X  (i = 1, 2, . . . , I)  Where the newly defined terms are as follows: x j , xj are,.respectively, the positive and negative deviation of x^ from x ° . ( (x, -  + x  i  =  ( ( - ( x  i  =  x°) i f  X i  - x° > 0  (  0 Xi  otherwise - x?) i f  x.  - x° < 0  (  (  0  otherwise  Thatiis, for each i at most one of x| and x!j can be positive. +  w^ are the weighted penalty associated with the positive or negative deviation, respectively, of item cluster i from the original amount.  In more detail, the linear objective minimized the weighted positive and weighted negative deviation of item clusters from i n i t i a l  levels.  Unlike the quadratic formulation the penalty associated with this deviation varies in direct proportion to the difference.  The modelling constraints  include, in addition to the nutrient constraints previously discussed, an equality constraint (Eqn. 4.4) which defines the positive or negative deviation of the revised diet to the i n i t i a l diet, and a non-negativity constraint (Eqn. 4.7) which restricts entry of non-negative quantities of variables in the solution. Although linearized solutions were not developed, with which the quadratic results could be directly compared, the two models would be expected to generate different solutions.  Compared to the linear version  the quadratic term introduces larger penalties as the prescribed diet  144  deviates more widely from the i n i t i a l diet, but smaller penalties for very small changes, as illustrated in Figure 4,2.  Thus, the quadratic objective  should moderate extreme fluctuations in any particular item by suppressing single large deviations between i n i t i a l and revised values in favour of numerous smaller changes.  The linear model should be less sensitive to  large changes in specific items.  Consequently, one can speculate that the  linear model may provide dietary solutions with undesirably large changes in some single items, unless these changes are restricted by specific constraints on the amounts of these items.  Correspondingly while the  quadratic model moderates these extreme fluctuations, its solutions are more likely to include a large number of small decrements or increments. These small deviations may be considered as cosmetic problems which can be remedied by rounding values to the nearest usable portion, but consequently losing some accuracy of the minimization in the process.  Alter-  natively, either constraints on minimum entry levels to prevent unusably small increments from zero, or integer programming, could be used to overcome this difficulty with more accuracy.  However, the possibility of many  items entering at minimum levels s t i l l exists unless specific are applied to the number of items entering the solution  Figure 4.2  restrictions  set.  Graph of quadratic and linear functions reflecting for deviation from i n i t i a l consumption of food i .  penalties  145  4.3.2.2  First Term:  Penalty Coefficient,  The relative acceptability of change for different items should also be considered, in that alterations in the quantities of different dietary items may not be equally acceptable to the client.  Consequently, in  developing a recommended diet some items should be preferentially maintained at the i n i t i a l amount whereas other items may be favourably altered to a greater or lesser extent.  By altering the penalty-coefficient terms,  w^, of the equation which affect the relative slope of the exponential minimization curve, the value of increments for an item can be exaggerated or depressed. The numerical value assigned to the penalty coefficient can correspond to one of, or the product of, a number of componenti:values used to represent perceived acceptability,  (i)  The weighting coefficient can be normalized  for i n i t i a l comsumption level, that i s , the inverse of the client's i n i t i a l consumption of an item as determined from questionnaire data, as fol1ows: w. <* l/(x? + e) n  '  Here a small increment, e > 0, has been included in the denominator so that percentage change can s t i l l be represented when x° = 0.  Values of e = 1 and  e = .01 were chosen for test purposes, and are not intended to represent established values.  Although the value chosen for e is in some sense  arbitrary, the values should be small enough so that solutions will not be appreciably altered at typical portion-sized values of x ° .  At zero  values of x ° , the small epsilon value of e = .01 weights heavily against the entry of the item into the diet since percentage increase of x° would necessarily be large.  Alternatively, i n i t i a l consumption could be given a  146  lesser weighting by using for example the inverse of the square root of initial consumption, as follows:  Thus, as consumption level increases the weighting penalty does not decrease proportionately. (ii)  Deviation could be normalized for standard serving size by  weighting deviation with the inverse of serving size, as follows: w  i  oc  1/grams per serving of items cluster i  Standard serving sizes are specified in Appendix A. (iii)  Further information which may be used in formulating a weighting  coefficient to represent client willingness to deviate from original consumption includes:  data on Canadian average consumption for normalizing  deviations to Canadian norms; weighting coefficients  obtained from  nutritionists to massage solutions towards specific ends; or some other measure which may possibly represent an individual's acceptability dynamics, such as penalties to inhibit the entry of initially-zero item quantities. Two of the possible variations of the weighting coefficient, mentioned above, have been explicitly considered in the following material.  These  involve weighting deviation on the basis of i n i t i a l consumption levels, and preferential weighting of consumed versus non-consumed items in the i n i t i a l inventory.  4.3.2.2.1  Penalty Coefficient, w^ Based on Amount Consumed  The objective function already considered (Eqn. 4.1) can be defined to weight the squared deviation with a coefficient of 1 (Eqn. 4.9).  A  second approach is to weight deviations on the basis of i n i t i a l consumption  147  level, or to produce percentage squared deviations (Eqn. 4.9).  The cor-  responding summed expressions for the objective functions are, as follows:  (4.8)  minimize  I z 1 (x i=l  2 i  - x°)  1  1  (4.9)  minimize  z  1  2  1 —  1=1 ( x ° + D  T  (x. - x?) 1  1  The value e = 1 has been included in the denominator so that percentage change can be represented when x-j = 0, as previously discussed (p.145). Using the above objectives (Eqns. 4^8, 4.9), revised diets were developed from the two standard i n i t i a l diets, SID-1 and SID-2, as shown in Table 4.4.  The corresponding nutrient compositions are tabulated in Table 4.5.  148  Table 4.4  GROUP-ITEM CODE #  Revised diets developed from SID-1 and SID-2 using quadratic objective functions (Eqns. 4.8, 4.9) FOOD ITEM  SID-1 SID-2 (grams/week)* (grams/week) Initial Eqn.4.8 Eqn.4.9 Initial. Eqn.4.8 Eqn.4.9  ENTREE-DAIRY 001-003 001-004 001-005  CHEDDER CHEESE. COTTAGE CHEESE. CREAM CHEESE...  002-006 002-008  SOUR CREAM YOGHURT  003-010  EGG  354  182  519  495  117  28  210 221  108 88  270  187 24  207 21  618  .13  ENTREE-CEREALS 004-012 004-013  CORN CEREAL WHEAT CEREAL...  005-015 005-016  OATMEAL WHEAT CEREAL...  006-017  PANCAKES  007-018 007-019  NOODLES SPAGHETTI.  008-020 008-021 008-023  RICE, brown.... RICE, white WHEAT GERM  009-024 009-027 009-028  FRENCH BREAD... WHITE BREAD.... WHOLE WHEAT BREAD  010-031 010-032 010-033 010-035  BISCUITS HAMBURGER BUN.. MUFFIN ENGLISH MUFFIN.  011-037 011-040  SALTINES RYE KRISP  203  203 50 14  135  9375  7773  3407 25  47  75 39  . 29 11  63 451  62 752  138  66  280 L  49 82  40 414  182 82 182 34 50  70 46 40 46 24  184  2 73 28 186  6 221  54 88  83  ENTREE-MEATS 012-041 012-042 012-043 012-046 012-047 012-048  BEEF, BEEF, BEEF, PORK, PORK, BACON  30% f a t . . 20% f a t . . 15% f a t . . lean cuts a l l hams.  013-049 013-051  CHICKEN, steame d CHICKEN, fried.  014-055 014-056  FRIED FISH BROILED FISH...  170 85 170 85 43 64 340  Values rounded to the nearest grams/week, except values below 0.5 grams/ week which are rounded to three decimal places.  .149 Table 4.4 . (Continued) .1  014-058 014-062 014-063 014-065  OYSTERS SARDINES SHRIMP TUNA'.  015-067  LIVER  017-069 017-070 017-071  FRANKFURTERS FRESH SAUSAGES.... LIVERWURST  018-073 018-075  BEANS, dried SOYBEANS  126 33  019-076 019-077 019-079 019-080 019-081  ALMONDS CASHEW NUTS PEANUT BUTTER PEANUTS PECANS  73  43 60  22  37  98  61  40 25  88 79  15 28 120  29 152  100 100 200 180  15 10 85  43 112  82 218  125 443  253 363 180 178 572 374 182  199 277 175 271 924 95 214  518  159 146 70 109 448  178 166 113 49 577  275  210  8  119  230 1 48  158  52  380  84 20 30  22  T  ENTREE-VEGETABLES 020-082 020-083 020-084 020-085  POTATOE, baked.... POTATOE, f r i e d . . . . POTATOE, mashed... SWEET POTATOE  021-089 021-091 021-092 021-095 021-098 021-099 021-101  BEANS, green BROCOLLI. CABBAGE LETTUCE PEAS PEPPERS SPINACH  022-103 022-104 022-105 022-106 022-111  BEETS CARROTS, cooked... CARROTS, raw CORN TOMATOE  023-113 023-114 023-115  CUCUMBER MUSHROOMS ONIONS  39 37 117  024-116  SUCCOTASH  210  025-117 025- 118 025.119  OLIVES PICKLES, sweet.... PICKLES, sour  324 438 199 599 663 20  1443  235 352 178 83 355 304 231  65 63 65 180 170  163 205 181 25 59  80 76 50  374 93 82  85  20 34  ENTREE-FATS 026- 121 026-124  LARD SOYBEAN OIL  152  70 67  5  027-125  BUTTER.  104  52  125  028-127 028-128  CHEESE SAUCE GRAVY  59 72  .003 28  150 Table 4.4 029-132 029-133  (Continued) MAYONNAISE 24  SALAD DRESSING...  27'. 49  105 90  21 70  732  632  195 8  93  44 74  BEVERAGES-DAIRY 030-135 030-137 031- 140 031- 141  WHOLE MILK SKIM MILK TABLE CREAM  5658  4197  1977 13 23  WHIPPED CREAM  60  BEVERAGES-FRUIT 032-143 032-144 032-145 032- 148  APPLE JUICE GRAPEFRUIT JUICE. LEMON JUICE ORANGE JUICE  033- 150  TOMATOE JUICE BEVERAGES-MISC.  034- 151  COLA-TYPE.  035- 153 035- 154  COFFEE TEA BEER. DISTILLED SPIRITS DRY WINES  036- 155 036-156 036-158  6 120 5 360  17:4  44 -  39  339  328  332  2800 1600  2771 1572  306  3240 43 400  3229 34 324  4588 41 159  198 200  145 99  155 128 3  SOUPS 037-160 037-161 037-163  CREAMED SOUPS..._ PEA SOUPS MEAT + VEGIE SOUP >  17 20  DESSERTS-CEREALS 038-166 038-168 038- 170  COFFEE CAKE FRUITCAKE ICED CAKES  22  039- 171 039-172  FRUIT PIES PUMPKIN PIES  97 82  040- 173 040- 174  COOKIES FRUIT COOKIES....  041- 175 041- 177  CAKE DOUGHNUTS... DANISH PASTRY....  75  11  60  28  160 150  219 123  307 202  400  27 106  60 320  447 60 38  DESSERTS-DAIRY 042-178 042- 179 043- 181  270  SHERBERT  21 20  195 15  137 6  PUDDINGS  9  185  107  143  222 120  300 185  485 284  1103 419  ICE-CREAM  DESSERTS-FRUIT 044-183 044-184  APPLE APPLESAUCE  328 118  151  Table 4.4  (Continued)  004-185 044-186 044-187 044-188 044-192  BANANA CANTALOUPE GRAPEFRUIT ORANGE PINEAPPLE  045-193  DRIED FRUIT  206 113  38 29 27 122 90  100 100 200 150  44 135 159 37  174 55 159 226 18  33  DESSERTS-SWEETS 047-195 047-197  HONEY SUGAR  8  47 68  105  33 109  18 151  048-198 048-199  JAMS SYRUP  107 332  57 94  100 40  99 129  123 95  049-201 049-202  CHOCOLATE CANDY.. MARSHMALLOW  8  33 68  44  11  MISCELLANEOUS 051-205  POT PIES  227  199  261  063-217  SPAGHETTI + MEAT.  330  179  107  067-221  COCOA MIX Total Grams/Week.  15313 14679  52  7  12891  18871  19463  16831  20  Total # of Items.  3  22  75  83  70  82  #..of Initial Items  3  2  2  83  53  61  152  Table 4.5  Nutrient composition of SID-1 and SID-2, and the revised diets developed using quadratic objective functions .: (Eqns. 4.8, 4.9)  153 Nutrient  Nutrient Minimum (/week) PROTEIN(gm) 392.00 CHO-T(gm) 2735.4 T-FAT(gm) 44.209 KCAL 18899. CHO-F(gm) 79.376 SFA(gm) .0 SUCR(gm) .0 PUFA(gm) 44.209 VIT-A(iu) 35000. VIT-D(iu) 700.00 VIT-E(mg) 63.000 VIT-c(mg) 210.00 THIA(mg) 9.9470 RIBO(mg) 11.936 NIAC(mg) 131.30 VIT-B6(ug) 14000. FOLATE(mg) 1400.0 POTAS(mg) 9800.0 CAL(mg) 5600.0 PHOSP(mg) 5600.0 IRON(mg) 70.000 MAG(mg) 2205.0 CA/P .8 P/S 1.8  Constraints Maximum (/week) 558.60 4476.1 663.13 20889. 199.15 221.04 746.02 663.13 140000. 4200.0 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  SID-1. Initial (/week) 512.54 2809.9 92.428 14210. 35.125 9.6000 30.925 22.400 0.0 2319.8 56.550 56.580 16.329 13.963 148.67 20777. 2019.5 17082. 8172.8 15349. 73.195 4471.8 0.53245 4.0000  SID-1 Eqn. 4.8 (/week) 445.72 3186.9 511.45 18899. 79.576 139.79 291.22 279.59 78874. 1765.4 63.000 1440.5 11.669 12.413 131.30 19814. 1682.1 19600. 8960.0 11200. 110.24 3777.1 0.8000 2.0000  SID-1 Eqn. 4.9 (/week) 520.27 2808.5 663.13 18899. 79.576 201.04 591.58 372.48 92705. 1612.1 97.001 1105.0 11.831 12.062 131.30 14000. 1649.3 19600. 8960.0 11200. 110.56 3401.8 0.8000 1.8528  Nutrient  Constraints Maximum (/week) 558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 140000. 4200.0 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  SID-1: Initial (/week) 713.82 1872.4 981.39 20195. 32.250 367.46 549.60 543.42 69952. 1111.3 75.273 886.80 10.292 13.803 178.25 12942. 1460.9 22490. 6137.3 11258. 120.08 2386.5 .54514 1.4789  SID-2 Eqn. 4.8 (/week) 540.24 2735.4 612.94 19239. 79.576 221.04 746.02 332.18 103460. 705.06 93.305 1436.5 12.694 13.931 131.30 14000. 1745.2 19600. 8960.0 11200. 140.00 2875.1) .80000 1.5028  SID-2 Eqn. 4.9 (/week) 550.68 2735.4 638.72 19754. 79.576 221.04 746.02 359.57 120730. 700.00 110.41 1224.0 12.408 13.479 133.04 14000. 1977.3 19600. 8960.0 11200. 136.93 2800.3 .80000 1.6267  Nutrient Minimum (/week) PROTEIN(gm) 392.00 CHO-T(gm) 2735.4 T-FAT(gm) 44.209 KCAL 18899. CHO-F(gm) 79.576 SFA(gm) .0 SUCR(gm) .0 PUFA(gm) 44.209 VIT-A(iu) 35000. VIT-D(iu) 700.00 VIT-E(mg) 63.000 VIT-C(mg) 210.00 THIA(mg) 9.9470 RIBO(mg) 11.936 NIAC(mg) 131.30 VIT-B6(ug) 14000. FOLATE(mg) 1400.0 POTAS(mg) 9800.0 CAL(mg) 5600.0 PHOSP(mg) 5600.0 I RON (nig) 70.000 MAG(mg) 2205.0 CA/P .8 P/S 1.0  154  Although each of these formulations produces revised diets which are optimal solutions in the sense that they minimize total squared deviation, the revised diets are expected to be different because of the different penalty coefficients, w^.  Using Equation 4.9, the size of the  absolute deviation of items from i n i t i a l levels should increase as i n i t i a l consumption level rises.  In contrast, the absolute deviation associated  with different i n i t i a l amounts should not vary with the equation using direct squared deviation (Eqn. 4.8), since the penalty coefficient does not incorporate a term for i n i t i a l consumption levels. This phenomenon is illustrated in Table 4.6 which compares the average absolute deviation of items in the revised diets from their i n i t i a l consumption levels*  Where the penalty coefficient includes a term for i n i t i a l  consumption (Eqn. 4.9), the average absolute deviation rises from 61 grams at a consumption level of 0 grams/week to 5968 grams at a consumption of 9351-9400 grams/week for the SID-1 revised diet, and from 22 grams at a consumption level of 0 grams/week to 2800 grams at a consumption of 27512800 grams/week for the SID-2 revised diet.  This extent of rise is not  evident with the other equation (Eqn. 4.8)/ The average absolute deviation of items at each i n i t i a l consumption level is not exactly proportional to the penalty weighting applied.  If  they were, for example, the objective using direct squared deviation (Eqn.. 4.8) would be expected to produce the same average absolute deviation for items at each of the i n i t i a l consumption levels, but i t does not. This discrepancy is caused by the unequal distribution of nutrients across items.  Those which are efficient sources of nutrients for a given  situation deviate more widely.  If all items made the same nutrient  contribution, the quadratic objectives would spread deviations over all items in exact proportions to the penalty coefficient's  relative weighting,  155  in order to satisfy nutrient constraints.  On the other hand, i f each  item contributed only one unique nutrient, the solution outcome would be insensitive to different weighting coefficients.  Each item would be a  most efficient source of one particular nutrient.  Table 4.6  Average absolute deviation from i n i t i a l consumption levels for items in revised diets developed from SID-1 and SID-2 using quadratic objective functions (Eqns. 4.8, 4.9)  Initial Consumption (grams/item) 0 0-50 51-100 101-150 151-200 201-250 251-300 301-350 351-400 401-450 451-500 501-550 701-750 1551-1600 2751-2800 3201-3250 5651-5700 9351-9400  Average Absolute Deviation From SID-1 # of Eqn. 4.8 Eqn. 4.9 Items (grams) (grams) 124  46  61  1  280  280  1 1  1461 1602  Average Absolute Deviation From SID-2 # of Eqn. 4.8 Eqn.4.9 Items (grams) (grams) 44 22 22 8 14 1 4 3 2 1 1 1 1 1 1 1  56 43 82 74 129 28 102 167 131 37 378 7 7JO  100 28 29 11  22 32 77 81 191 34 261 190 303 338 495 59 732 1294 2800 1348  3681 5968  The different treatment- of deviations by these two objective functions results in marked differences in the solution profiles.  With respect to  the number of items contained in the diets developed from SID-1 (foot of Table 4.4), both objectives have retained two of the three original items. However, the objective function using percentage squared deviation (Eqn. 4.9)  156  has added 73 new items to the diet for a total of 75 items in the revised diet, whereas the other objective (Eqn. 4.8) has only added 20 new items for a total of 22 items in the revised diet.  Thus, each objective has  enlisted new items, but the percentage formulation has enlisted a much greater number. This difference is to be expected since the percentage formulation applies a heavy penalty load for deviation of small-quantity items in the initial diet compared to an equivalent deviation in a large-quantity item, while the other formulation does not.  Thus, the efficiency of any small-  quantity item to supply nutrients to satisfy the constraints is quickly exhausted with the percentage formulation. longer sequence of alternate efficient  The consequence is that a  sources must be incorporated into  the revised diet based on minimum percentage deviation, i f these cannot be claimed from items already included in the diet. With respect to the number of items in the diets developed from SID-2. the percentage formulation has retained 61 of the original 83 items and added 21 new items for a total revised solution of 82 items.  The solution  using direct squared deviation has retained 53 of the original 83 items and added 17 items for a total revised solution of 70 items. Unlike the revised diets developed from the SID-1, those developed from SID-2 do not demonstrate such a large difference in the number of new items entering the solution.  In fact, the revised diet developed from  SID-2 using the percentage formulation enlists only a few more new items than the revised diet based on direct squared deviation - - 17 and 21, pectively, as compared, to 20 and 73, for the SID-1 "solutions.  res-  The change: in  relative proportion of new-item entry for the two formulations is due to the availability of additional consumed items in the SID-2.  In this instance,  157  additional items have provided the percentage formulation with a relatively greater number of efficient nutrient sources, and consequently relatively fewer new items have entered the solution.  Because larger-quantity items  can deviate widely compared to smal1-quantity items with the percentage formulation, less concentrated sources of nutrients can be used efficiently to meet the constraints provided these items are present in the diet in appropriate amounts. Although, in terms of the number of items, there is not a great disparity between the revised diets developed from SID-2 with these two formulations, as compared to those from SID-1, differences are apparent. The revised diet developed from SID-2 by the percentage formulation has more liberally included both original items - - 61 as compared with 53 from the direct formulation  arid new items - - 21 as compared with 17 from  the direct formulation.  It is not clear that the percentage formulation  should necessarily incorporate more items in revised solutions developed with every diet.  Both formulations, direct and percentage squared  deviation are quadratic and hence tend to exhaust the efficiency of any particular item to act as an efficient nutrient source. deviation over items.  This spreads  In this instance, however, the nutrient distribution  over items and the quantities of the items consumed has resulted in the percentage formulation incorporating a greater number of both non-consumed and consumed items in the revised diet. The percentage formulation does not necessarily have a conceptual advantage over the dilrect formulation.  The formulation incorporating  percentage squared deviation provides the apparent advantage of adjusting the unit value of deviations relative to their i n i t i a l consumption level. Thus, regardless of the i n i t i a l amount consumed, the penalty associated  158  with decreasing to zero or doubling in amount is the same.'  Viewed one  way, with the percentage formulation small-quantity items are not sacrificed to meet nutrient constraints while large-quantity items are pre- „ ferentially maintained, as is the case with the formulation using direct squared deviation.  On the other hand, the formulation with percentage  squared deviation implies that the penalty for seemingly large deviations in the quantities of large-quantity items should be penalized the same as an equal-percentage, but relatively small, absolute change in smallquantity items.  4.3.2.2.2  Penalty Coefficient, w^, Based on Initial Consumption  Penalty.coefficients were chosen to differentially penalize i n i t i a l l y consumed versus non-consumed items, as follows: (4.10)  minimize  I E w-(x-) ( • - x-) i=l 1  2  1  I 2 minimize E w.(x.) (x. - x.) i=l  (100 i f x° = 0 where w.(x-) = ( (.1 i f x? > 0 (10,000 i f x? = 0  0  (4.11)  1  1  1  1  1  (4.12)  minimize  1 £ —— i=l (x? + .01) !  T  ( . x  1  where w-(x-) = ( o (1 i f x° > 0 (  1  o - xV)  1  2  1  For the formulations using direct squared deviation (Eqns. 4.10, 4.11), penalty weightings of 100 and 10,000. respectively, were applied to items not consumed in the client's original diet.  1  For the formulation using  percentage squareddeviation (Eqn. 4.12), a penalty value of 10.000 was applied to non-consumed items by defining epsilon as .01 over all items. Although this is not strictly equivalent to the expression:  for originally consumed items, the error introduced for portion-sized itemsrs'hould be small.  Revised diets (Table 4.7, 4.8) were developed  from SID-1 and SID-2 using these objectives (Eqns. 4.10-4.12).  The  nutrient composition of the i n i t i a l and revised diets are tabulated in Tables 4.9 and 4.10.  160  Table 4.7  GROUP-ITEM CODE #  Revised diets developed from SID-1 using quadratic objective functions (Eqns. 4.9, 4.12). FOOD ITEM [nitial  SID-1 (grams/week)* Eqn. 4.12 Eqn; 4.9  ENTREE-DAIRY 001-003 001-004 001-005  CHEDDER CHEESE... COTTAGE CHEESE... CREAM CHEESE  002-006 002-008  SOUR CREAM YOGHURT  003-010  EGG  354  358  13  14  203 50  202 51  14  15  3407 25  3403 25  49 82 182  49 82 182  82  82  182  182  34 50  34 51  ENTREE-CEREALS 004-012 004-013  CORN CEREAL WHEAT CEREAL  005-015 005-016  OATMEAL CEREAL... WHEAT CEREAL  006-017  PANCAKES  007-018 007-019  NOODLES SPAGHETTI  008-020 008-021 008-023  RICE, brown RICE, white WHEAT GERM  009-024 009-027 009-028  FRENCH BREAD WHITE BREAD WHOLE WHEAT BREAD 1  010-031 010-032 010-033 010-035  BUSCUITS HAMBURGER BUN.... MUFFIN '-" ENGLISH MUFFIN...  011-037 011-040  SALTINES RYE KRISP  9375 280  1  ENTREE-MEATS 012-041 012-042 012-043 012-046 012-047 012-048  BEEF, BEEF, BEEF, PORK, PORK, BACON  30% f a t . . . . 20% f a t . . . . 15% fat lean cuts.. a l l hams...  013-049 013-051  CHICKEN, steamed. CHICKEN, f r i e d . . .  014-055 014-056 014-058  FRIED FISH BROILED FISH OYSTERS  Values rounded to the nearest gram/week.  Table 4.7 (Continued) 014-062 014-063 014-065  SARDINES SHRIMP TUNA  015-067  LIVER-.-*...i'uY  017-069 017-070 017-071  FRANKFURTERS FRESH SAUSAGES... LIVERWURST  018-073 018-075  BEANS, dri-ed SOYBEANS  019-076 019-077 019-079 019-080 019-081  ALMONDS CASHEW NUTS PEANUT BUTTER.... PEANUTS PECANS  1  126 33  126 33  73  74  88 79  88 79  ENTREE-VEGETABLES 020-082 020-083 020-084 020- 085  POTATOE, baked... POTATOE, f r i e d . . . POTATOE, mashed... SWEET POTATOE....  29 152  30 152  021- 089 021-091 021-092 021-095 021-098 021-099 021-101  BEANS, green BROCOLLI CABBAGE LETTUCE PEAS PEPPERS SPINACH  235 352 173 83 355 304 231  235 352 173 84 354 304 232  022-103 022-104 022-105 022-106 022-111  BEETS CARROTS, cooked.. CARROTS, raw CORN TOMATOE  163 205 181 25 59  162 204 181 26 59  023-113 023-114 023-115  CUCUMBER MUSHROOMS ONIONS  39 37 117  39 37 117  024-116  SUCCOTASH.  210  210  025-117 025-118  OLIVES PICKLES, sweet...  025- 119  PICKLES, sour....  374 93 82  373 93 82  ENTREE-FATS 026- 121 026-124  LARD SOYBEAN OIL  70 67  69 66  027-125  BUTTER  52  52  028-127 028-128  CHEESE SAUCE GRAVY  59  60  162  Table 4.7 (Continued) 029-132 029-133  MAYONNAISE SALAD DRESSING  27 49  27 49  1977  1 1936  13 23  14 23  6  6  44  44  39  39  17  18  20  20  BEVERAGES-DAIRY 030-136 030- 137  WHOLE MILK SKIM MILK  031- 140 031-141  TABLE CREAM WHIPPED CREAM  5658  BEVERAGES-FRUIT 032-143 032-144 032-145 032- 148  APPLE JUICE GRAPEFRUIT JUICE.. LEMON JUICE ORANGE JUICE  033- 150  TOMATOE JUICE BEVERAGES-MISC.  034- 151  COLA-TYPE  035- 153 035- 154  COFFEE TEA  036- 155 036-156 036-158  BEER DISTILLED SPIRITS. DRY WINES SOUPS  037-160 037-161 037-163  CREAMED SOUPS PEA SOURS MEAT + VEGIE SOUPS DESSERTS-CEREALS  038-166 038-168 038- 170  COFFEE CAKE FRUITCAKE ICED CAKES  22  22  039- 171 039- 172  FRUIT PIES PUMPKIN PIES  97 82  97 82  040- 173 040- 174  COOKIES FRUIT COOKIES,,, ,  447  445  041- 175 041-177  CAKE DOUGHNUTS DANISH PASTRY  21 20  22 20  9  10  222 120  221 119  J.  DESSERTS-DAIRY 042-178 042- 179  ICE-CREAM SHERBERT  043- 181  PUDDINGS DESSERTS-FRUIT  044-183 044-184  APPLE APPLESAUCE  163  Table 4.7 (Continued) 044-185 044-186 044-187 044-188 044- 192  BANANA CANTALOUPE GRAPEFRUIT ORANGE PINEAPPLE  38 ?29 27 122 90  39 '30 27 122 91  045- 193  DRIED FRUIT  33  35  DESSERTS-SWEETS 047-195 047- 197  HONEY SUGAR  47 68  47 68  048- 198 048- 199  JAMS SYRUP  57 94  57 94  049- 201 049-202  CHOCOLATE CANDY... MARSHMALLOW  33 68  34 68  52  54  MISCELLANEOUS 051-205  POT PIES  063-217  SPAGHETTI + MEAT..  067-221  COCOA MIX  -V  Total Grams/Week..  15313  12891  12862  Total # of Items..  3  75  76  # of Initial Items  3  2  2  164  Table 4.8  Revised diets developed from SID-2 using quadratic objective functions(Eqns. 4.8-4.12)  GROUP-ITEM FOOD ITEM CODE #  SID-2 (grams/week)* Initial Eqn.7.8 Eqn.4.1C Eqn.4.11 Eqn.4.S Eqn.4.12  ENTREE-DAIRY 001-003 001-004 001-005  CHEDDER CHEESE COTTAGE CHEESE CREAM CHEESE...  002-006 002-008  SOUR CREAM YOGHURT  003-010  EGG  182  + 495  519  117  553  546 •  618  696  6  ENTREE-CEREALS 004-012 004-013  CORN CEREAL.... WHEAT CEREAL...  28  210 221  421 13  431 0.14  108 38  42 0.03  005-015 005-016  OATMEAL WHEAT CEREAL...  270  187 24  185 3  174 0.03  207 21  125 0.01  006-017  PANCAKES  135  007-018 007-919  NOODLES SPAGHETTI  008-020 008-021  RICE, brown.... RICE, white  008-021  WHEAT GERM  009-024 009-027 009-028  FRENCH BREAD... WHITE BREAD.... WHOLE WHEAT BREAD  40 414  010-031 010-032 010-033 010-035  BISCUITS HAMBURGER BUN.. MUFFIN ENGLISH MUFFIN.  70 46 40 56  011-037 011-049  SALTIMES RYE KRISP  24  47  75 39  6 1  0.06 0.02  3  0.04  63 451  50 370  138  5  83  2  44 357 0.05  184  467  478  6 221  146 13  147 0.14  29 11 62 752 66 2 73 28 186 54 88  0.01 0.01 44 981 0.02 16 53 34 95 27 0.01  ENTREE-MEATS 012-041 012-042 012-043 012-046 012-047 012-048 *  BEEF, 30% f a t . . BEEF, 20% f a t . . BEEF, 15% f a t . . PORK, lean cuts PORK, a l l hams. BACON.  170 85 170 '85 43 64  21  Values rounded to the nearest gram/week, except values below 0.5 grams/ week which are rounded to two decimal places.  165 Table 4.8 (Continued) 013-049 013-051  CHICKEN, steamec CHICKEN, fried.  350  014-055 014-056 014-058 014-062 014-063 014-065  FRIED FISH BROILED FISH... OYSTERS SARDINES SHRIMP TUNA  43 60  015-067  LIVER  017-069 017-070 017-071  FRANKFURTERS... FRESH SAUSAGES.. . LIVERWURST  019-076 019-077 019-079 019-080 019-081  ALMONDS CASHEW NUTS PEANUT BUTTER.. PEANUTS PECANS  il.  22  167  0. Oil.  211  2  0. 03  22  0. 03  37  40 98 15 28 120  115 10 85  19  133 0. 21  61 15 43 112  17 67  5 0. 03 15 20 36 0. 04  ENTREE-VEGETABLE S 020-082 020-083 020-084 020-085  POTATOE, baked.. POTATOE, fried. POTATOE, mashed SWEET POTATOE..  100 100 200 180  82 218  324  021-089 021-091 021-092 021-095 021-098 021-099 021-101  BEANS, green... 