DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR FOOD SAFETY AND APPLIED NUTRITION

NUTRITION SUBCOMMITTEE MEETING

TOTAL FAT AND TRANS FAT

Tuesday, April 27, 2004

12:30 p.m.

Leows L'Enfant Plaza Hotel

480 L'Enfant Plaza

Ballroom A and B

Washington, D.C.

PARTICIPANTS

Norman I. Krinsky, Ph.D., Chair

Jeanne Latham, Executive Secretary

MEMBERS:

Susan S. Baker, M.D., Ph.D.

R. Jean Hine, Ph.D. (Consumer Representative)

Guy Johnson, Ph.D. (Industry Representative)

Susan T. Mayne, Ph.D.

Suzanne Pelican, M.S., R.D.

Barbara M. Shannon, Ph.D.

TEMPORARY VOTING MEMBERS:

Alice Lichtenstein, Ph.D.

Eric Rimm, Ph.D.

C O N T E N T S

Welcoming Remarks, Virginia Wilkeneng, R.D. 4

Conflict of Interest Statement, Jeanne Latham 6

Background, Questions and Charges,

Dr. Kathleen Ellwood 10

Remarks by Chairperson and Member Introductions,

Norman I. Krinsky, Ph.D. 20

Public Comment:

Robert Earl, National Food Processors

Association 167

Martin Hahn, Hogan and Hartson 175

Subcommittee Discussion 195

Public Comment:

Mary Enig, Ph.D., Weston A. Price Foundation 200

Subcommittee Discussion 215

P R O C E E D I N G S

Welcoming Remarks

MS. WILKENENG: I would like to start the meeting. I am Virginia Wilkeneng, the Deputy Director in the Office of Nutritional Products, Labeling and Dietary Supplements. I am here to welcome the committee. This is the first meeting of the Nutrition Subcommittee of FDA's Food Advisory Committee. We thank you very much for your willingness to serve the agency in this way and for the service that you will be providing as you go through this meeting and future meetings.

We thank you so much for helping us address some of the major issues of the day, in this particular case primarily health claims and nutrition labeling. In the case of health claims, we have been facing increasing issues of whether the total fat limitations we have on health claims is still relevant in light of the other limitations on saturated fat and cholesterol and we welcome your input on this.

In the case of nutrition labeling, the issue of trans fat labeling has been one the agency has been dealing with for a great deal of time now. We do have our final regulations for putting the quantitative amount of trans fat on the label and you are starting to see those appear in the marketplace now. But our concern is how consumers will be able to put that information into action. They need some context for the numbers.

Numerically, the amounts of trans fat in the label will be small numbers. Some of our research has shown that people react to the magnitude of the number. Whether it is grams of fat or milligrams of sodium or micrograms of other things, they can't be expected to know what amounts of those various nutrients are needed in a diet.

For that reason, FDA established, with the new nutritional labeling regulations, daily values so there are to be percentages there so that the consumer doesn't need to know how much of any particular nutrient they need but can look at the percent DV and determine whether it is a high number or a low number and, therefore, know the magnitude of that nutrient in the serving of a food. So, we need some help with the trans fat in trying to figure out the best way possible to address those issues.

So, I welcome you. I am a stand-in. You will see on your agenda that Dr. Laura Tarantino's name had been down here. She is our Acting Office Director. I expect many of you knew Dr. Christine Taylor who had been our director for a number of years since our Office was put together, and she has just recently, in fact this week, finally got to Geneva where she will now be in a two-year program with the World Health Organization. If you don't already know, I am happy to announce that starting next Monday we will have a new Office Director, Dr. Barbara Schnieman, from UC Davis, so she will be here to welcome you the next time you have a meeting but, in the meantime, thank you very much. We very much appreciate your service. I will turn it over to Jeanne Latham.

Conflict of Interest Statement

MS. LATHAM: Good morning. I am Jeanne Latham, the Executive Secretary for the FDA Nutrition Subcommittee of the Advisory Committee. I would like to welcome everyone, particularly our members and our temporary voting members. With that, I would like to read the conflict of interest statement into the meeting record.

The following announcement addresses the issue of conflict of interest with respect to this meeting, and is made a part of the record to preclude even the appearance of such at this meeting.

The authority to appoint temporary voting members is granted to the Center Director and, relying on that authority, Dr. Robert E. Brackett, Center Director of the Center for Food Safety and Applied Nutrition, has signed letters stating, by authority granted under the Food Advisory Committee Charter, that he appoints, Dr. Lichtenstein and Dr. Rimm as temporary voting members of the Nutrition Subcommittee of the Food Advisory Committee for the April 27th and 28th, 2004 meeting on total fat and trans fat.

All members and temporary voting members have been screened for financial conflicts of interest. Because of the breadth of topics to be discussed at this meeting, all of the members and temporary voting members have been screened for any and all financial interests associated with regulated industry. Based on this review, FDA has determined, in accordance with 18 USC Section 208(b)(3), to grant genera matters waivers to Dr. R.G. Hine, Dr. Norman Krinsky, Dr. Alice Lichtenstein, Dr. Jay Michael McGinnis, Dr. Susan T. Mayne, Ms. Susan Pelican and Dr. Eric Rimm that permits them to fully participate in the matters before the committee. Copies of the waiver statements may be obtained by submitting a written request to the agency's Freedom of Information Office, Room 12A-30 of the Parklawn Building.

In an effort to enhance consistency within FDA, the agency has recently adopted a policy whereby all public commentors will be asked to report any personal financial interests that could be affected by the committee's deliberations. A copy of this policy was provided to all the individuals who registered to make a public comment. I believe that additional copies of the policy may be obtained at the registration desk.

I have one additional administrative announcement. We have received six written submissions from individuals, some of whom were not able to appear at this meeting. The submissions are from Barbara Sakow, Dr. Mary Enig of the Western Price Foundation, Alison Krester, of the Grocery Manufacturers of America, two from Dr. Fred Kumerow at the University of Illinois Urbana, and Ann Beckman of Hogan and Hartson. The submissions have been provided to our members and copies are available at the registration desk.

Thank you and, again, welcome to the committee members and everyone who has joined us today. I will turn the meeting over now to Dr. Kathleen Ellwood, FDA's Office of Nutritional Products Labeling and Dietary Supplements, the Center for Food Safety and Applied Nutrition, who will give us some background information and talk about questions and charges to the committee. Kathy?

Background, Questions and Charges

DR. ELLWOOD: I would like to thank the members of the committee for agreeing to serve on this committee and work with us on these important issues.

I am going to go through and give you just a little bit of background as to why we are asking these questions. The first question is addressing the eligibility criterion the FDA has applied to most health claim regulations pertaining to heart disease risk. Foods bearing these claims must be low in total fat. We would like to know what the current evidence suggests in terms of total fat intake and risk of coronary heart disease. As you heard from Virginia Wilkeneng, we do have various disqualifying levels for many new trans and total fat is among those. Keep in mind that we still have low unsaturated fat and cholesterol.

But as a background to this, I want to go through some of the health claims that do bear a low total fat criterion and some of the exemptions that we have made. Now, the first two that you see listed here are sat fat and the fruit and vegetable one. They came in under NLEA, the Nutritional Labeling and Education Act. The second one was a little bit of a modification for what Congress asked for. Then, the latter two were through the petition process. These are all health claims that meet significant scientific agreements so they are authorized.

In the first health claim on saturated fat and cholesterol effect on heart disease, FDA concluded that the scientific evidence linking diets low in saturated fat and cholesterol to reduce the risk of CHD was strong. The agency also noted that while total fat was not as strongly or directly linked to increased risk of coronary heart disease as sat fat and cholesterol, it may have significant indirect effects.

The justification was that low fat foods generally help individuals in reducing their intake of sat fat and cholesterol and that excess calories, of which fat contributes more per gram than the other energy nutrients, is associated with two health-related conditions that can be risk factors for heart disease.

FDA concluded that the scientific evidence linking diets low in saturated fat and cholesterol to reduce risk of coronary heart disease was strong and, as such, the subject of the claim is dietary saturated fat and cholesterol and dietary total fat is addressed as an additional criterion that must be met by a food before it may use the dietary saturated fat and cholesterol health claim.

But there was an exemption because it was noted that you can have diets with meat and poultry, as long as they were lean, as part of a healthful diet. Meats and poultry are recognized or regulated by FSIS. However, FDA put in a claim for game meats and fish, which we do regulate, that they meet the extra lean requirements for saturated fat, cholesterol and total fat or they were not as stringent as the definitions for low sat fat, low cholesterol and low total fat.

Now, in the fruits, vegetables and grain products, the second health claim that I wanted to discuss, the low total fat criterion was included because low total fat is one of the characterizing features of diets rich in fiber containing fruits, vegetables and grain products. It is also a characteristic of dietary patterns which relate to lower heart disease risk and it facilitates the ability of consumers to achieve diets low in saturated fat and cholesterol. So, that remained a criterion.

In the soy protein and coronary heart disease claim there was an exemption for a lower total fat criterion for food products consisting of or derived from whole soy beans as long as those products contained no other fat. The rationale for this was that whole soy beans are useful sources of soy protein and can be appropriately incorporated in a diet low in fat, saturated fat and cholesterol.

For a more recent health claim, plant sterol and stanol esters and coronary heart disease, we have an interim final rule that came out in 2000. FDA required that foods bearing this health claim be low in saturated fat and low in cholesterol but not low in total fat. The rationale for this was that because the emphasis in the Dietary Guidelines for 2000 was on diets moderate in total fat and, additionally, there were a number of foods such as margarine and salad dressings that could be formulated to contain plant sterol or stanol esters by still qualifying as low saturated fat and low cholesterol.

More recently, we just released a qualified health claim for walnuts and coronary heart disease. This health claim is not one through authorization, through a letter of enforcement discretion. We did not impose a requirement that a food meet the definition of a low fat food as walnuts are not because walnuts have a good ratio of unsaturated fat to saturated fat and they contain significant amounts of other substances, such as dietary fiber and phytosterols, which may reduce the risk of coronary heart disease. They are also exempt from the high fat disqualifying level which is 13 grams per reference amount customarily consumed.

We have seen a lot of current scientific evidence in recent years and the shift that you need to look more at the type of fat, such as the Dietary Guidelines for Americans 2000 which I have mentioned, as well as the recent Institute of Medicine/National Academy of Sciences report, macronutrient report that was recently released, as well as the American Heart Association Dietary Guidelines and the National Cholesterol Education Program's ATP III report.

Now I am going to shift to our second set of questions dealing with trans fatty acids, and our first one was that the dietary guidelines committee may suggest that less than 1 percent of energy should be obtained from trans fatty acids, and this calculates out to about 2 grams per day for a 2,000 kilocalorie diet. We would like to know if the scientific evidence supports this level.

It is very important to FDA that we set a dietary value so we can have a percent DV. If you have a percent DV it gives the consumer an idea as to how much there is of that nutrient. They can put it in the context of their total daily diet. For instance, if you have a 5 percent daily value, that would be a low amount of that nutrient versus 20 percent which would be a high amount. So, it is a little bit more than just looking at a gram amount of a particular nutrient.

The second question related to trans fatty acids we would like answered is, when compared to saturated fatty acids, are trans fatty acids considered to be more, less or similarly adverse with respect to coronary heart disease?

A little bit of background on trans fat, in 1993 there was a final rule that came out entitled Food Labeling Mandatory Status of Nutrition Labeling and Nutrient Content Revision Format for Nutrition Label. In this, the agency required declaration of total fat and saturated fat in the nutrition label. At the same time, the agency said it was premature to include trans fatty acids in nutrition labeling because we lacked scientific agreement on the dietary implication of trans fat intake, but we also acknowledged that we would be revisiting this in the future as scientific evidence became more evident in this area.

A final rule was released just last July. This was a partial response to a petition from the Center for Science in the Public Interest. But, at the same time, FDA was very aware of an increasing body of evidence that showed that dietary trans fatty acids raised blood cholesterol levels, specifically LDL cholesterol, thereby increasing the risk of coronary heart disease. Scientific evidence included authoritative statements from the Dietary Guidelines for Americans 2000.

In this final rule the declaration of this nutrient must be on a separate line, immediately under the declaration for saturated fat, but it would not include a percent daily value. This is required for some of the other mandatory nutrients such as saturated fat. This final rule will go into effect as of January 1st, 2006.

At the same time, an advance notice of proposed rule-making was released and this is a way for FDA to go out and ask several questions that we need to have addressed prior to going to a proposed rule. Some of the things we asked about trans fat statements were whether it should be alone or in combination with saturated fat and cholesterol, or should it be a footnote in the nutrition facts panel or disclosure statement in conjunction with claims. There are other issues on this but I am not going into those. I am going to just deal with the first statement here.

Now, the use of a footnote--we had several statements but essentially, due to several reports that have come out, trans fats should be kept as low as possible. But even coming up with a percent daily value would also help us in nutrient content and health claims.

Some of the authoritative reports I have already discussed are the Dietary Guidelines 2000 that recommended that Americans cut back on trans fats, reducing fat intake; a recent report in the IOM/National Academy macronutrients report, as well as the NCEO ATP III report, all stating that trans fat should be kept as low as possible. The IOM/NAS report said it should be kept as low as possible while maintaining a nutritionally balanced diet. It may not provide a dietary reference intake for trans fat or information the agency needs to establish the daily reference value for nutrition labeling purposes. In addition, a more current report on guiding principles was just released suggesting that the saturated fat and trans fat be on separate lines but have a joint percent daily value.

The authoritative body, as you all know, the Dietary Guidelines Committee, is currently meeting and, as of last month, discussed setting less than 1 percent of energy as what would be obtained from trans fatty acids. As I have said, this calculates out on a 2,000 kilocalorie diet, which is what we base our daily values on, to be about 2 grams per day.

So, what we would like for the committee to do is to provide a succinct reply to the three questions which we have proposed based on the most current scientific evidence and on what you will have heard during the deliberations of this meeting. And, we look forward very much to your deliberations. Thank you very much.

Remarks by Chairperson and Member Introductions

DR. KRINSKY: Thank you very much, Kathy, and this officially starts the meeting of the Subcommittee on Nutrition of the Food Advisory Committee. I would like to do a little housekeeping first of all. We have microphones in front of us and, as you will note, while I am speaking there is a red light on the microphone. That will indicate to me when you want to say something. There is a semicircle on the bottom that you push on or off and that will turn the mike on and off.

I think it would be worthwhile at this time, since we know what our charge is, if we would just briefly introduce ourselves for the sake of the audience behind us. I would like to begin. I am Norman Krinsky and I am an Emeritus Professor of Biochemistry.

I want to step back for a moment. I should have mentioned my name at the very beginning because this is being transcribed and the transcriber will find it most useful when you mention your name and then we will be able to write this down. So, Norman Krinsky. I am an Emeritus Professor of Biochemistry at Tufts University School of Medicine, in Boston, and a scientist at the Human Nutrition Research Center on Aging, also at Tufts University. My area of specialization is carotenoids, both chemistry, metabolism, action, function--whatever they do in the human body, and also what their metabolites may be doing. Most of you know that the primary metabolite of the provitamin in carotenoids is vitamin A but there are other metabolites which have been becoming increasingly interesting with respect to what their functions may be. I am a member of the Food Advisory Committee of the FDA and I have also chaired two committees for the IOM/NAS, one the DRI committee on antioxidant vitamins, and the other a committee that was evaluating the scientific evidence for making claims. Let's go around from my right.

DR. MAYNE: I am Susan Mayne. I am an Associate Professor of Epidemiology and Public Health at Yale University School of Medicine, and I am also the Associate Director of the Yale Cancer Center where I am responsible for cancer prevention and control at the Yale Cancer Center. My research interest is largely in nutritional epidemiology of cancer, with a broader interest in nutritional epidemiology of chronic disease. I have some specialized expertise in the antioxidant nutrients and the carotenoids, and I did serve on the DRI panel for antioxidant nutrients that Dr. Krinsky chaired. My background is in nutritional biochemistry and toxicology which later merged into epidemiology and clinical trials.

DR. RIMM: I am Eric Rimm. I am an Associate Professor of Epidemiology and Nutrition at the Harvard School of Public Health. I have interests in nutrition and cardiovascular disease, specifically focusing on fatty acids, antioxidants, B vitamins, alcohol and a number of other interesting things related to cardiovascular disease, as well as obesity and diabetes. I also was on the DRI panel for macronutrients, and that is it.

DR. SHANNON: I am Barbara Shannon. I am Professor Emeritus in Nutrition for the Pennsylvania State University and also Dean Emeritus of the College of Health and Human Development. My area is nutrition and education across the entire life span but most particularly nutrition and education in schools and cholesterol education in heart disease prevention.

DR. HINE: My name is Jean Hine and I am from the Arkansas Cancer Research Center in Little Rock, where I am a Professor in the Department of Nutrition. My interests are in B vitamins, specifically folic acid and vitamin B-12, and more recently the effects of carbohydrates on development of cancers.

DR. LICHTENSTEIN: My name is Alice Lichtenstein. I am a Professor of Nutrition Science and Policy at the Freedman School, Tufts University, and Director and Senior Scientist of the Cardiovascular Nutrition Laboratory at the USCA Human Nutrition Research Center on Aging at Tufts University. My primary interest is the effect of diet on cardiovascular disease risk factors in postmenopausal women and older males. As Dr. Rimm, I served on the IOM Macronutrients Panel. I also served on the 2000 Dietary Guidelines Committee and currently chair the nutrition committee of the Heart Association.

DR. JOHNSON: My name is Guy Johnson. My background is nutrition and, after a brief stint in academia, I spent 20-some years with different food companies; 11 years with Gerber, 9 years with Pillsbury and, most recently, 2 years with Kellogg as VP of Nutrition for Kellogg USA. I saw the light a few years ago and decided to start my own consulting business so I have been deliriously happy helping a variety of interesting stakeholders address nutrition-related scientific, regulatory and communications issues, and I am very much looking forward to participating on this panel.

DR. BAKER: My name is Susan Baker. I am a Professor of Pediatrics at the State University of New York at Buffalo and I am co-Division Chief of the Digestive Diseases and Nutrition Center at Buffalo Children's Hospital. I am a clinician. I am also a researcher. In the past my research has focused on the barrier function of the gastrointestinal tract and factors such as nutritional free radicals, and so on, that tend to disrupt it. More recently, we have had a big change in our laboratory and we are now looking at polymorphism in inflammatory mediators and the effect that it has on diseases in childhood. I also do a fair amount of clinical research, most recently in nutrition and CF, but my love is nutrition and pediatrics and that is where I continually try to focus myself. Thank you.

MS. PELICAN: My name is Suzanne Pelican and I am a food and nutrition specialist with the University of Wyoming Cooperative Extension Service. I am also the co-principal investigator on a three-state project called "Wellness in the Rockies." We are seeking to promote healthy weight among individuals of all ages using community and individual level approaches. We are looking to change people's attitudes and behaviors related to food, physical activity and body image. My area of work and training has been in nutrition education and public health. Prior to coming to University of Wyoming I worked for the Indian Health Service for the Navajo Nation for Pennsylvania State University and Colorado State University. Thank you.

DR. KRINSKY: Thank you all for describing your interests and activities. I might point out that Michael McGinnis, whose name is at the table to my left, is not here today. He is not able to attend today but he will be here tomorrow but will not be in a position to vote on our actions because he will not be privy to the discussions that we have today.

Speaking of voting, we have two temporary members of this committee, Eric Rimm and Alice Lichtenstein. They are voting members of this committee. We also have our consumer representative, Jean Hine. She also is a voting member. Guy Johnson is the industrial representative and he is a non-voting member. So, that sets us all straight, I hope, as far as who can vote and who can't vote. It is clear that we can all vote except for Guy.

The issue that confronts us are the three separate questions, and we might as well begin with the very first one. The first one is what does the current evidence suggest in terms of total fat intake and risk of coronary heart disease?

We have been supplied with readings, both in the notebook and readings that were sent to us later. There certainly is a long history of the relationship of fat intake and coronary heart disease. I think what this demonstrates is that we are beginning to get a little more sophisticated with respect to fat, that it is not a single entity but that it is composed of many different kinds of compounds. I think that is a general aspect of nutrition, that we deal with broad compounds and we see in many cases that we have to look at what the individual components are to see what their effects are on human nutrition.

So, I would invite any comment with respect to total fat intake, particularly those of you that have had more direct experience than I. So, again, what does the current evidence suggest in terms of total fat intake and risk of coronary heart disease? Again, if you can give your name to begin with.

DR. JOHNSON: Johnson. I just wanted to add one thing to the presentation that Dr. Ellwood made. Correct me if I am wrong, but I think there is another example of regulatory action that the agency has taken to provide more flexibility in a health claim with respect to fat, and that is the whole grain heart disease claim. It is a claim under the FDA Modernization Act. Originally the criteria for that health claim included a low fat criterion and, in response to a petition not too long ago, the agency relaxed that so that whole grain products that are not low fat are eligible to bear that claim as long as they meet the other criteria.

DR. LICHTENSTEIN: I guess I will start the discussion off. I think the data support the impact of different fatty acids or classes of fatty acids in cardiovascular disease risk via effects on primarily lipids and lipoprotein levels and, to a lesser extent, on coagulation factors. But independent of the effect of low fat, high carbohydrate diets and triglyceride levels, I think the effect of the total fat content of the diet is far less, if not at all, relative to the individual fatty acids or classes.

DR. KRINSKY: May I ask the question with respect to triglyceride levels and whether, in fact, that is a relative factor with respect to risk for coronary heart disease.

