UNITED STATES
DEPARTMENT OF AGRICULTURE

Pages: 293 through 516
Place: Washington, DC
Date: March 9, 1999

HERITAGE REPORTING CORPORATION
Official Reporters
1220 L Street N.W., Suite 600
Washington, DC
(202) 628-4888

The proceeding in the above-entitled matter was reconvened, pursuant to Notice, at 9:02 a.m.

P R O C E E D I N G S

(9:30 a.m.)

(Laughter.)

Good morning. Very pleased to see so many of you return despite the weather forecast, which for Washington, I know, strikes terror and horror in everyone's heart. It reminds me a little bit of Houston where two snowflakes and people run for cover. So I've adjusted, you can tell, from after 10 years. Now, 12 inches and it looks normal to me.

Okay, we have the pleasure this morning of having Dr. Tim Byers, Professor of Preventive Medicine, University of Colorado School of Medicine. We've asked Professor Byers to come and review with us the links between fat fiber and carbohydrate intake. That was the original title. Then I called Tim back about a little bit over an week ago and said, "Tim, can you add alcohol to that, and help us look at the links between those four and the risk of cancer?" And he very graciously agreed.

And so with that very brief introduction, Professor Byers.

(Laughter.)

So got a little added sugar to the presentation as well.

So this is my assignment really, to review findings just in the last four years relating cancer risk to these factors. It is snowing outside, I notice. I'll keep you all informed because I do have a little bit of a view of the outside from here.

And to comment as well, I was asked to comment as well on my views about food, the implications for food policy, which is what you all are about, and research approaches, so I'll try to do that.

My approach to looking at these four factors was sort of the usual thing, look at publications largely in the English literature in the past four years. My presentation, however, will not be intended to be comprehensive in the 20 minutes with these four topics, and I'm going to focus instead on two things: the larger pool studies, or read analyses, and focus on specific controversial areas. So we're sort of more of the same, I'll just say that, and focus really on what I think are the more cutting edge issues.

Prior to 1995, there were a few studies about dietary fat and prostate cancer. The emphasis was pretty, I think, weak and inconsistent, and I think since 1995, it's more of the same. I think the range of kinds of findings for prostate cancer is essentially unchanged; that is, that there is a lot of heterogen in either cross studies, there are some very good studies, including one published just a month or two ago indicating fats, in particular, saturated fats, are a risk factor, or a review written by Larry Colonel, published just this last month, I think, covers the topic well and makes a case that there are a range of findings. And if there are relationships with fats, there is probably more evidence for animal sources or saturated fats.

The other thing that I think is important is that there is continued to be evidence that there are effects of fats, especially saturated fats, on circulating androgens and perhaps androgen conversion in tissues, and I think that that's an important continuing development.

As far as colorectal cancer before '95, there were fairly consistent findings for associations with total and/or saturated fats. Most of the research at that point came from case control studies. Since 1995, the associations have generally been weaker from the larger prospective studies and it's pretty clear that various methods of caloric adjustment in epidemiology largely adjust away fat effects for colorectal cancer.

But this is problematic really because the kind of factors we adjust for are not only total calories but also relevant factors for colon cancer risk, including fruits and vegetables, physical activity and red meats, which are themselves associated with fat and saturated fat intake, really, I think, create problems with multi-variant adjustments such that it's really difficult given our current epidemiologic technique to be certain that we're not overly adjusting some of these models, and the new issues, I think, will be screening. In the past, there has been very little screening activity for colorectal cancer. In the future, as we increase screening in this country, it's likely that those people who are screened for colon cancer will also be on more hard, healthy diets, so that will be a new compounder in future studies, I think.

With regard to fats and breast cancer before 1995, the evidence was pretty inconsistent and really weak, and came from case control studies. In the last four years the evidence has been more consistent and more consistently now, and the larger cohort studies really are null for the dietary fats/ breast cancer hypothesis.

I think there is a new -- as we scramble to try to resurrect the hypothesis, there is a new area of confusion that I want to comment on, and that is some confusion about different types of fats.

These are the findings from the pooled analysis of seven large cohort studies conducted around the world, and these four were total, saturated, poly and mono, and I didn't label them because it doesn't matter. As you can see, they are all the same. And essentially across a fairly wide range of dietary fat intake findings are null for breast cancer risk.

And this range really does, I think, even accounting for measurement error, really does include the range from about between 25 or 30 percent of calories from fat, in the high thirties, 35 to 40 percent of calories from fat, so across the range of intakes that are typical in the diet, but probably not below 25 percent. We can conclude that there is not much for relationship between dietary fat and breast cancer. Of course, the women's health initiative is an experimental study to try to test the effect, possible effectiveness of lower levels.

I wanted to comment though on a new problem, which is an analytic problem, I think, and that is, as we try to separate out the effects of different kinds of fats using multi-variant models, there are particular problems that emerge. I think as an example here is a case control study within a cohort in Sweden published just last year in which these are the relationships between saturated, mono and polyunsaturated fats, and across, again, a fairly wide range of intakes, really not much evidence of a relationship. But in multi-variant models in which one type of fat is adjusted for the other, then these fats can separate, and I think that some of this separation of, in this case, mono fats came to be protective, and saturated and polies looking to be slightly risk factors is an unfortunate artifact of a statistical colonary. This is before adjustment, this is after, and the headlines from this were that olive oil or monofats protect against breast cancer.

I think, as we scramble and try hard to squeeze effects out of fats using multi-variant models, we're going to create some artifacts that I think are, unfortunately, probably not a reflection of true biology.

DR. BYERS: All of the risks are below 1.4.

Well, the model fat was a protective relationship that was marginally statistically significant.

The increase in risk for saturated and polies in that study was not statistically significant, and it was certainly below 1.4 as an observation.

Alcohol and breast cancer before '95, they were fairly consistent, but weak associations. There were a lot of questions though at that time about not only beverage specificity for the effect, but dose response relationship, biological mechanisms and latency. By latency, I mean at what point in life might alcohol really be relevant. Is it in later adulthood? Is it teenage years or whatever?

Subsequent to '95, the findings have really been more consistent, and I think we now have some good early answers to all the above questions that I want to comment on. But the new questions and the questions that policy panels such as you have to deal with are the trade-offs for heart disease. So I'll comment on all those things.

Here is a pooled analysis from the same seven large cohort studies showed before for dietary fats, indicating that above -- this would be 15 grams or so would be about a drink a day, and certainly above a drink or two a day you start to see a linear increase in risk for breast cancer.

The question of latency, I think, has been addressed nicely in three studies that have been published in the last two - three years, and they all agree that alcohol intake -- it appears that alcohol intake later in life is more relevant to breast cancer risk than alcohol intake early in life. One of the questions was, was maybe this weak, inconsistent relationship with alcohol might be sort of a residual effect of heavier drinking earlier in life, and it looks like that probably is not the case; that alcohol, in fact, does increase breast cancer risk -- because I believe it does -- then the effect appears to be more proximal to the breast cancer itself.

There are a couple of studies of alcohol and stage of breast cancer, and both studies agree really that alcohol drinkers, women who drink alcohol tend to present at later stages of breast cancer. The explanations for that aren't clear cause there is really two possibilities. One is that there is a biological effect of promotion by alcohol and breast cancer. I think that is plausible by affecting estrogens. The other possibility that needs to be teased out is that there may in fact be diagnostic delay, and that women who drink alcohol may be less attentive to breast cancer in terms of screening mammography. That should be fairly easy to answer pretty soon, I would think.

The real problem is, as you're well aware, is the trade-off. Here is data from two large prospective studies, the Nurses' Health Study and the American Cancer Society -- Cancer Prevention Study II. The solid lines from the two studies indicate this U-shape relationship with total mortality driven largely by benefits to cardiovascular health, of drinking one or two drinks a day in this range. The breast cancer risks shown in the dotted lines, and when you get up to one or certainly two drinks a day, then I think there are appreciable breast cancer risks.

The trade-off then, I think, has to be one, and the tough thing for you all to consider is that there is not only just gender-specific specificity for possible adverse effects of low drinking alcohol for cancer; that is, affecting women with regard to breast cancer, but also age-specificity as well.

Here are the U.S. mortality data showing the breast cancer relationship with age, showing the heart disease relationship with age, and a line that I've added, which is breast cancer mortality times two; two times breast cancer risk is something that a lot of women can fairly easily compute for themselves based on their family histories, other risk factors, previous biopsies or, in fact, previous breast cancer.

And for breast cancer, it's actually similar or maybe even a higher cause of death premenstaulpausily than is heart disease. For women at higher risk for breast cancer, the breast cancer times two curve, that cross-over really doesn't happen until about age 60. So the tough thing for you is that it's not only issues about gender-specificity, but also age-specificity as well.

