QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dixon, L. B. Articles by Ernst, N. D. Search for Related Content PubMed PubMed Citation Articles by Dixon, L. B. Articles by Ernst, N. D.

---

Supplement

Choose a Diet That Is Low in Saturated Fat and Cholesterol and Moderate in Total Fat: Subtle Changes to a Familiar Message¹

Division of Cancer Control and Population Sciences, National Cancer Institute, National Institutes of Health, Bethesda, MD and ^* Division of Clinical Applications and Epidemiology, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD

²To whom correspondence should be addressed. E-mail: ld120i{at}nih.gov.

	ABSTRACT

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
"Choose a diet that is low in saturated fat and cholesterol and moderate in total fat," issued in Nutrition and Your Health: Dietary Guidelines for Americans in the year 2000, has an interesting and lengthy history. The first guideline, for which there was extensive scientific data to show that dietary excess increased chronic disease risk, prompted much scientific discussion and debate when implemented as dietary guidance. Three major changes in the guideline are noted since it was issued in 1980, i.e., numerical goals for dietary fats; the applicability of recommended fat intakes for all individuals 2 y old; and rewording to emphasize reducing saturated fat and cholesterol intakes. The shift in emphasis includes the terminology moderate fat, which replaces the phrasing low fat. National data about the food supply, the population’s dietary intake, knowledge, attitudes and behaviors, and nutritional status indicators (e.g., serum cholesterol levels) related to dietary fats help to monitor nutrition and health in the population. Experts consider that national data, although not without limitations, are sufficient to conclude that U.S. intakes of fats, as a proportion of energy, have decreased. The lower intakes of saturated fat and cholesterol are consistent with decreases in blood cholesterol levels and lower rates of coronary mortality over the past 30 years. Strategies are needed and some are suggested, to further encourage the population to achieve a dietary pattern that is low in saturated fat and cholesterol and moderate in total fat. Other suggestions are offered to improve national nutrition monitoring and surveillance related to the guideline.

KEY WORDS: • cardiovascular • cholesterol • dietary fat • dietary guideline • saturated fat

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Historically, the Dietary Guideline (hereafter referred to as the guideline) to "choose a diet that is low in saturated fat³ and cholesterol and moderate in total fat," has evoked substantial interest and even controversy. This paper outlines the history of this guideline, discusses surveillance methods available to assess the intake of dietary fats⁴ and associated health outcomes in the U.S. population, presents data showing how the U.S. population fares today compared with the past and concludes with strategies for improving the dietary fat intakes of the U.S. population.

	Dietary fats and chronic disease risk

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Dietary fat was recognized as an important organic macronutrient in 1827 (1) . The first recommendations related to dietary fat were published in 1894 by W. O. Atwater, a scientist at the USDA who recommended that 33% of energy come from fat, 52% from carbohydrates and 15% from protein (2) . The earliest evidence linking diet and atherosclerosis was obtained when an animal-experimental model of human atherosclerosis was produced by feeding eggs, milk and meat to rabbits in the early part of the 20th century (3) . It was not, however, until the 1950s that the findings from such studies were considered generally applicable to humans. Today, we recognize that the major dietary principles that link dietary lipids and blood cholesterol are anchored by clinical studies reported during the 1960s (4) . Investigators such as Connor et al. (5 ,6) , Keys et al. (7) and Hegsted et al. (8) established the hypercholesterolemic effect of dietary cholesterol in humans. The serum cholesterol-raising effect of saturated fatty acids (SFA)⁵ was established by Keys et al. (9) and Ahrens et al. (10) . Probably the earliest study in humans to conclude that a high fat diet is not always associated with an increase in serum cholesterol was reported in 1952 (11) .

The elucidation of the basic relation between diet and blood lipids culminated in the seminal studies of Keys et al. (7) and Hegsted et al. (8) . These two nutrition scientists constructed equations to quantify the changes in blood total cholesterol produced by dietary change. In brief, SFA raise serum cholesterol twice as much as polyunsaturated fatty acids (PUFA) lower it. These equations remain reasonably predictive of the serum cholesterol response of populations to shifts in diet (12) .

More than four decades ago, a population study report compared the diet, serum cholesterol and coronary heart disease (CHD) rates in the U.S. with those in other countries (13) . Positive relationships were noted among dietary patterns, rich in total fat and SFA, and serum cholesterol and CHD mortality rates within and across populations. In 1970, the Seven Countries Study confirmed and extended these findings (14 ,15) .

Evidence for the effectiveness of blood cholesterol-lowering therapy in reducing the incidence of heart attack and other clinical sequelae of CHD was first provided in January 1984 with the Lipid Research Clinics Coronary Primary Prevention Trial (16) , and subsequently by numerous large, randomized trials using a variety of cholesterol-lowering diets, drugs and other interventions (17 18 19) .