DR. LICHTENSTEIN: Well, there is an inverse relationship between HDL cholesterol levels and triglyceride levels with rare exceptions, things like estrogen therapy which will increase triglycerides and not HDL cholesterol levels. Are you asking essentially is there an independent effect of triglyceride levels? I don't think it can be separated from the effect on HDL cholesterol levels. There are some data that would suggest that it is independent, but then there are other data to suggest that it is very difficult to control for one and not the other. Then, the other confounding factor is frequently, at least in a metabolic situation, that if you drop the fat content to the diet and the caloric density of the diet subjects tend to lose weight, at least in the short term. We are not clear about the long term. Then that complicates the effect of the macronutrient component of the diet on triglyceride levels because the weight loss itself will lower the triglyceride levels.

DR. KRINSKY: I ask because it wasn't clear whether the trans fatty acids and saturated fatty acids, if they have an impact on triglycerides, whether that is going to be a direct relationship to altering the risk of coronary heart disease. Because I think basically what we are dealing with is what are the factors of trans and sat fat that impact on coronary heart disease. So, I gather with the triglycerides it is ambiguous. Thank you.

DR. RIMM: I think one of the big issues that we stumbled over for a long time in the IOM committee was that by just talking about fat you miss what is being replaced in the diet, and that is if you move from a low fat diet to a high fat diet or, conversely, from a high fat diet to a low fat diet in an isocaloric situation, if you are putting in very refined carbohydrates you can say one thing, or if you are putting in whole grains or other carbohydrates which are absorbed more slowly then it makes it that much more difficult to interpret the effects of a high fat versus a low fat diet.

But, in general, I think I concur completely with what Alice said, that the evidence suggests that the fatty acid component is much more important than the total amount of fat, and much of what we have struggled with is that in this country it is very hard to have a diet which is 60 percent of calories from fat and still be below the threshold for saturated fat and cholesterol.

Obviously, that is an extreme but as you move down I think the challenge becomes, given the current food supply, can you eat 40 percent of calories from fat and still stay below the saturated fat threshold and cholesterol threshold? So, given that, I think that there really is not strong evidence at all that variations in fat impact on coronary heart disease.

DR. KRINSKY: Any other comments then? Alice?

DR. LICHTENSTEIN: Just to follow-up on that, I think initially some of the reasoning behind recommending diets low in total fat, saturated fat and cholesterol was that if you decrease total fat, at least within the United States, you decrease saturated fat. There was a reassessment of that, at least during the deliberations of the 2000 Dietary Guidelines Committee, because what happened as a result of that recommendation is that the low fat recommendation sort of took on a life of its own independent of the low saturated fat recommendation so that there was a proliferation of products that were low in total fat and had not been high in saturated far to begin with so that the adoption of a low fat diet using a lot of those products really resulted just in a net increase in carbohydrate intake, primarily refined carbohydrate intake and simple carbohydrates, and didn't have much of an impact on saturated fat.

So, I think that some of the data that we have is partially confounded by that trend. I don't think it necessarily changes the conclusion that the total fat content of a diet doesn't seem to have a major effect on cardiovascular disease risk. But it needs to be taken into consideration because I think when you make recommendations they have to be scientifically sound but then one also needs to take into consideration how they are going to be interpreted by the public and by manufacturers.

DR. KRINSKY: So, can we answer this first question with respect to total fat intake? If we are concerned about the component of the fat, how valid is the question with respect to coronary heart disease if it is not describing what the components are of the total fat? Because unless we assume that there is some given formula for total fat--such-and-such percent sat, such-and-such percent polyunsaturated, such-and-such percent monounsaturated and trans, what is meant by total fat?

DR. LICHTENSTEIN: What I think we have seen is that it de-emphasizes the components that probably have the major effect on cardiovascular disease which is saturated fat and trans fatty acids. Then we get into a situation where there are foods where perhaps there is adequate scientific data to make health claims as far as the examples that were given that actually required exemptions, so the bar was essentially higher for them.

Another example would be something like salmon which is relatively high in fat but it is unsaturated fat, specifically a long chain of omega-3 fatty acids. So, the current emphasis on total fat really, I think, is diverting us from where probably the biggest bang for a health claim would be, and that would be the saturated fat and trans fatty acids.

DR. RIMM: I think that if we look specifically at what the question asks, which is what I will try to do, the last line is what does the current evidence suggest in terms of total fat intake and risk of coronary heart disease? I think the current evidence from clinical trials that have been done and from prospective cohort studies--all that evidence suggests that there is essentially no association within the range of what people eat. There have been two large clinical trials that have been done, one back in 1965 and one back in 1989, both of which altered fat intake and followed people for two to six years. There were secondary prevention trials. Both found no effect of change in fat of the diet.

Subsequently, we and others have done prospective cohort studies looking across diets among men in the health professional follow-up study and nurses in the nurse health study, as well as a few other studies across a pretty wide range of diets in our cohort, ranging from 24 percent to 42 percent in calories from fat and within that, holding constant the composition, as we said, there really is no association between total fat and coronary heart disease.

So, I think we really should be focusing on the composition, otherwise there may be more and more exemptions as more and more foods come up which, as you say, have higher monos and higher polys and lower sats and trans and would fit the definition and should be exempt.

DR. JOHNSON: I wanted to also focus on the question that FDA asked. Primarily it is in the context of health claims. As those of us who lived through NLEA can attest, NLEA was spawned in the ocean where fat was public enemy number one and the dietary recommendations from the Heart Association and the dietary guidelines committee focused in on limiting total fat as part of a strategy to deal with that issue. Now we have a whole new chapter of the Heart Association, with guidelines written by the person sitting on my left I suspect, that came to the conclusion that heart disease risk can be perfectly well managed with a moderate fat diet assuming that it contained poly unsaturated and monounsaturated fatty acids. The dietary guidelines committee changed its tune. And, with respect to health claims there would still be a disqualifier level for total fat so that products that are very high in fat would be limited by that unless FDA granted them exclusion.

So, there are still some controls on total fat with respect to health claims and it just seems to me that based on the information that has accumulated during the last 10 or 12 years, to have a carte blanche requirement that foods be low in fat in order to qualify for a health claim is based on yesterday's science.

In addition, as has been pointed out, it is saturated fat and high levels of trans fat that are implicated in heart disease risk now and the agency can very easily control those as criteria for new health claims that come around. So, to me, this one is a slam dunk.

DR. KRINSKY: Any other comments on total fat intake?

MS. PELICAN: Suzie Pelican. Just to follow-up on Guy's slam dunk, it is really not necessarily a comment on the research but more looking in the future, and maybe this is beyond not only our committee's role but FDA's role, but thinking about such a change like this, would it be worthwhile to look at the kinds of changes that we might see in the food supply based on a change like this? Should we just kind of explore what that might mean for what we saw with, as I think Guy said NLEA, and saw this huge proliferation of low fat foods? Should we think at all that by dropping this requirement for total fat--would that be worth exploring at all?

DR. KRINSKY: Well, we know it has happened, the results of the low fat craze, if you will, and that is the present epidemic that we have of obesity. So, these recommendations have the most serious consequences. Alice?

DR. LICHTENSTEIN: I think we have seen some of what will happen already because a lot of the products that were non-fat or low fat have actually gone off the market, in some cases due to consumer non-acceptance and in some cases because of some of the other diet trends that have occurred since that major proliferation. But I think if we look at what the consequences were and what types of foods came on the market, essentially if you take fat out of the diet or out of a food what are you putting in? Well, it appears that for a lot of those products it was sugar, and if it wasn't sugar it was refined carbohydrates. So, I think that is an important point to consider but I am not sure that, from what we have seen, it would be a disadvantage to drop those components of the diet. I think when the original low fat recommendations were made it was sort of envisioned that people would be eating a lot of fruits and vegetables, and legumes and whole grains. However, you know, we learned a lesson. We learned that although that may have been assumed and may have been what was intended, it turns out that is not exactly what happened. So, I think it is a point we need to really think about.

DR. KRINSKY: That is "the road to hell is paved with good intentions" result. Certainly, the intention was laudable and the result is horrible, if you will, for the country.

DR. LICHTENSTEIN: Unfortunate.

DR. JOHNSON: I wanted to respond to Suzanne's question. I think that one ramification of what we are talking about would be to have the food industry focus in on cholesterol raising fatty acids and that would give them an incentive to develop products that are lower in those if there was more liberal use of health claims that provided some flexibility with respect to total fat.

DR. LICHTENSTEIN: One other point that I think we learned from that lesson of just recommending low fat is that there was a certain perception of the consumer that low fat was synonymous with low calory, and that certainly was very dangerous because, with very rare exceptions, with something like whole milk and skim milk it definitely is not the case and, yet, there was a feeling that if it is low fat, non-fat, if that is what it says, it is good for me; I can eat lots of it. And, we see the consequences of it.

Also, because I think the issue of overweight and obesity has really risen as far as a major health problem in the United States, in addition to perhaps putting less emphasis on the total fat content of the diet, I think what we really need to do is put a lot more emphasis on the caloric value of the specific foods and portions of the foods as presented on the food label. Because we are sort of seeing now that we have seen the pendulum swing towards the higher fat, higher protein foods, again to a certain extent the assumption is, well, these are the diet foods; these are lower in calories and, again, that is not the case. So, I think we have to perhaps recalibrate where we are putting the emphasis with respect to food health claims and nutrition education.

DR. KRINSKY: The "we" is correct but the "we" is not around this table with respect to that kind of issue. The only calories that we are supposed to deal with is the percent trans of total calories. But your statement is heard and I hope it will be heard very widely. Thank you very much, Alice.

DR. MAYNE: Just one comment, I know that we are talking about this in the context of cardiovascular but once foods make it on the marketplace we have to think about the global public health, not just cardiovascular disease. So, at some point I think the committee needs to think a little bit about the relationship between total fat intake in diabetes or total fat intake in cancer and total fat intake in other things because I think we get into trouble when we look at public health recommendations disease by disease rather than looking at the global burden of chronic disease.

DR. KRINSKY: Well, let's go down that road a little bit. Can somebody give me the relationship between the total fat of the diet and its relationship to diabetes, cancer, rheumatoid arthritis--you name it, whatever chronic disease?

DR. MAYNE: As a cancer researcher, I think I will start with the cancer one, also the second leading cause of mortality in the U.S. It is still somewhat equivocal. The strongest evidence is thought to be related to things like breast cancer which was the primary rationale for launching the low fat diet part of the women's health initiative. As you know, that part of the trial is continuing so we don't have a final answer yet on whether a reduced fat diet will significantly impact on the incidence of breast cancer. We have some cohort studies, things like the nurses cohort and the pooling project which has pooled data from approximately seven different cohorts together and seems to not indicate a strong association between total fat intake and breast cancer risk. But the data is still somewhat equivocal and I think we just need to be a little cognizant.

I think it is probably unlike cardiovascular disease but the type of fat is probably rather important. We have talked a little bit about omega-3, long chain omega-3 fatty acids. The evidence suggests, if anything, that they appear to be protective and not adverse in terms of cancer risk. That is pretty much a summary of where it is; the jury is still out at least until we have results. The only clinical trial is really the women's health initiative.

There is one other secondary prevention trial under way, the WIN study which is looking at secondary intervention of breast cancer. No results are available yet.

DR. RIMM: I think the evidence for diabetes also, if anything, points to specific fatty acids and not so much as that total fat may be an association for trans fat, independent of total fat. The one that we may argue the most about is looking at total fat and obesity. That maybe is what led to us to focusing on coronary heart disease because there have been so many trials on low fat diets and three months, six months, one year and two year loss of obesity. I think that area of trials is quite interesting and people argue both sides. I think part of the issue is that you have to follow long enough to see the long-term impact of a low fat diet. In some case people adapt to the low fat diet and in some cases people get fed up and can't stand the low fat diet. Short-term trials seem to show that low fat diets are pretty good at having people lose weight, much like the Atkins diet. In fact, after a year people were much better staying on the Atkins diet than they were on the low fat diet.

But I think looking at a few of the longer studies, if you stay on a low fat diet your weight starts to creep up to your baseline weight. So, it is another area where I think that the evidence is not strong that total fat leads to increased obesity. I think, if anything, it is total calories and not total fat.

DR. KRINSKY: I am sure your comment was not meant to be construed as an endorsement of any particular diet.

DR. RIMM: No, it was not.

DR. LICHTENSTEIN: I would agree with both of the prior speakers, especially because a lot of the relationships between diabetes, cancer, especially breast cancer, and dietary fat and just dietary calories--really there is a lot of confounding because, from what I understand, women that are fatter are at increased risk of breast cancer because of increased estrogen production. So, there is sort of a biological basis for it, and the same thing with diabetes, we are seeing a tremendous increase in rates of type 2 diabetes which is tied to body weight.

DR. KRINSKY: But is the body weight due to total dietary fat or total dietary calories?

DR. LICHTENSTEIN: I would personally say that the data support calories as opposed to dietary fat, especially because there are so many different types of diets that could be low or high in fat, as the two examples that I used of a diet that is high in fruits, vegetables and whole grains, which I don't think anybody would argue against independent of the fat content, versus an extremely low fat diet that is made up primarily of diet soda and non-fat brownies and non-fat ice cream and potato chips made with olestra. They are just non-comparable.

DR. KRINSKY: I think we have worked over the total fat issue. Do we have any other comments on it?

[No response]

We will return to this later on in our session but let's move on to our second question. The dietary guidelines committee may suggest--now, I need some help with this, at what level are the dietary guidelines committee ready to suggest? Does somebody know?

DR. JOHNSON: I have been attending the meetings and my understanding of where that recommendation is that the subcommittee that is working on fatty acid recommendations is deliberating that. The last sort of discussion about this topic was that 1 percent of energy as trans fatty acids as a limit was kind of a stake in the ground but they were going to talk about it more. So, this is certainly not a done deal at this point. I suspect there will be some more conversation about it.

DR. KRINSKY: Thank you very much. But whether it is a done deal or not we have the question, and that is, the dietary guidelines committee may suggest that less than 1 percent of energy should be obtained from trans fatty acids, about 2 grams per day for a 2,000 kilocalorie diet. Does the scientific evidence support this level of less than 1 percent or 2 grams per day for trans fatty acids? Yes, go ahead.

DR. LICHTENSTEIN: I am not sure actually that the scientific evidence is there. There are few dose-response studies on trans fatty acids and I am not aware of any that are really honed in on the range, let's say, between zero and 3 or 4 percent of calories to indicate that 1 percent is more efficacious than 0.5 percent or 1.5 percent. But I am a little concerned because if you reduce the fatty acid content of the diet you, by definition, need to change the fats that are used to make the foods or create a lot of low fat foods. If you change the fats, then you have to ask, well, if you dramatically decrease the trans fatty acids what are the types of fats that are going to be substituted? If they are fats that are primarily unsaturated, well, I think it would be difficult to argue against doing that. If it is a tradeoff and you decrease the trans fatty acid content by half but you triple the saturated fatty acids, then the net effect may not be what is hoped for.

DR. KRINSKY: That begins approaching our third question and that is the equivalency, if you would, of trans versus saturated fatty acids, equal, more or less, or not being able to measure it. Can you give some examples of studies where, in fact, the trans fat has been specifically altered?

DR. LICHTENSTEIN: Well, I probably can't find it right now in the booklet but the first study was Mensick and Katan in 1990, The New England Journal article, but that was not a dose-dependent study where dosage was assessed. There was one study published by Joe Judd, in The American Journal of Clinical Nutrition, in 1994, but they looked at 3.8 percent calories per trans and 6.6 percent of calories per trans and statistically they couldn't distinguish between them with respect to the increase in LDL cholesterol levels, although they were significantly higher than in a diet that was high in oleic acid.

Then, in 1999 we published a study in The New England Journal where we didn't look specifically at trans fatty acids but looked at different types of hydrogenated fat that had different levels of trans fatty acids so they were diets that had 0.5, 0.9 or 1 of calories as trans fatty acids, 3.3 percent, 4.15, 6.72 and 1.25 was butter, and among the lower levels of trans fatty acids statistically for LDL cholesterol and for HDL cholesterol the differences were not statistically distinguishable. At the higher levels they were but that was up at 6.7 percent. And, I am not aware of any other studies that specifically looked at different levels of trans fatty acids within the same study.

I think there is in the book a compilation of studies that was published by Ascherio in The New England Journal of Medicine in 1999, where he looked at the relationship of LDL to HDL cholesterol levels with increasing levels of trans fatty acids but that was across studies, and there seemed to be a dose-dependent relationship but there was essentially one point under 1 percent and then a number of points at 3 percent and then above. So, there are just not a lot of data there as far as just the dose of trans fatty acids and change in lipoprotein levels.

DR. KRINSKY: Is that that figure 8.4? What page number is that? It is section V, page 8-61.

DR. LICHTENSTEIN: Yes, that is the one I am referring to. But, remember, that is the LDL/HDL cholesterol ratio. Some people feel that total HDL or LDL, THDL is the critical factor, whereas others--for example, NCEP just looked at LDL cholesterol levels. So, you need to be somewhat careful about that but, yes, that is the figure I am referring to.

DR. KRINSKY: Could you clarify for me the relative significance of the LDL/HDL ratio as opposed to the LDL cholesterol and the HDL total cholesterol?

DR. LICHTENSTEIN: Right now I don't think we have enough data to determine whether the total to HDL or LDL to HDL ratios are different or equivalent. In the 1999 paper they used the LDL to HDL because they didn't have adequate data to look at the total to HDL cholesterol ratio, or that is what was told to me.

Now, as far as looking at the ratio with respect to risk of cardiovascular disease versus looking at just LDL cholesterol level really depends on which side of the Atlantic you are on because it seems that on the other side of the Atlantic it is felt that the emphasis should be on the ratio, whereas on this side of the Atlantic we still feel the ratio should be--and this is just sort of general and I am using it as an example, but NCEP decided to use as their criteria--they rely more on the LDL cholesterol levels.

DR. RIMM: I am not as much of an expert in this area as Alice, but I think part of the issue is that it is much easier to treat LDL than it is to treat the ratio. So, in epidemiological studies I think the ratio is more predictive of coronary outcomes than LDL is per se. If you go to your clinician and they measure your ratio, there are not clinical trials, thousands of clinical trials that have been done to lower someone's ratio. Essentially, there are thousands of clinical trials that have been shown to lower someone's LDL. So, I think that is part of the issue.

I think that if you focused on the total to HDL ratio or the LDL to HDL ratio it would probably give you better guidance on the impacts of tans on lipids because we do know that trans impacts both LDL and HDL.

DR. KRINSKY: Well, if trans impacts both LDL and HDL the guideline, as we all know, is that LDL is bad and HDL is good and what then does trans do to HDL in comparison to what it does to LDL?

DR. LICHTENSTEIN: Well, it depends actually on the absolute levels. Just for the record, table 8.10, which is on page 8-60 of section V, has a summary of the studies where there were some kind of dose-dependent relationships with trans fatty acids. I think it is quite clear that trans raise LDL cholesterol levels. They probably raise them slightly less on a mole for mole basis than saturated fatty acids but, whereas saturated fatty acids raise HDL, trans fatty acids don't. However, you need also to take into consideration what the current levels of trans and saturated fatty acids are and what they potentially could be, and that is that the estimate of trans fatty acid intake in the United States now is between about 1.5 and 2.6 percent of energy, and saturated fats are running between about 11-13 percent of energy. So, if you look at the data on HDL given those relatively low levels, although at very high levels of trans they could not currently be achieved in the U.S. diet without extensive manipulation or somebody probably just eating out of fast food restaurants breakfast, lunch and dinner, there is not that much effect on HDL but you still do see the effect on LDL so the ratio is probably slightly worse with trans fatty acids but the LDL is probably not particularly worse or a little bit better. So, it is sort of a wash. Then you have to come down to trying to predict what the effect would be of having a specific recommendation for trans fatty acids, again, back to what would happen to the food supply for very little differences.

DR. RIMM: I am sorry to go back to our meetings of the IOM but our charge for the macronutrient panel was to sort of take the best science and dictate what the science says and not necessarily make guidelines or give advice. It was to try to interpret the science. The interpretation of the science was that we could not come up with a health reason for having trans in the diet, and essentially that is what our final message was, that your trans should be as low as possible.

So, I guess taking Ascherio's summary of all the trials that have been done on trans and LDL to HDL ratio, I guess we went down as far as there are data, getting down to the 1 percent level and realizing also that it would be very hard to have a diet in this country without having some trans because you get trans in ruminant animals. So, I am not quite sure beyond that where the exact 1 percent came from. Maybe Guy can provide guidance. Maybe it was sort of a place to start and to work forwards or backwards from that, and I don't know if there is going to be better evidence to say that 1.5 is better than 1 or 0.5 is better than 1. I think it should be as low as possible, especially for children where we don't have as good data and there is the potential likelihood that there may be much higher levels of trans consumption among children or individuals who are more likely to eat at fast food restaurants or large-scale donor purveyors. You know, the two largest sources of trans in our diet are from fried foods and from donuts. So, I think starting with 1 as a starting point is a good place to shoot for but I think the overall summary was that there was no reason to have trans in our diet, no health reasons to have trans in our diet.

DR. LIECHTENSTEIN: I would agree with what Eric said, although I think part of the dilemma of the IOM committee was that we were asked to set an AI or an RDA for nutrients that were not essential, and that was somewhat new and I think the rubric that we were asked to use had really been developed for essential nutrients. Essentially, if trans fatty acids is an essential of saturated fatty acids, then it becomes very difficult to come up with any specific numbers and I think that is what the recommendation was, that for both saturated, trans and I also believe cholesterol the intake should be as low as possible because the data indicated that with any increase there was an increase in the indicators of risk.

