|
|
In the Matter of:
Dietary Guidelines Advisory Committee Meeting Transcript
Thursday, June 17, 1999
Economic Research Service
1800 M Street, N.W.
Waugh Auditorium
Washington, D.C.
Pages: 318 through 581
HERITAGE REPORTING CORPORATION
Official Reporters
1220 L. Street, NW, Suite 600
Washington, D.C.
(202) 628-4888
The hearing in the above-entitled matter was convened, pursuant to Notice, at 9:03 a.m.
IN ATTENDANCE
CUTBERTO GARZA, M.D., Ph.D.
Vice Provost and Professor, Cornell University
Associate Director, Food and Nutrition
Programme, United Nations University
RICHARD J. DECKELBAUM, M.D.
Director, Institute of Human Nutrition
Columbia University College of Physicians and Surgeons
JOHANNA T. DWYER, D.Sc., R.D.
Director, Frances Stern Nutrition Center
New England Medical Center
Professor of Medicine (Nutrition) and Community Health
Tafts University School of Nutrition
SCOTT M. GRUNDY, M.D., Ph.D.
Chair, Department of Clinical Nutrition
Director, Center for Human Nutrition
University of Texas Southwestern Medical
Center at Dallas
RACHEL K. JOHNSON, Ph.D., M.P.H., R.D.
Interim Associate Dean, College of Agriculture and Life Sciences
Associate Professor Nutrition and Food Sciences
University of Vermont
SHIRIKI K. KUMANYIKA, Ph.D., M.P.H., R.D.
Associate Dean for Health Promotion and Disease Prevention
University of Pennsylvania School of Medicine
Center for Cinial Epidemiology and Biostatistics
ALICE H. LICHTENSTEIN, D.Sc.
Professor, Tufts University School of Nutrition Science and Policy
Senior Scientist
Jean Mayer USDA Human Nutrition Research Center on Aging
Tufts University
SUZANNE P. MURPHY, Ph.D., R.D.
Researcher, Cancer Research Center of Hawaii
University of Hawaii
MEIR J. STAMPFER, M.D., Dr. P.H.
Professor of Epidemiology and Nutrition
Harvard School of Public Health
Associate Professor of Medicine
Harvard Medical School
LESLEY FELS TINKER, Ph.D., R.D.
Assistant Member, Fred Hutchinson Cancer Research Center
Affiliate Assistant Professor
Department of Health Sciences
University of Washington
ROLAND L. WEINSIER, M.D., Dr. P.H.
Chair and Professor, Department of Nutrition Sciences and Medicine
School of Medicine
University of Alabama at Birmingham
DR. XAVIER PI-SUYNER
Columbia University
DR. CLAUDE BOCHARD
Director, Pennington Center
EILEEN KENNEDY
SHANTHY BOWMAN
CAROLE DAVIS
ALYSON ESCOBAR
JOAN LYON
KATHRYN McMURRY
LINDA MEYERS
CAROL SUITOR
DR. GARZA: Yesterday, we heard from Dr. Ludwig on the glycemic index and health disease relationships and out of a concern that there are certainly more than one viewpoint on this issue, we tried to find representatives that might give us a good view of the spectrum. So, we are very lucky that Dr. Xavier Pi-Suyner agreed to come, with much prodding, from a very busy schedule. We're very pleased that he made some time to join us this morning.
So, without my taking any further time, Dr. Suyner is from Columbia University, extremely well known researcher in the area of diabetes, obesity, energy metabolism. Thank you very much for joining us.
DR. PI-SUYNER: Well, it's a pleasure for me to be here and speak with you about the question of carbohydrate and glycemic index. There's been a lot of debate over this subject over the years, and I think it's been primarily directed at the diabetic population. But, what you're talking about here is directing it to the population at large, and I think that brings up some issues that I would like to discuss.
I'm sorry for my overheads, but I was travelling and I didn't have a chance to copy them. But, what I'm going to talk about today first is a diet with high glycemic-index food detrimental to health? And, is a diet high in carbohydrate detrimental to health?
And, the most often described health risks with regard to both of these topics have been obesity, diabetes mellitus, dyslipidemia, primarily, and hypertriglycidemia. I'm uncertain, scattered reports of cancer, particularly of colon cancer, lung and breast. Now, I'm not going to deal with the cancer issue, because I think there isn't really enough data to deal with that, and I don't think it's relevant to your discussions at the present time, given the paucity of information.
I want to start with the second point first, namely, is a diet high in carbohydrate detrimental to health? And, I'm going to spend very little time on that and go on to the glycemic index. I think my reading of the literature is that there is no convincing data that a high-carbohydrate is detrimental to health. If you look at the epidemiological studies over the years, if you look at traditional studies in traditional societies, the Japanese, the Chinese, the Mexicans, many others, East Indians, they generally were eating a low-fat diet, high carbohydrate, and low calorie, and their incidence of diabetes mellitus and of cardiovascular disease was really much lower than Western countries.
When you go to transitional societies that are moving from the traditional to the more Western society, we've had really quite a lot of studies over the years. We've had studies of the Japanese, of the Chinese, of the Mexicans and of the East Indians. The East Indian data mostly from the United Kingdom, and in all of those societies, they're moving to a higher fat, lower carbohydrate, higher calorie diet, which, with an increased incidence of both diabetes and cardiovascular disease.
Now, there's a confounder here between higher fat and higher calories, so that one can't ascribe it all to fat. Much of it has to do with a higher caloric content of the diet.
And, then, finally, the highest incidence of diabetes and cardiovascular diseases in Western countries with a higher fat and lower carbohydrate and higher calorie diet. But, essentially, the lower caloric intake of most traditional societies worked against the appearance of diabetes and cardiovascular disease.
Now, aside from these isolated studies, there are also quite a number of comparative international studies, that is, studies done by one group or individuals across the spectrum. There have been the studies in Japan and Japanese-Americans in Hawaii and the Japanese-Americans in the United States that show that as carbohydrate is decreased and fat is increased, the risk of diabetes and cardiovascular disease goes up, and it's highest in the U.S.-Japanese.
Now, again, this is confounded to some extent by increasing caloric intake and increasing body weight. In Mexicans, the San Antonio group at the University of Texas, Steve Haffner and his group, have done studies in Mexico, retrospectively, and then in Mexico now, and in Texas now, similar genetic pools, showing that with increasing fat and decreasing carbohydrate and increasing calories, again, we get a much higher incidence of diabetes and cardiovascular disease.
So, I see no data in the epidemiological literature to suggest that taking a high carbohydrate diet is detrimental to health. Now, taking a high caloric balanced diet may be detrimental to health, but that is across the board in terms of calories and not focusing on carbohydrates.
Now, I want to move onto the glycemic index and the first thing I'd like to do is just to define the glycemic index for you in case you don't remember it. The glycemic index really deals with a meal test. A meal test is to compare 75 grams of glucose with 75 grams of any food that you're trying to test, and you're testing the food as a pure food, as that single item of food.
What you then do is do essentially a tolerance test and follow the glucose over time, over a period of two hours, and the glycemic index is the curve of the food over the curve of glucose times 100, okay. So, you take whatever curve you get for glucose, whatever curve you get for the food, you put one on top of the other, and then you multiply by 100. It's a percentage of the response you get from glucose.
Now, there's been a lot of argument over the years about whether the glycemic index should be measured as the area under the curve from the fasting, which is what I've shaded here, or the whole area under the curve. And, I can tell you, volumes have bee written about which way this should be done. The proponents of the glycemic index as something that is kind of pathological and you have to worry about, have argued that you should take the area above the fasting level, the difference between this and this. Other proponents have argued that insulin is insulin, and you can't detect this molecule from this molecule, and that the area under the curve is what is important when you calculate the glycemic index.
This makes a big difference because we're talking about very small statistical differences between one glycemic index food and another, okay, and if you're going from a glycemic index of 65 to a glycemic index of 70, you may find a difference statistically if you do the area above the curve, but you may not find it if you do the whole thing. So, there's a confusion right from the start as to what the glycemic index really is and how you define it. And, there are proponents to the idea that it should be the area above the curve. There are proponents to the idea that it should be all of the insulin, since all the insulin is the same and all the insulin is circulating and why make an arbitrary decision by the amount of insulin that comes out after a meal?
Now, just to give you some example, I've put here a glycemic index of some foods. You can see that white bread is 100. That means that the glycemic index of white bread is exactly the same as that of glucose. Dark bread varies very widely, 58 to actually 100. Mashed potatoes, 104. If you don't mash the potatoes, it's much less, goes down into the 70s. Cold breakfast cereal can vary anywhere from 72 to 127. The cola beverage is around 87, apples, 65, orange juice, 65, yogurt 35, broccoli, 45, peanut butter, 40. Chocolate is also very low, but peanut butter and chocolate, as you know, have a significant amount of fat.
The point I want to make with this slide is that there are foods, and I'll come back to that, where there's a very wide variation between different foods and the way they're processed, the way they're cooked, with regard to the glycemic index. The more processed the food, the more disrupted you say, the less natural the food, generally the higher the glycemic index.
The more unobstructed the food, the less cooked the food, generally, the lower the glycemic index. Now, if you'd talk about high and low glycemic index effect, the individuals who proposed the glycemic index hypothesis on health suggest that with a low glycemic index, you would have a lowered glucose response. As a result, you will have a lower insulin response. You will have less suppression of free fatty acids, so your FAA will be elevated, will be more elevated.
As your glycemic index increases, you're going to have most higher post-prandial glucose. As a result, you'll require more insulin and you'll suppress fatty acids more distinctively. Now, the carbohydrate insulin hypothesis thing goes that with increased insulin that you get from the high glycemic index foods, you're going to have an increased food intake. In other words, it will drive the hypothalamus to make you hungry, you will eat more, and you're more prone to obesity.
The other hypothesis which has been proposed by the group at Harvard, the Nutrition Department, is that the high insulin would lead to beta cell exhaustion over time, because of the greater insulin demand, and that would lead to diabetes.
And, the third that is proposed by some, is that the increased insulin leads to an insulin resistant state, which then can lead to diabetes and cardiovascular disease.
Now, you probably heard yesterday from Dr. Ludwig, who is a proponent of this theory, the arguments for, and so what I would like to do is to give you some arguments essentially against this hypothesis. Now, the first study I want to show you is a study that was published by Keens and Richter in Scandinavia, in Denmark, I believe, in which they tested high glycemic and low glycemic index diets in individuals. And, the reason I'm showing you this amongst the many, many studies that are done is that I think it's a mistake to look at this question of glycemic index in terms of a single meal.
Most of the data dealing with increased food intake with high glycemic index deals either with a single meal, where you give a high glycemic index meal or a low glycemic index meal, and then you follow food intake over a period of time. Or, like Dr. Ludwig did it, he did two meals, breakfast and lunch and then went for five hours after lunch.
Now, whereas I have no problem with the design of that experiment as an experiment, I think what we're really talking about when we talk about giving recommendations to a population, what we want to talk about is the long-term effect. Now, there are very few -- if you go through the literature, there are almost no long-term effects, long-term studies of high glycemic index versus low glycemic index in the literature. They are rather low.
But, I found this one. This is done by a group in, as I said, in Europe, and they provided food to these individuals for 30 days. This was published in the American Journal of Clinical Nutrition in 1996. And, what I show you here is -- first of all, I'll show you the glucose. Before and after 30 days, you'll see here the dotted line is the low glycemic index, the solid line is the high glycemic index. This is a 20-hour test. After 30 days on a high glycemic and a low glycemic index diet, that was carefully monitored, and you can see here the high glycemic index, the low glycemic index, and this is the average standard area, the mean, and there's absolutely no significant difference, no difference after 30 days on a high glycemic index diet and a low glycemic index diet.
Here is the results of the insulin. Again, the solid is the high glycemic index. The dotted is the low glycemic index. No difference across here in terms of the insulin values before and after, and you'll notice importantly that the fasting values are not different, either. There's been some talk about insulin resistance with a high glycemic index diet. If you expect a lean normal individuals in insulin resistant state, after 30 days, you would expect a higher fasting insulin level responding to the insulin resistant state.
Now, these individuals, investigators not only did glucose and insulin, but they did hyperglycemic, glycemic hyperinsulinanemic clamps on these individuals. And, so, before the end of 30 days, at the end of 30 days, they did clamps. You notice here that the glucose uptake, this M stands for the glucose uptake, the whole body glucose uptake on a clamp in these young men were equivalent. In other words, there was no increased insulin resistance at a physiological level of insulin, 370 pica amounts per liter, between the high glycemic index and the low glycemic index individuals.