65 BROCOLLI +• 63 CABBAGE 65 LETTUCE 190 PEAS 178 PEPPERS 85 SPINACH  253 363 180 178 572 374 182  368 683 377 97 1134 11 202  022-103 022-104 022-105 022-106 022-111  BEETS CARROTS, cooked CARROTS, raw CORN TOMATOE  518  158 146 70 109 448  165 11 44 9 327  023-113 023-114 023-115  CUCUMBER MUSHROOMS ONIONS  275  210  52  29  8  119  357  381  024-116  SUCCOTASH  159  4  025-117 025-118 025-119  OLIVES PICKLES, sweet. PICKLES, sour..  380  7  80 76 50  7 387 368 708 401 81 1169 0. 11 212  125 449  84 539  199 277 175 271 924 95 214  141 181 130 321 1247 0. 04 187  178 166 113 49 577  151 137 77 0. 02 724  230 1 48  198 0. 00 9  0 04  52  0. 02  0 07  84 20 30  0. 03 21 33  153 34 0. 10 310  20 34  ENTREE-FATS 026-121 026-124  LARD SOYBEAN O I L . . . .  027-125  BUTTER  5 125  22  63  78  29  29  5 1  16  166 Table 4.8 (Continued) 028-127 028-128  CHEESE SAUCE... GRAVY  72  029-132 029-133  MAYONNAISE SALAD DRESSING.  105 30  21 70  65  65  446  421  0.2? 28  46  44 74  50 78  BEVERAGES-DAIRY 030-136 030-137  WHOLE MILS SKIM MILK  732  632  031-140 031- 141  TABLE CREAM.v.. WHIPPED CREAM..  185 8  93  60 8  BEVERAGES-FRUIT 032- 143 032-144 032-145 032- 148  APPLE JUICE.... GRAPEFRUIT JUICE LEMON JUICE ORANGE JUICE...  033- 150  TOMATOE JUICE..  120 5 360  174  339  328  350  350  332  5  BEVERAGES-MISC. 034-151  COLA-TYPE  035-153 035-154  COFFEE TEA  ?800 1600  2771 1572  2711 1516  2705 1508  306  036-155 036-156 036-158  BEER 3240 DISTILLED SPIRITS 43 DRY WINES 400  3229 34 324  3363 28 241  3384 28 235  4588 41 154  145 99  25 37  7 22  155 128 3  347  7878 42  SOUPS 037-160 037-161 037- 163  CREAMED SOUPS... 198 PEA SOUPS 200 MEAT + VEGIE SOUP  112 81 0.00  DESSERTS-CEREALS 75  11  27  60  28  43  307 202  361 224  038- 166 038-168 038-170  COFFEE CAKE FRUITCAKE ICED CAKES  039-171 039- 172  FRUIT PIES PUMPKIN PIES....  160 150  040- 173 040- 174  COOKIES FRUIT COOKIES....  60  041- 175 051-177  CAKE COUGHNUTS... DANISH PASTRY....  60 38  219 123  346 45  400  60 9  352 18 58 0.09  27 106  42 0.04 22 26  DESSERTS-DAIRY 042- 178 042-179  ICE-CREAM SHERBERT  270  195 15  85 1  66 0.01  137 6  043-181  PUDDINGS  185  107  35  19  143  9 0.00 116  167 Table 4.8 (Continued) DESSERTS-FRUIT 044-183 044-184 044-185 044-186 044-187 044-188 044-192  APPLE APPLESAUSE BANANA CANTALOUPE GRAPEFRUIT ORANGE PINEAPPLE  045-193  DRIED FRUIT  300 185 100 100 200 150  485 284 44  644 423 370  135 159 37  43 1  643 430 430 24 0.01  1105 419 174 55 159 226 18  1920 552 184 66 137 253 0.01  DESSERTS-SWEETS 047-195 047-197  HONEY SUGAR  105  33 109  2 190  0.02 197  18 151  0.01 185  048-198 048-199  JAMS SYRUP  100 40  99 129  157 283  162 312  123 95  136 54  049-201 049-202  CHOCOLATE CANDY.. MARSHMALLOW  44  2  0.03  11  0.01  MISCELLANEOUS 051-205  POT PIES  227  199  299  312  261  063-217  SPAGHETTI + MEAT. 330  179  45  28  107  067-221  COCOA MIX  7  Total Grams/Week.18871 18463  286  20  8  18748  18748  16831  19850  83  70  70  67  82  86  # of Initial Items 83  53  50  47  61  66  # of New Items...  17  20  20  21  20  Total # of Items.  168  Table 4.9  Nutrient  PROTEIN(gm) CHO-T (gm) T-FAT(gm) KCAL CHO-F(gm) SFA(gm) SUCR(gm) PUFA(gm) VIT-A(iu) VIT-D(iu) VIT-E(mg) VIT-C(mg) THIA(mg) RIBO(mg) NIAC(mg) VIT-B6(ug) FOLATE(mg) POTAS(mg) CAL(mg) PHOSP (mg) IRON(mg) MAG(mg) CA/P P/S  Nutrient composition of SID-1, and the revised diets developed using quadratic objective functions (Eqns. 4.9, 4.12).  Nutrient Minimum (/week) 392.00 2735.4 44.209 18899. 79.576 .0 .0 44.209 35000. 700.00 63.000 210.00 9J9470 11.936 131.30 14000. 1400.0 9800.0 5600.0 5600.0 70.000 2205.0 .8 1.0  Constraints Maximum (/week) 558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 .14000E+06 4200.0 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  SID-1 Initial (/week) 512.54 2809.9 92.428 14210. 35.125 5.6000 30.925 22.400 0.0 2319.8 56.550 56.580 16.329 13.963 148.67 20777. 2019.5 17082. 8172.8 15349. 73.195 4471.8 0.53245 4.0000  SID-1 Eqn. 4.9 (/week)  SID-1 Eqn. 4.1 (/week)  520.27 2808.5 663.13 18899. 79.576 201.04 591.58 372.48 92705. 1612.1 97.001 1105.0 11.831 12.062 131.30 14000. 1649.3 19600. 8960.0 11200. 110.56 3401.8 0.8000 1.8528  520.46 2808.9 663.13 18899. 79.576 20.179 593.21 371.84 92840. 1617.1 97.018 1105.3 11.824 12.039 131.30 14000. 1649.6 19600. 8960.0 11200. 110.72 3408.1 0.80000 1.8427  169  Table 4.10  Nutrient composition of SID-2, and the revised diets developed using quadratic objective functions (Eqns. 4.8-4.12)  170  Nutrient  PROTEIN(gm) CHO-T(gm) T-FAT (gin) KCAL CHO-F(Gm) SFA(gm) SUCR(gm) PUFA(gm) VIT-A(iu) VIT-D(iu) VIT-E(mg) VIT-C(mg) THIA(mg) RIBO(mg) NIAC(mg) BIT-B6(ug) FOLATE(mg) POTAS(mg) CAL(mg) PHOSP(mg) IRON(mg) MAG(mg) CA/P P/S Nutrient PROTEIN(gm) CHO-T(gm) T-FAT(gm) KCAL CHO-F(gm) SFA(gm) SUCR(gm) PUFA(gm) VIT-A(iu) VIT-D(iu) VIT-E(mg) VIT-C(mq) THIA(mg) RIBO(mg) NIAC(mg) VIT-B6(ug) FOLATE(mg) POTAS(mg) CAL(mg) PHOSP(mg) IRON(mg) MAG(mg) CA/P P/S  Nutrient Minimum (/week)  Constraints Maximum (/week)  SID-2 Initial ((/week)  SID-2 Eqn. 4.8 (/week)  SID-2 Eqn. 4.10 (/week)  Eqn.4.1 (/week)  558U60 2735.4 575.66 19026. 79.576 221.04 746.02 308.42 94146. 896.00 102.51 1361.2 11.699 12.421 137.02 14000. 1755.0 19600. 8419.0 10524. 140.00 2772.8 .80000 1.3953  558.60 2735.4 572.11 18999. 79.576 221.04 746.02 305.72 91999. 980.41 101.04 1382.6 11.452 11.936 139.70 14000. 1718.1 19600. 8266.1 10333, 140.00 2766.9 .80000 1.383  392.00 2735.4 44.209 18899. 79.576 .0 .0 44.209 35000. 700.00 63.000 210.00 9.9470 11.936 131.30 14000. 1400.0 9800.0 5600.0 5600.0 70.000 2205.0 .8 1.0  558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 140000. 4200.0 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  713.82 1872.4 981.39 20195. 32.250 367.46 549.60 543.42 69952. 1111.3 75.273 886.80 10.292 13.803 178.25 12942. 1460.9 22490. 6137.3 11258. 120.08 2386.5 .54514 1.4789  540.24 2735.4 612.94 19239. 79.576 221.04 746.02 332.18 103460. 705.06 93.305 1436.5 12.694 13.931 131.30 14000. 1745.2 19600. 8960.0 11200. 140.00 2875.1 .80000 1.5028  Nutrient 'Minimun (/week) 392.00 2735.4 44.209 1.8899. 79.576 .0 .0 44.209 35000. 700.00 63.000 210.00 9.9470 11.936 131.30 14000. 1400.0 9800.0 5600.0 5600.0 70.000 2205.0 .8 1.0  Constraints Maximum (/week) 558.60 4476.1 663.13 20889. 159.15 221.05 746.02 663.13 140000. 4200.0 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  SID-2 Initial (/week) 713.82 1872.4 981.39 20195. 32.250 367.46 549.60 543.42 69952. 1111.3 75.273 886.80 10.292 13.803 178.25 12942. 1460.9 22490. 6137.3 11258. 120.08 2386.5 .54514 1.4789  SID-2 Eqn. 4.9 (/week) 550.68 2735.4 636.72 19754. 79.576 221.04 746.02 359.57 120730. 700.0 110.41 1224.0 12.408 13.479 133.05 14000. 1977.3 19600. 8960.0 11200. 136.93 2800.3 .80000 .80000  SID-2 Eqn. 4.12 (/week) 538.24 2735.4 586.78 20075. 79.576 221.04 746.02 310.87 12T670. 700.00 88.360 1145.9 11.339 13.373 148.24 14000. 1939.2 19600. 8960.0 11200. 128.56 2702.6 .80000 1.4064  171  The solution dynamics for formulations which selectively penalize change in non-consumed versus consumed items are analogous to those previously discussed which penalize on the basis of percentage or absolute deviation.  In this instance, however, the penalty curves slope is further  altered as a function of the penalty weighting applied to selected items. If a change in a particular item is heavily or lightly penalized, the slope of the penalty curve is proportionately increased or decreased, respectively.  Consequently, the size of the average absolute deviation for  an item tends to decrease as the penalty weighting increases, and increase as the penalty decreases.  This phenomenon is illustrated in Table 4.12  which compares the average absolute deviation from i n i t i a l consumption levels for each of the revised diets developed from SID-2. applied to non-consumed items increases  As the penalty  relative to that  for consumed items, the average absolute deviation of non-consumed items decreases.  For example, consider the objective functions using direct  squared deviation (Eqns. 4.8, 4.10. 4.11).  With ehese the absolute  deviation from zero for non-consumed items has decreased from 56 to 3 to 0.03 with penalty ratios for non-consumed/consumed items of 1/1, 100/1, and 10,000/1, respectively.  Interestingly, the effect is roughly proportional  to the increase in penalty assignment.  Similarly, for the percentage  formulations (Eqns. 4.9, 4.12) an increase in relative penalty on non-con-u sumed items of 1 to 10,000 has reduced the average absolute deviation from 22 to 0.01.  The decreased deviation in non-consumed items has been reflec-  ted by a reasonably consistent increase in the deviation of the categories of consumed items, as shown in Table 4.12. Unlike the revised diets developed from SID-2, those developed from SID-1 (Table 4.11) using the percentage formulation (Eqns. 4.9, 4.12) are  172  virtually insensitive to increased penalty weightings on non-consumed items.  This different outcome can be explained by the large number of non-  consumed items in the i n i t i a l diet relative to consumed items.  The uniform  increase of the penalty over all non-consumed items does not appreciably alter the solution dynamics among the vast majority of items in the diet. Also, i t is apparent that the u t i l i t y of the three initially-consumed items to supply nutrients has not changed much, even with the 10,000 fold increase in penalty on the non-consumed items.  Table 4.11  Initial Consumption (grams/week) 0 >0-5O 51-100 101-150 151-200 201-250 251-300 301-350 351-400 401-450 451-500 501-550 701-750 1551-1600 2751-2800 3201-3250 5651-5700 9351-9400  Average absolute deviation from i n i t i a l consumption levels for items in revised diets developed from SID-1 using quadratic objective functions (Eqns. 4.9, 4.12).  Average Absolute Deviation From SID-1 # U  f m  s  ^  4  g r a m s )  J  4  1 2  g r a m s )  124  61  61  1  280  280  1 1  3681 5968  3722 5972  173  Table 4.12  Average absolute deviation from i n i t i a l consumption levels for items in revised diets developed from SID-2 using quadratic objective functions (Eqns. 4.8-4.12).  Initial Consumption ^grams/week; 0 >0-50 51-100 101-150 151-200 201-250 251-300 301-350 351-400 501-450. 451-500 501-550 701-751 1551-1600 2751-2800 3201-3250 .  #  Q  f  I t e m s  44 22 22 8 14 1 4 3 2 1 ]  1 1 1 1 1  Average Absolute Deviation From SID-2 _ . i (grams) (grams) (grams) (grams) (grams)  E a n  4  >  56 43 82 74 129 28 102 167 131 37 378 70 100 28 29 11  g  E q n >  4  /  3 102 129 109 248 72 209 212 260 44 489 191 286 84 89 123  m  £ q n >  4  >  1  1  0.03 104 136 116 256 85 222 218 263 57 495 208 311 92 95 144  E q n >  4  g  22 32 77 81 191 34 261 190 303 338 495 59 732 1294 2800 1348  E q n  4 >  0 11 59 95 :260 59 526 226 380 567 495 256 732 1600 2800 4638  2  174  Penalizing deviation of non-consumed items relative to consumed items has resulted in only minor changes in the solution profiles.  With respect  to the number of items contained in the diets developed from SID-1 (Table 4.7), both objectives original items.  (Eqns. 4.9, 4.12) have retained two of the three  The percentage objective weighting against non-consumed  items (Eqn.. 4.12) has added 74 new items; one greater than the other percentage objective (Eqn. 4.9).  The greater penalty load on non-consumed  items has altered the relative u t i l i t y of one additional item to provide nutrients to meet the constraints.  This item has subsequently entered  the solution at 1 gram per week. With respect to the number of items in diets developed from SID-2 (Table 4.8),  some of the differences between the revised diets developed  with the percentage formulation  (Eqn. 4.9) arid with the direct formulation  (Eqn. 4.8), observed previously, have been amplified with the formulations (Eqn, 4.10-4.12) penalizing non-consumed item deviations.  In particular,  the number of consumed items retained in the direct formulation solutions have decreased further, from 53 without additional penalty on non-consumed items (Eqn. 4.9), to 50 and 47 with penalties of 100 (Eqn. 4.10) and 10,000 (Eqn. 4.11) on non-consumed items, respectively.  The percentage formulation  with additional penalties on non-consumed items (Eqn. 4.12) has increased the number of consumed items retained - - 66 as compared to 61 from the formulation with no additional penalty applied to zero items (Eqn. 4.9). Again, i t is not clear that the outcome results from other than the c i r cumstances associated with this one particular diet. The number of i n i t i a l l y non-consumed items in the revised diets developed with the direct formulation have increased from 17 with no penalty on non-consumed items (Eqn. 4.8) to 20 with penalties on non-  175  consumed items (Eqns. 4.10, 4.11).  Conversely, the revised diet developed  with the percentage formulation which penalizes non-consumed item deviation (Eqn. 4.12) has 20 new items as compared to 21 from its counterpart without the additional penalty (Eqn. 4.9).  Although all the above  outcomes meet the objective of minimizing the summed deviation of items, the slight reduction in new items entering the solution developed with the percentage formulation is more consistent with the intent of penalizing the deviation of non-consumed items; that i s , the penalty should reduce the number of new items entering the solution rather than increase them, as is the case with the revised diets from the formulation using direct squared deviation. This discussion of phenomena associatediwith differently weighting the deviations of specific items has been restricted to a few examples where non-consumed versus consumed items have been differentially penalized. However, these observations should be applicable to other specific situations.  4.3.2.2.3  Penalty Coefficient, w., Further Comments  As discussed above, penalty coefficients the deviation of items from i n i t i a l levels.  can be used to influence How a particular penalty  coefficient will affect the solution is a product of a number of factors, namely:  the nutrient distribution in the items considered, the quantities  of the items consumed in the i n i t i a l diet, and the nutrient constraints applied.  The complex interactions inherent in these factors underlines  the real problem in trying to establish penalty coefficients which function to provide useful revisions for many different  situations.  176  4.3.2.3  First Term:  Further Modifications of the Algorithm  As presently formulated the objective function (Eqn. 4.1) does not control the direction or extent of change.  It would appear realistic  that the desirability of an increase or decrease in a particular item may be different, and hence that the penalty coefficient applied should reflect this difference.  In this regard, a quadratic objective can be  formulated to consider, separately, positive and negative deviations in item clusters.  In more detail, this algorithm finds the combination of  item clusters which minimizes the total, squared, weighted-positive and weighted-negative deviations in the amounts of each item cluster from i n i t i a l consumption levels, while satisfying nutrient constraints.  The  linear form of this objective function has been given already. (Eqn. 4.3). The formulation is as follows: (4.13)  . . .  minimize  1  ,  +  +  1  (4.14)  .-.2  -  E (w. x- + w. x.) i=l  subject to  1  1  1  xj - xT = x^ - x°  (i = 1, 2,  I)  (q = 1, 2,  Q)  I (4.15)  m  3  q ^  qi  X  i * "q  I (4.16)  r  u v  *  ^  (4.17  a  .\  111 a  x-, x., x. n  ui  x  i  t  > t  u v  v i * i  >0  where the terms are as previously defined (p. 142 ).  (u,v = any specified set of nutrient pairs)  (i = 1, 2,  I)  Further elaboration  of this algorithm could include separable programming to assign different penalty weighting to different portions of the curve. To ensure that changes fall within specified limits, constraints could  177  be appropriately included to provide minimum and maximum limits on the deviation in the amounts of specific item clusters.  This expression is  as follows: (4.18)  d. » (  Xi  - x?) » e,  where the newly defined terms are: d.j and e^ are, respectively, the upper and lower bound on the deviation in specific item clusters.  If d^ = 0 then the revised  consumption of the item cluster cannot be greater than the original consumption level.  If  = 0 then the revised consumption of the  item cluster cannot be less than the original consumption level. The constraints used should be carefully applied so as not to interfere with solution feasibility. The formulations just described (Eqns. 4.13-4.18) have not been tested, but provide the basis for desirable future work.  4.3.2.4  Second Term of the Objective Function  The first term of the objective function represents the acceptability of an item deviating from its i n i t i a l dietary levels.  In this f i r s t term  the acceptability of this deviation is described as independent of speci f i c concurrent changes in other items.  Further, the f i r s t term does not  consider changes in other dietary attributes, only in terms of the item clusters themselves. Presumably, however, the acceptability of an item deviating from initial dietary levels depends not only upon the nature of the deviation of that food i t s e l f , but also upon the nature of concurrent deviations of the other foods in the consumption pattern, and of changes in other characteristics of the diet and in the diet as a whole.  In any case,  178  such an assumption would appear more realistic than the assumption that food preferences are not modified by changes in the consumption levels of the different foods. Two basic types of compensatory relationships between foods have been described (Smith 1963), namely, incompatabilities and complementarities.  Incompatible foods are those which clash or disagree, absolutely  or in some proportion.  They typically vary in inverse proportion. On  the other hand, complementary foods either enhance one another, or their coexistence may be essential to palatability or acceptability. mentary foods tend to vary in direct proportion.  Comple-  These relationships are  applicable to item clusters and attribute groups. The second term of the model's objective function contains the mathematical expression used to minimize the deviation of attribute groups from their i n i t i a l levels, and to establish the relationships of concurrent change between item clusters and attribute groups.  At present, only  incompatibility between item clusters and attribute classes has been considered in establishing the relationships of concurrent change.  This  term minimizes the summed weighted square of differences between the amounts of item cluster groups - - called attribute groups - - in the i n i t i a l and recommended diets, as follows:  (4.19)  minimize  K z  J L  k-lj-!  k  o (x- - xV) )  P.. ( £ J  k  KG  2  J k  The complete formulation incorporating this additional term  given in  Equations 3.2 to 3.5. The addition of this second term is expected to modify the output profiles as compared to those producedwithout i t . . First, i f the deviations of attribute groups are increasingly penalized, then the quantity of these  179  attribute groups in the revised diet should approach i n i t i a l  levels.  Thus, this term allows for monitoring general food and diet characteristics considered important to the diet's acceptability.  Second, i f emphasis  on maintaining i n i t i a l levels of attribute groups increases, the attribute structure encourages the item clusters'within each attribute group to vary in inverse proportion. Thus, the second term tends to cause substitution between the presumed nearest acceptable alternative for an item cluster or attribute group. This relationship of concurrent change between item clusters and attribute groups i s , in a sense, an artifact of assigning those elements with like characteristics into the hierarchy of successive subgroups, groups, and supergroups, outlined in Appendix G. The number of item clusters contained in any attribute group and the number of attribute groups to which any item is assigned will whether substitution between items will be apparent.  influence  Where only two items  are contained in a group, the shift between these grouped partners should be more apparent than when a large number of items is contained in a group. Where an item cluster has simultaneous membership in many attribute groups, as is the case above, (Eqn. 4.19), the exact outcome is dependent on the relative emphasis on the deviation of each attribute group considered. '.It should be noted that the dynamics between the attribute groups on each hierarchical level, k=l to k=7, should be similar to those previously discussed for item clusters in the zero^ hierarchical level.  180  4.3.2.5  Second Term:  Shape of the Curve  The second term, as expressed in Equation 4.19, also describes a quadratic curve centered at the i n i t i a l consumption level of each attribute group defined, as shown in Figure 4.3.  Thus the relative unacceptabi1ity  of differences between the i n i t i a l and revised values for any attribute element is strongly intensified by this quadratic objective.  Figure 4.3  Graph of quadratic function for deviation from i n i t i a l consumption of attribute group Gj^.  Other non-linear or linear functions may be appropriately used to describe the acceptability of these attribute elements deviating from i n i t i a l levels, and to describe the interactional properties of individual items and of attribute groups.  The second term of the objective function  (Eqn. 4.19) can be readily transformed to the linear equivalent.  This  minimizes the total, weighted-positive and weighted-negative deviation in the amount of specified groups from i n i t i a l consumption levels.;  With  the addition of this linear second term the linear formulation (Eqns. 4.34.7) becomes: (4.20)  minimize  I z  + + - ^^k + + {vi. x. + x- v^) + E E (p- z n  k  k  + p- + z- ) k  k  181  (4.21)  subject to z*  - zT =  k  z  k  i € G  (xj - xT) (j = 1, 2, . . . , J ) k  Jk  (4.22)  xj - xT = x. - x°  (4.23)  I m ^ i  a  q  (k = 1, 2, . . . , K) (i . 1, 1,  x-  q i  n  I)  (q = 1, 2, . . . , Q)  q  I (4.24)  U 1  r  1  ^ i=1 > t I z a . x.  (u,v = any specified set of nutrient pairs)  y  u v  u v  x., xt, xT, ztu, zT, >, 0 i ' i ' i ' jk' jk >  (4.25)  x  x  x  z  z  Where the newly defined terms are as follows: Z  z  a  jk' jk  r  e  t  1 e  P  o s  1  t l v e  a n c  ' negative deviation, respectively, th of attribute group j in the k hierarchical level.  z  '  _ fi G ik " ^ (0  (xj  k  - xT )  if  k  J K  (i€G. J  (xt - xT) > 0  z 1 f e b  ik J  (X k  X  J k - jk)  (0  i  f  *  otherwise (4 " i) < X  0  .  otherwise  P j , p j are the weighted penalty associated with the positive k  k  or negative deviation, respectively, of attribute 1  group j in the k^* hierarchical level. Since the linear and quadratic formulations of the objective functions first term are expected to generate different solutions, we can correspondingly expect changes with these formulations incorporating both the first and second terms.  The quadratic objective would be expected to  moderate extreme fluctuations in any particular item cluster or attribute group in favour of numerous smaller changes.  The linear model, however,  182  should be insensitive to large deviations in specific items, provided these changes are consistent with the overall objective of reducing aggregate deviation of item clusters and attribute groups.  For example,  i f all foods provided the same nutrient contributions, the quadratic objective (Eqn. 3.2) would spread deviations over all items and attribute groups in exact proportion to the penalty coefficients applied. The linear objective (Eqn. 4.20) would be insensitive to how the total deviation was apportioned over item clusters and attribute groups.  These  expectations are not illustrated here since the linear solutions were not developed for comparison with the quadratic results.  4.3.2.6  Second Term:  Penalty Coefficients, w. and P.. 1  JK  The impact on the solution of altering the relative weighting of the penalty coefficients, w- and P.. , should also be considered. these penalty coefficients,  By altering  the value of increments for an item cluster  or attribute group can be exaggerated or depressed, and with i t the emphasis on concurrent change of items.  As previously discussed, the  numerical value of the penalty coefficients can be based on various rationales for perceived acceptability of change. can represent:  The coefficient chosen  differences in i n i t i a l consumption levels for each item  or group; whether an item or group is i n i t i a l l y consumed or not consumed; differences in standard serving sizes; the average consumption levels observed in a population; or some other measure of personal preference. In this instance, all penalty coefficients were arbitrarily assigned as "1" for deviation of any item cluster, and "1" or "0" for deviation of attribute groups depending on whether the deviation of a particular group was to be considered in the computation. In order to observe the impact of including the second term of the  183  objective function (Eqn. 4.19), three different objective functions (Eqns. 4.26-4.28) were defined.  Each of these objective functions minimizes the  summed, squared deviation of both item clusters and specified attribute groups from i n i t i a l levels.  A numerical value of 1 has been assigned for  weighting penalty coefficients,  \n. and P ^ .  The f i r s t objective function  (Eqn. 4.26) sums the deviations of all item clusters and of attribute groups on the first hierarchical level, whereas the second objective function (Eqn. 4.27) sums the deviations of all item clusters and of attribute groups on both the f i r s t and second hierarchical levels.  The  third objective function (Eqn. 4.28) sums the deviations over all item clusters and attribute groups on levels, k=l to k=7, and the total diet.'s amount, k=8.  The algebraic expressions corresponding to these objective  functions are as follows:  (4.26)  minimize  1 n 2 1 ^1 z 1(x. - xV) + • & z 1( z i=l k=l j=l i£G  I 2 ^2 z l(x. - x?) + z z i=l k=l j=l  n 2  ( x- - x.) ) 1 j k  2  (4.27)  minimize  1  (4.28)  minimize  1  I z Kx- - xV) i=l  Using these objectives  2  8 J z z k=l j=l  1( z i€G  (x, - x°.) )  2  1 jk  8  +  l(i • (xiCG jk 1  - x .) )2 u  1  (Eqns. 4.26-4.28) revised diets were developed  from Standard Initial Diet 2, and compared with the solution obtained from an objective function (Eqn. 4.8) previously described, which minimizes only the summed, squared deviation of item clusters from i n i t i a l quantities (Table 4.13).  Table 4.14 gives nutrient composition of the i n i t i a l and  revised diets for each of these objective functions (Eqns. 4.26-4.28).  T84  Table 4.13 GROUP-ITEM CODE #  Revised diets developed from SID-2 using quadratic objective functions (Eqns. 4.8, 4.26-4.28). FOOD ITEM Initial  SID-2 (grams/week)* Eqn. 4.8 Eqn. 4.26 Eqn.4.27  Eqn.4.28  ENTREE-DAIRY 1001-003 001-004 001-005  CHEDDER CHEESE. COTTAGE CHEESE. CREAM CHEESE...  002-006 002-008  SOUR CREAM YOGHURT  003-010  EGG  182  519  538  508  495-  117  96  7  499  ENTREE CEREALS 004-012 004-013  CORN CEREAL... WHEAT CEREAL..  23  210 221  196 87  167 112  005-015 005-016  OATMEAL WHEAT CEREAL..  270  187 24  115 96  35 103  006-017 007-018 007-019  PANCAKES NOODLES SPAGHETTI  135  998-020 008-021 008-023  RICE, brown... RICE, white... WHEAT GERM....  75 39  69 9  81  15  009-024 009-027 009-028  FRENCH BREAD.. WHITE BREAD... WHOLE WHEAT BREAD  40 414  63 451  409  263  104  138  198  278  327  010-031 010-032 010-033 010-035  BUSCUITS HAMBURGER BUN. MUFFIN ENGLISH MUFFIN  70 46 40 46  83  64  184  267  306  276  011-037 011-040  SALTINES RYE KRISP  24  6 221  183  176  133  170 85 170 85 43 64  74  67  340  5  401 17  ENTREE-MEATS 012-041 012-042 012-043 012-046 012-047 012-048  BEEF, BEEF, BEEF, PORK, PORK, BACON  013-049 013-051  CHICKEN- steame j CHICKEN, fried.  014-055 014-056  FRIED FISH.... BROILED FISH..  30% fat. 20% fat. 15% fat. lean cut all hams  Values rounded to the nearest grams/week.  100  145  Table 4.13 (Continued) 014-058 014-062 014-063 014-065  OYSTERS SARDINES SHRIMP TUNA  015-067  LIVER  017-069 017-070 017-071  FRANKFURTERS... FRESH SAUSAGES. LIVERWURST  018-073 018-075  BEANS, d r i e d . . . SOYBEANS  019-076 019-077 019-079 019-080 019-081  ALMONDS CASHEW NUTS PEANUT BUTTER.. PEANUTS PECANS  267 43 60  22  69  127  102  98  84  3b  99  40 83  15 28 120  10 85  16 147  155  184  82 218  88 446  411  393  330  385  481  23 525 316 3  34 471 329  90 427 388  193 172 112 70 341  165 160 162  150 168 194  ENTREE-VEGETABLES 020-082 020-083 020-084 020-085  POTATOE, baked. POTATOE, fried. POTATOE, mashed SWEET POTATOE..  100 100 200 180  021-089 021-091 021-092 021-095 021-098 021-099 021-101  BEANS, green... BROCOLLI....... CABBAGE LETTUCE PEAS PEPPERS SPINACH  65 63 65 180 170  022-103 022-104 022-105 022-106 022-111  BEETS CARROTS, cooked CARROTS, raw... CORN TOMATOE  80 76 50 518 -  158 146 70 109 448  -.023-113 023-115  CUCUMBER ONIONS  275 8  210 119  137 238  201  97;  024-116  SUCCOTASH  149  215  127  50  025-117 025-118 025-119  OLIVES PICKLES, sweet. PICKLES, sour..  380  475  475  507  21  51  85  253 363 180 178 572 374 182  216  20 34  ENTREE-FATS 026-121 026-124  LARD SOYBEAN OIL  027-125  BUTTER  028-127 028-128 •  CHEESE SAUCE... GRAVY  5 125 72  22  186  Table 4. 