Specifically, we focused in this case on cardiovascular disease but I think we have hopefully learned a lesson, and that is that we have to take into consideration how it becomes interpreted and translated into the food supply since we certainly got a surprise with the low fat message. Certainly, I have responded to people saying, well, does that mean they are poisonous because we are recommending to consume them, you know, at levels as low possible? Well, no, that is not exactly what we are saying. We are saying that there is no reason to include them in the diet and then we just have to make some educated guess on what the impact will be of getting them as low as possible in the diet relative to what they would be replaced with.

With respect to ruminant animals, if people adhere to the recommendation that saturated fat intake be as low as possible, then the trans from the ruminant animal fat will essentially come down. So, then it becomes translating the recommendation to minimize intake into one that doesn't result in an increase in saturated fat intake.

DR. SHANNON: I wanted to ask a question to follow-up on that. You said our best guess as to how it will be interpreted. Are there people around the table who can give us a best guess or has that best guessing been done with regard to how lowering the trans fat recommendations are likely to affect the food supply?

DR. JOHNSON: Clearly, the food industry has gotten the message about trans fats and there are earnest efforts under way to try and eliminate trans fats. Unfortunately, the technology to do that is still somewhat limited at this time. I am told there are four different approaches to reduce trans fats in foods. Let's see if I can remember them off the top of my head.

One is to use naturale sources that don't contain trans fats but tend to have more saturated fats, like palm oil and that kind of thing.

Another is to interesterify. What you do is to use sources of potentially totally hydrogenated fats which basically eliminate all the double bonds. So, if you eliminate all the double bonds you don't have any trans ones. Then you interesterify that with other sources of lipids so you get the functional attributes of hydrogenated fat that will allow you to have baked goods and things like that. That is a much more expensive process and it is still being developed.

I am also told that there is an issue there because consumers, many of them, have been attuned to the word hydrogenated in the ingredient panel. So, if there is something that says hydrogenated fat, there are some people who will, you know, go to great lengths to avoid that. Peanut butter is an example. It is basically a trans fat saturated food. The nutrition fact panel is going to say zero gram trans fat but there is a little bit of hydrogenated fat to provide functionality of peanut butter and there are a lot of people who are avoiding peanut butter for that reason.

So, my industry friends would love to see that issue addressed. It is not exactly our question but they would love to see some alternate nomenclature in the ingredient deck that would allow the use of totally hydrogenated fat in order to employ the interesterification technique without red flagging consumers. So, that is one.

Another technique is use of different fats, like medium oleic, canola oil, and so forth, so that fats that have been bred to have a different fatty acid distribution that would provide the functionality for some things.

The final one is genetically engineered fats. So, there are fats in the pipeline that alter the fatty acid content that would be a replacement for trans fats under some circumstances. But those are a ways off.

I wanted to respond again. I think it was Dr. Rimm who had raised the issue about how the dietary guidelines committee had gotten to this thing. I am not a member of the committee but I have been attending the meetings so I know what the discussion was. When the fatty acid committee was making their report about where they are at in this recommendation for trans fats, an initial thing that they considered was to basically recommend that sources of trans from hydrogenated fats be discouraged or eliminated and that trans fats from natural sources be accommodated in the diet. The discussion around that point was, well, that may be laudable but it is not practical given the state of replacing trans fats. So, then the discussion was, well, I think that the current level of intake of somewhere around 6.2 percent of calories is one reasonable target to shoot for.

I think that was based on more of an arbitrary thing than on hard science. Since then I have been given some information that is really interesting, put together by a guy named Ed Hunter at the University of Cincinnati. What he has done is take all of the studies in the figure that we have been looking at and look at the individual results, plotting percent energy as trans fats to changes in LDL and changes in HDL. What he did was he compared the control group and the trans fatty acid group from those studies, and the ones that were significant he gave a yellow dot and the ones that weren't significant he gave a black dot. Then he summarized them in a plot and, interestingly, the black dots or the level of trans fatty acids that resulted in insignificant intakes of LDL--and nobody is arguing that trans fats raise LDL; they do. I mean, you can take that to the bank. But at low levels the increases were not statistically significant.

DR. KRINSKY: I am sorry, what do you mean by low levels?

DR. JOHNSON: Less than 1 percent of calories, and there was another study at around 3.8 percent. That might have been Judd too, I am not sure. But those two are the lowest two points that around 1 and a little over 3 were insignificant. His conclusion was that there really was no data to show that intakes of trans fatty acids up to 4 percent of total energy significantly increased LDL. That is just based on the data that we have and there is a gap there, as Dr. Lichtenstein pointed out.

Then, if you look at HDL, which I think is really interesting, he did the same thing. What he found was that until you got up to like 6 percent energy from trans or roughly twice of what the current intake is there was no significant impact on HDL. So, their conclusion is that really there isn't good data to suggest that 1 percent is any better than 2 percent, is any better than 3 percent--hard data. So, I just thought I would pass that on because I thought it was a pretty interesting analysis.

DR. KRINSKY: Then how does that jibe with the earlier statement that the lower the percent trans fatty acid the lower the risk for coronary heart disease? Am I misquoting you, Alice?

DR. LICHTENSTEIN: Yes, I think so. I am not sure. Remember, the risk for heart disease that would be an association would be the epi data. In that case, pardon me, but I think you are guesstimating trans fatty acid. Essentially you are looking at dietary patterns and you are guesstimating it because you are relying on food frequency data and then you are relying on incomplete dietary tables. It certainly sort of points you in a direction to assess and it turns out that those data are very consistent with the clinical data.

But as far as the actual clinical data, there just aren't a lot that are within that range of 0-3 percent, but in the data that I am aware of, which came from our group, it was between 0.55, which is about as low as you can get, up to 1.25 and there was no significant effect on LDL cholesterol. It wasn't until you got to the 3.3, and the Judd study was--

DR. KRINSKY: I am sorry, you are looking at table 8.10?

DR. LICHTENSTEIN: Yes.

DR. KRINSKY: And your butter is the one with 1.25 percent trans fatty acid?

DR. LICHTENSTEIN: Yes.

DR. KRINSKY: And the value--

DR. LICHTENSTEIN: In the original paper that was reported separately because the saturated fat was so different and the cholesterol was so different. So, it is the 0.55 and the 0.91, and that was relative to the next which was the 3.3. That is because we were looking at products that would be commonly available to the consumer.

But there is one other issue and that does have to do with what would possibly be a substitute if you were going to try and get the trans fatty acids as low as possible. One of the issues that has come up is total hydrogenation so you end up primarily with standard stearic acids since most of the fatty acids in the vegetable oils are 18-0 and stearic acid does appear to have very little effect on lipoprotein levels and most of it is metabolized to oleic acid, which is a mono. But from a labeling perspective, it gets lumped with saturated fatty acids so from a manufacturer's perspective I would assume that there wouldn't be any advantage because their saturated fat would likely be very high. That comes then to a comment that somebody made about LDL raising fatty acids as opposed to trans or saturated fatty acids. And, it is difficult because we know with the other saturated fatty acids the magnitude of effect on LDL cholesterol is different. So, it would be quite complicated to think in terms of cholesterol raising fatty acids but, because of this anomaly of stearic acid and the potential advantage of stearic acid to substitute for trans fatty acids, it is something that ultimately we need to really think about.

DR. RIMM: I have several comments, if I can remember them. It is an important issue that Guy and Alice have brought up that, you know, no single study will be able to shed light on differences between very small increments in trans, which is why I think that it was useful that the figure was put together where Alberto Ascherio put together all the trials that have been done to show that there was a dose response which would allow us to look across the whole range of the percentages of trans fatty acids. So, I think that does give us more information than any one study.

Related to that and actually going back to Barbara's question that she put forward to Guy, you know, how would industry respond? I wish I knew more about how they responded in the Netherlands, and maybe you know more about this, but I heard they have essentially removed almost all of trans fatty acids from their diet there. I would assume that production methods aren't that much better in the Netherlands than they are here, but it does seem that we are much slower in adapting methods which could bring us foods that have a lower amount of trans fatty acids.

I know that two years ago, or whatever, McDonald's announced that they were reducing almost all of their trans fatting acids out of their diet and they sort of rescinded that when they realized it was too hard to retrofit the oil that gets spread on their French fries.

I am sure you know much more about this, Guy, than I do but I think we need to look carefully around the world and what other people are doing because I know that there are technologies out there, and the second method you described, maybe that is what they are doing in the Netherlands where they are essentially folding the fats together, esterification--is that what it was called? Interesterification. Maybe that is what they are doing in the Netherlands to get rid of the trans in the diet. But I know that it is existing and I would hope that if we did make a statement about saying 1 percent of trans it would push industry to move that much quicker. You know, it may be our one effort to help things along.

Finally, actually going back and thinking about the 1 percent and where the 1 percent guideline came from, part of it actually may have come from the nurses health study because the bottom 20 percent of the population of the nurses health study, their average intake was about 1 percent trans and those are the women that were at lowest risk of coronary heart disease. So, I think somewhere along the line someone may have been influenced by that although there are not many other studies where the lowest group has 1 percent energy from trans.

DR. JOHNSON: I just want to respond to that a little bit. I am not an expert about what is going on in Finland but my understanding is that trans fatty acids have been such a bad word there that a lot of the margarines have used saturated fats to replace trans fatty acids. I don't think it had been the interesterification thing primarily. So, I think we need to be careful about that.

With respect to the way that industry has responded, I think that the trans fatty acid rule that is out there requiring companies to list the grams of trans fatty acids has certainly elicited a big response. You have seen the snack food industry totally eliminate trans fatty acids. It is happening all over the place and I am almost wondering if by the time all of this gets done it will be a moot point because the industry is driven by consumers. Consumers have gotten the word about trans fatty acids and they will continue to get it. So, I do think the industry has been pretty responsive here.

With respect to the company you mentioned, shame on them for making an announcement before they were sure they could do it. I'll bet you they won't do that again.

DR. LICHTENSTEIN: As far as the trans fatty acid intake of Western European countries, there was a study, I think it may have been called the trans fat study. It involved 14 countries and it was published in the European Journal of Clinical Nutrition, in 1990, volume 53, page 143. Their range was 0.5-2 percent of energy so it was lower than was estimated in the United States, or at least the 2.6 percent.

But as far as the Netherlands and Finland as far as really getting the trans down, I think we need to keep in mind that their food intake patterns are different. The major sources of their trans fatty acids are not from fast foods. A higher percentage from fast foods is in the United States. So, that allowed them to manipulate margarine and have a bigger impact on the trans fatty acid intake than we would be able to by just manipulating the margarine in this country.

DR. KRINSKY: Can somebody tell me what the trans fatty acid consumption would be for a person who eats no hydrogenated fat containing trans fatty acid but will eat animal protein three or four times a week? Is there any idea of evaluating what percent of their fat will be trans fat or is it so small that it is insignificant with respect to what comes from hydrogenated fat? Is there data available?

DR. LICHTENSTEIN: There are data available. You have to do the calculations. I think the best data on the distribution of the major sources of trans fatty acids in the U.S. diet actually come from the FDA web site. There is a great pie chart and you can actually see that. But that is for the whole conglomerate. I think we really need to keep in mind that you are describing one type of person. If you have somebody else who has lunch and dinner with commercially fried foods and baked foods that would be very different. The calculation certainly can be done and probably is represented in the range of intakes from the health professionals and the nurses.

DR. KRINSKY: So, I don't even hear a consensus.

DR. RIMM: The question was can you achieve 1 percent by just eating a normal diet with four servings of meat and would you still be below the 1 percent threshold? That is really probably the direct question you are getting at. I am sure there are some really smart people that are sitting behind me right now who probably could answer that question. She has probably looked at that much more than I have but I don't know if she is allowed to say anything or not. I would imagine it could be below 1 percent. Oh, it is already in the book we already read? Thank you, smart person!

DR. KRINSKY: On page 8-45, it has been brought to my attention by a smart person that the middle paragraph which, in fact, I see has yellow marker in my particular copy meets 0.01-0.21 grams per serving. So, four servings of meat could contain close to a gram, and there are people who eat more than four servings of meat.

DR. LICHTENSTEIN: That is less than half a percent. But you would hope people would be consuming fat-free milk.

DR. KRINSKY: Hope burns eternal.

DR. RIMM: The 1 percent DV is 2 g/2000 kg. So, you have to be below 2 grams so you could achieve that easily.

DR. HINE: I have to ask a question about serving sizes of meat. When people talk to me about meat, how many grams, how many ounces would this be a serving? It is a question I had when I read the information because 8 oz of meat is quite a bit more than 3 oz of meat.

DR. LICHTENSTEIN: One would assume it would either be the serving size defined by the food pyramid or by the nutrient facts.

DR. HINE: Can you tell us what that size is?

DR. LICHTENSTEIN: No, I can't but I know the difference between the two systems.

DR. RIMM: I think there is also some evidence out there to suggest that the trans fat achieved from hydrogenation has different health effects than necessarily the trans fat from milk or meat products. There is a case-control study that was done in Costa Rica looking at hydrogenated fats consumed in the diet and hydrogenated fats in the adipose and those hydrogenated fats that were made through hydrogenation of oils that were found in the adipose were much more strongly related to risk of non-fatal heart disease--it was a case-control study--were much more strongly related to risk of coronary heart disease than the naturally occurring trans fats that are found in ruminant animals. Now, that would require 15 more studies that Alice would need to do on LDL and HDL to show the differences, but I think there is some evidence to suggest that there are health differences.

DR. LICHTENSTEIN: True, but you don't know what those two sources of trans fatty acids--whether they deposit equally in the adipose tissue. There may be a selective metabolism of one isomer versus another isomer. We don't know. You can't get the purified isomers so it is really difficult to do those studies but, again, it raises an issue that needs to be considered and reconciled as some point.

DR. KRINSKY: So, as I read this and hear you, we have the epidemiological evidence that the lower the trans fatty acid consumption is the safer the diet is with respect to coronary heart disease.

DR. JOHNSON: I don't want to be imposing or passing myself off as an epidemiologist but through the wonders of the Internet I took a look at some of the epi studies that looked at trans fat intake and cardiovascular disease instance or mortality, and all of the ones that I could find that I had access to showed that the effect didn't become significant at the 95th percentile until the upper quintile of trans fatty acid intake.

So, for instance, there was a case-control study reported just this year by Clifton. It was a small study, only 209 cases and 179 controls but the relative risk for heart disease instance in the quintiles is 1, 1, 1.3, 1.23 and then 2.25, with the upper quintile being significant.

A case-control study reported by Mensick, in Boston, with heart patients, the same thing. After you controlled for all of the confounding variables, including high blood pressure, smoking, alcohol intake, family history, physical activity, MUFAs, PUFAs and saturated fats and cholesterol, the relative risk was 1, 0.81, 0.4 which, interestingly, was protective at intakes of 3.35 grams a day, 0.72 and then the upper quintile was significant at 2.03.

The nurses health study and, Eric, correct me if I am misinterpreting this, but ultimately when the data on trans fatty acid intake were controlled for all of the variables that they could think of the relative risk was 1, 1.09, 1.16, 1.24 and 1.53 at intakes of 2.9 grams per day, and only the last one was significant.

I won't bore you with the other one but there is another one that shows the same thing. To me, that kind of jibed with these intervention data that showed that there really wasn't a significant change in LDL or HDL at low levels of intake. So, I didn't think that the epidemiology gave us a lot of data at the lower intake but it certainly does show that there is a correlation at the higher intake, which I don't think is a point of contention.

DR. LICHTENSTEIN: I would also like to point out that we are looking at associations there and that there are a lot of factors that covary with trans fatty intakes which are corrected but there could be residual confounding. I think those studies are important and they are all very consistent, and I think that is the important point, that trans fatty acids are associated with increased risk but I think you need to take into consideration that there are other confounders that may be operating there.

I forgot the exact words you used but I am not sure that I would be as confident in those specific words but certainly we come down to there is no biological reason for people to consume trans fatty acid so you can't recommend them. By having a daily value, that really puts scientists in a difficult position because it implies that maybe there should be a certain minimum level the way there is for something like ascorbic acid, and that is just not the case.

DR. KRINSKY: The question that we are dealing with is less than a certain percentage. That is not recommending that it should be a certain percentage, and maybe this question is really referring to the reality of the American diet and not to the scientific evidence with respect to what the exact relationship is between trans fatty acids and coronary heart disease risk. Because to get to zero trans fatty acids would eliminate meat products, animal products for example, and I don't think that we should be in a position to advocate that position at all. So, I wonder if, in fact, we are not in the same position that the dietary guidelines committee was in. That is, that there is a reality imposed upon the scientific evidence and that you can say you shouldn't do this but it is impossible to live in a world where you don't do that.

DR. LICHTENSTEIN: Well, I don't think anyone is suggesting that there be zero trans fatty acids, just because of what you are saying. The IOM report was specifically very careful about the wording because it is essentially impossible to have nutritionally adequate diets without trans fatty acids. I guess what I am concerned about is what the general interpretation of the DV is and how it was originally defined, and that is the same concern that I had with the IOM committee where a system was set up for essential nutrients that below a certain level would cause deficiency and then you were imposing that system on nutrients that there is no biological requirement for. I think that is something we can't ignore.

DR. KRINSKY: That is very similar to what we encountered with carotenoids for example. We were not able to find a biological basis for recommending an AI or minimum amount. Eric, I have a lot of trouble seeing your light. I don't want to prohibit you from using your computer.

DR. RIMM: Sorry, I was too busy stewing over here over what Guy just said so I chose not to turn my light on. You know, I think it is true that any one study will not find highly significant associations between the second quintile and the first quintile. Usually, by definition, the variation between quintiles two, three and four is much less than quintiles one and quintiles five just because you have to cut the distribution that way.

I think any one study when it is an association study is unlikely to find significant impacts in every increasing quintile. However, I think by putting together the intervention studies, it does suggest that there is an impact on LDL cholesterol across a range of trans fatty acids. I think now there is evidence that Alice has that it increases LDL; there is evidence to suggest that trans impacts on adipose-derived inflammatory markers. I mean, I think there is a whole number of other factors we haven't even started to address related to trans so I don't think we should get too lost in the fact that one single study may not find a highly significant impact of trans at any given level because there is measurement error and there is going to be random error so that studies will jump back and forth. But I think if you sort of take the totality of evidence it is a pretty strong suggestion of a dose response between increasing trans fat and increasing biological markers and increasing coronary heart disease.

DR. MAYNE: I agree with Eric's comment. When you were reading the point estimates from the nurses health study, what epidemiologists tend to use is not necessarily each individual point estimate but we also do what is called a test for linear trend. So, while Q3 versus Q1 may not be significant, if there is a linear relationship across those quintiles that is something we also look at, not just one quintile versus another. So, just be aware that we also consider the test for linear trend.

Back to Dr. Lichtenstein's comments about how it seems awkward to use this, a non-essential nutrient that presumably has these adverse effects, using it in the setting of a daily value where I think we are struggling, I was struck that it would almost make more sense to set an upper limit for trans fatty acids rather than say this is your daily value of this. To many consumers, that kind of implies that it is a useful nutrient. So, in trying to communicate to consumers it is something we ought to think about--would it be more useful to say the upper limit of trans fatty acids we recommend is no more than X grams per day, rather than trying to use a DV which I think is becoming confusing to the consumer?

DR. LICHTENSTEIN: I guess the issue--and I don't know if anyone from the FDA can respond--with respect to the current regulations as far as labeling is does that help? Can we do that? I mean upper limits is a term set for the DRI but as far as the FDA, I am not sure they even recognize that term and that is a conundrum that we really face.

DR. KRINSKY: That is what the FDA is there for, to deal with conundrums. You are absolutely right, Alice, that is, the UL is a DRI. It is an IOM/NAS term--FNB/IOM/NAS term, to be more precise, and it is not an FDA term, if I am not mistaken. Is that correct? So, even though it is a very good idea and may, in fact, be more reflective of the way one should approach trans fatty acids, we may not be in a position to make that recommendation. So, there is DV. You don't have an upper DV, a UDV? No. May we create one?

MS. PELICAN: I guess this is not the first time that we face this because I am looking at a label that I brought on the plane with me yesterday. We have cholesterol which in fact reflects the same kind of direction for sodium. You know, these tend to be nutrients for which we are trying to put an upper limit on. But I would agree, I have struggled with looking at the tans fat materials that we were given before the meeting. But I know on the same package--and, you know, this is off fried chips which were pretty tasty on the plane--this is like one of these products where I thought I am really looking forward to trans fat labeling to see how much trans fats are in some of these foods that I like in a serving. I guess one package has total fat as 7 grams; saturated fat is 1 gram; and there is no trans fat, which kind of goes to what Guy is saying.

You know, I think there are a lot of industry folks that are looking to eliminate trans fat. But it has--and I think this is optional, voluntary labeling--a double asterisk and at the bottom and it says intake of trans fat should be as low as possible. I am just wondering if we are kind of faced with what FDA was faced with 13 or 14 years ago of was there enough information at that point to label trans fats and they thought no. With all of our briefing materials we know that over time there has been enough information and I am just wondering if we would want to discuss the possibility of saying we don't have enough information but might recommend some kind of a statement like this, which I found informative but, again, we only have 275 million people reading the same food labels and we always have to think about how will the majority of people interpret that. But I did find this informative, both the label and the footnote.

DR. KRINSKY: Thank you very much, Suzie. I just want to ask a question. How high can trans fat be to be declared zero percent trans fat, 0.5?