By the 1980s, several authoritative histories reflected scientific agreement that supported a relation between dietary total fat and certain types of cancer, especially cancer of the breast, colon and prostate (20) . Evidence also suggested an association between dietary fat and the development of obesity in animals and possibly in humans (19) .

	Emergence of dietary guidance on fat intake

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
It was from the perspective of this accumulating scientific evidence linking dietary fats and disease that the impetus arose to recommend a lowering of fat intake by Americans. The Dietary Goals for the United States, published by the U.S. Senate Select Committee on Nutrition and Human Needs (21) , suggested numerical goals, including reduction of total fat intake to 30% of energy, 10% of energy from SFA and 300 mg/d of cholesterol. Similar targets for intake of fats were first promoted by the American Heart Association (AHA), a voluntary health association (22) .

A recent historical account of the development of the Dietary Guidelines, presented by Michael McGinnis at a Dietary Guidelines Advisory Committee (DGAC) meeting in 1998, suggests that the Dietary Goals for the U.S., and in particular their quantified nutrient targets, prompted the two departments, the USDA and the Department of Health and Human Services (DHHS), to consider descriptive (but not numerical) Dietary Guidelines for the American people (23) .

Several national and expert scientific committee activities have influenced Federal recommendations regarding advice to lower intake of total fat, SFA and cholesterol. These include the consensus of a Task Force sponsored by the American Society for Clinical Nutrition that evaluated the strength of evidence for the relationships among dietary fats, dietary cholesterol and heart disease (24) , the furor created by the National Academy of Science Report on Healthful Diets (25) , national prevention goals disseminated through the DHHS Health Objectives for the Nation (26 27 28) , the National Cholesterol Education Program (NCEP) (29) , The Surgeon General’s Report on Nutrition and Health (17) , the National Research Council’s report on Diet and Health: Implications for Chronic Disease Risk (19) and the Nutrition Labeling and Education Act (NLEA) of 1990. In addition, the nutrition-related research of the NIH on diet and chronic diseases has continued to grow exponentially in concert with federal efforts to reduce the nation’s burden of chronic diseases. Simultaneously, consumers have shown an ever-increasing interest in nutrition and health. These activities are interwoven and interact; it is not possible to separate the independent effect of each one.

Table 1 summarizes how the recommendations for dietary fats have changed over time. Anumber of perspectives on the history of these recommendations are available (30 31 32 33 34 35 36 37 38 39) .

	Dietary Guidelines for Americans: Advice on dietary fats

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

In this paper, the primary sources of the rationale for changes in the guideline are the DGAC reports (40) and the transcripts of the public meetings (23) .

	Major changes in the guideline for dietary fats

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
At the end of two decades, 1980–2000, three notable changes in the guideline are evident: the numerical goals for dietary fats; the age of applicability of recommended fat intakes for children; and the rewording of the statement for the guideline.

Numerical goals for dietary fats.

The 1990 DGAC report indicates agreement with numerical goals for dietary fats with the conclusion that widely based population data supplemented the evidence from clinical trial research, which at the time had been conducted predominantly in white males. The totality of the evidence was such that the goals were considered to be appropriate for the total population, not just persons at risk for heart disease.

The 1990 DGAC concluded that the potential benefits in reducing risk of CHD and some types of cancer by reduction in total fat, SFA and cholesterol were sufficient to support numerical goals for all healthy adults. The numerical goals for total fat and SFA related to the two food components considered to be of greatest significance to the health of Americans. The goals adapted, those of the National Research Council, the NCEP, and by history, the AHA, are total fat 30% of energy and SFA < 10% of energy.

The 1990 DGAC noted that the exact numerical goals reflect both science and pragmatism. The numerical goal for SFA is based on epidemiologic evidence that populations with such intakes have less CHD and on a large body of evidence from many types of studies that reducing intake of SFA lowers blood cholesterol and CHD risk. The 1990 DGAC report noted the virtually unanimous agreement for SFA < 10% of energy as the appropriate level. Less than 10% of energy is advised for practicality of U.S. dietary habits, consistency with other expert recommendations and comprehensibility if dietary guidance is provided.

The numerical goal for total fat is based on expected beneficial effects. A reduction of total fat would facilitate reduction of SFA and benefit CHD reduction; a limit on total fat was thought to be useful to reduce the risk of some types of cancer and help manage energy intake, thus helping to curb overweight and obesity.