13 (Continued) 029-132 029-133  MAYONNAISE.-..'...:. SALAD DRESSING....  105 90  21 70  79  76  42  BEVERAGES-DAIRY 030-136 030- 137  WHOLE MILK SKIM MILK.  732  632  652  702  711 2  031- 140 031-141  TABLE CREAM..  195 8  93  81  150  191  86  151  267  219  10 212  158  34  53  WHIPPED CREAM BEVERAGES-FRUIT  032032032032-  143 144 145 148  033-150  APPLE JUICE GRAPEFRUIT JUICE.. LEMON JUICE ORANGE JUICE  120 5 360  174  TOMATOE JUICE BEVERAGES-MISC.  034- 151  COLA-TYPE  035- 153 035- 154  COFFEE TEA BEER DISTILLED SPIRITS. DRY WINES  036- 155 036-156 036-158  339  328  323  328  352  2800 1600  2771 1572  2788 1581  2799 1581  2814 1581  3240 43 400  3229 34 324  3251 68 321  3264 72 324  3304 71 302  198 200  145 99  184 59 74  233  248  106  121  219 123  248 86  306 88  237  400  496  554  585  270  185 15  135 85  145 102  6 218  185  107  116  147  104  300 185  485 284  608 288  661 225  741 165  SOUPS 037-160 1 CREAMED SOUPS' 037-161 PEA SOUPS 037-163 MEAT + VEGIE SOUPS DESSERTS-CEREALS 038-166 038-168 038- 170  COFFEE CAKE FRUITCAKE ICED CAKES...  75  039- 171 039- 172  FRUIT PIES PUMPKIN PIES  160 150  040- 173 040- 174  COOKIES FRUIT COOKIES.....  60  041- 175 041-177  CAKE DOUGHNUTS.... DANISH PASTRY  60 38  60  DESSERTS-DAIRY ICE-CREAM  042-178 042- 179  SHERBERT  043- 181  PUDDINGS DESSERTS-FRUIT  044-183 044-184  APPLE APPLESAUCE  187 Table 4.13 (Continued) 044-185 044-186 044-187 044-188 044-192  BANANA CANTALOUPE GRAPEFRUIT ORANGE PINEAPPLE  045-193  DRIED FRUIT  .,.  100 100 200 150  44  40  83  135 159 37  217 18  242  189  DESSERTS-SWEETS 047-195 047-197  HONEY SUGAR  105  33 109  19 125  51 110  123 25  048-198 048-199  JAMS SYRUP  100 40  99 129  49 172  18 223  23 179  049-201 049-202  CHOCOLATE CANDY... MARSHMALLOW  44  30  38  106  MISCELLANEOUS 051-205  POT PIES  227  199  195  155  168  063-217  SPAGHETTI + MEAT..  330  179  208  213  240  067-221  COCOA MIX  7  Total Grams/Week..  18889  19461  19288  19117  18933  Total # of Items..  83  70  68  61  55  # of Initial Items  83  53  48  43  36  # of New Items....  17  20  18  19  # of Items Dropped  13  15  22  28  188  Table 4.14  Nutrient  Nutrient composition of the SID-2, and the revised diets developed using quadratic objective functions (Eqns. 4.26-4.28).  Nutrient Minimum (/week)  PROTEIN (igm) CHO-T (gm) T-FAT(gm) KCAL CHO-F(gm) SFA(gm) SUCR(gm) PUFA(gm) VIT-A(iu) VIT-E(mg) VIT-C(mg) THIA(mg) RIBO(mg) NIAC(mg) VIT-B6(ug) FOLATE(mg) POTAS(mg) CAL(mg) PHOSP(mg) I RON (trig) MAG(mg) CA/P P/S  392.00 2735.4 44.209 18899. 79.576 .0 .0 44.209 35000. 63.000 210.00 9.9470 11.936 131.30 14000. 1400.0 9800.0 5600.0 5600.0 70.000 2205.0 .8 1.0  Constraints Maximum (/week)  SID-2 Initial (/week)  SID-2 SID-2 Eqn.4.26 Eqri.4.27 (/week) (/week)  SID-2 Eqn.4.28 (/week)  558.60 4476.1 663.13 20889. 159.15 221.04 746.02 663.13 .140E+06 700.00 3500.0 19.894 23.873 262.60 28000. 2800.0 19600. 11200. 11200. 140.00 4410.0 1.2 2.0  713.82 1872.4 981.39 20195. 32.250 367.46 549.60 543.42 69952. 75.273 886.80 10.292 13.803 178.25 12942. 1460.9 22490. 6137.3 11258. 120.08 2386.5 0.54514 1.4789  558.60 558.60 2735.4 2735.4 663113 658.14 19832. 19810. 79.576 79.576 221.04 221.04 746.02 746.02 360.24 364.07 0.107E+06 95816. 91.075 90.631 1248.2 1231.6 11.496 11.190 12.289 13.038 131.30 131.30 14000. 14000. 1400.0 1400.0 19600. 19600. 8960.0 8960.0 11200. 11200. 135.86 130.87 2717.7 2751.3 0.80000 0.80000 1.6297 1.6471  558.60 2735.4 663:13 19881. 79.576 210.90 738.58 362.29 94650. 94.316 1375.4 11.448 13.194 131.30 14000. 1400.0 19600. 8960.0 11200. 140.00 2562.4 0.80000 1.717  J  189  As noted, the second term in the objective function is expected to reduce the deviation of attribute groups from their i n i t i a l levels.  This  deviation should be roughly proportional to the penalty coefficient applied - - in this instance 1 or 0.  In order to illustrate this phenomenon, the  average absolute deviations of item clusters and of attribute groups in each hierarchical level have been determined for the revised diets of Table 4.13 and compared in Table 4.15. of an attribute group is penalized - -  As expected, where the deviation = 1 - - average absolute deviation  is consistently lower relative to the condition where the deviation of that levels groups is not penalized - - P^^ = 0.  For example, the  average absolute deviation of groups in each hierarchical level, k=3 to k=8, is lower when the deviations jof attribute groups are penalized by the objective function (Eqn. 4.28) as compared with the case when the deviations are not penalized (Eqns. 4.8, 4.26, 4.27).  Where the objective  functions have equally penalized the deviations of groups on any particular level, the outcome is roughly dependent on the amount of emphasis on other group deviations.  Thus, as emphasis shifts towards attribute  groups in higher numbered hierarchical levels, there is a coincident alteration in deviation of item clusters and attribute groups at lower hierchical  levels.  190  Table 4.15  Average absolute deviation of item clusters and of attribute groups in each hierarchical level from i n i t i a l levels, and the penalty coefficients assigned for each hierarchical level.  Hierarchical # of Items Levels of Groups/ L e v e i  Average Absolute Deviation From SID-2* and values of w- and P . . . Eqn.4J8 Eqn.4.26 (gm)(w /P ) (gm)(w./Pj ) i  127 50 39 34 23 20 12 4 1  75 167 190 207 297 308 490 763 572  1 0 0 0 0 0 0 0 0  k=l k=2 k=3 k=4 k=5 k=6 k=7 k=8 values rounded to nearest gram  jk  k  86 136 163 176 252 253 402 571 399  1 1 0 0 0 0 0 0 0  Eqn.4.27 (gm)(w /Pj ) i  95 130 136 149 214 210 340 394 288  k  1 1 1 o: 0 0 0 0 0  Eqn.4.28 (gm)(w./P ) jk  106 149 124 132 176 155 235 244 44  The different treatment of deviation by these objective functions (Eqns. 4.26 - 4.28) shows in the character of the solution profiles.  With  respect to the number of items contained in the SID-2 revised diets, as more attribute groups are considered, the total number of items in the revised diet and the number of items included from the i n i t i a l diet has consistently decreased (foot of Table 4.13).  Although this phenomenon need  not occur in every instance, i t is not an unreasonable outcome.  In order  to meet nutrient constraints the objective seeks to change the limited set of most efficient nutrient sources.  As indicated previously, the  quadratic penalty function restricts the extent to which any particular item can act as an efficient nutrient source.  Consequently, the limited  set changed may correspond to a larger number of items than would be predicted from comparing the nutrient concentration of items alone.  As  the deviation in attribute groups in penalized, the deviation of item clusters which act as efficient nutrient sources is amplified.  The result  191  is in this instance, that as emphasis on attribute group deviation has increased, progressively more initially-consumed items have declined to zero and thus dropped from the revised diets; that i s , thirteen initially-consumed items have deviated to zero when no penalty was applied on attribute group deviations (Eqn. 4.8), whereas 15, 22, and 28 items have deviated to zero as the penalty was increased (Eqns. 4.26-4.28).  In  these revised diets the number of new items which have entered the solution varies only slightly, between 17 and 20. Another outcome, of particular interest, is the impact of the attribute structure on concurrent deviation of item clusters.  That i s ,  does the attribute structure influence compatibility relationships between item clusters by causing these items to vary in inverse proportions, or to substitute?  It is apparent from previous results that when attribute  group deviations are penalized, item clusters within these attribute groups tend to substitute.  That this occurs i s , ipso facto, a condition of changes  in the item profile of the revised diet and the simultaneous reduction in the attribute group deviation in these revised diets. It is obvious that these shifts have not occured item for item, since there has been a general reduction from the i n i t i a l condition in the total number of items in the revised diet.  However, i f the revised diet from  the objective function (Eqn. 4.8), which does not penalize deviation of attribute groups, is compared with the objective function (Eqn. 4.26) that penalizes deviation of all attribute groups on hierarchical level "1", one slight substitution effect can be illustrated (Table 4.16); that i s , 41 attribute groups represented in SID-2 hierarchical level "1" are more often represented, 38 to 35, in the revised diet when those attribute groups are penalized (Eqn. 4.26) as opposed to when the attribute groups  192  are not penalized (Eqn. 4.8).  Table 4.16  Effect of penalty assignment on the number of attribute groups containing consumed item clusters.  Hierarchy Levels  # of Groups/ Level  k=l  50  # of Groups/Level Containing Consumed Item Clusters SID-2 Initial  SID-1 . Eqn. 4.8  SID-2 Eqn, 4.26  41  35  38  Although the addition of the second term to the objective function has altered the solution in expected ways, the specific details of these changes are not easily monitored.!  If such an approach for modelling concurrent  changes between items is r e a l i s t i c , its success depends on appropriately weighting the relative acceptability of each attribute group's deviation from i n i t i a l levels.  4.3.2.7  Second Term:  Further Modifications of the Algorithm  Apart from affecting the solutions by altering the penalty weighting for attribute group deviation, i t may be useful to consider:  selectively  penalizing positive and negative deviation of attribute groups; and transforming the objective's second term into a constraint which restricts attribute group deviation. (i)  As presently formulated, the objective function's second term  (Eqn. 4.19) does not provide an opportunity to specify different  penalties  for positive, as opposed to negative, deviation of attribute groups. A quadratic second term can be formulated which minimizes the total, weighted-positive and weighted-negative deviation in the amounts of  193  attribute groups from i n i t i a l consumption levels, as follows: K (4.29)  J  minimize ^  (4.30)  subject to  zj  £ k  (pj z j > p" k  k  - zT = z  (4.31)  ^k-'j'k*  z~.^  (x - - x?)  k  i t G  k  n  jk  (j = 1, 2, . . . , J ) k  (k = 1  5  2, . . . , K)  0  Where the terms are as previously defined (p. 181 ). (ii)  To ensure that changes in attribute groups fall within speci-  fied limits, a constraint can be appropriately included in the algorithm (Eqns. 3.2-3.5), or substituted for the second term of the obj>ective function (Eqn. 4.19).  This constraint provides upper and lower bounds on  the deviation in the amounts of attribute groups, as follows: (4.32)  b  j k  *  z  (x. - x?) * c  l f G jk  j k  (j = 1, 2,  J ) k  (k = 1, 2, . . . , K)  Where the newly-defined terms are: bj  k  and C j are respectively, the upper and lower bounds on the k  deviation of attribute group j in the k ^ hierarchy level. If b.. = 0 then the revised consumption of the attribute JK group j cannot be more than the original consumption of group j .  If C j = 0 then the revised consumption of k  attribute group j cannot be less than the original consumption (iii)  of group j .  The second term of the objective function (Eqn. 4.19) only  considers compatibility relationships among item clusters.  Complimentarity  between items has been largely neglected in the algorithm's design,  1-9.4  with the exception of choosing items with low complementary characteristics where possible.  For example, bread has been chosen for the food item l i s t  rather than its ingredients - - flour, water, yeast.  Therefore, i t may be  worthwhile to add a term to the objective function (Eqn. 3.2) which considers complementary relationships by minimizing the total, weighted squared deviation in the ratio of the amounts of specified item clusters from i n i t i a l consumption levels.  The formulation is: x  (4.33)  minimize  E (h,l)fis  f., hl  x  (n )— ^1  2  h ) { x? ;  (h,l = any specified pair of item clusters)  Where the newly-defined terms are: x^, x° are the amounts, in grams, of specified item clusters h and 1, respectively, contained in the client's i n i t i a l  diet.  x^, x-| are the amounts, in grams, of specified item clusters hnand 1, respectively, contained in the revised diet. s^i  is the set of item cluster pairs (h.l) for which the x° deviation from the i n i t i a l ratio _h is being penalized. o l x  f^l  is the penalty associated with the deviation of item cluster pairs (h,l) from the i n i t i a l ratio _h .  *° (iv)  The above term (Eqn. 4.33) can be transformed to a constraint  which restricts the deviation in the ratios of specific item clusters.  The  expression is: x. (4.34)  3. , > m  x° -  x  l  * \  i  x°  (h,l = any specified pair of item clusters)  Where the newly-defined terms are as follows: 3^1 and  are respectively, the upper and lower bounds on how much  195  the ratio of item clusters h and 1 will be allowed to x  °  change from the i n i t i a l ratio _h . o l x  196  4.4  Summary and Comments on Testing the Algorithm ?  As indicated (p. 1:28), testing the diet planning phase has been limited to a descriptive evaluation of some of the algorithm's characteristics - specifically, the premises and assumptions which define the acceptability of altering a diet, and the revised diets developed when these assumptions are modified. (i)  To reiterate, these assumptions include:  A concept of attributes which form the basic practical and conceptual entities that the individual is conditioned to respond to and uses to identify a diet's character.  These attributes are  used to categorize foods into a matrix of attribute groups - - an attribute map. (ii)  The acceptability or perceived extent of change is considered dependent on the significance of maintaining each element ofuthe map at its i n i t i a l amount with respect to i t s e l f and to other elements, and on how this significance is altered when the relative or absolute proportions of the map elements change.  Within the context of the premises established for the algorithm's desi (p. 136), modifications in the assumptions about acceptability of dietary change can be proposed, and incorporated in the design of the model's objective function or constraints.  Here, only some features of the  objective function were explicitly modified - - specifically, the attribute elements defined and the penalty coefficients defined-items.  for deviation of these  The results illustrate that altering these features pro-  duces marked variations in the revised diets with respect to the observed parameters - - that i s , deviations in the amounts of items and changes in the number of items.  It is expected that the other modifications in the  algorithm's objective function and constraints, described above, should also have an impact on the revised diets developed.  197  The descriptive evaluation was undertaken to explore the conceptual and technical feasibility of using this type of mathematical model for approximating an individual's nutrient-constrained food choice, given his or her diet inventory.  Thus, the evaluation provides insight into the  ability of the model to adapt to a variety of output demands. Although this explorative evaluation was not intended to verify the diet-planning model's operation, the implicit question throughout is which, i f any, of these approaches provides the closest approximation to the nutrient-allowed food choice an individual would make given that the individual could understand, accept, and utilize nutrient information.  It  would seem unlikely that all variations of the revised diets developed from a given dietary inventory using different objectives would be equivalent in acceptability to the client.  However, any judgement about the  most appropriate formulation must be purely subjective at present. manipulation of the algorithm and observation of the solution's  Hence,  sensitivity  to these modifications, do not immediately provide information on the perceived acceptability of the dietary revision. It may appear that the most appropriate model can be easily determined by selecting the most limiting solution - - that i s , the one that deviates least from the original diet.  However, as the results demonstrate, this  is not a mathematical concept which can be objectively appraised without choosing one of the many possible definitions of "minimum deviation". Given any such definition the algorithm provides a mathematically-optimal solution.  Therefore, evaluations to determine which algorithm provides  the least-changed solution, can only serve to corroborate or refute whether the algorithm and test procedure use the same optimality criterion. In order to determine the validity of any particular algorithm for  198  approximating the client's nutrient-constrained food choice, an empirical basis is required.  The mathematically-optimal deviation of the revised  diet from the i n i t i a l diet must be equated to a behavioral or perceived optimum; or at least an explicitly-stated tradition or professional guidelineron which to justify revisions.  In this instance, as indicated in  pagel36no appropriate tradition or guidelines exists for model development or testing.  Appropriate empirical methods for evaluating the efficacy of  the diet-planning model would include, f i r s t , comparison of the system output with client-generated or nutritionist-generated solutions, including observation of the client's behavioral response.  Second, a less rigorous  evaluation of the algorithm canubecundertaken by considering the response of a client - - or a nutritionist acting as the client's agent - - to the system outputs. Such definitive evaluation was judged premature, pending further explorative evaluation and system development, because of two major difficulties. First, establishing a test situation where the client or nutritionist could effectively use the same nutritional objectives as the computational method was not presently possible, even i f the nutritional objectives could be agreed upon.  This equivalent ability is important in order that solutions  generated from different sources be comparable; or so that comments by client's and nutritionist's on computationally-produced diets be tempered with recognition of available nutrient-restricted options.  Computer-  assisted diet planning has been specifically employed in this project to ensure data handling capabilities unavailable to either clients or nutritionists without these means. Second, even i f the difficulty of providing a framework for comparison of computational and human methods can be overcome, the computationally-  199  generated solutions would be difficult to appraise constructively, either by comparison with human-generated solutions or by consideration of the responses of client's or nutritionist's.  Although such evaluation could confirm  or deny the acceptability of solutions, the relative acceptability of different computational solutions cannot be determined, except for exact or near-exact solution replicates, or certain acceptance of computational recommendations. While difficulties are as relevant now as they were previously, the inquiry and observation undertaken have provided a basis on which to explore more usefully the validity of the algorithm's assumptions.  The observations  illustrate that the algorithm can be modified to accommodate a variety of different solution requirements.  Although iterating the options available  may be a lengthy process, these options can be systematically evaluated to establish which, i f any, of the algorithms approaches the nutrientconstrained food choices of individuals.  Also, such testing may provide  for further understanding of the factors which influence the mathematical formulation in producing a solution judged to be useful.  Perhaps of more  importance, i t may provide an opportunity to explore measurably the tradeoffs between the nutritional goals of presenting accurate nutrient information, and the educational objectives of presenting simple, straightforward information.  200  CHAPTER 5 SUMMARY, RECOMMENDATIONS, AND CONCLUSIONS  5.1  Summary  5.1.1  The Thesi s Goal  The overall thesis goal was to develop a system to provide information on dietary practices for adults who are motivated to apply nutritional principles to their eating habits, and who have sufficient resources to make use of the information required to define healthful dietary practices for them.  This development was undertaken bearing in mind two primary  problems, namely:  the developing of nutritional guidelines suitable for  health promotion, and the communication of these guidelines to individuals.  5.1.2  First Objective of the Thesis  The prototypical computerized system developed has two major functions: f i r s t , to assess the dietary intake of individuals (diet-assessment), and second, to recommend changes in food intake for those individuals with nutrient intake which do not meet specified limits (diet-planning).  5.1.2.1  Diet-Assessment Function  Standard dietary assessment procedures of data collection, analysis, and evaluation, were adopted for the system's design.  In the data  collection phase, the i n i t i a l diet of an individual is estimated by using an intake questionnaire.  This information is then translated into nutrient  data, in the data analysis phase, and evaluated in the evaluation phase by  201  comparing calculated nutrient-intake with nutrient limits which represent an "acceptable" range of intake for the individual.  The l i s t of food consumed,  and the results of data analysis and evaluation are displayed in the computer output. The nutritional guidelines used for evaluating diets are largely based on the dietary standards of various countries and international agencies, and on dietary goals proposed by recognized scientific agencies.  These  guidelines are adopted for the system's design because they provide a technically accurate rationale for healthful food selection practice. However, it is recognized that this complex information on nutrients cannot easily be used to recommend dietary practices for individuals, and therefore would require translation into a statement of food and meals.  5.1.2.2  Diet-Planning Function  A revised diet is developed by a constrained-optimization model which determines the combination of foods, subject to nutrient constraints, that minimizes the total squared deviation of food items and item groups from their original amounts as specified on the client questionnaire.  The  revised diet is provided on the computer output. This model has been developed to provide a recommended dietifor/motivated individuals which: (i) (ii)  accurately reflects nutritional guidelines, and facilitates adoption of recommendations by providing a selfexplanatory statement of foods to consume and by limiting the changes from present or desired food patterns.  Presumably, this would be similar to the diet which the individual would choose i f she or he understood, accepted, and used nutritional knowledge  202  - - that i s , the individual's nutrient-constrained food choice.  The use  of mathematical modelling provides an effective means of collating the vast amount of data required to develop dietary recommendations which are both nutritionally accurate, straightforward, and hopefully, acceptable to the client.  5.1.3  Second Objective of the Thesis  Following formulation of the diet-planning algorithm, the major task was to test this algorithm in order to explore the conceptual details of this diet-planning approach.  Testing was restricted to a descriptive  evaluation of some of the algorithm's characteristics - - specifically, the assumptions which define the acceptability of altering a diet and the revised diets developed when these assumptions are modified. The assumptions defined included:  f i r s t , a concept of attribute  elements which form the basic practical and conceptual entities that the individual is conditioned to respond to and uses to identify a diet's character; second, the acceptability or perceived extent of change is considered to be dependent on both the significance of maintaining each element at its original amount with respect to itself and to other elements, and on the extent to which this significance is altered when the relative or absolute proportions of the map elements change. A variety of modifications in the assumptions about acceptability of dietary change can be incorporated in the algorithm's design.  In this  instance, only certain features of the objective function - - specifically, the attribute elements defined and the penalty coefficients deviation of these attribute elements - - were explicitly  weighting  modified.  The  results illustrate that altering these features produced marked variations  203  in the revised diets with respect to the observed parameters, that i s , deviation in the amounts of items and changes in the number of items. Other modifications in the algorithm's objective function and constraints would also be expected to have an impact on the revised diets developed. Although the explorative evaluation was not intended to validate whether the diet planning model can provide acceptable revisions of diets for clients, i t does provide a foundation for more systematic evaluation of the validity of different approaches.  As noted the algorithm can be modified  to accommodate a variety of different solution requirements.  These options  can by systematically tested, by considering client or nutritionistgenerated solutions or their responses, to establish which, i f any, of the possible algorithms approximate the nutrient-constrained food choice of individuals.  5.2  Recommendations Since iterating the available approaches may be a lengthy process, some  recommendations which may facilitate exploration and development of a "commercially" viable model are now given.  5.2.1  First Recommendation  Emphasis should be given to reducing the absolute number of changes in revised diets rather than just minimizing deviation of item amounts.  It  became apparent while analyzing the results, that even though the revised diets.deyiate minimally and in fact could represent the real adjustments that individuals would make, these revisions were difficult to appreciate because so many changes had occured.  In a nutrition education setting,  realistic solutions may need to maintain many items at i n i t i a l levels.  204  Some alterations in the algorithm that may reduce changes in the revised diet to a manageable level include: (i)  The use of a linear rather than a quadratic formulation.  Although  the quadratic penalty function may seem more realistic than the linear function, the quadratic function tends to spread deviation over all items.  A linear function may provide fewer changes and  thereby a more easily understood solution, (ii)  A formulation which restricts the items in the solution set to the originally consumed items and to a limited number of alternatives for guaranteeing solution feasibility may simplify outcomes; as might one which concentrates changes to specified types of items, such as low nutrient foods.  Although the total deviation  of item quantities may be greater using these means, the alterations may be easier for the client to monitor and appreciate, (iii)  Constraints on the deviation of items could be included to control the solutions, and would be particularly useful for a linear format. Some items may usefully be retained at i n i t i a l levels.  (iv)  In place of an integer formulation which would be difficult to develop, a routine to average item quantities in the revised diets to their nearest serving or half-serving should be considered. The loss of accuracy may be considerably offset  by the increase in  u t i l i t y from removing inordinately small increments and decrements, and from 'simplifying the' serving sizes reported;.) (v)  Solutions based on a more abbreviated set of item classes or food groups may be easier to comprehend.  Aggregating foods into groups  and averaging over nutrient values would necessarily reduce the nutritional accuracy of the solutions.  However, the increased  205  fulness of the information for illustrating dietary change may substantially offset this disadvantage.  5.2.2  Second Recommendation  Only information on the client's i n i t i a l diet inventory, or alternatively a food plan identified as desirable was used to define modifications of that inventory.  This was done to limit the amount of information required from  the client on the questionnaire.  