DR. LICHTENSTEIN: I think there is no one that has argued that we should encourage trans fatty acid intake or that we need trans fatty acids and I am just wondering if a way of getting around the issue of DVs--we seem to accept it for saturated fat although there is no biological basis for recommending a minimum amount of saturated fat--would be to just combine sats and trans with the DV that has already been established for saturated fatty acids. Since they both increase risk of cardiovascular disease, you may argue on a little bit more than the other but it may be compensated because you consume less of one than another but those, to me, sort of seem to be somewhat minor arguments. The major health advantage would be to minimize intake of both. So, what would the disadvantage be of just having one DV for trans and sats?

DR. RIMM: There are about 20 pages of the Federal Register that are in our pamphlet which talk about why we don't want to do that. I think--well, I don't even want to go into it because it looks like it has probably been discussed for four straight years at the FDA but there are a lot of reasons that they have already decided that that is not the case for at least putting it on the label. They decided to separate trans and sats on the label and I think they decided for a good reason. These are different chemicals; they have different biological effects; there are probably a lot of effects related to trans that we don't even know yet that probably don't impact sats. So, to then say, okay, let's separate the gram amount on the label but put together the DV values, to me, seems a little incongruent. I think if we think it is important enough to separate the trans from the sats, then we should probably separate the daily values also.

DR. JOHNSON: My industry pals are divided on this issue. They are as confused as everybody else I think. However, the recent IOM report on applications suggests that they be combined. So, I think, like you say, Eric, it opens up a whole morass because then you get into the physiological versus the chemical differences and the stearic acid issue emerges and it all becomes very complicated.

I wanted to respond to Suzanne's comment about the footnote. The footnote is another thing that my industry folks are not highly supportive of. It, again, was discussed in the Federal Register, whether by calling attention to any one item or multiple items in a footnote people who read that footnote assign it an unrealistic priority, I guess, and by putting that in the footnote--some consumer research was done showing that when you put that footnote in people will make incorrect choices for things that are minimum in trans fatty acids but may have way more saturated fat. So, it kind of gives them a focus on trans fatty acid and dilutes the importance of saturated fats. So, there are some issues associated with that too.

DR. LICHTENSTEIN: I understand that there couldn't potentially be biological differences between trans and saturated fatty acids but biological differences among saturated fatty acids, and I think the issue that has come up and gone down, come up and gone down still comes back to potentially looking at something like cholesterol-raising fatty acids. We knew and we explored and we wrote in the IOM report that there were a lot of different criteria that could be used to set an AI or a DA or DRI for any of the micronutrients, and what we did was go through them all and then chose the one that we felt either had the most evidence or was the most important as far as health outcomes.

I think with respect to fatty acids, probably the most evidence is available for either LDL cholesterol or the ratio. But if we really want to see a decrease in trans fatty acid intake, and I think there is certainly good reason to want that, and if one alternative is to increase the saturated fatty acid intake, if that is actually really correct and will result in the expected outcome, then, again, maybe one needs to consider coming up with a different designation to put on the food label, and maybe that is close to erasing fatty acids.

DR. KRINSKY: Would you also have cholesterol-lowering fatty acids as part of a label?

DR. LICHTENSTEIN: You mean the unsaturated fatty acids?

DR. KRINSKY: Yes.

DR. LICHTENSTEIN: Right now I believe it is optional to include monos and polys so that is already in place. I certainly don't think that that should be eliminated. I think obviously it is something that should be considered and one would need to understand how the public would interpret that, as they would interpret cholesterol-raising fatty acids, but the bottom line within the nutrition facts panel is that you want to give consumers as much clear information that they can use to make educated decisions on what items they are going to purchase and, hopefully, that will be in favor of health promotion as opposed to the opposite. So, there has been a huge shift in what the consumer is aware of. There have been a lot of terms that weren't in the lay press ten years ago that are in the lay press now. I think there has been a lot more on the labeling and on regulations and health claims than there has been on education but ultimately I think we need to understand more about how the consumer is going to use this information and whether those types of designations would actually be useful in communicating any kind of information that the scientific community feels is important to communicate to the consumer because that is where the bottom line is.

DR. KRINSKY: Does the consumer know the difference between cholesterol and LDL cholesterol?

DR. SHANNON: That is the very point I was going to bring up. I think the consumer is aware of "good cholesterol" and "bad cholesterol." If I am reading our material correctly, trans fats do lower good cholesterol. So, coming back to your point, Alice, if you are going to use this terminology of cholesterol-raising and cholesterol-lowering, what are you going to do with good versus bad? What kind of message are we sending to the consumer when we get into that?

DR. LICHTENSTEIN: Well, within the context of the amounts of trans fatty acids that are currently consumed in the United States, the effect of HDL cholesterol is not significant. It is significant in the clinical trials where you artificially increase the trans fatty acid of the diet so that you are really pushing the system to understand what goes on and, in that case, clearly, there is a big difference between trans and saturated fatty acids and it is something that shouldn't be lost. But it appears that trans fatty acid intake in the U.S. has gone down with the introduction of the trans-free margarines. Although one company said they were going to eliminate trans and then didn't, another company actually has done this. So, at least you see industry going towards less trans.

But what is on the label is not "good cholesterol" or "bad cholesterol" though, you are right, I mean I think it is what the consumer understands. What is on the label is the composition of the food and I think what we really need to understand is how best to communicate that information so that consumers will make the best choice, or how that information, by presenting it, will nudge industry into providing healthier alternatives. From what I understand, it is not clear whether the nutrient facts panel is most important because people are actually comparing and making a decision on this food or that food, or that industry wants their food to look pretty good on the nutrients facts label. And, that is fine. In either case it benefits the consumer and with nutrient labeling that is what we are talking about. That is very different than the guidelines or the IOM reports.

DR. HINE: I have two questions that are unrelated to one another but the first one we have touched on already. First of all, what are the data about who reads food labels? That is question number one. Question number two is regarding differential effects of age, gender and ethnicity and the duration of the clinical trials and how the subjects were fed during those clinical trials.

DR. LICHTENSTEIN: Well, I can certainly address the second question; I can't the first As far as the clinical trials, most of them were actually done with moderately hypercholesteremic individuals so the individuals would be candidates for dietary modification but not drug modification. So, a lot of the clinical trials were done in individuals with LDL cholesterol levels above about 130. Most of them were done with both males and females. There is very limited evidence that females might be slightly less responsive than males but that evidence is equivocal at this point.

While most of the studies were done with older individuals, we know that women after menopause are at increased risk for cardiovascular disease. Their LDL cholesterol levels go up. So, if you use a minimum LDL cholesterol level most of the young females wouldn't qualify for the clinical trials anyway.

As far as the duration of the feeding trials, most were done for at least four weeks. There certainly has been a lot of work on how long it takes for lipid levels to stabilize after you alter the fat content of the diet, and that certainly seems to be an adequate period of time. If you think about the half-life of lipoprotein particles, it certainly is more than adequate. But there are very good data that if you keep people on controlled trials for up to 24 weeks and monitor every 4 weeks you get a maximum effect at 4 weeks. If you are providing all the food so that you are reasonably assured that they are maintaining the diet, the line is straight. You don't see any going up or down.

As far as the individual trials, I think most of the ones that are summarized in this book were controlled feeding trials where all the food was provided. I mean, there are a couple of studies, but they are so designated or separated out, where just the major fats were provided to the study subjects and they followed their habitual diet. But I think what we are talking about now is really the controlled clinical trials.

I am not aware of any data on comparisons of young individuals versus older individuals, but I think that that probably is a moot point. I think what is critical is those individuals that are candidates for dietary modification or for pharmacological intervention. If you change the diet, can you have a clinically significant effect on the standard "risk factors" for cardiovascular disease? So, the studies that are summarized here pretty much meet those criteria.

DR. HINE: You didn't mention ethnicity. Are there any data regarding that?

DR. LICHTENSTEIN: With respect to trans fatty acids, not to my knowledge. I assume the NHANES data does break it down by ethnicity but not as far as the controlled feeding trials. They tend, by definition, to be small because they are extraordinarily expensive. The trials that are currently ongoing--I think it is either a 15 or 20 percent minority representation and that is mandated by the NIH, but that wouldn't give you the power to actually look at response on the basis of ethnicity.

DR. KRINSKY: With respect to Jean's first question, we only have an N of one, and that is Susan Pelican who reads the labels. I wonder if somebody from the FDA might have some information on the efficacy of labeling. Have any studies been carried out on the reading of your labels? If you have an answer, if you could use the podium we would appreciate it so that we and the audience can hear it, and if you could identify yourself for the recorder.

MS. WILKENENG: Virginia Wilkeneng. I don't have any of the numbers with me but certainly the agency has done a lot of studies, as well as the food industry and groups such as the American Dietetic Association over the years since we implemented the new nutritional labeling regulations in '93.

Most of those studies show that the majority of consumers do look at the nutrition facts panel, particularly the first time they are buying a product. Of course, they are looking at different things. Some are looking at calories; others at fat; others at sodium, depending on what their personal needs are. But they look at that information and they tend to use it to compare products or just to find out information about that particular food. They don't particularly use it to add up numbers and plan a daily diet, but in making the purchasing decision all that research that has been done does show the majority of people are familiar with the nutrition label and look at it.

DR. KRINSKY: Thank you very much. Jean?

DR. HINE: How were the data collected, if you recall?

MS. WILKENENG: Let's see, within the FDA studies we have done telephone surveys and we have done mall intercept surveys. I couldn't speak for ADA's studies and maybe the industry, when they come up and make some comments later, can indicate how some of their studies were done.

DR. KRINSKY: Eric?

DR. RIMM: I have a question for the FDA, as well as bringing us back to the original question here. The original question was does the scientific evidence support the level of 1 percent? I think what we are tripping over is 1.1 or 1 percent. I wonder if we could use the analogy of saturated fat or cholesterol. Right now there is a daily value for saturated fat. I don't know if it was in this packet or anywhere about how that was derived and why 1 gram less for saturated fat was okay and 2 grams more for saturated fat was not okay. Is there anybody who can give us some guidelines of how the exact daily value for saturated fat came to be?

MS. WILKENENG: The current daily value for saturated fat was based on recommendations and dietary guidelines and other places for less than 10 percent of calories coming from saturated fat.

DR. RIMM: Again, that was guidance, not necessarily hard science saying that 10 is that much worse than 9.9 or that much better than 10.1. So, it is somewhat of a guidance as opposed to taking the exact hard science and saying this is the threshold.

MS. WILKENENG: Right, but we did have those recommendations in reports such as the Dietary Guidelines or American Heart Association and others. The exact recommendation is less than 10 percent. If you do the mathematics for a 2,000 calory diet it comes out to 22 grams. So, we rounded it out to meet the less than, and make it easier for consumers to use.

DR. KRINSKY: Thank you. Alice?

DR. LICHTENSTEIN: That is actually a really good point. My understanding is that less than 10 percent, less than 300 mg of cholesterol per day are actually somewhat historical and were established before the criteria for setting numbers was really as stringent or has come under as much scrutiny as now. It dates back to NCEP ATP I, and my understanding is that before that it is from the American Heart Association. I don't know which actually came first. The point is well taken. I am not sure, were it to be done today, that one would come up with those numbers but I know there have been efforts to try to trace back where they actually came from and it hasn't been solid.

DR. KRINSKY: Well, historical is okay; hysterical though is to be avoided and I think that that is what we want to avoid with respect to trans fatty acids. We are asked for the scientific evidence and it is not clear that we have what people might define as scientific evidence, that is, rigorous studies that can help us come to a value of less than 1 percent, less than 1.1, less than 0.9. Alice?

DR. LICHTENSTEIN: Well, there is one other factor, and that is I am not sure we are ever going to have that evidence because when you alter the trans fatty acid level of a diet, if you decrease one thing, if you are going to keep it isocaloric which you need to do so you don't have other confounding factors, you have to increase something else. And, if you increase carbohydrate or if you replace it with saturated fat or you replace it with mono or replace it with polys you are going to get slightly different results. So, it is almost as though we may be asking for something that probably we can't really deliver.

DR. RIMM: I think we should also keep in mind what current health claims are allowed and that, in fact, we don't always have to have a ton of data and 16 clinical trials to make a point. Right now we have said that you can claim that fruits and vegetables or a whole grain product and, in fact, we don't have large long-term clinical trials showing that fruits and vegetables are better or that whole grains are better. You can say the same for exercising and you can say the same for smoking. So, I think we have to be a little bit careful about saying what is going to be the threshold before we set a number for trans. You know, I think if you ask can you eat meat and drink dairy products and still be well under the 1 percent, I think the answer is yes.

DR. LICHTENSTEIN: Taken from the other perspective, we have something like fish which sometimes can be high in fat and high in omega-3 fatty acids and we have reasonably good data that that would be beneficial but we don't have a health claim for that.

DR. KRINSKY: That is because of the PCBs. We are scheduled for a break at 2:45. It is now 2:40 and, with the committee's approval, I am going to call for a 15-minute break. I would just like to remind the committee not to have private meetings on this topic until we return because otherwise that will make the whole process invalid. So, we will reconvene at 2:56.

[Brief recess]

DR. KRINSKY: Thank you very much. I hope you are all refreshed and ready to continue the discussion. Let me just make the suggestion that when on occasion you want to make a point on some of the material that is in the notebook that was handed out to us, if you could just draw our attention to the section, the page number, paragraph, line, word, etc. so that we can all follow on it.

We are still in the midst of deciding, since we can't change the world and change the definition of fatty acids that increase cholesterol levels although there is certain merit to that, we are still left with the situation dealing with the trans fatty acid. What I hear is that there is not adequate scientific evidence to set a figure for trans fatty acids in the diet as far as a high level or a low level, and that we are going to have to try to answer this, as they say, by flying by the seat of our pants, an expression that everyone understands and I have used frequently and I hope that you people understand what I mean.

So, is there an argument with me when I say that there does not appear to be scientific evidence to support the level? Dr. Rimm?

DR. RIMM: I am sorry, the argument is I guess I don't know what you mean by scientific evidence. We have scientific evidence to support a daily value for saturated fat at some level, which I guess was set based on the fact that 30 years ago HA flipped a coin and said let's take 10 percent of calories from fat. So, I think if that is our definition of scientific evidence, then I think we do have scientific evidence to suggest that at some level of tans fat there is harm.

DR. KRINSKY: I am not sure that I agree that flipping of a coin is considered to be scientific evidence but, in fact, a claim has been made--a statement has been made and there has been agreement at some point in the past as far as saturated fat is concerned. What that was based on or not based on I am not sure. We certainly have, in terms of the epidemiological data, associations between trans fatty acid levels and risk for coronary heart disease. Certainly, associations are considered scientific evidence. I will grant that to the epidemiologists, out of the kindness of my heart, but we don't have support from intervention studies for that particular figure. What I have heard is that there does not seem to be a great deal of difference between levels below 1 percent and levels above 1 percent with respect to LDL cholesterol levels or its effect on LDL cholesterol levels and that, in fact, this is not the easiest kind of experiment to run because it you alter the level of trans fatty acids you are altering something else, and that something else that you are altering may not be as important as the trans fatty acid but it has to be taken into consideration. So, I am in a bit of a dilemma and I am willing to be argued into any position; I am easy.

DR. RIMM: Well, I don't want to rehash our discussion from before but, again, I think it would be very hard but there is no one trial that tested 1 percent, 2 percent, 3 percent, 4 percent of trans and looked at a change in something which we would deem scientifically important. But if you put all of the clinical trials together, all metabolic studies together they do suggest a trend that with increasing trans intake there are increasing levels of LDL or LDL to HDL ratio.

Having said that, could we really set a level at 4 percent? I don't know if we could. Could we set it at 6 percent? If we can't set it at 1 percent, why do we think we can set it at 4 percent or 6 percent because there is not one trial that said, okay, 5 is worse than 4 and 4 is worse than 3. So, I think what we have to do is put all the evidence together and see what it shows us. It shows us that there is a pretty straightforward dose response. Obviously, there is some variability around that. So, I don't know how we will ever be able to specifically pick any point on that line.

DR. JOHNSON: I think I agree with what Eric just said. I mean, it seems pretty clear, based on the intervention trials as well as the epi stuff, that at some point trans fatty acids raise LDL and lower HDL. The devil is in figuring out when that occurs and, based on the data that is available right now, I don't think there is evidence to set it at 1 percent, 2 percent, 4 percent or whatever. I don't think we are being asked to do that. The question says is there evidence to support 1 percent and in my mind the answer is a resounding no.

DR. LICHTENSTEIN: I agree with the previous two speakers and I think were a level to be set one would have to take into consideration the balance of the diet and what is reasonable within the context of what is available in the United States. Having said that, given the range of dietary intakes, it would be pretty darned hard to do.

DR. KRINSKY: Could you remind me what the dietary intake is for trans fatty acids in the U.S.?

DR. RIMM: It was about 2.6 percent I think the last time it was measured. That includes hydrogenation, and donuts, and fast food restaurants. That is the average. But I think it is very easy to attain 1 percent.

DR. KRINSKY: The question I want to know is whether that 2.6 percent has been changing with time.

DR. LICHTENSTEIN: We don't have any data. I suspect it is and actually I think in some of your studies the average is lower than that. Isn't it closer to 2 or 2.1?

DR. RIMM: Yes, right. For the nurses and the health professionals the average is in the low 2's.

DR. KRINSKY: They are not normal!

[Laughter]

Well, no, they don't represent the average U.S.--

DR. RIMM: They represent a very wide range but, you are right, nurses and health professionals are not average; they are normal though.

DR. LICHTENSTEIN: Also, the data were collected differently because that was food frequency. That was interpolated for food frequency whereas the 2.6 I think was from the NHANES data, or something like that, but we know that the food supply has changed.

DR. RIMM: Right. So, it is already going down and we can maybe estimate that the average is around 2 right now. Although I don't know, you know, a few of the companies have taken trans out of their products but I don't know how quickly that is impacting on the overall diet considering that a lot of the trans comes from fried bakery products and from fast food potato products and I don't think those have changed that much yet, although, as you said, one fast food restaurant I think is attempting to take it out.

DR. KRINSKY: So, it seems that in the last post-break conversation I have heard three people say that we don't have the scientific evidence to support this level. Although we may feel, which is not scientific, we may believe that this would be a good level for the U.S. population to achieve, as far as scientific evidence is concerned, it is at this stage lacking.

DR. RIMM: I think the scientific evidence is as good as what we have for cholesterol and saturated fat.

DR. KRINSKY: But we are not asked about cholesterol and saturated fat.

DR. RIMM: But they are going to be on the same label.

DR. KRINSKY: We are not even asked about the label--well, yes, this does go to the labeling. I am sorry. This is something that has to be mulled over individually and maybe with a little re-reading of pages 8-40 to 8-50, or something of that nature, in tab 5--and, as you know, we can't discuss this privately. We are not obligated to vote on this issue at this particular moment and I am not sure if we have any more discussion. Suzie?

MS. PELICAN: Just to follow-up with this thought about saturated fat, kind of weighing that issue, is it appropriate for us to ask for any kind of summary? I mean, it seems like we have some good folks around the table here who have a sense that that saturated fat was historical, maybe just based on the best policy recommendations and not quite as much science maybe but it seems like we really don't have a sense around this table for being able to say firmly this is how that saturated fat DV or that 10 percent recommendation, which is historical, came about. Could we ask for that kind of research? Is that appropriate? That could be information that we could then have and discuss?

DR. KRINSKY: It would be appropriate because the process is going to be the same and, Alice, you said you tried looking at how that figure came up?

DR. LICHTENSTEIN: Indirectly because the Heart Association had been asked to justify the figure of that and the cholesterol so, of course, they bounced a telephone call to me and I then tried to get some information from them. I don't know, maybe the standing NCEP committee would have that. It may have come from data from other countries that had intakes of less than 10 percent of calories and less than 3 mg of cholesterol a day and they had better outcomes with respect to cardiovascular disease, but that is speculation. It is not something that I had the time to really pursue.

DR. KRINSKY: Yes, but if it came from, say, other countries that had the dietary information, is that really comparable to what is going on here? Don't they have different life styles? How can you base whatever their health outcome might be solely and exclusively on the cholesterol level in the diet?

DR. LICHTENSTEIN: I was just speculating on where it came from. It has to be at least 20, if not 30, years old and I think people thought about things differently then.

DR. RIMM: Yes, I think at the DRI committee meeting Scott Rundy and Ron Krause were reflecting when they were discussing this for the HA 20 or 30 years ago, and I think there were just sort of a few people sitting in a room and picking a level which they thought seemed appropriate from the available evidence in the 1970s or 1960s before they actually spent a lot of time talking about LDL and HDL. At that time they were just talking about total cholesterol. So, I think it definitely is not based on as much fact as we have now.

DR. LICHTENSTEIN: Also, keep in mind that when it was first proposed, it was proposed as a target for the population that had extraordinarily high risks for cardiovascular disease and very high intakes of saturated fat. It was not proposed to be used as criteria for labeling or anything like that. So, the whole basis for it was very different. It was a target.

DR. KRINSKY: But targets tend to move and we have seen that with total plasma cholesterol recommendations. That is a moving target which has been moving down and down and soon will reach my level.

DR. LICHTENSTEIN: Well, it has been a moving target for saturated fat because ATP I and ATP II had less than 10 percent of calories from saturated fat, whereas ATP III got rid of the step one diet and just now has the TLC diet, the therapeutic life style diet, and it is less than 7 percent of calories from saturated fat. So, that was the prior step two because they felt that the population had moved so close to the step one diet that that was no longer needed, that that should be considered the norm and then the TLC was the step two diet and, in the same way, to total fat was changed from the first two, ATP I and ATP II, less than 30 percent of calories from fat and ATP III is 25-35 percent of calories as fat, again de-emphasizing the total fat.