The policy to base recommendations on reduction of fat intake as a proportion of energy intake, rather than absolute intake (g/d), is consistent with historical practice. In 1968, an AHA report stated that the desirable intake of fat for the general U.S. population was 30–35% of energy, of which one third was to be SFA, one third monounsaturated fatty acids (MUFA) and one third PUFA (22) . In subsequent statements, the Nutrition Committee of the AHA recommended that SFA be < 10% of total energy, that PUFA be < 10% of total energy and the remainder be supplied by MUFA (31) . Establishing dietary guidance for fat intake as a proportion of energy also follows clinical dietetics, e.g., dietary management of hyperlipoproteinemia (41) . There was no historical precedence to suggest that the DGAC considered recommending grams of fat per day.

A recommended numerical goal for dietary cholesterol has been suggested indirectly since 1995 by reference to the Daily Value (DV) on the Nutrition Facts label (i.e., 300 mg/d). The 2000 DGAC’s public discussion and report note that SFA and cholesterol occur together in the foods that are major sources of SFA.

Age for use of numerical goals in children.

A second major change in the guideline is the advice on the age in childhood at which the numerical recommendations of dietary fats are relevant. The 1980 and 1985 Dietary Guidelines are generally considered appropriate for all healthy Americans. The 1990 edition is the first to state that the Dietary Guidelines are advice for healthy Americans 2 y old, not for younger children and infants. The guideline on dietary fats restates the introduction, saying that the goals for fats are not for children <2 y old.

The 1990 DGAC report notes the decision to use the numerical goals for children, stating that evidence indicates that children 2 y old can grow and develop normally when consuming 30% of energy as fat. The DGAC concluded from the body of evidence that health advantages would probably occur among children through guidance for low fat (30% of energy) dietary patterns without being unduly restrictive. Thus, the DGAC recommended flexibility in applying the numerical goals to children, recognizing the importance of a diet to support growth and the need to develop acceptability for a lower fat diet among children.

The evidence to support a reduced intake of fats, especially SFA and cholesterol, is documented in the NCEP report that focuses on the treatment of high blood cholesterol levels in children (42) . Members of the 1990 DGAC who represented the pediatric community were knowledgeable about this report. The major points include the following: 1) entire populations have low intakes of SFA and cholesterol and total fat, and these populations have low levels of serum cholesterol and CHD; 2) the children are healthy and intelligent; 3) CHD risk factor levels track from childhood into adulthood; and 4) food purchases and preparation feed the whole family, i.e., children and adults. Also, the 1990 DGAC report noted that the NCEP was preparing recommendations for children, and that new legislation required the USDA and DHHS to develop a publication, "Nutrition Guidance for Child Nutrition Programs," and apply this guidance to child nutrition programs.

In 1995 the DGAC report stated that in keeping with recommendations by other authorities, the guideline applied to children 2 y old. For logical and practical reasons, the DGAC recommended gradual adoption of the guideline from age 2 to 5 y, so that by the time children were in school, they should be consuming diets that follow the Dietary Guidelines. The 1995 DGAC report also stated that the advice was appropriate for health reasons and would enable uniformity in directions for National School Lunch Programs. The report noted a recent study, the Dietary Intervention Study in Children, indicating no adverse effect among children, ages 8–10 y, of a diet that contained 28.5% of energy from total fat, 10% from SFA and 90 mg cholesterol/1000 kcal (43) .

The 2000 guideline changes the recommendation back to the advice of 1990, clearly stating that the recommended intakes of dietary fats apply to children beginning at age 2 y. The DGAC concludes that recent studies among children ages 8–10 y (43 ,44) and children ages 7–36 mo (45) support the safety of diets that are low in SFA and cholesterol for children who are 2 y old,. Views in support and in dispute of this recommendation have been expressed (46 47 48 49 50 51) .

Statement wording.

The third major change in the guideline is the statement wording. The 2000 language is the first national recommendation to prioritize type of fat over amount of fat. The wording, choose a diet low in saturated fat and cholesterol and moderate in total fat, reflects the DGAC’s conclusion that for total fat, the evidence of causality of chronic disease is less conclusive than it is for SFA, cholesterol and CHD.

The 2000 DGAC report provides several explanations for the word moderate in total fat vs. low in total fat. First, a change in perception in the meaning of the two terms is recognized. The report suggests that the term high fat refers to total fat of 40% of energy, moderate fat to 30% and low fat to 20%. Second, the DGAC cites evidence that suggests that low fat diets (20% of energy) are by necessity high in carbohydrate and may cause an unfavorable pattern characterized by low HDL and elevated triglycerides. Also, an excessive carbohydrate intake might enhance postprandial response in glucose and insulin concentrations that may predispose persons with insulin resistance to type 2 diabetes mellitus. In contrast, moderate fat intakes allow more unsaturated fatty acids, which, along with low SFA, are associated with reduced risk of heart disease. A final argument for using the word moderate fat is to move away from an incorrect viewpoint that a low total fat intake per se (as opposed to making additional changes to develop healthful dietary patterns) will result in a healthful diet.