However, the client (or the client's  counsellor) could usefully provide information on the relative acceptability of deviating from i n i t i a l levels in order to more acceptably direct revisions of the diet.  For example, the client could indicate whether the relative  acceptability of either an increase or decrease in the amount presently consumed (or not consumed) is:  high, indeterminate, or low.  These responses  can then be incorporated into the algorithm's penalty coefficients constraints to adjust the solution outcomes. for the penalty coefficients  and/or  The initially-assigned values  and/or constraints will be arbitrary until the  relationship of individual response to coefficient assignment is appropriately indexed.  5.3  Conclusions The use of mathematical modelling and computer technology has provided  an effective means of collating the vast amount of data required to develop cogent dietary recommendations which are nutritionally accurate, straightforward, and potentially acceptable to the individual.  The concept of  minimizing deviation from an i n i t i a l dietary inventory seems a feasible approach to establish the link between these mathematical programming  206  techniques and the objectives of diet planning - - that i s , to approximate the individual's nutrient-constrained food choice.  207  LIST OF REFERENCES  Action B.C. Nutrition Evaluation Program.  Vancouver, B . C . , 1976.  Adams, C.F. Nutritive Value of American Foods in Common Units. United States Department of Agriculture, Agriculture Handbook No. 456, 1975. Adelson, S.F. Some problems in collecting dietary data from individuals. J . Amer. Dietet. Assoc. 36: 453, 1960. Ahlstrom, A. and Rasanen, L. Review of food grouping systems in nutrition education. J . Nutr. Educ. 5: 13, 1973. American Dietetic Association. Position paper on nutrition education for the public. J . Amer. Dietet. Assoc. 62: 429, 1973. American Heart Association.  Diet and coronary heart disease.  1973.  Andrews, J . T . , Moore, A.N. and T u t h i l l , B.H. Electronic data processing in intra-departmental food cost accounting. J . Amer. Dietet. Assoc. 51: 332, 1967. Anon. Anon.  Nutrition education for youth.  J . Home Econ. 64:  34, 1972.  Food labeling, Federal Register 38: 2124, 1973; Food labeling, Federal Register 38: 6950, 1973; Food nutrition labeling, Federal Register 38: 6959, 1973; Labeling of foods in relation to fat, fatty acid, and cholesterol contant, Federal Register 38: 6963, 1973.  Armstrong, R.D. and Balintfy, J . L . A chance-constrained multiple-choice programming algorithm. Operations Research 23: 494, 1975. Armstrong, R.D.'and Sinha, P. Application of quasi-integer programming to the solution of menu planning problems with variable portion size. Management Science 21: 474, 1974. Arroyave, G. Standards for the diagnosis of vitamin deficiency in man. In: Proceedings of the Special Session, Ninth Meeting of PAHO Advisory Committee on Medical Research, 16 June, 1970. Pan American Health Organization, Washington, D . C . , 1971, p. 88. Asenjo, C.F. Variation in nutritive value of foods. 11: 368, 1962. Astrand, P-0. Nutrition and physical performance. 16: 59, 1973.  Amer. J . Clin. Nutr. World Rev. Nutr. Diet.  Astrand, P-0. and Rodahl, K. Textbook of Work Physiology. Company, Toronto, 1970.  McGraw-Hill Book  Atwater, W.0. Methods and Results of Investigations on the Chemistry and Economy of Foods. United States Department of Agriculture, Bulletin No. 21, 1895.  208  Baker, M.J. Influence of nutrition education on fourth and fifth graders. J . Nutr. Educ. 4: 55, 1972. Balintfy, J . L . Mathematical Programming of Menu Planning in Hospitals. Tulane University, School of Business, New Orleans, 1963. Balintfy, J . L . Menu planning by computer. Communications of the ACM 7: 255,1964. Balintfy, J . L . Linear programming models for menu planning. In_: Hospital Industrial Engineering. Edited by H.E. Smalley and J.R. Freeman. Reinhold, New York, 1966, p. 402. Balintfy, J . L . System/360 Computer Assisted Menu Planning. IBM Contributed Program Library, 360D-15.2.013, IBM Corporation, PID, Hawthorne, New York, 1969. Balintfy, J . L . Computer-assisted menu planning and food service. In: Handbook of Biomedical Information Systems. Edited by W. Spector Encyclopedia Britannica, Inc., Chicago, 1971, p. 85. Balintfy, J . L . Mathematical modeling and human nutrition. 581 , 1973a.  Science 181:  Balintfy, J . L . Nonlinear Programming and the Food Price Index. Technical Report No. 11, School of Business Administration, University of Massachusetts, Amherst, Massachusets, 1973b. Balintfy, J . L . Large-scale programming properties of menu planning and schedu l i n g . / In: Optimization Methods for Resource Allocation. Edited by R.W. Cottle and J . Krarup. The English Universities Press L t d . , London, 1974a, p. 81. Balintfy, J . L . Mathematical models for menus. 1 , 1974b.  Naval Research Reviews 27:  Balintfy, J . L . A mathematical programming system for food management applications. Interfaces 6: 13, 1975. Balintfy, J . L . Applications of mathematical programming to optimize human diets. Istituto Nationale di Matematica. Symposia Mathematica. XIX: 313, 1976a. Balintfy, J . L . A scientific and democratic approach. Quarterly 17: 33, May 1976b.  The Cornell H.R.A.  Balintfy, J . L . , Duffy, W.J. and Sinha, P. Modelling food preferences over time. Operations Research 22: 711, 1974. Balintfy, J . L . and Nebel, E.C. Experiments with computer-assisted menu planning. Hospitals 40: 88, June 16, 1966. Balogh, M., Kahn, H.A. and Medalie, J . H . Random repeat 24-hour recalls. Amer. J . Clin. Nutr. 24: 304, 1971.  2Q9  Balogh, M . , Medalie, J . H . , Smith,'H. and Groen, J . J , The development of a dietary questionnaire for an ischaemic heart disease survey. Israel J . Med. Sci. 4: .195; 1968. Beal, V.A. The nutritional history in longitudinal research. Dietet. Assoc. 51: 426, 1967  J . Amer.  Beaton, G.H. The use of nutritional requirements and allowances. J^n: Western Hemisphere Nutrition Congress III. Edited by P.L. White. Futura Publishing Co., New York, 1972, p. 356. Beaton, G.H. Evaluation of nutrient intake: New statistical Arch. Latino, de Nutricon 25: 31, 1975.  approaches.  Beckman, M.J. On the determination of an adequate diet at minimum cost. Trabajos de Estadisica 11: 139 1960. (Misprint corrected in Smith, V . E . Electronic Computation of Human Diets. MSU Business Studies, Bureau of Business and Economic Research, Graduate School of Business Administration, Michigan State University, East Lansing, Michigan, 1963.) Bell, C.G. and Lamb, M.W. Nutrition education and dietary behavior of fifth graders. J . Nutr. Educ. 5: 196, 1973. Berg, A. The Nutrition Factor. 1973.  The Brookings Institution, Washington, D . C . ,  Berlo, D.K. The Process of Communication. 1960.  Rinehart and Winston, New York,  Bogert, L . T . , Briggs, G.M. and Calloway, D.H. Nutrition and Physical Fitness, 9th edition. W.B. Saunders Company, Toronto, 1973. Bowman, J . L . and Brennan. Computer-assisted menu planning provides control of food service. Hospitals 43: 107, August 16, 1969. Boysen, S.C. and Ahrens, R.A. Nutrition education and lunch surveys with second graders. J . Nutr. Educ. 4: 172, 1972. Bransby, E . R . , Daubney, C.G. and King, J . Comparison of nutrient values of individual diets found by calculation from food tables and by chemical analysis. Brit. J . Nutr. 2: 232, 1948. Bransby, E.R., Daubney, C.G. and King, J . Comparison of results obtained by different methods of individual dietary surveys. Brit. J . Nutr. 2: 89, 1948. Bransby, E . R . , Berry, W.I.C. and Taylor, D.M. Study of the Vitamin-D intakes of infants in 1960. Brit Med. J . i : 1661, 1964. Briggs, G.M. The need for nutrition education.  J . Nutr. Educ. 1:  Brisbane, H.N. Computing menu nutrients by data processing. Dietet. Assoc. 44: 453, 1964.  7, 1969.  J . Amer.  210  Browe, J . H . , Morlley, D.M., McCarthy, M.C. and Gofstein, R.M. Diet and heart disease study in the cardiovascular health center. II. Construction of a food composition table and its use. J . Amer. Dietet. Assoc. 48: 101, 1966. Brown, E.L. College students look at the basis for their food habits. J . Home Econ. 59: 784, 1967. Brown, J.A.C. An experiment in demand analysis: The computation of a diet problem on the Manchester computer. J_n_: Conference on Linear Programming, May 1954. Edited by Anonymous. Ferranti, L t d . , London, 1954, p. 41. Burke, B.S. The dietary history as a tool in research. Assoc. 23: 1041, 1947.  J . Amer. Dietet.  Burke, B.S., Beal, V . A . , Kirkwood, S.B. and Stuart, H.G. Nutrition studies during pregnancy. Amer. J . Qbstet. Gynec. 46: 38, 1943. Burkitt, D.P. and Painter, N.S. Assoc. 299: 1068, 1974.  Dietary fiber and disease.  J . Amer. Med.  Buzina, R., Ferber, E . , Keys, A . , Brodarec, A . , Agneletto, B. and Horvat, A. Diets of rural families and heads of families in two regions of Yugoslavia. Voeding 25: 629,1964. Buzina, R., Keys, A . , Brodarec, A. and Fidanza, F. Dietary surveys in rural Yugoslavia. III. Comparison of three methods. Voeding 27: 99, 1966. Calavan, M.M. A model of food selection in northern Thailand. Nutr. 5: 63, 1976.  Ecol. Food  Canada. Nutrition Canada, Department of National Health and Welfare. Nutrition: A National Priority. Information Canada, Ottawa, 1973. Canada. Bureau of Nutritional Sciences, Department of National Health and Welfare. Dietary Standards for Canada. Information Canada, Ottawa, 1975. Canada. Department of National Health and Welfare. Report of the Committee on Diet and Cardiovascular Disease. Ottawa, 1976a. Canada. Statistics Canada. Ottawa, 1976b.  Vital Statistics 1974.  Catalogue 84-206,  Canada. Bureau of Nutritional Sciences, Department of National Health and Welfare. Food Consumption Pattern Report. Information Canada, Ottawa, 1977. Canada. Bureau of Nutritional Sciences, Department of National Health and Welfare. Canada's Food Guide Handbook. Ottawa, 1977.  211  Cantoni, M . , Paffenbarger, R.S. and Krueger, D.E. Methods of dietary assessment in current epidemiological studies of cardiovascular diseases. Amer. J . Pub!. HIth. 51: 70,1961. Carmel, J . The prediction of diets of high energy and protein value by linear programming. Ecol. Food Nutr. 5: 161, 1976. Caso, E.K. Calculation of diabetic diets. 575, 1950.  J . Amer. Dietet. Assoc. 26:  Chalmers, F.W., Clayton, M.M., Gates, L . O . , Tucker, R . E . , Wertz, A.W., Young, CM. and Foster, W.D. The dietary record - How many and which days? J . Amer. Dietet. Assoc. 28: 711, 1952. Chamber!in, J.G. and Stickney, R.E. Improvement of children's diets in developing countries: an analytic approach to evaluation of alternative strategies. Nutr. Rept. Intern. 7: 71, 1973. Chandler, C.A. and Perloff, B.P. Updating food composition values - A progress report. U.S. Department of Agriculture, Nutrition Program News, September - December, 1975, p. 1. Chappell, G.M. Long term individual dietary studies. 323, 1955.  Brit. J . Nutr. 9:  Cheney, M.C. Food Enrichment: Nutritional Standards for Synthetic and Modified Foods. Paper presented at Current Topics in Food and Nutrition - Workshop. University of British Columbia, Vancouver, B . C . , July 27, 1976. Christakis, G. Nutritional assessment in health programs. Hlth. 63 Part II (Suppl.): 1, 1973. Church, C.F. and Church, H.N. Food Portions Commonly Used, J.B. Lippincott Company, New York, 1975.  Amer. J . Publ. 1.2th edition.  Clements, F.W. Nutrition 11 - Recommended dietary allowances. Austr. 2: 24, 1975.  Med. J .  Combs, G.F. and Wolfe, A.C. Methods used in dietary survey of civilians in Ecuador. Publ. Hlth. Rep., Wash. 75: 707, 1960. Conner, M.M. 1967.  A guiding star for modern nutrition.  J . Home Econ. 59:  Comely, P.B. There is more to health than just paying b i l l s . Public Health 64: 845,1974.  734,  Am. J .  Cornfield, J . Letter to Tore Dalenius. Nov. 29, 1951. Unpublished (See Dorfman, R., Samuelson, P. and Solow, R.M. Linear Programming and Economic Analysis. McGraw-Hill Book Co., Inc., New York, 1958; and Smith, V.E. Electronic Computation of Human Diets. MSU Business Studies, Bureau of Business and Economic Research, Graduate School of Business Administration, Michigan State University, East Lansing, Michigan, 1963.)  212  Correa, H. Measured influence of nutrition on socio-economic development. World Rev. Nutr. Diet. 20: 2,1975. Cottle, R.W. and Dantzig G.B. Complementary pivot theory of mathematical programming. Linear Algebra and Applications. 1: 103, 1968. Dantzig, G.B. Linear Programming and Extensions. Press, Princeton, New Jersey, 1963.  Princeton University  Davenport, E. Calculating the Nutritive-Value-of Diets. ARS-62-10-1. Agriculture Research Service. USDA, Washington, D . C , 1964. Davis, D.R. and Williams, R.J. Potentially useful criteria for judging nutritional adequacy. Amer. J . Clin. Nutr. 29: 710, 1976. Dawber, T . R . , Pearson, G . , Anderson, P., Mann, G.V., Kannel, W.B., Shirtleff, D. and McNamara, P. Dietary assessment in the epidemiologic study of coronary heart disease: The Framingham study. II. Reliability of measurement. Amer. J . Clin. Nutr. 11: 226, 1962. Dorfman, R., Samuel son, P. and Solow, R.M. Linear Programming and Economic Analysis. McGraw-Hill Book Co. Inc., New York, 1958. Durnin, J.V.G.A. Appetite and the relationship between expenditure and intake of calories in man. J . Physiol., Lond. 156: 294, 1961. Eagles, J . A . , Whiting, M.G. and Olson, R.E. Dietary appraisal - Problems in processing dietary data. Amer. J . Clin. Nutr. 19: 1, 1966. Eckstein, E.F. Menu planning by computer: Dietet. Assoc. 51: 529, 1967.  The random approach.  J . Amer.  Eddison, P. Nutrimetrics 07. Cybus Systems Limited, Suite 604, 350 Sparks Street, Ottawa, Canada, 1975. Emmons, L. and Hayes, M. Nutrition knowledge of mothers and children. J . Nutr. Educ. 5: 134, 1973. Evans, S.N. and Gormican, A. The computer in retrieving dietary history data. I. Designing and evaluating a computerized diabetic dietary history. J . Amer. Dietet. Assoc. 63: 397, 1973. Fidanza, F . , Fidanza-Alberti, A . , Ferro-Luzzi, G. and Proja, M. Dietary surveys in connection with the epidemiology of heart disease: Results in Italy. Voeding 25: 502, 1964. Fleetwood, M . , Moore, A . N . , Andrews, J , T . and T u t h i l l , B.H. Food production and control system at the medical center, UMC. Computer-assisted Food Management Systems. Edited by A.N. Moore and B.H. Tuthill. University of Missouri, Columbia, Missouri, 1974, p. 47. Flook, M.G. and Alfred, B.B. Cost comparison of systems for nutrient analysis - Computer vs. manual calculations. J . Amer. Dietet. Assoc. 64: 62, 1974.  213 Florencio, C.A. and Smith, V.E. Efficiency of food purchasing among working-class families. J . Amer- Dietat. Assoc. 55: 239, 1969. Florencio, C.A. and Smith, V.E. Towards more efficient expenditure for food among Columbian families. Nutr. Rept. Intern. 4: 207, 1970. Flores, M., Flores, Z. and Lara, M.Y. Estimation of family and mother's dietary intake comparing two methods. Trop. Geogr. Med. 17: 135, 1965. Food and Agricultural Organization, World Health Organization (FA0/WH0). Dietary Surveys - Their Techniques and Interpretation. FAO Nutr. Studies. Washington, D . C . , 1949. Fremes, R. and Sabry, Z. Eating Right: New food guide needed. Sun Newspaper, October 13, 1976, p. 90.  Vancouver  Gelpi, M . J . , Balintfy, J . L . , Dennis, L . C . and Findorff, I.K. Integrated nutrition and food cost control by computer. J . Amer. Dietet. Assoc. 61: 637, 1972. George, J.D. A nutrition education unit for a sixth grade. 2: 111, 1971.  J . Nutr. Educ.  Gifft, H.H., Washbon, M.B. and Harrison, G.G. Nutrition, Behavior and Change. Prentice-Hall, Inc., Englewood C l i f f s , New Jersey, 1972. Goodhart, R.S. Criteria of an adequate diet. In: Modern Nutrition in Health and Disease, 5th edition. Edited by R.S. Goodhart and M.E. Shils. Lea and Febiger, Philadelphia, 1973, p. 403. Goodhart, R.S. and Shils, M.E. editors. Modern Nutrition in Health and Disease, 5th edition. Lea and Febiger, Philadelphia, 1973. Goodloe, M.H.R., Waters, G . , Quinlan, C.B and Barrows, J . G . Studies in atherosclerosis. 4. Electronic computers in the processing of dietary data. Amer. J.. Clin. Nutr. 13: 304, 1963. Groover, M.E., Boone, L . , Houk, P.C. and Wolf, S. Problems in the quantitation of dietary surveys. J . Amer. Med. Assoc. 201: 8, 1967. Gue, R.L. and Liggett, J . C . Mathematical programming models for hospital menu planning. J . Industrial Engineering 17: 395, 1966. Guggenheim, K., IIan, J . , Vago, T. and Mundel, G. Nutritional status and food consumption of pregnant women in a development area of Israel. Brit. J . Nutr. 14: 347, 1960. Gussow, J . Counter-nutritional messages of TV ads aimed at children. J. Nutr. Educ. 4: 48, 1972. Hall, R.L. Food additives.  Nutr. Today 8:  20, 1973.  214  Hall, R.L. Is nutrition a stagnating science? Hall, R.L. Safe at the plate.  Nutr; Today.  New Scientist 65: 12:  7, 1975.  6, 1977.  Hampton, M.C., Huenemann, R . L . , Shapiro, L.R. and Mitchell, B.W. Caloric and nutrient intakes of teenagers. J . Amer. Dietet; Assoc. 50: 385, 1967. Hankin, J.H. and Huenemann, R.L. A short dietary method for epidemiologic studies. I. Developing standard methods for interpreting seven-day measured food records. J . Amer. Dietet. Assoc. 50: 487, 1967. Hankin, J . H . , Reynolds, W.E. and Margen, S. A short dietary method for epidemiologic studies. Amer. J . Clin. Nutr. 20: 935, 1967. Hansen, R.G. An index of food quality.  Nutr. Rev. 31:  1 , 1973.  Hanson, W.A. Medical computerized nutritional analysis. 21: 74, 1969.  J . Applied Nutr.  Harper, A.E. Recommended Dietary Allownaces: Are they what we think they are? J . Amer. Dietet. Assoc. 64: 151, 1974. Harper, A.E. U.S. Dietary Goals - Against.  J . Nutr. Educ. 9:  Harris, R.S. Reliability of nutrient analysis and food tables. Clin. Nutr. 11: 377, 1962.  154, 1977. Amer. J .  Hart, M.L. and Cox, A.G. A comparison of dietary analysis methods during a computer. Nutrition, Lond. 21: 146, 1967. Hartog, C. Den, Van Schaik, T . F . S . M . , Dalderup, L . M . , Drion, E.F. and Mulder, T. The diet of volunteers participating in a long term epidemiological field study on coronary heart disease at Zutphen, The Netherlands. Voeding 26: 184, 1965. Hayes, O.B., Abraham, S. and Caceres, C A . Computers in epidemiologic dietary studies. J . Amer. Dietet. Assoc. 44: 456, 1964. Hayes, O.B., Trulson, M.F. and Stare, F . J . Suggested revisions of the basic 7. J . Amer. Dietet. Assoc. 31: 1103, 1955. Head, M.K., Weeks, R.J. and Gibb, E. Major nutrients in type-a lunch. J . Amer. Dietet. Assoc. 63: 620, 1973. Hegsted, D.M. Nutrition, bone, and calcified tissue. Assoc. 50: 105, 1967. Hegsted, D.M. Dietary standards.  J . Amer. Dietet.  J . Amer. Dietet. Assoc. 66:  13, 1975.  Hertzler, A.A..and Anderson, H.L. Food guides in the United States. J . Amer: Dietet. Assoc. 64: 19, 1974.  215 Hertzler, A.A. and- Hoover, L.W. Development of food tables and use with computers. J . Amer. Dietet. Assoc. 70: 20, 1977. H i l l , M.M. Nutrition committees and nutrition education. 1: 14, 1969.  J . Nutr. Educ.  Hinton, M.A., Eppright, E . S . , Chadderton, H. and Wolins, L. Eating behavior and dietary intake of girls 12 to 14 years old. J . Amer. Dietet. Assoc. 43: 223, 1963. Hjortland, M.C., Duddleson, W.G., Porter, C. and French, A.B. Using the computer to calculate nutrients in metabolic diets. J . Amer. Dietet. Assoc. 49: 316, 1966. Hoover, L.W. Computers in dietetics: Dietet. Assoc. 68: 39, 1976.  State-of-the-art,  1976.  ;  J . Amer.  Huenemann, R . L . , French, F.E. and Bierman, J.M. Diets of pregnant women in Kauai, Hawaii. J . Amer. Dietet. Assoc. 39: 569, 1961. Huenemann, R.L. and Turner, D. Methods of dietary investigation. Dietet. Assoc. 18: 562, 1942.  J . Amer.  Interdepartmental Committee on Nutrition for National Defense (ICNND). Manual.for Nutrition Surveys, 2nd edition. ICNND, National Institute of Health,"-Bethesda, Maryland, 1963. Jansen, C , Dupont, J . and Blaker, G.G. A tool for individualized management of fat-controlled diets. J . Amer. Dietet. Assoc. 67: 28, 1975. J e l l i f f e , D.B. The Assessment of the Nutritional Status of the Community. World Health Organization, Geneva, 1966. J e l l i f f e , D.B. 1 , 1973.  Nutrition in early childhood.  World Rev. Nutr. Diet. 16:  J e l l i f f e , D.B. and J e l l i f f e , E . F . P . , editors. Symposium: human milk. Amer. J . Clin. Nutr. 24: 968, 1971.  The uniqueness of  Johnson, N . E . , Nitzke, S. and VandeBerg, D.L. A reporting system for nutrient.adequacy. Home Economics Research J . 2: 211, 1974. Kansra, A.C. Srinivasan, N. and Gopalda, T. Least-cost multi-mix supplements: Development of formula to f i l l dietary deficits of pre-school children. Ind. J . Nutr. Dietet. 11: 276, 1974. Keen, H. and Rose, G.A. Diet and arterial disease in a population sample. Brit. Med. J . i : 1508, 1958. Keys, A . , Aravenis, D. and Sdrin, H. The diets of middle-aged men in two rural areas of Greece. Voeding 27: 575, 1966. K i n d  e r , F.  Meal Management, 4th edition.  The Macmillan Company, N.Y., 1973.  216  Kitchen, A.H., Passmore, R., Pyke, M. and Warnock, G.M. Studies of the diet of students at Edinburgh University. Brit, J . Soc. Med. 3: 10, 1949. Klevay, L.M. Coronary heart disease and dietary fiber. 27: 1202, 1974. Krehl, W.A. The evaluation of nutritional status. 1129, 1964.  Amer. J . Clin. Nutr.  Med.Clin. N. Amer. 48:  Kugler., H.J. Nutrition, Health and Activity Profile. Pacific Research Systems, 2222 Corinth Avenue, Los Angeles, California, 1976. Lachance, P.A.  Food guides.  J . Nutr. Educ. 4:  44, 1972.  Lalonde, M. A New Perspective on the Health of Canadians. Canada, Ottawa, 1974. Latham, M.C. and Stephenson, L.S. 9: 152, 1977.  Information  U.S. Dietary Goals - For.  J . Nutr. Educ.  Leichsenring, J.M. and Wilson, E.D. Food composition table for short method of dietary analysis (2nd revision). J . Amer. Dietet. Assoc. 27: 386, 1951. :  Lemke, C.E. On complementary pivot theory. In^: Mathematics of the Decision Sciences. Edited by G.B. Dantzig and A.F. Veinott. American.Mathematical Society, Providence, Rhode Island, 1968, p. 95. Leverton, R.M. What is nutrition education? T.7, 1974.  J . Amer. Dietet.  Leverton, R.M. The RDA's are not for amateurs. 66: 9, 1975.  Assoc. 64:  J . Amer. Dietet. Assoc.  Lewis, H.C. and Peng, P.H. The three-consideration diet revisited. Amer. Dietet. Assoc. 70: 270, 1977.  J_.  Litman, T . J . , Cooney, J.P. and Stief, R. Views of Minnesota school children on food, J . Amer. Dietet. Assoc. 45: 433, 1964. Lorstad, M.H. Recommended intake and its relation to nutrient F.A.O. Nutr. Newsletter 9: 18, 1971.  deficiency.  Lowenberg, M.E., Todhunter, E . D . , Wilson, E . O . , Savage, J.R. and Lubawski, J . L . Food and Man, 2nd edition. John Wiley and Sons, Toronto, 1974. Lund, L.A. and Burk, M.C. A Multidisciplinary Analysis of Children's Food Consumption Behavior. Tech. Bull. 265. Agriculture Experimental " Station, University of Minnesota, St. Paul, Minnesota, 1969. MacMillan, J.B. and Smith., E.-B. Development of a lacto-ovo vegetarian food guide. J . Can. Dietet. Assoc. 36: 110, 1975.  217  Mann, G.V. Diet-heart:  End of an era.  New Engl. J . Med. 297:  644, 1977.  Mann, G.V., Pearson, G . , Gordon, T. and Dawber, T. Diet and cardiovascular disease in the Framingham study. I. Measurement of dietary intake. Amer. J ; Clin. Nutr. 11: 200,1962. Manoff, R.K. Potential uses of mass media in nutrition programs. Educ. 5: 125, 1973. Manoff, R.K, Towards a new "American Mentality". 1975. Marr, J.W. Individual dietary surveys: Nutr. Diet. 13: 105, 1971.  J . Nutr.  J . Nutr. Educ. 7:  Purposes and methods.  •  139,  World Rev.  Marr, J.W., Heady, J.A. and Morris, J.N. Repeat individual weighed dietary surveys. Proc. Nutr. Soc. 18: x i i , 1959. Mayer, J .  Towards a national nutrition policy.  Science 176:  Mayer, J .  Forward.  xxi, 1973.  Mayer, J .  Time for reappraisal.  World Rev. Nutr. Diet. 16:  J . Nutr. Educ. 7:  237, 1972.  8, 1975.  McBean, L.O. and Speckmann, E.W. A Review: The importance of nutrition in oral health. J . Amer. Dent. Assoc. 89: 109, 1974. McCann, M.B., Trulson, M.F. and Stare, F . J . Follow-up of serum cholesterol, diet and physical findings of Italian-American factory workers. Amer. J . Clin. Nutr. 9: 351 , 1961.' McClinton, P., Milne, H. and Beaton, G. An evaluation of food habits and nutrient intakes in Canada: Design of effective food guides. Can. J . Public Health 62: 139,1971. McKenzie, J . L . and Mumford, P. Evaluation of nutrition education programs: A review of the present situation. World Rev. Nutr. Diet. 5: 21, 1965. McNabb, M.E. 90-day nonselective menus by computer. May 16, 1971.  Hospitals 45:  88,  Meneely, G.R. and- Battarbee, H.D. Sodium and Potassium. In: Present Knowledge of Nutrition, 4th edition. Nutrition Foundation, New York, 1976, p. 259. Mertz, W. Human requirements: 191 , 1972.  Basic and optimal.  Ann. N.Y. Acad. Sci. 199:  Milne, H . , Kerr, C. Trenholme, M. and Beaton, G.H. Studies of teenage eating in Ontario. II. Evaluation of a diet scoring method. Can. J . Public Health 54: 463, 1963. Mitchell, D.  The Politics of Food.  James Lorimer and Company, Toronto, 1975.  218  Mongeau, E. Dietary assessment of nutritional status. r_n: Assessment of Nutritional Status, Miles Symposium '73. Miles Laboratories, Limited, 1974, p. 13. Moore, J . L . and Wendt, P.F. Nutrition labeling - A summary and evaluation. J . Nutr. Educ. 5: 121, 1973. Morris, J . N . , Marr, J.W., Heady, J . A . , Mills, G.L. and Pilkington, T.R.E. Diet and plasma cholesterol in 99 bankmen. Brit. Med. J . i : 571, 1963. Morrison, S.D., Russel, F.C. and Stevenson, J . Estimating food intake by questioning and weighing: A one-day survey of eight subjects. Brit. J . Nutr. 3: v, 1949. Neter, J . and Wasserman, W. An experimental comparison between fixed weight and linear programming food price indexes. J . Amer. Stat. Assoc. 65: 49, 1970. Ohlson, M.A. Experimental and Therapeutic Dietetics, 2nd edition. Publishing Co., Minneapolis, Minnesota, 1972.  Burgess  Page, L. and Phipard, E.F. Essentials of and adequate diet. USDA Agriculture Research Service, Home Economics Research Report 3, 1957. Pao, E.M. and Burke, M.C. A computer-assisted approach to meal patterning. J . Amer. Dietet. Assoc. 65: 144, 1974. Passmore, R., Nicol, B.M., Narayana Rao, M . , Beaton, G.H. and Demayer, E.M. Handbook on Human Nutritional Requirements. World Health Organization, Geneva, 1974. Paul, 0., Lepper, M.H., Phelan, W.H., Dupertuis, G.W., MacMillan, A . , McKeen, H. and Park, H. .A longitudinal study of coronary heart disease. Circulation 28: 20, 1963. Pauling, L. Vitamin C and the common cold and the f l u . Company, San Francisco, 1976.  W.H. Freeman and  Payton, E . , Crump, E.P. and Horton, C P . Growth and development. VII. Dietary habits of 571 pregnant Southern Negro women. J . Amer. Dietet. Assoc. 37: 129, 1960. Pekkarinen, M. Chemical analysis in connection with dietary surveys in Finland. Voeding 28: 1,1967. Pekkarinen, M. Methodology in the collection of food consumption data. World Rev. Nutr. Diet. 12: 145, 1970. Pekkarinen, M. andRoine, P. Studies on the food consumption of the rural population in east and west Finland. Ann. Med. Exp. Fenn. 42: 93, 1964. Pekkarinen, M . , Kivioja, S. and Jortikka, L. A comparison of the food intake of rural families estimated by one-day recall and precise weighing methods. Voeding 28: 470, 1967.  219 Pennington, J.A. Dietary Nutrient Guide. The AVI Publishing Company, Incorporated, Westport, Connecticut, 1976. Pike, R.L. and Brown, M.L. Nutrition - A n Integrated Approach, 2nd edition, John Wiley and Sons Incorporated, New York, 1975. Poolton, M.A. Predicting application of nutrition education. Educ. 4: 110, 1972. Robinson, C H . Nutrition education - what comes next? Assoc. 69: 126, 1976.  J . Nutr.  J . Amer. Dietet.  Rosenstock, I . M . , Hoefner, D.P., Kegeles, S.S.. and Kirscht, J.P. Public knowledge, opinion and action concerning three public health issues: radioactive fallout, insect and plant sprays, and fatty foods. J . Health and Human Behavior. 7: 91, 1966. Sager, J . F . The United States Army Hospital food service data processing system. J_n: Computer-assisted Food Management Systems. Edited by A.N. Moore and B.H. Tuthill. University of Missouri, Columbia, Missouri, 1974, p. 75. Schaum, K.D. Patient-oriented dietetic information system. III. A computer-assisted program for calculating nutrient intake. J . Amer. Dietet. Assoc. 63: 154, 1973. Schaum, K.D. and Sharp, J . L . Patient-oriented dietetic information system. I. A computerized menu print and diet order system. J . Amer. Dietet. Assoc. 63: 35, 1973. Smith, E.B. A guide to good eating the vegetarian way. 7: 109, 1975.  J . Nutr. Educ.  Smith, P.E. The diet problem revisited: A linear programming model for convex economists. J . Farm Econ. 43: 706, 1961. Smith, V . E . Electronic Computation of Human Diets. MSU Business Studies, Bureau of Business and Economic Research, Graduate School of Business Administration, Michigan State University, East Lansing, Michigan, 1963. Smith, V.E. A. diet model with protein quality variable. Science 20: 971, 1974. Sorenson, A.W. and Hansen, R.G. Index of Food Quality. 