DR. KRINSKY: We have talked here and a number of you have raised the daily value and Suzie was good enough to give us a chart that she had taken with her off the airplane, with permission I assume. We have here a copy of a nutrition facts label and Virginia Wilkeneng is going to be willing to talk to us about deciphering this code. There are copies outside for the audience. I don't know if everyone in the audience has received a copy. It would only be fair to our audience if they could--could somebody go--they are doing it? Thank you. We will get you copies of this, although I am sure all of you at one point or another have seen a nutrition fact label but we are going to get the inside dope on it. Virginia?

MS. WILKENENG: Let me give a little backup first. Within your packets is a copy of the Federal Register Notice, dated March 1. It is in tab 7, a little bit behind there. It is the announcement when we reopened the comment period on our ANPRM for trans fat labeling to try and deal with the idea of percent daily value for trans or perhaps the use of a footnote that would help consumers put the quantitative number into the context of how much they should eat in a day's time.

In that Federal Register Notice we made mention of a recent NAS report that had been funded by FDA, USDA and Health Canada to give us some guiding principles on how all of the new National Academy dietary reference intake values should be used to update the daily values on the nutritional label.

In regard to saturated fat and trans, that report had recommended that we perhaps reconsider the daily values and then put them together to have one joint daily value for saturated fat and trans. In our Federal Register Notice we were going out and asking for comments specifically on that recommendation, as well as in general any ideas that would help the agency address trans fat.

Back when we first did the new nutrition facts label, we had some graphic artist that worked pro bono to help the agency design the nutrition facts panel. We think that was a very successful activity and we have even gotten presidential awards, all sorts of things, for the nutrition facts box and now you see it for all sorts of related things, that idea of the graphic.

So, the graphic designer who did that, when he read the report that there be a joint daily value, felt some need to again get involved in the process and he put together this graphic that we have handed out to you just as a way to show how you could have a joint daily value with separate quantitative listings for saturated fat and trans. That is what this shows just to give you the idea of what it might look like to bring a halt to the line that divides them under the column for percent daily value.

Again, I would just point out to the committee the importance that we place on the percent daily value information is a way to help consumers get a feeling for what those numbers, for whatever nutrient, represent in terms of dietary recommendations. In the case of the fats, fatty acids, cholesterol, sodium, generally those are maximum recommendations. So, our advice to consumers is don't go over 100 percent of your daily value for those nutrients, whereas for the vitamins and minerals, the others that are helpful in a diet, we want them to reach at least 100 percent of the daily value.

So, this would give them an impression in this particular label--you know, we have jumbled with the numbers some because the graphic artist didn't have much concept of what the numbers were for trans fat, but if you had 1 gram of saturated fat and just 0.5 grams of trans fat in a serving of the product it would amount to 8 percent of the daily value so that they could know how that fits it in the context of the whole diet.

If there are any questions on this particular rendition or anything else about the daily value, I would be glad to help. Yes?

DR. RIMM: I have a few questions. The first one is the daily value is based on just summing up the total amount of saturated fat and trans fat in grams and then basing that on the total grams at the bottom, where you say 20 grams for 2,000 and 25 grams--

MS. WILKENENG: Actually, in this case we did that, yes. We summed them up and we used the 20 grams but were we to do this, we would need to go back and reassess the daily value for saturated fat as well as create one for trans fat. Now, the IOM report has recommended we do that through menu modeling and using food composition data tables. The biggest problem with that, as alluded to a little bit earlier, is that there aren't really good food composition data tables at this point in time.

DR. RIMM: It is a moving target so any time a guy goes back and reconstitutes a potato chip we have to go back and measure it. So, it is a moving target. There are a number of foods that have been measured and I don't think we should completely throw our hands up. I think for trans the values are probably a lot better than zinc, for instance, but that is a separate issue.

So, the proposal would be to have total fat, sat fat and trans fat in the bottom little section under the 2,000 and 2,500. You would actually have a separate line for trans fat saying you should have less than X number of grams.

MS. WILKENENG: I would think that would be a logical outcome of it, yes.

DR. RIMM: So, you actually could calculate a percent daily value for trans fat. You wouldn't necessarily have to sum them.

MS. WILKENENG: Well, this bottom table gives consumers what the daily value is for those macronutrients.

DR. RIMM: Correct.

MS. WILKENENG: So, if we had a daily value for trans fat it could be added into that list.

DR. LICHTENSTEIN: But from the way it is depicted here, there is a percent daily value for trans plus fats and it would appear that the value on the bottom would have to be trans plus fats because otherwise you would end up with individual daily values.

DR. RIMM: That was my point that I was trying to communicate. Thank you, Alice.

DR. LICHTENSTEIN: Anytime, Eric!

MS. WILKENENG: That is certainly an option, yes. These are the types of things we hope people will comment on as a result of that ANPRM. I would let the committee know, if they are not aware of it, that because of the discussion of this committee we extended the comment period on that ANPRM until I believe June 15th.

DR. RIMM: I also apologize, you can voluntarily put in polyunsaturated fat and monounsaturated fat, and is there a daily value that gets assigned to that also?

MS. WILKENENG: No, there is not.

DR. RIMM: Those are just separate as grams?

MS. WILKENENG: The value lines are blank for mono and poly.

DR. RIMM: I have only looked at a food label a thousand times but I just forget. Thank you.

DR. KRINSKY: To follow-up on Eric's point, it seems to me that there is a conflict here between the total fat with saturated fat and trans fat underneath it in terms of the percent daily value and not having that trans fat on the bottom of the table because--

MS. WILKENENG: Oh, I think the bottom of the table would definitely need that.

DR. KRINSKY: Would that have that? Okay.

MS. WILKENENG: It is just that the artist didn't think about it; we didn't give him any guidance. This just came in as a comment.

DR. KRINSKY: So, in our discussion here and with our question number two, if in fact we arrived at the conclusion that there was evidence that 2 grams of trans fatty acid would be the appropriate level, as this nutrition fact label is set up, if it had trans fat of 0.5 gram that would then be 25 percent of the daily value.

MS. WILKENENG: If it was listed separately.

DR. KRINSKY: If it were listed separately. Would that frighten the average consumer?

MS. WILKENENG: That is an area that needs study.

DR. LICHTENSTEIN: The question really is would it be useful. The way this mock label is set up, it looks to me like the logical thing would be to have a DV that was--and I think this is what the assumption is, that the DV is 10 percent, that it is combined trans and sat. So, first of all, if it is under 0.5 grams it could be listed as zero. As you indicated if it was 2 grams it would be 25 percent and I don't think that would necessarily be useful information to the consumer relative to the 8 percent of saturated fat. So, again, one goes back to potentially combining them and then using a label something like this and communicating that you should pick the product that has the lowest number.

DR. KRINSKY: But we can't combine it. That is what the Federal Register says?

MS. WILKENENG: No, in terms of not combining them the discussions in the past Federal Registers is in regard to the quantitative amount, and that is why you see trans fat on a separate line under saturated fat. Our first proposals had put them all on one line using the term saturated fat.

DR. MAYNE: I think the most compelling reason not to have the specific DV for trans fat is the measurement issues. If we can't measure any more precisely than, you know, less than 0.5 grams and we set it at 2, you are saying that we can't differentiate zero and 25 percent of a DV and that is extremely misleading to a consumer. So, I just don't think it is practical if you are talking about 2 grams as the limit there.

DR. RIMM: I think we can differentiate it, it is just that the labeling laws allow anything less than 0.5 grams not to be put on the label. We definitely can measure 0.2 and 0.1.

DR. MAYNE: Right, so then it would make no sense because manufacturers would make them all at 0.4 grams and then they would be exceeding their 2 grams and not realize they are consuming any at all. So, I don't think that would make any sense.

DR. RIMM: Unless the daily value can be based on the true amount of trans fat. Even though the grams may say zero, the daily value could state 10 percent if it really was 0.3 grams but that is too much detail.

DR. MAYNE: It confuses them even further.

DR. LICHTENSTEIN: Again, it goes back to this is a reasonable proposal for presenting the information, at least at this point, and obviously none of this is chiselled in stone because nutrient facts panels do change if there are data to suggest that they should in the future.

MS. WILKENENG: I think this format also clearly addresses your third question, whether trans fat and saturated fat have similar or different effects. If you are going to recommend that they have similar effects, this is reasonable. If they have different effects then perhaps it is not.

DR. KRINSKY: Well, in the simplest world if they had identical effects, then they could be combined as sat and trans and we wouldn't have anything to worry about. So, we might at this point move into question three. I think you have already raised it for us, Virginia, and thank you very much, unless the committee has questions on the nutrition facts. Yes, Suzie?

MS. PELICAN: It is not really a question, just a comment. I think Eric brought it up and Virginia kind of confirmed that we don't have anything in the DV column for monounsaturated and polys in the case of those foods that decide to declare them. So, there is a precedent for having no DV.

DR. JOHNSON: I just wanted to touch on the analytical thing again. In a perfect world no doubt you can be precise in determining 0.4, 0.5, but in the reality of a food-based system there are manufacturing variabilities, there are variabilities in raw materials and all kinds of things like that, and the current label does have a window for compliance, which many of us feel is a challenge in and of itself, of 20 percent. So, I think when you get down to that level of half a gram or less it is tough to get much more precise than that.

DR. KRINSKY: I really don't see why, Guy, because you are not dealing with an individual serving; you are dealing with a huge pot of product that is being produced. Certainly, industry can determine what the percent trans is in that pot and then break it down to the individual serving. No? Go ahead.

DR. JOHNSON: I mean, there are just variabilities within any kind of system and processing, distribution, raw materials, analytical variability--there are a lot of factors that enter into that.

DR. KRINSKY: And you are allowed 20 percent over and under in the process?

DR. JOHNSON: That is true.

DR. KRINSKY: The whole thing would be simplified if trans was the same as sat. So, let us then move to the third question. The third question is when compared to saturated fatty acids, are trans fatty acids considered to be more, less or similarly adverse with respect to coronary heart disease?

DR. LICHTENSTEIN: I will start the discussion. Probably they are not identical. We have just talked about some of the differences that are well established and some of the potential differences that there are. I think from a practical perspective what we also need to take into consideration is within the context of the quantities that they are likely to be consumed in, and how different are then saturated fatty acids.

One could say that for practical purposes, because the intake is around 2 percent of energy so that if you break it down to a variety of foods, each food doesn't have that much but, you know, cumulatively it probably, from a biological perspective, would be difficult to distinguish between 2 percent trans and 6 percent sats or 3 percent trans and 5 percent sats as far as whatever biological measures we chose to use as the criteria.

DR. KRINSKY: So, does that mean you couldn't differentiate 3 percent trans from 3 percent sat?

DR. LICHTENSTEIN: Three percent trans and 3 percent sat you might be able to, but that is not what is likely to happen. We know that is not what is in the diet. What we know is that there is two to three times or four times more sat relative to trans so you are shifting in trans. Even if you cut your trans in half you are going from 2 percent to 1 percent, whereas your saturated fat, if you went down an equivalent amount in weight the percent shift of the total is going to be smaller. So, again, it comes down to what you can measure with respect to your biological outcomes and, from a practical perspective, I am not sure that we have the data to really assign a relative factor to each so it may, to a certain extent, be a moot point if we go back to what we want to encourage people to do, minimize intake of both.

DR. SHANNON: Could I ask a question here? If you did, like you say, cut trans in half and replaced that with a totally hydrogenated product what does that do? Do you have any sense of what that does in terms of raising the percent of saturated in the diet?

DR. RIMM: It would all be saturated so if you take 2 percent and go down to 1 percent that 1 percent would no longer be trans, it would be all saturated. So, that would not increase your sats dramatically because your sats are at 10 percent and going to 11 percent so it would only be a small increase in overall sats. I think that is what you are asking.

DR. SHANNON: That is what I am asking from a biological point of view.

DR. LICHTENSTEIN: From a biological point of view, well, that is a unique example because in that case you are putting in a saturated fatty acid that is relatively neutral. If you are going to use your outcome measure to be some LDL cholesterol levels, it would probably have a very small biological outcome if those are the quantities you were talking about. You certainly could design a metabolic study where you had very high levels of trans; you cut it in half; you substitute stearic and you probably could see some beneficial effects biologically. But within the context of what is likely to happen with food, I don't think there is going to be that big an effect.

DR. SHANNON: Then that is what I think is the most important thing for us to look at. When we are talking about similarity, what is likely to happen if we dropped the trans to the extreme of replacing it? Is it going to have a biological effect? If not, then perhaps we can consider them similar.

DR. RIMM: I think we are not the first committee to ponder this. Obviously, when they spent four years trying to come up with a quantitative assessment for trans fat they spent a lot of time thinking about this and decided that biologically trans fat is very different from saturated fat so let's make a separate line for trans fat. I think that is one of the reasons that the trans fat line is now going to be in the labels.

But I think in the paper we were given, Mensick does a meta-analysis of all the metabolic studies and it is very clear that the LDL or LDL/HDL raising effects of trans is about two- to three-fold that of saturated fat. If you look at the observational epidemiology, the risks associated with the 1 percent increment in trans is about three or four times that you would get for a 1 percent increase in saturated fat. If you look at the diabetes research, it seems that there is an impact for trans and not for sat. If you look at insulin sensitivity research, it suggests that there is an adverse effect for trans and maybe not necessarily for saturated fat. If you look at the inflammatory markers--I mean, there are a lot of biological studies out there, some observational and some experimental, which would suggest that trans is very different from sat. You know, I don't know how important the small range that you are talking about is in our diet because there are people that have 1 percent of trans and there are people who have 6 percent of trans, and within that range I think there is a very big difference in the impact of trans--than on saturated fat.

DR. KRINSKY: So, in a normal American diet 2 to 3 percent trans; 2 percent; 2.6?

DR. LICHTENSTEIN: But that is probably high, probably closer to 2.

DR. KRINSKY: We will round it off, 2 percent trans. If you drop that 2 percent trans down to 1 percent and replace it with any saturated fatty acid will there be any biological effect? Any measurable change in CVD risk?

DR. RIMM: Well, you are talking about the average if you took the top 25 percent.

DR. KRINSKY: No, no, no. I am talking about--

DR. RIMM: A single 1 percent change?

DR. KRINSKY: Yes.

DR. RIMM: Again, if you are looking at the observational studies it would suggest that there is a slight difference. You know, we do it estimating on the whole range of the nurses health study, looking at a 1 percent change in trans versus--exchanging it for 1 percent unsaturated fat and there would be a difference, small but there would be a difference.

DR. LICHTENSTEIN: If you looked at the intervention data, no, there wouldn't be. Even if you look at things like glucose levels or insulin levels, even with LP(a) where some studies that show a change; some studies don't. But even those studies where we saw a change it is not in the range that would be physiologically relevant. It just turns out that in the intervention studies you see less of an effect than you do in the observational studies. So, in the interventional studies that magnitude of change is unlikely to have a measurable biological effect, at least in the short term. I think all bets are off in the long term.

DR. RIMM: I think that is part of it. You are talking about short term where we are really only measuring one outcome. I think there are a lot of other things and if you put them altogether and measured insulin sensitivity, measured lipids--if you put them altogether you would get a bigger difference than if you do if you have a four-week trial with a single outcome.

DR. LICHTENSTEIN: Well, no, even with multiple outcomes in the same study you still get very little effect.

DR. RIMM: I don't know if we know all the outcomes yet that are related to trans.

DR. LICHTENSTEIN: We may not know all the outcomes related to sats.

DR. KRINSKY: So, under some circumstances they are equivalent; under other circumstances they are not equivalent.

DR. LICHTENSTEIN: Primarily under other circumstances where the intake levels are relatively low they are representative of the mean intake in the United States. If you go to higher levels, then I don't think one would be confident saying that. I think it is only in the context of the average current dietary intakes.

DR. KRINSKY: And does that mean at the higher levels you are seeing more pronounced detrimental effects than you are at the lower levels? Is it a gradual change in risk as you go from 1 to 6 percent trans?

DR. LICHTENSTEIN: Well, those data for the intervention studies are really just available for lipoprotein levels and a little bit for glucose insulin.

DR. KRINSKY: When I say coronary vascular disease I mean the association with it.

DR. LICHTENSTEIN: The epi data?

DR. KRINSKY: Well, what you have, Alice, is change in LDL cholesterol. LDL cholesterol is associated with coronary vascular disease so even in the intervention trials aren't you, in fact, dealing with an association? You have data--

DR. LICHTENSTEIN: A relationship, yes. Well, we have relationship with glucose, with insulin, with Lp(a), with HDL primarily--

DR. KRINSKY: Right.

DR. LICHTENSTEIN: --with actually postprandial particles with triglyceride levels, and in that case it is hard to distinguish in the lower levels. You can certainly distinguish 6 percent from the 0.55 percent, which is about the lowest that we ended up with in our experimental studies. But within that range of 0.55 and about 3, we weren't picking up statistically significant differences. In some cases there were small trends in LDL, not in HDL but in LDL.

DR. KRINSKY: That is table 8-9? Is that the one you are quoting, page 8-59? No, no, no, it isn't. I am sorry, it is page 60, the next page. It is table 8-10.

DR. LICHTENSTEIN: But those data are just for LDL, HDL and Lp(a).

DR. KRINSKY: Right. I guess what I am trying to do is to relate these two worlds that I am hearing. I am hearing one world where the intervention trials are, in fact, getting hard data as far as LDL cholesterol is concerned as modified by the level of trans fatty acid, although that is not the only thing that is varying in those studies, it appears.

DR. LICHTENSTEIN: It was for polys. That is because of the products that are currently available as far as spreads in the United States. It seems that that is what the effect is. If you hydrogenate, you decrease the polys and increase the trans.

DR. KRINSKY: And in your studies, Eric, as you see a variation in the trans levels what is replacing it?

DR. RIMM: Well, the way we modeled it is that you are reducing trans for carbohydrates so you are holding calories constant and you are moving trans down and carbohydrate up. Or, you can model it the other way. Statistically you can trade trans for sat also.

DR. KRINSKY: So, are we dealing with apples and pears in terms of trying to compare these?

DR. RIMM: No, I think that Alice has a dose response. If you put all the metabolic studies together there is a dose response between trans and LDL. If you look at our studies, as Susan was saying, there is a significant dose response between trans and coronary heart disease. The point is that if you are at 1 percent and 2 percent there is a difference but is it tangible? Well, I would argue that, of course, every little bit helps, and I think there is a pretty linear effect but no one has shown me that there is an absolute threshold effect based on the metabolic studies and I think if you put all the epidemiology together you are looking at ranges that are different in different countries and it would suggest there is a pretty linear association.

DR. LICHTENSTEIN: That would actually argue to combine the two because if less than a half gram is considered zero trans you would actually lose those small effects, especially if it was cumulative over multiple foods throughout the day. Whereas, if it had to be added to the sats, then at least it would all be inclusive and you would hold together all those little potential dietary effects that might occur.

DR. MAYNE: We have this Mensick review article from AJC and they specifically address this issue of isoenergetic replacement of trans AT-1 with saturated fats. The quote is that isoenergetic replacement of trans AT-1 constituting 1 percent of energy with saturated fats decreases total HDL cholesterol by 0.019, and those are the standard units. So, it is clear that there is an effect; that they are not comparable. Then, when they go on and look at the prediction on coronary heart disease risk, they find that they underestimate. With the LDL changes they underestimate the coronary-artery disease risk, saying that there may be other effects, like Eric was alluding to, that maybe trans fats may affect more than just the LDL cholesterol because the standard prediction equations underestimate the chronic disease risk.

DR. KRINSKY: But that puts me in a bind because how do you evaluate the other effects of the trans?

DR. RIMM: We do it the same way we do the lipids. We are just a year or two behind. I mean, I think there is a lot more data that will come out in insulin sensitivity or inflammation. There are already a number of association studies looking at trans and inflammation. So, I think it will just be a few more years until Alice has time to do some more trials to look at all these outcomes, unless she has some saved blood that we can look at.

DR. JOHNSON: It seems to me that the meta-analysis, and so forth, are based on the studies that are there, most of which have been done at higher levels than are reflective of the current diet in the U.S. So, then you crunch the numbers and you come up with those things so I still think there really isn't hard evidence to show that at the levels that we are talking about in the diet there is a meaningful difference between trans and saturated fats. I think we also need to begin to feather in how this is going to be interpreted by the consumers. If we make a big difference between trans fats and saturated fats, and so forth, unless we really have to, I think we run the risk of really confusing people and pointing them in the wrong direction, emphasizing either trans or saturated fats unduly. So, we need to be cognizant of that I think.

DR. KRINSKY: Let me just try to put this together. If we are dealing with low levels of trans fatty acids, then does that seem to be much of a change if we try to change the trans for the saturated? So, you go from 2 percent to 1 percent trans, it may be difficult to demonstrate that there is an increase, decrease or it looks as though it is equivalent. But if you are starting at 6 percent trans, then you can begin seeing a difference if you replace that with saturated fatty acids.

So, in fact, maybe the thing that we should be doing here should be working even more energetically to try to get the trans level down to a level where there may not be much of a difference between trans and saturated fatty acids. That is, if we have people that are consuming 6 percent trans fatty acids, then that may be a greater risk than a comparable level of saturated fatty acids--they may be at greater risk than if they had a comparable level of saturated fatty acids. Whereas, if they are consuming 2 percent and they go to 1 percent it is going to be a little bit more difficult to evaluate the effect. Am I overstating the case? If I am, please correct me.