The glucose was not different, the insulin was not different. There was some increased triglycerides in the high glycemic index, and there were lower free fatty acids in this group.
Now, there are other studies that are somewhat similar to these, but they're not as focused and not as long, that tend to suggest that over time, if there is a temporary increase in insulin, short-term, in a one-meal set up over a longer time, this is not borne out.
Now, I want to go to epidemiological studies from here and review the two studies that I think have been seminole in getting people to think that maybe high glycemic index is detrimental to health, leads to diabetes and we should do something about it in our population. And, they're the two studies by Salmeron and the epidemiological group at Harvard. And, I'd like to go through them a little carefully, because I think that they're important to your deliberations.
Now, the first study was the Salmeron study that was published in the JAMA in 1997, Volume 277, page 472, and it reported on 65,173 U.S. women. Now, these are not ordinary women. They are health professionals, okay, so they're not representative of the population as a whole. They are nurses, Ph.D.'s, medical workers of one kind or another.
They were followed for six years. They were given a semi-quantitative food frequency questionnaire, which consisted of 131 items. Over a period of six years, until the time of this report, 915 of them became diabetic. That is, 1.5 percent of the sample, or .25 percent per year became diabetic.
The second report was in diabetes care. Again, 1997 -- that's not a seven, that's a nine -- and 42,759 men, six years, 523 diabetic patients over the six years, 1.2 percent conversion rate, .2 percent per year, and again, semi-quantitative food frequency questionnaire.
Now, let me just describe for you this questionnaire. The questionnaire was a questionnaire that, as it says, had 131 items. It was mailed to the individuals. The individuals responded on their own. They put down what they had eaten on a -- and they also had to have some element, I assume, of the portion sizes that they were eating. There was no direct interview between the person who was collecting the data and the individuals who were filling out the questionnaire. Okay, they did it at home. It was mailed to them and they mailed it back.
Now, in the text of the original article, it says, "A full description of the food frequency questionnaire in its abbreviated form, 61 items, and of the procedures for calculation of nutrient intake, as well as data on reproducibility and validity in this cohort have been previously reported."
Okay, they reported that validity, a questionnaire that had 61 items -- this questionnaire had 131 items. The second thing is that they accepted questionnaires that had less than 60 blanks. In other words, you could have -- they have to fill at least 70 blanks out of 131. So, essentially, you could fill out half the questionnaire and be considered part of this study.
Okay, my problem with this kind of questionnaire is not that this isn't worth doing, but that it's a very inaccurate way of measuring food intake in individuals, okay. You're asking people to do this independently, you're giving them a questionnaire, they're filling it out at home. You have no real idea about their ability to measure portion sizes and you're going to calculate total carbohydrates and then total individual items in that carbohydrate meal. And, so, you're depending on them being extraordinarily good reporters of what they're eating.
Now, we know from other data unrelated to this article that getting accurate dietary information out of people is extraordinarily difficult. It gets more difficult as people get fatter and we know that Americans are getting fatter. And, that the inaccuracy as BMI increases, so that I personally don't have a lot of faith in the very small numbers that we're dealing with, as you'll see when we get to the conclusions from these reports.
Now, to start off with, I just want you to note some things that came out of this study, these two studies. First of all, sugar is never mentioned in either of the studies. Now, we're talking about two big articles on carbohydrates and sugar is never mentioned, okay. It's not mentioned because there was no significant effect of sugar.
Total carbohydrate, there was no significant effect. When you get to the glycemic index score and the glycemic load score, which I will come back to, the glycemic index score was significant in women but it was not significant in the men. The glycemic load score was not significant in either the men or the women and it only became significant after adjustment for cereal fiber intake, okay. In other words, they did not get an effect from the glycemic index, then they added in a second component and said, well, what if we add cereal fiber to that, and when they added cereal fiber to that, they got an effect.
So, we've got three non-significants and only one significant. Now, let me tell you how they calculated this, because I think it's important for you to understand this. They used what is called the average dietary glycemic index, which comes from Wolever and his group, Jenkins, the group in Toronto. It's published in the American Journal of Clinical Nutrition in 1994. And, the way this calculation works is, you take the carbohydrate content of each food from the USDA Handbook, you multiply by the number of servings per day and then you multiply by the glycemic index. Then you take the whole thing and you divide it by the total carbohydrate intake for that day.
So, you have the carbohydrate content times the servings per day times the glycemic index over the total carbohydrate, and you get the average dietary glycemic index.
Now, this is a USDA meeting and I don't want to be disrespectful, but we recently put a bagel in a bomb calorimeter and compared what we got to what is in the USDA Handbook, and our bagel was two and a half times, had two and a half times as much calories as was in the USDA Handbook. Now, I'm not saying this to criticize --
DR. GRUNDY: Was that a New York bagel?
DR. PI-SUYNER: That's a New York bagel.
(Laughter.)
DR. GRUNDY: Well, that's not fair.
DR. PI-SUYNER: Now, I'm not saying this to criticize the USDA. I'm saying it that we're talking about a rather inaccurate calculation. You take it from the USDA Handbook and there's a lot of variety between foods, variety in geographical areas, variety between countries, etc.
Then you take the number of servings per day of that particular food, which is dependent on the recall of the individual that you've sent the questionnaire to, and then you multiply it by another inaccurate number, the glycemic index, which is also variable from test to test.
So, you know, white bread is not always 100 when you test it. Sometimes it's 110 and sometimes it's 90. So, you've got a variation around the glycemic index. So, we're essentially multiplying four, three very inaccurate determinations and dividing it by a fourth inaccurate determination, and we come up with a number.
The second thing that they used besides the glycemic index is the global dietary glycemic load, and that's the carbohydrate content times the servings per day times the gastric, the glycemic index, without dividing it by the total carbohydrate. And, that, they used as they state, as a measure of insulin demand.
So, again, quoting from their paper, "The glycemic index as a relative measure of glycemic response to a given amount of carbohydrate, does represent the quality of carbohydrate, but does not take into account the quantity." And, then, "In contrast, the total glycemic load represents the combination of quality as well as quantity of carbohydrates consumed, and may be interpreted as a measure of insulin demand."
So, what they're saying is that this is quality, this is quantity and this may be more related to the insulin demand, because it gives the total amount, the total load, you might say.
Now, I want to show you their results. We take first the men, and this is the glycemic index and this is the glycemic load, divided by quintals. Okay, this is the relative risk, using the lowest quintal as one, and you can see these are 95 percent confidence intervals. You can see here that the confidence interval here is .86 to 1.5, .85 to 1.5, .86 to 1.5, .96 to 1.62, P value not significant at .12.
Glycemic load, .8 to 1.36, .73 to 1.28, .75 to 1.36, .77 to 1.43, P value .83 not significant for trend. Okay, so we got nothing in the men. Here's the women.
Glycemic index, we get a significant value of .04. You can see the confidence intervals, though, .92 to 1.45, 103 to 161, 102 to 159, 125 to 154, very close to non-significance. And, the glycemic load is non-significant at .09. Okay, so the glycemic load which they, themselves, quote as being the important item for insulin demand is not significant in both men and women.
Now, I'd like to go on now and talk a little bit about the practicalities of telling a population that they should not have a high glycemic foods and they should have low glycemic foods. My first item here is foods in the questionnaire contributing most to carbohydrate variation in our cohort. This is, again, from the Salmeron study. Okay, these are the foods that they list as being most important contributing to carbohydrate variation. Cooked potatoes, french fried potatoes, cola beverages, jams, pasta, white bread, English muffins, white rice, cold breakfast cereals.
Now, I already showed you that cold breakfast cereals have a very wide variation, okay, in glycemic index. They can go from the 60s all the way to the 120s, 110s. We know that rice has wide variation, according to the kind of rice, the way it's processed, how much amylose is in it, etc. We know that potatoes, there's a big difference how they're cooked and also whether they're mashed or not or whole, and I'm just interested in knowing whether the recommendation of the dietary guidelines would tell people not to eat potatoes and not to eat rice, certainly, and possibly not to eat breads when these are staples of a carbohydrate intake.
And, please remember that it's not just white bread, but also whole wheat bread has a high glycemic index. There's also another complication for populations trying to figure out a high glycemic index and that is that many diverse common foods have similar GI values. Now, here's a whole group of foods that I've picked out for you have a high value range, 94 to 106. You know, they don't have a lot of relationship -- Melba toast, bagels, white bread, 100 percent whole wheat bread, angel food cake, graham crackers, whole wheat crackers, cous cous, corn chips, oatmeal muffins, french fries, mashed potatoes, canned green pea soup, breakfast cereals, such as Cream of Wheat, Cheerios, Golden Grahams.
So, as you tell the population to change to a lower glycemic index diet, you're going to have a lot of confusion about the food items that are included or not included. Another way to put it here is, I've put on your left high GI foods and the right, low GI foods. Bread, you'd have to tell them that they have to take bread that contains a lot of whole grains. Processed breakfast cereals versus unrefined cereal, oats, muesli or porridge. Also, some processed materials which have a low GI factor that have been put in, such as Kellogg's All-Bran, that has extra fiber. Plain cookies or crackers, you'd have to tell them to have cookies with dried fruit and whole grains. Same with cakes and muffins.
Tropical fruits such as bananas are high glycemic index, so are you going to cut out bananas? And, temperature climate fruits are better, namely things like apples. Potatoes are high and pasta is lower, but are you going to cut out potatoes? Rice is generally high, but there are kinds of rice like Batsmati rice and other high amylose rice which are low, and so, are you going to define for people the kind of rice that they're going to eat?
Now, the other thing that happens is the GI values of some foods can vary widely. And, that, you notice you can take the same food and the glycemic index can vary, depending on the variety, the processing and the preparation.
As I mentioned, there are many varieties of rice with different types of starch, processed in different ways, that result in a very variable value of glycemic index. As an example, the glycemic index of one-inch cubes of boiled potato can be increased 25 percent by mashing them. Just by mashing these one-inch cubes, you can increase the glycemic index by 25 percent. So, are you going to tell people not to mash their potatoes?
And, there are subtle differences in some fruit. For instance, in bananas, according to the ripeness, you can double the glycemic index of a banana, just by how ripe that banana is or is not. So, are you going to tell the population how ripe of a banana they should eat in their dietary guidelines?
I'm pointing this out to just remind you how difficult this is going to be for the individual citizens who you're giving the recommendations to. Now, I put this on as an exercise of what's going on with the glycemic index, okay? This is taken from a paper by a very respected food technologist, Chatrevette, and he tested different starches in India. And, he writes at the end of his paper, "The glycemic index of grain amaranth, wheat and rice preparation was studied in non-insulin dependent diabetic subjects. Diets containing 50 grams carbohydrate equivalent were given and post-prandial blood glucose estimated at different intervals of glycemic index, calculated for different experimental diets, showed that the glycemic index of amaranth wheat composite flour, 25 to 75, was the least, 65.5 percent, followed by the wheat diet, 65.7 percent, the rice diet, 69.2 percent, the amaranth wheat diet, 50/50, 75.5 percent and popped amaranth in milk, 97.3 percent. Therefore, a 25/75 combination of amaranth and wheat, wheat and rice, can be considered low GI food, 50/50 grain amaranths in wheat medium GI food, and popped amaranth in milk combination, high GI food." Now, tell that to your citizens.
What I'm saying is, first of all, this fellow did not find that the rice that they're using is any worse than the wheat. In fact, you can see here, he got rice 69.2 and the wheat, 65.7. And, in the 50/25/75/65.5. So, when you get into the question of how individuals are going to construct their diet and you try to tell them what is high glycemic index and what is low glycemic index and what starches they can use and what starches they can't, you're going to have problems.
Now, to bolster that argument, I'm going to give you a quote from Wolever himself, who is the champion of high glycemic index diet -- low glycemic index diet. "Being on a low glycemic index diet does not require elimination of all high glycemic index foods. Indeed, there are situations where high glycemic index foods may be appropriate or even desirable." So, you're going to tell your citizens, don't have high glycemic index foods, but if you want some, it's okay and may even be desirable for you to have some.