Just a couple comment on fiber because I'm pretty much going to dismiss it in my conclusions, and a lot of studies have shown this kind of thing. This is a large study out of Italy of about 2,000 cases, 2,000 controls, showing that fiber from grain sources is unrelated or perhaps actually associated with increased risk, but I think most conservatively judged as unrelated to risk, whereas fruit, and especially vegetable fiber, associated with lower risk.

Now, a little sprinkling of sugar. Before '95, there were a few studies. After '95, there are still a few studies. Most of the interest in added sugar and cancer has been with regard to colorectal cancer, but I think really the findings are weak. Even as I look at the papers that ostensibly from their titles and their abstracts report a positive finding with sugar or sucrose and colon cancer, I think the findings are even weak within those analyses. I think there has been a tendency to feature on subgroup findings and the big picture is there's not much of a relationship there.

There is problems though with studying added sugars in cancer. The biological hypotheses are pretty compelling that either caloric load or maybe stimulation of insulin might help to drive the promotion of neoplasia, and it's biologically appealing. The problems, of course, in epidemiology is it's difficult to quantify many things in the diet, especially things like added sugar, and that sugar, choice of the sweet foods associated with other aspects of behavior and it may, in fact, be an indicator food. So if you ask somebody how often they eat candy, you can make a pretty good guess about some of their behaviors as well.

After reviewing about 100 studies on sugar in the diet, Burley actually published a pair of papers in The European Journal of Cancer Prevention, the most recent one last year, and concluded, after reviewing all of them, it's apparent that there is insufficient evidence to conclude whether sugar has a role in cancer at any site, and I would agree with that even though I didn't go through as many studies as Burley did.

So just to conclude my comments, first, with research recommendations. These are obviously in a nutshell. I think, with regard to alcohol, the main -- the effect of alcohol in upper elementary cancers is pretty clear, and those are higher doses. The only real cutting edge question is low dose alcohol and breast cancer, and there I think the research really needs to focus on estrogen effects as it has, but I think there needs to be more work on estrogen effects of alcohol.

With regard to fats, I think the only critical question now really is fats, especially saturated fats, in prostrate cancer, and there I think we need more studies on the effects of fats on androgens in men.

I think with regard to fiber, we just need to get over it and focus on foods and not fiber per se in cancer.

And with regard to sugar, I think studying insulin resistance syndrome as related to neoplasia, since that's really the primary method, mode that the hypothesis holds for sugar, is probably the way to go, to better understand insulin and missal-like growth factors as they relate to neoplasia, and then to back up from that and make inferences about sugars and starches.

So since 1995, there have been two important dietary guidelines issued apart from the ones that your predecessor panel did in 1995. The American Cancer Society a year later, in '96, and The World Cancer Research Fund, in conjunction with the American Institute for Cancer Research a year later, in 1997, issued dietary prevention, or cancer prevention guidelines, and they really overlapped substantially, and the substantial areas of agreement between these two guidelines are just summarized here: Eat more plants, eat less animals, avoid obesity, be physically active, drink little, if any, alcohol. Perhaps you just want to take this exact wording for your next guidelines.

Food policy implications of the finding in the last four years, I've summarized here. For alcohol, I think the policy challenge for you all is that I think the best evidence now is that there does need to be some gender and age specificity now in alcohol guidelines, how you translate that into words or how we translate it into bumper stickers or other messages for the U.S. population is indeed a challenge, but I think that science dictates this.

For fats, I think the types of fats to avoid to lower cancer risk, either for colon or prostate cancer, where I think there are still some questions especially about saturated fats are, fortunately, consistent with heart disease prevention recommendations. So I see no problem or no conflict here. I think cancer can take a back seat to heart disease when it comes to types of fats, so I think that's appropriate, just as in the past.

With regard to fiber, I think we should really avoid the term in recommendations and focus instead on the foods that contain fiber and, I think, other nutritional aspects of those foods that are more relevant for cancer anyhow.

And for sugar, I see really no need to add cancer to reasons to limit sugar intake.

So those are my comments. I'd be happy to take any questions or comments.

Dr. Meir?

The comment is I agree with you on the fiber, but just to emphasize that this was for cancer and that there may be other health benefits for fiber besides cancer.

My question is, Tim, for the breast cancer what's your -- the current guideline is one drink per day for women. That's the current dietary guideline, and what's your take on the level of risk for breast cancer at that level of consumption?

So at exactly one drink a day, I think, I think there is some elevation of breast cancer risk but it's probably of that order.

DR. GRUNDY: Obviously, there are people who believe very strongly about diet and cancer, and it was, you know, believed so strongly that they initiated that major study and millions of dollars are invested. What was the scientific data or has that changed? 

DR. BYERS: Which study are you referring to? 

DR. GRUNDY: Well, I'm taking about like the women's health study where they are going to have a low fat diet and obviously that was -- to mount a study like that there has to be a lot of presumptive evidence and cross-culture and all that. 

DR. BYERS: Well, I think the best summary is -- Roy provided by the National Cancer Institute itself, the rationale for it. There has been repeated papers. Most recently, Peter Greenwall restated the rationale just a month or so ago. I think it's in JNCI or one of the national journals.

I think the rationale for that experiment is that you need to get to lower levels. There may be a threshold below which there is an effect, that's it hard to really measure that in observational studies. I wouldn't want to argue strongly for the rationale, but this is sort of water over the dam from a decade ago, and the trial is well underway, and the difficulty in interpretation of the trial is it is not just a low fat trial; it's sort of a total diet trial. But that's good enough for me.

I actually think the Women's Health Initiative Study is a reasonable thing to do.

DR. BYERS: I think if we would have had these data in hand 10 years ago when the debate -- I guess it was 10, it seems like 10 years ago when the debate was going on -- that trial may not have gone forward. But at the time I think there was uncertainly, certainly uncertainty in the extent to which we should believe observational epidemiology, uncertainty that there may have been a lower threshold.

My guess is that the trial is not going to be particularly positive, but nonetheless I think it's -- I thought when the decision was being made that it was a reasonable trial to do.

CHAIRMAN GARZA: Dr. Lichtenstein. 

DR. LICHTENSTEIN: A very informative presentation. You said that there had been some work done on the relationship between alcohol intake and breast cancer as a function of, I guess, age of women, age of diagnosis of the breast cancer.

I'm just wondering, has there any work been done on tumors that are estrogen-sensitive to estrogen-nonsensitive and any alcohol intake?

I'm sure somebody has looked at it. Meir, do you know the estrogen on that? I'm not sure what the literature -- what research has been done on that, to tell you the truth.

DR. BYERS: You mean it hasn't been studied or there is no difference? 

DR. STAMPFER: No, it has been studied in -- it has been studied, although not thoroughly, but the studies that have been done don't show a difference. 

DR. LICHTENSTEIN: So then that would argue against the relationship between alcohol and estrogen or estrogen metabolism? 

DR. STAMPFER: No, not necessarily, because estrogen sensitivity changes in the course of tumor progression. 

DR. BYERS: So there may be effects on a tumor that as it becomes apparent looks to be estrogen receptive negative. 

CHAIRMAN GARZA: Dr. Dwyer? 

DR. DWYER: Just a quick question, Tim, on the alcohol breast cancer. Has anyone seen any relationships between hormone replacement therapy and the effect or lack of effect of alcohol? 

DR. BYERS: I think that was included in the pooled analysis, HRT, and the report I'm understanding in general is that there was not much difference across a number of other risk factors.

Was HRT one of those?

CHAIRMAN GARZA: Dr. Johnson. 

DR. JOHNSON: Thanks, Tim.

I was curious about what you said about added sugar because I know in reading the World Cancer Research Fund and American Institute for Cancer Research, that book that came out, that they came to the conclusion in that section on colorectal cancer that sugar, particularly sucrose, was associated with colorectal cancer.

DR. JOHNSON: Yeah. And, in fact, among the recommendations one of the -- in the end, when they give their recommendations, one of them is to limit consumption of refined sugar, and I'm going to show that in a minute. So I'm just curious if you could comment that you've clearly come to a different conclusion that that group did about sugar. 

DR. BYERS: Yes. As I look at the studies, especially the bigger, stronger studies that ostensibly have themselves concluded that there is a relationship, it seems to be just within subgroups, so you get an effect in young men, but not older men, and no effect in women and so forth.

And so I think if there is a relationship, it's pretty weak.

Now, the weakness of that relationship may be a function of the difficulties I listed as how to measure sugar and analyze it and so forth, but that's my own take. I would be interested to hear yours.