DR. RIMM: Well, I think there is no question in getting someone to go from 6 percent to 1 percent. Even if they were switching to saturated fat I think there would be benefit because I think the trans is steeper than the saturated fat in terms of cholesterol and heart disease risk. So, sure, I think going from 2 to 1 is not as important as going from 6 to 1. But I think going from 2 to 1 is as important as going from 6 to 5. You know, the bigger the change, the better.

DR. MAYNE: If you go back to the figure from Ascherio, the two regression lines there, it looks like the magnitude of the change is much greater if you go from 6 to 5 than if you go from 2 to 1 because they are not parallel lines. So, if you look at the total net change in LDL/HDL cholesterol, as Norman suggested, it would predict a much greater impact at higher levels of trans fatty intake if you agree with that line.

DR. LICHTENSTEIN: I think the difference between those two lines is there was one study that was omitted for one of the lines because it was so out of range. So, I think you should look at the line that did not include that study because that was very unusual data.

DR. RIMM: I think the line is estimated based on saturated fat change. The dotted line is saturated fat and the straight line is trans fat.

DR. LICHTENSTEIN: Okay, I am sorry.

DR. RIMM: There clearly are differences based on the estimate. The slope clearly is different. So, you can argue if you are way down at the 1 and 2 percent, you know, the slope is the same, there is going to be a difference but, obviously, the higher you get up the absolute difference will be greater.

DR. KRINSKY: To get back to our question, when compared to saturated fatty acids at certain levels, because this is really the answer to this question is going to depend on the level of the trans fatty acid in the diet. So, can we modify our responses to the question by adding something to it? Are we allowed? Do we have dispensation? We are going to do it anyway. So, we can play around somewhat with this. I think the Ascherio figure really is a very important figure. What are the squares and the circles in that figure, please?

DR. LICHTENSTEIN: I think they refer to the individual studies. The original figure was actually done in color. Oh, no, I am sorry, that is the controls versus the intervention. They were actually all color coded so you could match the study with the dots.

DR. JOHNSON: Could I make just one observation about that figure? It is important to remember that what they are plotting here is the energy from saturated or trans fat versus the ratio of LDL to HDL cholesterol. If you look at the individual components, which my friend Ed Hunter at Cincinnati did, he showed that there is a clear relationship between trans fat intake and LDL over 4 percent of the diet, not under, but there are few points. When you look at HDL individually there is no effect in any of these studies until you get up to around 6 percent of the diet which is where the effect kicks in.

So, I think it is important to recognize that the ratio here can be a bit misleading if you are simply looking at that and extrapolating it through the whole range and forcing it through the origin.

DR. KRINSKY: If I follow Ed Hunter's argument I would ignore the percentage of energy from fat above 6 percent because if the HDL is not changing under those circumstances, if trans is not changing the HDL level, that ratio is the same as the effect on LDL if your denominator is constant. It looks to me, and I would need somebody who has had more experience in manipulating data, that is the epidemiologists, what those two lines looked like if you ignored everything above 6 percent energy. Again, I don't have the colored points so that makes it a little bit difficult to read this. For example, at 3 percent are those two points for trans fatty acids? At 3.5 percent are those two points--so, that looks like there is, in fact, a real effect of trans fatty acid on LDL cholesterol from 6 percent down.

DR. LICHTENSTEIN: The question is between 1 and 2 percent and, remember, that is on the ratio. For LDL the effect of trans is less than for saturated fatty acids, it is just that there is an HDL effect also.

DR. KRINSKY: But if we are told that there is no effect on HDL I am ignoring the ratio.

DR. LICHTENSTEIN: But that is data are just plotting the ratio, this figure.

DR. KRINSKY: But the denominator is constant.

DR. LICHTENSTEIN: It is not. I don't think you can assume that. I think you have to look at the individual studies.

DR. KRINSKY: I thought this is what Guy just said.

DR. LICHTENSTEIN: Maybe you should reproduce that, or something.

DR. JOHNSON: I would be happy to share this with the group.

DR. KRINSKY: Whether the HDL is constant--

DR. JOHNSON: It is extracted from all of the studies that are published.

DR. RIMM: I mean, if you look at the circles, regardless whether this is the ratio or LDL, the circles are the trans and the squares are the placebo group, the control group for that particular diet. The circles make a pretty nice, straight line with the exception of the one really high one, which is why I think people felt comfortable extrapolating down to zero, plus there is Alice's really low point, 0.5 percent. But if you look at just the circles it would suggest that there clearly are differences all the way down to the origin.

DR. LICHTENSTEIN: But some of them must be for saturated. Some of those squares must be saturated fats because how would there be a line here for saturated fat?

DR. RIMM: That is estimated based on all the other trials that have ever been done; that is not based on these eight trials. There have been a lot of trials done on saturated fat.

DR. LICHTENSTEIN: But it says it represents the best fit regression for saturated fatty acids.

DR. RIMM: This is just a meta-analysis of the trans studies, and that is the dark line. The dotted line is what you would expect if you took Mensick and Katan's review of all the 30 or 60 studies that have been done on saturated fat. So, you can see there are clear differences in the trend between trans and what would have been expected if you plotted all the saturated fat trials.

DR. LICHTENSTEIN: Which Mensick and Katan review meta-analysis?

DR. RIMM: Maybe it was the '92 one, I believe, the first one.

DR. KRINSKY: We are having a data searching pause.

DR. JOHNSON: Dr. Lichtenstein has slides from her study that show that as you increase trans there is a clear impact on LDL but there isn't on HDL, and that is what my friend Dr. Hunter showed by looking at Dr. Lichtenstein's study as well as the other studies that were published in this document.

DR. LICHTENSTEIN: It is on 8-60. So, it is one study. I mean, you know, I think we have to take that into consideration.

DR. KRINSKY: And that study is included in this figure? We are trying to get the methods out of Ascherio's paper with respect to how the lines were established. Any luck, Eric? The pressure is on.

DR. RIMM: As usual, Alice is right. The figure summarizes the randomized trials that directly compared the effects of trans fatty acids with isocaloric amounts of cis fatty acids. When the data are available, the figure also shows the effects of saturated fatty acids in the same studies. So, of those same studies that have looked at trans, some of them also looked at saturated fat and that is the four or five squares that are there. So, that is in the same studies.

DR. LICHTENSTEIN: Right, but the data are incomplete because not all the studies looked at that.

DR. RIMM: Correct. However, you can see that in most of them the saturated fat effect is much lower than the trans fat.

DR. LICHTENSTEIN: Right, on the ratio but not on LDL and on LDL the trans tend to have less of an effect than the saturated fact, but it is compensated for because of the effects on HDL. So, at low levels, again, it seems sort of a wash.

DR. KRINSKY: To get back to our question, you just said at low levels it seems like a wash.

DR. LICHTENSTEIN: At 1 and 2 percent of energy. When you go higher the interpretation would be different.

DR. KRINSKY: So, we could consider an approach to answering the question. That is, at low levels of trans it tends to be approximately equivalent to that of the saturated fatty acids. We may have to define what low levels mean but if you just accept that, for the moment at low levels there does not seem to be much difference between saturated and trans. But as we go to high levels we begin seeing a very distinct and presumably significant--I don't want to use that term, we begin seeing a very marked difference between the trans fatty acids and the saturated fatty acids. In fact, the trans fatty acids are more adverse with respect to coronary heart disease, which very simply means that there should be recommendations that we have very low trans fatty acids in our diet.

DR. LICHTENSTEIN: We do. That is the IOM report.

DR. KRINSKY: I understand that, but now I am thinking right back to question two because question two is asking can we support the suggestion that a diet should contain less than 1 percent of its energy from trans fatty acids. If you look at figure 8.4, it seems that that would be a good thing to have.

DR. RIMM: And that is just looking at the lipid effect. There are all sorts of other effects.

DR. KRINSKY: But we are only asked with respect to--yes, one and three are asking with respect to coronary heart disease and I know that there are other markers that are coming up and trans is having more and more effects and in the future it may be very profound, but the future lies before us and we are here today and tomorrow. So, we are going to have to evaluate what is available today and tomorrow, what is available to us, not what is available in unpublished or unstudied material.

DR. JOHNSON: I was just going to reiterate the fact that all but one of these studies used levels of trans fatty acids over the mean intake. I mean, the fact that we would like to see trans fatty acid intake lowered doesn't mean that we have data to suggest that it would be a meaningful difference between saturated fat and trans fatty acid levels at usual dietary intakes. We just don't.

DR. KRINSKY: I understand that and, you know, the usual dietary intake is between 2 and 2.6 grams of trans fatty acid per day.

DR. LICHTENSTEIN: Percent of energy.

DR. KRINSKY: Of energy, sorry. Thank you. But, in fact, we have a significant portion of the population that is consuming both more and less than that, and that portion of the population that is consuming more than 2 to 2.6 percent of energy--if we look at this figure, is it risk? Is it increased risk? Now, do we then only think of the average intake of the population or do we think much more broadly, that, in fact, that average represents a very broad spectrum of intakes?

DR. LICHTENSTEIN: Eric, do you remember what is the range in the nurses and health professionals? Because what we are looking at here are metabolic studies where it has been manipulated to be extremely high. It is hard to get 6 percent of energy. Do you remember what the range is?

DR. RIMM: I think for the nurses the tenth percentile is 1-3.5 percent, or something like that. In the health professionals the lowest quintile is 1.5 grams trans fatty acids per day. The highest quintile was 4.3.

DR. LICHTENSTEIN: Which is 9 percent energy so we are switching units here.

DR. RIMM: Right. In the health professional the lowest quintile is 1.3 grams per day; the highest quintile is 2.9.

DR. LICHTENSTEIN: So, that is 2 percent of energy.

DR. KRINSKY: We may have some figures in one of our tables on page 8-67.

DR. RIMM: Yes, so the nurses goes from 1.3 percent of energy up to 3.2 percent of energy in the median of that first and fifth quintile. The health professionals is in grams so the two of them are difficult to interpret. So, it is 1.7 grams up to 6.5 grams, which would be about 0.5 percent up to 3.5 percent.

DR. LICHTENSTEIN: I think for them it is probably a little bit lower because the males are consuming more than 2,000 calories.

DR. RIMM: Right.

DR. LICHTENSTEIN: So, it is probably closer to--I don't know, 2, 2.5 percent. So, that is what we are talking about for the current range on the basis of these data.

DR. KRINSKY: But for the men there is like a four-fold range from the lowest--

DR. LICHTENSTEIN: That is grams. That is grams per day.

DR. KRINSKY: It is four-fold--

DR. LICHTENSTEIN: Okay, it is still four-fold.

DR. KRINSKY: Four-fold as opposed to the women that have about a 2.5-fold. So, that would indicate that you are going to have a higher percentage of men. If, in fact, the highest quintile of men is equivalent to 3.5 percent energy intake--is that what somebody just said?

DR. RIMM: Yes, the nurses are 3.2 percent in the highest quintile.

DR. LICHTENSTEIN: We don't know what the men are because we don't know what the caloric intake is.

DR. RIMM: The average caloric intake is about 2,100 kilocalorie for the men. For these men we estimated it was 2,100 kilocalorie. So, if they are an underestimate, then everything is an underestimate so they would be comparable. Dr. Krinsky is now doing a back of the envelope calculation.

DR. KRINSKY: It looks like it is about 3.2071. So, it is similar to the 3.5--it is between 3-3.5 percent energy intake.

DR. RIMM: These are adults that are 40-75 male health professionals so we have to keep that in mind. So, again, I think the general population definitely at the high end will be much higher than these men and women, and I think that is who we want to be directing our message to. I think that the food label really needs to get out to people who are eating 5 and 6 percent of energy from trans.

DR. KRINSKY: Kids, adolescents, college students. So, what is the message, Eric, that we want to get out to them?

DR. RIMM: I think we need to make the point that trans is different than saturated fat and we have already put that on the quantitative label and I think we have to acknowledge the fact that it is important for people to understand that trans, maybe at very low levels, is not different than saturated fat but I think the biological effects of trans are different from saturated fat and the label should denote that.

DR. JOHNSON: It sounds like we are in agreement that at high levels trans are different than saturated fat. I think the harder part of that question is should the label reflect that if most of the population is not going to benefit from that information.

One of the things I am an advocate for is putting the E back in NLEA, and we need to remember that the trans fat message needs to be communicated by more than just the nutrition label. Particularly people that are consuming high amounts of trans I think should be given that information. I mean, that should be the target for educational efforts. I think we do a disservice if the bulk of the population need not be--if that distinction is not significant.

DR. KRINSKY: Well, I am not sure what position we are in to put out a message as you are suggesting, Guy. I think it is a very important message and it would be very useful, and I don't know whose responsibility that is. I don't know if it is the FDA's responsibility or who, but I think your point is very well taken that that message should come out. But how do we then try to reconcile that with trying to answer a question of this nature where the question is not addressing the educational value of this information? In fact, it is addressing the scientific aspects of this question and we are having a terrible time trying to answer it scientifically--a difficult time trying to answer it scientifically because not so much that it varies but the difference may become insignificant at low levels of trans, whereas it may be very significant in terms of coronary heart disease risk at higher levels of trans.

I am saying low and high and I am obviously not defining them. If any of you would like to define it, that would be fine. I can't define this. If we were to come out and try and answer this question by talking about high and low levels, I don't think that would be terribly useful for the FDA. I think it would be more useful if we could come out with some level and if I begin thinking about coming out with some level, then I begin thinking that this is a DV for trans fatty acids and maybe that is what direction we should go in. We can't go in that direction but we can make recommendations in that direction.

DR. JOHNSON: Let me just say that I made the point about education because I don't think that should be lost. I think it is important and I think it is a shared responsibility not only with FDA and other government programs and the Dietary Guidelines but the industry plays a very active role in educating people about trans fatty acids. So, I think that is something that shouldn't be lost.

With respect to where we are at, it seems like we all agree that at high levels, whatever that is, trans fatty acids are different than saturated fat. At low levels we don't have the data. So, the big question is where do you draw the line. The problem is that we really don't have good data. If you go back and look at the individual studies, HDL isn't impacted until you get something like 6 percent of calories and LDL is roughly the same between the two. So, that is one potential answer. But without good, solid kind of data at levels below the mean intake, we are kind of guessing.

DR. KRINSKY: Could you hold up Alice's figure on LDL cholesterol as a function of trans fatty acid?

DR. LICHTENSTEIN: Again, remember that was one study. That is not a compilation of studies.

DR. KRINSKY: I understand that.

DR. LICHTENSTEIN: A nice study, I must say.

DR. KRINSKY: We will make the judgment call, if you don't mind! But as I look at that figure, Guy, there is a trend, isn't there? Even though there is not a significant difference, there is a trend. If you were to drop butter and do a trend analysis would it be significant?

DR. LICHTENSTEIN: We just did an analysis of variance with all of the diets. I don't remember dropping the butter because that was just part of the experimental design but it certainly could be done. You can see a clear effect of HDL, with the butter being really high, and then at the highest level of trans, which was 20 percent of calories of stick margarine which was pushing the system.

But getting back to the original questions, I don't think any of us are arguing that we should encourage the consumption of trans fatty acids and I don't think any of us arguing that if the intake was down to 1 percent of energy that that wouldn't be advantageous to the general population.

So, in some senses it seems to me that the three questions are in one realm and then there is another realm which is how to deal with the label, and it seems sometimes that we are crossing over in that discussion or maybe you are bringing us back to it. Because when we started looking at the mock label, which is somewhere in my papers--

DR. KRINSKY: Actually, I think you brought up the DV for the label and I just thought that would be useful.

DR. RIMM: Let me go back to a point I made before about what the threshold for what we call scientific evidence is for all the other things in this label because if you look at dietary fiber, there is a DV for dietary fiber and I am sure if you exchange 1 gram of dietary fiber for 1 gram of sugar you also would not see any biological difference. Yet, you know, I think there still is an accepted dose response between dietary fiber and LDL cholesterol if it is soluble fiber. So, I think maybe we are raising the bar too high for saturated fat compared to everything else in this label and that for most of these factors there is a dose response; there are metabolic studies where you can draw the line; but there are not 30,000 metabolic studies at all levels of sodium and all levels of fiber. So, I think we should just uphold the same--of course, I don't know this for a fact but we should uphold the same evidence that we use for all these other factors that we do for trans fat and saturated fat.

DR. LICHTENSTEIN: I am not sure the criteria for fiber was LDL cholesterol. I don't know if we could get Paula to comment, if she is allowed to. But it still doesn't diminish your point that we still have the saturated fat and the cholesterol where the data are not what we would expect them to be.

MS. PELICAN: Just to follow-up on what Alice had said and also Guy, I think I really appreciate what Guy has said more than once about education, and I think Alice said, you know, somehow we are getting into kind of a grey area, kind of a difficult point, and I think this is sort of where the science meets education and I am one who doesn't think that the nutrition label itself is an education tool. It informs people but it needs to be used in the hands of educators, and I think our best role as this group is to give guidance on the science and then we have educators--and I applaud FDA for this wonderful educational tool that in the hands of educators, with any number of programs, this can then be taken but we need to have this based on good science so that educators can interpret it correctly and help consumers interpret the foods they buy correctly.

DR. RIMM: I think it really is meant to be an informative tool. So, the fact that we have grams of trans fat on there and grams of saturated fat is really quite a step forward and I applaud the FDA for being able to push that forward. My concern is that we need to inform people and educate them, and my concern is that forever people will think that a gram of trans fat is the same as a gram of saturated fat and I think the growing body of evidence would suggest that is not the case, that a gram of trans is probably worse than a gram of saturated fat. But we will have to leave it up to the educators to teach us that.

DR. KRINSKY: Unless we had ULs for them, which we don't and we won't. I have a slight sense of wheel spinning, at least in my head. I have just been trying to find out whether it might not be worthwhile to see if we could move forward our public comment. I don't know if everybody is here for the public comment. We are trying to see if the speakers are ready so why don't we take a ten-minute break and reconvene at 4:35?

[Brief recess]

Public Comment

DR. KRINSKY: Thank you, all. We are going to move into our public comment session and we have three speakers. I even have a list of them and they each have ten minutes and we are free to ask them any question that we want to. But before they begin, I have a statement that the FDA requires me to read. It deals with the disclosure of financial relationships for public commentors.

Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision-making. To ensure such transparency at the open public hearing session of the advisory committee meeting, FDA believes that it is important to understand the context of an individual's presentation. For this reason, the FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with the sponsor, its product and, if known, its direct competitors. For example, this financial information may include the sponsor's payment of your travel, lodging or other expenses in connection with your attendance at the meeting. Likewise, FDA encourages you, at the beginning of your statement, to advise the committee if you do not have any such financial relationships. If you choose not to address this issue of financial relationships at the beginning of your statement it will not preclude you from speaking.

So, I would like to call on Robert Earl, and if you could go to the podium, please, Mr. Earl. Robert Earl is the Senior Director of Nutrition Policy at the National Food Processors Association and he will cover the Association's position and principles about disqualifying levels of nutrients for health claims, and perspectives on data needs and utility of a DV for trans fat. Mr. Earl?

MR. EARL: Thank you, Dr. Krinsky. Good afternoon to you and to members of the Nutrition Subcommittee and thank you for this opportunity to present the views of the National Food Processors Association on these important issues.

The National Food Processors Association--I think this next paragraph will cover the disclosure of all things financial related to my employment--in the National Food Processors Association we are the voice of the 500 billion food processing industry on science and public policy issues involving food safety, food security, nutrition, technical and regulatory matters and consumer affairs. We have three scientific centers and one international office in Bangkok, Thailand. Our scientist and professional staff represent the food industry interests on government and regulatory affairs and provide research, technology assistance, education and communications, and crisis management support for the Association's U.S. and international members.

Our members produce processed and packaged fruit, vegetable and grain products, meat, poultry and seafood products, snacks, beverages and juices, or provide supplies and services to food manufacturers.

There are two topics before the subcommittee today, disqualifying levels for total fat and heart health claims and evidence supporting a daily value or DV for trans fatty acids. The National Food Processors Association has been active for over a decade regarding issues related to food nutrition labeling under the Nutrition Labeling and Educational Act. Under the NLEA implementing rules, the current health claim disqualifying level for total fat is predicated on the 2,000 calory diet in which 3 percent of calories is derived from fat.

As evidenced by the content of the report from the Institute of Medicine's dietary reference intake panel on macronutrients, the ongoing discussions of the dietary guidelines advisory committee preparing the edition for 2005 and evolving peer-reviewed scientific literature, we believe it is no longer prudent or acceptable for public health to advocate a one-size-fits all dietary pattern for protein, fat and carbohydrate. Existing national dietary guidelines advocating moderate in total fat diet and FDA believe the evidence is clear that ranges of fatty intake in varying ratios of protein to fat to carbohydrate can promote health, weight maintenance or weight loss among the public, assuming a wide variety of food choices to meet nutrient needs and balance between food intake and physical activity, with the negative energy balance required for weight loss.

Based on this, NFPA believes that FDA should reconsider the total fat disqualifying level of heart health claims. The agency has already waived this criterion and replaced it with a disclosure statement for several heart health claims. Many of them were mentioned this morning and examples include health claims for stanol esters and sterolesters and the qualified health claims for walnuts and nuts. These actions acknowledge that a variety of fat levels in foods are acceptable in an overall dietary pattern, and that individual fat content of food is less important than the levels of saturated fat in reducing the risk of heart disease. The Institute of Medicine's dietary reference intake report on macronutrients also concurs that total dietary fat, within acceptable macronutrient distribution ranges, is not associated with chronic diseases, including heart disease.