Now, I quote you another quote from Wolever, because it makes the point about short-term studies. He says in one of his articles, "Glycemic responses are quite variable from day to day within subjects." Okay, this is the champion of low glycemic index foods. "Glycemic responses are quite variable from day to day within subjects. Thus, when the number of subjects studied is not large and the expected difference is small, there is a large possibility of not detecting a difference that really exists, Type II error." Okay, these differences are so small that he's worried that half the time you're not going to be able to pick them out, okay, you're not going to find a difference.
He goes on to say in the same article, "If a significant difference is not observed, the responses are considered to be the same. This is an inappropriate approach because the fact that no difference is detected does not necessarily mean that no difference really exists." Okay, so he's saying, even if you don't get a difference, there's a difference there. You just aren't picking it up.
Well, if the difference is so small that in repeated tests, it can't be picked up, it can't be made, it's not even statistically significant, not to say clinically significant, then I think we have a problem recommending that to the American people.
So, I want to finish up with this quote, which is from Gerry Reaven. "The results of epidemiological studies provide hypothesis to be tested, not definitive answers to biological questions." And, I think that's where we are today. We have hypotheses to be tested, but we don't have definitive answers to this biological question.
I can tell you that the Diabetes Association has been struggling with this question for years now and has never come up with the idea that they should recommend a particular glycemic index for diabetic individuals, and the reason they have not is two-fold. Number one, the long-term data is essentially non-existent, and the difficulties of providing a clear message to the diabetic individuals about what he or she cannot eat and the caution, the worry that we would be eliminating foods from the diet that are absolutely good foods, for no particular reasons.
For instance, you know, take the minority population, the Hispanic population in America, which eats a lot of rice, okay. Are we going to take away rice from the Hispanic population from America and tell them they all have to switch to potatoes with the kind of data we've got? I think it would be an enormous mistake to do that. Maybe in the future, we would have more data to base the public health recommendation of this sort on, but I think at the present time, given the data that we have, which I think is considerably flawed, I would not recommend that you recommend that the glycemic index suggestion on the dietary guidelines.
I think there already are in the guidelines recommendations for having a high fiber diet. I have to tell you that in the Salmeron study, the only fiber that was significant in terms of the diabetes was cereal fiber. Soluble fiber was not significant, made no difference. The difference in the women, between the high cereal fiber and the low cereal fiber was something, I think it was 2.4 grams a day. So, how an individual is going to switch by 2.4 grams a day, I'm not sure. But, I think giving a recommendation of high fiber would be a good thing to do, maybe more for the GI effects than for the effect it might have on diabetes and lipids.
So, with that, I'll stop. I think I've gone over. I'll be happy to answer any questions you may have.
DR. GARZA: Okay, thank you very much. Are there any questions? Roland?
DR. WEINSIER I think I know the answer, but just to have it on record, unless there's a third choice you can think of, if you had the choice as a recommendation to us, would you suggest we address the issue, but downplay, we don't address the issue, ask for more research or await more research?
DR. PI-SUYNER: I would not address the issue. I think the average citizen is not going to know what you're talking about if you talk about high glycemic index. And, unless you actually are going out to make a recommendation, I don't see any point in doing it.
You will have addressed the issue of fiber, I assume, and that takes care of a good bit of that, because the higher fiber foods, generally, are lower glycemic index. So, you already, in effect, have a recommendation for a lower glycemic index diet, by putting in a fiber recommendation. Although, as I mentioned to you, in the studies, the Salmeron studies, dietary fiber, soluble fiber was not significantly affected. Yes?
DR. DWYER: Xavier, in very, very high carbohydrate diets, take very high rice diets, where it's a staple and people are eating a lot, like 55, 60 percent, just from one staple, refined cereal, what is your view on utility of these concepts there?
DR. PI-SUYNER: Well, see, I think what's happening is people are confusing glycemic index with excess calories. I mean, people have been taking that kind of diet in Japan for centuries without any problem, you know, no high incidence of heart disease or diabetes. All right, so it's not a question of having a very high carbohydrate diet, it's a question of having a lot of carbohydrate calories.
All right, what you're doing is, you're overeating. Now, you can take the same glycemic index food and if instead of taking 75 grams of it, you take 150 grams of it, you're going to get a much greater insulin response. Same glycemic index food. Okay, you'll get triple or four times the response of insulin if you take double -- instead of 75 grams, you take 150 grams.
So, it's not a question so much of taking carbohydrates, it's a question of taking too much of everything, including carbohydrate. So, I think the message to the population would be better being to limit your calories than to limit your carbohydrates. And, certainly, within the carbohydrates, as I mentioned, total carbohydrate was not significant in the Salmeron study. Sugar was not significant. We're left with a small portion of the whole, the total carbohydrate intake.
DR. GARZA: Scott?
DR. GRUNDY: Two questions. Would this, would you extend this to sugars, as well, on the glycemic index? Would you think it should be a recommendation?
DR. PI-SUYNER: Well, you know, sugar is a low glycemic index food, it's not a high glycemic index food, because it has fructose, that's 50 percent of its carbohydrates. So, when you test it, sucrose is lower than glucose by quite a bit.
DR. GRUNDY: Is there any reason to moderate sugar intake, then, on the glycemic basis?
DR. PI-SUYNER: Not on the glycemic basis. There's other reasons, which you've heard about, you know. So-called empty calories and the fact that you're getting no vitamins and minerals when you're eating the sugar and so forth. So, personally, I think 18 percent sugar intake is very high in this country and it contributes very significantly to the caloric load that we're eating. So, I personally thing for other reasons, cutting back on sugar is probably a good thing, but not from a glycemic index approach, since it's a low glycemic index food.
DR. GRUNDY: Okay, second question was, you through up that one hypothesis about the insulin exhaustion theory from overstimulation, whether it be from excess calories or whatever it's from. Do you think that's a valid hypothesis, or what's your view on that?
DR. PI-SUYNER: Well, I think insulin demand as a stimulus to the bringing on of diabetes is a perfectly valid hypothesis. I think you have to have the right genes, you know. We have obese people who are 500 pounds that have been 500 pounds for ten years, that are non-diabetic, because they don't have the diabetic gene, so you have to have gotten the appropriate gene that will cause a defect in the beta cell. The beta cell will eventually exhaust.
But, if you have those genes, yes, I think the insulin demand theory is a reasonable theory, but we don't have any long-term testing of it as a hypothesis. And, the beta cell is actually, as we see from these very obese individuals, is able to manufacture lots of insulin over time. So, generally, I think it has to be an interaction between the environment and the genetic predisposition to get this.
The question is, how much insulin demand is enough to cause the problem? And, you know, when you're talking about a different -- I have to tell you, I was trying to find the -- the difference between the first quintal and the last quintal in that Salmeron study, I think the first quintal was a glycemic index of 64 and the last quintal was 71, okay. So, I have that in -- I think it was seven points, 64 to 71.
Now, to get to that kind of shaving, of moving somebody from 71 to 64, you know.
DR. GARZA: Yes, Shiriki?
DR. KUMANYIKA: Does your comment about over-consumption of foods at one occasion have implications for meal size recommendations? I mean, because we will be getting a weight guideline, and what you just said could be interpreted as eating a large serving of, say, a high glycemic index food could have the effects that we heard about yesterday, for example.
DR. PI-SUYNER: Yes, I think portion sizes and meal sizes are one of the big problems in America. I think people have gotten used to extraordinarily portions and the common now is something that is very uncommon in much of the rest of the world. So, I think some alert to the fact that it's not so much what kind of food that you eat, but the portions overall, if you could cut them back, you would cut back on your calories.
I still think that the primary problem is calories in the United States, and clearly, you can have a very low glycemic index diet, and if you take a lot of it, you're still going to get a big insulin demand.
DR. GARZA: Rachel?
DR. JOHNSON: Thank you very much. You've mentioned what you just said about the total calories and the fact that you don't think we should recommend limiting carbohydrate. I wondered if you'd like to weigh in on the total fat message and total calories, and it's relationship to obesity, and whether you think it's prudent to keep a message to limit total fat to 30 percent, or whether you think total calories is really the ultimate message that we should try to get across?
DR. PI-SUYNER: Well, I think total calories is the ultimate message, but I really -- I feel very comfortable with the 30 percent fat recommendation. I think zigzagging around all the time is not good for us and it's not good for the American public. And, I think the NCP guidelines have been widely distributed around the country. I think they've been accepted quite well.
Proportional fat has gone down, but I guess total fat has stayed up, so we haven't made much of a dent in the total fat. But, and so therefore, I think we still need to do that. By far, the biggest killer for metabolic disease is cardiovascular disease, heart disease, and clearly, there's a relationship between total fat and saturated fat, at least with the way I read the literature and coronary heart disease. So, I have no problem with the NCP guidelines for fat.
But, I would also strongly recommend caloric, you know.
DR. JOHNSON: Thank you.
DR. GARZA: Johanna?
DR. DWYER: Just two other questions. One is, do you think we should talk about beverages and foods high in added sugar, and could you also elaborate a little bit on this glucose toxicity here?
DR. PI-SUYNER: Yes, I would say something about beverages, because I think they're growing at an enormous rate and it really is something that is not necessary for the American public to eat at that very large amount, and it brings, in a sense, truly empty calories into play in the American citizenry.
So, I think some cautionary comment about sugar-containing beverages would be a reasonable thing, from a caloric point of view.
With regard to glucose toxicity, glucose toxicity is something that occurs in diabetic individuals. It doesn't occur in normal individuals. Your normal population will not have anything, even if they take a high glycemic index diet, glucose toxicity will not come into play. Glucose toxicity is when the glucose level gets so high that it affects the ability of the beta cell to secrete insulin, and the beta cell gradually decreases its response. It's as if it gets tired from this continual stimulus at a very high glucose level.
So, if you get people who have glucoses above 250 milligrams percent continually over time, you get a glucose toxicity effect and the beta cell does not respond appropriately, and your insulin secretory response goes down.
But, that is strictly in diabetic individuals who are out of control, and has nothing to do with normal metabolism and glucose in the normal population.
DR. GARZA: I have one or two questions. If one looks at the U.S. population and the high prevalence of obesity, do we have enough data to say anything about the role of high glycemic index foods and the metabolic responses among the obese versus the lean, or from your read of the literature, there are no differences in the metabolic responses between obese and lean individuals?
DR. PI-SUYNER: No, what happens as people get overweight and very obese, the metabolic response which I would translate you to say as the insulin response, goes up four or five-fold in an obese individual anyway. And, so, then, the differences between a high glycemic index food and a low glycemic index food become less, because your total amount of insulin coming out is much higher.
It's much easier to get differences in glycemic index foods in lean individuals than it is in obese individuals, because you're starting at a much higher threshold with a much greater response to the meal.
DR. GARZA: So, the concern that we've heard expressed, that individuals who choose to lower their fat intake, either because of the desire to lower their risk to cardiovascular disease, or because they use it as a strategy to lose weight, replace those calories to some degree with a high-carbohydrate diet, and that in employing that strategy, we put them at risk because of the supposed metabolic responses to glycemic index foods. In those scenarios, you would then say for the same reasons you just gave us, that that's not a solid view?
DR. PI-SUYNER: No, I don't think that's true. First of all, and there are a couple of things, one is that, you know, I've shown you the data of glycemic index taken a single meal -- just the food. You take pasta and you boil it and you tell somebody to eat 50, 75 grams of pasta and you measure the glucose. Nothing on the pasta, that's it.
As soon as you go to mixed meals, you change the glycemic index, okay, because of many factors. One is, the fat slows down, in the meal, slows down the carbohydrate transfer. The carbohydrate transfer is much faster when it's taken alone, out of the stomach, goes out of the stomach into the duodenum and is absorbed much faster than when fat is mixed with it and when protein is mixed with it.
So, the glycemic index differences drop when you get a mixed meal. So, right away, you have a difference between a single, pure carbohydrate you're testing and the carbohydrate as you're eating it, which is mixed with vegetables, fat, protein, etc. And, the differences shrink when you do that.
I think that the trouble with people continuing to gain weight on a low-fat diet is that if they only restrict, if they substitute the fat, carbohydrate for the fat, then they will take the same calories and continue to gain weight. I think the message is not to substitute. It's not to say you can eat as much fat as you want. It's to say, don't substitute for the fat you've cut out with extra carbohydrates. Because, I think it's not the carbohydrates that are driving this, I think it's the calories that are driving it. And, what people have done is, they've switched from fat to carbohydrates. And, so, they're eating the same number of calories and so they're continuing to gain weight.