You know, I'm wondering is it the sugar or --

CHAIRMAN GARZA: Dr. Kumanyika. 

DR. KUMANYIKA: Tim, do you know if there has been any dietary pattern analysis on any of these cancers? Because since obviously these are all related, has anyone done any of the scoring, index -- 

DR. BYERS: Yeah, there have been some, and some attempt at sort of cluster or factor analysis, and you can come up with clusters or factors, and you can attach names to them, but just what they mean, I mean, it takes somebody smarter than me to figure out what those clusters really represent.

There have also been some other analyses looking at patterns, but you get sort of a predicable thing like with fruits and vegetables especially. So I think that the food pattern approach and analysis that's been done to date has not really added to the field very much.

DR. WEINSIER: I realize that you're charted to not include the relationship of physical activity, but you mentioned AICR's recommendation to be more physically active, and, in fact, it's part of our charge to consider in the weight guidelines.

Do you know of any evidence outside of the relationship of physical activity to obesity and then to cancer, particularly breast cancer, do you know of any direct evidence of physical activity and cancer?

For breast and prostate cancer, the other two cancers where there has been thought to be a relationship, I think it's either not there or very much weaker that colon cancer.

DR. DWYER: No. 

CHAIRMAN GARZA: Any other questions?

Tim, there is one other one, and it wasn't in your charge either but perhaps you ran across the whole issue of selenium and prostate cancer.

How strong is that relationship, and obviously it has some implications then for how one balances between animal and non-animal food given the sources of selenium? Any observations you want to share with the group?

I think the possible effect of selenium on cancer is one of the most exciting things that's happened in the last decade in cancer research, but it still is possible and needs to be confirmed. I think if subsequent trials indicates even half the benefit of selenium, as Larry Clark's found in the secondary data points in his trial, then I think that's the biggest finding since tobacco and cancer.

So are we excited about it? I'm very much enthusiastic about new trials that are going to be underway, and I've got my fingers crossed.

But another point of interest is that Finland, which is the lowest selenium country based on their early studies, decided to fortify and raise their selenium levels quite dramatically 10 years ago, and there hasn't been one iota of suggestion that their prostate cancer rates have decreased.

So, yeah, I'm hopeful too, and it would be great is the Larry Clark data were replicated, but I think we need to be cautious.

The other aspect of Finland is very interesting, and that is that their lung cancer rates began to dip before the rest of Europe, and so I think the jury is out on selenium.

For those of you who missed the meeting yesterday, we did not get to grain products, vegetables and fruits despite heroic efforts on everybody's part, so we're going to go back to that portion of yesterday's agenda and take it up from there.

So first slide, please. Can we lower that a bit because most titles will be missing then. That's fine.

DR. DECKELBAUM: So the grains, vegetable and fruits working group was consisted of myself, Alice Lichtenstein and Meir Stampfer, but we really had very substantial help and input from USDA and other staff, including Elta Salton, Shanthy Bowman, Andrea Lindsey, and Kathryn McMurry, Carol Davis, and they gave us a lot of materials and lot of substantive advice in formulating some of the things you are going to hear ago. And as well, I'd like to thank Carol Suitor and Suzanne Murphy and Burt Garza, who also took place in our meetings and had a role in what you're about to see.

Next slide, please.

So the 1995 guideline here is at the top of this slide, and the three major sections that were discussed in the previous guidelines are listed here underneath. Our charge was really to review the science base and add to it, focusing on literature since the previous report. If supported by new evidence, we were asked to make appropriate revisions, and as well, we looked at modalities to suggest approaches for better implementation of the content of the grain, fruits and vegetables guidelines.

Next slide.

So these are the options that we mainly focused on during our deliberations. We asked if there should be an increased emphasis on whole grains in the guideline itself and/or in the text. Should there be clearer definition of different types of carbohydrates?

Are carbohydrates in potatoes as good as carbohydrates in broccoli? And we also had discussions related to the potential role of the glycemic index in choosing carbohydrates, and Dr. Stampfer will be talking more on this in a few minutes.

Should there be more emphasis on quality versus the quantity of grains, vegetables and fruits ingested, and that's related to what I just said? But this could also include should we better point out grains, fruits and vegetables that are rich in certain macra nutrients, say fiber or micro nutrients, certain antioxidant vitamins?

Should we have an increased emphasis on not ingestion, and, again, Dr. Stampfer will be addressing that? And we also reviewed potential ways for clearer implementation guidance for grains, fruits and vegetables, and along these lines Dr. Lichtenstein will be giving a short talk as to the question should the grain guideline be separated from vegetable and fruits.

Next slide.

I'm not going to spend much time on this because the good news is that since 1995 there's an increasing body of literature showing beneficial effects of fruits and vegetables in decreasing cancer, cardiovascular disease, cataracts, diverticular disease and likely Type 2 diabetes, and the references for these will be provided in our updated report.

Next slide.

So I'd like to turn now to the question whether there should be increased emphasis on whole grain products and look at some evidence relating to coronary heart disease risk, cancer risk and Type 2 diabetes and insulin resistance.

Next slide. This is where slides didn't -- got messed up.

This is a meta-analysis study of Dr. Jacobs and his group, "Whole Grain Intake May Reduce the Risk of Ischemia Heart Disease, Death and Post-Menopausal Women," the Iowa Women's Health Study. So I will be focusing on the first few slide actually on coronary heart disease.

They studied almost 35,000 post-menopausal women. The relative risk was about .6 confidence integrals shown for the top versus lowest quintile of whole grain intake, and this was not explained when adjustments were made for fiber, Vitamin E and Folic, suggesting that whole grain intake is protective for ischemic heart disease.

Next slide.

The next two slides have been borrowed from Dr. Stampfer. This is unpublished data from the Nurses' Health Study, looking at whole grain foods and the risk of coronary heart disease in the Nurses' cohort, and looking at exposure of at least five to six servings per week. You can see that cereals, bran and brown rice, all markedly decrease the relative risk of coronary heart disease in women, but I don't know if this is good news or bad news, popcorn had no effect.

(Laughter.)

Next slide.

What about men? These are two studies that include men, Rimm paper published in JAMA is on the health professionals study; Pietinen study where it was the ATB study where the primary end point was cancer, but they looked at coronary heart disease and found that there was substantial decreases in risk associated with whole grain intake. Now, this is in addition to the effects that whole grains could have on improve lipoprotein profiles and separate, and along those lines I'd like to point out that the effects of carbohydrates, for example, and HDL cholesterol may not be -- may not be applicable to all types of carbohydrates.

We published a paper, Tom Stark and our group, last June in the American Journal of Clinical Nutrition, showing that in hypercholesterol anemic children when they went on to fat low, cholesterol lowering diets that HDL decreased only when simple sugar increased but not when complex carbohydrate intake was increased.

But, in general, the papers also seem to indicate that risk reduction is associated with higher levels of whole grain intake, and they cannot entirely be explained by adjustments for fiber intake.

Next slide.

This is another slide borrowed from Dr. Stampfer which shows that the relative risk for coronary heart disease is decreased in general across a number of studies looking at fiber consumption and relative risk for coronary heart disease. I think the important point here that it looks like whole grains do have an important effect, but we still can't dismiss fiber, as we just heard we might have to do in terms of cancer, in terms of coronary heart disease risk.

Next slide.

Now, what about whole grain intake and cancer, and I thought I was showing this slide earlier but it came up now? This is the study of -- another study of Jacobs, et al., looking at meta-analysis of 40 case controls studies between 1984 and 1997, looking at a variety of cancers, 20 in all, and colonic polyps, and, again, there is a lot of data in this paper and I'd be interested in Dr. Byers' comments on them.

But the pooled odds ratio, looking at the entire cohort, was about 0.66 for high versus low whole grain intake. And again, this was maintained in general through most cancers. Of interest, breast and prostate had lower correlations with whole grain intake in terms of reduced risk after -- this was maintained after adjustments for social-economic status, age, sex, BMI and other things listed here and not listed on the slide.

Next slide.

An interesting paper published by Chanteoud in the International Journal of Cancer looked at 10,000 cases and 8,000 control cases of hospitalized cases -- hospitalized patients, mainly in Italy, with different kinds of cancer, and they reported risk ratios, again, with high whole grain food intake reduced by substantial amounts for GI tract cancer, bladder and kidney, lymphomas and myelomas, but not for breast cancer. And again, these were maintained after adjustments for a number of potential other variables, other confounders.

Next slide.