7: 53, 1975. '  Management J . Nutr. Educ. '  Sorenson, A.W., Wyse, B.W.., Wittwer, A . J . and Hansen, R.G. An index of nutritional quality for a balanced diet. J . Amer. Dietet. Assoc. 68: 236, 1976. Spiller, G.A. and Amen, R.J. Dietary fiber in human nutrition. Reviews in Food Science and Nutrition. 6: 39, 1975.  Critical  220  Stevens, H.A., Bleiler, R.E. and Ohlson, M.A. Dietary intake of five groups, of subjects: 24-hour recall diets vs. dietary patterns. J . 'Amer. Dietet. Assoc. 42: 387, 1963. Stigler, G.J. The cost of subsistence.  J . Farm Econ. 27:  Taggart, N. Diet, activity and body-weight. woman. Brit. J . Nutr. 16: 223,1962. Thompson, E.M. and Tucker, H. Assoc. 40: 308, 1962. Thompson, J . J .  303, 1945.  A study of variation in a  Computers in dietary studies. J . Amer. Dietet.  Instructional Communication.  Van Nostrand, Toronto, 1969.  Thomson, A.M. Diet in pregnancy. I. Dietary survey technique and the nutritive value of diets taken by primigravidae. Brit. J . Nutr. 12: 446, 1958. Tie, L . T . , Lian, O.K., Liong-Ong, T.W. and Rose, C.S. Health, development, and nutrition survey of preschool children in Central Java. Amer. J . Clin. Nutr. 20: 1260, 1967. Todhunter, E.N. Approaches to nutrition education. J . Nutr. Educ. 1969. Travers, R.M. Essentials of Learning.  1: 8,  The MacMillan Co., New York, 1963.  Trenholme, M. and Milne, H. Studies of teenage eating in Ontario. I. Evaluation of teenage food selection. Canadian J . Publ. HIth. 54: 455, 1963. Trulson, M.F. Assessment of dietary study methods. methods for obtaining data for clinical work. 30: 991, 1954.  I. Comparison of J . Amer. Dietet. Assoc.  Tuthill, B.H. Computer-assisted food management systems at the medical center, UMC. In: Computer-assisted Food Management Systems. Edited by A.N. Moore and B.H. Tuthill. University of Missouri, Columbia, Missouri, 1974, p. 9. T u t h i l l , B.H. and Moore, A.N. Inventory control system, Department of Nutrition and Dietetics, Medical Center, UMC. Irr. Computer-assisted Food Management Systems. Edited by A.N. Moore and B.H. Tuthill. University of Missouri, Columbia, Missouri, 1974, p. 19. Tyler, K. The influence of mass media on children's food habits. lrr. Proceedings of the Nutrition Education Conference. USDA Misc. Publ. No. 913, Washington, D.C., 1962.  .221  Ullrich, H.D.  Teaching tools.  J . Nutr. Educ. 2:  Ullrich, H.D. 1973a.  The right way:  In whose judgement?  Ullrich, H.D. Needed: 5: 224, 1973b.  80,  1971.  J . Nutr. Educ. 5:  A definition, legislation, action.  Ullrich, H.D. Nutrition education is survival education. 84, 1974a. Ullrich, H.D. Energy crisis today, food crisis tomorrow? 6: 4, 1974b. Ullrich, H.D.  What's nutritious?  J . Nutr. Educ. 7:  48,  184,  J . Nutr. Educ. J . Nutr. Educ. 6: J . Nutr. Educ. 1975.  United States. Bureau of Human Nutrition and Home Economics, in cooperation with the National Research Council. Tables of Food Composition in Terms of Eleven Nutrients. USDA Misc. Publ. No. 572, 1945. United States. Nutritive.Value of Foods. No. 72, 1960.  USDA Home and Garden Bulletin  United States. Food and Nutrition Board. National Research Council. Toxicants Occuring Naturally.in Foods, 2nd Edition. National Academy of Sciences, Washington, D.C., 1973. United States. Food and Nutrition Board, National Research Council. Recommended Dietary Allowances, 8th edition. National Academy of Sciences, Washington, D.C., 1974. United States. Consumer and Food Economics Institute, Agriculture Research Service, USDA. Data published in Composition of Foods - Raw, Processed, Prepared. Rev. USDA Agriculture Handbook No. 8, 1963. Data sets 8-1-0, 2-2-0, and 8-3-0, CFE Form 52, August 1977a. United States. Expansion 1 (March 1972) of data published in Composition of Foods - Raw, Processed, Prepared. Rev. USDA Agriculture Handbook No. 8, 1963. Data sets 8-1-1 and 8-2-1, CFE Form 52, August 1977b. United States. Data published in Home and Garden Bulletin No. 72. Rev., 1971. Data sets 72-4-0, CFE Form 52, August 1977c. United States. Data published in Nutritive Value of American Foods in Common Units, USDA Agriculture Handbook No. 8, 1975. Data sets 456-1, CFE Form 52, August 1977d. United States. Data published in Agriculture Handbook No. 8-1, Rev. 1976. Data sets 8-1 and 8-2, CFE Form 52, 1977e. United States. Select Committee on Nutrition and Human Needs, U.S. Senate. Dietary Goals for the United States. Government Printing Office, Washington, D.C., 1977f.  222  United States Department of Agriculture (U.S.D.A.). USDA Leaflet No. 288, 1946.  The Basic 7 FOod Groups.  United States Department of Agriculture-(U.S.D.A.). Dietary Levels in the United. States. Spring 1965. U.S. Govt. Print Office, Washington, D . C , 1968. United States Department of Agriculture (U.S.D.A.). USDA Leaflet No. 424, 1971.  A Daily Food Guide.  United States Department of Health, Education, and Welfare. Ten-State Nutrition Survey, 1968-1970. DHEW Publ. No. (HSM) 72-8133, 1972. Vajda, S. Readings in Linear Programming. York, 1958.  John Wiley and Sons, Inc., New  Watt, B.K. and Merrill, A.L. Composition of Foods - Raw, Processed, Prepared. USDA Agriculture Handbook No. 8, 1950. Watt, B.K. and Merrill, A.L. Composition of Foods - Raw, Processed, Prepared. Rev. USDA Agriculture Handbook No. 8, 1963. Watt, B.K., Gebhardt, S . E . , Murphy, E.W. and Buntrum, R.R. Food composition tables for the 70's. J . Amer. Dietet. Assoc. 64: 257, 1974. Werko, L. Risk factors and coronary heart disease - facts or fancy? Amer. Heart J . 91: 84, 1976. White, P.L. Why all the fuss over nutrition education? 8: 54, 1976.  J . Nutr. Educ.  Whiting, M.G. and Leverton, R.M. Reliability of dietary appraisal: Comparisons between laboratory analysis and calculation from tables of food values. Amer. J . Publ... Hlth. 50: 815, 1960. Widdowson, E.M. and McCance, R.A. Lancet i : 230, 1943.  Food tables.  Their scope and limitations.  Wiehl, D.G. and Reed, R. Development of new or improved dietary methods for epidemiological investigations. Amer. J. Publ. Hlth. 50: 824, 1960. Wilson, C . S . , Schaefer, A . E . , Darby, W.J., Bridgforth, E . B . , Pearson, W.N., Combs, G . F . , Leatherwood, E . C , Greene, J . C , Teply, L . J . , Plough, I . C , McGanity, W.J., Hand, D.B., Kertesz, Z . I . and Woodruff, C.W. A review of methods used in nutrition surveys conducted by the Interdepartmental Committee on Nutrition for National Defense (ICNND). Amer. J . Clin. Nutr. 15: 29, 1964. Winarski, M. Food guides. Abused and misused. Econ. 19: 225, 1976.  Illinois Teacher of Home  Witschi, J . , Porter, D., Vogel, S., Buxbaum, R., Stare, F . J . and Slack, W. A computer-based dietary counselling system. J . Amer. Dietet. Assoc. 69: 385, 1976.  223  Wittwer, A . J . , Sorenson, A.W., Wyse, B.W. and Hansen, R.G. Nutrient densityEvaluation of nutritional attributes of food.' J . Nutr. Educ. 9: 26, 1977. Wolff, R.J.  Who eats for health?  World Health Organization (W.H.O.). 32: 377, 1978.  Am. J . Clin. Nutr. 26: Chemicals and health.  438,  1973.  WHO Chronicle.  Wyse, B.W., Sorenson, A.W., Wittwer, A . J . and. Hansen, R.G. Nutritional quality index identifies consumer nutrient needs. Food Technology. 30: 22, 1976. Yano, K., Rhoads, G.G. and Kagan, A. Coffee, alcohol and risk of coronary heart disease among Japanese men living in Hawaii. New Engl. J . Med. 297: 405, 1977. Young, C M . , Berresford, K. and Waldner, B.G. What the homemaker knows about nutrition. I l l : Relation of knowledge to practice. J . Amer. Dietet. Assoc. 32: 321, 1956a. Young, C M . , Berresford, K. and Waldner, B.G. Nutrition knowledge and practices. Pub. Health Rept. 71: 487, 1956b. Young, C M . , Chalmers, F.W., Church, H.N., Clayton, M.M., Hagan, G . C , Steele, B . F . , Tucker, R.E. and Foster, W.D. Subject's ability to estimated food portions. Univ. of Mass.,.Agric. Exp. Stat. Bull. 469: 63, 1952a. Young, C M . , Chalmers, F.W., Church, H.N., Clayton, M.M., Tucker, R . E . , Wertz, A.W. and Foster, W.D. A comparison of dietary study methods. I. Dietary history vs. seven-day record. J . Amer. Dietet. Assoc. 28: 124, 1952b. Young, C M . , Hagan, G . , Tucker, R.E. and Foster, W.D. A comparison of dietary study methods. II. Dietary history vs. seven-day records vs. 24-hour recall. J . Amer. Dietet. Assoc, 28: 218, 1952c. Young, E.G. Dietary standards. J_n: Nutrition - A Comprehensive Treatise. Edited by G.H. Beaton and E..W. McHenry. Academic Press, New York, 1964, Volume 2, p. 299. Young, M.A.C Reviews of research and studies related to health education practice. Health Education Monographs 23: 12, 1967. Yudkin, J . Dietary surveys: Variations in weekly intake of nutrients. Brit. J . Nutr. 5: 177, 1951.  . 224  APPENDIX A FOOD-ITEM FILE  Two hundred and twenty-one item clusters with attribute group code number, item cluster code number, standard portion sizes, and gram equivalents per portion.  ATTRIB. GROUP CODE #  ITEM GRAMS CLUSTER PER CODE # PORTION  STANDARD PORTION SIZE  FOOD ITEMS AND DESCRIPTION  ENTREE.ITEMS - DAIRY AND EGGS  001 001 001 001 001  001 002 003 004 005  28 28 28 114 28  r-r-iy  d oz.)^  V-V-lh"  (1 oz.)"  002 002 002 002  006 007 008 009  15 244 246 246  1 tbsp. 1 cupd 1 cupd 1 cup"  003 003  010 on  55 140  d egg 2 eggs  r-r-1%" (i oz.r h cup (4 o z . ) . d  2 tbsp. (1 oz.)  1  Q  AMERICAN PROCESSED CHEESE. BLUE or ROQUEFORT CHEESE. CHEDDAR, JACK or SWISS CHEESE. COTTAGE CHEESE, creamed or uncreamed, any curd. CREAM CHEESE or CHEESE SPREADS. SOUR CREAM or YOGHURT, base YOGHURT, base YOGHURT, base  d  CHIP DIP. of whole milk (3.5% b.f.) of skim milk. of part skimmed and 2% nonfat milk solids  EGG: raw, boiled, poached, fried (add fat). EGG: scrambled, omelets, souffles, spoon bread.  d 0  ENTREE ITEMS - CEREALS  004 004  012 013  28 28  005 005 005  014 015 016  120 120 120  006  017  90  007 007  018 019  80 75  008 008 008 008  020 021 022 023  75 75 120 28  1 1  (1 (1  cup cup  % cup h cup h cupd  oz.) oz.)  d  CORN CEREAL, enriched, eg. corn flakes: ready-to-eat. WHEAT CEREAL, enriched, eg. rice krispies: ready-to-eat. WHEAT CEREAL, more refined, eg. cream of wheat: cooked. OATMEAL, a l l types: cooked.  d  d  WHEAT CEREAL, less refined, eg. rolled wheat: 2 at 4" d i a .  cooked.  d  PANCAKES, WAFFLES or FRITTERS, made with milk and eggs.  d  h cup h cupd  NOODLES, EGG TYPE, enriched: cooked. SPAGHETTI, MACARONI or NON-EGG PASTAS, enriched: cooked.  M  ro  d  h cup h cupV h cup 3 tbsp. (1 oz.)  RICE, BROWN: cooked. RICE, WHITE, enriched, unenriched or parboiled: CORNMEAL or CORN GRITS, enriched: cooked. WHEAT GERM.  cooked.  009 009 009 009 009 009 009  024 025 026 027 028 029 030  20 23 23 23 23 40 30  1 1 1 1 1 1 1  s l i c e (9/16")J I slice (9/16")° I slice (9/16")° I s l i c e (9/16"ft I s l i c e (9/16") . square ( 2 " - 2 " - l V ) at 6" d i a .  010 010 010 010 010  031 032 033 034 035  35 46 40 40 46  1 at 2" d i a . id l l l  011  036  14  Oil  037  6  Oil Oil 011  038 039 040  21 30 14  u  d  FRENCH or SOURDOUGH BREAD, enriched: fresh or toasted. RAISIN or RAISIN-NUT BREAD, enriched: fresh or toasted. RYE BREAD, l i g h t or dark: fresh or toasted. WHITE or CRACKED WHEAT BREAD, enriched: fresh or toasted. WHOLE WHEAT BREAD: fresh or toasted. CORN BREAD or JOHNNY CAKE. CORN TORTILLAS or FLOUR TORTILLAS. BISCUITS, ROLLS or POPOVERS. HAMBURGER BUN, HOT DOG BUN, or BAGEL, enriched. MUFFIN, white, bran, blueberry, brownbread. SOURDOUGH ENGLISH MUFFIN. WHOLE WHEAT ENGLISH MUFFIN.  e  d d  d  1 at 5"-2%"-3/16"£  2 at lV'-14"-l/8"  t  10 at 3 V - 1 / 8 " d i a . 1 cups . 15 chips (1 oz.)J 2 at 3 V - 1 4 " - V 3^ 8 e  GRAHAM CRACKERS, honey-coated or whole wheat. SODA CRACKERS, s a l t i n e s , holland rusk, matzoth. PRETZELS. POPCORN: popped (add s a l t and fat i s used). POTATOE CHIPS, FRITOS, CORN PUFFS or TORTILLA CHIPS. RYE KRISP. TRISCUITS. WHEAT THINS. ENTREE ITEMS - MEATS AND PLANT PROTEIN  012 012 012 012 012 012 012 012  041 042 043 044 045 046 047 048  85 85 85 85 85 85 85 16  013 013 013 013  049 050 051 052  85 200 85 85  1/3 cup (3 1/3 cup (3 1/3 cup (3 1/3 cup (3 1/3 cup (3 1/3 cup (3 1/3 cup (3 2 slices  oz.jjj oz.) oz.) oz.) d  d  d  oz.f  oz.) oz.)  d a  d  d  1/3 cup (3 o z . ) 1 cup . 1/3 cup (3 o z . n 1/3 cup (3 o z . ) d  a  BEEF, 30% fat cut, eg. chuck r i b , v e a l : cooked. BEEF, 20% fat c u t , eg. regular ground hamburger: cooked. BEEF, 15% fat cut, eg. round: cooked. CORNED BEEF, fresh or canned. LAMB, a l l muscle c u t s , eg. l e g , shoulder, chops: cooked. PORK, lean c u t s , eg. chops, shoulder, roast: cooked. PORK, a l l hams, fresh or canned. BACON, t h i c k cut, t h i n cut or s l a b : cooked, drained. CHICKEN: steamed, stewed, b r o i l e d , baked or canned. CHICKEN: prepared with sauce, eg. f r i c a s s e e , c a c c i a t o r e . CHICKEN: f r i e d flesh and s k i n . TURKEY, DUCK, RABBIT or SQUAB GOOSE: roasted.  d  5 small (3%oz.) . 4 to 5 sticks (3%oz.) (3%oz.r (3 o z . ) , 5 to 8 medium (3%oz.)Jj 5 to 8 medium (3*5 oz.) (3 oz.)d (2 o z . ) (3 o z . ) . 2 medium (1 o z . ) (3 o z . ) (3% oz. ) . 1/3 can.(2 oz.) (3 oz.)  CLAMS: canned. FISH STICKS, FISH CAKES, FISH LOAFS: cooked. FRIED FISH, eg. fried haddock: breaded, fried. BROILED FISH, eg. broiled halibut. OYSTERS: canned. OYSTERS: breaded, fried. SALMON, all types: fresh or frozen, cooked. SALMON, canned. SMOKED FISH, all types, eg. smoked salmon. SARDINES or KIPPERS: fresh or canned. SHRIMP, LOBSTER, CRAB or ABALONE: fresh or canned. SHRIMP: fried. TUNA: canned, oil or water-pack. RAW FISH, eg. raw tuna.  014 014 014 014 014 014 014 014 014 014 014 014 014 014  053 054 055 056 057 058 059 060 061 062 063 064 065 066  100 100 100 85 100 100 85 56 85 28 85 100 60 85  015  067  85  (3 o z . )  d  LIVER, beef, calf, hog, chicken or lamb: fried.  016  068  85  (3 o z . )  d  KIDNEY.  017  069  45  ALL LUNCH MEATS EXCEPT LIVERWURST. 1 at 5"-^" FRANKFURTER: cooked, xl .4(1 link at 4"-^ ") * KNOCKWURST. x l . l ( l slice at 4"-3"-fc") HEAD CHEESE, xl .6(1 slice at 4%"-4V'-#') BOLOGNA. 3 at 2"-$' d i a . VIENNA SAUSAGE. 1 at 3%'-3%'-k" SALAMI. x2.8(l slice at 4V'-4V'-&') LOAF MEAT, eg. ham loaf, olive loaf. 3 tbsp. MEAT SPREADS, eg. deviled ham. 2 links at 3"-h" dia. FRESH COOKED SAUSAGES, eg. pork sausages. 1 slice at 3" d i a . - V ' LIVERWURST or PATE DE FOIS GRAS.  a  d  d  d  d  d  f  T  f  T  8  p  e  P  e  e  e  f  f  017 017  070 071  40 30  018 018 018 018 019  072 073 074 075 076  125 100 85 75 15  e  d  h cup . h cup (3% oz.) h cup ,(3 o z . ) h cupd 12 to 15° 10 to 12 2 tbsp.; t tbsp. " d  e  e  019  077  15  1  BEANS, WHITE, RED, PINTO, KIDNEY: canned. BEANS, white, red, pinto or kidney: boiled and drained. COWPEAS or BLACK-EYED PEAS: boiled and drained. SOYBEANS: boiled and drained. ALMONDS. FILBERTS or HAZELNUTS. SESAME SEEDS, CASHEW BUTTER.  ro  d  019  078  30  6 to 8 2 pieces at l"-l"-%" cup shredded . 2 tbsp. (1 o z . ) ° 1 tbsp. , 2 tbsp., 8 to 10 halves  CASHEW NUTS. COCONUT: fresh, shredded or dried,  e  019 019 019  079 080 081  28 15 15  PEANUT BUTTER and OTHER NUT BUTTERS EXCEPT CASHEWS. PEANUTS or SPANISH PEANUTS. WALNUTS, persian, black or english: PECANS.  d  0  ENTREE ITEMS - VEGETABLES e  020 020 020 020 020  083 084 085 086 087  50 100 180 100 100  round: 1 at 2%" to 2%" d i a . round: xl.5(1 at 3k" d i a . ) long: x2.5(l at 2%' dia-4^') 10 pes. at h"-k"-2:, cup h cup (3% oz.)" 1 at 2" d i a . - 5 : , or ^cup . 1 at 2" dia.-4"e, or h cup 2/5 cup  021 021 021 021 021 021 021 021 021 021 021 021 021 021  088 089 090 091 092 093 094 095 096 097 098 099 100 102  100 65 65 63 43 62 85 37 100 85 85 62 85 27  h h h h h h \ h h h h h h h  022 022  103 104  80 76  020  082  100  POTATOE, all white types: baked or boiled,  e  022  105  50  e  f  a  e  3 f of 3V dia.-4" , or%cup cup cup cup . cup cup cup cup chunks cup cup cup cup. cup cup  i  0  d  d d  d  d  d  d  d  0  0  d 0  ,  slg.(9"-lV)»|med.(7¥ -lV) 1 sml. (5" ) >4cup di ced^ h cup shredded'  POTATOE, all white types: fried, eg. french fried. POTATOE, all white types: mashed with fat and milk. SWEET POTATOE: baked in skin. SWEET POTATOE: canned in syrup. YAM: baked or boiled. AVACADO, raw or avacado dip. BEANS, snap green or wax: fresh or frozen, boiled. BEANS, snap green or wax: canned, boiled and drained. BROCCOLI: boiled and drained. CABBAGE, CAULIFLOWER or SAVOY: raw. CABBAGE, BRUSSEL SPROUTS or CAULIFLOWER: boiled, drained. COLLARDS or KALE: boiled, drained. LETTUCE, crisp head, romaine, iceberg or endive: raw. GREENS, mustard or turnip: fresh or frozen, boiled, drained. PEAS, LIMA BEANS or SNOW PEAS: canned, boiled, drained. PEAS, LIMA BEANS or SNOW PEAS: fresh or frozen, boiled. PEPPERS, sweet green: raw, canned or boiled. SPINACH, BOK CHOY, or BEET GREENS: canned, drained. SPINACH: raw.  d  h cup h cup  f T  f T  BEETS or ARTICHOKE: raw or canned, cooked, drained. CARROTS or WINTERSQUASH: boiled and drained. MATURE RED PEPPERS or HOT CHILI PEPPERS: canned or fresh. CARROTS: raw.  £  CO  022 022 022 022 022 022  106 107 108 109 no 111  85 126 85 100 100 148  % cup (3 oz.) % cup . % cup (3 oz.) , h cup (3% oz.) % cup (3% oz.) T 1 med.(2^'), x^sml.(2^') x 1% l r g . ( 3 " ) h cup piece (24" dia.-l%") or f  CORN: fresh or frozen, boiled and drained. CORN, cream-style: canned CORN, kernal: canned, boiled and drained. SUMMER SQUASH, ASPARAGUS or ZUCCHINI: boiled, drained. TOMATO: canned solid and liquid. TOMATO: raw  d  d  f  f  022 023  112 113  77 50  TURNIP ROOT or TUTABAGA: boiled and drained, CUCUMBER: raw.  d  (14" dia.-2")  1 Irg. stalk(8"-lV) or  CELERY: raw  3  3 sml. stalks(5"- 4')f 10 W to 1") or 6(1" h cup (3% oz.) k cup d  , T  to 1V') RADISHES: raw MUSHROOMS or WATER CHESTNUT: fresh or canned, cooked. ONIONS, LEEKS, GARLIC or GREEN ONIONS: raw or boiled.  023 023  114 115  100 45  024  116  85  h cup (3 o z . )  025  117  13  4(%" dia.9fc"),-3(V dia.-^), OLIVES, green, black or stuffed. 2(%" dia.-l^") " 1(1" d i a . - 3 V ' ) , 1(1" dia.- PICKLES, sweet, bread and butter or pickle relish. 1^"), 3(1%" dia.-V), overful tbsp. k l r g . ( l % ' dia.-4"), %cup, PICKLES, d i l l or sour. h med.(lV 5(1%" dia.-%")  d  MIXED VEGETABLES or SUCCOTASH: frozen, boiled.  d  1  025  118  20  025  119  34  dia.-3V)\ f  ENTREE ITEMS - FATS AND OILS 026 026 026 026 026  120 121 122 123 124  15 15 15 15 15  1 1 1 1 1  tbsp.j tbsp ° tbsp tbsp. tbsp. d  d  VEGETABLE FAT, eg. crisco or hardened shortening. LARD, SUET or SALT PORK. COTTONSEED OIL. OLIVE OIL. SOYBEAN or CORN OIL.  027 027  125 126  5 5  1 tsp.^ 1 tsp.  BUTTER, sweet or salted butter. MARGARINE, all brands, whipped or diet.  028  127  38  2 tbsp.  CHEESE SAUCE or FONDUE.  d  ro ro  NO  028 028 028 028  128 129 130 131  72 33 17 100  4 1 1 h  tbsp., or h cup tbsp.d tbsp. cup  029 029 029 029  132 133 134 135  15 15 15 14  1 1 1 1  tbsp tbsp ° tbsp. , tbsp.  GRAVY, all types. WHITE SAUCE, thick or thin, or HOLLANDAISE SAUCE. TOMATO CATSUP or BARBECUE SAUCE. TOMATO SAUCE.  d  d  MAYONNAISE, SANDWICH SPREAD or TARTAR SAUCE. SALAD DRESSING, oil types, eg. Italian. SALAD DRESSING, mayonnaise type, eg. thousand island. ROQUEFORT or BLUE CHEESE SALAD DRESSING.  d  0  BEVERAGES - DAIRY d  030 030 030  136 137 138  244 246 246  1 cup (8 f l . o z . ) 1 cup (8 f1. o z . ) 1 cup (8 f1. o z . )  031 031 031  139 140 141  121 15 8  031  142  15  1 cup . 1 tbsp. 1 tbsp. whipped, 2 tbsp. whippedd 1 tbsp.  a a  d  d  WHOLE MILK, 3.5% b.f. SKIM MILK, BUTTERMILK or NONFAT INSTANT MILK. PART SKIM and 2% NONFAT MILK SOLIDS. HALF AND HALF. LIGHT TABLE CREAM and NONDAIRY COFFEE WHITENERS. WHIPPED CREAM, COOL WHIP or IMITATION TOPPINGS. CANNED EVAPORATED or CONDENSED MILK, sweet or unsweet. BEVERAGES - FRUIT and VEGETABLE  d  032 032 032 032  143 144 145 146  120 120 15 226  h cup (4 f1. h cup (4 f 1.  oz.) oz.) 1 tbsp, or '6 lemon 1 cup (8 f1. o z . ) °  032  147  120  a  oz.)  a  032 032  148 149  120 120  hcup (4 f 1. h cup (4 f l . h cup (4 f1.  oz.) oz.)  a  033  150  100  ^cup (4 f 1. o z . )  d  APPLE, GRAPE, PRUNE, or PEAR JUICE: fresh, frozen, canned GRAPEFRUIT JUICE: canned and unsweetened. LEMON, LIME or ACEROLA JUICE: fresh, frozen or canned. FRESH LEMONADE, ALL FRUIT ADES and ARTIFICIAL LEMONADES. ORANGE JUICE: canned. ORANGE JUICE: frozen.  a  ORANGE JUICE: fresh. TOMATOE JUICE or VEGETABLE JUICE: canned. 034 034  151 152  339 339  12 f 1. oz.j! 12 f 1. oz.  BEVERAGES - MISCELLANEOUS COLA-TYPE BEVERAGES, caffeine-containing, carbonated. GINGER ALE, non-caffeine containing, carbonated.  COFFEE, all types. TEA, all types.  035 035  153 154  200 200  1 cup. 1 cup  036 036 036 036  155 156 157 158  360 43 60 100  12 f l . o z . . 1.5 oz. jigger 2 f l . oz., sherry glass 3.5 f1. oz., wine glass  d  BEER, ALE, or STOUT. DISTILLED SPIRITS, eg. gin, rum, whiskey or vodka. DESSERT or SWEET WINE TABLE or DRY WINE. SOUPS  037 037 037 037 037 037  159 160 161 162 163 164  198 198 200 198 203 200  8/10 8/10 8/10 8/10 8/10 8/10  cup, cup, cup, cup, cup, cupd  NOODLE or RICE SOUPS WITH MEAT: canned, dry or instant. CREAMED SOUPS MADE WITH MILK: canned, dry or instant. PEA or BEAN SOUP: canned, dry or instant. TOMATOE SOUP WITH OR WITHOUT RICE. MEAT AND VEGETABLE AND VEGETARIAN VEGETABLE SOUPS. ONION SOUP or CONSOMME.  can can can can can  DESSERTS AND SWEETS - CEREALS 038  165  45  I V arc of cake 9V dia.-4 or xl .35 of 2H a r c 1 piece 3"-2J "-l% 1 " arc of cake 8" dia.-3" xl%(2V d i a . ) , xl.65(2V dia.) square 2"-2"-l% 1 piece 3"-3"-h" * 1 piece 3"-3"-h" I V arc cake 8" dia-3"  ANGEL FOOD CAKE, SPONGE CAKE or TWINKES.  ALL FRUIT PIES, eg. apple, pecan, lemon meringue. PUMPKIN or SWEET POTATOE PIE.  l  038 038  166 167  75 55  f  llf  2  f  3  f ,T  e  038 038 038  168 169 170  40 30 60  039  171  160  4V arc 9" d i a . , l/6tht  039  172  150  4V arc 9" d i a . , l/6th  040  173  40  2 cookies  040 041 041  174 175 176  29 30 30  2 bars 1 average^ 1 average  f  f  0  d  0  COFFEE CAKE, BANANA BREAD, DATE-NUT BREAD or SNAKIN CAKE CHOCOLATE CAKE WITH ICING. CUPCAKE WITH FROSTING. BROWNIE. FRUITCAKE, light or dark. POUND CAKE. ALL OTHER COMMERCIAL, HOMEMADE or FROZEN CAKE WITH ICING  ALL COOKIES WITHOUT FRUIT FILLING, eg. sandwich-type. ICE CREAM CONE or COOKIE OF ICE-CREAM SANDWICH. ALL FRUIT-FILLED COOKIES, eg. figbars or raisin cookies. CAKE DOUGHNUT, iced or uniced. YEAST-RAISED DOUGHNUT, iced or uniced.  041  177  DANISH PASTRY or HOT CROSS BUN. CINNAMON BUN or SWEET ROLL.  38 x 1.6(1)  (  DESSERTS AND SWEETS - DAIRY 042 042 042  178 179 180  90 97 339  d % cup V cup l  043 043  181 182  130 246  % cup 1 cup  ICE CREAM, ICE MILK or ICE CREAM BARS, all flavors. SHERBERT, all flavors. POPSICLE.  d  d  TAPIOCA and RICE PUDDING, JUNKET, CUSTARD or PIE FILLING. YOGHURT, add fruit and sugar i f included.  d  DESSERTS AND SWEETS - FRUIT 044  183  150  1(2*" d i a . ) , x l ^ 3 V d i a . ) , or x * ( 2 V d i a . ) . 1(3%" long-2%" d i a J 1 cup, 20(%" dia.) " 1 cup, 3(2£" d i a . ) , 4(1%" dia.)f  APPLE, raw.  f  e  T  1  2of  f  1/3 (10" d i a . - l " s l i c e ) %(6" d i a . - l V s l i c e ) -h: cup nd e x- r u~.. d e s cup 1 sml.(6"), x l%med.(8") %cup h of 5" dia. % of 4" dia, dia.) 1 sml. x 1% med.(3" d i a . ) x 2.35 lrg.(3%" dia.) 10 l r g . , 2/3 cup % cup 1 at 2%' d i a . , 1* (2%'Y or ™ 3 f % cup % cupd e  044  184  130  2  1n  044  185  100  e  044 044 044  186 187 188  100 100 100  PEAR, raw. GRAPES, raw. PLUMS, raw. CHERRIES, raw. WATERMELON, raw. HONEYDEW MELON or CASABA MELON, raw. APPLESAUCE: fresh or canned. PEARS, GRAPES, PLUMS or CHERRIES: canned. BANANA or PLANTAIN.  f  (2%"  (  CANTALOUPE. GRAPEFRUIT, fresh or unsweetened. ORANGE, raw.  c  (  e  044 044  189 190  126 100  044 044 045  191 192 193  139 80 28  e  2  d  2 tbsp. (1 oz.)'  STRAWBERRIES, raw. PEACHES, APRICOTS or FRUIT COCKTAIL, heavy syrup pack. PEACHES, PERSIMMONS or NECTARINES, raw. APRICOTS, raw. PINEAPPLE, MANDARIN ORANGE or BOYSENBERRIES, heavy syrup. PINEAPPLE, TANGERINES or BOYSENBERRIES, water or juice pack. RAISINS, PRUNES or OTHER DRIED FRUIT, not peaches or apric.  046  194  80  d  cup  h  GELATIN DESSERT,  w i t h or w i t h o u t fruit.  DESSERTS AND SWEETS - SWEETS 047 047 047  195 196 197  21 20 5  1 1 1  tbsp.J tbsp, tsp.  HONEY, all t ype. MOLASSES, all t ypes. SUGAR.  048 048  198 199  20 20  1 1  tbsp.j! tbsp.  JELLIES, PRESERVES, JAM or MARMALADES. CORN SYRUP or MAPLE SYRUP.  049  200  28  1 1  201  28  049  a  d  f  p i e c e (1"-1"-1V), (loz.) CARAMEL CANDY, TAFFY or VANILLA-COATED CARAMELS. p i e c e (2"-2"-V) 3 c a r a m e l 1 p i e c e (2"-2"-V), (1 oz.) CHOCOLATE CANDY, MILK CHOCOLATE or FUDGE. 1p i e c e (r-l"-lV) 2(1%" dia.-V), 10(V d i a . % CHOCOLATE DISK. ")f f  f  f  d  049  202  15  049  203  20  (1 oz.) 2 lrg. W dia.) 1 p i e c e (1"-1"-V), 2 e 3 lrg., 6 sml. 7 2  f  CHOCOLATE SYRUP o r FUDGE TOPPING. MARSHMALLOW. hard6 HARD CANDY. MARASHINO CHERRIES. CHEWING GUM JELLY BEANS. f  lrg. % ( " dia.),  16  T  sml.  GUM DROPS  or  JELLY CANDIES.  MISCELLANEOUS ITEMS 050  204  125  051  205  227  052  206  238  053  207  220  054  208  220  055  209  224  h  d  cup  1(4%"  dia.), or d  1  cup  1  cup  1  cup  1  cup  d  d  d  BEANS WITH PORK AND TOMATOE SAUCE, 1/3(9"  dia.)  POT PIES,  c a n n e d or h o m e m a d e .  c h i c k e n or tuna: c o m m e r c a il or h o m e m a d e .  MEAT AND VEGETABLE STEWS. CHOW MEIN  or  CHOP SUEY:  c a n n e d or frozen.  CHOW MEIN  or  CHOP SUEY:  h o m e m a d e .  CHILI CON CARNE WITH OR WITHOUT BEANS,  c a n n e d .  •G£ O  FROZEN DINNER: fried chicken, mashed potatoes and peas.  d  056  210  312  1 dinner, (11 o z . )  057  211  312  1 dinner, (11 oz.) '  058  212  300  1 dinner, (11 o z . )  059  213  224  h cup  060  214  225  h cup  061  215  200  3/8 of 14" pizza 3 (5V dia.) sector  FROZEN DINNER: meatloaf, mashed potatoes and peas.  0  FROZEN DINNER: roast turkey, mashed potatoes and peas.  d  HASHES, CANNED C0RNBEEF or ANY HOMEMADE HASH.  d  MACARONI AND CHEESE: homemade, packaged or frozen. .PIZZA, any kind.  f  d  062  216  200  1 cup  063  217  220  d 1 cup  064  218  155  2 from can of 6  065  219  4  1 cube  066  220  1  1 tsp.  067  221  7  1 tbsp.  )  f  SPAGHETTI IN TOMATOE SAUCE, CANNED RAVIOLI or NOODLE-0'S. SPAGHETTI WITH MEAT BALLS: homemade or packaged. d  TAMALES: homemade or canned. BOULLION CUBE.  d  TABLE SALT or MONO-SODIUM GLUTAMATE. d  COCOA MIX.  ro oo  For the legend of group codes, see Appendix G of this text. A food item cluster is one item or two or more specific items, item varieties, or prepared variations of an item with common nutrient characteristics. Item clusters are numbered in the sequence they appear in the food item f i l e . The food items l i s t is adapted from Mini List Foods and Food Substitutions in Pennington, J . A . , Dietary Nutrient Guide, AVI Publishing Co., Westport, Connecticut, 1976. Gram equivalents per portion are from Mini List Foods, Appendix C, Pennington, J . A . , Dietary Nutrient Guide, AVI Publishing Co., Westport, Connecticut, 1976. Standard portion size derived from Mini List Foods, Appendix C, in Pennington, J . A . , Dietary Nutrient Guide, AVI Publishing Co., Westport, Connecticut, 1976. Standard portion size derived from Church, C.F. and Church, H.N., Food Values of Portions Commonly Used, 12th edition, J.B. Lippincott Co., New York, 1975. Standard portion size derived from Adams, G . F . , Nutritive Value of American Foods in Common Units, Agriculture Handbook No. 456, U.S.D.A., 1975. Description of standard portions per gram quantity differ between  e  and  ro Co ui  236  APPENDIX B ABRIDGED FOOD-ITEM FILE  Listing, by group code and item cluster code, of 127 item clusters selected from Appendix A.  237 a  ATTRIBUTE GROUP CODE # 001 001 001 002 002 003 004 004 005 005 006 007 007 008 008 008 009 009 009 010 010 010 010 on on 012 012 012 012 012 012 013 013 014 014 014 014 014 014 015 017 017 017 018 018 019 019 019 019 019  ITEM CLUSTER CODE #  ATTRIBUTE CLUSTER CODE #  ITEM CLUSTER CODE #  ATTRIBUTE GROUP CODE #  ITEM CLUSTER CODE #  003 004 005 006 008 010 012 013 015 016 017 018 019 020 021 023 024 027 028 031 032 033 035 037 040 041 042 043 046 047 048 049 051 055 056 058 062 063 065 067 069 070 071 073 075 076 077 079 080 081  020 020 020 020 021 021 021 021 021 021 021 022 022 022 022 022 023 023 023 024 025 025 025 026 026 027 028 028 029 029 030 030 031 031 032 032 032 032 033 034 035 035 036 036 036 037 037 037 038 038  082 083 084 085 089 091 092 095 098 099 101 103 104 105 106 111 113 114 115 116 117 118 119 121 124 125 127 128 132 133 136 137 140 141 143 144 145 148 150 151 153 154 155 156 158 160 161 163 166 168  038 039 039 040 040 041 041 042 042 043 044 044 044 044 044 044 044 045 047 047 048 048 049 049 051 063 067  170 171 172 173 174 175 177 178 179 181 183 184 185 186 187 188 192 193 195 197 198 199 201 202 205 217 221  Refer to Appendix A for description of item clusters, standard portion sizes, and gram equivalents per portion associated with each of the coded items.  2.38  APPENDIX C FOOD-COMPOSITION FILE  Table of 41 nutrient values, in nutrients per 100 grams of edible portions of food, for 221 item clusters indexed by attribute group code and item cluster code number.  