NFPA believes it is appropriate to reconsider total fat disqualifying criteria for heart health claims because the nutrition label clearly discloses amounts of total fat and other nutrients. NFPA has long advocated disclosure of information over criteria that would prohibit speech. NFPA advocates a how to eat appropriate to diet and health rather than a good food, bad food appropriate that would stigmatize moderate fat foods that can contribute to overall health and dietary patterns that prevent chronic disease. NFPA believes that total disqualifying levels for heart disease health claims should be converted to a disclosure.

The second topic of this subcommittee's meeting relates to the ability of the agency to proceed with establishing a daily value for trans fat for food labeling. Given the needs for harmony and coordinating mandatory food label changes, we question whether sufficient data exists to introduce a daily value for trans fat to the nutrition facts panel. It is unclear that sufficient data exists about current trans fat content of food and food consumption patterns to make meaningful estimates to establish a daily value. Studies of dietary intake of trans fat conclude that trans fat consumption is far below the consumption of saturated fat as a percentage of energy.

FDA should also not proceed rapidly or piecemeal with further required changes to the nutrition label following implementation of the trans fat quantitative declaration final rules. NFPA believes that it is important to avoid the prospect of several sequential mandatory nutrition label revisions within the span of a few years as frequent changes to the nutrition label structure could undermine consumer confidence in the label.

Careful consideration and coordination of mandatory label changes are essential because contextual information on trans fat related to information about saturated fat and cholesterol which, in turn, relates to the future prospect of daily values for other fatty acids.

The nutrition label should not be a field for experimentation but should be a tool for conveying reliable information to consumers. NFPA believes that consideration of a daily value for trans fat should be coordinated into the future single set of changes to all daily values that FDA plans to initiate in the near future.

An additional comment is that this would also allow to take advantage of future data from food composition databases and also information from the ongoing data collection of food consumption and health indices from the integrated USDA and National Center for Health Statistics survey that was formerly best known as the NHANES survey.

Further, we believe that there is a paucity of data to demonstrate that daily values are either understood or widely used by consumers in assembling foods for a dietary pattern. A strategy should be developed for educating consumers about daily values. NFPA has been a leader in food label education and continues to believe in the nutrition label and the E in the NLEA, education. More educational opportunities need to be provided by the Dietary Guidelines for Americans, the food guide pyramid, and these are all critical for consumer understanding about the contributions of foods to diet and health.

FDA appears poised to recommit to consumer education about food and nutrition labeling, as is mentioned and evidenced in its "Calories Count" report. NFPA concurs that further education about the understanding and use of daily values is warranted.

In summary, we offer these comments to remind the subcommittee that your scientific considerations do not exist in a vacuum. They would be applied in nutrition laboratory. Because of this application, the practical effects of your deliberations should also be considered.

A couple of other points based on some of the discussion that we heard this afternoon, some of the discussion related to the pros and cons and related to positive and negative nutrients on the nutrition facts panel and the use of daily values and now considering the dietary reference intakes for labels. One thing to remember is that ULs are upper levels, not limits, and that risk begins at the establishment level of the upper level, and that daily values for saturated fat, cholesterol and sodium are upper targets currently on the label. And, these are amounts that are recommended, as this discussion has evidenced, not to exceed in planning an overall daily diet.

Another point is that perhaps the subcommittee might benefit from some glance back in history to the proposed rule for daily values from 1991 and the final rules for daily values, to look back at what were the considerations in picking the numbers and how the agency proceeded forward when there was not information in the form of an RDA to create the daily values on the label.

Dr. Lichtenstein is correct that the source of the 10 percent daily value for saturated fat on the label is from the ATP report by the National Heart, Lung and Blood Institute. But if I recall correctly, that value is still for the normal population and primary prevention I think is still 10 percent for saturated fat and 300 mg for cholesterol, with the lower numbers you mentioned that were relevant for those who had had heart disease events, stroke or something else and were on a different level of dietary intervention.

With those comments, thank you very much for the opportunity to comment on the important issues before you.

DR. KRINSKY: Mr. Earl, thank you very much for your very timely presentation. Are there any questions from the members of the committee?

DR. RIMM: If we are looking at our made-up label here, are you advocating to remove anything related to the daily value assigned to trans fat and essentially you would like to leave the daily value for saturated fat and leave a line blank next to the trans fat?

MR. EARL: Well, at present the final rule allows you to have the quantitative listing of trans fat on a separate line and, given the discussions of the subcommittee this afternoon and the level of evidence related to developing a daily value, it may be prudent to wait and roll that into the entire portfolio of daily value revisions that the agency is about to embark on and as further science unfolds related to diet and health food composition and food consumption patterns.

DR. RIMM: What we have in front of us is what is proposed currently. Is that correct?

MR. EARL: I believe what is on the label as was described by Virginia Wilkeneng was relating to what the Institute of Medicine's diet reference intakes in food labeling and fortification reports laid out as one of its principles, not as what the final rule states for labels.

DR. KRINSKY: Any other questions or comments? If not, thank you very much, Mr. Earl.

MR. EARL: Thank you.

DR. KRINSKY: Our next presenter is Mr. Martin Hahn, of Hogan and Hartson and he will be speaking on behalf of his client, GFA Brands. We have a handout that was supplied by Mr. Hahn and he is going to summarize the scientific evidence establishing the importance of considering the blend of fatty acids in the total diet when considering risk factors for heart disease.

MR. HAHN: Well, I certainly want to thank you for the opportunity to make these comments today. I noticed that I was very transparent saying that I was here on behalf of GFA brands. I was less transparent in saying that I am an attorney with Hogan and Hartson so thank you for clarifying that. For purposes of the FDA guidelines, I can assure you that I do have a financial relationship and I am being paid for being here today.

[Laughter]

That is just in case there was any doubt! A little bit about GFA, GFA is a manufacturer of many fat-containing products including buttery spreads, cheese products, microwave popcorn and other foods. Their entire product line is based on the research that is out of Brandyce University, by Dr. Casey Hayes who has done a lot of research on the importance of balancing the various fatty acids that are in the diet.

The presentation today is based on materials that Dr. Hayes has prepared I have to say it is a little bit intimidating for someone who has legal training and, although I have a degree in food technology and took basic biochem and nutrition classes, the area of expertise of this group that I am addressing far surpasses what I am accustomed to so I ask you to bear with me. I will try my best to answer any questions that you do have but when it comes to questions about specific studies, I have not read all the studies in detail and, unfortunately, Dr. Hayes could not be here to present this presentation so I am standing in his place.

I wanted to address the questions that were raised, and to address the first question--it was a very fascinating debate to date and I think the debate really underlines the complication of trying to address this whole trans fatty acids issue. But we believe, based on the science that is out there, that trans fatty acids are considered more adverse with respect to coronary heart disease risk factors than saturated fat, particularly when you look at comparing comparable levels and higher levels of trans fatty acids in the diet. We think that that biological activity is something that this group needs to consider when addressing that question.

With regard to question one, we also think that coronary heart disease risk factors can--the data will show that it is appropriate to include or allow products that may not be low in total fat or low in saturated fat to be eligible for health claims regarding coronary heart disease, and a lot of what we are going to focus on today is some of the research that talks about the importance of balancing the fatty acids in the diet. When we are talking about the correct balance, the balance that we are talking about and the numbers you are going to see throughout the presentation is, first, the percent energy of saturated fatty acids, followed by percent energy of monounsaturated fatty acids, followed by percent energy of polyunsaturated fatty acids, and the correct balance that we are looking for is a balance in the relationship of 1:1:1.

Let's first address the trans fatty acids issue. Here is a graph that was based on research out of The New England Journal of Medicine that tried to quantify the risk associated with coronary heart disease with trans fatty acids. You can see they have 132 percent; saturated fats; monounsaturated fats; and poly unsaturated fatty acids. I am not going to spend a lot of time on this, I just thought it was relevant to illustrate the point that trans fatty acids can, in fact, have a very important role in CHD risk factors.

I did, however, want to spend a little bit of time on a study by Sandram that evaluated trans fatty acids head-to-head with saturated fatty acids and compared the effect that trans fatty acids and saturated fatty acids have on the cholesterol profile. It also addresses one of the questions on this slide as well as other ones, of what happens when you substitute trans fatty acids and you replace them with saturated fatty acids. I would say it will depend in part on which saturated fatty acid is being replaced and what is the level of polyunsaturated fatty acids and monounsaturated fatty acids in that food.

When you look at this particular slide you will see that the trans fatty acid had the worst impact on LDL and depressed the HDL. It also shows that different saturated fatty acids have different effects on cholesterol profiles. If you look at what we would pretty much categorize as the control here, when you look at a product that has a 1-carbon chain saturated fatty acid, if you look at the numbers here, they are the ratio of LDL to HDL. So, you can see that total cholesterol remains about the same. The ratio to LDL and HDL remains about the same as well.

When you start looking at what happens with diets when you have a 12- and 14-carbon saturated fatty acid, you start to see a gradual decrease in the HDL and increase in the LDL. The ratio starts going up. Then when you look at the trans fat diet--and I will note that this number is a diet that contained 7 percent trans fatty acids so it is a relatively high trans fatty acid diet but, certainly based on the nurses survey or the professional survey, a diet that could be typical for those on the higher end of the trans fatty acid consumption. Here you can see how the trans fats definitely are depressing the HDL and increasing the LDL level.

In conclusion, when we look at the effect of trans fatty acids, as has been echoed by many of the people today, we really do believe that trans fats have been shown to be biochemically different and have a more adverse effect on coronary heart disease risk factors.

I want to then look at the importance of balancing the fatty acids and considering the various ratios of saturated fat to polyunsaturated fat to monounsaturated fat. Here is a diet that looked at what happens when you start removing saturated fat from the diet. They did it in two steps. They removed 4.5 percent energy in the form of saturated fat at each step and then they evaluated what happened to the blood cholesterol profiles in these individuals.

It shows that a progressive removal of saturated fatty acid does decrease LDL by 12 percent, but it also decreases HDL proportionately. So, what you end up with is you may have a decrease in total cholesterol as you increase the saturated fatty acid content but today it is important to focus on the HDL/LDL ratio. You can see that, yes, total cholesterol goes down but so does LDL and so does HDL, and the ratio of LDL to HDL remains relatively constant throughout those diets.

With the fatty acid balance we can selectively lower the LDL but not the HDL which is, of course, a goal. We are trying to increase that ratio of HDL to LDL. Here is a study that shows that by balancing the ratio you can, in fact, have a positive effect on the coronary heart disease risk factors.

Taking a look here, the regular fat diet, 37 percent energy, and here you see that 18 percent of the energy is from saturated fats whereas only 6 percent is from polyunsaturated fats. When you balance the ratio so we have 12:12:12 balanced saturated fat, monounsaturated fat and polyunsaturated fat, you can see a reduction in total cholesterol. The asterisk means it is statistically significant. A statistically significant reduction in LDL. HDL remains the same and, importantly, that HDL/LDL ratio improves.

We believe that saturated fatty acids and polyunsaturated fatty acids are required for the best LDL/HDL ratio. Here is a study that evaluates three different fat blends in 23 young men with normal cholesterol. One was a balanced fat, one was a high monounsaturated fatty acid blend and one was a high saturated fatty acid blend with low polyunsaturated fatty acids. Although all three of these diets yield about the same total cholesterol, the balanced diet produced the highest HDL and the lowest LDL.

Here you can see the balanced diet. It says the AHA because this is based on the former American Heart Association's step one recommendation. We are not exactly at 1:1:1 here but when you compare that to the other diets that are here you are certainly seeing that that is by far the most balanced of the diets, and the diet that has the best HDL to LDL ratio.

Other research has pointed to the fact that fatty acid balance can actually be as critical or as important as the total amount of fat in the diet. Here is a study that evaluated the lipoprotein profiles in 22 nuns with mildly elevated total cholesterol. There were three dietary fat mixes for six weeks each were consumed. The three diets included a high level saturated fat with 42 percent energy, a high fat with a balanced polyunsaturated and saturated fatty acid ratio, and then we had a moderate fat diet, one that goes under 32 percent of energy from fat.

What this study showed is that by balancing the polys and the saturated fats you can, in fact, improve the LDL/HDL ratio even in these relatively high fat diets. The study also shows that a moderate fat diet did not further improve total cholesterol.

Once again, here is the high fat diet. When you go to the 42 percent energy, one that is balanced with the polys and the saturated fats that are 1:1 in the monos and saturated slightly higher, you can see a statistically significant improvement in the LDL/HDL ratio. Total cholesterol goes down. Interestingly, when we look at a moderate fat diet with 32 percent energy from fat, once again the polys and the saturated fats are balanced and you can see a very limited difference in terms of the cholesterol profile.

The research conducted by Dr. Hayes and others, and these are just a few of the numerous studies that are out there, points to the fact that not all saturated fats are in fact the same. Typically, those saturated fatty acids in the 12- and 14-chain saturated fatty acids have a tendency to raise LDL the most. When you start looking a palmitic acid, which is the 16-carbon saturated fat, it has more of an intermediate effect and, when combined with a proper balance of polyunsaturated fats and monounsaturated fats, has been most successful in improving cholesterol profiles.

This is a slide that Dr. Hayes asked that I put on and display before this group. It is new research, as of yet unpublished. The preliminary results have been made public but one of the points that Dr. Hayes wanted to make is that there is a need for trans fats and saturated fats in the food supply. It is important to have a solid fat when you are making various types of foods. As you move away from saturated fats, as you move to trans fats one of the things that people are starting to look into is the interesterification effect because it also allows you to have a solid fat at room temperature.

Here is a study that just evaluated the biological effects of these various diets. Once again, it is an unpublished study but it did show that use of these interesterified fats started to show some adverse effects in terms of fasting blood glucose level. So, it was just something that we do think that the industry and the agency needs to be aware of because as we start looking for replacements for trans fats or saturated fats some of the alternatives may not necessarily put us in a better place than where the trans fats have us today. That would apply to some of the saturated fats as well.

In conclusion, I think the data that we have presented here shows that the right type of saturated fats, when balanced with polyunsaturated and monounsaturated fats, can in fact improve the cholesterol profile. Because of that, we think it is important that when the agency is looking at coronary heart disease health claims that they should continue to allow for the use of CHD health claims on products that exceed the low fat definition.

We also believe that flexibility is warranted for products that may exceed definition of low unsaturated fat because when products are formulated to have the correct balance of fact you can, in fact, have a powerful effect on the overall blood lipid profile.

Finally, we believe that the data does show that trans fats have been shown to be more adverse with respect to coronary heart disease risk factors, particularly when compared to the palmitic acid saturated fatty acid.

If there are any questions, I would be more than happy to try to answer them for you.

DR. KRINSKY: Thank you very much for your presentation. Do we have questions? Alice?

DR. LICHTENSTEIN: Yes, I have a question. Not all saturated fatty acids are created equal and their palmitic is intermediate relative to 12-0 and 14-0. I am just looking at a meta-analysis that was recently published and that doesn't seem to be consistent with the aggregate data and I was just wondering if you have any comments.

MR. HAHN: Unfortunately, I cannot comment on that recent meta-analysis, sorry.

MS. PELICAN: Could you tell us a little bit more about the study that is unpublished, by Sandram, in terms of number of subjects and sort of the makeup of those folks?

DR. HAHN: I don't have any details in terms of the number of subjects. These are only the preliminary results and he is waiting until the study is published. When everything is finalized he will be able to get it published.

DR. RIMM: Guy, do you know about any other studies that are similar to what he talked about with the interesterification causing adverse outcomes?

DR. JOHNSON: No.

DR. LICHTENSTEIN: There is one study that was published by Judd that did not make that observation with interesterified fatty acids. I am just wondering about that study that you mentioned, with Sandram, with the saturated relative to trans, their results were quite different than any others as far as the magnitude of the change. Although I guess you may not be the right person to ask, was there any explanation for that because from the data that we have on page 8-61, that is the one that showed the unusually large differential effect which is not consistent with any of the other data.

MR. HAHN: That was 7 percent energy from trans fatty acids, if that is the study that you are referring to.

DR. LICHTENSTEIN: Absolutely, but there are similar studies with 6, 8, 10 percent and you still didn't see that level of difference.

MR. HAHN: When compared to diets with saturated fats?

DR. LICHTENSTEIN: Yes.

DR. KRINSKY: If there are no other question, thank you very much, Mr. Hahn.

MR. HAHN: Thank you.

We are in the process of trying to determine whether our third speaker, Dr. Enig, is out in the hallway waiting to come in. Since we don't know if she is here, if she is here, she has not made her appearance in this room, we will defer her presentation and we can continue with our discussion.

Subcommittee Discussion (Continued)

DR. KRINSKY: On trans versus sat, would anybody like to comment on whether we can determine these differences in the isolation of other dietary changes, if we change trans for saturated fatty acids. For example, Alice, in your studies, were those the only variable in the groups prior to the butter group?

DR. LICHTENSTEIN: Yes. It turns out they were and the difference, though, was dictated by the products that were available. But it turns out that the monos were essentially the same. The sats were essentially the same except for the butter, so it was a poly-trans difference.

DR. KRINSKY: In this study, although it is only one study, the change in the LDL cholesterol was not significant as you went up to what percent trans?

DR. LICHTENSTEIN: It was at the lower levels. I think by 3 percent it was significant.

DR. KRINSKY: By 3 percent.

DR. LICHTENSTEIN: But certainly there was a trend there.

DR. KRINSKY: So what we are left with with Question 3, I think one thing that I have used in the past is whenever things get sticky and tacky is maybe to divide the question and let's see whether we can't come up with some kind of division of this question so that we might be able to come up with a reasonable answer.

DR. LICHTENSTEIN: I don't think we are necessarily disagreeing that they are not different. I think they are different. I think the only issue, perhaps, and maybe that oversteps the question, is whether, at the current levels of intake, they are different enough to make the distinction. That is really not what the question is. The question is do they have different biological effects or not. From my perspective, the data is yes; they do have different biological effects.

DR. KRINSKY: But the question is not just different biological effects but adverse with respect to coronary heart disease.

DR. LICHTENSTEIN: Using the data that is most prevalent which is lipoprotein levels, yes, because of differential effects on HDL cholesterol. Then the other potential effects, I think the data are scarcer and somewhat less consistent although--and we are somewhat constrained because of the paucity of data.

DR. KRINSKY: So a response to the question that trans fatty acids are more adverse--that is, consumption of trans fatty acids are more adverse with respect to coronary heart disease would be a reasonable response to this question.

DR. LICHTENSTEIN: Qualifying on the basis of the total-to-HDL ratio.

DR. KRINSKY: The LDL-to-HDL ratio.

DR. LICHTENSTEIN: Either LDL-to-HDL or total-to-HDL.

DR. KRINSKY: Okay. We do have Dr. Enig here.

Would you be prepared to speak now, or would you like a moment?

So we seem to be moving towards agreement at some level with respect to this question, with the adverse effect of trans versus saturated fatty acid. Then we come into the fact that just that statement, by itself, is not going to be the most useful statement for the public interest. I think that we would should not neglect the fact that we are an advisory committee to the FDA and the FDA is really responsible to the public.

So would you like to comment on that, Guy?

DR. JOHNSON: I was going to clarify what the consensus that was emerging is, and that is that trans fatty acids do have a different effect than saturated fats based on the big caveat that the HDL fact is taken into account. That means at higher levels. That is where, I think, we are struggling, is kind of where is the dividing line and is that significant from a public-health perspective based on current intake levels.

DR. KRINSKY: Anybody have a level that they would like to propose? Does anybody have the temerity to say 2 grams per day about that would be adverse?

DR. JOHNSON: See, I just don't think we have the data to do that. You can look at this HDL stuff and say it is 6 percent. You can say, well, one molecule is undesirable. So it is just kind of looking at the limited data we have and reacting with our own perspectives, I think. That is the problem.

I guess if you had a nice randomized, double-blind, placebo-controlled trial for every answer, you wouldn't need committees like this. I think that is where we are at.

DR. KRINSKY: We need a committee to verify that the data that was being presented was, in fact, valid data. but we may have an opportunity to hear a little bit more.

I would like to introduce Dr. Mary Enig. Dr. Enig is the Vice President and Science Advisor of the Weston A. Price Foundation. She is going to talk to us about the second and third questions, n the level of trans fat in the diet and the health effect of trans fat on risk levels of coronary heart disease.

Public Comment (Continued)

Do you have a Powerpoint presentation?

DR. ENIG: No, I didn't bring one because, given ten minutes, there was hardly very much that you can put into a Powerpoint presentation.

DR. KRINSKY: Oh; you can actually get much more in than talking. You can have your slide in a very minute font and really present an immense amount of material. But we are grateful that you have not tried to do that.

DR. ENIG: I remember very distinctly that Dave Kritchevsky indicated that he didn't bring any slides to the Cholesterol Consensus Conference because, if he did, he wouldn't have enough time to even say what he wanted to say. So--

DR. KRINSKY: May I interrupt you for a moment, Dr. Enig. I read a disclaimer about your financial relationship to the Foundation and to sponsors, but have you presented before the FDA?

DR. ENIG: Yes; I have.

DR. KRINSKY: So you have had this disclaimer read to you?

This is not a new disclaimer, is it?

MS. LATHAM: It is relatively new.

DR. KRINSKY: I am going to give you the privilege of listening to this before you start your talk.

DR. ENIG: Sure.