Briefly, I'm just going to show an overhead. Here on whole grain intake in non-insulin dependent diabetes, two papers by Salmeron and Group. The top one, "Diabetes are," and the bottom one -- sorry, the top one is "Diabetes Care On Men." The bottom one is in JAMA on women, large cohorts again, the Nurses' Health Study and The Health Professional Study.

And the bottom line here is that glycemic index, they suggested foods that have a high glycemic index are associated with about a 1.4 to 1.5 increased risk of non-insulin dependent diabetes, and, again, whole grains are associated with a decreased risk, about .7 relative risk for non-insulin dependent diabetes.

Next slide.

So what are some of the issues that could be involved in considering adding whole to the guideline itself or increasing its emphasis in the text going with the guideline?

Well, there could be implications that grains that are not whole are not part of a healthy diet. That's actually related to something a little lower on the slide that we don't want to label necessary good versus bad food or should we.

Could there be a cost factor involved which might affect certain classes, lower SES classes who might not be able to afford some of the good whole grain products?

Will there be decreased intake of enriched and fortified foods and how would this affect, especially micro nutrient intake, and actually in some analysis we did we really didn't see any effect on this that we could sort of tease out relating to folate as well?

Would there be a decrease in food choices that could affect certain segments of the population?

Might certain important nutrients actually have decreased absorption? For example, might fitates in whole grains affect iron and zinc absorption? And again, there is really no evidence yet on this, but this is something that certainly would have to be looked at.

I mentioned good versus bad foods, and again, is the scientific evidence really strong enough to make these changes because as we've seen in a number of examples of where studies that came out, for example, in case control fashion a number of years ago are not supported by larger cohort studies or clinical intervention studies, so that we still have to consider whether the -- for example, even the papers I showed you are sufficient to allow us to make these substantial changes.

Next slide, please.

Just a couple of words on our plans, our thoughts towards better guidance provisions in the guidelines, and actually if we can just -- we have to actually raise the bottom of this slide. This is Box 9 from the current recommendations which sort of tells us how to go about getting the diet better with more grains, fruits and vegetables.

At the bottom of the box you've got to go to Box 2, page 7 for what counts as a serving. So, in other words, it's not a total user friendly, take one look, it's all there type of box.

Next slide.

So we're going to be looking at as a working group ways to improve the message, and this would be just one example for Box 9 where here would be some of the messages that we might want to include, and paralleling it right next to it would be a "how to" box with different kinds of messages. You know, pack X and X fruit in your purse or in your handbag for lunch or your afternoon snack, et cetera. So these are the kind of things we are considering.

I would like to ask Dr. Stampfer to come up and talk about glycemic index and different kinds of carbohydrates.

Nuts are basically recommended against, if you take the dietary guidelines at their face value, because nuts are a high fat food and it says choose foods low in fat. And a lot of people have been doing that, and it's not a good thing I want to tell you about that. Although nuts are a very high fat food, most of the fatty acids are unsaturated and it's a good source of protein and some other good things.

Next slide, please.

Consequently, as you might expect, nuts have a favorable impact on the lipid profile because of their mostly unsaturated fat content, so with a high walnut diet the LDL/HDL ratio was substantially reduced. This is a very good predictor of risk of heart disease, diet supplement with almonds also lowered LDL substantially, and this is just what you'd expect from what we know about the impact of high unsaturated fats on the end -- monounsaturated fats on the LDL and HDL ratio.

Next slide, please.

Well, that's all well and good, but what about clinical end points, and I think this is something we need to keep coming back to and not just rely on the influence of diet on intermediate markers, but we need to look at actual disease outcomes, and there has been three published studies so far looking at nuts. All three find substantially reduced risks of coronary heart disease with nut consumption. We're not talking about mega doses here; just a handful of nuts a couple times a week was enough to reduce risk in the range of 30 or so percent, a very big decrease.

So I think nuts should not have this astigmia attached to them, but to the contrary. Their consumption should be promoted.

Next slide, please.

Okay, that's all I'm going to say about nuts. Now onto glycemic index. It's a complicated concept, and I don't know if I'm going to succeed in getting it across in the short time I have, but I'll try.

The basic idea is very simple. Different foods have a different propensity to raise blood sugar following their ingestion, and the glycemic index is a way of quantifying that. Typically, in the traditional sense nutritionists have divided carbohydrate into simple and complex. Simple being mono and diet sacarides, sugars, and complex being everything else. But this is not a physiologic distinction as the glycemic index is. The glycemic index is really based on what happens to real people who eat food, so that they're given various kinds of foods and blood sugar is actually measured, and it goes up sharply with foods that have a high glycemic index, and less so for foods that have a flow glycemic index.

Next slide.

So this shows, for example, what happens if you have the same caloric intake for three different kinds of carbohydrates, glucose, amylose pectin or amylose based on either the glycemic response -- let's see over here -- so you can see how glucose, blood sugar shoots up very fast with amylose pectin goes up but less so, and with amylose, which is less readily broken down, its much flatter, and consequently the same pattern emerges with insulin, and this has physiologic effects. It's not a good thing for the system to have your glucose and insulin shooting up and down very sharply that way.

Next slide, please.

There is -- in muscle, this is in animal studies, insulin sensitivity of muscle, glycogen synthesis is impaired in amylose pectin fat rats. That's a type of carbohydrate that has a higher glycemic index than amylose.

Next slide, please.

And post-meal lipogenesis, also another animal study looking at, again, the different kinds of carbohydrate here. The high glycemic index fed animals had a higher capacity for lipogenesis after a meal. So these are all adverse physiologic outcomes of high GI diet.

Next slide, please.

Now, it's important to distinguish between glycemic index and glycemic load. The glycemic index is the property of the food. It's the property of how the food can raise blood sugar. Glycemic load is what that food -- is basically taking into account the amount of carbohydrate that is in there, so it's not just the quality of the carbohydrate but the amount. So you can think of it as glycemic index is sort of the nutrient composition of the good whereas the glycemic load is the amount of that nutrient that you get from eating a normal portion size.

Now, everything is calculated in terms of percent of white bread. So you can see, for example, carrots have a high glycemic index, 131 percent compared to white bread is 100 percent, but there is very little carbohydrate per serving, so it only accounts in a typical diet to one percent of the glycemic load whereas potatoes are similar to white break, they have more carbohydrate, of course, and they account for eight percent of the glycemic load of a typical diet, and obviously it's going to change, but you can see those differences.

Next slide, please.

Well, again, what about clinical end points? Does this really matter for real people?

Well, these are data from the Nurses' Health Study looking at glycemic load and risk of coronary heart disease, unpublished data from Dr. Sima Lu in our group. And you can see that the high glycemic load in various statistical models, either adjusting for fat or adjusting -- without adjustment for fat, you can see in the best model, which adjusts for fat intake, the highest level of glycemic load is associated with about a doubling of risk of coronary disease, highly statistically significant, and this is taking into account all the other coronary risk factors.

Next slide, please.

Now, what is contributing to the glycemic load in the Nurses' Health Study diet? The number one contributor is potatoes. And so we looked specifically at potatoes and after adjusting for all the coronary risk factors what one finds in these data, again, is about a doubling in risk with high intake of potatoes. It's not -- the confidence intervals are broad and the trend is of borderline statistical significance, but it's clear that I don't think we can consider potatoes as a health food here.

Next slide, please.

Finally, you can ask why, if glycemic index and glycemic load is so important, why don't we have an epidemic of coronary disease in China, for example, where white rice makes up a big part of the diet? And that's a fair question.

And the answer is that the impact of a high glycemic load diet is mainly among people who are already marginally glucose intolerant; that is, overweight and inactive, and those qualities are uncommon in China, although they are getting more common, but in the U.S. they are very common.

So these are data showing the relation between glycemic load in relation to risk of coronary diseases by body mass index. So among people who are lean, glycemic load really doesn't matter very much. Unfortunately, most Americans don't fit into this lean category. Most Americans are here where high glycemic load will double the woman's risk of coronary heart disease.

I think I've gone over. I'll stop here.

Oh, let me just say one more thing on popcorn. Although it wasn't statistically significant, popcorn did have the same trend as all the other whole grains.

(Laughter.)

DR. DECKELBAUM: It has corn syrup, and Dr. Lichtenstein will now discuss the possibility or the question to splitting the guideline. 

DR. LICHTENSTEIN: Okay. Well, as indicated, I would like to suggest that we consider splitting the grains from the fruits and vegetables and having two separate guidelines. One of the primary reasons I think we should consider this is I think we should really think a lot about what we are recommending people should do. We've spent a lot of time in the guidelines recommending what people should not do, and there's been word smiting over the years as far as, you know, consumer diet low in something while consider diet moderate in something, trying to make it sound more positive. But I really think what we need to go is give individuals more guidance on what they should do and what makes a healthy diet.