239  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS:b KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) (MG) LEU (MG 1 LYS MET (MG) CYS (MG) (MG J PHE TYR ( MG) (MG) VAL (MG) HIS (GM) FAT-T (GM) SFA PUFA (GM ) CHGLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM) (MG) THIA (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) (UG) PANTO B I O T I N (MG) VIT-A (IU ) VIT-D (IU) VIT-E (MG ) CA (MG » P (MG) MG (MG) (MG) FE (MG) I ZN (MG) NA (MG ) K (MG ) CU (MG)  1 1 370.000 23.20C 320.000 840.OOC 1540.CCC 2 2 3 0 .OCO 1670.OCC 600.000 130.OCC 1200 . 0 0 0 1110.OOC 1640.000 760.000 3 0 . OCC 15.000 10.OCC 0.150 1.9 00 0 .0 0.0 0.02C 0.410 0.0 80.OCC 11.000 1.000 0.0 500.000 5.OOC 1220.OCC 30.OOC l.OOC 697.000 771.000 48.000 0.90C 11.OCC 4.IOC 1136.OCC 80.000 0 . 17C  2 1 368.000 21.500 290.000 810.000 1480.000 2140.000 1550.000 580.000 120.000 1200.000 1030.000 1580.000 700.000 30. 500 17 . 0 0 0 11.000 0.150 2. 000 0.0 0.0 0. 030 0.610 1. 2 0 0 170.000 11.000 1.400 0.0 1800.000 3.000 1240.OCO 30.000 0. 800 315.000 339.000 20.000 0.500 11.OCO 2.200 666.OCO 78.000 0 . 160  3 1 398.000 25.000 320.000 860.000 1560.000 2260 .000 1700.000 600.000 130.000 1300.000 1110.000 1670.000 760.000 32.200 18.000 12.000 0.120 2.100 0.0 0.0 0.030 0.460 0 . 100 80.000 6.000 1.000 0.0 500.000 2.000 1310.000 30.000 1.300 750.000 478.000 37.000 1.000 11.000 0.900 700.000 82.000 0.130  4 1  5 1  374.000 106.000 13.600 8.000 150.000 70.000 640.000 350.000 460.000 790.000 800.000 1470.000 ' 1140.000 630.000 380.000 200.000 120.000 80.000 760.000 500.000 730.000 360.000 780.000 470.000 440 .000 250.000 4.200 37.700 2.000 21.OOC 1 .000 13.000 0.015 0 . 120 2.900 2.100 0.0 0.0 0.0 0.0 0 . 030 0.020 0.240 0.250 0 . 100 0 . 100 40.000 60.000 27.000 16.000 1.000 0 . 220 0.0 0.0 200.000 300.000 2.000 1.000 170.000 1540.000 4.000 30.000 0.100 1.000 62.000 94.000 L52.000 95.000 8.000 5.000 0.300 0.200 6. 000 4.000 1.400 0. 800 229.000 250.000 85.000 74.000 0.020 0.040  See Appendix A of this text for definition of item cluster and the items associated with the attribute group code and item code numbers. Values of nutrient composition obtained from Mini List Foods, Appendix B, Pennington, J . A . , Dietary Nutrient Guide, The AVI Publishing Co., Westport, Connecticut, 1976.  24Q  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) (MG) ISO LEU (MG) LYS (MG ) MET (MG) CYS ( MG) PHE (MG) TYR ( MG) VAL (MG) HI S (MG) FAT-T (GM) SFA (GM) PUFA (GM) (GM ) CHOLE CHO-T (GM) SUCR (GM) CHO-F (GM) THIA (MG) RIBC (MG) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) (MG ) VIT-C PANTO (UG) B I O T I N (MG) VIT-A (I U) VIT-D ( IU) VIT-E ( MG) CA (MG) P ( MG) MG (MG) FE (MG) I (MG) ZN (MG) NA (MG) K (MG) (MG) CU  6 2 211.000 3 . OOC 40.OCC 140.000 1 9 0 . OCC 300.OCC 230.CCC 70.000 30.OCC 150 . 0 0 0 150.OCC 210.OCC 80.000 20.6CC 11.000 8 . OOC 0 .070 4 . 30C O.C 0 .0 0.030 0.150 0.100 30 . 0 0 0 20.000 0.250 1.000 300.OOC 4.00C 840.000 15.000 0.70C 102.000 80.OCC 10.OOC 0.0 6.OOC 0 .300 43.OCC 122.OOC 0.17C  7 2 65.000 3. 500 50.000 160.000 230.000 350.000 280.000 80.000 30.000 170.000 180.000 240.000 90.000 3. 500 2.000 1.000 0.014 4.900 0. 0 0.0 0. 030 0 . 170 0 . 100 40.000 9.000 0.400 1.000 300.000 4.000 140.000 41.000 0. 100 118.000 93.000 13.000 0.0 7. 000 0.400 50.000 144.000 0. 150  8 2 36.000 3 .600 50.000 160.000 190.000 360.000 280.000 90.000 30 . 0 0 0 170.000 180.000 250.000 100.000 0 . 100 0.0 0.0 0.002 5.100 0.0 0 .0 0.040 0.180 0 . 100 40.000 9.000 0.400 I .000 400.000 2.000 0.0 41.000 0.0 121.000 95 . 0 0 0 15.000 0.0 7.000 0.400 52.000 145.000 0.020  9 2 59.000 4.200 60.000 190.000 270.000 420.000 330.000 100.000 40 . 0 0 0 200.000 200.000 290.000 110.000 2 .000 1.000 1 .000 0.002 6.000 0. 0 0.0 0.040 0.210 0.100 40.000 9 .000 0.400 1 .000 400.COO 3.COO 80.000 41.000 0 .100 143.000 112 . 0 0 0 17.000 0 . 100 80.000 0.400 61 . 0 0 0 175.000 0 .020  10 3 163.000 12.900 210.000 640.000 850.000 1100.000 820.000 400.000 300.000 740.000 550.000 950.000 310.000 11 . 5 0 0 4.000 6.000 0. 550 0.900 0.0 0.0 0.090 0.280 0.100 90.000 30.000 2.000 0.0 2000.000 23.000 1180.000 50.000 1.000 54.000 205.000 12.000 2 . 300 14.000 1.400 122 . 0 0 0 129.000 0.070  .241  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PR07 (GM) TRY (MG) THR (MG) (MG) ISO LEU (MG) (MG) LYS MET (MG) CYS (MG) PHE (MG) TYR (MG) (MG) VAL HIS (MG) FAT-T (GM ) (GM) SFA PUFA (GM) CHOLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM) THI A (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 (MG) (UG.) FOLIC VIT-B12(UG > VI T - C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D ( IU) (MG) VIT-E CA (MG) P (MG) MG (MG) FE (MG) I (MG) ZN (MG) NA (MG) K (MG) (MG) CU  11 3 173.000 11.200 180.OCC 550.000 7 4 0 . OCC 970.000 710.000 350.000 260.OOC 690.000 480.0CC 820.OOC 270.OCC 12.900 5.OOC 7.OOC 0 .380 2.40C 0.0 0.0 0.08C 0 .280 0.100 90.000 19.0C0 2 . OCC 0.0 2300.OCC 17.OOC 1C80.00C 50.000 l.OOC 80.000 189.OOC 12 . 0 0 0 1.70C 13.OOC 1 .20C 257.OCC 146.OOC 0.05C  12 4 386.000 7.900 60.000 290.000 320.000 1090.000 160.000 140.000 150.000 3 6 0 . 000 280.000 400.OCO 220.000 0.400 0.0 0.0 0.0 85.300 23.6C0 0.700 0.430 0.080 2 . 100 70.000 5.000 0. 0 0.0 200.000 1.000 0. 0 0. 0 0 . 100 17.000 45.000 14.000 1.400 14.000 0.400 1005.OCO 120.000 0.130  13 4 354.000 10.200 120.000 340.000 470.000 840.000 340.000 120.000 180.000 500.000 300.000 540.000 220.000 1.600 0.0 1 .000 0.0 80.500 20.100 1.600 0.640 0 . 140 4.900 290.000 18.000 0.0 0 .0 500.000 1.000 0.0 0.0 0.500 41.000 309.000 96.000 4.400 14.000 2.400 1032.000 120.000 0.450  14 5 42.000 1 .300 20.000 20 . 0 0 0 70.000 100.000 20.000 20.000 20.000 60.000 50.000 20.000 30.000 0.100 0. 0 0.0 0. 0 8 .700 0.0 0.0 0.040 0.030 0.400 10.000 0.0 0. 0 0.0 100.000 16.000 0.0 0.0 0 .0 4.000 12.000 3 . 000 0 .300 1. 000 0.100 144.000 9.000 0 .030  15 5 55.000 2.000 30.000 70.000 130.000 150.000 70.000 30.000 40.000 110.000 70.000 120.000 40.000 1 .000 3.000 6.000 0.0 9.700 0.0 0.200 0.080 0.020 0.100 20.000 11.000 0.0 0.0 200.000 24.000 0.0 0.0 0.200 9.000 57.000 24.000 0.600 1.000 0.900 218.000 61.000 0.03 0  242  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) ( MG) THR (MG) ISO LEU (MG) (MG) LYS MET (MG) CYS (MG ) PHE (MG ) TYR (MG) (MG) VAL HI S (MG) FAT-T (GM) SFA (GM) (GM) PUFA (GM ) CHGLE CHO-T (GM) (GM ) SUCR CHO-F (GM) THIA (MG) RI BC (MG) N I A C I N (MG I V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C ( MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) CA (MG) P (MG) MG (MG) FE ( MG) (MG ) I ZN (MG) NA (MG) K (MG) (MG ) CU  16 5 75.000 2.200 30.000 70.OCC 140.OCC 160.OOC 40.OOC 30 . 0 0 0 40.OCC 10 . 0 0 0 90.000 130.OCC 500.000 0.4CC 0.0 0.0 0.0 16.900 0 .0 0.50C 0.C7C 0.03C 0.90C 90.000 7.OOC 0 .0 0.0 200.OCC 16.000 0.0 0 .0 0.300 8.000 76.OCC 31.000 0.70C 3.OOC 0.90C 0.0 84.OOC 0.28C  17 6 225.000 7.200 100.000 280.000 400.OCC 610.000 340.000 150.000 190.000 390.000 310.000 410.000 170.000 7.300 3.000 5. 0 0 0 0.070 32.400 0 . 100 0 . 100 0 . 150 0.240 0 . 800 40.000 8. 000 0.0 0.0 700.OCO 5.000 250.OCO 7.000 0.900 215.000 260.000 14.000 1. 200 6 . OCO 0.600 564.000 154.000 0.050  18 7 125.000 4.100 50.000 170.000 200.000 270.000 140.000 70.000 80.000 200.000 100.000 240.000 100 . 0 0 0 1.500 0.0 I.000 0.040 23 .300 0.100 0.100 0 . 140 0.080 1.200 20.000 2.000 0.0 0.0 200.000 2.000 70.000 1.000 0.400 10.000 59.000 23.000 0.900 1.000 0.600 2.000 44.000 0.010  19 7 111.000 3.400 40.000 130 . 0 0 0 160.000 210.000 100.000 50.000 70.000 180.000 110.000 180.000 80.000 0.400 0.0 0.0 0.0 23.000 0.100 0.100 0. 140 0.080 1 . 100 20.000 2.000 0.0 0 .0 100.000 0.0 0. 0 0.0 0.400 8.000 50.000 13.000 0.900 1.000 0 . 100 1 .000 61.000 0.050  20 8 119.000 2.500 20.000 100.000 120.000 220.000 100.000 50.000 20.000 130.000 100.000 180.000 30.000 0.600 0.0 0.0 0.0 25.500 0.300 0.300 0.090 0.02C 1 .400 170.000 7.000 0.0 0.0 400.000 12.000 0.0 0.0 0.200 12.000 73.000 29.000 0.500 2.000 0.300 282 .000 70.000 0.110  243  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KC AL PROT (GM) (MG) TRY THR (MG) (MG) ISO (MG) LEU (MG) LYS MET (MG) (MG) CYS PHE (MG) TYR ( MG) (MG) VAL (MG) HI S FAT-T (GM) SFA . (GM) PUFA (GM ) (GM ) CHCLE (GM ) CHO-T (GM ) SUCR CHO-F (GM) THIA (MG) (MG) RI BO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) CA (MG) P ( MG) MG (MG) FE (MG) I (MG ) ZN (MG) (MG ) NA K (MG) (MG) CU  21 8 109.OOC 2.OCC 20.000 80.OCC 90.OCC 170.OCC 80.OCC 40.000 20.000 100 . 0 0 0 SO.000 140.OCC 30.OCC 0 .IOC 0.0 0 .0 0.0 24.200 0.100 0 . ICO 0 . 11C 0.01C l.OCC 40.OCC 1 .000 0 .0 0.0 200.OCC 5.000 0.0 0.0 O.iOC 10.000 28.OCC 6.000 0.9CC l.OCC 0.20C 374.OOC 28 . 0 0 0 0.05C  22 8 50.000 1. 1 0 0 10.000 40.000 30.OCO 140.000 30.OCO 20.000 10.000 50.000 70.000 60.000 20.000 0.2C0 0.0 0. 0 0.0 10.700 0. 200 0 . 100 0 . 06C 0.040 0. 500 30.000 1.000 0.0 0.0 100.000 7.000 60.000 0.0 0 . 100 1.000 14.000 8.000 0.400 2 . OCO 0.200 0. 0 16.000 0.030  23 8 363.OOC 26.600 270.000 1410.000 1250.000 1810 .000 1620.000 430.000 310.000 1000.000 940.000 1440.000 730.000 10.900 2.000 8.000 0.0 46.700 1.000 2 .500 2.010 0.680 4.200 920 .000 305.000 0.0 0.0 2200.000 20.000 0.0 0.0 13.500 72.000 1118.000 323.000 9 .400 2.000 13.200 3.000 827.000 2.910  24 9 290.000 9.100 110.000 260.000 410.000 700.COO 210.000 120.000 200.000 500.000 260 .000 390.000 340.000 3. 000 1.000 1 .000 0.004 55.400 1-000 0 .200 0.280 0.220 2.500 50.000 9.000 0.0 0.0 400.000 1 .000 0. 0  o.c 0.100 43.000 85.000 22.000 2 .200 9.000 1.4C0 580.000 90.000 0.230  25 9 262.000 6.600 80.000 130.000 280.000 440.000 170.000 100.000 150.000 360.000 190.000 300.000 250.000 2.800 1.000 1.000 0.004 53 . 6 0 0 4 . 100 0.900 0.050 0.090 0.700 40.000 17.000 0.0 0.0 400.000 1.000 0.0 0.0 0.100 71.000 87.000 24.000 1.300 9.000 1.200 365.000 233.000 0.230  -244  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT TRY THR ISO LEU LYS MET CYS PHE TYR VAL HIS FAT-T SFA PUFA CHOLE CHO-T SUCR CHO-F TH I £ RIBO NIACIN VIT-B6 FOLIC VI T - E 1 VIT-C PANTO BIOTIN VIT-A VTT-D VIT-E CA P MG FE I ZN NA K CU  (GM) (MG) (MG) (MG) (MG) (MG) (MG) (MG) (MG) (MG) (MG) (MG ) (GM ) (GM) (GM) (GM) (GM) (GM) (GM ) (MG) (MG) (MG) (MG) (UG) 2(UGI (MG) (UG) (MG) (IU) ( IU) (MG) (MG) (MG) (MG) (MG) (MG ) (MG) (MG) ( MG) (MG)  26 9  243.OCC 9 . IOC 100.OCC 290.CCC 390.000 620.000 290.000 140.OCC 2CO.0CG 500 .OOC 260.CCC 480.OCC 350.COG 1 .100 0.0 1.000 0.0C4 5 2 . ICC 1 .000 0.4CC 0.180 0.07C 1.400 ICO.OOC 38.CCC 0 .0 O.C 5C0.0C0 l.OCC 0*0 0.0 0.300 75.OCC 147.OCC 42.OOC 1.6CC 9.000 1.60C 557.000 145.OCC 0.22C  27 9  28 9  270.000 8. 7CO 100.000 270.OCC 420.000 690.000 260.000 130.000 2 CO.OOC 480.000 250.000 400.000 330.000 3.200 1.000 2.000 0.005 5 0 . 5 CO 1.000 0. 200 0.250 0. 210 2.400 40.000 17.000 0.0 0. 0  243 .000 10.500 140.000 320.000 480.000 760.000 290.000 160.000  400.000 1.000 0.0 15.000 0. 100 84.000 97.OCC 26.000 2. 500 9.000 1.300 507.000 105.000 0. 23 C  800 .000 2.000 0.0 8 .000 0.400 99.000  24 57 40 50 22  0.000 0.000 0.000 0.000 0.000 3.000 1 .000 1.000 0.005 47.700 1.000 1.6C0 0. 260 0.120 2.800 180.000 38.000 0 .0 0.0  228.000 45.000 2.300 1 1 . OCC 2.800 527.000 273 .000 0.220  29 9  207.000 7.400 70.OCO 270.000 350.000 660.000 360.000 130.000 70.000 400.000 330.000 380.000 150.000 7.200 2 .000 4. 000 0.006 29.100 l.OOC 0.500 0 . 130 0.190 0.60C 110 .000 4.000 O.C 1 .000 300.000 1.000 150.000 5 .000 0.200 120.000 211.000 15.000 1 .100 5.000 0.700 628.000 157.000 0. 080  30 9  210.000 5.000 30.000 200.000 290.000 810.000 130.000 70.000 50.000 210.000 90.000 260.OOC 70.000 2.000 1.000 1.000 0.0 45.000 0.700 1 .000 0. 13C 0.050 1.000 70.000 1.000 0.0 0.0 100.000 2.000 20.000 0.0 0. 100 200.000 140.000 107.000 3 .000 6.OOC 0.100 110.000 16.000 0.190  .245  ITEM CLUSTER ATTRIBUTE GROUP NUTRIENTS: KC A L PROT TRY THR  31  33  32 10  10  34 10  10  325.000 7.100  270.000 8.700  294.000 7.800  290.000 9.100  100.000 270.000 420.OCO 690.000 260.000  100.000 280.000  110.000 260.000  420.0 650.0 330.0 140.0  CYS PHE TYR  (MG) (MG) (MG) (MG) (MG ) (MG) ( MG)  90.000 210.OOC 330.OCC 550.OCC 160•OOC 90.OOC 140.OCC 390.000 230.OCO  VAL HIS FAT-T SFA  (MG)  310.000  2 4 2 4  410.000 700.COO 2 10.000 120.000 200.000  (MG) (GM) (GM)  130.OOC 9.300 2.OCC  ISO LEU LYS MET  PUFA C HOLE CHG-T  (GM) (MG) ( MG)  (GM) (GM ) (GM)  SUCR CHO-F  (GM) (GM) THIA (MG) (MG) RI BO ( MG) NIACIN VIT-B6 (MG) (UG) FOLIC VIT-B12(UG) (MG) VIT-C PANTO (UG) BIOTIN VIT-A VIT-D VIT-E CA P  (MG) (IU ) ( IU ) (MG) (MG ) ( MG)  MG FE  (MG ) (MG)  I ZN  (MG) (MG)  NA K  (MG) (MG)  CU  (MG)  7.OOC 0.006 52.3C0 1 0 0 0 2  .000 .200 .27C .250 . OCC  40.000 8.OOC 0 .0 0.0 400.OOC 1 .000 0.0 0.0 0.200 68  .000  232.OCC 24.0GC 2.3C0 9.OCC 1 973. 116. 0.  .200000 000 310  130.000 0 8 5 0  0 0 0 0  . . . .  0 0 0 0  0 0 0 0  0 0 0 0  330.000 3. 200 1.000 2 . OCO 0.005 50.500 1.000 0. 200 C.250 0.210 2. 400 40.000 17.000 0.0 0.0 400.OCO 1.000 0. 0 15.000 0. 100 84.000 97.000 26.000 2. 500 9 . OOC 1.300 507.COO 105.000 0.230  0 0 0 0  0 0 0 0  170.000 430.000 290 .000 40.0 150.0 10.1 2.0 7.0 0.0 42.3  0 0 0 0 0 0 0  500.000 260.000  0 0 0 0 0 6 0  390.000 340.000 3. 000 1.000 I . 0 0 0 0.004 55.400  1.000 0.100 0.170  1. 000 0 .200 0.280 0.220 2.500  0.230 1 .400 50.000 8.000 0.160 0.0 500.000 1.000 100.000 8.000 0.200 104.000 151 . 0 0 0 24.000 1 .600 9.000 1.200 441.000 125.OOC 0.220  50.000 9 .000 0. C 0.0 400.000 1 .000 0.0 0.0 0.100 43.000 85.000 22.000 2 .200 9. 000 1.400 580.000 90.000 0.230  35 10  243.000 10.500 140.000 320.000 480.000 760.000 290.000 160.000 240.000 570.OOC 400.000 500.000 220.000 3.000 1.000 1 .000 0.005 4 7 . 7 0 0 1.000 1.600 0.260 0. 120 2.800 180.000 3 8 . 0 0 0 0.0 0.0 800.000 2.000 0.0 8.00 0 0.400 99.000 228.000 4 5 . 0 0 0 2. 300 11.000 2. 800 527.000 273.OOC 0.220  .246  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR ( MG) ISO (MG) (MG) LEU (MG) LYS MET (MG) CYS (MG) PHE (MG) TYR ( MG) VAL (MG) HIS (MG) FAT-T (GM ) SFA (GM) PUFA (GM) (GM ) CHOLE CHO-T (GM) (GM) SUCR CHO-F (GM) THI A (MG) RIBC (MG) N I A C I N (MG ) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT^D (IU) VIT-E (MG) CA (MG) P ( MG) (MG) -MG FE (MG) (MG) I ZN (MG) NA (MG ) K (MG) (MG) CU  36 11 384.000 8.OCC 100 . 0 0 0 230.OCC 370.OCC 620.OOC 180.OOC 100.000 160.OOC 440.000 270.OCO 3 4 0.CCC 160.OCC 9.4CC 2.OOC 7.00C 0 .0 73.300 24.100 1.1CC 0 .040 0.21C 1. 500 70.000 24.000 0 .0 0.0 500.OCC 1.000 0.0 0 .0 0.100 40.000 149.OOC 36.CCC 1.50C 2.OOC 0.800 670.OOC 384 .000 0.180  37 11 433.000 9.000 110.000 260.000 420.OCC 690.000 210.000 120.000 180.000 490.000 310.000 390.000 180.000 12.000 3.000 8. 000 0.0 71.500 0 . 200 0.400 0. 010 0.040 1.000 70.000 24.000 0.0 0.0 500.OCO 1 .000 0. 0 0.0 0 . 100 21.000 90.000 25.000 1.200 2 . OCO 0.700 1100.000 120.000 0.170  38 11 386.000 12.700 80.000 510.000 590.000 1670.000 370.000 240.000 130.000 1000.000 620.000 660.000 330.000 5.000 1.000 4.000 0.0 76.700 1.000 2.200 0.420 0 . 120 2.200 200.OOC 0.0 0.0 0.0 400.000 1.000 0.0 0.0 4.400 11.000 281 . 0 0 0 173.000 2 .700 14.000 2.000 3.000 200.000 0.310  39 11 568.000 5.300 60.000 210.000 230.000 270.000 280.000 70.000 50.000 250.000 100.000 290.000 80.000 39.800 10.000 28.000 0.0 50 . 0 0 0 0.200 1 .600 0.210 0.070 4.800 180.000 9.000 0. 0 16.000 500.000 7.000 0.0 O.C 4.300 40.000 139.000 43.000 1 .800 13.000 2.500 1000.000 1130.000 0.220  40 11 354.000 10.200 120.000 340.000 470.000 840.000 340.000 120.000 180.000 500.000 300.000 540.000 220.000 1.600 0.0 1 .000 0.0 80.500 20.100 1.600 0.640 0.140 4.900 290.000 18.000 0.0 0.0 500.000 1.000 0.0 0.0 0.500 41.000 309.000 96.000 4.400 14.000 2.400 1032.000 120.000 0.450  247  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR ( MG) ISO (MG) LEU (MG) (MG) LYS MET (MG) CYS (MG) PHE (MG ) TYR ( MG) (MG) VAL HI S (MG) FAT-T (GM) SFA (GM) PUFA (GM) (GM) CHOLE CHO-T (GM) (GM) SUCR CHO-F (GM) THIA (MG) (MG) RI BC N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) (MG) VI T-C PANTO (UG) B I O T I N (MG) VIT-A (I U) VIT-D (IU) (MG) VIT-E CA (MG ) P (MG) MG (MG) FE (MG) I (MG) ZN (MG) NA (MG) K (MG) ( MG) CU  41 12 377.000 24.2CC 300 .000 1100.OCC 1300.OCC 2100.OOC 2200.OCC 640.OCO 320.OCC 990 .000 9C0.00C 1500.CCC 890.OOC 3 0 . 300 14.000 14.000 0.070 0.0 0 .0 O.C 0.040 0.190 3.80G 350.000 3.OOC 1.800 0.0 4CO.0CC 3.OCC 60.OCC 0.0 0.200 10.000 121.CCC 23.OCC 3.ICO 6.000 2.500 60.OOC 370.000 0.08C  42 12 286.000 24.200 250.000 960.000 1130.OOC 1770.000 1890.000 540.000 270.000 1000.000 740.000 1190.000 750.000 20.300 10.000 9.000 0.070 0. 0 0.0 0.0 0. 090 0.210 5.400 460.000 5.000 0.900 0.0 400.OCO 3.000 40.OOC 0.0 0.400 11.000 194.000 21.OCO 3.200 6 . OCO 4.300 47.000 450.000 0.080  43 12 261.000 28 . 6 0 0 350.000 1300.000 1500.000 2500.000 2600.000 760.000 380.000 1170.000 1060.000 1800.000 1050.000 15.400 7.000 7.000 0.07C 0.0 0.0 0.0 0.080 0.220 5.600 410.000 4.000 2.650 0.0 500.000 3.000 30.000 0.0 0.200 12.000 250.000 28.000 3.500 6.000 3.000 60.000 370.000 0.080  44 12 216.000 25.300 290.000 1100.000 1300.000 2100.000 2200.000 630.000 320.000 1040.000 900.000 1400.000 8 70.000 12.000 6.COO 5.000 0.070 0 .0 0. 0 0 .0 0.020 0.240 3.400 100.000 3 .000 1. 840 0.0 600.000 3.OOC 20.000 O.C 0 . 100 20.000 106.000 27.000 4.300 6.000 3.100 1010.000 87.000 0 .210  45 12 266.000 25.800 330.000 1180.000 1330.000 1990.000 2060.000 630.000 340.000 1000.000 890.000 1260.000 720.000 17.300 10.000 5 .000 0.070 0.0 0.0 0.0 0.150 0.270 5.600 320.000 3.000 3.100 0.0 600.000 6.000 0.0 0.0 0.200 11.000 212.000 22.000 1.800 3.000 5.400 70.000 290.000 0.240  248  ITEM CLUSTER A T T R I E U T E GROUP NUTF-IENTS: KCAL PR07 (GM) (MG) TRY THR (MG) ISO (MG) (MG) LEU (MG) LYS MET (MG) CYS ( MG) PHE (MG) TYR (MG ) V AL (MG) HI S (MG) FAT-T (GM ) S FA (GM ) PUFA (GM) CHOLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM) (MG) THIA (MG) RIBO N I A C I N (MG ) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (I U) VIT-D (IU) VIT-E ( MG) CA (MG) P (MG) MG (MG ) FE (MG) (MG ) I ZN (MG) NA (MG) K (MG) ( MG) CU  46 12 373.000 22.600 3 GO.OCC 1 0 5 0 .OCO 1160.OCC 1650.000 1850.OCO 570.000 210.OCC 890.000 810.000 1190.CCC 780.OCC 30.600 12.000 15.000 0 .070 0.0 0.0 0 .0 0.500 0 .230 4.900 320 .000 3.OCC 0.500 0.0 500.OCC 5 .000 0.0 0 .0 0.20C 10.CCC 232.000 23.CCC 2 .900 10.OOC 2.700 65.OOC 3^0.CCC 0.09C  47 12 193.000 18.300 180.000 760.000 920.000 1430.000 1550.000 450.000 300.000 700.000 720.000 960.000 600.000 12.300 4.000 6.000 0.070 0.900 0. C 0.0 0. 530 0 . 190 3. 800 360.000 3.000 0. 500 0.0 400.OCO 5.000 0. 0 0.0 0. 200 11.000 156.000 15.OCO 2.700 8.000 1.900 1100.000 340.000 0.440  48 12 611.000 30.400 320.000 1000.000 1300.000 2400.000 2000.000 470.000 360 .000 1400.000 780.000 1400.000 820.000 52.000 17.00 0 30.000 0.C80 3 .200 3.200 0.0 0.510 0 . 340 5.200 100.000 2 .000 0.700 0.0 1300.000 8.000 0.0 0.0 0 .300 14.000 224.000 25.000 3.300 27.000 5.100 1021 .000 236.000 0.520  49 13 198.000 21.700 250.000 890.COO 800.000 1100.000 1300.000 400.000 210.000 600.000 540.000 750.000 440.000 11.700 4.000 6.000 0 . 070 0.0 0. 0 0.0 0.040 0 . 120 4.400 300.000 2 .000 0. 790 4.000 900.000 11.000 230.000 0.0 0.100 21.000 247.000 18.000 1.500 5.000 4.900 400.000 138.000 0 .230  50 13 161.000 15.300 230.000 790.000 1000.000 1340.000 1630.000 480.000 250.000 490.000 650.000 920.000 530.OOC 9.300 3.000 6.000 0.040 3 . 200 0.0 0.0 0.020 0.07C 2 .400 250.000 4.000 0.260 0.0 900.000 6.000 70.000 0.0 0 . 200 6. 000 113.000 11.000 0.900 6.000 2.400 154.000 140.000 0 . 180  249  I T E M CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PPOT (GM) TRY (MG ) THR (MG) (MG) ISO LEU (MG) ( MG) LYS MET (MG) CYS (MG) PHE (MG • TYR (MG) (MG) VAL HIS (MG) FAT-T (GM) SFA (GM) PUFA (GM) (GM) CHOLE CHO-T (GM) (GM • SUCR CHO-F (GM ) THIA (MG) RIBO (MG ) N I A C I N (MG) V I T - B 6 (MG) (UG ) FOLIC VIT-B12(UG> (MG) VIT-C (UG) PANTO B I O T I N (MG ) VIT-A ( IU) VIT-D (IU) VIT-E (MG) CA (MG ) P (MG) MG (MG ) FE (MG) I (MG ) ZN (MG) NA (MG) K (MG) CU (MG)  51 13 250.OCC 30 . 6 0 0 370.OCC 1290.000 1600.000 2200 .000 27C0.CCC 8 CO.000 410.000 1200.OCC 1080.OCC 1500.OCC 880.000 11.900 3.000 8.OCO 0.C7C 2 .800 0.0 0 .0 0.060 0.360 9.200 4C0.0CC 6.OCC 0.42C 0.0 9C0.C0C 11.000 170.OCC 0.0 0.200 12.000 243.OCC 19.OCC 1.800 7.000 4.600 88.OOC 428.OCC 0.33C  52 13 263.000 27.000 330.OCO 1140.000 1420.000 2050.000 2450.000 750.000 370.000 1080.OCO 1080.000 1320.000 730.000 16.400 4.000 10.000 0. 075 0.0 0. 0 0.0 0 . 110 0. 200 11.400 400.000 10.000 0.420 2.000 9 0 0 . 000 11.000 170.000 0.0 0.300 11.000 300.000 25.OCO 2.100 6. 000 2 . 800 93.000 443.CCO 0 . 180  53 14 98.000 15.800 160.000 680.000 800.000 1200.000 840.000 460.000 210.000 580.000 430.000 480.000 30.000 2.500 1.000 I .000 0.080 1.900 0.0 0.0 0.010 0.110 1 .100 80.000 3 .000 19.100 11.000 300.000 20.000 110.000 3.000 0.300 55.000 184.000 113.000 4.100 90.000 1.60C 1010.000 140.000 0.0  54 14 176.000 16.600 170.000 720.000 840.000 1250.000 1460 .000 480.COO 220.000 620.000 450.OCO 880.000 760.000 8.900 2.000 6 .000 0. 070 6.500 0.0 0.0 0.040 0.070 I .600 50.000 16.000 1. 0 0 0 2.000 300.000 3.000 0.0 0. 0 0.600 11.000 167.000 18.000 0.400 34.000 0.300 180.000 390.000 0.140  55 14 165.000 19.600 200.000 840.000 1000.000 1490.000 1730.000 570.000 260.000 730.000 530.COO 1040.000 900.000 6.400 2.000 4.000 0. 060 5 . 80C 0.0 0.0 0.040 0.07 0 3.200 140.000 16.000 1.000 2.000 300.000 3.000 320.000 0.0 0.600 40.000 247.000 36.000 1.200 62.000 0.300 177.000 348.000 0 . 150  250  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) (MG) ISO LEU (MG) LYS (MG) MET (MGI CYS (MG) PHE (MG ) TYR (MG) (MG) VAL HIS (MG) FAT-T (GMl (GM) SFA PUFA ( GM) (GM) CHOLE (GM) CHG-T (GM) SUCR CHO-F (GM) THIA ( MG) (MG ) RIBO N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) (MG ) VIT-C PANTO (UG) B I O T I N (MG) VIT-A ( I U) VIT-D ( IU ) VIT-E (MG) (MG) CA P (MG) MG (MG) (MG) FE I ( MG) (MG) ZN NA (MG) K (MG) (MG) CU  56 14 171.OCC 25 . 2 0 0 2 5 0 . CCO 1080 .000 12 9 0.OOC 1920.000 2220.OOC 730.OCC 340.OOC 940.OCC 690.000 1330.OCC 1160.000 7 . OOC 3 .CCC 3.000 0.06C 0.0 0.0 0 .0 0.05C 0.07C 8.3CC 340.OCC 16.000 l.COC 4.000 3 CO.OCC 8 .000 680.000 0.0 0 .60C 1 6 . OOC 248.OOG 23.OOC 0.800 46.OCC 1.000 134.OCC 525.OCC 0 .190  57 14 76.CCO 8.500 90.000 370.000 430.000 640.000 280.000 250.000 110.000 320.000 230.000 450.000 960.000 2. 200 1. 0 0 0 1.000 0.230 4.900 0.0 0 . 100 0 . 020 0. 200 0.800 40.OCO 3.000 18.OCO 26.000 200.000 9.000 260.000 10.CCO 0.300 28.000 124.000 17.000 5.600 48.000 5 2 . CCC 62.000 70.OCO 3.430  58 14 239.000 8.600 90.000 370.000 440.000 650.000 280.000 250.000 120.000 320.CCO 230.000 460.000 960.000 13 . 9 0 0 4.000 8.000 0.230 18.600 0.0 O.C 0.170 0.290 3 .200 40.000 3.000 18.000 39.OCO 200.000 9.000 440.000 5.000 0.600 152.000 241.000 17.000 8.100 69.000 52.000 206.000 203.000 4.270  5S 14 182.000 27.000 270.000 1160.000 1350.000 2030.000 2350 .000 780.000 380.000 1000.000 730.000 1430.000 700.000 7 .400 2.000 3 .000 0. 060 0.0 0.0 0. 0 0.160 0. 060 9.800 300.000 7.000 1 .000 5.000 500.000 12.000 160.000 400.000 1.400 80.000 414.000 41.000 1. 2 0 0 37.000 1. 700 116.000 443.000 0 . 800  60 14 210.000 19.600 200.000 840.000 980.000 1470.000 1710.000 570.000 280.000 730.000 530.000 1040.000 510.000 14.000 4.000 4.000 0.060 0.0 0.0 0.0 0.030 0.140 7 . 300 300.000 7.OOC 6.900 9.000 600.000 12.000 230.000 370.000 0.500 154.000 289.000 27.000 0.900 51.000 0.700 407.OOC 366.000 0.290  .251  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KC AL (GM) PROT TRY ( MG) THR (MG) (MG) I SO LEU (MG ) LYS (MG) MET (MG) CYS (MG) PHE (MG ) TYR (MG) ( MG) VAL (MG) HIS FAT-T (GM) SFA (GM) PUFA (GM) CHOLE (GM ) CHO-T (GM ) (GM) SUCR CHO-F (GM) THIA ( MG) RIBO (MG ) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG ) (MG) VIT-C PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) CA (MG) P (MG) MG (MG) FE (MG) (MG) I (MG) ZN NA (MG) K (MG) (MG ) CU  61 14 176.CCC 21.600 220.OCC 930.000 1040.OOC 1620.OCC 1890.OCC 6 3 0 . OCC 310.OOC 8 CO.OOC 580 .000 1150.OCC 5*0.OCC 9.30C 3 . OCC 3.OCC 0.06C 0.0 0.0 0 .0 0.21C 0.08C 12.70C 7 C O . COG 7 .000 7.OCC 5 .OOC 7C0.C0C 12.OCO 190.OOC 4CO.0CC 1 .400 14.OCC 245.000 33.CCC 1.4CC 37.OOC 1.3CC 134.000 512.OCC 1 .300  62 14 203.000 24.000 210.000 920.000 1070.000 1590.OCO 1850.000 610.OCO 280.000 880.000 570.000 1120.000 990.OCO 11.100 5. OCO 5.000 0.070 0.0 0.0 0.0 0.030 0. 200 5.400 180.OCO 32.000 10.000 0. 0 900.000 24.OCC 220.000 5C0.000 0.600 437.000 499.000 39.000 2 . 9C0 37.000 2.9C0 823.000 590.000 0.040  63 14 94.000 21.000 240.000 1040.000 1230.000 1840.000 2130.000 700.000 320.000 850.000 650.000 1280.000 530.000 0.800 0.0 0.0 0 . 140 0.500 0.0 0.100 0.020 0.030 3.300 50.000 2.000 0.690 11.000 200.000 10.000 40.000 150.OCO 0.500 78.000 208.000 42.000 1.700 65.000 1.500 126.000 203.000 0.57C  64 14 223 .000 12.300 150.000 560.000 690.000 1050.000 1030.000 360.000 180.000 460.000 360.000 450 .000 270.000 11.OCC 7. 000 . 0.0 0.140 18.600 0 .0 0. 100 0.030 0 . 030 2.000 60.000 2.000 0.720 7. 000 300 . 0 0 0 10.000 30.000 150.000 0.60C 38.000 111.000 61.000 1 . COO 66.000 1 .000 213.000 197.000 0.37C  65 14 197.000 28.800 290.000 1240.000 1470.000 2160.000 2 5 3 0 . COO 840.000 390.000 1000.000 780.000 1530.000 1550.000 8 . 200 3.000 4.000 0.060 0.0 0.0 0.0 0.050 0 . 120 11.900 430.000 1.000 2.200 10.000 300.000 3.000 80.000 250.000 0 . 500 8.000 234.000 27.000 1.900 16.000 0.400 662.000 249.000 0 . 120  252 ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY ( MG) THR (MG) (MG) I SO (MG ) LEU (MG) LYS MET (MG) CYS (MG) ( MG) PHE (MG) TYR (MG) VAL (MG) HIS FAT-T (GM) SFA (GM) (GM) PUFA (GM) CHOLE CHO-T (GM) (GM) SUCR CHO-F (GM ) TH IA (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG> VIT-C (MG) PANTO (UG) B I O T I N (MG ) VIT-A (IU) VIT-0 (IU) VIT-E (MG) CA (MG) P (MG) MG (MG) FE (MG) (MG) I ZN (MG) NA (MG) K (MG) (MG ) CU  66 14  67 15  68 16  245.000 133.OCC 229.000 27.400 26.400 27.500 370.000 250.OCC 390.000 1240.CCC 1C60.0GC 1270.000 1260.OOC 1430.000 1360.000 1850.OCC 24CO.OC0 2450.000 2170.OCC 1950.000 2275.000 720.OOC 610.000 685.000 330 .000 350.000 320.000 1235.000 860.OCC 1300.000 980.000 1020.000 670.000 1745.000 1310.OOC 1690.000 1 3 3 0 . CCC 1240.000 1145.000 3.OCC 10.600 13.000 l.OCC 3 . OCO 5.000 1.000 6.000 7.000 0.060 0. 300 0.190 0.0 5.300 2.700 0.0 0.0 0.0 O.C 0. 0 0.0 0.02C 0.170 0.260 0.05C 4 . 190 2.210 6.600 16.500 1 1 . 100 900.000 670.000 540.000 3.000 294.000 149.000 3.OOC 41 . 3 3 0 80.000 7.OCC 14.000 27.000 500.OOC 7100.000 3800 .000 50.000 3 . OCC 96.OCO 50.OOC 5 3 4 0 0 . 0 0 0 2 6 7 1 5 . 0 0 0 2 50.000 50.000 25.000 0 .200 0.600 0.400 4 . OOC 11.000 12.000 177.OCC 476.000 363.000 29.OOC 25.000 22.000 6.200 8. 800 1,300 23.OOC 19.000 13.000 0.50G 5.000 7.000 37.OOC 184.000 122.000 181.OOC 375.000 380.000 0.50C 1.890 3.700  6S 17 304.000 12.400 120.000 510.000 620.OCC 970.000 1100.000 300.000 200.000 460.000 480.000 650 .000 400.000 27.20C 10.000 15.000 0.070 1.600 I .600 0. 0 0 . 150 0.200 2.500 110.000 4.000 I .300 0.0 400.000 5.000 0.0 0.0 0 . 100 5 .000 102.000 9 . COO 1. 500 8.OOC 1 .500 1060.000 212 .000 0. 080  70 17 476.000 18.100 160.000 720.000 880.000 1290.000 1460.OOC 380.000 230.000 600.000 590.000 920.000 490.000 44.200 16 . 0 0 0 23.000 0.070 0.0 0.0 0.0 0.790 0.340 3.700 190.000 4.000 1.400 0.0 600.000 5. 0 0 0 0.0 0.0 0.200 7.000 162.000 16.000 2.400 8.000 0.600 958.000 269.000 0.150  253 ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS : KCAL (GM) PROT TRY ( MG ) THR (MG) (MG) ISO LEU (MG) LYS (MG) MET (MG) (MG) CYS PHE (MG) TYR (MG) (MG) VAL (MG) HIS FAT-T (GM ) (GM) SFA (GM) PUFA CHOLE (GM) (GM ) CHO-T (GM) SUCR CHO-F (GM ) THI A (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 ( MG) (UG) FOLIC VIT-B12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG ) VIT-A (IU) VIT-D (ID VIT-E (MG) CA (MG) P (MG ) (MG) MG FE (MG) (MG) I (MG) ZN (MG) NA K (MG) (MG) CU  71 17 307.OCC 16.2CC 240.OCC 760.CCC 840.OOC 1470.OCC 1180.000 3 80.OCC 200 .000 820.OOC 6CO.OCC 1 0 1 0 .OCC 760.OCC 25.60C 10.000 14.000 0.346 1.800 0.0 0.0 0.2CC 1.3CC 5.700 930.OCC 6.OCC 2.36C O.C 5900.OOC 111.OOC 6350 .000 1 5 . OCC 0.700 9.000 238.OCC 23.OCC 5.4CC 23.000 7.5CC 291 . 0 0 0 232.CCC 3.05C  72 18 90.000 5.700 50.000 250.OCO 320.000 490.OCO 420.000 60.000 60.000 310.000 220.000 340.000 160.000 0.400 0. 0 0.0 0.0 16.400 0.700 0.900 0.050 0.040 0. 600 330.000 6.000 0.0 0. C 100.000 4.000 0. 0 0.0 0 . 100 29.000 109.000 27.000 1. 8 0 0 2.000 1.100 236.000 264.000 0 . 100  73 18 118.000 7.800 70.000 340.000 450.000 670.000 580.000 80.000 80.000 400.000 300.OOC 480 .000 220.000 0.600 0.0 0.0 0.0 21.200 0.700 1.5CC 0 . 140 0.070 0.700 140 .000 8.000 0 .0 0.0 200.000 6 . OCO 0.0 0 .0 0.200 50.000 148.000 37.000 2.700 3 . OCC 1 .500 7.000 416 .000 0.240  74 18 108.000 8. 100 80.000 320.000 390.000 600.000 530.000 120.000 90.000 450 .000 200.000 450.000 270.000 0.8C0 0.0 0.0 0.0 18.100 1 .300 1. 8 0 0 0 . 200 0.110 1.400 50 . 0 0 0 26.000 O.C 17.000 300.000 10.000 350.000 0.0 0.200 24.000 146.000 19.OOC 2.100 7.000 0.800 1. 000 379.000 0.280  75 18 118.000 9.800 100.000 370.000 490.OOC 680.000 620.000 140.000 100.000 530.000 380.000 540.000 330.000 5 . 100 1.000 4.000 0.0 10.100 3.400 1.400 0.31C 0.130 1.200 40.000 38.000 0.0 17.000 700.000 30.000 660.000 0.0 0.700 60.000 191.000 194.000 2 .500 4.000 1.100 2.000 487.000 0. 