An emphasis on type of fat reflects the DGAC’s conclusion that the weight of evidence concerning the contribution of a high fat diet to cancer incidence is not presently sufficient to justify promoting low fat intake for the U.S. population. Also, the DGAC’s conclusion reflects their persuasion of the theory that a high percentage of fat contributes to obesity and may warrant placing a ceiling on total fat intake, but is not sufficient in itself to warrant a definitive recommendation for a low fat diet. The use of the term moderate is part of a strategy that is intended to shift the emphasis away from total fat. Whether consumers make appropriate food choices reflecting this subtle shift in message remains to be seen.

	Other message changes in the guideline

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Several additional changes in the guideline for fat are observed with close review of the wording in the Dietary Guidelines brochures. These changes may be viewed as three categories of guideline messages: 1) What are important sources of SFA, cholesterol and total fat? 2) Why is there a health concern? and 3) How can consumers reduce their intake of fat, SFA and cholesterol? These issues are discussed in detail in the Appendix.

	SURVEILLANCE METHODS TO ASSESS THE GUIDELINE

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Data from the National Nutrition Monitoring and Related Research Program are used to develop the Dietary Guidelines for Americans, and to monitor the trends and progress toward achieving the guidelines (52 53 54) . These include data from the food supply, and national surveys of food and nutrient intake. National surveys also provide data about knowledge, attitudes and behaviors, and nutritional status indicators (e.g., blood cholesterol) related to dietary fats and other nutrient and food intake. The sources of the data and issues related to the interpretation of population estimates of and trends in fat intake are discussed below.

Food supply data

Estimates of foods and nutrients available for consumption from the food supply, collected by the USDA since 1909, are updated and published annually in the U.S. Food and Nutrition Supply Series (53) . Annual estimates of foods available for consumption are determined from the difference between the total available supply, which is the sum of production, inventories from the past year, and imports of food, and foods that are exported, used by industry or as farm inputs, or in the current year-end inventories (55) . These estimates, calculated in pounds per person per year, are often referred to as food disappearance data because they represent the disappearance of food into the market place. Estimates of nutrients from food available for consumption are calculated by multiplying the nutrient value per pound by the quantity of each food (in pounds) and summing across all foods. These food and nutrient data are used as proxies to estimate human consumption. However, the data overestimate consumption because they include inedible refuse (e.g., bones), ingredients in processed foods that are exported (e.g., soft drinks, baked goods, cereal products), and amounts lost in processing, marketing or spoilage, or foods thrown away or fed to pets (55) .

The use of food supply data to monitor fat intakes of the population has been strongly cautioned, primarily because of the unaccounted waste (56) . Over the past two decades, the waste portion of fats and oils has increased due to growth in away-from-home eating places, in particular, fast-food places (57) . Amounts of fat used for frying and discarded from foodservice operations have been estimated to be 50% (58) . Estimations of fat intake from food supply data are further complicated by the production of fat substitutes made from fats (e.g., olestra made from vegetable oils) that are consumed but not digested (59) . Recently, the USDA has developed methods to adjust the data for spoilage and other waste including retail, food service and consumer losses, changes in weight due to cooking and the discarding of nonedible food parts to more closely reflect intake data (60) .

Survey data of food and nutrient intake

Several major surveys help monitor food and nutrient intakes (Table 2 ). The primary dietary assessment tool used to monitor intake is the 24-h dietary recall. Multiple 24-h recalls are considered to be the "gold standard" for nutrition monitoring (61) and will produce stable estimates of mean nutrient intakes from groups of individuals. Provided that the sample size is large enough, single 24-h recalls will also produce reasonably accurate estimates of mean nutrient intakes from groups of individuals (62) . However, if additional 24-h recalls have been collected from a representative subsample of participants, methods that adjust for the within-person day-to-day variation in dietary intake from one 24-h recall are encouraged (63) . Such methods have been applied to analyses of 1-d dietary data from the Third National Health and Nutrition Examination Survey (NHANES III) using data from a second 24-h recall collected from 5% of participants (64) .

In addition, 24-h recalls have limitations related to systematic bias in underreporting of energy intake (67) and possibly underreporting of foods containing fat (68) . Methods have been developed to assess the degree of underreporting in national survey data (69 ,70) , allowing for comparisons of dietary intakes of low energy reporters with those of others (71) .