DR. GARZA: Is there any data that suggests that certain ethnic groups in the U.S. that are predisposed to Type II diabetes because of genetic predisposition are at any higher risk from diet that have a high glycemic index?
DR. PI-SUYNER: I don't think so. I don' think there's any data. Clearly, the population that is most at risk, I would say, is the Hispanic population, the Mexican-American population, which has a very high incidence of diabetes and growing rapidly in Mexico itself and in Texas. We don't have much data from California. Most of the data that we have is from Texas, but I assume it's the same in California.
We have very little data from Caribbean Hispanics. Very little data from Puerto Ricans, Dominicans, etc. Most of the data is from Mexican-Americans from Texas, and clearly, they have a high risk for diabetes and clearly, their diabetes incidence is going up.
Now, what has changed in their lifestyle? Well, all kinds of things have changed. You know, they're much more sedentary than they were. They eat much higher caloric intake and that caloric intake, the extra calories that they've added to their traditional food are generally fat calories, okay. In Mexico, traditionally, they ate rice and beans, okay. Now, they've added cheese, they've added sour cream to their -- they've added, they fry the tacos. They used to just steam the tacos. So, the changes have been not so much adding extra carbohydrate as adding extra fat to this diet, which added extra calories, which increased their BMI, which increased their risk for diabetes and cardiovascular disease.
So, I think it's the calorie effect rather than a carbohydrate effect.
DR. GARZA: Any other questions? Johanna?
DR. DWYER: Would you continue the guidelines such as the one in the '95 guidelines, adding beverages and foods? Add more emphasis to added sugars or get rid of the whole thing, or do something else, if you were on this committee?
DR. PI-SUYNER: You mean, to give a recommendation on the beverages?
DR. DWYER: One of the dietary guidelines last time involved sugars and I wondered how you would change it, expand it --
DR. PI-SUYNER: Well, I guess I don't recollect it exactly. I wouldn't want to say how I would change it. If you could read it to me, I'll tell you what I think.
DR. JOHNSON: Choose a diet moderate in sugars.
DR. PI-SUYNER: Pardon?
DR. JOHNSON: Choose a diet moderate in sugars.
DR. PI-SUYNER: I think that's fine. I wouldn't be, I think that's very acceptable. I think, as I mentioned, I believe that that's a way Americans are getting calories that they don't really need, and it's sort of empty calories and it's perfectly reasonable to suggest that they be moderating how much they take of that.
DR. GARZA: Meir?
DR. STAMPFER: I think it's easy to see that the glycemic index evokes not only insulin, but almost sort of a religious zeal. I'm not going to take time for a point by point discussion of the Salmeron papers, but I would encourage the committee to have a look at them and draw their own conclusions.
Just to make a couple of comments, though, of course diet is hard to measure and if you're going to study chronic disease, you have to measure it in a large number of people, and that means measuring it with a certain amount of error. But, in a prospective study, that kind of misclassification generally leads to an underestimate of the effect.
My second comment relates just to the very part of the talk, where the implication was made that as societies go from a traditional high carbohydrate to a more high-fat diet that they get more diabetes. That is certainly true, but clearly, this can be mostly, if not entirely, attributed to obesity and physical activity. In the traditional society of Crete, for example, with about 48 percent of calories from fat, also had very, very low prevalence of diabetes when they were active and lean.
DR. GARZA: Richard?
DR. DECKELBAUM: Xav, can you comment on the potential differences in complex versus simple carbohydrates on satiety?
DR. PI-SUYNER: I can't comment on it, because there's no real data that I know of, good data, on the difference. Generally, you know, the thought has gone that simple carbohydrates are more rapidly absorbed. However, some of the simple carbohydrates that humans eat are not sucrose or glucose. A lot of them, they eat quite a lot of fructose, too, which doesn't raise the glucose very much, and certainly almost doesn't raise the insulin at all.
So, there have been specific studies comparing fructose and glucose and sucrose, showing that glucose gives you the highest insulin, sucrose the next and fructose the least, as you would expect.
Lactose is pretty close to glucose. I don't know of any, I don't remember recalling studies with lactose, itself. That's not a big player. If you go head to head with complex carbohydrates, taking complex carbohydrates as a starch, starch is a series of glucose molecules and has almost identical glycemic index to sugar if it's cooked. So, the glucose molecules are really hydrolysed very rapidly and absorbed just as rapidly as glucose alone.
Flour that is not cooked, of course, takes much longer, but who eats uncooked starch? Not many people. The rest of the complex carbohydrates, probably the slowing down of the glucose absorption is more related to the fiber content, although this has always been a difficult issue than it is to any inherent characteristic of the carbohydrate itself.
DR. GARZA: Xavier, could you comment on the point that Meir just raised of the differential effects of biases in methodologies versus noise in the methodology? Most of the points that you made in terms of your review of the paper spoke to noise, and generally when there is noise in data, then relationships of the type that you looked at would obviously be attenuated by noise. If there are biases in the data or the methodology, then obviously, one can then look at spurious relationships.
DR. PI-SUYNER: No, I don't think -- I wouldn't call it bias and I wouldn't call it religiosity. I would call it looking at the data and seeing what you see. I think there is a lot of noise in the data. I think it's a very imperfect measure, glycemic index, it's very variable. As I said, it will change even with a given food, according to the ripeness of the food, the way it's processed, the way it's cooked, etc. So, I think there's a lot of noise and it's how you interpret the noise, I would guess, would be the answer.
And, the second is that there are very few long-term studies. I mean, I think that's the big problem. Most of these studies are one-meal studies, and I don't think you can recommend something for a population on single-meal studies. I think you've got to go to longer-term kinds of trials.
DR. GARZA: In the six-month study that you reviewed, the European paper, were those done in lean individuals or obese individuals?
DR. PI-SUYNER: They were done in lean individuals.
DR. GARZA: Scott?
DR. GRUNDY: So, in, sort of in summarizing what you say, is that from a metabolic point of view, all carbohydrate is the same for practical purposes? You can dissect out little tiny things with different foods and so forth, but taken as a whole, there's not a lot of difference from a metabolic point of view from different kinds of carbohydrate, is that what you're saying?
DR. PI-SUYNER: I think you've hit it just right. There's not a lot of difference. There is some difference. There's not a lot of difference. I think all the arguments have come around that there's not a lot of difference.
As I mentioned, that trial, the Salmeron trial, the difference between the lowest quintal and the highest was like nine index points. So, I think the problem comes in the confusion of how you're going to change what you're going to recommend and not recommend, what foods you're going to call good foods and bad foods, given the variety and the variability between these foods.
I would be happier if your committee would deal with calories and portion sizes and fiber and not deal with glycemic index. I think you'll hit a lot of it with fiber.
DR. GRUNDY: There used to be a lot of discussion about complex versus simple carbohydrate and their different metabolic effects, but you kind of downplay that, too?
DR. PI-SUYNER: Yes, well, starch, that was, years ago, we used to tell diabetic patients not to eat sugar, they should eat complex carbohydrates. That's where the term came from.
DR. GRUNDY: Right.
DR. PI-SUYNER: The recommendations to diabetic patients. And, then, some people actually did some experiments and found out that, you know, that starch could be absorbed just as rapidly as sugar, so the recommendation changed. When you cook starch, glucose absorption is as rapid as with glucose.
DR. GARZA: Alice will have the last question.
DR. LICHTENSTEIN: Isn't there also another factor with sugar versus starch, that now a lot of the things we traditionally thought of as high sucrose or glucose are now fructose, so the high-fructose corn syrup is being used instead of corn syrup which was high-glucose? So, there's been a shift in the available simple carbohydrate in the diet?
DR. PI-SUYNER: Well, that's true. In the marketplace, there are now more high-fructose foods than there were ten or 20 years ago, as people have shifted to using fructose in syrups and jams and jellies and so forth.
DR. GRUNDY: And, fructose is not a bad thing?
DR. PI-SUYNER: Well, fructose, you know, you get into a whole -- if you feed rats fructose long enough, they become insulin resistant. And, Reaven showed that in a series of papers.
So, if you eat a lot of fructose, it probably is not good for you, but small amounts of fructose, as people would generally eat -- I mean, you're not going to eat tons of syrup, in terms of a glycemic index and insulin demand, is much lower than sucrose or glucose, because fructose goes right to the liver and it doesn't ever get out into the general circulation most of it. So, there's not a stimulus for the beta cell to respond.
DR. GARZA: Okay, thank you very much. We're going to move onto a discussion of the sugar guideline, because Dr. Johnson has a plane to catch, because of some important events in the family.
DR. JOHNSON: Okay, thanks very much. What I had planned to do was to go over each of the public comments and give response, and then raise three or four issues that the working group would like the total committee to reflect on.
Thanks very much to the staff for providing this list, and I believe I addressed everyone of those in my preparation for today, except for the last one. So, I'll put that up and then we'll go through it. But, what we have first is the response to the public comments and I will start out with the first comment, which is to strengthen the advice.
There was some testimony that we should define moderate, define the word moderate as it applies to sugars. And, the only definition that I could find currently is that in some of the food guide pyramid materials, they do state that added sugars should balance calories, but not exceed current recommendations. And, I have been told by USDA staff that these are not meant to be recommendations. What they are is that when they created the pyramid and they met all of the requirements from the various food groups, this was what was left over in terms of sugar calories.
But, that is, you know, one thing that's out there is this six, 12 and 18 teaspoons, depending on the different calorie levels. These are substantially lower than current intake, and I think we do describe current intake in the document that you have. But, those teaspoon levels are substantially lower than current intake.
Another recommendation was to change the wording and choose a diet low in added sugars and the work group had recommended focusing the guidelines on added sugars.
DR. GARZA: Rachel, before you take that off, could you give us a sense of what six teaspoons would be equal to in terms of a soft drink or jam?
DR. JOHNSON: How many teaspoons of sugar in a soft drink, 12? Thirteen? So, it would be half. I believe there's about 13 in a 12-ounce soda, so that would be half a soda. That would be six ounces of soda, so it is quite low. Oh, I'm sorry, the last one, no, I did that one.
Okay, the next one on the list was added versus naturally occurring sugars and there was testimony that we should distinguish between foods with added and naturally occurring sugar, and I would say that the work group concurs with this and has done that in the draft document.
The next on the list is consumer perceptions. I think we've had three or four consumer focus group studies that have been presented to us. The first that we've already discussed at this meeting, or no, the first was that consumers believed sugar-containing foods in moderation are part of a healthy diet. I don't think that the guideline or any of the materials that have been developed to date contradict that.
The second was that consumers are confused by the message to moderate sugar intake but eat plenty of fruit, and we've already discussed that. That was from our early consumer focus group study that we had.
The next was a recent consumer focus group study we were provided, that said that consumers don't understand the added sugar term, and they used an example that somebody asked if that meant that you add sugar to a Coke? So, there does seem to be some consumer confusion about what we mean when we use the term "added sugar".
And, the last was some work that said there's no further benefit to consumers when a distinction between added and naturally occurring sugars is made. And, I would urge that the Consumer Dietary Guideline Study that we were provided with -- is that a Carole, of the questions?
MS. DAVIS: It is a draft.
DR. JOHNSON: It's a draft, so if we wanted to make some suggestions, I would think, at this meeting on the questions regarding added sugar --
MS. DAVIS: No, no, to be clear, I used the term draft because it hasn't been out of OB clearance.
DR. JOHNSON: Okay, because I did note that the terminology in that report is not what -- I can talk to you about it. It's not exactly what the committee had used for the suggested wording.
But, I would think that this consumer group study could focus on the best terminology to communicate the intent of the guidelines.
Okay, sugar and BMI. There was no association shown between sugar intake and BMI, using the CSFII database. Whenever you look at this, it's always very critical when you look at these studies on sugar and BMI, to notice whether they're talking about total sugar or added sugar, because as we know, total sugar includes lactose in milk and dairy products and fructose in fruits. So, you need to look at it critically in that way.
This one CSFII analysis was based on one-day dietary intake data, and I think it's important to remember that self-reported height and weights were used. Madelyn Sigmund Grant from the University of Nevada presented the committee with a good description of some of her work and she talked about the weaknesses of any comparison with the CSFII data, the one-day food intake data and BMI, because of the nature of the intra-individual variability you have in one-day intake, as well as the fact that the BMI was self-reported.