810  254  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS : KCAL PROT (GM ) TRY (MG) THR (MG) ISO (MG) LEU (MG) LYS (MG) MET ( MG) CYS (MG) PHE (MG) TYR (MG) VAL (MG) HIS (MG) FAT-T {GM ) SFA (GM) PUF* (GM) CHOLE (GM) CHO-T (GM * (GM ) SUCR CHO-F (GM ) TH IA (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 ( MG) FOLIO (UG) VIT-B12(UG ) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A ( IU) VIT-D (IU) VIT-E (MG) CA (MG) P (MG) MG (MG) FE (MG) (MG) I ZN (MG) NA (MG) K (MG) CU (MG )  76 19 598.OOC 18.6CC 180.OOC 6 10 .OCC 870.000 1450.OCC 580.000 260.OCC 380.CCC 1100.OOC 620.OCC 1120.OOC 5 2 0 . CCC 54.200 4 . OCC 47.OOC 0.0 19.5CC 2.300 2.600 0.240 0.92C 3.500 ICO.CCC 45.OCC 0.0 0.0 500.000 8.000 0 .0 0.0 15.CCC 234.000 5C4.C0C 269.000 4.7CC 2.000 1.50C 4 .CCC 773.OOC 0.68C  77 19 561.000 17.2CC 430.000 690.OCO 1140.000 1410.000 740.000 330.000 480.OCO 950.000 660.000 1480.000 390.OCO 45.700 8.000 35.000 0.0 29.300 3.000 1.4C0 0.430 0.250 1.800 400.000 2 5. CCC 0.0 0. 0 1300.000 30.000 100.000 0. 0 5. ICC 38.000 373.OCO 267.000 3. 800 3.000 1.000 15.000 464.000 0. 760  78 19 346.000 3.500 32.000 126.000 175.000 260.000 148.000 69.000 60.000 170.000 106.000 205.000 670.000 35.300 30.000 2.000 0.0 9.400 4.700 4.000 0.050 0.020 0 . 500 40.000 27.000 0.0 3.000 200.000 6.000 0.0 0.0 1.000 13.000 95.000 44.000 1.700 2.000 3.000 23.000 256.000 0.390  79 19 581.000 27.800 360.000 870.000 1300.000 2000.000 1100.000 280.000 480.000 1600.000 1200.000 1600.000 780.000 49.400 9.000 39.000 0 .0 17. 200 4 . 500 1 .900 0 . 130 0 . 130 15.700 330.000 13.000 0.0 0.0 1700.000 39.000 0. C 0.0 6. 700 63.000 407.000 174.000 2.000 12.000 2 .200 607.OOC 670.000 0. 570  80 19 568.000 26.300 340.000 820.000 1250.000 1850.000 1090.000 270.000 450.000 1500.000 1090.000 1520.000 740.000 48.400 10.000 34.000 0.0 17.600 4.500 1 .900 0.990 0.130 15.800 400.000 25.000 0.0 0.0 2800.000 34.000 0.0 0.0 6.500 59.000 409.000 168.000 2.000 11.000 2.100 5.000 674.000 0.690  255  ITEM C L U S T E R A T T R I E U T E GROUP NUTRIENTS: KCAL (GM ) PROT (MG) TRY THR (MG) (MG) ISO (MG) LEU (MG) LYS MET (MG) (MG) CYS PHE (MG) TYR (MG) (MG) VAL (MG) HIS FAT-T (GM) SFA (GM) (GM) PUFA ( GM) CHOLE (GM ) CHO-T (GN ) SUCR CHO-F (GM) THIA (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) (UG ) FOLIC VIT-E12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG) (IU) VIT-A VIT-D (IU) VIT-E (MG) CA (MG ) P (MG ) MG (MG) (MG) FE (MG) I (MG) ZN (MG) NA K ( MG) (MG) CU  81 19 651.OCC 14.8CC 170.OCC 580.000 760.000 1220.OCC 490.000 3C0.0CC 320.OCC 760.OOC 580.COC 950 .000 4C0.00C 64.000 4.0CC 50.CCC 0.0 15.8CC 3.OCC 2 . ICC 0 .330 0.13C 0 .900 730.OCC 58.OCC 0 .0 2.OCC 900.000 3 7.000 30 . 0 0 0 0.0 22.CCC 99.OOC 380.OCC 1.34.000 3 . ICC 3.OOC 2.8CC 2 . OOC 450.OCC 0.90C  82 20 93.000 2. 600 30.000 ICO.CCO 120.000 130.000 140.000 30.000 30.CCC 120.000 50.OCO 140.000 40.000 0. 100 0.0 0.0 0.0 21.100 0.100 0 . 6C0 0.100 0.040 1. 7 0 0 200.000 12.OCO 0.0 20.000 400.000 2. 000 0.0 0.0 0. C 9.000 65.COO 22.000 0.700 4.000 0.200 4. 000 503.000 0. 150  83 20 274.000 4.300 40.000 170.000 180.OCO 210.000 220.000 50.000 40.000 190.000 80.000 200.000 60.000 13.200 3.000 10.000 0 .020 36.CCO 0.200 1.000 0 . 130 0.080 3 . 100 180.000 9.000 0 .0 21.000 500.000 1 .000 0.0 0.0 0.300 15.000 111.000 17.000 1.300 11.000 0.200 6.000 853.000 0.270  84 2C 94.000 2 . 100 20.000 90.000 90.000 110.000 110.000 30.000 20.000 90.000 40.000 110.000 30.000 4.300 2.000 1. 000 0.015 12.300 0 . 100 0.400 0.C80 0.050 1.000 90.000 12.000 O.C 9.000 200.000 2.000 170.000 0.0 0.200 24.000 48.000 14.000 0.400 3.000 0.100 331.000 250.COO 0 . 100  85 20 141.000 2.100 40.OOC 100.000 100.000 120.000 100.000 40.000 30.000 120.000 100.000 160.000 40.000 0.500 0.0 0.0 0.0 32.500 7.20C 0.900 0.09C 0.070 0.700 170.000 19.000 0.0 22 . 0 0 0 700.000 2.000 8100.000 0.0 2.000 40.000 58.000 12.000 0.900 3.000 0.700 12.000 300.000 0.170  256  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS : KCAL PROT (GM) TRY (MG) THR (MG) I SO (MG) (MG) LEU LYS (MG) MET (MG) CYS (MG) PHE (MG) TYR (MG ) VAL (MG ) HIS (MG) FAT-T (GM) SFA (GM) PUF/S (GM) CHOLE (GM) CHO-T (GM) (GM ) SUCR CHO-F (GM) THIA (MG) RIBO ( MG) N I A C I N (MG) V I T - B 6 ( MG) (UG) FOLIC VIT-B12(UG) VIT-C (MG) PANTO (UG ) B I O T I N (MG) VIT-A (IU) VIT-D (IU) (MG ) VIT-E C/i (MG) P (MG ) MG (MG) FE (MG) I (MG) ZN (MG) NA (MG ) K (MG) (MG) CU  86 20 114.OCC l.CCC 20.000 50.OCC 50 . 0 0 0 60.0GC 50.OOC 20.OOC 20.OCC 60 .OOC 5 0 . OCC 80.000 20.OCC 0 .200 0.0 0.0 0 .0 27.5CC 14.9CC 0 . 6 CC 0.030 0.030 0 . 6 CC 70.OOC 19.OCC 0.0 8.CCC 400.OOC 2.000 5000.000 0.0 0.2CC 13.OCC 29.OCC 18.000 0.7CC 3.000 0.5CC 4 8 . CCC 120.OCO 0.06C  87 20 93.000 2. 600 30.000 100.000 120.000 130.000 140.000 30.000 30.OCC 120.000 50.CCO 140.000 40.000 0.100 0.0 0.0 0.0 21.100 0 . 100 0 . 600 0 . 100 0.040 1. 700 200.000 12.COO 0.0 20.OCO 400.000 2.000 0. 0 0.0 0. 0 9.000 65.CCO 22.000 0. 700 4.000 0.200 4 . OCO 503.000 0 . 150  88 21 167.000 2 . 100 14.000 1.000 1.000 I .000 74.000 12.000 1.000 1.000 I.000 1.000 1.000 16.400 3.000 9 . OOC 0.0 6.300 1.600 1.600 0.110 0.200 1.600 420.000 30.000 0.0 14.000 1100.000 6.000 290.000 0.0 1.500 10.000 42.000 37.000 0.600 2.000 2.400 4.000 604.000 0.390  89 21  90 21  25.000 1.600 20.000 60.000 70.000 90.000 80.000 20 . 0 0 0 20.000 60.000 30.000 80.000 30.000 0. 200 0.0 0. 0 O.C 5.400 0.400 1.000 0 . 07C 0.090 0. 500 60.000 5.000 0.0 12 . 0 0 0 200.OOC I .000 540.000 0.0 0 . 800 5O.C0C 37.000 22.000 0.600 3.000 0.300 4.000 151.000 0. 090  24.000 1.400 20.000 50.000 70.000 80.000 80.000 20.000 10.000 40.000 30.000 70.000 30.000 0.200 0.0 0.0 0.0 5.200 0.400 1 .000 0.030 0.050 0.300 40.000 12.000 0.0 4.000 100.000 1 .000 470.000 0.0 0.0 45.OOC 25.000 14.000 1.500 1.000 0 . 300 236.000 95.000 0.090  257  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM! TRY (MG) THR (MG) ISO (MG) (MG) LEU LYS (MG) MET (MG) CYS (MG) PHE (MG) TYR (MG) VAL (MG) HIS (MG) (GM) FAT-T SFA (GM) (GM ) PUFA (GM) CHCLE CHO-T (GM ) (GM) SUCR CHO-F (GM) THI A (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) (MG) VIT-C (UG) PANTO B I O T I N (MG) VIT-A (IU) VIT-D ( IU) (MG) VIT-E CA (MG) P (MG) (MG) MG FE (MG ) (MG ) I ZN (MG) NA (MG ) K (MG) (MG ) CU  91 21 26.OOC 3 . ICC 30 . 0 0 0 120.OCC 120 .000 150.OCC 140.CCC 50.OCC 50.OCC 100.OOC 110.OCC 160.000 180.OOC 0.30C 0.0 0.0 0.0 4.5CC 0.20C 1 .50C 0.090 0.20C 0.8CC 170.OOC 22.CCC 0 .0 90.OOC 500 .000 l.OCC 2500.000 0.0 1.9CC 88.OCC 62.OOC 21.000 0.8CC 4.000 0.2CC 10.CCC 267.OOC 0 . ICC  92 21  93 21  24.000 1.3 CO 10.000 40.000 50.000 50.000 60.000 10.000 30.OCO 70.000 30.OCO 40.000 20.000 0. 200 0.0 0. 0 0.0 5.400 0.300 0. 800 0.050 0.050 0.300 160.000 55.OCO 0.0 47.000 200.000 2.000 130.000 0.0 7. 8CC 49.000 29.OCO 15.000 0.400 3 . 000 0.300 20.000 233.000 0. 120  20.000 1.100 10.000 30.000 40.000 40.000 50.000 10.000 20.000 60.000 20.000 30.000 20.000 0.200 0.0 O.G 0.0 4.300 0.200 0.800 0.040 0.040 0.300 130.000 11.000 0.0 33.000 200.000 1 .000 130.000 0.0 7.600 44.000 20.000 12.000 0.300 2.000 0.200 14.000 163 . 0 0 0 0.C4C  94 21 33.000 3.600 50.000 160.000 150.000 290.000 140.000 40.000 60.000 120.000 140.000 210.000 80.000 0.700 0.0 0. 0 0.0 5 . 100 0.200 1 .000 0.110 0.200 1.200 200.000 24.000 0.0 76.000 500.000 1.000 78C0.000 0.0 5. 900 188.OOC 52.000 42.OCO 0.800 3 . 000 0.700 25.000 262.000 0.310  95 21 13.OOC 0.900 10.000 40.000 40.000 70.000 50.000 0.0 10.000 40.OCC 30.000 50.000 20.000 0.100 0.0 0.0 0.0 2.900 0.200 0.500 0.060 0.060 0.300 60.000 200.000 0.0 6.000 200.000 3.000 330.000 0.0 0.300 20.000 22 . 0 0 0 11.000 0.500 10.000 0.100 9.000 175.000 0.09 0  258  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS : KCAL PROT (GM ) TRY (MG) THR (MG) ISO (MG) ( MG) LEU LYS (MG) MET (MG) CYS (MG) PHE (MG) TYR (MG ) VAL (MG) HIS (MG) FAT-T (GM) SFA (GM) (GM) PUFA CHCLE (GM) CHO-T (GM ) (GM) SUCR CHO-F (GM) THIA (MG) RIBO (MG) N I A C I N (MG) V I T - E 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) (UG) PANTO B I Q T I N (MG) VIT-A (IU) VIT-D ( IU) VIT-E (MG ) CA (MG ) P (MG) KG (MG ) (MG) FE (MG) I ZN (MG) NA (MG ) K (MG) (MG) CU  96 21  97 21  98 21  99 21  100 21  23.OCC 2.2CC 40.000 60.OCC 70.000 60.OOC 110.OCC 20.OOC 30.OOC 70.OCC 120.000 100.OOC 4 0 . OCC 0.400 0.0 0.0 0.0 4.OOC 0 .300 0.9CC 0.080 0 . 14C 0.6CC 1 3 0 .OCC 8.CCC 0.0 48.OCC 200 .000 1.000 8CO.CCC 0.0 1 .7 C C 138.OOC 32.OOC 17.OOC I. 8 0 0 4.000 O.2C0 18.CCC 220.OOC 0.09C  88.000 4 . 7C0 40.000 180.000 220.000 300.000 220.000 40.000 50.000 180.000 110.OCC 190.000 80.000 0.400 0.0 0. C 0.0 16.800 6. 4C0 2 . 300 0. 090 0. 060 0 . 800 50.000 15.OCO 0.0 8. 000 200.000 2.000 690.000 0.0 0. 0 26.000 76.OCO 13.000 1.900 2 . 000 1.400 236.CCO 96.000 0. 170  68 . 0 0 0 5.100 40.000 80.000 240.000 320 .000 240.000 40.000 60.000 200.000 120.000 210.COO 80.000 0.300 0.0 0.0 0.0 11.800 4.500 1.900 0.27C 0.090 1.700 130 . 0 0 0 15.000 0.0 13.000 300.000 2 .000 600.000 0.0 0.300 19.000 86.000 21.CCC 1.900 3.000 0.900 115.000 135.000 0.210  18.000 1.000 10.000 40.000 40.000 40.000 40.000 10.000 20.000 50.OCC 40.000 30.000 10.000 0.20C 0.0 0. 0 0.0 3.800 0 . IOC 1.400 0. 060 0.070 0. 500 210.000 2 .000 0.0 96.000 200.000 I .000 420.000 0.0 0. 500 9.000 16.000 12.000 0.500 9. 000 0.100 9.000 149.000 0.070  24.OCO 2.700 40.000 120.000 130.000 210.000 170.000 50.000 50.000 130.000 90.000 150.000 60.000 0.600 0.0 0.0 0.0 3.600 0.300 0.900 0.020 0 . 120 0.300 70.000 29.000 0.0 14.000 100.000 2.000 8000.000 0.0 0.100 118.OOC 26.000 43.000 2.600 14.000 0.500 236.000 250.000 0.100  259  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) (MG) LEU LYS (MG) MET (MG) (MG ) CYS PHE (MG) TYR (MG) VAL (MG ) ( MG) HIS FAT-T (GM) SFA (GM) PUFA (GM) CHOLE (GM) CHO-T (GM ) (GM ) SUCR CHO-F (GM) THIA (MG) RIBO (MG) N I A C I N (MG) V I T - 8 6 (MG) FOLIC (UG ) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-C (IU) VIT-E (MG) CA (MG ) P (MG) (MG) MG FE ( MG) I (MG) ZN (MG) NA (MG) K (MG) (MG) CU  101 21  102 21  23.000 3.OCC 50 .OOC 130.OCC 140 . 0 0 0 230.000 190.CCC 50.OCC 60.OCC 140.OOC 90.000 170.OOC 60.OCC 0.300 0.0 0.0  26.000 3 . 2CO 50.000 140.000 150.000 250.000 200.CCC 50.000 60.000 150.000 100.000 180.000 70.000 0. 300 0.0 0. 0 0.0 . 4.300 0.300 0.600 0.100 0. 200 0 . 6CC 280.000 75.OCC 0.0 51.000 300.000 7.000 8100.OCO 0.0 2.9C0 93.000 51.000 57.000 3. 100 9.000 0.700 71.OCO 470.000 0 . 140  o.c  3.700 0.300 0.800 0 .07C 0.150 0 .4CC 190 .OCC 29.OOC 0 .0 19.OOC 100.000 2 . OCC 7900.OCC 0.0 1.1CC 113.OOC 44.OCC 42.OOC 2 . ICC 3.CCC 0.50C 52.OCC 3 3 3 .OCC 0 . 8 00  103 22  104 22  31.000 37.000 0.900 1.000 10.000 10.000 20.000 30.000 40.000 30.COO 30.000 50.000 50.000 40.000 10.000 10.OCC 10.000 20.000 20.000 30.OCC 30.000 20.000 30.000 40.000 20.000 10.000 0.100 0.200 0.0 0 .0 0.0 0. 0 0.0 O.C 8.800 7.100 1 .300 1.200 0.800 I.000 0.C10 0 . 050 0.030 0.050 0 . 100 0 . 500 50.000 30.000 20.000 3 .000 0.0 0.0 3.000 6.000 100.000 300.000 1 .000 2.000 20.000 10500.000 0 .0 O.C 0.0 0.500 19.000 33.COO 18.000 31.000 6.000 15.000 0.600 0.700 2.000 5.000 0.300 0.400 236.000 33.000 167.000 222.000 0.210 0.090  105 22 42.000 1.100 10.000 40.000 40.000 60.000 50.000 10.000 30.000 40.000 20.000 50.000 20.000 0.200 0.0 0.0 0.0 9.700 1.700 1 .000 0.060 0.050 0.600 150.000 15.000 0.0 8.000 300.000 3.000 1000.000 0.0 0. 500 37.OOC 36.000 18.000 0.700 2.000 0.500 47.000 341.000 0.090  260  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS : KCAL PROT (GM ) TRY (MG) THR (MG) (MG) ISO LEU (MG) LYS (MG ) MET (MG) CYS (MG ) PHE (MG ) TYR ( MG) VAL (MG) HIS (MG) FAT- T (GM) SFA (GM) PUFA (GM) (GM ) CHCLE CHO- T (GM) (GM) SUCP (GM) CHO- F THI /S (MG) RI BO (MG) N I A C I N (MG) VI T - B6 (MG) FCLIC (UG ) V I T - 812(UG) VIT-C (MG) PANTO (UG) B I O T I N ( MG ) VIT- A (IU) VIT- D (IU) VIT- E (MG) CA (MG) P (MG) (MG) MG FE ( MG) (MG ) I ZN (MG) NA (MG) K (MG) (MG) CU  106 22  107 22  108 22  109 22  110 22  83.0GC 3.2CC 20.000 130.OCC 120.OCC 250.OCC 120.OCC 60.OOC 50.OCC 180 . 0 0 0 110.OCC 200.000 80.OCC l.OOC 0 .0 0.0 0.0 18.8CC 0.300 0.7CC 0.11C - 0.100 1.300 2 9 0 .OCC 2 . OCC 0.0 7.OOC 400.000 2.OCC 4C0.0CC 0.0 1.2CC 3 .000 89.OOC 31.000 0.600 4 . OCC 0.3CC 0.0 165.OOC 0.090  82.000 2 . 100 10.000 90.000 8 0 . OCO 230.000 80.OCO 40.000 40.000 120.000 70.000 130.000 50.000 0. 600 0.0 0. 0 0.0 20.000 0.300 0.500 0. C30 0 . 050 1. OCO 200.000 2.000 0.0 5.000 300.000 2.000 330.OCC 0.0 0. 100 3.000 56.000 20.000 0.600 2 . OCC 0.300 236.000 97.000 0. 060  84.000 2.600 20.000 110.000 100.000 290 . 0 0 0 ICO.COO 50.000 40.000 150.000 90.000 170.000 70.000 0.800 0.0 0.0 0.0 19.800 0.300 0.800 0.030 0.050 G.900 200.000 2.000 0.0 4.000 200.000 2 .000 350.OCO 0.0 C.100 5.000 49.000 21.000 0.500 2.000 0 .300 236.000 97.000 0.060  14.000 0.900 10.000 20.000 30.000 40.CCO 30.000 10.000 10.000 20.000 20.000 30.000 10.000 0.100 O.C 0.0 0.0 3.100 0. 700 0.600 0. 050 0.080 0. 800 60.000 2.000 0.0 10 . 0 0 0 200.000 2 .000 390.000 0.0 2.400 25.OOC 25.000 15.000  21.000 I .000 10.000 30.000 30.000 40.000 40.000 10.000 10.000 20.000 10.000 30.000 20.000 0.200 0.0 0.0 0.0 4.300 0.300 0.400 0. 050 0.C30 0.700 90.000 26.000 0.0 17.000 200.000 2.000 900.000 0.0 0.0 6.OOC 19.000 11.000 0.500 0.0 0 . 100 130.000 217.OOC 0.130  0 .400 4 . 000 0.400 1 .000 141.000 0.080  261 ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) LEU (MG) (MG) LYS MET (MG) CYS (MG ) PHE (MG ) TYR (MG) VAL (MG ) HIS (MG) FAT-T (GM) SFA (GM) PUFA (GM) (GM) CHOLE (GM) CHO-T (GM) SUCR CHO-F (GM ) THI A (MG ) RIBC (MG) N I A C I N (MG ) V I T - B 6 (MG) FOLIC (UG) VIT-E12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG) VIT-A ( IU) VIT-D ( IU) (MG) VIT-E CA (MG) P (MG) MG (MG) (MG) FE (MG) I (MG) ZN NA (MG ) K (MG) (MG) CU  111 22 22.OOC l.iOO 10 . 0 0 0 40.CCC 30.CCG 50.OCO 50.OCC 10.OOC 10.CCC 30 .000 20.OCC 30 . 0 0 0 20.OOC 0 .200 0.0 O.C 0.0 4.70C 0.3CC 0.5CC 0.060 0.040 0.700 ICO.OCC 18.OOC 0.0 23.OCC 300.OOC 4 . OOC 900.000 O.C 0.3CC 13.OCC 27.OOC 13.000 0.5CC 2.000 0.20C 3 . OCC 244.OCC 0.11C  112 22 23.000 0. 800 10.000 20.000 10.CCC 40.000 40.000 10.000 10.000 10.000 20.OCO 30.000 10.000 0. 200 0.0 0.0 0.0 4. 9C0 0.700 0 . 9CC 0.040 0.050 0. 300 70.000 1. OCC 0.0 22.OCO 100.000 I.000 0.0 0.0 0. 0 35.000 24.OCC 14.000 0. 400 3. 000 0. 100 34.CCC 188.000 0 . C40  113 23 14.000 0.600 0.0 20.000 20.000 30.000 30.000 10.000 10.000 10.000 20.000 20.000 0.0 C. 100 0.0 0.0 0.0 3.200 0.100 0.30C 0.C30 0.040 0.200 40.000 14.000 0 .0 11.000 300.000 3.000 0.0 0.0 0 . 100 17.OOC 18.00 0 10.000 0.300 2.000 0.100 6.000 160 . 0 0 0 0.050  114 23  115 23  17.000 1.900 10.COO 40.000 420.000 220.000 60.000 130.000 20.000 20.000 30.000 300.000 20.000 0.100 0.0 0. 0 0 .0 2.400 O.C 0.600 0.020 0.250 2 . 000 60.000 8. 000 O.C 2.000 1000.000 7.000 0. 0 40 . 0 0 0 0.0 6.000 68.000 8 . OCC 0.500 O.C 0.400 400.000 197.OOC 0.260  29.OOC 1.200 20.000 20.000 20.000 30.000 60.000 10.000 20.000 40.000 40.000 30.COO 10.000 0 . 100 0.0 0.0 0.0 6. 500 2.200 0.600 0.030 0.030 0.200 100.000 25.000 0.0 7.000 100.000 2.000 40.000 0.0 0.200 24.000 29.000 8.000 0.400 3.000 0.600 7.000 110.000 0.080  .262  ITEM CLUSTER A T T R I B U T E GFOUP NUTRIENTS: KCAL FPGT (GM) TRY (MG ) THR (MG) ISO (MG) ( MG) LEU LYS (MG) MET (MG) CYS (MG ) PHE (MG) TYR (MG) V AL (MG) HIS (MG) (GM) FAT-T SF A (GM) PUFA (GM) (GM) CHCLE CHO-T (GM ) (GM ) SUCF CHO-F (GM) THIA (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-E12(UG) (MG) VIT-C (UG) PANTO B I O T I N (MG) VIT-A (IU) VIT-D (IU) (MG) VIT-E CA (MG) P (MG) (MG) MG (MG) FE (MG) I 2N (MG) NA (MG ) K (MG) (MG) CU  lie 24 64.OCC 3.2CC 20.000 80.OOC 140 .000 ISO.CCC 150.OCC 30.OOC 50.OOC 100 . 0 0 0 8 0.OCC 150.OOC 60.OOC 0.30C 0.0 0.0 0.0 13.4CC 1.600 1.2 CO C.12C 0.07C 1 .ICC 100.OCC 15.000 0 .0 8.000 300 .000 2 . OCC 4950.OCC 0.0 0.5CC 25.OOC 6 3 . OCC 25.000 1.3CC l.OCC 0 .6CC 53.OCC 191.000 0.12C  117 25  118 25  119 25  120 26  116.000 1.4CC 10.000 40.000 40.000 60.000 60.CCC 10.000 20.OCO 30.000 60.000 50.000 0. 0 12.7CO 2.000 10.CCC 0.0 1.300 0. 0 1.300 0. 0 0.0 0. 0 20.000 13.OCO 0.0 2.000 0.0 1.000 3 0 0 . OCO 0.0 0. 0 61.000 17.000 12.000 1.600 17.OCC 0.300 2400.OCC 55.000 0. 370  146.000 0.700 10.CCC 20.000 20.000 30.000 30.000 10.000 10.000 20.000 30.000 20.000 0.0 0.400 0 .0 0.0 0.0 36.500 33.4CC 0.500 0.0 0.02C 0.0 10.000 3.000 0.0 6.000 200.000 1 .000 90.000 0.0 O.C 12.000 16.000 1.000 1 .200 17.000 0.500 823.000 200.000 0.210  11.OCC 0.700 10.000 20.000 20.000 30.000 30.000 10.000 10.000 20.OOC 30.000 20.000 0.0 0. 200 O.C 0.0 O.C 2.200 0.0 0.500 0. 0 0.020 0. 0 10.000 3.000 O.C 16.000 200.000 1.000 100.000 0.0 0.0 26.000 21.000 1.000 I .000 17.000 O.5C0 1420.000 200.000 0.020  384.CCC 0.0 0. 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.000 23.000 72.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0. 0 0.0 0.0 0.0 0.0 O.C 2 .300 0.0 0.0 1.000 0.0 24.000 0.800 0.0 0.0 0.030  263 ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR ( MG) ISO (MG) LEU (MG > LYS (MG) MET (MG) CYS (MG) PHE (MG) TYR (MG) VAL (MG) HI S (MG) FAT-T (GM) SFA (GM) PUFA (GM ) (GM ) CHGLE CHO-T (G M) (GM) SUCR CHO-F (GM) THIA (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 (MG ) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) CA (MG) P ( MG) MG (MG) FE (MG) (MG) I ZN (MG) NA (MG) K (MG) ( MG) CU  121 26  122 26  123 26  124 26  125 27  902.000 0.0 0.0 O.C O.C 0.0 O.C 0 .0 0.0 0.0 0.0 0.0 0 .0 ICO.OCC 38.OOC 56.OCC 0.090 0.0 0 .0 0.0 0 .0 0.0 0.0 20 .OOC 0.0 0 .0 0.0 0.0 0.0 0.0 0 .0 1.200 0.0 0.0 l.OCC 0.0 3 . OCC 0.500 O.C 0.0 0.020  884.000 0.0 0.0 0.0 0. 0 0.0 0. 0 0.0 0.0 0.0 0. 0 0.0 0.0 100.OCC 25.000 71.OCO 0.0 0.0 0.0 0.0 0. C 0.0 0. 0 0.0 0.0 0.0 0.0 0. 0 0.0 0. 0 0. 0 43.600 0. 0 0.0 1.000 0.0 4 . COO 0.500 0. 0 0.0 0.070  884.000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 O.C 0.0 0.0 0.0 0 .0 1 CO.000 11.000 83.000  884.000 0.0 0. c 0.0  716.000 0.600 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 81.000 46.000 29.000 0.270 0.400 0. 0 0.0 0.0  o.c  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 .0 0.0 0.0 0.0 0.0 14.400 0.0 0.0 1.000 0.0 7.000 0.500 0.0  o.c  0.070  o.c o.c  0.0  o.c 0.0 0. c 0 .0 0. 0  o.c  100.000 15.000 72 . 0 0 0 0.0 0 .0  o.c  0 .0 0. 0  o.c  0.0 0.0 0.0 0. 0 0 .0 0.0 0.0 0.0 0.0 12.100  o.c  0 .0 1. 000  o.c  4.000 0.500 0.0 0.0 0.070  o.c  0 .0 0.0 0.0 0 . 100 0.0 0.0 10.000 3300.000 40.000 1 .900 20.000 16.000 2.000 0.0 9.000 0.30C 987.000 23.000 0.030  264  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) LEU (MG) (MG ) LYS MET (MG) (MG) CYS PHE (MG) TYR ( MG) (MG) VAL HI S (MG) FAT-T (GM ) SFA (GM) PUFA (GM) CHOLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM ) THIA (MG) RIBC (MG) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (I U) VIT-D ( IU) VIT-E (MG) CA (MG) P (MG) MG (MG) FE (MG) I (MG) ZN (MG) NA (MG) K (MG) (MG) CU  126 27 720.000 0.6CC O.C 0.0 0.0 0.0 O.C 0.0 0.0 0 .0 O.C 0.0 0.0 8 1 .000 19.000 60.OCC 0 .0 0.4CC O.C 0 .0 0.0 0.0 0.0 0 .0 0.0 O.C 0.0 0.0 0.0 33 CO.CCC 0 .0 12.5CC 20.000 16.OCC 2 . OCC 0.0 7 . OCC 0 .300 987.OCC 23 . 0 0 0 0.03C  127 28  128 28  129 28  173.000 7.900 110.OCO 290.000 510.CCC 760.000 570.000 200.000 50.000 410.000 390.000 560.OOC 260.000 13.OCO 7.000 5.000 0.070 6.400 0. 0 0.0 0. 030 0.210 0. 0 40.000 9.000 0.4CC 1.000 200.000 4.000 550.000 1.000 0.500 234.000 172.000 1 7 . OCO 0.300 8. OCO 0.900 518.000 106.000 0. 070  228.OOC 1 .700 20.000 50.000 80.000 130.000 40.000 20.000 30.000 90.000 60 . 0 0 0 70.000 30.000 . 19.500 9.000 9.000 0.01C 11 . 1 0 0 O.G 0.0 0.060 0.040 0.0 50.000 1 .000 0 . 160 0.0 200.000 0.0 0.0 0.0 0.200 0.0 11 . 0 0 0 2.000 0.600 1.000 0.500 1000.000 106.OOC 0.010  162.000 3.900 90.000 230.000 3 50.OCO 580.OCO 260.000 110.000 120.000 190.000 270.000 350.000 150.000 12.500 7.OCC 4.000 0 . C40 8.800 0. 0 O.C 0.040 0 . 170 0.200 50.000 I .000 0.160 0.0 600.000 4.000 460.000 O.C 0.100 115.OOC 93.000 14.000 0.200 7.000 0.400 379.000 139.000 0.040  130 28 106.000 2.000 20.000 70.000 60.000 80.000 80.000 10.000 20.000 50.000 30.000 60.000 30.000 0.400 0. 0 0.0 0.0 25.400 16.600 0.500 0.090 0.070 1.600 110.000 27.000 0.0 15.000 200.000 4 . 000 1400.000 0.0 0.200 22.000 50.000 21.000 0.80C 6.000 1.100 1042 .000 363.000 0.510  265  ITEM CLUSTER A T T R I B U T E GROUP NUT F I E N T S : KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) LEU (MG) LYS (MG) MET (MG) ( MG) CYS PHE (MG ) TYR (MG) VAL (MG) HIS (MG) FAT-T (GM ) SFA (GM) PUFA (GM) CHOLE (GM) CHO-T (GM) (GM ) SUCR CHO-F (GM) THIA (MG) RI BC (MG) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C ( MG) PANTO (UG) B I C T I N (MG) VIT-A (I U) VIT-D (IU) VIT-E (MG) CA (MG ) P (MG ) MG (MG) FE (MG) I (MG) ZN (MG ) NA (MG) K (MG) ( MG) CU  131 28  132 29  21.000 I.CCC 10.OCC 30 .OCC 30.CCC 40.OOC 40.OCC 10 . 0 0 0 10.OOC 20 .OOC 10.OCC 30.OCC 20 .OCC 0.2CC 0 .0 0.0 0 .0 4.300 0.3CC 0 .4CC 0.G5C 0.030 0 . 7 00 90.000 26.OCC 0.0 17.OCC 2 CO.OCC 2.000 900.OOC 0 .0 0.0 6 . OCC 19.OOC 11.OCC 0.5CC 0.0 0.10G 130.OCC 217.000 0.13C  718.000 1. 1 0 0 0.0 0.0 0. 0 0.0 0. 0 0. 0 0.0 0.0 0.0 0. 0 0.0 79.9C0 14.000 57.000 0.050 2.200 0.0 0.0 0. 020 0.040 0.0 0. 0 0.0 0.0 3.000 100.000 3.000 280.000 8.000 11.900 18.000 28.000 2 . OOC 0.500 27.COG 0 . 500 597.000 34.000 0.240  133 29 552.000 0.200 0.0 0 .0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 60.000 10.000 44.000 O.C 6.900 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ' 0.0 0.0 9.100 10.CCO 4.000 7.000 0.200 2.000 0.40C 2092.000 15.000 0.040  134 29 435.000 1 .000 0. 0 O.C 0.0 0.0 0.0 0.0 0.0 O.C 0 .0 0. 0  o.c 42.300 8.C0C 30.000 0.050 14.400 10.000 0.100 0.010 0.030 0.0 0.0 0 .0 0. 0 3.000 100.000 3.000 220.000 8.000 5.300 14.000 26 . 0 0 0 2.000 0.20C 27.000 0.500 586.000 9.000 0.240  135 29 368.000 21.500 290.000 810.COO 1480.000 2140.000 1550.000 580.000 120.000 1200.000 1030.000 1580.000 700.000 30.500 17.000 11.000 0.150 2.000 0. 0 0.0 0.030 0.61C 1 .200 170.000 11.000 1.400 0.0 1800.000 3.000 1240.000 30.000 0.800 315.000 339.000 20.000 0.500 11 . 0 0 0 2.200 666.000 78.000 0 . 160  266  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PRQT (GM) TRY (MG 1 THR (MG) ISO (MG) LEU (MG) ( MG) LYS MET (MG ) CYS (MG) PHE (MG ) TYR (MG) (MG) VAL HIS (MG) (GM) FAT-T (GM) SFA PUFA (GM) (GM) CHOLE (GM) CHC-T (GM ) SUCR CHO-F (GM ) (MG) THIA RIBG (MG) N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) (MG) VIT-C (UG) PANTO B I O T I N (MG) VIT-A (IU) VIT-D (IU) (MG) VIT-E CA (MG ) P (MG) MG (MG) FE (MG ) (MG) I ZN (MG) NA ( MG) K (MG) (MG) CU  136 3C  13 7 30  128 30  139 31  140 31  65.OCC 3.500 5 0.OCC 160.000 230.OCC 3 50 . 0 0 0 280.OOC EO.OCC 30.OOC 1 7 0 . OCC 180.OCC 240.000 90.000 3.5CC 2.OOC I.OCC 0.C14 4.90C 0.0 0 .0 0.03C 0 .170 0.1CC 4 0 . CCC 9 . OOC 0.4CC l.OOC 3C0.CGC 4.000 140.OCC 41.000 0.1CC 118.OOC 93 .OCC 13.000 0.0 7 . OCC 0 .400 50.CCC 144.OCC 0.15C  36.000 3.600 50.CCC 160.000 190.000 360.000 280.000 90.000 30.000 170.OCC 180.000 250.OCO 100.000 0.100 0. 0 0.0 C. CC2 5.100 0. 0 0.0 0.040 0. 180 0.100 40.000 9.000 0.4CC 1.000 400.OCO 2.000 0.0 41.000 0.0 121.OCO 95.000 15.000 0.0 7.000 0.400 52.000 145.OCO 0.020  59.000 4.200 60.000 190.000 270.000 420.000 330.000 100.000 40.000 2C0.000 200.000 290.000 110.000 2.000 1.000 1 .000 0.002 6.000 0.0 0.0 0.040 0.210 0.100 40.000 9.000 0.400 1.000 400.000 3.000 80.000 41.000 0.100 143.00 0 112.000 17.000 0.10C 8.000 0.400 61 . 0 0 0 175.000 0.020  138.000 3.300 45.000 150.000 210.000 325.000 255 .000 75.000 30.000 160.000 165.000 225.000 85.000 12.100 7. 000 1 .000 0 . C08 4.600 0.0 O.C 0.030 0.160 0.100 35.000 15.000 0.330 1.000 300.000 4.000 495.000 28.000 0 .400 110.000 87.000 12.000 0.0 7.000 0.400 47.000 133.000 0.160  211.000 3.000 40.000 140.000 190.000 300.000 230.000 70.000 30.000 150.000 150.000 210.000 80.000 20.600 11.000 8.000 0.070 4.300 0.0 0.0 0.03 0 0 . 150 0.100 30.000 20.000 0.250 l.OCC 300.000 4.OCC 840.000 15.000 0.700 102.000 80.000 10.000 0.0 6.000 0.300 43.000 122.000 0.170  267  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PR 07 (GM) TRY (MG) THR (MG ) ISO (MG) LEU (MG) LYS (MG) MET (MG) CYS (MG) PHE (MG ) TYR (MG ) VAL (MG) HIS (MG ) (GM) FAT-T (GM) SFA PUFA (GM) CHOLE (GM ) CHO-T (GM) SUCR (GM) CHO-F (GM) THIA (MG) RIBC (MG ) N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12CUG) (MG) VIT-C PANTO (UG) B I O T I N (MG) VIT-A ( IU) VIT-D (IU ) VI T - E (MG) CA (MG ) P (MG) MG (MG) FE (MG) ( MG) I ZN (MG) NA (MG) K (MG) CU (MG )  141 31  142 31  143 32  144 32  145 32  252.OCC 2.20C 20.OCC 100.000 140.OOC 22 0 . CCC 170.000 50.OCC 20.OOC 110.000 110.OOC 150.OCC 60.000 37.6CC 21.OCC 13.OCC 0.12C 3 .100 2 . OOC 0 .0 0.02C 0.11C 0.0 20.CCC 15.000 O.IOC 0.0 2C0.00C 3 .OCC 1540.OCC 11.OCC 4.9CC 7 5 . CCC 59 . 0 0 0 7.OCO O.C 4.OOC 0.2CC 32.000 £9.OCC 0 .120  137.OCO 7.000 100.000 320.000 450.000 690.000 550.000 170.OCO 60.000 340.COO 360.000 480.000 190.000 7.900 7 . OCC 3.000 0. 110 9.700 0.0 0. 0 0. 040 C.34C 0.200 50.OCO 1.000 0. 160 1.000 600.000 8.000 320.000 79.OCC 0.200 252.000 205.000 33.000 0. 100 16.000 C. 7 0 0 118.000 303.000 0.090  47.00 0 0.100 0.0 3.000 5 .000 5.OCC 4.000 2.000 1.000 3 . OOC 2 . COO 3.000 2.000 0.0 0.0 0.0 O.C 11.900 5.500 0 . 100 0.010 0.020 0.100 30.000 0.0 0.0 l.OCC 100.000 O.C 40.000 0.0 0.0 6.000 9.000 4.000 0.600 2 .000 0 . 100 1.000 101.000 0.020  41.000 0.50C 1 .000 1.000 1 .000 1 . OCO 6.000 0. 0 l.OCC 11.000 6.000 I.000 1. 0 0 0 0 .100 0.0 0.0 0.0 9 . 800 2.700 0. 0 0 .030 0. 020 0.20C 10.000 1.000 0.0 34.000 100.000 1.000 10.000 0. 0 0.0 8.000 14.OOC 7.000 0.400 1 .000 0.0 1.000 162.000 0.010  23.000 0.400 2 .000 1.000 I.000 1. 000 8.000 1.000 1.000 11 . 0 0 0 6.000 1.000 1.000 0 . 100 0. 0 0.0 0.0 7.600 0.100 0.0 0.030 0.010 0 . 100 50.000 2.000 0.0 42.000 100.000 0.0 20.000 0.0 0. 0 7.000 10.000 9.000 0.200 5.000 0.200 1.000 141.000 0.080  268  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY ( MG) THR (MG ) (MG) I SO LEU (MG ) (MG) LYS MET (MG) CYS (MG) (MG) PHE TYR (MG) ( MG) VAL (MG) HIS FAT- T (GM) SFA (GM) PUF/ (GM) CHOLE (GM ) (GM ) CHO- T (GM) SUCR (GM) CHO- F THIA ( MG) RIBG (MG) N I A C I N (MG) V I T - B6 (MG) FOLIC (UG ) V I T - B12(UG) VIT- C (MG) PANTO (UG) B I O T I N (MG ) VI T (IU) VIT- D ( I U) VIT- E (MG ) CA (MG) P (MG) MG (MG) FE (MG ) (MG) I ZN (MG) NA (MG) K (MG) (MG ) CU  146 32 4 4 . OCC O.lOO 0.0 0.0 0.0 0.0 0 .0 0.0 0.0 0.0 0.0 0.0 0.0 0 .0 0.0 0 .0 0.0 11 . 4 0 0 4.9CC 0 .0 0.0 0.01C 0.100 10.OCC 2.OOC 0.0 7 .000 O.C 0.0 0 .0 0.0 0.0 1.000 1.000 l.OCC O.C 7.OOC 0.10C 0 .0 1 6 . OCC 0 .010  147 32  148 32  149 32  150 33  48.000 0. 800 3.000 1.000 1.000 1. OCC 21.000 2.000 1.000 9. 000 15.000 1.000 1. 0 0 0 0. 200 0. 0 0.0 0.0 11.200 3.200 0 . 100 0.070 0 . 02 0 0.300 40.OCC 4 . 000 0.0 40.000 200.000 1. 0 0 0 200.000 0. 0 0.0 10.000 18.000 11.000 C.4CC 1.000 0.2C0 1. 000 199.000 0.050  45.000 0.700 3 .000 1.000 I.000 1.000 21.000 2.000 1.000 9.000 15.000 1 .000 1.000 0.100 0.0 0.0 O.C 1C.7CC 3.200 O.C 0 .090 0.010 0.300 30.000 4.000 0.0 45.000 200.000 0.0 200.000 0.0 O.C 9.000 16.000 12 . 0 0 0 0.100 1.000 0 . 100 1.000 186.000 0.050  49.000 1.000 3 .000 1.000 1.000 1. 0 0 0 24.000 3.000 1.000 12.000 21.000 1 .000 1. 000 0.200 0. 0 O.C 0.0 12.200 4.200 0 . 500 0.100 0.040 0.400 60.000 45.OOC 0.0 50.000 300.000 2. 000 200.000 0.0 0.200 41.000 20.000 11.000 0.400 O.C 0.100 1.000 200.000 0.090  21.000 1.000 10.000 30.000 30.000 40.000 40.000 10.000 10.000 20.000 10.000 30.000 20.000 0.200 0.0 O.C 0.0 4.300 0.300 0.400 0.050 0 . 030 0.70C 90.000 26.000 0.0 17.000 200.000 2.000 900.000 0.0 0.0 6.000 19.000 11.000 0.500 0.0 0.100 130.000 217.000 0.130  269  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL (GM) PROT TRY (MG) THR (MG) (MG) I SO (MG ) LEU LYS (MG) MET (MG) (MG) CYS PHE (MG) TYR (MG) VAL (MG) (MG) HIS FAT-T (GM) SFA (GM) PUFA (GM) CHOLE (GM) (GM) CHO-T (GM SLCR CHO-F (GM ) THIA (MG ) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG ) (MG) VIT-C PANTO (UG) B I O T I N (MG ) VIT-A (IU) VIT-D (IU) VIT-E (MG) CA ( MG ) P (MG) MG (MG) FE (MG) (MG) I ZN (MG) NA (MG) K (MG) CU (MG)  151 34 39.OCC 0.0 0.0 O.C O.C 0.0 0.0 0.0 0.0  o.c  0.0 0.0 0.0 0 .0 0.0 0 .0 0.0 10.OCC 10.OCC 0.0 0.0 0.0 0 .0 0.0 0.0 0.0  o.c 0 .0 CO 0.0 0.0 0.0 8.000 15.CCC 1.000 0.400 1 .000 0.1CC 6 .000  o.c 0.04C  152 34 31.000 0.0 0.0 O.C 0.0 0. 0 0.0 0.0 0.0 0.0 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 8. 000 8.000 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 0. 0 0.0 0. 0 0.0 0. 0 0.0 8.000 15.OCC 1.000 0.4CC 1.000 0.100 6.000 0.0 0.03C  153 35  154 35  I .000 0.0 0.0 0.0 0.0  2.000 O.C 0 .0 0. 0 0.0 0.0 0.0 0.0 0. 0 0 .0  o.c  0.0 0.0 O.C 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0.300 10 . 0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2 .000 4.000 5.000 0.100 4.000 0.0 1.000 36 . 0 0 0 0.02C  o.c o.c 0.0  o.c  0.0  o.c  0.0 0.400  o.c  0.0  o.c  0.010 0.0 0 .0 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 0. 0 0.0 0.0 2.000  o.c 16.000 0.0  o.c  25.000 0. 010  155 36 42.000 0.300 0.0 0.0 0. C 0.0 O.C 0.0 0.0 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0.0 3 . 800 0.0 0.0 0.0 0.030 0.600 60.000 0.0 0.0 0.0 100.000 0.0 0.0 0.0 0. 