Measuring change in dietary intakes is further complicated by differences in procedures and databases used in national surveys (62 ,72) . Interpretation of change in dietary intakes across national surveys requires knowledge of survey design characteristics, i.e., sampling frames, response rates of the surveys, age ranges of respondents, number of days of dietary data collection, days of the week selected for dietary data collection, and interviewing and coding methods. Comparisons of fat intakes are particularly affected by the content of the food and nutrient databases. For example, USDA and manufacturers’ composition data were used in NHANES I and II, whereas an updated version of the USDA database with increased numbers of ethnic and brand name foods was used in NHANES III (73) . Also, the procedure for estimating fatty acid intake has changed. The 1977–1978 National Food Consumption Survey nutrient composition tables included all PUFA, whereas NHANES I and II nutrient composition tables included the major PUFA, linoleic acid (12 ,74) . Differences in survey estimates of dietary cholesterol may reflect changes in the cholesterol content of eggs as well as improvements in analytic technology to assess cholesterol content of animal products in the late 1980s (75) . Improvements in measurements of intake (e.g., including the contribution of alcohol to energy intake, using additional probes to improve food recall) are also factors to consider when evaluating trends in dietary data expressed as a percentage of total energy. Many of these differences should be resolved with one national survey of dietary intake administered on a continuous basis (76) .

In 2000, the guideline for dietary fats recommends that consumers also reduce their intake of trans-fatty acids; these were not available in the national food composition database until recently. In 1995, the USDA released a special data set that provides g/100 g trans-fatty acids in 214 foods (77) . These foods are major sources of trans-fatty acids in the U.S. population. Designated food codes allow linkage to the national surveys of dietary intake (78) . Such analyses are limited by the number of foods that have values for trans-fatty acids and require judgements when matching codes of these 214 foods to codes of many similar, but not identical, foods in the national surveys. However, previous estimates of trans-fatty acid intake in the U.S. are from Food Supply Data or food frequency data and are less precise.

Survey data of knowledge, attitudes, and behaviors

The Diet and Health Knowledge Survey (DHKS), administered to a subsample of participants in both the 1989–1991 Continuing Survey of Food Intakes by Individuals (CSFII) and the 1994–1996 CSFII, provides information about consumers’ knowledge, attitudes and behaviors related to total fat, SFA and cholesterol (79 ,80) . Respondents were asked whether they were aware of health problems related to fat, SFA and cholesterol, how important it is to avoid too much fat, SFA and cholesterol (which was changed to how important it is to choose a diet low in fat, SFA and cholesterol when the wording of the guideline changed in 1990), and whether they thought their diets should be lower or higher or were just right in terms of the amounts of fat, SFA and cholesterol.

Survey data of nutritional status indicators

NHANES I, II and III provide national data on nutritional status and health indicators, such as serum cholesterol and lipoprotein levels. Methodologies used to measure serum lipids and lipoproteins differ across surveys, but all have been standardized to the criteria of the Centers for Disease Control and Prevention (CDC) or the CDC-National Heart, Lung, and Blood Institute Cholesterol Standardization Program (81) .

	RESULTS FROM SURVEILLANCE DATA

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Survey data of food and nutrient intake

Dietary intake data from the 1994–1996 CSFII provide current population estimates of dietary fats. These 1-d dietary data show mean intake of total fat as 32.8% and SFA as 11.3%, as a proportion of energy, and mean cholesterol intake as 256 mg for all participants (Table 3 ). Absolute mean intakes of total fat and SFA intake are 74 and 26 g, respectively (Table 3) . These values correspond to about half of the estimated consumption derived from food supply data in 1994 (i.e., 159 g of total fat and 52 g of SFA) (Table 4A). This is not surprising, given that at least half of fats and oils in the U.S. food supply are estimated to be discarded before consumption (58) . Dietary data from the 1989–1991 CSFII, in combination with USDA food data on trans-fatty acids, show the mean intake of trans-fatty acids as 5.3 g, accounting for 7.4% of total fat and 2.6% of total energy, for participants 3 y old (Table 5 ).

View larger version (24K): [in this window] [in a new window]   Figure 1. Trends in fat intakes, 1971–1996, and the "Healthy People 2010" objective. The national Health and Nutrition Examination Survey (NHANES) series and the Nationwide Food Consumption Survey (NFCS), together with related USDA surveys, show a decline in the percentage of kilograms from fat, although the NHANES trend appears to be more gradual. Data for the surveys are as follows: 1) 1-d data from individuals 2–74 y old (unpublished data); 2) 1-d data from individuals of all ages (109) ; 3) 1-d data from individuals ages 6 mo-74 y (unpublished data); 4) 1-d data from individuals 2 mo old (110) ; 5) 1-d data for individuals 2 y old (111) ; and 6) 1-d data for individuals 2 y old (112) . Adapted from Ref. 113 .