There is a paper by Rip Troiano that is impressed right now with AJCN, using the NHANES data, measured height and weight, and he found that soft drinks contributed a significantly higher portion of energy for overweight than for non-overweight children and adolescents. And, we were provided with that paper fairly early on in our deliberations, but I have a copy of it if anybody would like to see that specifically. So, there's some information on sugar and BMI.
The next was no separate sugar guidelines. The testimony said to focus on total diet, do not single out sugar and to emphasize physical activity. I believe that the work group is recommending retaining a sugar guideline and the justification is in the draft document, and we certainly are considering the addition of a physical activity guideline.
The next was to clarify terms. There was a request that we define sugar and currently, in the current draft, the work group is recommending the USDA definition that's used for added sugars in the CSFII database. So, that might clarify that a bit, in terms of providing a definition.
The next was glycemic index testimony. We had testimony to include the glycemic index and to not include the glycemic index, and the work group is recommending more research at this time to determine if diets high in sugar or glycemic index are linked with the etiology of non-insulin dependent diabetes mellitus, but the work group did not feel that they wanted to move forward with including glycemic index at this time.
Eating out, there was testimony to note the high levels of added sugars in restaurant food. And, although the work group did not single out restaurant foods, per se, we have included a list of foods high in added sugar to be added to the booklet.
Sugar substitutes, there was testimony that the sugar substitute asasolstane potassium be added to the list of sugar substitutes in the document. This sugar substitute does have FDA approval for use in the U.S. Currently, sorbitol, saccharin and aspartame are listed, and I have no idea what the consumption of that particular sugar substitute is. I thought that maybe we could get some data on whether it's a widely used sugar substitute or not, because I'm not sure. I looked in several textbooks that I had to try and get some information about it, but didn't come up, you know, quickly with an answer. So, I think we could do some background work on that.
Soft drinks. There was testimony about soft drinks regarding fluid intake, hydration and energy needs. The gist of the testimony was that children and adults need fluids. Children have been shown to drink more fluids when beverages are flavored, and that soft drinks can play a role in maintaining or improving weight, because the original '95 guideline said that we should maintain or improve body weight.
So, my question to the committee is, should the guideline mention fluid requirements and the role of soft drinks, both regular and low-calorie, in meeting fluid requirements? And, I was assured that we could get some data on fluid intake to see if dehydration is a problem in the population.
Okay, sugar and chronic diseases or obesity. The testimony was that there is no link between sugar and chronic disease and obesity. Our draft document does discuss some links with weight and carbohydrate-containing beverages. There's been some work by Barbara Rolls and Jim Hill, as well as Rick Mattes, showing that carbohydrate-containing beverages, that energy intake is less well regulated when the calories occur in beverages versus food. So, we have mentioned some of that.
The dental carries piece has not been changed from the '95 guidelines. In the report to the Secretaries, we did add something about fructose and colorectal cancer, primarily based on recommendations from the World Cancer Research Fund and the American Institute for Cancer Research saying that sucrose seems to play a role in colorectal cancer.
Coronary heart disease and NIDDM, the work group is recommending more research in this area and does not make any statements linking sugar with either heart disease or diabetes.
Added sugar, again, and coronary heart disease. The work group believes the evidence is inconclusive and is recommending further research on added sugar and hypertriglyceridemia and dyslipidemias.
Proposed wording, I had picked out, I'm sorry, the nutrient density is not on this list. When I went through all the testimony, I did pick up some testimony about nutrient density, so I thought it was important that we look at that. Some of the testimony said that added sugar displaces more nutrient-dense foods and other testimony said that sugars were not linked to nutrient density of the diet.
The draft documents cite research demonstrating that high intakes of soft drinks which are clearly the number one source of added sugars are negatively associated with intakes of milk and these are papers by Gunther, Harnack and Skinner, as well as being negatively associated with intakes of riboflavin, folate, Vitamin A and C, calcium and phosphorous, and this is in children who are high consumers of soft drinks. And, this was the paper by Harnack. And, that soft drinks were also positively associated with energy intakes, again, in children who were high consumers of soft drinks.
The proposed wording, i think I've discussed already and the last testimony, which I just received yesterday, was something about sucrose and calcium, high levels of sucrose leading to increased calcium excretion, and I can't really comment on that without doing my homework, unless there's somebody on the committee that may be more familiar with this literature than I am.
I think that pretty much sums up the public testimony. Some of the issues that the work group has raised, that I think would be useful to have some input from the whole committee on is the change in emphasis of the guideline. As Carole Suitor nicely pointed out for us, the focus of the draft document is now on moderation, using that term, "go easy on" beverages and foods that are high in added sugar. And, this really is a different, a shift from the '95 guideline that focused on a total diet moderate in sugar.
So, I think we need to talk about whether we want to focus on particular foods that are very high in added sugar and their effects within the diet, or do we want to continue this message of a total diet that's moderate in total sugars? And, I think it would be very useful to have some input on that.
In terms of consumer confusion, I want to go back to that, that consumers don't seem to understand the term, "added sugar," and whether that's a matter of education or whether it doesn't really matter. When we say sugars, I'm not sure consumers are really thinking about milk and fruits or other foods that are high in sugar. But, we should probably have some discussion about that.
This idea of whether or not foods with added sugar have a displacement effect on the diet, the published literature demonstrates this displacement effect in children and adolescents. I have not found published literature in adults that -- I'm not saying that the effect is not there. I'm just saying that the literature is not available. But, we have clearly seen it in children. Morton and Guthrie published a paper saying that increased energy intake in between the last two surveys, between '89 and '95, have been associated largely with carbohydrate intakes largely attained from soft drinks. Carbohydrate intakes have gone up, primarily from soft drinks. Micronutrient intakes have stayed the same, except for calcium, which has gone down.
Soft drinks have been shown to displace milk in children's diets. Again, three papers have shown that. And, high soft drink consumers having this lower riboflavin, folate, Vitamin A and C, calcium and phosphorous intakes and higher energy intakes. Again, all these data are in children or adolescents.
Again, this issue of sugar and BMI, the CSFII data showing no association, using self-reported height and weight. The NHANES data showing an association between soft drinks contributing a higher proportion of the energy in overweight than non-overweight children and adolescents, and then this work that I've already mentioned about energy intake regulation with carbohydrate-containing beverages.
And, then, lastly, we did address this a little bit yesterday, but in the draft document, it alluded to the fact that sugar might be in a second tier, and I just wanted to point out what has happened in terms of consumption, that core sweetener consumption has increased 16 percent from 1982 to 1996, going from 27 to 32 teaspoons per person per day during this time, so there is a clear increase in caloric sweetener consumption, that I thought it was important that we know about as we think about where to place these guidelines.
So, that is my review of the public comments and some of the queries from this work group and I would be happy to take comments, questions, discussion.
DR. GARZA: Any questions for Rachel? Yes?
DR. TINKER: Rachel, thank you. Do you know from the literature on the increased sources in the beverage that have the sugars, if that's mainly consumed outside the home or inside the home, and whether there needs or would benefit from a comment in the guidelines that direct people to where they may put their attention? Because if they keep thinking about in the home and going out to eat is something special or out to fast food is something special, and they just blow that off --
DR. JOHNSON: I can't answer specifically about soda. There is a good booklet that USDA did about eating away from home. We know an increasing number of calories are eaten outside of the home. I'd have to examine that to see if they single out soda as a food. I believe I've seen some data that said that meals at home are more likely to have other beverages than soda, but I'd have to check that. I know I've seen something. Good question, to emphasize the eating away from home.
Joanne?
JOANNE: Can I say something about eating out data?
DR. JOHNSON: Yes, thanks.
JOANNE: Isn't (inaudible) easier across time, multiple databases, but calcium density of most away from home food sources was significant in school meals, where milk is a required beverage, weren't much lower. So, I think you can draw a sort of obvious inference.
DR. JOHNSON: Thanks, Joanne.
JOANNE: I think there have been some smaller studies that did say what we said about women's propensity to consume milk versus other beverages outside the home.
DR. JOHNSON: Thanks. Joanne is the author of that paper so we know it's reliable.
DR. GARZA: During the confusion, the consumer confusion over the word added, what other terms have the work group considered, other than added sugars?
DR. JOHNSON: I can't say that we've really considered anything else, just whether to add the word "added," would you say, Alice?
We just learned of the confusion. I got that report in a Fed Ex package that I think we got on Tuesday, so we didn't really have time as a work group to address it.
DR. LICHTENSTEIN: When you're trying to ferret out the sugar that's consumed with meals versus out of the house, if it could be separated also out by school, because we saw some information at the last minute, suggesting that some school systems now were offering and even encouraging the consumption of soft drinks.
DR. JOHNSON: Well, in fact, there is a bill in the Senate right now introduced by my Senator Lahey and some other Senators. Apparently, there is a loophole in the law, because in elementary and middle schools, the machines that sell foods with minimal nutritional value, if you're participating in the USDA School Nutrition Program, can't be turned on until after the last lunch period, so there are some soft drink companies that are giving soda away free during the lunch, during the noon hour, because they can't sell it. I would imagine that it's done to develop brand loyalty on the part of children fairly early, and there has been a bill introduced in the Senate to close that loophole in the law that says they can't give away free soda at noon. But, I haven't heard anymore other than just that the bill was introduced.
DR. GARZA: Richard?
DR. DECKELBAUM: I would like some guidance on the role of focus groups in how we present our documents, both in terms of the booklet and in the green book. Because, if we're being asked to use evidence-based documentation for what we say, I can understand the focus groups are pretty important in whether perception of what we think is good wording translates into what the public thinks is good wording.
But, in terms of the science, because here we're debating a major term, "added sugar" versus "sugar," and, you know, maybe it's not the perfect word, but we sort of know what the concept is. What's the role, then, of the focus groups in guiding our exact wording and this scientific basis for putting down a recommendation in both the brochure and the green book?
DR. GARZA: I'm not sure what the status of the focus groups were. Carole, can you brief us?
MS. DAVIS: Well, what is happening is the focus groups moderators guide is in clearance now and we're expecting to have focus groups probably in July, and the results of that would not be available to you until sometime in August, and this would be the first round, where we're trying to get some information about concepts that just would guide you, to some extent, and then later on, when you have another draft, we will be testing sections of that, like maybe you think would be a problem, the weight guideline, parts of that. Part of whatever it's called, variety guideline.
And, this would be probably in September, almost. Some of these results are going to be helpful to us in putting the publication together. But, there's a lot of material out there already.
DR. GARZA: So, as a guide, but not necessarily anything beyond that.
DR. DWYER: Rachel, thank you for what I think was a very thorough presentation. I think there are a couple of other artificial sweeteners, suculose and so forth, that didn't get mentioned in the draft.
The energy density is mentioned, that whole argument is mentioned in the sugar one, but then it isn't mentioned in the fat one and I think that's a mistake. We need something on total fat and energy density.
And, finally, I'm totally confused by the different submissions. I think the first dietary guidelines meeting, we had a man present consumption and I didn't really follow it very well. Now, there's additional. What is the true increase one can, if you adjust for the definitional differences, what is the increase, because it seems to me that the size of that increase, if it can be determined, would be critical in making a decision between focusing on added sugars and total sugars?
DR. JOHNSON: The numbers that I took were from an ERS document. I believe the first author is Cantor, that I have with me and I can pull out, that looks at caloric sweeteners. There has been some debate about this definition of added sugar and how it's operationalized. I think clearly that when you look at consumption data, you look at disappearance data. You can, in that document I was referring to, it has some excellent charts showing what's happened to soft drink consumption, which, I believe, is contributing about a quarter of all the added sugar in U.S. diets. There's been a clear increase in consumption.
I think it's pretty hard to argue against an increase in consumption of total added sugars.
DR. DWYER: The question is how big?
DR. JOHNSON: Well, from that Cantor article I showed you the last slide, it was at 16 percent since '82.
DR. DWYER: Is that disappearance or is that consumption?
DR. JOHNSON: It's disappearance.
DR. DWYER: Well, I guess consumption would be better to look at.
DR. JOHNSON: We can get consumption of particular foods. I'm not sure we can get consumption increases using the USDA added sugar definition, because as the older surveys were not coded in that way, is that correct, Shanthy? It's only with the newest '94, '95, '96 survey that you have the ability to pull out added sugars, because that's the first time they used that definition. Otherwise, we'd have to look at total.