I think if you read the literature on the predictors of grain intake are different from the predictors of fruit and vegetable intake, so this is one reason for splitting them because people view them differently, and I also think the barriers to grain intake are different than the barriers to fruit and vegetable intake, and that's taking into consideration cost, storage, preparation, perishability so that -- and safety also, so that one needs to give different guidance and advice to individuals with respect to fruits and vegetables versus grains.

I tried to see if this issue had been addressed directly because we are supposed to propose changes on the basis of scientific -- a scientific basis for proposing changes, but the question has never really been addressed directly.

Now, there is an error. The first focus group that I'm going to mention was actually in 1995.

But what I did is I looked at the focus groups, and although that specific question had never been posed or considered, we can get certain nuggets of information about it, and when the fruit, vegetable and grain guideline were considered one of the comments was that the suggestion that increasing fruit and vegetables was challenging because of cost, and this speaks to the issue of different barriers to fruit and vegetable intake versus grain intake.

There was another focus group conducted in September of 1988, and one of the comments there, again since the question specifically was not posed, was "I like to eat more fresh fruit and vegetables but I can only shop once a week. In two or three days the stuff is no good. The rest of the week is going to have to be canned or frozen."

Well, clearly, we didn't get the message across because canned and frozen fruits and vegetables are quite acceptable, and it's not just that someone has to consume fresh fruits and vegetables to get the nutrient value.

Another very telling comment was in a focus group that was published in August of 1998, and in this case it had to do with the whole guideline, and the guideline has the word "You should consumer a diet that has plenty of grains, fruits and vegetables." One of the comment was, or on the term "plenty" because it was being equated actually with the "Five-a-Day Program," which is another federal program; that the comments were related to how much is "plenty." Well, it's five and that someone actually mentioned the Five-a-Day Program, so I think that there is a program specifically that focuses on fruits and vegetables. The "plenty" is not a quantitative term, although it's certainly quantitative in the food pyramid which, by the way, also distinguishes between grains, fruits and vegetables. So I think separating the two would be quite consistent with current programs that actually encourage increased consumption of fruit, vegetables, and then grains.

Again, just to reiterate, I think the message should really emphasize what people should be doing as opposed to what they shouldn't be doing, and that the message is if fruits and vegetables were separated from grains within each of those categories, I think we could be more focused and clearer in what guidance we're giving for grains, what guidance we're giving for fruits and vegetables.

Also, as mentioned, that Box 9 that's actually in the current guidelines is quite big. There is a lot of information, and I think that's where the difficulty lies in actually distinguishing between how to give advice for those different groups, so I think it's something that we should consider.

Are there any questions of any of the presenters?

Roland?

Meir, help me with the glycemic index and trying to do something practice or make a practical, safe and reasonable recommendation to the public based upon these findings. If in fact the -- I don't know what you're calling it on the right side of the slide. What is it, "The relative glycemic impact of the diet which takes into account the glycemic index as well as the glycemic load." Potatoes are eight-fold greater impact on the diet than carrots. And I'm trying to envision populations in the world, whether it's China, or Papuans or other, you know, potato/rice eating populations that subsist in these foods, I guess, are healthy.

Their glycemic impact factor would be what, 100 percent? Ninety percent? You know, it would be extraordinary.

So we'd have to argue then, based upon the date you're representing, well, it doesn't have an impact on them because perhaps of their normal weight, and it's only with BMIs that are getting in the higher range with insulin resistance that's having an impact. And I don't know what the answer is, but I'm having trouble from a rational standpoint trying to separate that we would do something different from a dietary standpoint that prevents a disease, a chronic disease, than we do in people who have the chronic disease, i.e., if you're lean, we should have one diet prescription; if you're getting above a certain BMI, we have a different diet prescription.

Help me come up with a plan that would make sense.

DR. WEINSIER: A guideline that makes sense. 

DR. STAMPFER: It will take a minute or two.

First, just let me explain that slide with the eight percent and the one percent. That slide represented different contributors to the total glycemic load in the Nurses' Health Study diet. So of the total glycemic load, eight percent was contributed by potatoes. That's what that eight percent means. It doesn't mean that potatoes are eight times as bad as carrots or something. It just means that the way the diet is distributed of the total glycemic load of the diet, eight percent was contributed by potatoes, which was the number one contributor.

Now, the question you raised about populations that seem to do fine, and certainly have very low rates of coronary disease despite a high glycemic load diet, I believe the reason for that is that they have -- do not have by and large the levels of insulin resistance that we have by virtue of physical activity, because muscle decreases insulin resistance and lean body mass, so that the adverse effects of the high glycemic load diet are manifest where there starts to -- starts to be insulin resistance.

Now, I wouldn't characterize that as a disease state because if we did, you know, three-quarters of the population in the U.S. would be characterized as diseased. Well, maybe they are. But to the extent that we as a country are fat and slothful in our physical activity patterns, there is a lot of insulin resistance and this is being exacerbated by the high glycemic load diet.

Now, how to implement that is another issue because it's kind of complicated to get across in a dietary guidelines, and also I think -- although I think this is an exciting area of research, I don't think the findings are completely proven or conclusive, so we need to decide, you know, if this is ready for prime time.

My take on it is that this lends strong support to what our group has been trying to get across, which is an emphasis on whole grains and minimally processed foods, and I think it also lends support to taking potatoes out of the vegetable group and maybe thinking of it as a starchy food group where it might be more appropriate. So those would be a couple ways to implement it.

The ADA, the work of the Diabetes Association looked at the glycemic index and they found it difficult to deal with.

How in the studies that you've looked at do you look at pattern and say, well, maybe it's the pattern we have to be concerned about, to say, well, you know, you don't get to choose one guideline over another, you've got to take them all? And so that if you just take potatoes without variety or you just take this without the other, that in fact you can run into the sorts of problems that you've uncovered.

Is that something we need to be concerned about or do you really feel that, gee, we need to focus in on potatoes and other starchy vegetables because they are the culprit and not the pattern?

I think the details of getting across the glycemic index concept may be too difficult, but in broad brush strokes I think it's actually not difficult, and that would be an emphasis on whole grain, minimally processed grains; get away from this concept that just because white bread is low fat that therefore it's healthy.

And I think the potato issue is just that right now, according to the guidelines, it's considered a vegetable, and if you have a large McDonald's french fries, you've got four out of your a day vegetables according to the guidelines, and I don't think this is right, and I think we ought to consider -- we oughtn't to consider potatoes along with broccoli and carrots and other things that we think of as vegetables.

DR. GRUNDY: Ask a follow-up about the glycemic question. It seems like there might be two issues here.

One is the total carbohydrate load in the diet, which if you have a very high percentage of carbohydrate in the diet, then the problem, I think, with the American population that you point out, which tends to be sedentary, creates a metabolic stress on the insulin metabolism and so forth. And then the second is that the type of carbohydrate can accentuate that.

Is that what you're saying?

CHAIRMAN GARZA: Johanna. 

DR. DWYER: Meir, I'm very much interested in the glycemic load concept, but I must admit to ignorance on much of it.

First of all, how many foods have glycemic indices experimentally determined? Are there a lot or a few?

DR. DWYER: How many is a lot? 

DR. STAMPFER: Hundreds. 

DR. DWYER: Hundreds.

And how does the glycemic index of an individual food affect the glycemic index of meals. I thought years ago that it changed depending on the total meal for example, if you mixed all the foods together in a meal.

DR. DWYER: Just one other -- we just put together an issue, a journal on this topic, and it's very popular in Australia, for example. Australia doesn't have food labels like we do so, you know, it's harder to find out these things. I just wonder if in the future if we're going to consider this it might be possible to get a representative of the Diabetes Association or some other group of endocrinologist who deal with this every day. 

CHAIRMAN GARZA: I've got three people, Richard, Ellis and Rachel, so I want to assure everyone that I will get to the three of you in just a minute.

Richard.

And I think when we deliberate this afternoon, we'll be bringing this in to, you know, where it should be, if it should be, but where should it be and in what manner because I think we have to make sure that we look at the whole and not, you know, have some kind of topic that may bring up a major controversy which will affect the whole report.

We spent a lot of time on this, but I did bring up some other questions, so I'd prefer, if it's okay, not to discuss glycemic index anymore. We can bring it up again tomorrow when we report, but there were some other issues that we looked at as I reviewed, and that included nuts, and that included separation of grains from vegetable and fruits, and I'd like to get some comments from the committee before we break into our working groups this afternoon.