Food supply data from 1970 to 1996 provide information about changes in major food sources of dietary fats, including meats, eggs, dairy foods, and added fats and oils (83) . During these years, Americans consumed less red meat but more chicken, fish and turkey; fewer eggs, particularly eggs purchased in cartons in the grocery store, but more egg products used in processed foods or in the form of nonfat, no-cholesterol liquid egg substitutes; less milk, particularly whole milk, despite increases in fat-free, light and reduced fat milk; more yogurt (predominantly lower fat), but also more higher fat dairy foods including cheese and fluid cream; and more fats and oils, particularly salad and cooking oils, although a recent slight decrease has been seen.

Food supply data also suggest that Americans have consumed more total fat and less SFA and cholesterol since the early 1900s, reflecting an increase in production and use of vegetable oils and a decrease in animal fats (Table 4AA ) (84) . Although these trends continue from 1970 to 1994, the increase in grams of total fat and the decrease in grams of SFA are much less than that observed from 1909 to 1970. Food supply data from 1970–1996, adjusted for waste and other losses, show that Americans have consumed increasingly more added fats and oils relative to the Food Guide Pyramid recommendations (Table 4B ) (60) .

Two-day dietary data from 1994–1996 CSFII show that approximately one third of adult men and women met the dietary goals for total fat and SFA, whereas more than two thirds met the DV for cholesterol (Table 6A ). Data from the 1994–1996 DHKS show that the majority of adult respondents thought their diet was "too high" in fat but "about right" in SFA and cholesterol content, indicating that consumers may be able to judge their total fat and cholesterol intakes but have difficulty evaluating their SFA intakes. The majority of adult respondents also perceived the guideline of "choose a diet low in fat, saturated fat, and cholesterol" as "very important" (Table 6B ).

It is important to note that differences between the two DHKS surveys may be due to differences in sample populations and sample weighting (i.e., 1989–1991 respondents were household main meal-planner/preparers vs. 1994–1996 respondents who were individuals 20 y old), differences in question wording and/or response scales, and differences in methodologies (i.e., the 1989–1991 CSFII included three consecutive days of dietary intake using a 1-d recall and a 2-d record; the 1994–1996 CSFII included two nonconsecutive days of dietary intake from two 24-h recalls) (85) . It is also possible that changing attitudes about the importance of total fat intake reflect a shift in scientific thinking as well as conflicting media reports that total fat intake is not as important as was previously considered. Conflicting health messages are often disseminated with overinterpretation of the data from each research finding, rather than with a perspective of how the new finding fits with the totality of research on nutrition and health.

Survey data of nutritional status indicators

In NHANES III, the mean serum cholesterol level for adults, 20 y old, was 206 mg/dL, with 21% having concentrations > 240 mg/dL (defined as high blood cholesterol according to the NCEP) (28) . Approximately 41% of all deaths are attributed to cardiovascular disease in current data (86) . Since the 1940s, cardiovascular disease continues to be the leading cause of death in the U.S. population.

Age-adjusted mean serum total cholesterol levels decreased by 12–17 mg/dL among women and men, ages 20–74 y, who were examined in one of four separate national surveys conducted by NCHS between 1960 and 1991 (87) (Fig. 2 ). Age-adjusted LDL cholesterol levels decreased by an average of 8 mg/dL and age-adjusted HDL cholesterol levels increased by an average of 1 mg/dL, according to data from adults who participated in NHANES II 1976–80 and NHANES III, phase I (1988–1991). Mean serum total cholesterol levels also decreased in adolescents, ages 12–17 y, between 1966 and 1994 (88) . Population decreases in age-adjusted serum total and LDL cholesterol levels parallel changes in coronary mortality rates in the U.S.A., which decreased by 54% between 1963 and 1990 (87) .

View larger version (14K): [in this window] [in a new window]   Figure 2. Trends in age-adjusted mean serum cholesterol levels. NHES I indicates first National Health Examination Survey; NHANES refers to the National Health and Nutrition Examination Survey. Sources include the Centers for Disease Control and Prevention/National Center for Health Statistics. Adapted from Ref. 87 .

	STRATEGIES FOR IMPROVEMENT

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Why many Americans do not meet the guideline

Although the majority of Americans report that it is important to reduce their fat intake, beliefs and knowledge often do not translate into behavior change (90) . Many other factors influence food choices and eating behavior (91) . People choose foods according to personal preferences (taste in particular), cultural values, what they learned or were exposed to in childhood, what is available in the community and cost. People’s food choices are also influenced by their family and friends, and especially by the media.