Oh, you did? So, we could look at the difference between '89 and '95? Okay, for added? Okay, that would be useful. We could do that.
DR. GARZA: Shiriki?
DR. KUMANYIKA: With this sugar guideline, I'm wondering if you can see a way to relate it to the other guidelines rather than or in addition to having it as a separate guideline, because the issues relate to weight control, they relate to replacement for fat intake, fruits and vegetables and so forth. So, I think also with the sodium guideline, these ones that have to do with qualitative differences within a particular food category and things that are added. Have you thought about the sugar guideline as phrased, mainly in terms of how it breaks out for the other guidelines?
DR. JOHNSON: I guess I have. I guess my concern with folding it in with the other guidelines is, it will lose some attention in and of itself, and when you think about the public policy implications of these guidelines and what the sugar guideline means in terms of the WIC program, the school nutrition programs, unless we're very clear and it's very, very clear within the context of some other guideline, I think it could have some really, really profound policy implications.
DR. KUMANYIKA: I didn't mean folding it in.
DR. JOHNSON: Oh, okay.
DR. KUMANYIKA: I meant expressing it. In other words --
DR. JOHNSON: Keeping it as a guideline, but --
DR. KUMANYIKA: You have these other guidelines, then you get to sugar and this tends to be written as a guideline to stand by itself when, in fact, what it's doing is giving you information about choices related to other categories, and I'm wondering if there's a way to make it clearer so that people can integrate the information better, rather than see this as yet another consideration, because they've already passed through sugar considerations in the other guidelines. It's not linear, but I think if you pass through the weight guidelines and you've thought about all your weight issues, that's when you should be thinking about the contribution of sugar and total calorie intake, not as a separate issue or in the same with fat.
DR. JOHNSON: So, I guess I'm not sure. Are you arguing, you're still
saying keep a sugar guideline, but somehow integrate it?
DR. KUMANYIKA: The way it's presented and framed to keep it from
seeming like an additional consideration, to express it more in terms of how it relates to
the advice they've had about food group choices and total calorie intake.
DR. JOHNSON: Yes, I think that's a good point and Carole and the work
group, maybe we could work on integrating it better?
DR. GARZA: Roland?
DR. WEINSIER As usual, Rachel, you do such a great job and you're so
organized. Stepping way back to see if I can play consumer here, the feeling I got from
reading this and see if these are the points you wanted to make, because these are the
points that I took from it, is that the primary problem is particularly among children and
adolescents getting too much sugar, particularly from soft drinks, and the associated
behavior problems, weight problems and health problems. That's the problem.
So, it's a pediatric focus, soft drink focus. The solution is that use less soft
drinks and substitute with milk, low-fat milk. I may be interpreting wrong, but that's the
flavor that comes across and my question is, number one, is this really a primary
pediatric issue? And, it may be, and number two, is the behavior changes you want to
primarily switch from soft drinks to milk? What about water? What about -- I mean,
cause there's no mention of water. Is that not a substitute that you want to recommend,
particularly if there's a concern about calories?
DR. JOHNSON: No, I think you've made two very good points. One is,
the data exists in children and adolescents. That's the focus on that group in the
document. I'm not saying the problem doesn't exist in adults, but there's no data. That's
why the focus.
The second, I think, is an excellent point, because I think we have seen
this displacement of facts with soda and milk, and clearly, children's calcium needs are
very high. They're getting 75 percent of their calcium from milk and dairy products, so
we need to emphasize that, but there may be children that are meeting their calcium
needs, that are also adding calories with soda, that certainly we could encourage another
type of fluids, particularly water, as a good substitute. And, I think that's a good point
that we could put in there about encouraging water consumption, sure.
DR. WEINSIER Can I follow up? This is more of a worrying thing, but I
don't think wordsmithing. Under Box 15, one of the bullets is, "Eat a variety of foods
from the Food Guide Pyramid for healthy teeth and gums." Does this get back to what
we're trying to do with variety? It's the issue that Suzanne and our subcommittee have
tried to deal with, what point do you want to make when you say eat a variety of foods
from the Food Guide Pyramid and how does that help with healthy teeth and gums?
Are you trying to say that the foundation of a healthy diet which is going
to reduce risk of dental disease is the five basic food groups in the Food Guide Pyramid?
Is that what you're trying to say?
DR. JOHNSON: Now, this is not changed from the '95 book. We didn't
make any changes in the dental area, is that correct, Carol?
DR. SUITOR: Very minor.
DR. JOHNSON: Very minor.
DR. WEINSIER So, just tell me what the point is you're trying to make
and then see if maybe I'm just misreading it? Is the emphasis on variety, is the emphasis
on choosing the right foods? This is on page --
DR. JOHNSON: Right, page 44.
DR. WEINSIER -- 44, under Box 15 for healthy teeth and gums. It's the
first bullet. Tell me what the point is you want to make and then we can figure out if the
words do say it or not?
DR. JOHNSON: My guess, since I didn't write this and yes, this is from
the '95 booklet, is that the -- oh. It's not, so did we add that line? We must have.
DR. WEINSIER You probably stole it from our variety guidelines.
(Laughter.)
DR. JOHNSON: Okay, so in the '95 booklet on page 35, it just says, eat
fewer foods containing sugars and starches between meals. And, then we added, "Eat
few foods or beverages containing sugars and starches." I think, Johanna, at one point,
you wanted something added about stickiness, didn't you?
DR. DWYER: Well, I don't think that's what the Dental Association calls
it. I think it's called retentive fermentable carbohydrate. Now, how you put that into
nickel and dime words instead of fanciness, I don't know, but the basic notion is.
DR. LICHTENSTEIN: The dried fruit, that you brought up.
DR. DWYER: Things like dried fruits, that you said before.
DR. WEINSIER Well, you don't need to resolve this now for my sake.
DR. JOHNSON: I wasn't trying to make a point. I don't really know how
that line appeared there, but I'm not married to that line.
DR. WEINSIER If the subcommittee could just think about it, I just
wasn't clear on what the message was.
DR. JOHNSON: Okay.
DR. GARZA: Alice?
DR. LICHTENSTEIN: I guess I'm a little perplexed. When I read this, I
don't see where we're actually recommending milk intake as a substitute, although I
might personally be a proponent of it in addition to water. But, as I read this, I don't see
where you're getting that?
DR. WEINSIER I don't know, it's fairly short, concise and the way it
should be, it's pointed, and the first paragraph and the closing paragraph allude to that.
DR. LICHTENSTEIN: No, they allude to displacement of milk by soft
drinks.
DR. WEINSIER The first paragraph --
DR. LICHTENSTEIN: Yes, displacement.
DR. WEINSIER And, the second and the last paragraph.
DR. LICHTENSTEIN: No, but the last paragraph is talking about all the
food groups, and the first paragraph is just stating the observation of displacement. But,
I actually, in reading this and being on that subgroup, I don't see where we're specifically
saying that you --
DR. JOHNSON: Well, actually, I'd like to make the point that there have
been rumblings, you know, that we need a calcium guideline because of what has been
perceived as a calcium crisis.
DR. LICHTENSTEIN: Right, but I'm just responding to what's written
here.
DR. JOHNSON: I'm just wondering if this is the place and the guideline
to raise some of these issues about calcium, unless it's going to happen in the formerly
called variety guidelines.
DR. LICHTENSTEIN: But, again, the last sentence is really just talking
about displacement.
DR. WEINSIER "Take care not to let soft drinks crowd out low-fat
milk."
DR. LICHTENSTEIN: Right, so it's displacement.
DR. WEINSIER Right, I mean, I'm just one person. That's the way I read
it, because it hit me in the first paragraph and hit me in the closing sentence, that that's a
very important message I should be left with, and if that's the message it should be, then
it should be. I was just curious why, you know, water, perhaps other foods. Maybe you
want them to drink sugar-substituted sodas or maybe you want them to drink juice. I
don't know, I just didn't feel that there were options presented, other than if you're going
to take away the soft drinks, which is important, then what am I left with? And, you're
left with milk.
DR. JOHNSON: Actually, that's just based on the data that that's the food
that is primarily being displaced. That, and there is, I believe it was Skinner's paper
showed that fruit juice is also displaced to a certain extent.
DR. LICHTENSTEIN: Are there any data regarding water intake and
whether water -- one would think it would be, but water is displaced with soft drinks?
DR. JOHNSON: I haven't seen anything.
DR. LICHTENSTEIN: It may just be that the data aren't there. I think
that's the way it was worded the way it was, to be very specific.
DR. WEINSIER I understand and Rachel's point is well-made. I
understand that that's what the data point to, but is that the advice you want to give?
DR. GARZA: Johanna and then Shiriki?
DR. DWYER: Rachel, on the tooth decay thing, maybe we could say
something like both sugars added to foods and those occurring naturally in starches can
promote food decay, especially if they're retained in the mouth, so, you get all those
things. Under Box 15, one thing that isn't mentioned is drinking fluoridated water and,
of course, I think that's important with the sealants and the rinses and stuff.
DR. JOHNSON: And, actually, I have heard there's a concern with all the
high usage of bottled water that it's not fluoridated, that we're kind of worried about an
increase.
DR. DWYER: And, the other thing is, you mentioned that in adults, there
were no relationships -- I don't know if that's after taking into account alcohol intake or
before?
DR. JOHNSON: After taking into account what?
DR. DWYER: Alcohol intake or before? I mean, very often, we don't
want to encourage adults to substitute alcohol and added sugars.
And, finally, I favor a separate guideline. I don't agree with the focus,
particularly, but I think we have to have something about monitoring of sugar intakes,
and maybe there's a way to do it under a rubric. Dr. Murphy, I think, at the end of the
day talked about not grouping in terms of collapsing various guidelines, but just putting
them in some kind of an order, so that we had some simpler sort of framework for
looking at them, and that may be one way to cope with that one, and also whatever
comes up on the sodium and salt.
DR. GARZA: And, the classification, so that everybody remembers, was
moderate adequacy and balance.
Shiriki?
DR. KUMANYIKA: I wanted you to clarify what data you said were not
available in adults related to sources of sugar? I didn't catch all of that, and I know there
are data on how much soft drinks and so forth contribute in adults.
DR. JOHNSON: The data on consumption. There is not published data
on displacement effects of added sugar in adults that I've found. There is published work
on total sugar, which we, I presented early in this process.
DR. KUMANYIKA: There are data, the one paper on sources, which
shows the ranking of soft drinks and cakes and cookies. So, it's not displacement, but it
certainly shows that they are high contributors, and it seems that that's support for
pointing that out to people.
DR. JOHNSON: Right.
DR. KUMANYIKA: And, the things that are below that you might want
to put up higher are --
DR. JOHNSON: Right.
DR. KUMANYIKA: -- potentially displaced.
DR. GARZA: Meir?
DR. STAMPFER: It's interesting in the nurses' study, sugar and alcohol
are, as Johanna alluded to, are inversely related, at least in the women, it looked like that
was being displaced. I don't think that has any relevance for our deliberations.
DR. DWYER: The men take both.
DR. GARZA: Pardon me?
DR. DWYER: The men take both.
DR. STAMPFER: Right, the men have it.
(Laughter.)
DR. GARZA: The only other comment I would make is under the
heading of sugar substitutes, you may want to look at the language so that it isn't -- it
doesn't warn people that you can't replace the calories. I mean, we could probably find a
way to say it more directly. It suggests that if you want a sweet taste without the
calories, but then it doesn't go the next step to say, that doesn't give you license, then, to
replace those calories and still expect to lose weight.
DR. JOHNSON: Okay, great, thanks.
DR. DECKELBAUM: Just a couple of points that came up. One is when
we did have the working group proposed, it was problematic that most of the data that we
have I think was related to children and adolescents and it was a paucity in the adult
population. Roland, I guess, you picked that up, because we had to base the sort of
science on the pediatric or children and adolescent population.
Any suggestions on how to broaden that would certainly be welcome.
And, the other point I think we should mention is a good one that right now, I guess,
we're sprinkling calcium through the guidelines and there could be a greater emphasis so
that, for example, if we, in terms of the replacement of, say, soft drinks, with low-fat
milk, it could be low-fat milk or other calcium-rich foods. It might be an approach to put
in in a few places.
DR. GARZA: All right, then, let's take a break.
DR. JOHNSON: Thank you.
DR. GARZA: Thank you very much, Rachel. We'll come back at 11:15.