DR. LICHTENSTEIN: I was going to comment on the glycemic index, but what I'll do is first ask my question about nuts -- free speech. But just with nuts, are peanuts included with the nut group? 

DR. STAMPFER: Yes. 

DR. LICHTENSTEIN: Okay. Okay, peanuts are a legume and then nuts, most of the other nuts grow on trees.

(Laughter.)

DR. STAMPFER: Well, if you look at the composition of peanuts, it looks a lot like nuts, and so the common parlance of nuts, peanuts being considered nuts actually makes more sense than the laneon classification of where they all came from. I don't think we should be hung up on the -- 

DR. JOHNSON: Which components of the nuts? I guess, what aspects of the nuts? 

DR. STAMPFER: In terms of the protein and fatty acid composition of peanuts. They look like other nuts and peanuts, in our study we separated out peanuts from other nuts, and they are both have the similar effect. And you might say, well, what about peanut butter? And peanut butter, if it's just made from peanuts, presumably has the same effect. But a lot of peanut butter has trans added to it, trans fatty acids to keep the fats from separating, so that probably detracts from the benefits. 

DR. JOHNSON: But I guess that's -- I'm getting at something a little bit different. Is it specifically that people should increase their nut consumption or should they increase their consumption of a diet that's consistent with the fatty acid profile of the nuts as far as advice? 

DR. STAMPFER: Well, I think -- I don't know. I mean, these are the observations that people who eat nuts have a lower risk of coronary disease, and they also have a better lipid profile. You could get that lipid profile by feeding them oils instead of the nuts.

But I think the main message is a simple one, that we should remove the astigmia from nuts. Instead of considering them bad, because they are a high fat food, we should consider them according to what their health effects really are.

(Simultaneous conversation.)

DR. DECKELBAUM: Well, let me ask it in a different way. Is anyone against including nuts, you know, somewhere -- 

DR. DWYER: I have a question. 

CHAIRMAN GARZA: Let's make sure that of the ones are targeted. 

DR. DWYER: The question is, you know, I've heard about a lot of single groups, some of which apparently risks enormously in these analyses, some of which decrease risk by large amounts. Risk go down to .6, .4, whatever. How much of this is confounded? 

DR. STAMPFER: Oh, these are -- these are adjustments for -- 

DR. DWYER: I know they are, but I'm still asking the question. 

DR. STAMPFER: I think -- I mean, we measure diet, we try to assess diet, we assess coronary risk factor, we adjust as well -- you know, is there some residual confounded? Yeah, probably there is. Is it all correct? Well, it's a guess. These analysis I presented are after multiple adjustments. Obviously, you can't fully adjust -- 

CHAIRMAN GARZA: Meir, can you get closer to the microphone because people cannot -- 

DR. STAMPFER: The question is are these findings due to confounding, and obviously that's our bread and butter. We pay a lot of attention to confounding and try to avoid it as much as we can recognizing that there is residual confounding which could explain part of it. But it seems very unlikely that it can explain these effects to a very great extent, either the adverse or the beneficial ones. 

CHAIRMAN GARZA: All right, we are going to do everything. Right now every single committee member has their hand up, so I'm just going to go around the table.

Scott?

CHAIRMAN GARZA: Dr. Murphy. 

DR. MURPHY: Since I'll only get one turn, probably I'm going to make a couple of comments.

First, on the nuts issue, I think certainly nuts are a nutritious food, and I have no problem encouraging consumers to eat more nuts. I'm not sure I think they're a fruit or a vegetable or a grain, and I would prefer, if we can, to see the nut issue addressed as a protein food, and in the context then of variety or maybe in our introduction that is now going to be as long as the original report, but I don't -- I don't think it's necessary that it be in with fruit, vegetable and grains.

A second comment I'd like to address, I think, is whether "whole" should be in the grain guideline, and whether it's separate or whether it's combined. I think we need to be careful about discouraging consumption of non-whole grain products for many of the reasons that Richard has already summarized. But I do think it would be good to focus more on variety. If we're going to take it out as a separate guide, it should be emphasized more in these guideline or guidelines. And certainly if we could just get consumers to do 50/50 whole grain/non-whole grain, we'd be many times better off than we are right now.

So my personal preference would be to see some focus, some additional focus on specificity, particularly with whole grains. Could we even say try to make half your grains whole grains? That would be my preference.

And I would also like to see potatoes on the fruits and vegetable because the pyramid is based on certain calculations that assume that all your vegetables are not potatoes, and we don't want people to eat all of any one fruit or any one vegetable. And so at least in the text we say things like eat dark green or colored vegetables frequently or more or whatever. Maybe we need to be more specific about that, at least once a day or whatever. So I would encourage the group to think about both variety and specificity within these guidelines.

DR. JOHNSON: No.

Roland?

On the second issue very briefly, Meir suggests that perhaps potatoes, which seems to be the standout in terms of the vegetable group, be considered as part of the grain, perhaps the starch group, and one of the compelling reasons picks up on what Suzanne is saying, and that is that I think the figure yesterday was about 25 percent of the vegetable intake comes as potatoes, and I think the group should strongly consider putting it in the grain/starch group with consideration from people such as Suzanne, whether it in fact is more like a starch grain or is it more like a vegetable from a nutritional standpoint. So I'm just raising this for consideration. I think it's a good point to consider of people such as Suzanne, you know, think that it is more comparable to the grain group in terms of nutritional content than the vegetable group.

DR. KUMANYIKA: I think separating the fruit and vegetable guideline is a good idea. I wanted to emphasize the need to encourage fruit consumption because the data suggests that that's much more of a problem, vegetable consumption, I think, even if you subtract the potatoes.

The other issue that comes up, I think, most because of these extreme guideline is the range of servings. I'm now convinced, based on totally anecdotal evidence, that most people don't understand the servings as they appear in the pyramid, and it's very easy to elicit a conversation with a consumer who thinks that they should try to get 11 servings of grains, even if they're appropriate at the 1600 calorie level.

And the way the information is put in the book aggravates it in the same way that you have to go hunting. You have to really hunt for the information that that range means different calorie levels because I went looking for it, and I almost thought it wasn't in here for a minute, and then I found it. But it's very submerged. So however we do the servings, I think we should pick up that issue of what this number means for people who are eating different -- one base range and then say something else later.

And I would go for moving the potatoes. I think we have a tendency to use the guidelines to reenforce traditional wisdom rather than really tell the public things that make sense based on the way we look at the data, and I think moving potatoes would be a good sign.

DR. DECKELBAUM: Well, I think we would like some input from USDA. First of all, are we allowed to move potatoes?

(Laughter.)

I mean, I can just see -- I don't know --

(Laughter.)

(Simultaneous conversation.)

CHAIRMAN GARZA: You may not like the answer. 

DR. DECKELBAUM: The other point, I think the working group would like some input just to remind us on the history of why grains, vegetables and fruits are together. And I think, you know, this may have come up at previous meetings. And why was a decision made not to separate them. We need to know, we need to have that kind of information. 

CHAIRMAN GARZA: Let me, before turning it over to USDA, let me try once again to remind the group the pyramid is not our responsibility. You can make recommendations certainly. I mean, you can move potatoes, add t-bone steak, but that's what it is; it's a recommendation. 

DR. DECKELBAUM: Put steak with milk.

(Laughter.)

(Laughter.)

(Laughter.)

DR. KENNEDY: Can I before I answer that question talk about -- I think there is a fairly straightforward reason why potatoes are where they are, and it's based on the -- I think, Richard, your question or comment -- the nutrient profile of potatoes fits more in the vegetable group than it does in the grain group.

Now, having said that, Shiriki's issue, and I think it's one that as we in both departments think about how we're actually going to promote the guidelines once the next edition is released, I think there are serious issues related to how consumers see the guidelines.

I mean, an example: We all know that botanically tomatoes are fruit, yet we put them in the vegetable group. And why do we do that? The overriding reason, I think, we do it is because that's the way most consumers see tomatoes. They see them as vegetables rather than fruits.

But from the point of view of Dr. Garza's comment, well taken, that the guidelines as they will emerge guide the one part of any revision on the food guide pyramid that would take place, but they are only one part of it. The other two parts I keep coming back to are what will emerge from the new DRI because that clearly is the second key building block of the pyramid, and the third part is American's latest consumption patterns, which will be based on the '94 - '96.

So get back to comments made yesterday, Alice's point about what do you do with calcium fortified orange juice. Well, to the extent, in proportion to how it's showing up in consumption patterns, it gets fed back in our algorithm into the composites that are used to look at here is where we are, here is where we need to be as far as having an adequate diet. So it is very complicated.