Advertisements, developed by marketing experts to persuade consumers of food appeal, are widely disseminated. In 1997, the food industry spent $11 billion on advertising: $7.1 billion by food manufacturers, $3.1 billion by restaurants and $860 million by food retailers (92) . Food manufacturers spent 44.4% of their advertising budget on prepared convenience foods, confectionery and snacks, and bakery goods compared with only 2.2% on fruits, vegetables, grains and beans. For perspective, the budget to develop the Food Guide Pyramid was $1 million (93) . The annual budget of the Five-A-Day for Better Health program, aimed at increasing consumption of fruits and vegetables, is $5 million (94) .

The proportion of meals and snacks eaten away from home has increased from 16% in 1977–1978 to 27% in 1995 (95) . The consumption of foods prepared away from home often results in less control over the amount of fat eaten. Busy people may not find it convenient, nor make it a priority to select food choices that are low in SFA and cholesterol. Take-out foods generally do not have a Nutrition Facts label so consumers rarely know the content of SFA, cholesterol, trans-fatty acids or total fat.

Consumers’ food choices are also affected by federal policies that may be internally inconsistent. For example, the USDA is responsible for both dietary guidance policy aimed at health promotion and economic policy that provides incentives to enhance production of animal food commodities (96) .

Strategies for meeting the guideline

Nutrition education is the common strategy used to combat barriers to dietary change at the individual level. A variety of theoretical models or frameworks are used by educators in academia, government and private industry to lower dietary fats (91) . Those that have been most successful focus on behavioral changes (90) and incorporate self-evaluation tools, social support, contractual agreements, principles of social marketing and increased access to lower fat foods that taste good (91) . Many programs teach specific food-choice strategies such as choosing skim milk and lean meats instead of higher fat counterparts, increasing intake of fruits, vegetables and whole grains, and reducing portion sizes (91) . Many programs involve the media, schools, work-sites and the general community (36) . With the increasing diversity of the U.S. population, future programs will have to target vulnerable subgroups in culturally sensitive ways (97 ,98) .

Several federal dietary guidance policies aim to reduce the population’s intake of SFA, cholesterol and total fat. The passage of the NLEA in 1993 and the revised Nutrition Facts label, jointly agreed upon by the Food and Drug Administration (FDA) and USDA, comprise a daily source of information for consumers (99) . Nutrient content and disease-specific health claims related to dietary fat have been approved by the FDA and appear on many food labels (99) . A USDA policy to require the Nutrition Facts label on packaged meat and poultry products was announced at the 2000 National Nutrition Summit in Washington, DC (100) . The inclusion of the content of trans-fatty acids with SFA on the Nutrition Facts label has been proposed by the FDA (101) . The Food Guide Pyramid, released in 1992, demonstrates proportionality, variety and moderation of foods, including added fats, and is widely recognized (102) . In 1996, the National School Lunch and Breakfast Programs were mandated to serve lunches and breakfasts that meet the guideline for dietary fats (103) . Several education initiatives from the NIH have adapted diet and risk factor information that is culturally sensitive and targeted to specific population groups. Industry has also introduced a plethora of nonfat, low fat and fat-modified foods intended to reduce dietary fats in the U.S. population. Although these products are lower in fat, many of these products are not lower in energy. Thus, dietary guidance policy that would mirror, in the U.S. population, both the lower fat intakes of other countries and their lower prevalence of overweight and obesity, has not succeeded.

What next steps might be encouraged?

The concepts and information that provide the foundation for the Food Guide Pyramid must be updated to reflect current intakes and recent Dietary Reference Intakes (104 105 106) . In addition, messages on types of dietary fats are not well conveyed by the Pyramid. Consideration should be given to the design and implementation of a project that will help consumers use the Pyramid to recognize different types of fats and their food sources, and to choose a diet in the market place that is low in SFA and cholesterol and moderate in total fat in the context of the total diet. Research initiatives should be developed to identify practices that are most effective in assisting people with lower levels of income and education to choose diets that reflect the guideline.

Further consideration must be given to messages or food product labels at the point of purchase (wherever the purchase) to help consumers identify foods that are low in SFA and cholesterol and compare energy content among food products. This is especially important for foods prepared away from home. Evaluation of endeavors related to recent food labeling initiatives is also warranted. For example, what do consumers understand about % DV, what refinements in the Nutrition Facts label can improve the message that type of fat is important to consider, and if implemented, what is the effect of trans-fatty acid labeling on food purchases and fat intake?

Some academic scholars provide provocative ideas for changing the environment, such as the provision of economic incentives for the promotion of healthful foods or placing putative taxes on unhealthful foods (36 ,96 ,107) . Such policy would require an agreed upon and enduring definition of a healthful food, an especially challenging task (e.g., not all nonfat or low fat foods might be considered healthful and vice versa, not all full-fat foods might be considered unhealthful). Ways to modify Federal policies to enhance public health goals have also been suggested (96 ,108) .