(Whereupon, a short recess was taken.)
DR. GARZA: We will move on to our 11:15 presentation, if everyone
will please take their seats. I am very pleased to welcome Dr. Claude Bochard who is at
LaValle, but is going to be shifting from French very soon to a Cajun French, and those
of us that were privileged enough to live close to the Louisiana border can appreciate the
fine time he's about to start having with our Louisiana brethren.
But, Dr. Bochard is going to be the Director at the Pennington Center, and
so he's going to be adding quite a bit to the scientific environment in the Southeast. Dr.
Bochard is going to talk to us about fat and health disease relationships and without my
taking anymore of his time, I think the crowd has especially settled down now that you
could start.
DR. BOUCHARD: Thank you, Mr. Chairman. I have seen the proceedings
of the previous sessions and I hope that you have a linguist going over my presentation
so that it is in proper English.
I repeatedly said to the chairman that I was not the right person to talk
about this topic and you are about to discover why today. It is a very complex issue and
it covers so much of the health and metabolism area that obviously I will be very
selective, and focus on what I believe is a very key problem in this country, and that is
the relationship between dietary fat and obesity, in particular, some of the consequences
of obesity.
I will first of all say a word about the controversies. I think that you can
find in the literature research supporting both sides of each argument or conclusions in
the literature. You will find, and I have removed about 20 slides from my presentation,
but I have data to show you, yes, you can support Hypothesis A, no, you cannot support
Hypothesis A.
So, I will say a word about the controversy. I will spend a bit of time on a
key issue and that is the potential interactions between dietary fat and a sedentary
lifestyle in fostering the current epidemic of obesity. And, finally, a word about the
importance of individual differences, which should always be kept in mind when we talk
about the role of dietary fat.
I guess that a key element in my presentation is summarizing this slide.
There is a major public health problem in the prevalence of obesity. These are based
largely on the IOTF data put together by Jack Seidell and he estimated, based on the data
from almost 100 countries of the world that are accumulated in the IOTF report, that 7
percent of the world population has a BMI of 30, compared to U.S. prevalence of about
21 percent.
Now, I am about to change team and I will be -- I feel like a hockey
player being traded. I will now be joining the winning team.
(Laughter.)
DR. BOUCHARD: Moving from the second best in the world at 14 percent
to the superstar, 21 percent. So, in the U.S. and pretty soon in Canada, being overweight,
that is, having a BMI of 25 and higher, is becoming the norm. It is now already the norm
in this country and will be soon in Canada, so this is a major public health problem that
we should not lose sight of when we talk about the role of dietary fat.
Okay, recently George Bray and Barry Popkin put together a very nice
overview of the arguments in favor of a role of dietary fat in obesity. Like I said, I'm
going to go through that very quickly. Like I said, you can always find a paper and more
than one, with very few exceptions, that will show the opposite. Experimental animals
did not become obese on low-fat diets. That is pretty much the very strong trend. There
may be some exceptions to that, but they will be very rare, indeed.
Experimental animals become obese on high-fat diets, but not all of them.
There are strains that are resistant to dietary fat, and supposedly, we also have human
beings that are resistant to becoming obese on a high-fat diet, as well.
Fat oxidation does not increase in proportion to fat intake, the implications
being that with a high fat intake, there would be plenty of calories left for storage, fat
calories left for storage. I will come back to this. It is an important issue, particularly
when this is confounded by a sedentary lifestyle. A high-fat meal is not compensated by
later decreases in food intake. This is a fairly controversial area. Here, we can find
many studies that are not supporting this, but we do find a body of evidence supportive
of the statements.
Post-obese individuals have an impaired ability to oxidize fat, which is
exacerbated when exposed to high-fat diets. This comes fairly repeatedly across
sectional studies. We don't have many intervention studies or experimental studies, but
the few studies that are available tend to support this statement. High-fat diets can
increase fat cell numbers, which may limit weight loss. This is based mainly on animal
data. There is no equivalent experimental human studies that support that at this time.
Indigestible fat substitutes cause weight loss. There are some studies
showing this, but limited duration. We don't have very long-term studies yet, and some
people on low-fat diets lose weight. Not all of them, but some lose about 20 to 30 grams
per day of weight loss, but nobody is gaining weight on a low-fat diet, and that is also a
fairly solid statement.
Now, the other view was also summarized in Bray's paper, and it pertains
to the arguments, again, of the role of dietary fat in obesity. The first one is a very strong
observation. It is obviously based on ecological data type. In Western nations, the
prevalence of obesity has increased despite a reduction in dietary fats. I will have more
to say about that, and it's one of the controversial topics that I would like to say a word
about. Cross-sectional prospective studies are showing no relation or a very low
relationship between obesity and dietary fat -- yes, in some, no, in others. You can find
supporting both arguments.
When the relationship is there, it tends to be fairly low or moderate. It's
not a very strong correlation and, of course, these studies are confounded by many other
factors which are, or are not, properly controlled in these studies. And, reduction in
dietary fat has little effect on body weight loss. It seems to have in the short term and the
long term, the effect is generally relatively small, but again, it depends what are the other
factors, what are the other determinants of energy balance that are controlled and not
controlled?
So, let me stay with the controversy for awhile. This is an example of a
study where we have differences between obese and lean people in fat intake. For
example, here, we have in the obese, 33 grams of fat in the males versus 29 in the lean,
36 versus about 30 in the female. They are the opposite trends in terms of percentage of
energy from sugar between lean and obese. That represents a whole series of studies that
are supportive of a difference in the amount of fat in the diet.
In contrast, you find those where, ten individuals, normal weight
individuals and overweight individuals were compared, we don't see the expected trend.
No differences in percent of energy from fat and actually, the grams of fat are in the
opposite direction from that which could be predicted.
So, and there are certainly dozens of studies available in literature and you
can select the ones that you prefer, depending on the conclusions that you want to reach
on those.
In an ecological perspective, we, for a long time, we have observed
differences in the fat, portion of fat in the diet and the mean body weight of a population,
or mean BMI, or, in this particular case, the percentage of those who had a BMI above
25. That is the prevalence of overweight. This was put together in the review by Bray
and Popkin, and we have this fairly strong correlation between the percentage of energy
from fat in the diet and the prevalence of overweight, the R2 being almost .8.
Now, we must be careful not to overinterpret these data, because, of
course, they are certainly confounded by a large number of other environmental and
lifestyle factors and other prevailing disease conditions. It's very hard to compare, for
example, conclude on the basis of contrasting the percentage of the diets in India, Maui,
China and the Philippines versus Australia and the USA or Italy, because there are many
other factors besides the percentage of fat in the diet which is controverting to this
prevalent issue.
And, again, if you look closely at the literature, then you can find the
opposite. This is the Monica study. This is the percentage of fat, and here you have the
opposite trend, exactly the opposite trend. This is the medium BMI and the high-fat diet
and the lowest BMI level. So, we can go on and on and on, and depending on the
conclusion that you want me to reach for you, I can select a study and build my case.
So, we have to take some other approaches. In this particular case, I'm not
sure we're gaining much from comparisons of countries, comparisons of culture,
culturally different nations and/or purely epidemiological approaches. Now, an
important component of this debate is what is happening when you reduce fat in the diet?
Again, Bray and
Popkin did a review of that literature, and they plotted here the weight loss in grams per
day versus the reduction in percentage points of percent of calories from fat. And, as you
can see, there is a slope, there is a slope. The R2 is .45. The difference here between this
level and that level is approximately 20 to 30 grams per day of weight loss.
So, it makes a difference, but we're not sure whether, and in this particular
case we're certainly not sure, if it is in the presence of everything else being taken into
account. It's just the dietary fat and many of the studies are not controlling for calories or
other components of the determinants of energy balance.
In contrast to the preceding, we are often shown this, this one. This is part
of the Rockefeller intervention study in which some obese people are being exposed to a
low-calorie, weight reducing diet to reduce their body weight by 10 percent, and then
kept on a weight-stable diet to be studied at both levels of body mass. And, here we have
a 60-year old woman fed 7,300 kilo joules per day on a 10 percent fat or on a 70 percent
fat, and there is no real difference in body weight.
Well, this is my definition, isocaloric, so you would not expect much
difference over time in body mass, if
the diet is isocaloric. If body fat, if dietary fat is playing a role, one of the ways through
which it's going to exert its influence is through the caloric density and, of course,
through the caloric intake. And, an example of the impact it may have on caloric intake
is shown here, where we have a cumulative energy intake under three conditions. Low
density food available, medium density food available, and high density food available,
over a period of 48 hours, two days of three meals and a snack in both cases. And, the
energy consumption on a high calorie, high density items available is clearly translating
into a greater consumption of calories over a two-day period.
So, I guess if we had a large body of data in which both were clearly
controlled, the density and the amount of food and the proportion of fat within this
context, we could conclude that by and large, these are not controlled, except in a few
instances, which I will rely upon later on.
The debate is also fueled by the fact that we see an increase in the
prevalence of overweight and obesity, both conditions, in this country, in the presence of
a decrease in the proportion of fat in the diet. These are data from the recent past
covering about 18 years of observation, and as you can see, decrease is typical of other
assessments of the changes in the proportion of fat in the diet of the U.S. population.
And, at the same time as we registered this, we have the opposite trend, the prevalence of
obesity.
So, those who are a proponent of a role of other macronutrients in the
position to dietary fat can use this to conclude that, indeed, a low-fat diet is not, on the
average, is not fueling the current epidemic of obesity -- or, rather, it is fueling the
current epidemic of obesity.
The arguments on the role of dietary fat is largely based on this cartoon
summarizing the nutrient oxidation issue. It goes this way. When carbohydrate intake is
increased, we have a proportional increase in carbohydrate oxidation. When lipid intake
is increased, we do not have a proportional increase in lipid oxidation, therefore, there is
more left for storage -- the difference between the identity line here and this particular
line is what's left for storage. Therefore, individuals on a high-fat diet will have a greater
risk of having progressive fat disposition until they reach a body mass and body fat
content sufficiently high to foster further increase in lipid oxidation.
We know that obese individuals have an elevated rate of lipid oxidation,
and therefore, obesity is one mechanism by which we adapt to a high-fat diet, to balance
fat intake and fat oxidation rates.
This is an important element and it has been tested in many studies and it
is by and large, by and large quite robust a finding. I draw this line purposefully low, so
I would exaggerate the effect and make sure that the message is understood, but in
reality, it's not as low as it is here. There are differences and there are some studies that
allow us to address this. This is a study we did in our lab, my colleague, Tremblay, led
the project several years ago, in which it's only a 48-hour experiment, but I think it
begins to show the impact it may have.
We have people exposed to high-lipid food available for the day before
and for two days after an exercise bout. Low fat mixed and a high fat, approximately 20,
40 and 60 percent of the calories. Then, they all exercised for a period of 60 minutes and
we subtracted the resting energy expenditure from that and the exercise was the same
under all three conditions. And, we let them eat over the next 48 hours, and under the
low-fat diet over that period of time, we see that we had a negative energy balance in the
post-exercise period. Same under the mixed diet, but on the high-fat diet, the subjects
were in positive energy balance, such that the overall energy balance over two days was
about six mega joule and the low fat and the positive energy balance of about one mega
joule on the high fat.
Now, this is short term and, of course, we must not over-interpret these
results. But, fortunately, there are other data and better data on this issue and with great
humility, I have to admit that they are much better than those from our lab. They
originate from James Stubbs and Andrew Prentice and their colleagues at the Dunns in
Cambridge. And, here, they registered, they obtained a contribution of seven young
adult men. They kept them in a metabolic chamber for seven days, and they were
exposed to three different dietary conditions, low fat, medium fat and high fat, 20, 40 and
60 percent, pretty much as in our previous study.
And, these were fed ad libitum. They could eat as much as they wanted,
but it was covert manipulation of the fat content of the diet, but keep in mind they are in
a confined metabolic chamber, so it's not the normal pattern of daily activities. And, here
we have the food intake, kilograms of food over seven days, cumulative, no difference
between the two groups. However, when we look at the energy intake, purely, the high
fat condition leads to a greater amount of calories, about 85 mega joules versus about 60
for the low-fat diet, under ad libitum conditions over seven days.
Now, then, we can separate out the carbohydrate intake and the lipid
intake. Here, the high-fat diet consumed less carbohydrates. The low-fat diet consumed
more carbohydrates, but the opposite is true for the amount of fat consumed, the 20
percent at about 10 mega joules, the 60 percent at about 15 mega joules of fat over that
week.