DR. KRINSKY: Both the Food and Drug Administration and the Public believe in a transparent process for an information-gathering decision making. To ensure such transparency at the Open Public Hearing session of the advisory committee, FDA believes that it is important to understand the context of an individual's presentation.

For this reason, FDA encourages you, the Open Public Hearing speaker--that is a general "you." I am not directing this specifically at you--the Open Public Hearing speaker, at the beginning of your written or oral statement, to advise the committee of any financial relationship that you may have with the sponsor, its product and, if known, its direct competitors.

For example, this financial information may include the sponsor's payment of your travel, lodging or other expenses in connection with your attendance at the meeting. Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships. If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.

So you are free to speak to us.

DR. ENIG: My response to that is that I am a member of Weston A. Price Foundation, a volunteer member. It is, I guess you could call it, like, a volunteer organization. I am the science advisor because I am one of the few people who actually happens to have the type of scientific training that is needed by the organization to be able to make sure that the information that we are putting out for the public is accurate.

My expertise happens to come from many years of being at the University of Maryland in the Lipid Research Group working on trans fatty acids, saturated fatty acids, other types of lipids. Since I left the University of Maryland, which was in the 1990s, I have followed the literature extremely closely. I have gone to meetings even though I, myself, have been the one that sponsored the expense.

I just went to the Trans Fatty Acid meeting in the Chicago area in February that was put on by the American Oil Chem Society. I sponsored myself, I guess you could say. At that time, I noticed that there was an awful lot of spinning being done, that the information that was being put together, some of it extremely interesting, was either incomplete or else it was just with a little spin here and a little spin there, and that included the amounts of trans that are actually in the diet.

I happen to know from my contacts with the people in Canada, that the amounts that I predicted as probably being in the diet were exactly the same as the amounts that they ended up predicting that some of my colleagues at the University of Maryland also came up with the same amounts in the diet that the people in Canada did.

You don't have a lot of trans fatty acids in the diet of a formula-fed infant. You may have a tremendous amount of trans fatty acids in the diet of a nursed infant if the mother takes in 30 grams of trans, as the Canadians found she did. They ended up discovering, both in Canada, research from Sheila Innis, and in research in Holland with Hornstra, that these infants, as they had more trans in their diets, ended up with visual-acuity problems.

So, as we have gone from the 1970s, when I first became concerned and interested and first reported to the FDA that there were potential problems with the trans fatty acids, we have ended up seeing not only that there definitely is an effect, if you want to consider an increase in LDL as an effect on the heart-disease issues or a decrease in HDL, an increase in lipoprotein (a); in other words, all of the things that the clinicians say are not good the trans fatty acids cause.

When you look at the research, you also discover that the saturated fats do exactly the opposite. So I have been following this in the literature. We will have a couple of people following things at the University of Maryland and I have, as I said, been in touch with people in Canada and, to a certain extent, also in Europe. Gradually, more and more of the health problems that we didn't even predict have cropped up.

One particular one, Europe has discovered and reported, and I think there has been some reporting in this country, that you see asthma increasing from the trans fatty acids in children.

Well, that is not surprising because the children who are consuming partially hydrogenated vegetable fats in their crackers and cookies are probably not getting enough of the carbohydrates to make their own saturates and they are not getting enough saturates in the foods that they are eating because they have been replaced by the trans. They need saturates in their lungs for lung surfactant. I think Alice and I had a discussion on that a couple of years ago, as a matter of fact, at the American Heart Association meeting.

So the trans are very definitely a problem and the amounts that were taken in can range anywhere from, I guess, 2 grams for people who are not taking in very much to 40 grams. In Canada, they have verified this. In Scotland, they have verified this. At the University of Maryland, we verified this.

I actually had a student that I was teaching who had to put down all of what he ate and then do the calculations, come up with over 100 grams of trans fatty acids. So I have no difficulty thinking that there are a lot of diets out there that have a large amount of trans fatty acids.

My own personal feeling is that, except for the ruminant trans, which are a totally different type of fatty acid, react differently in the body, and that research has been done and the committee had a note from Dr. Kummerow to make sure that that was kept separate. Except of that, my feeling is 0 is the amount because we have no need for trans fatty acids in our tissues. We have no need for trans fatty acids in our phospholipids. We have no need for trans fatty acids in our lungs.

We have gotten along very well without the trans fatty acids for many years. The only people who didn't get along very well without the trans fatty acids were the industry because they had more expense with the more saturated animal fats. So, as far as I am concerned, we can have a 0 on that. But then, of course, you can't do some of your menu-modeling if you have got 0.

Now, with respect to the saturates, I was around when the anti-saturate agenda started. I think that almost all of the researchers who say that saturates are bad have never read the original literature. They just simply don't know who did what, who pulled what. I happen to have known, for quite a long time, and I just recently started to see some of the research coming out where, in fact, the saturates were showing up good.

But the people who were writing the papers sort of tucked that under the bottom of some of the tables and graphs and so forth. There are three papers that came out from--it is Dreon, et al., so that is Krauss's group out in California--where they lowered the total amount of fat in the diet in their cohort that they were studying. What did they end up getting? They ended up getting things going the wrong direction.

They ended up with the LDL going from the light and fluffly LDL that everybody has now decided is okay and they are not quite sure what they are going to do with reporting it to the public, and they got an increase in the small, dense LDL. Then they did the same thing with saturated fat. They lowered the saturated fat and they also got the wrong kind of a response.

Then they looked at some of the children of the people that they were studying and they discovered, in fact, that the children who should have been like the (a) type became a (b) type, who should have been like the (b) type became an (a) type. So, removing saturated fatty acids from somebody's diet does not do something good.

DR. KRINSKY: You have about one more minute of your time.

DR. ENIG: I had sort of prepared just three comments to the three questions. I think I have actually more or less covered them. What I wanted to say was basically that the current scientific evidence indicates that the idea of lowering total unsaturated fat is going to be counterproductive in the long term because there are enough people like myself and others, and especially in Europe, who are very, very much aware of this.

Then, the scientific evidence supports a level below 1 percent of the energy, although I said, of course, 0, for the commercially produced trans fatty acids. But, as pointed out by others, Dr. Kummerow today, this should not apply to ruminant trans fatty acids.

Then, the third thing was a comment that, compared to saturated fatty acids, which have no adverse effects with respect to coronary heart disease, the commercially produced trans fatty acids are more adverse in their effects and ruminant trans are not a problem.

I will be happy to take any questions.

DR. KRINSKY: Thank you very much, Dr. Enig. Do we have some questions for Dr. Enig, please?

DR. RIMM: I am not sure, at the beginning, did you mean to imply that our estimates, our food-composition estimates of trans fatty acids greatly underestimate what we are really eating? Before you came in, we had discussed that, on average, in this country, or at least five years ago, consumed 2.6 percent of energy from trans. Do you think that is a dramatic underestimate based on our food-composition tables?

DR. ENIG: I think it is because if you go in and take a look at how much an infant is taking in, or a child is taking in, and do the weighted average of some things of that sort, you will get a totally different amount. The amounts that people come up with when they start doing the types of analysis depend on what they want to find as an answer.

I spent years, literally starting in the late '70's, having this happen to my work from what the industry wanted reported. All I can tell you is, right now, today, after having provided some information to people in Europe, didn't tell them how they should think, didn't tell them what they should do, but I gave them the type of information that they wouldn't have otherwise had access to and they could run with that.

Denmark has said, "It is going to go out of our food supply and anybody that puts it in," this is the commercial trans," is going to go to jail.

DR. RIMM: What have they done instead? In Denmark, what have they done instead to create foods that need to have solid components?

DR. ENIG: They have replaced the partially hydrogenated vegetable fats with the normal saturated animal and tropical-oils fats; in other words, there exists in the world, and have existed in the world since the beginning of who knows how long, certainly the beginning of this century, adequate amounts of these types of fats, and we never had any of the problems that we have been having when they were being used.

DR. SHANNON: Dr. Enig, is it your position that all saturated fats are appropriate in the diet and that we should not be concerned with reducing saturated fats of any kind?

DR. ENIG: Basically, that is what my position is. If you are talking about palmitic acid, palmitic acid is very essential, dipalmitoyl(?) phosphotidalcholine is what the lung surfactant is. Palmitic acid is absolutely essential for growth in infants and children. Myristic acid is what the body uses to do what is called myristilation(?); in other words, it attaches myristic acid to a type of protein, protein molecule, and that is used as a messenger type of protein.

You really don't very much in the way of myristic acid until you have animal fats in the diet. It is almost nonexistent the vegetable fats. Lauric acid, which the body needs, used to get small amounts of it on a regular basis, is made by the mammary gland because infants need it for their development. You can get it in coconut oil or palm-kernel oil, and you can get some of it in milk fat. But you don't get it in any of the other vegetable oils although there were some attempts to make, to grow, a vegetable oil that would have it because the company recognized that it was something that was going to be a product down the road to be making money from. And then they ran into trouble growing it.

Stearic acid, everybody who has been anti-saturate for as long as I can remember, is something that everybody says, oh; well stearic acid is not a problem. Stearic acid is found in very small amounts in most fats and oils except cocoa butter and some of the animal fats.

DR. KRINSKY: Dr. Baker?

DR. BAKER: Thank you so much for mentioning children. I have been waiting to hear their name brought up and I actually had some comments. What I really like, in your handout, do you have references, or will you share with us the calculations that you have done on their intake and outcomes that you know about? I would really love to see that.

DR. ENIG: I have a paper from 1990 published in the Journal of the American College of Nutrition.

DR. BAKER: On trans fats?

DR. ENIG: On trans fat, and so forth. I had another one that was published in Cereal Foods World.

DR. BAKER: Okay. Is that part of your handout? Do you reference it? We could find it.

DR. ENIG: I don't know if it is in the references that we put together, the ones that I put together. No; I don't have that, but I can certainly, if you want me to put together that particular information, I can certainly do that and--

DR. BAKER: Just the references. Thank you so much.

DR. KRINSKY: Any other questions? If not, thank you very much, Dr. Enig.

DR. ENIG: My pleasure.

Committee Discussion (Continued)

DR. KRINSKY: We return to our committee discussion. We are on Question 3. We are dealing with a quantitative issue that may be irresolvable unless we go to nonquantitative terms like big and little, unless somebody can help me. Can we get anything semi-quantitative, qualitative--qualitative we can talk about the fact that trans fatty acids can increase the risk of coronary heart disease and are detrimental. But how much?

Do you know how much Eric?

DR. RIMM: I think we actually all did sort of agree that trans fat can be more detrimental than saturated fatty acid. The only issue is that Guy had the caveat that it has to be at high enough intake.

But the answer to the question is, is trans more adverse than saturated fat. The answer to that would be Yes. There are caveats for everything we are going to discuss here, and I am sure there's caveats for everything that is currently on the Nutrition Facts label. But, based on what we discussed and based on the published evidence, it would suggest that it is more detrimental.

DR. KRINSKY: Does anybody feel an obligation to come up with when does it become more detrimental? At what level of intake does it become more detrimental or is it more detrimental at all levels of intake.

DR. MAYNE: It seems to me that it is a continuum, so I would be more comfortable saying something to the effect that saturated fatty acids are considered to be more adverse--I'm sorry; trans fatty acids are considered to be more adverse when compared to saturated fatty acids with respect to coronary heart disease, and then say this is particularly true in looking at LDL, HDL ratios as the outcome of the interest and the difference is more noted, or more magnified as intake levels increase.

I think we are all in agreement with that because it is a continuum and I am not really comfortable setting thresholds because the data doesn't look like a threshold to me.

DR. HINE: To further muddy the waters, I raise the question of duration of exposure because, as pointed out by Susan Baker, the issue of children who are now exposed to a lot more trans fatty acids than many of this in this room have been exposed to. I think that is an important scientific question as well.

DR. RIMM: I will third that motion. I think, in addition to being concerned about trans in children, I think that when we looked at our nurses' data and you looked at trans intake at baseline, at estimated baseline, we found some increased risk of coronary disease. But, if you updated their diets every four years so you had a better, tighter estimate, you got a stronger relative risk which suggests that there probably is a long-term effect that becomes to accumulate that maybe you can't capture in a four-week metabolic-ward study.

DR. JOHNSON: I am not sure, based on the data that we have, I can buy the continuum conclusion since there just aren't any datapoints at levels around the mean intake or below, other than 1. So it is anybody's guess as to what is going on there. We don't know whether there is a threshold or not. The HDL data would suggest that there is but it is not at the level where we really are interested. So we just don't have the data, I don't think.

DR. MAYNE: Can I respond to that? Susan Mayne. True. Clearly, we are lacking datapoints. We all agree on that. But I think the more conservative estimate is to assume a linear relationship. I think it would be erroneous to assume a less conservative relationship, given the lack of data. So, given that we are trying to protect the consumer, I am much more comfortable assuming a linear relationship based on a lack of data than assuming a threshold when there is really no evidence.

In my view of that data, there is no evidence for a threshold effect there.

DR. JOHNSON: Then we get into sort of a philosophical thing and maybe it comes a consumer research issue because it is a conservative estimate. Maybe your position is conservative under the assumption that we should alert consumers to potential increased adverse effects for trans fatty acids, regardless of whether there is a threshold or not, on the assumption that that is going to help them make the appropriate changes in their diet.

However, we are doing a disservice by making that point. We are giving people an unrealistic perspective on trans fatty acids and, perhaps, encourage them, unwittingly, perhaps, to focus on trans rather than saturated. We all know that saturated fats are present in the diet at much higher levels and we don't want them to make trade-offs based on what could be a false distinction between the two at the levels where people really are now.

That wasn't very articulate, but I think you know what I am trying to say.

DR. KRINSKY: Would there be any damage if somebody moved from a 2 percent intake of trans fatty acid to 1 percent with the substitution of a saturated fatty acid?

DR. LICHTENSTEIN: Unlikely.

DR. KRINSKY: Unlikely?

DR. LICHTENSTEIN: Or, no.

DR. KRINSKY: So we don't have any threshold data. That is for sure. The linear data, certainly it looks linear from the Ascherio plot that we had presented to us. And, if that is the case--well, let me back off for a second. Can you present an argument that there may be a threshold because I would assume that we either have a linear response or a threshold response. Do we have any indication whatsoever that there is a threshold that, below a certain level, there is no effect.

Now, I know that is dangerous because the ability to measure these effects get more and more constrained as you go to lower and lower levels.

DR. LICHTENSTEIN: That is what the IOM committee came--the decision they came to is that, because there is no advantage to consumer trans fatty acids, you can't come up with any sort of lower limit because the lower you go, it appears, the better off you are.

DR. KRINSKY: I know that Guy is concerned about the educational aspect, or the information aspect, of transmitting that to the public, that the lower you go in trans fatty acid, the better off you are going to be, and how that will be resolved in terms of dietary selections as they look at the nutrition label.

DR. LICHTENSTEIN: I mean, that is something that I brought up at the beginning. I don't want to really repeat a lot, but the issue becomes how do you communicate it to the consumer. From my perspective, what you want to communicate is that you would minimize your intake of trans and saturated fatty acids. You need to do that in the most direct way possible.

If you communicate to them that you have got to keep track and balance your sats and trans, I don't know what they are going to do with that. I mean, that is what we need consumer, I guess, research for. But I tend to be minimalist when it comes to communicating information.

DR. JOHNSON: Can I respond to your question about is there any evidence for a threshold. I feel like I am beating a dead horse and forgive me if I am, but you could answer that question yes, if you take studies, as my friend did, in the Ascherio plot and, instead of plotting the percent intake, energy intake, of trans and saturated fats versus the ratio of LDL to HDL, you tease those out, and you simply look in the papers and see which changes in the trans group compared to the control group were significant, statistically significant.

He is not doing any statistical smoke in mirrors here. He is just saying, was it significant or not. And, when you put points on that dot, you see that the studies below the current mean intake of trans fats did not raise LDL statistically significantly. Maybe we don't know if that is biologically significant or not, but statistically, they didn't. If you look at the changes in HDL, there were none of the studies up until over 4 percent of energy and then only one was statistically significant when they looked at trans diets versus the control diet with respect to HDL.

So it looks like there is a threshold there.

Then, and I apologize to Eric again, but you look at the epi studies and they all--maybe this isn't fair, but they all show nonsignificant relative risks until you get to the upper quintile. Now, maybe that is unfair and maybe not, and I understand how you are unlikely to get a nice quantitative dose response in each one of those quintiles, but the fact of the matter is, they weren't significant.

You can assume that there is a trend there, but it is an assumption. So I am coming back to the fact that there is some evidence for a threshold if we are looking for that. Is it iron-clad? No; but there is some evidence that certainly injects some doubt into the conclusion that there isn't.

MS. PELICAN: Pelican. Guy, just to follow up on that. If you were to, assuming that you have said that you think there is a threshold, what would you suggest if the threshold could lend itself to a D.V.; what would you see that D.V. being?

DR. JOHNSON: I am not comfortable coming up with a number. The IOM couldn't come up with a number. I am certainly not going to do one. If you look at the HDL data, you would say 5 to 6 percent. But I am not going to say that there is enough data to put that on the Nutrition Facts panel as a daily value.

That decision is a profound one that is going to have to be defended. I just think we need a more thorough assessment of all this. Maybe we need new data. I just don't think we are at the point where we can bet on the outcome with the Nutrition Facts panel.

DR. KRINSKY: Eric, can you respond, again, to the quintilization of the data for the epidemiological surveys?

DR. RIMM: Yes. I mean, we are sort of getting down to mincing--fighting over small words here. I think the fact of the matter is that Guy agrees with Question 3, that trans fat is worse than saturated fat. Now, he may think there is a threshold or maybe there are other subtle things in there, but, the fact of the matter is I think everybody agrees that trans fat is worse than saturated fat.

Dr. Enig has done some of the premiere research on this area and thinks it should be 0 percent from hydrogenated oils. Maybe some of us think that. Obviously, there is going to be some cutoff above that if there is any cutoff at all.

But, again, no single study, I think, should be used to summarize the data, no single metabolic study, no single observational epidemiological study should be used. But if you take the four or five large observational studies on trans fatty acid and coronary heart disease--that is not just from here. It is from Costa Rica. It is from the Netherlands as well as several from the U.S. and Finland--and plot them, rather than just doing it be quintiles, if you actually plotted by the median of each quintile so the fifth quintile would end up being further out because of the fact there is a much wider range in that highest quintile, and draw a line, you are not too far off. You are not exaggerating too far off to say that there is a linear association as you go up with intake.

I don't think anybody would be surprised that the second quintile is significantly different than the first quintile. The studies aren't powered to do that. There are not enough cases to do that. You would need to have 3 million people to follow for a long time.

So that is why we use an estimate dose response and to a test for trend rather than counting on each quintile being significant different from the first. So I think that if each one of the studies had found the relative risk of 1, 1, 1, 1, 1 and 1.7 in the highest quintile, we would be concerned. But, in fact, that is not what we see for the dietary data and that is not what we see for the data where trans fatty acids are isolated from adipose.

I think, if you were to say, I am definitely sure about one, no one would say that. I am definitely sure that there is a threshold. No one would say that. But I think the evidence would support that there is more of a dose response. I think there is more evidence to support there is a dose response than there is evidence to support there is a threshold.

DR. KRINSKY: We all have the opportunity to express our opinions. As a matter of fact, one policy that the FDA--some of the advisory committees have done--I don't know whether, in fact, we are obligated to, is that when the time comes for voting, the members have to tell why they say yes or no so that there is some support in the record of the position that the individual members of committee take.

But I am not sure that, at this time--no; I am sure. At this time, I think, in fact, we can take a vote on Question 3 as it is formulated to us. That formulation is, as we all have in front of us, when compared to saturated fatty acids, are trans fatty acids considered to be more, less or similarly adverse with respect to coronary heart disease.

I will start out by asking for a show of hands for those that feel that trans fatty acids are more adverse.

(Show of hands.)

Thank you. Trans fatty acids are less adverse?

(No response.)

And trans fatty acids are similarly adverse.

(No response.)

Guy you don't like the question?

DR. JOHNSON: I am not allowed to vote.

DR. KRINSKY: Oh; sorry. Look what I just gave you, the opportunity.

Well, with that momentous vote, I think that it is, in fact, time to adjourn our meeting until tomorrow morning where we will have an opportunity to go back to Question 2 and even vote on Question 1 which we didn't vote on.

I just want to remind the committee that the meeting tomorrow meeting--the official meeting will begin at 8:00 in the Monet Room. The Monet Room is up on the second floor.

DR. LICHTENSTEIN: Can I just ask the Chair a question?

DR. KRINSKY: You certainly may.

DR. LICHTENSTEIN: Why don't we vote on Question 1 since the discussion is sort of fresh in our heads? Is there a reason for waiting until tomorrow for Question 1?

DR. KRINSKY: No. Do you feel that we are prepared to vote on Question 1? Yes; as a matter of fact, considering the discussion, what does the current evidence suggest--okay. The question is, what does the current evidence suggest in terms of total fat intake and risk of coronary heart disease.

That is not a yes or no. So I think the question will have to be rephrased to give it a yes or no answer. That is going to be done by the Chair and the Exec Secretary between now and 8:00 tomorrow morning. At 8:00 tomorrow morning, this committee reconvenes in the Monet Room. But, prior to that, from 7:15 to 8:00 a.m., you will be eligible to receive a continental breakfast in the Montcalm Room.

I hope to see all of you before 8:00 eating our high trans foods in the Montcalm Room.

Thank you all for your attendance and thank you all for your patience.

(Whereupon, at 6:00 p.m., the meeting was recessed, to resume on April 28, 2004, at 8:00 a.m.)