Lessons from community studies aimed at health promotion and risk factor reduction show that through identification and working with community leaders, healthful behavior changes occur. The challenge is to maintain healthful behaviors once the funding source no longer provides support. It may be time to reconsider research support for the development of model centers of change. These model centers could identify examples of successful change in adopting lower intakes of SFA and cholesterol and moderate intakes of total fat by diverse cultures, and measure long-term intake and health indicators.

Finally, support for research, nutrition monitoring and surveillance must continue if we are to clarify further the relationships among dietary components, nutrients and chronic disease. Attention must be given to continued development of improved dietary intake methods to assess fat intake with accuracy. Types of fat vary substantially among and within brands of foods. Which fat is used in a food product depends on the price of oils and fats available at the time of production, the functional property of the fat and the requirement to show SFA content on the Nutrition Facts label. Systematic probes and questions to identify brand names as well as amounts of fats that consumers add to foods are necessary to obtain accurate information regarding dietary fats. Support for accurate and comprehensive food and nutrient composition databases that provide data on nutrients such as trans-fatty acids must also continue and be increased to keep pace with the thousands of new foods introduced each year (e.g., 16,863 were introduced in 1995) (92) . In addition, support for the collection of at least two independent days of dietary intake data in national surveys is important. Multiple days of dietary data are necessary to estimate "usual intake" in order to assess properly the prevalence of individuals with fat intakes above the numerical goals (63) . Such data are critical for food assistance, food regulatory and consumer education programs.

In conclusion, the success of dietary guidance to achieve a national dietary pattern that is low in SFA and cholesterol and moderate in total fat is dependent upon a multifaceted approach to dietary change in step with national nutritional monitoring that will facilitate analysis of what and where change is occurring and where additional resources are needed.

	APPENDIX

TOP ABSTRACT INTRODUCTION Dietary fats and chronic... Emergence of dietary guidance... Dietary Guidelines for... Major changes in the... Other message changes in... SURVEILLANCE METHODS TO ASSESS... RESULTS FROM SURVEILLANCE DATA STRATEGIES FOR IMPROVEMENT APPENDIX REFERENCES

 
Dietary sources of SFA, cholesterol, and total fat

The text of the guideline initially contained a dual message about food sources of fats (Table A1). Limited information about major food sources was accompanied by a reminder that these foods were also good sources of essential nutrients. In 1990 and subsequently, the text of the guideline offers suggestions to help identify the major sources of SFA and cholesterol. Emphasis is no longer on the nutritious quality of foods that are high in SFA and cholesterol. Instead, the discussion promotes low fat versions of meats and dairy products.

    Dietary fats and health concerns Shifts in the explanation of the link between dietary fats and health have occurred (Table B1 ).The initial concern, the relation between dietary fats and risk of CHD, was broadened to include an emphasis on total fat to reduce the risk of certain cancers, and potentially to avoid greater prevalence of weight gain and obesity. The current message returns the focus to reduction of CHD through reduced intake of SFA and cholesterol and use of unsaturated fatty acids within a moderate total fat intake.

Messages were dropped that tended to "medicalize" the guideline and diffuse the diet message, such as advice to have blood cholesterol measured and achieve 200 mg/dL, and the mention of multiple CHD risk factors. Also deleted were discussions about variation in blood cholesterol levels with intakes of SFA and cholesterol, and a comparison of the relative effect of SFA vs. dietary cholesterol. The intent is a change in national dietary patterns to lower the intake of SFA and cholesterol, to improve mean blood cholesterol and reduce the CHD risk of the population.

Two additional links between dietary fat and health are made in the guideline in 1995 and in 2000: an association between intakes of trans-fatty acids and (n-3) PUFA and CHD. The advice is that foods high in trans-fatty acids tend to raise blood cholesterol and that (n-3) fatty acids are being studied because of a possible association with decreased risk for heart disease in certain people.

    Dietary fats and food selection/preparation The advice about food selection and preparation has evolved into describing choices among types of fat (Table C1 ).Recent editions discuss SFA and cholesterol together, rather than make distinctions between the two food components. Although the 1995 guideline suggests using all types of fats and oils "sparingly," current readers are advised to limit animal fats, hard margarines and partially hydrogenated shortenings. Vegetable oils are mentioned as substitutes for SFA.

The Nutrition Facts label appears in the guideline to make the numerical goals more understandable and useful to readers at the point of food purchase. The % DV allows the Committees to signal a numerical goal for cholesterol, by referencing the cholesterol DV of 300 mg, at the 2000 kcal level.