In the confine of the chamber, knowing that people tend to be sedentary,
they don't move around as much as when they are free-living, we can see that under the
20 percent calorie content of the diet, they are in energy balance over 70. This is energy
balance. Those at 40 percent were at positive energy balance by the end of seven days.
They had about four mega joules of positive energy balance. But, those at 60 percent
had about 16, 17 mega joules positive energy balance. Clearly, under standardized
conditions, a high-fat diet, when people eat ad libitum, it leads to a positive energy
balance. It would be very hard to detect under population studies or ecological studies.
Now, this was artificial, in the sense that this is a metabolic chamber. But,
they did another experiment. They compared the same diet, three levels, this time for 14
days and they had the subjects coming to the metabolic chambers for the measurement,
spending periods, hours in the chamber, but most of the time, they were free living.
Again, the food intake, kilograms, no difference between the two groups.
Fat intake, of course, the low fat, very low fat. The high fat had a much higher fat
intake, and the energy intake was higher in the high fat. The upper line, here's the high
fat, so we reproduced pretty much the conditions of the previous study. All the meals
were controlled, though. They were eating their meals in the metabolic ward.
But, then, because we had free living people, we had people moving
around and about more, they were able to compare those who had been sedentary in the
previous study versus those who were considered active, that is, those who were free
living in the second study. And, under the 20 percent fat, there was energy balance as I
showed you in the previous people, even in the confinement of the metabolic chamber.
But, those who were moving around and about were clearly in negative energy balance
by about three mega joules per day. Under 40 percent fat, you begin to have positive
energy balance in sedentary conditions, but being physically active brings, keeps you in
negative energy balance. But, on the 60 percent fat, both become in positive energy
balance.
So, we have this interaction between activity and dietary fat. Because
physical activity is a great lipid oxidizer and people would be able to tolerate a fairly
high degree, fairly high level of fat intake, if they were physically active. Like marathon
runners, for example, or people who walk a couple of hours a day. They oxidize grams
and grams of fat. But, most people, and I think we have population data to show that, do
not exercise at that level, and therefore, even on the 40 percent fat diet, there is a great
risk of being in positive energy balance.
Now, we don't have many other studies that have attempted to document
the same phenomenon on the population scale. There is at least one study which has
taken this approach and tried to validate it in a larger setting. Here, what we have is a
Scandinavian study. The six-year weight gain in sedentary women versus those who are
somewhat active and those who are most active, as determined by questionnaires.
And, the weight gain is, on the low-fat diet, when people are sedentary, is
essentially zero. But, even on the high-fat diet it is positive, and therefore, in both cases,
the high-fat diet and the low-fat diet, you have an increase over the sedentary. And, the
active doesn't make much difference except when you go from sedentary to somewhat
active or active. It's not as clear as in the intervention metabolic ward data, but it points
somewhat in the same direction.
Now, if the population data are taken farther than in most studies I know
of, I think we get some of the same message, as well. This comes from the United
Kingdom and, again, from the Danes group, Andrew Prentice and Susan Jebb, in which
they plotted the prevalence of overweight in Great Britain in the high and the low social
classes, as against the energy intake in mega joules per day, and the percent energy from
fat. And, the data plotted as a deviation from the mean for the whole population. We
can see, of course, there is a very strong gradient. The high social classes have low
prevalence compared to the low social classes, very strong grade.
But, in this particular case, there is no tracking, no relationship between
intake and calories from fat. While, when we look at the activity, proxies for activity,
here, a proportion of inactive is derived from the number of cars they have in the
household and the number of hours watching TV. There seems to be some company
between the prevalence data with the surrogate for activity.
This is cross-sectionally. Then we have, over time, these data have been
made available for a period of about 40 years in the UK and over time, without all social
classes confounded, we had this change in the prevalence of obesity. And, here again,
we see that there is very little coupling between the prevalence data and the intake data,
but a better coupling between the surrogate for activity and the prevalence data.
So, I think taking, trying to isolate the specific contribution of dietary fat
without taking into account the mode of life, in this case, the level of sedentariness, may
lead to recommendations or perceptions that are probably right for a segment of the
population, but wrong for the rest of the population. Certainly, relaxing -- relaxing the
dietary fat recommendations to, with a view to increase dietary fat, would be unwise in a
population which is as sedentary as that of the United States or Canada or the Western
world, in general.
Now, let's, however, keep in mind that we have considerable differences
in responsiveness. At least, we believe that. We don't have much human studies to go by
here, but the animal data are very strong. This is an example of those. A study done by
David West at the Pennington Center, where he compared nine inbred strains of mice
under a 12 percent of calories from fat and 33 percent from fat. The dark bar here, as
you can see, under low-calorie, that's a low-fat diet, none of them have an elevated lipid
content in the carcass. But, when challenged by the diet which is 33 percent, almost
three times as many calories from fat, some of them are resistant -- they don't gain fat --
but these gain fat and become fairly obese. Only the AKR/J, only this one, consumed
more calories. The others did not consume more calories, therefore, the effect comes
from the dietary fat, itself. So, these are differences that I assume exist in human
populations. We don't have data, but there is other evidence that strongly supports the
view that we have individual differences in responsiveness.
For example, challenged with cholesterol load or a high-cholesterol diet
for a period of time, we know we have huge variation and responsiveness among people.
And, these variations are not random. When we do these studies with identical twins, we
find that there are differences between pairs of twins. Those members, brothers and
sisters who are genetically similar respond fairly similarly, compared to those who are
from different genetic backgrounds.
In the mouse study, the cross between -- David West crossed the highly
sensitive versus the highly resistant strain, the two extremes in responsiveness to the
high-fat diet. And, did some back cross and then began looking for the genes or the
regions in the chromosome. We call that QTL, quantitative trade locus. You look for
QTLs that were linked to body fat in the offspring of these divergent lines. And, as you
can see here, this is not the full picture, but these four genetic loci were linked with the
type of interest. That is, some markers of adiposity in the carcass.
And, he now claims -- that has not been published yet, but seven loci
account for all of the genetic variants that he has seen, determining the responsiveness to
high dietary fat versus low dietary fat. The genetic variance was about 50 percent of the
responsiveness and about seven genetic loci would account for the whole. So, this
supports the notion that we have individual differences and responsiveness. There are
people that can tolerate more than others a high-fat diet. In the case of humans, it might
be their skeletal muscular capacity, which is the higher rate, which make a greater use of
calories.
So, if this is correct, if this hypothesis is correct, then a given level of fat
intake will not have the same effect on body fat. It will depend on your innate
characteristics. And, it is very hard to predict whether you are sensitive to dietary fat or
resistant to either dietary fat, until you have been challenged by dietary fat. I know of no
long-term experiments in humans. We had some data over a week, over days, but not
over two to three years. And, I assume that these, it's like smoking, we will never have a
randomized controlled trial to document exactly what is the extent of the differences in
responsiveness to long-term exposure to a diet which is 60 percent fat or 70 percent fat.
Having said all that, that would be my last line, Mr. Chairman, it does
matter if you are on a high-fat diet and gain a few kilograms of fat, it does matter. I
didn't want to, I didn't have the time to spend on the implications of dietary fat on health
and morbidities and mortality rates and health outcomes, but let me show you just one
slide which I think demonstrates very strongly the fact that it will matter.
If dietary fat is conducive to greater density of food and creating risks for
being in positive energy balance, it will translate into a few extra kilograms, until you
have increased your lipid oxidation level to the point that you offset this tendency to eat
more fat in your diet. So, you might be in energy balance and weight stable at five, six,
seven, ten kilograms or more, because of the high-fat diet.
Then, the difference is, in my view, very important for health outcomes.
And, the reason I'm saying this is based largely from that study. In Finland, where you
have twin registry, a nice, well maintained twin registry, it was possible for a group of us
to interrogate that registry and ask the following question. There were about 1,500 pairs
of identical twins who were adult, middle-aged, that we could go to by questionnaire and
ask whether they'll, what is their current body weight and height? This was part of a
series of contacts with these pairs of twins, so there were data already in the bank, in the
data file, on height and weight of these twins.
And, we found that there were about 50 pairs who were divergent.
Identical twins, one is normal weight, lean, the other one is overweight or obese. Same
genes. We were able to bring to the laboratory 23 such pairs for investigation and study,
extensive study. And, we have published a few papers with this cohort. This is not
correct -- this is one of the papers and here I'm using some of the data from two papers.
The other one is in metabolism.
In the male, nine of these pairs that we were able to study in the lab, 44
years of age, the lean had a BMI of, a mean BMI of 22.7 and the obese had 28.8, about 5
percent fat difference. But, look at the implications? Incident, fasting incident almost --
DR. STAMPFER: Percent fat, is that percent in the diet or percent body
composition?
DR. BOUCHARD: Percent body composition. Fasting incident is much
elevated, almost double. Total cholesterol is higher, LDL cholesterol, triglyceride, I
have a long list. In the two papers, you will see a long list of markers. I didn't want to
get started. It's already crowded. I limited it to a few markers, but everything goes in the
direction of an iatrogenic profile, just with a few kilograms of difference, kilograms of
fat.
In the women, the same thing. The difference is even larger. BMI of 22
versus 30, and here again, we have all the negative effects, just with the addition of a few
kilograms. So, I would, in closing, I would certainly be one of those who would say it
may not be very critical for some people to, whether they have a diet which is 35, 38 or
40 percent of calories from fat, for some people, but for most individuals with the
sedentary lifestyle that we have, it makes a big difference in terms of the risk of
becoming obese, and, in turn, of being affected by the culpabilities of obesity.
I think I will stop there.
DR. GARZA: Alice?
DR. LICHTENSTEIN: Thank you for that incredible summary. Going
back to the Stubbs data where the subjects were on a 60 percent, 40 percent and 20
percent fat diet, and in the free living situation, it looked like for the 20 percent and 40
percent fat diets, that the caloric intake was not all that different, it was really only in the
50 percent fat diet that the caloric intake or energy intake was higher? Was that actually
the case? Because it's really the 20 to 40 percent that is sort of the variation that you tend
to see among the U.S. population.
DR. BOUCHARD: In the free living?
DR. LICHTENSTEIN: Yes, free living?
DR. BOUCHARD: Well, I have not shown you the whole cumulative
calories consumed, I know I have not shown you that. So, there is a difference between
the three. It may be not as clear cut as in the seven-day study in the metabolic chamber,
but there's a difference, graded, yes.
DR. GARZA: Scott?
DR. GRUNDY: I had two questions. First of all, it seems like the
Rockefeller study is somewhat at odds with the Prentice study, because at 70 percent fat,
there was no weight gain, but it seems like the Prentice would suggest that there should
have been?
DR. BOUCHARD: But, the daily caloric intake was clamped, so it's just
the composition.
DR. GRUNDY: Now, Prentice talks something about passive
hyperphagia, and I think you might agree with that, that people that eat a high-fat diet,
they eat more calories unawares, because the high caloric density of food. It's not that fat
causes obesity, but it's because they consume more and they don't realize they're doing it
and they put on extra pounds. Is that the basic hypothesis?
DR. BOUCHARD: That's a very clear summary of what I was trying to say.
DR. GRUNDY: Okay, the second question, I think you alluded to you
wouldn't recommend loosening up the recommendation to have a higher percentage of
calories. What do you think about the current recommendations that's been around for a
long time, about 30 percent of calories is fat? Is that a reasonable recommendation?
DR. BOUCHARD: I personally, based on, especially based on the obesity
issue, I feel comfortable with this recommendation.
DR. GARZA: Lesley?
DR. TINKER: Thank you very much. I have two questions. The first
one, to make sure I understood something that you said about being weight stable at a
higher fat diet potentially, because of the lipid oxidation, so does that imply there may be
a little more resistance to the weight loss if one is on a higher fat diet and other datas,
does it look at that?
DR. BOUCHARD: Well, let me take the first segment. It's a common
observation that when people gain weight, they oxidize more grams of lipids per day.
So, if they are consuming a high-fat diet, the only way for them to reach equilibrium
between their grams of lipid consumed and the amount of oxidized, is to gain weight
enough until that balance is achieved.
Now, there's some wobbling there, obviously, because you don't eat a
fixed amount of lipids every day, it fluctuates. But, if you are generally a high-fat
consumer, then either if you