I would suggest maybe if this plays to some of the issues which are being discussed, that maybe at the next Dietary Guidelines Advisory Committee we set aside a period of time where someone goes through what actually has to be done to modify the food guide pyramid because it is a very tedious, sometimes frustrating, but an awful lot of what the staff say to me is grunt work going into the revisions that are necessary for the food guide pyramid is not, as people who were involved in the '95 guidelines, it's not two or three people sitting down one day and pulling something out the air. There is a lot more thought that went into it.

DR. DWYER: I wanted to support provisionally the notion of thinking about two guidelines for fruits and vegetables and for grains, and express my reservations as a person of Irish descent at a time -- the potato suggestion. I'm not sure it isn't beyond our scope, and I'd like to hear a lot -- a lot more about the food composition, the reasons why it was put there originally before precipitous action is taken. 

CHAIRMAN GARZA: Well, remember -- 

DR. DWYER: I just don't have a -- 

CHAIRMAN GARZA: -- there is no precipitous action that will be taken at today's meeting. We are still in the information gathering stages, and all of these are just suggestions.

Linda and then Alice.

I'm blanking on what the 1980 said, whether it was the avoid times, but clearly it wasn't an avoid starch, but it was a starch and fiber emphasis at that time and I recall, so that was -- it was in the context of this was the first time you were making -- the government was making recommendations related at all to chronic disease prevention, so there was clearly attention to fats and lowering fats and lowering sodium and someone, sugars, I guess, and keeping carbohydrates and starch up, so everything was just sort of lumped there. But I'd be happy to bring those in so you can see it, unless Katherine may have them here.

CHAIRMAN GARZA: Okay. 

UNDER-SECRETARY WATKINS: I was not going to come to the table because I thought it was more fun sitting back listening to the comments, until you talked about separating potatoes out.

(Laughter.)

But I thought you better get to the table quick.

I think, Dr. Deckelbaum, you asked the question what kind of fire storm would this raise, and I think it would be a -- one that the Forest Service couldn't handle.

(Laughter.)

UNDER-SECRETARY WATKINS: I think the people on the other side of the house in the producing community, just knowing as much as I know about what they would say, I think you'd have tremendous opposition to moving potatoes from a vegetable/fruit category and putting it into some other category. I think you'd have -- if you think about how you're going to move the dietary guidelines, if that question is raised, and if anybody gets any inkling that you're talking about moving it, I think you're going to have a lot of controversy about the dietary guidelines.

And whatever you do on that issue, you would certainly want to make certain, if you talk about moving it, that it truly is science-based and that you have some real rationale for moving it.

The other issue that I will mention very quickly is how you connect the dietary guidelines and the food guide pyramid, and there seems to have been a disconnect. People really don't know. If you ask the question, do you know what the dietary guides are, they really don't. Do you know what the food guide pyramid is, yes. We need to connect the two, and what we're hoping is that when the dietary guidelines are approved that the food guide pyramid's revision will be right on the heels of that. I really would like to see them come out almost simultaneously. Then people start to connect the two. Then we can do something about the changes in diet of people.

But when they are two separate entities, and people are not connecting them, I think we have a real problem in this country, and we need to address that. So we'd like to look at how can we move the two simultaneously, and I know the staff is just about dying when we talking about that, but I think that's kind of the way it needs to move.

(Laughter.)

UNDER-SECRETARY WATKINS: I'm going to pass. 

DR. DECKELBAUM: In terms of separation of the guidelines, is that something that you would think would -- would be -- would there be -- do you envision opposition from certain groups there? 

UNDER-SECRETARY WATKINS: I don't think so. I think anything you can do to make it easier for people to understand the dietary guidelines is going to be acceptable. I think anything that helps make it easier and helps us to get the message out, I think, is going to be critical. 

CHAIRMAN GARZA: All right, on that positive note we have five minutes for our break.

(Laughter.)

(Whereupon, a recess was taken.)

VOICE: You could say 100 percent safely. 

CHAIRMAN GARZA: So we should have a lot of stressed individuals.

I'm going to challenge Dr. Johnson to get us through this one very, very efficiently.

CHAIRMAN GARZA: And the committee to be yet more insightful and inciseful in your comments. 

VOICE: Incise? 

CHAIRMAN GARZA: Concise.

(Laughter.)

DR. JOHNSON: Are we ready? Ready to roll? Okay, we're ready to roll with sugar.

First, I've chaired the subcommittee on the sugar guideline, and I'd like to thank Dr. Lichtenstein and Dr. Deckelbaum for their assistance and input, as well as the USDA staff, Dr. Garza and I'm very sorry that I left Shanthy Bowman off this slide because she was tremendously helpful in some work that we did that you'll see in a minute.

May I have the next slide, please?

The 1995 guideline said, "Choose a diet moderate in sugars," and the text elaborated by saying, "Sugar should be used in moderation by most healthy people and sparingly by people with low calorie needs."

Next slide, please.

the first thing the subcommittee did was grappled with the definition of what is a sugar, and according to the World Health Organization's 1995 report, or 1997, I'm sorry, report on carbohydrates in human nutrition, they say that "Sugars are conventionally described as the monoendysaccharides."

Next slide, please.

Unfortunately, this broad definition becomes problematic in the context of the dietary guidelines because it includes sugars like fructose, which are naturally present in high amounts in fruit, and lactose, which is present in dairy products. In fact, in the ILSI report done by Dr. Geiger, consumers reported being confused by what the guideline means by sugars, and they reported being uncomfortable with having one guideline that limits sugars and another in the same list of guidelines encouraging them to eat fruit which contain sugars.

Next slide, please.

Some of this dilemma, I think, led to the introduction of a number of terms which have been since developed to help further classify sugars. For example, in the U.K., the Department of Health uses the term "intrinsic" and "extrinsic" sugars to differentiate between those sugars which occur within the cell walls of plants and those which are added to foods.

Next slide, please.

The USDA has begun using the term "added sugars" when analyzing the nutrient intake of Americans, particularly with CSFII surveys. And added sugars have been defined by USDA as "all sugars used as ingredients in processed and prepared foods, such as breads, cakes, candies, soft drinks, jam and ice cream, as well as sugars eaten separately or added to foods at the table," and there is a list here specifically of which sugars are included in that definition of added sugars.

To me, the definition is quite clear cut and straightforward, and it has been well defined. It's important to note that added sugars do not include naturally occurring sugars, such as the lactose in milk or the fructose in fruit.

Next slide, please.

Next, I'd like to get into consumption or trying to answer the question exactly how much sugar is America eating. The Economic Research Service of USDA collects food supply data for caloric sweeteners which is comprised primarily of sucrose and corn sweeteners, including high fructose corn syrup, and total consumption has risen steadily since 1970, as you can see here. In 1997, Americans consumed on average 154 pounds of caloric sweeteners compared to the 122 pounds per person in 1970.

Next slide.

Using the USDA definition which I just gave you, added sugar intake varied with age and gender in the U.S. population. This chart shows the number of teaspoons of added sugar consumed by participants in the '94, '95, '96 CSFII surveys. Adolescent males had the highest intakes at 35 teaspoons of added sugar per day, and older females had the lowest intakes at 12 teaspoons per day.

Next slide.

Added sugar intakes ranged from 12 percent of total calories in females 51 and above, to 20 percent of total calories in adolescents, and this was true for both males and females between the ages of 12 and 18.

Next slide.

Now I wanted to move into looking at sources of added sugar. Clearly the most important source of added sugar in American's diets is regular calorie soft drinks, which accounts for one-third of all added sugar intake in the CSFII. Sugars and sweets were second in importance at 16 percent of added sugars and sweetened grains were third, contributing 13 percent of added sugars, regular calorie fruit aids and drinks were also important sources of added sugars, and together these four food categories were the source of three-fourths of all added sugar intake.

Next slide.

Okay, now, the next question is, okay, how does sugar intake relate to diet quality or the nutrient composition of the diet.

There were a couple of earlier studies which I show here that examine total sugar intake and nutrient adequacy, and the conclusion from these studies in both children and adults were that high amounts of total sugar do not necessarily lead to a poorer quality diet in comparison with consumers with low sugar intakes. It's important to realize that these studies included natural occurring sugar, such as fructose and lactose which are present in foods with generally high nutrient densities.

For example, in one study dairy foods contributed 31 percent of the total sugar intake in children, and fruits contributed 17 percent of the total sugar intake for all ages.

Next slide.

With the new CSFII database, there is now public access to data on the added sugar content of foods, and this allows for investigations into the impact of added sugar intake on nutrient quality. And with Dr. Bowman's assistance, we conducted analyses for the association of added sugar intake and diet quality in the '94, '95, '96 CSFII surveys. We statistically adjusted for age, sex and total energy intake to look at the association between added sugar intake and the total unsaturated fat, protein fiber, the essential vitamins and minerals, and we also looked at food groups.