0 5.000 30.000 10.000 0.0 1.000 0.100 7.000 25.000 0.070  270  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) (MG) ISO (MG ) LEU LYS (MG) MET (MG) (MG ) CYS PHE (MG) TYR (MG) V AL (MG) (MG) HJS FAT-T (GM) SFA (GM) PUFA (GM) CHOLE (GM) (GM ) CHO-T (GM) SUCR CHO-F (GM ) THIA (MG) RIBC (MG) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-E12(UG) VIT-C (MG) (UG) PANTO B I O T I N (MG) VIT-A (IU) VIT-D ( IU) VIT-E (MG) CA (MG) P (MG) MG (MG) (MG) FE I (MG) (MG) ZN (MG) NA K (MG) CU (MG)  156 36  157 36  249.000 0.0 0.0 0.0 0 .0 0.0 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0 .0 0.0 0 .0 0.0 0.0 0.0 0.0 0 .0 0.0 0.0 O.C 0.0 0 .0 0.0 0.0 O.C 0 .0 0.0 0.0 8.000 10.OOC 0 .0 0.400 1.000 0.1CC 1.000 2.OCC 0.C8C  137.000 0. ICC 0.0 0. 0 0.0 0. 0 0.0 0.0 0. 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7. 7CC 7.700 0. 0 0. 010 0.020 0.200 40.000 0. 0 0.0 0. 0 0.0 0. 0 0.0 0.0 0. 0 8.000 10.OCC 5.000 0.400 1.000 0.100 4.000 75.000 0 . 08C  158 26 85 . 0 0 0 0 . 100 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.200 4.200 0.0 0.0 0.010 0 . 100 40.000 0.0 0 .0 0.0 0.0 0.0 0.0 0.0 0.0 9.000 10.000 8.000 0.400 1.000 0.100 5.000 92.000 0.110  159 37  160 37  26.000 1. 4 0 0 10.OCC 70.000 80.000 80.000 180.000 40 . 0 0 0 60.000 60.000 30.000 60.OOC 20.000 0 . SCO 0 .0 0. 0 0 .003 3.300 O.C 0.100 0.010 0.010 0.300 30.000 0. 0 O.C 0.0 100.000 1 .000 20.000 0.0 0. c 4.000 15.000 4.000 0.200 1.000 0.300 408.000 23 . 0 0 0 0 . 130  73.000 3.000 40.000 130.000 290.000 230.000 260.000 90.000 120.000 120.000 50.000 170.OCC 40.000 4.200 I.000 2.000 0.009 5.900 0.0 0.100 0.20C 0.110 0.300 30.000 4.000 0. 200 1.000 200.000 1.000 250.000 I.000 0.0 70.000 62.000 9.000 0.200 4.000 0.400 430.000 106.000 0.050  271  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS: KCAL (GM) PROT TRY (MG) THR (MG) (MG) ISO ( MG) LEU (MG) LYS MET (MG) CYS (MG ) PHE (MG) TYR (MG) VAL (MG) (MG) HIS FAT-T (GM) SFA (GM) (GM) PUFA CHOLE (GM) CHO-T (GM ) (GM) SUCR CHO-F (GM) THIA (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) (MG) CA P (MG) MG (MG) (MG) FE (MG) I ZN (MG) NA .(MG) K (MG) (MG) CU  161 37  162 37  163 37  59.OCC 3.5CC 30.OCC 130.OCC 190.000 250.OCO 210.000 30.OCC 3 0 . CCC 8C0 .OCO 8 0 . OCC 190.000 80.000 1.300 0.0 O.C 0.002 8.40C 2 .6CC 0.2CC 0 .100 0.060 0 . 6 CC 50.OCC l.OCC 0.16C 0.0 100 .000 l.OCC 180 . 0 0 0 l.OOC 0 .0 12 .OCC 61.000 6.000 0.6CC 1.000 0.4CC 384.OCC 110.OOC 0.09C  36.000 0.8C0 10.000 30.OOC 30.000 30.000 80.000 30.000 10.CCC 20.000 10.OCC 30.000 20.000 1.000 0. 0 0. 0 0.002 6.4C0 0.400 0.200 0. 020 0.020 0.500 20.000 4 . OCC 0.0 5. CCO 100.000 3.000 410.000 I.000 0. C 6.000 14.OCC 7.000 0 . 300 3.000 0 . 300 396.OCO 94.000 0 . 160  32 . 0 0 0 2 . 100 20.000 90.000 70.000 170.000 230.000 40.000 30.000 90.000 30.000 120.000 30.000 C.900 0.0 0.0 0.003 3.900 0 . 100 0.200 0.020 0.020 0.400 30.000 4.000 0.0 2.000 100.000 1.000 100.000 0.0 0 . 100 5.000 20.000 11.CCC 0.300 2.000 0 .300 427.000 66 . 0 0 0 0.C90  164 37 3 .000 0.400 5.000 18.000 21.000 33.000 35.OOC 10.000 5.000 7.000 14.000 22.000 14.000 0.100 0.0 0. 0 0.0 0.100 0.0 0.0 0. 0 0.010 0 . 500 0.0 0.0 0.0 0.0 0.0 0.0 0. c 0 .0 0. 0  o.c  6.000 1.000 0 . 100 1.000 0.0 48.000 2.OOC 0.0  165 38 259.000 5. 700 90.000 210.000 330.000 550.000 170.000 90.000 100.000 310.COC 220.000 310.000 200.000 0.200 0.0 0.0 0.0 59.400 35.OOC 0.0 0.0 0.110 0.100 10.000 2 . 000 0.040 0.0 200.OCC 4.000 0.0 0.0 0.0 95.000 119.000 15.000 0.300 3.000 0 . 200 146.000 60.000 0.040  272-  ITEM C L U S T E R A T T R I B U T E GROUP NUTRIENTS : KCAL PROT (GM) TRY (MG) THR (MG) ISO (MG) (MG) LEU (MG) LYS MET (MG) (MG ) CYS PHE (MG) TYR (MG) (MG) VAL ( MG) HIS (GM) FAT-T SFA (GM) (GM) PUFA (GM) CHCLE CHO-T (GM) (GM) SUCR CHO-F (GM) THIA (MG) (MG) RIBO N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-E12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-0 (IU ) VIT-E (MG) CA (MG) P (MG) MG (MG) FE ( MG) I (MG ) ZN (MG ) NA (MG ) K (MG) (MG ) CU  166 38  167 38  168 38  169 38  170 38  322.OOC 6.30C 130.000 360.OOC 510.000 760.OCC 1150.OCC 200.OCC 180.OCC 340.000 350.OCC 540.000 2 CO.OCC 9.6CC 3.OCC 6.OOC 0.080 52.400 19.100 0 . ICC 0.180 0.16C 1.4C0 40.OOC 8 .OOC 0.0 0.0 200 .000 5 . OOC 160.CCC 6 .OCC 0.20C 61 .OOC 174.OCC 15.000 1.6 00 7 . CCC 0.6CC 4 2 1 .OCC 109.OCC 0.08C  380.000 4 . 3 CO 80.000 210.000 340.000 530.000 270.OCC 120.000 120.OCC 240.000 250.000 350.000 330.000 17.600 9.000 8 . OCC 0.070 55.60C 43.000 0. 300 0.020 0.080 C.2CC 50.000 22.OCC 0.0 0. 0 200.000 6.000 430.OCC 9.000 C. I C C 54.000 52.000 24.000 0.800 7.000 0 . 500 420.OCO 119.000 0.310  379.000 4.800 90.000 230.000 340.000 530.000 270.000 120.000 120.000 260.000 250.000 350.OCC 330 .000 15.300 6.000 9.000 0 . 100 59.700 26.700 0.600 0.130 0.140 0.800 80.000 3.000 0 . 130 0.0 400.000 8.000 120.000 13.000 0.700 72.000 113.000 16.000 2 .600 7.000 0.600 158.OCC 496.000 0.100  411.000 6.400 130.000 360.000 520.OCO 770.000 1160.000 210.OOC 180.000 350.OCC 350.000 550.COO 200.000 18.7CC 5 .000 12.000 0 . 160 54.700 28.100 0 . 100 0. 040 0.110 0.200 40.000 8.000 0.0 0.0 300.000 3.000 290.000 20.CCC 1 .100 40.000 104.000 13.000 0.800 7.000 0.600 178.000 78.000 0.060  337.OOC 4.100 80.000 210.000 300.000 470.000 240.000 110.000 100.000 230.000 220.000 310.000 290.000 11.300 5.000 6.000 0 . C9C 57.600 36.200 0.200 0.020 0.080 0.200 40.000 8.000 0.0 0.0 300.000 5.000 140.000 2. 000 0.500 91.OOC 182.000 20.000 0.600 7.000 0 . 500 227.000 109.000 0.100  273  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR ( MG) ISO (MG) (MG) LEU LYS (MG) MET (MG) (MG ) CYS PHE (MG) TYR ( MG) VAL (MG) HI S (MG) FAT-T (GM) SFA (GM) PUFA (GM) C HC LE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM) (MG) THIA RIBC (MG) N I A C I N (MG ) VI T - B 6 (MG) FOLIC (UG) VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D ( IU) VIT-E (MG) (MG) CA P ( MG) MG (MG) (MG) FE I (MG) (MG) ZN NA (MG) K (MG ) (MG) CU  171 39  172 39  173 40  174 40  175 41  256.OOC 2 .20C 30.000 60.000 ICO.OOC 170 .OCC 50.CCC 30 .OOC 40.OOC 120 .000 80.000 ICO.000 40.OOC 11.100 3.OCC 8 . OOC 0.030 38.ICC 10.900 0.4CC 0.02C 0.020 0.40C 40 .OOC 4.OCC 0.0 1.000 ICO.OCC 1 .OOC 20.OCC 0.0 0.900 8.000 22.000 4.CCC 0.30C 4.OCC 0 .40C 301.000 80.000 0.C8C  211.000 4.000 60.000 170.000 2 5 0 . CCC 340.000 260.OOG 110.000 70.000 220.000 160.000 260.000 90.000 1 1 . 2 CO 3.000 8. 000 0 . 100 24.500 15.100 0. 500 0 . C3C 0.100 0. 500 40.000 4.000 0. 0 0. 0 5 CO.CCO 5.000 24 7 0 . O C C 10.000 0. 900 51.000 69.000 6. 000 0 . 500 3 . OCC 0.400 214.OCO 160.000 0.050  480.COO 5.100 60.OOC 150.000 230.000 390.000 120.000 70.000 100.000 280.000 170.000 220.000 100.000 2 0 . 200 4.000 15.000 0.100 71.000 37.100 0 .100 0.030 0.050 0.400 50.000 11.000 0.0 0.0 400.OCO 5.000 80.000 1.000 0.500 37.OOC 163.000 15.000 0.7C0 10.000 1.700 365.OOC 67.OOC 0.150  .358.000 3.900 50.OCO 110.000 190.000 300.CCC 90.000 50.OOC 80.000 210.000 130 . 0 0 0 170.000 80 . 0 0 0 5. 60C l.COO 3.000 0.060 75.400 25.700 1 .700 0 . 040 0.07C 0.300 90.000 10.000  391.OOC 4.600 60.000 160.000 240.000 380.000 170.000 80.OCC 140.000 250.000 200.000 240.000 90.000 18.600 4.000 13.000 0.045 51.400 16.100 0 . 100 0 . 160 0 . 16C 1.200 40.OCC 9.000 0.0 0.0 400.000 3.000 80.000 3.000 0.400 40.000 190.000 16.000 1.400 7.000 0.700 501.000 90.OOC 0.110  0. C 0 .0 300.000 5.CCC 110.000 1 . COO 0.400 78.000 60 . 0 0 0 23.000 1 .000 10.000 0.900 252.000 198.000 0.190  274  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL (GM) PROT TRY (MG) THR (MG) (MG) ISO (MG) LEU LYS (MG ) MET (MG) CYS (MG) PHE (MG) TYR (MG) (MG > V AL HI S (MG) FAT-T (GM ) SFA (GM) PUFA (GM) (GM ) C HOLE CHO-T (GM) (GM) SUCR CHO-F (GM) (MG) THIA (MG) RI BO N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (10) VIT-D (IU) VIT-E (MG) CA (MG) P (MG) MG (MG) (MG) FE (MG) I (MG) ZN NA (MG) K (MG) (MG) CU  176 41  177 41  178 42  179 42  414.000 6.3CC 80.000 220.OCC 320.OCC 510.OCC 220.OCC 110.000 70.OCO 340.000 160.OCC 260.OOC 90.OCC 26.7CC 6.000 1 9 . CCC 0 .040 37.7CG 15.ICC 0.20C 0 .16 C 0.17C 1.3C0 40.000 8 . OOC 0 .0 0.0 5 CO.OCC 3.000 60.000 1.000 0.7CC 38.OOC 76.OCC 16.OCC 1.5CC 7.OCC 0.70C 234.OCC 80 . 0 0 0 0.11C  322.000 6.300 130.000 360.000 510.OCC 760.000 1150.OCO 200.000 180.000 340.000 350.000 540.CCO 200.000 9 . 6CC 3.000 6. 000 0.080 52.400 19.100 0 . 100 0. 180 0 . 160 1.4C0 40.000 8.000 0. 0 0.0 200.OCC 5.000 160.OCC 6.000 0.200 61.000 174.000 15.OOC 1.600 7.OOC 0.600 431.000 109.000 0.080  207.OCC 4.000 60.OOC 180 . 0 0 0 260.COC 400.000 310.000 100.000 40.000 190.000 200.000 280.000 110 . 0 0 0 12.500 7.000 4.000 C.C4C 20.600 15.6CC 0.0 0.C40 0 . 190 0. 100 30.000 1.000 0.250 1 .000 500.COO 4.000 520.000 5.000 0.100 123.OOC 99.000 18.000 0.100 12.000 0.500 40.000 112.000 0.020  134.000 0.900 10.000 40.000 60.000 90.COO 70.000 20.000 10.000 40.OCO 50 . 0 0 0 60.000 20.OOC 1 .200 0.0 0.0 O.C 30.800 29.300 0 .0 0. 010 0.030 0.0 30.000 3.000 0. 0 2 .000 300.000 1.000 60.000 0.0 0.100 16.000 13.000 9 . 000 0.0 2 .000 0.200 10.000 22.000 0.020  180 42 31.OOC 0.0 0.0 0.0 0. 0 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0. 0 0.0 0.0 0.0 0.0 8.000 8.000 0.0 0.0 O.C 0.0 0. 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.000 15.000 1.000 0.400 1 .000 0 . 100 6.000 0.0 0.030  275  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) (MG) ISO LEU (MG) (MG ) LYS MET (MG ) CYS (MG) (MG ) PHE TYR (MG) (MG ) VAL (MG) HIS FAT-T (GM) (GM) SFA PUFA (GM) CHOLE (GM ) CHO-T (GM ) (GM) SUCR CHO-F (GM) THIA (MG ) (MG) RIBC N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG> VI T - C (MG) PANTO (UG) B I G T I N (MG) VIT-A (IU) VIT-C (IU) VIT-E (MG) (MG) CA P (MG) (MG) MG (MG) FE (MG) I (MG) ZN NA (MG) K (MG) (MG) CU  181 43  182 43  183 44  184 44  185 44  124.000 3.4CC 60.CCC 190.OCC 270.OCC 4 0 0 .OOC 270.OCC 110.000 240.OCC 220.OCC 200.OCC 280.OCC 110.OOC 3 . OCC 2.000 l.OCC 0.C5C 22.80C 15.OCC 0.1CC 0.020 0.15C 0 . ICC 50 . 0 0 0 7.C0C 0.25C 0.0 5CC.0CC 5.OOC 130.OCC 2.000 0.7CC 1 0 2 . OOC 9 5 . OCC 2 3 . OCC 0.30C 7.OCC 2 .600 129.OOC 136.OOC 0.12C  59.000 4 . 200 6 0 . CCO 190.000 2 7 0 . CCC 420.000 330.000 100.000 40.000 200.OCC 200.000 290.OCO 110.000 2. 000 1.000 I.000 0.002 6.000 0. 0 0.0 0. 040 0.210 0.100 40.000 9.000 C. 4CC 1.000 4C0.OCC 3.000 80.000 41.000 0 . 100 143.OCC 112.000 17.CCC 0.100 8 . OCO 0.400 61.000 175.000 0.020  58.OCC 0.200 0.0 6.000 11.000 10.000 8.000 3.000 1 .000 6.000 3.000 7.000 3.000 0.60C O.G 0.0 O.C 14.500 3.300 1.000 0.030 0.020 0 . 100 30.000 2.000 0.0 4.000 100.000 1.000 90.000 0.0 0.600 7.C0C 10 . 0 0 0 5.000 0.300 3.000 0.0 I .000 110.OOC 0.080  91.000 0.200 O.C 6.OCO 11 . 0 0 0 10.000 8 .000 3.COO 1 .000 6 . COO 3.OCC 7.000 3.CCC 0.100 O.C 0 .0 O.C 23 . 8 0 0 12.600 0.5CC 0.020 0.01C 0.0 30.000 2 .000 0.0 1.000 100.000 l.OOC 40.000 0.0 0 .100 4 . 000 5.OCC 4.000 0.500 13.000 0.900 2.000 65.OOC 0 .010  85.000 I. 100 18.000 27.000 56.000 59.000 55.000 11.000 16.000 34.000 33.000 65.000 1.000 0.200 0.0 0.0 0. 0 22.200 8.700 0.50C 0.050 0.060 0.700 320.000 27.000 0.0 10.OOC 300.000 4.000 190.000 0.0 0.400 8.000 26.000 31.000 0.70C 8.000 0.200 1.000 370.000 0.130  276  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL FROT (GM) TRY (MG) THR (MG) ISO (MG ) LEU (MG ) LYS (MG) MET (MG) CYS (MG) PHE (MG ) TYR (MG) VAL (MG) HIS (MG) FAT-T (GM) SFA (GM) PUFA (GM) (GM ) CHCLE CHO-T (GM) (GM ) SUCR CHO-F (GM) THI A (MG) RI BO (MG) N I A C I N (MG) V I T - B 6 (MG) (UG) FOLIC VIT-B12CUG> VIT-C (MG) PANTO (UG) E I O T I N (MG) VIT-A ( IU) VIT-D (IU) VIT-E ( MG) CA (MG 1 P (MG) MG (MG) FE (MG) I (MG ) ZN (MG) NA (MG) K (MG ) CU (MG)  186 44  187 44  188 44  189 44  190 44  30.OCC 0.70C 1 .OCC 1 .OCC l.OCC 1.000 15.OCC 2 .000 l.OCC 21 .OCC 12.000 1 .OCC 1 .000 0.100 0.0 0.0 0.0 7.5CC 4.4CC 0.3C0 0.04C 0 .030 0.6CC SO.OCC 8.00C 0.0 33 .OCC 3 CO.CCC 3 .000 34C0.0CC 0 .0 0.1CC 1 4 . OCC 16.OCC 14.OCC 0 .400 2.CCC 0 .100 12.CCC 251.CCC 0.04C  41.CCC 0.500 1. OCC 1.000 1.000 1.000 6.000 0.0 1.000 11.CCO 6.000 l.OCO 1.000 0 . 100 0.0 0. 0 0. 0 10.600 2.9G0 0.200 0. 040 0. 020 0.200 30.000 2.000 0.0 38.000 300.000 3.000 80.000 0. 0 0.300 16.OCO 16.000 S. 0 0 0 0.400 1.000 0 . 100 1.000 135.OCC 0.040  49.000 1.000 3.000 1.000 1.000 l.OCO 24.000 3.000 1 .000 12.000 21.000 1.000 1.000 0.200 0.0 0.0 O.C 12.200 4.200 0.500 0.100 0.040 0.400 60.000 45.000 0.0 50.000 300.000 2.000 200.000 0.0 0.200 41.000 20.000 11.000 0.400 0.0 0.100 1.000 2C0.000 0.090  78.000 0.400 1.000 1.000 1 .000 1 .000 I .000 1. 000 1 .000 8.000 10.000 I .000 1.000 0.100 0. c 0 .0 0. 0 20.100 16.300 0.400 0.010 0.020 0.600 20.000 11.000 0. 0 3.OOC 100.000 0.0 430 .000 0. 0 0.0 4 . 000 12.000 6.000 0.30C 16.000 0.0 2 .000 130.000 0.07C  38.000 0.6CC 1.000 l.OOC 1 .000 1.000 1.000 1.000 I.000 12.000 15.OOC 1.000 l.OOC 0.100 0.0 0.0 0.0 9.700 5. 9 0 0 0.600 0.020 0.050 1.000 20.000 11.000 0.0 7.000 200.000 2.000 1330.000 0.0 0.500 9 . 000 19.000 11.000 0.500 6.000 0.0 1.000 202.000 0.050  277 I T E M CLUSTER A T T R I B U T E GROUP NUTR I E N T S : KC AL PROT (GM) TRY (MG) THR (MG) (MG) I SO (MG) LEU (MG) LYS MET (MG) CYS (MG) PRE (MG ) TYR (MG) VAL (MG ) HIS (MG) FAT-T (GM) (GM) SFA (GM) PUFA (GM • CHOLE CHO-T (GM) (GM) SUCR CHO-F (GM ) THIA ( MG) RI BO (MG) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VIT-B12(UG) (MG ) VIT-C PANTO (UG) B I O T I N (MG) VIT-A ( IU) VIT-D (IU) VIT-E (MG) CA (MG ) P (MG) MG (MG) FE (MG) ( MG) I ZN (MG) (MG) NA K (MG) CU (MG)  191 44  192 44  193 45  74.OCC 0 .3CC 5.OCC 1.000 1.000 1.000 9.OCC l.OCC 1 .OCC 8.OOC 8.OOC i.OOC 1.000 0.1CC 0.0 0.0 0.0 19 . 4 0 0 13•10 C 0 .300 0.080 0.C2C 0.20C 70.OCC 2 .OCC 0.0 7.000 2 CO.OOC 1.000 50.OCO 0.0 0.0 11 .OCC 5.OOC 8.000 0 .300 2.OCC 0.2CC 1 .OCC 96.OCC 0.150  52.000 0.400 5 . CCC 1.000 1.000 1.000 9.000 1. CCO 1.000 8. OOC 8.000 I.000 1.000 0.200 0.0 0.0 0. 0 13.700 7.400 0.400 0.090 0. 030 0.200 90.OCO 1.000 0.0 17.000 200.000 2.000 70.000 C. 0 0.600 17.000 8.000 12.000 0 . 500 16.000 C.20C 1.000 146.000 0.070  289.000 2.500 61.000 61.000 74.000 77.000 65 . 0 0 0 27.000 1.000 75.000 19.000 94.000 49.000 0.200 0.0 0.0 O.C 77.400 14.200 0.90C 0.110 0.C8C 0.500 240.000 9.000 0.0 1.000 100.000 5.000 20.000 0.0 0.300 62.000 101.000 31.000 3 . 500 3 .000 0.2C0 27.000 763.000 0.230  194 46 59.000 1.500 0.0 30.000 20.000 50.000 80.000 10.000 O.C 40.000 10.000 40.000 10.000 0.0 0. 0  o.c  0.0 14.100 14.100 0.0 0.0 0.0 0.0 0. 0 0.0 0.0  o.c  0 .0 0.0 0.0 0.0 0.0 0.0 0.0 I .000 0. 0 1 .000 0. 500 51.CCC 0.0 0.0  195 47 304.000 0.300 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 82.300 1.900 0.0 0.0 0.040 0.300 20.000 3.OCO 0.0 1.000 200.000 0.0 0.0 0.0 0.0 5.000 6.000 4.000 0.500 2.000 0.900 5.000 51 . 0 0 0 1.670  278  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR (MG) ( MG) ISO LEU (MG) LYS (MG) MET (MG) CYS (MG ) PHE (MG) TYR (MG) ( MG) VAL HIS (MG) FAT-T (GM) SFA (GM • PUFA (GM) CHOLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM ) THIA (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 (MG ) FOLIC (UG) VIT-B12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG ) VIT-A (IU) VIT-D (IU) VIT-E (MG ) CA ( MG) P (MG) (MG) MG FE (MG ) I (MG) (MG) ZN NA (MG ) K (MG) (MG) CU  196 47 2 32.CCC 0 .0 0.0 O.G O.C 0.0 0.0 0.0 0 .0 0.0 0 .0 O.C 0.0 0 .0 O.C 0.0 0.0 60.000 53.6CC 2 .CCC 0.09C 0.120 1.200 2C0.00C 10 . 0 0 0 0.0 O.G 4G0.0CC 9 . OCC O.C 0.0 0 .20C 290.OCC 6 9 . CCC 81.OCC 6.OCC 4.OCC 4.600 37.OOC 1063.OCC 1.170  197 47  156 48  199 48  200 49  385.000 0.0 0.0 0. 0 0.0 C. 0 0.0 0. 0 0.0 0.0 0.0 0.0 0. 0 0.0 0. c 0.0 0. 0 99.500 99.500 0. 0 0.0 0. 0 0.0 0.0 0.0 0. 0 0. c 0.0 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 8. 000 0.0 0. 0 1.000 3.000 0.030  273.000 0.100 0.0 0.0 0 .0 O.C 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0 .100 O.C 0.0 0.0 70.600 53 . 0 0 0 0.0 0.010 0.030 0 . 200 30.000 1.000 0.0 4.000 100.000 1.000 10.000 0.0 0.0 21 . 0 0 0 7.000 4.000 1.5CC 1.000 0.500 17.OOC 75.000 0 . 11C  290.000 O.C 0 .0 0. 0 0.0 0.0 0.0 0.0 O.C 0.0 0. 0 0.0 0. 0 O.C 0.0 0.0 0.0 75.000 4 .500 0.0 O.C 0.0 0.0 0.0 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 0.0 46.000 16.OCO 2.000 4 . 100 4.OOC 1.300 68.OOC 4.000 0 . C90  399.000 4.000 60.000 180.000 260.000 400.000 310.000 100.000 40.000 140.000 200.000 280.000 110.000 10.200 5.000 5.OOC 0.0 76.600 64.400 0.200 0.030 0 . 170 0.20C 20.000 4.000 0.0 0.0 0.0 5.000 10.000 40.000 0.300 148.000 122.000 0.0 1.400 7.CCO 1 .100 226.000 192.000 0.040  279  ITEM CLUSTER A T T R I B U T E GROUP NUTRIENTS: KCAL (GM) PROT TRY ( MG) THR (MG) ISO (MG) LEU (MG ) LYS (MG) MET (MG) CYS (MG) PHE (MG) TYR (MG) (MG) VAL HIS (MG) FAT-T (GM ) SFA (GM) PUFA (GM) (GM) CHOLE CHO-T (GM ) (GM) SUCR (GM ) CHO-F THI A (MG) R I BO (MG ) N I A C I N (MG) V I T - B 6 (MG) FOLIC (UG) VTT-B12(UG) (MG) VIT-C PANTO (UG) B I O T I N (MG ) VIT-A (IU) (ID VIT-D VIT-E (MG) OA (MG) P (MG) MG (MG) FE (MG) I (MG ) (MG) ZN NA (MG) K ( MG) CU (MG )  201 49  202 49  203 49  204 50  205 51  520.OCC 7.70C 40.OCC 90.OCC 120.OCC 240.CCC 130.OCC 20.OCC 80.OOC 160.OCC SO.CCC 140.OCC 40.OCC 32.3CC 19.OCC 12.OOC 0.015 56 . 9 0 0 43.OCC 0.4CC 0.06C 0.34C 0.300 20.OCC 8 .000 0.0 O.C 100.OOC 22.CCC 270.OCC 88.OCC 1.100 228.CCC 231.CCC 82.OCC 1 . ICC 14.OCC 2.6CC 94.000 3 8 4 . OOC 1 .000  385.000 0.0 0.0 0. 0 0.0 0. C 0.0 0.0 0.0 0. 0 0. 0 0.0 0. 0 0.0 0. 0 0.0 0.0 99.500 59.500 O.C 0.0 0. 0 0.0 0.0 0.0 0. 0 0. 0 0.0 0. 0 0.0 0.0 0.0 0.0 0.0 0.0 C. I C C 8.000 0. 0 1.000 3.000 0.030  273.000 0 . 100 0.0 0.0 0.0 0.0 0.0 0.0 O.C 0.0 0.0 0 .0 0.0 0.100 0.0 0.0 0.0 70.600 53 . 0 0 0 G.O 0 .010 0.030 0.200 30.000 1.000 0 .0 4.000 100.000 1.000 10.000 0.0 0.0 21 . 0 0 0 7.000 4.000 1.500 1.000 0.500 17.000 75.000 0.110  122.000 6. 100 50 . 0 0 0 260.000 350.000 520.000 450.CCC 60.000 20.000 310.000 180.000 370.000 200.000 2.60C 1 .000 l.OCC 0.001 19.000 3 .600 1.400 0.08C 0.030 0.600 380.000 10.000 0.0 2.000 100.OCO 6. 000 130.000 0.0 0 . 100 54.000 92.000 28 . 0 0 0 1.800 4.000 I .400 463.000 210.000 0.210  192.000 7.300 110.000 330.000 480.000 790.000 530.OOC 190.000 160.000 300.000 330.000 480.000 270.000 9 . 900 • 3.000 7.COO 0.02 0 18.000 0.300 0.100 0.C30 0. 060 1. 2 0 0 110.000 5.000 0.490 0. 0 400.000 2.000 410.COC 0.0 0 . 400 10.000 48.000 11.000 1. 0 0 0 3.000 1.000 366.000 93.000 0. 060  280  ITEM CLUSTER A T T R I E U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG) THR ( MG) (MG ) ISO (MG) LEU (MG) LYS MET (MG) CYS (MG) PHE (MG ) TYR ( MG ) (MG) VAL HIS (MG ) FAT-T (GM) (GM ) SFA (GM) PUFA CFXLE (GM ) CHO-T (GM) (GM) SUCR CHO-F (GM) THIA (MG) (MG) RIBC N I A C I N (MG ) V I T - E 6 (MG) (UG) FOLIC VIT-E12(UG) VI T-C ( MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D (IU) VIT-E (MG) (MG ) CA P ( MG) (MG) MG FE (MG) I (MG) ZN (MG) NA (MG) K (MG) (MG) CU  206 52  207 53  208 54  209 55  210 56  89.000 6.40C 80.000 260.OCC 470.OCC 590 .OCC 470.OCC 150.OCC 110.OCC 350 .000 250.OCC 360.OCC 690.OCC 4.3CC 2 .OCC 2.000 0.020 6.2CC 0 .300 0.4CC 0.06C 0.07C 1.9 CO 110.000 3 . CCC 0 .65C 7.OOC 5 CO.CCC l.OCC 980.OOC 0.0 0.3CC 12.000 75.OOC 20.CCC 1.20C 3.OCC 1 .OCC 37.OCC 250.000 0.02C  38.000 2.600 30.000 90.000 110.OOC 170.000 190.OCC 60.000 30.000 490.000 70.000 100.000 60.000 0 . ICO 0.0 0. 0 0.006 7.100 1.000 0.300 0. C2C 0.040 0.4C0 180.000 5.000 0.660 5. 000 500.000 2.000 60.CCO 0.0 0. 0 18.000 34.000 18.000 0. 500 3 . OCO 0.500 290.000 167.000 0. 110  102.CCC 12.400 120.OOC 440.000 540.COO 800.000 900.COO 270.000 130.000 260.000 360.000 500.000 300.000 4.000 1.000 2.000 0.C2C 4.000 0.50C 0.300 0.030 0.090 1.700 180.000 5.000 0.660 4.000 5C0.000 5.000 110.000 0.0 1.200 23.000 117.000 18.000 I.000 . 4.000 1.900 287.OOC 189.000 0.190  133.000 7.500 70.000 300.000 410.000 630.OCC 620.000 140.000 90.000 360.000 270.000 430.000 240.000 6.100 3.OOC 3.000 0.015 12.200 8. 900 0.600 0 . 030 0.070 1.300 100.000 9.000 0.230 2 .000 100.000 2.000 60.000 0.0 0.200 32.000 126 . 0 0 0 26.000 1 .700 5.OOC 1.8C0 531.000 233.000 0 .330  173.000 12.800 160.000 520.000 660.000 940.000 960.000 300.000 180.000 100.000 400.000 500.000 300.000 8.500 3 . 000 4.000 0.050 11.300 1.200 0.400 0.070 0 . 18C 5.200 300.OOC 6.000 0. 220 4.000 500.000 8.000 590.000 0.0 0.200 41.000 145.000 19.000 1.200 5.000 2.500 344.000 112.000 0.220  281  ITEM CLUSTER A T T R I E U T E GROUP NUTRIENTS: KCAL (GM) PROT TRY (MG) THR (MG) ISO (MG ) LEU (MG > (MG) LYS MET ( MG) (MG) CYS PHE (MG) TYR (MG) VAL (MG) (MG ) HIS FAT-T (GM) SFA (GM) (GM) PUFA CHOLE (GM) CHO-T (GM ) SUCR (GM) CHO-F (GM) THIA (MG) RIBO (MG) N I A C I N (MG) V I T - B 6 (MG) (UG ) FOL IC VIT-E12(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A (IU) VIT-D ( IU) VIT-E (MG) CA (MG) P (MG ) MG (MG) (MG) FE (MG) I (MG) ZN NA (MG) K (MG) (MG) CU  211 57  212 58  213 59  214 6C  131.OCC 8.OCC 80 .OCC 250.000 460.000 580.OOC 460 .000 140.CCC 110.OCC 330.OCC 2 5 0 . OCC 3 7 0 . OOC 680.OCC 6 .700 3 . OCC 3.CCC 0.02C 9.8CC 1 .40C 0.3CC 0.100 0 . 140 1 .700 230.OCC 7.OCC 0.36C 4 . OCC 300.000 2.OOC 4 3 0 .OOC 0.0 0.2CC 19.000 117.CCC 19.000 1 .3CC 4.000 1.900 393.OCC 115.OCC 0.17C  112.000 8. 4 0 0 100.000 5C0.OCO 630.000 920.000 1090.000 330.000 1 7 C . OCC 340.000 130.OCO 590.000 320.000 3.000 1. 0 0 0 1.000 0.020 1 2 . 7 00 1.200 0. 300 0 . 0 70 0.090 2.300 250.000 10.OCO 0.210 4 . CCC 600.000 4.000 130.000 0.0 C. 2CO 26.000 87.000 21.000 1. 100 3.000 1.200 400.OCC 176.000 0 . 14 C  181.OCO 8.800 100.000 390.000 460.COC 720.000 780.000 220.000 110.000 360.000 3C0.000 490.000 310.000 11.300 5 .000 5.000 0.020 1C.7C0 0.100 0.50C 0.C10 0.090 2 . 100 80.000 9.000 0.920 10.000 500.000 2.000 10.000 0.0 0 . 100 13.000 67.000 19.000 2.000 4.COO 1.200 540.000 200.000 0.140  215.000 8.400 170.000 520.000 750.000 1110.000 730.COO 280.000 170.000 440.000 560.000 860.000 370.000 11.100 5 .000 6 . OCO 0.040 20.100 0 . 100 0.100 0 . ICO 0.200 0. 900 40.000 6. 000 0.35C 0.0 200.000 2 .000 430.000 20.000 0.500 181.CCC 161.000 26.000 0.900 5 . OCO 0.300 543.000 120.OCC 0.040  215 61 245.000 9.500 160.000 400.000 690.000 1090.000 590.000 240.000 160.000 490.000 520.000 700.COO 330.000 7 . 100 2.000 3.000 0.040 35.400 3.900 0.300 0. 060 0.170 1.000 50.000 3.000 0 . 200 6. 000 300.000 1.000 440.000 8 .000 0.300 156.000 156.000 27.OOC 0.900 9.000 I. 100 647.000 114.000 0.340  282 ITEM C L U S T E R A T T R I E U T E GROUP NUTRIENTS: KCAL PROT (GM) TRY (MG J THR (MG) ISO (MG) LEU (MG) LYS (MG) MET (MG) CYS (MG ) PHE (MG) TYR (MG) (MG) VAL HI S (MG) FAT-T (GM) SFA (GM) PUFA (GM ) CHCLE (GM) CHO-T (GM) (GM) SUCR CHO-F (GM) TH IA (MG) (MG) RIBO N I A C I N (MG > V I T - B 6 (MG) FOLIC (UG) VITrei2(UG) VIT-C (MG) PANTO (UG) B I O T I N (MG) VIT-A ( IU) VIT-D (IU) VIT-E (MG) OA (MG) P (MG > MG (MG) FE (MG) (MG ) I ZN (MG) NA (MG ) K (MG) (MG) CU  216 62  217 63  218 64  76.OCC 2.2CC 30.OOC S O . OCC 110 . 0 0 0 150.CCC 70.OCC 30.OCC 4 0 . OOC 1 1 0 .OOC 70.OCC 130 .000 50.OCC 0.6CC O.C O.C 0.006 15.4CC 5.50C 0.2CC 0.140 0.11C 1.8CC 50 .OCC l.OCC 0.250 4.OCC 3 0 0 .OOC 0.0 370.OCC l.OOC 0.4CC 16.OCC 35.OCC 11.000 1 . IOC 5.CCC 0.10C 382.OCC 121.000 0.12C  134.000 7.500 80.000 300.000 350.000 530.000 520.CCC 160.000 100.OCO 310.000 220.000 350.000 220.000 4.700 2.000 3 . CCC 0.020 15.6CC 4.200 0. 300 0. 100 0.120 1.600 150.000 6 . CCC 0.220 9.000 200.000 1.000 6 4 0 . OCC 0.0 0.3CC 50.000 95.000 17.000 I. 5 0 0 3.000 1.400 4C7.0CC 268.000 0 . 17 0  140.000 4.50C 70.OCC 120.000 150.000 230.000 240.000 70.000 40.000 130.000 100.000 150.OOC 100.000 7 . 100 3.000 3.COO 0.C09 14.200 0.0 0. 0 O.C 0.0 0.0 200.000 1.000 0.0 0.0 400.OOC 10.000 0.0 O.C 0 . 100 20.000 39.000 9.000 1 .200 3.000 0 .900 665.000 0.0 0.050  219 65  220 66  120.000 O.C 20.000 0.0 230.000 0.0 880.000 0.0 1040.000 0.0 1630.000 0.0 1740.COO 0.0 490.000 0.0 0.0 250.000 0.0 340.CCO 680.000 0.0 1110.000 0.0 690 .000 0.0 3 . 000 0.0 I.000 0.0 1.000 0.0 O.C O.C02 5.000 0.0 0.0 0.0 1 .100 0.0 0.0 0.010 0.230 0.0 0.0 11.400 0.0 O.C 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 .0 0. 0 0.0 0.0 0.0 0.0 0.100 O.C 253.OOC 200.000 297.000 190.000 57.OOC 0 . 100 4.600 1 0 0 0 0 .000 43.000 0 . 500 0.800 2400.000 38758.000 4.OOC 100.000 0 .380 0.090  ITEM CLUSTER A T T R I E U T E GROUP NUTRIENTS : KCAL PROT TRY THR ISO LEU LYS MET CYS PFE TYR VAL HIS FAT- T S FA PUFA CHCLE CHO- T SUCR CHO- F  (GM) (MG) ( MG) (MG) (MG) (MG) (MG) (MG) (MG ) (MG) (MG ) (MG) (GM) (GM) (GM) (GM ) (GM) (GM) (GM)  THIA (MG) RIBC (MG) N I A C I N (MG) V I T - B6 (MG) FOLIC (UG ) VIT- E12(UG ) V I T -C (MG ) PANTO BIOTIN VIT- A VIT- D VIT- E CA P  (UG) (MG ) (I U ) (IU) (MG) (MG )  MG FE I ZN  (MG) (MG) (MG ) (MG) (MG)  NA K CU  (MG) (MG) (MG )  221 67  3 9 2 .OOC 9.4CC 40 . 0 0 0 110.OCC 150.CCC 3C0 .OOC 160.OCC 20 . 0 0 0 ICO.000 200.000 110.OOC 1 9 0 . CCC 50 . C C C 10.6CC 6 .000 4.OCC 0.0 73 . 9 C C 12. C C C 0 .800 0 . C8C 0.410 0.5CC 20.OCC 80 . O C C 0.0 1.000 ICO.OOC 32 .CCC 10.OCC 1C60.C0C 0 .4CC 275.OCC 290 .OOC 371.OCC 1.4CC 8 .CCC 2.60C 382 .OCC 605.OCC 3.69C  284  APPENDIX D ABRIDGED FOOD-COMPOSITION FILE  Listing of 22 nutrients selected from the food-composition f i l e (Appendix C) for each of the 127 item clusters designated in Appendix B.  ENERGY (kcal) PROTEIN  (gm)  TOTAL FAT (gm) SATURATED FAT (gm) POLYUNSATURATED FAT (gm) TOTAL CARBOHYDRATE (gm) SUCROSE  (gm)  FIBER (gm) THIAMIN  (mg)  RIBOFLAVIN  (mg)  NIACIN (mg) PYRODOXINE FOLACIN  (ug)  (ug)  ASCORBATE (mg) RETINOL  (iu)  CHOLECALCIFEROL TOCOPHEROL (mg) CALCIUM (mg) PHOSPHORUS (mg) MAGNESIUM  (mg)  IRON (mg) POTASSIUM (mg)  (iu)  286  APPENDIX E NUTRIENT-LIMITS FILE  Tables of daily minimum (Table E-l) and daily maximum (Table E-2) nutrient limits disaggregated for age, sex, activity pattern, and pregnancy status.  287 Table E - l . Minimum nutrient limits Code #  Sex  Age  111 112 113 114 121 122 123 124 131 132 133 134 141 142 143 144 211 212 213 214 221 222 223 224 231 232 233 234 241 242 243 244 205 206 207  Male  19-35  36-50  51-65  66 +  Female  19-35  51-65  66 + Preg. Preg. Lact.  .ab Protein Activity Energy' Pattern kcal/kg /day gm/kg/day  aef  42.6 40.6 35.7 32.6 41.4 39.3 34.4 D 31.3 40.2 A B 38.0 33.3 C D 30.1 A 38.9 36.9 B C 32.0 D 28.6 A 37.3 35.2 B C 33.8 D 30.7 36.3 A 34.1 B 32.7 C D 29.6 A 35.0 32.9 B C 31.4 D 28.4 A 33.9 B 31.8 30.4 C D 27.3 1st + 100 kcal/day 2nd + 3rd + 100 kcal/day + 500. kcal/day A B C D A B C  Histidine mg/kg/day  .57  .10  fh  Isoleucine mg/kg/day  fi  .10  .52  * 20 gm/day + 20 gm/day + 24 gm/day  Values for minimum limit (and for recommended energy intake) are obtained from: Bureau of Nutritional Sciences, Department of National Health and Welfare. Dietary Standard for Canada. Information Canada, Ottawa, 1975. To calculate lower and upper limit on energy intake per day for an individual, the recommended intake for a given age, sex, and activity level is multiplied by kilograms ideal body weight, then, the multiplied figure is summed with the additional caloric requirements for pregnancy or lactation, and lastly, this figure is reduced or raised five percent. For longer period of time, the daily limits are multiplied by the number of days considered. Body weight is measured as kilograms ideal body weight.  288 Table E - l . Minimum nutrient 1imits (cont'd) Code #  111 112 113 114 121 122 123 124 131 132 133 134 141 142 143 144 211 212 213 214 221 222 223 224 231 232 233 234 241 242 243 244 205 206 207  • fi Leucine mg/kg/day 14  1  Lysine^ ' mg/kg/day 12  Methionine fi + Cystine mg/kg/day 13  Phenylalanine + Tyrosine mg/kg/day  fi  Threonine mg/kg/day  fi  14  Where increased minimum allowance associated with pregnancy and lactation is not stated, the increase, i f any, is calculated on the basis of increased body weight or caloric requirement. Maximum intake limits for pregnant and lactating females are the same as for nonpregnant females. Minimum protein limit for the abridged nutrient consumption f i l e (Appendix F) is 0.80 gm/kg/day for males, and 0.73 gm/kg/day for females. This figure is based on the average amino acid composition of the Canadian diet as stated in the Dietary Standard For Canada, Bureau of Nutritional Sciences, Department of National Health and Welfare, Information Canada, Ottawa, 1975.  289 Table E-1 Code #  Minimum nutrient limits (cont'd)  Tryptophan mg/kg/day  1  1  Valine mg/kg/day  a:i  Fat %kcal/day  k  Saturated^ P/S Fat Ratio  Polyunsat. Fat %kcal/day  aj  %kcal/day 111 112 113 114 121 122 123 124 131 132 133 134 141 142 143 144 211 212 213 214 221 222 223 224 231 232 233 234 241 242 243 244 205 206 207  3.5  10 0  For determining limits on intake per day, multiply the table value by the individual's ideal body weight in kilograms. Additional minimum protein requirement for pregnancy and lactation based on the composition of the average daily protein intake of Canadians as per the Dietary Standard for Canada, Bureau of Nutritional Sciences, Department of National Health and Welfare, Information Canada, Ottawa, 1975. Value for minimum limit on histidine obtained from: cation, Dr. P . J . Stapleton, 1978.  personal communi-  Values for minimum limit obtained from: FA0/WH0, Ad Hoc Expert Committee. Energy and Protein Requirements. WHO Tech. Rep. Ser. No. 522, Geneva, 1973.  • 29Q Table E - l . Minimum nutrient limits (cont'd) Code #  111 112 113 114 121 122 123 124 131 132 133 134 141 142 143 144 211 212 213 214 221 222 223 224 231 232 233 234 241 242 243 244 205 206 207  Cholesterol mg/day  9  1  Carbo-J hydrate % kcal/day  Jm  0  Sucrose Fiber" % kcal/day gm/100 kcal/day  55  0.4  30