And as you can see on this slide, after statistically controlling for age, sex and total caloric intake, added sugar intake was negatively associated with intakes of total unsaturated fat, protein fiber, the vitamins listed there, A, E, C, riboflavin, niacin, B 6 volute, B 13 and the minerals, calcium, phosphorus, irons, zinc and magnesium. It was also negatively associated with the number of servings of grains, fruits, vegetables, meats and dairy products.

And Meir and I were talking just before this, clearly what this shows is that when you add added sugar to the diet, which is not accompanied by any nutrients, any other nutrient will be negatively associated because you're simply adding empty calories, and so the impact will be negative on really any other nutrient, which is different than the analyses that you get when you look at total sugar intake, which is included -- which accompanies food that have some nutrient density.

Okay, I did want to point out one last thing about that slide, which was -- sorry -- that note the total unsaturated fat intakes are inversely related to sugar intake, and this is true both with total and added sugar intake. This has been referred to as the fat sugar see-saw such that as sugar goes up, fat goes down. But it's also important to notice that high consumers of added sugars were also more likely to have low intakes of shortfall or problem nutrients such as fiber, Vitamins A, C, folate, calcium, iron and zinc. So we have that interesting disparity there.

Okay, next slide, please.

I wanted to move into looking at beverage patterns among U.S. children because they've changed remarkably over the past decade, and I'm showing some data here from a paper by Morton and Guthrie that in the Family Economics and Nutrition Review just late in '98.

I know you can't see this well but it's just important to see the different bar graphs, and this shows what's happening in terms of dairy product intake; that basically low fat milk is remaining stable, whole milk consumption is going down, skim milk and other dairy product intake is going up slightly.

But it's important to note that overall milk consumption did decline in the period between '89 and '91, from 422 grams per day down to 396 grams a day in the '94, '95 surveys, so milk consumption is going down, and there has been some change in the type of milk that is consumed as well.

Next slide.

At the same time that milk consumption was declining major changes occurred in other beverage patterns. The largest increase occurred in the soft drink category, which increased from 198 grams per day up to 279 grams per day in the '94, '95 surveys. Male adolescents increased their consumption of soft drinks from a mean intake of 352 grams in '89 - '91, to 580 grams, which is almost 20 ounces a day of soda in '94-95, and this just shows the change and this is tea and breakfast drinks; soft drinks, you can see the big jump. This is fruit, aids and non --

DR. JOHNSON: Thank you, Joanne. Other drinks. The author is over here so I knew she would know.

And on that note, I wanted to share with you an article which Dr. Lichtenstein very kindly brought to me. This is from the Boston Globe, March 1. it says, "Here is the so-called problem, the kids in the Colorado Spring schools just aren't drinking enough Coke, or so says John Bushy, an area superintendent, for 13 schools who signs his correspondence `The Coke Dude.' The Colorado District was hard up for money for extras like band competitions and debates, so in 1997, they signed a 10-year contract in which it would get eight to 11 million dollars from Coca-Cola in return for giving the soft drink giant exclusive rights to sell Coke and other beverages in school vending machines. Sales of Coke products have been so sluggish that Bushy wrote to school officials in September, and I quote `We need to all work together to get next year's volume up to 70,000 cases,'" and the articles goes on. But this is the situation that's occurred in some school districts in our country with regards to soda and access by your young people.

Next slide, please.

Why are we concerned about this change in beverage patterns in U.S. Children? Calcium is concerned a problem nutrient among most age and sex groups in the U.S. Particularly problematic are adolescent and adult women, the majority of whom do not meet current calcium recommendations.

There has been research, some research by Gunthur, who was with USDA, established that carbonated beverages tend to displace milk in the diets of teenagers with negative implications for diet quality. This displacement effect has also been shown in adults. Joanne Guthrie found adult women whose diets failed to meet calcium recommendations, drank significantly more regular calorie sodas than those with diets meeting recommendations.

And as we've discussed yesterday, the DRIs did recently increase calcium recommendations over and above the 1980 RDA. So in my view or in the subcommittee's view, this ongoing trend for calcium-rich beverages to be displaced by beverages high in added sugars is a concern.

Next slide.

There is little evidence suggesting diets high in total sugar promote weight gain when consumed in amounts that do not exceed energy requirements. There is some evidence, however, that soda consumption is a major factor in the increased energy intakes of children and adults between the '89, '94 and '94-95 USDA surveys. And, in addition, a meta-analysis by Dr. Rick Matters at Purdue suggested that beverages high in carbohydrates have a low society effect, leading to a poor regulation of energy intake and subsequent weight gain.

Next slide.

In terms of sugar and diabetes, we looked at the Nurses' Health Study report which has already been talked about today, so I won't elaborate on that. Basically, at this time the sugar subcommittee felt that there was a paucity of evidence making it difficult to determine diets high in sugar ar linked with the etiology or causality of non-insulin dependent diabetes. There are many papers on the use of glycemic index for the treatment of diabetes but a real scarcity of papers on the actual etiology.

I did want to point out there is a paper in this month's issue of Pediatrics, which I will get and circulate to the committee. It was done in Susan Roberts lab in Tufts, and she demonstrated that when teenage boys consumed a lunch with a high glycemic index, they consumed nearly twice as much food afterwards in comparison with a low glycemic index lunch. And they suggested that meals with a high glycemic index set off a chain of actions that caused overeating and potentially could lead to subsequent obesity, so that's a new paper that I think we'll want to consider in our deliberations.

Next slide.

Again, I won't spend a lot of time on this. Dr. Byers mentioned this earlier, and I mentioned it in my comments that the World Cancer Research Fund is recommending limited consumption of refined sugars for cancer prevention.

Next slide.

Very quickly, because I mentioned this in September when we got together, there has been a meta- analysis of 23 studies over a 12-year period, leading to the conclusion that there is little evidence that sugar has any significant influence on either behavior or cognitive performance in children. So this idea of sugar and hyperactivity in children has pretty much been put to rest, at least scientifically, with these data.

Next slide, please.

Clearly, there is a role for dietary sugars in the development of dental carries and between milk consumption of sugar remains a risk factor for occurrence of dental carriers, and the recommendation now is that we focus on fluoridation, adequate oral hygiene and not just on sucrose intake alone.

Next slide.

Wanted to show you some data on low calorie sweeteners. The USDA Economic Research Service collected food supply data on low calorie sweeteners from 1970 to 1992. There are no data available after '92, primarily because low calorie sweeteners are used as constituents in other products like soft drinks and food manufacturers consider this proprietary information and it's difficult to get. But between 1970 and '92, consumption increased from an average of five pounds per person to 24 pounds per person per year, so there has been a dramatic jump in the use of sugar substitutes.

Next slide.

When we did a literature study for sugar substitutes, we find this one study in George Blackburn's study done in George Blackburn's lab that showed some evidence that multidisciplinary weight controlled programs that included Asparte, a sugar substitute, enhanced or facilitated long-term weight maintenance, but we could only find that one study so the evidence, again, is fairly sparse.

Next slide.

So, in review, I'd just to review these key points that I've raised. There is consumer confusion about what we mean when we say to eat diet moderate in sugar, and it particularly seems to be problematic as it relates to the fruit group. Since '975, USDA has defined, created a definition of "added sugar" which now allows us to do analyses of food consumption data of the USDA database which look at added sugar and its impact on diet quality. Sugar intake is clearly increasing. A third of all added sugars now comes from soft drinks. I showed you some new data that we had on sugar and diet quality, the sugar/fat see-saw which I talked about, and the fact that added sugar intake was negatively associated with a number of those problem or scarcity nutrients in the food supply.

Next slide.

And I also reviewed the evidence that we've looked at with sugar and weight, diabetes, cancer, behavior, dental carries and sugar substitutes. So thanks very much.

I think in terms of the committee and what it would be nice to have some comments about are what we're really grappling with is this question of added sugars and whether or not the guideline needs to reflect the nature of some of the data that I've showed.

Thanks.

DR. LICHTENSTEIN: I think you did a nice job of summarizing the group's work.

I think that -- I agree, I think that there needs to be a mechanism for distinguishing sugar that comes from fruit and milk from other kinds of sugar, and added sugar really sounded like a way of doing that.

However, what I think we really need to know is what the public's perception is of the word "added sugar," if it's just sort of sprinkling it on some breakfast cereal, and it seems like the major contributor in soft drinks, and I wonder how that's actually perceived terminology was because that may not be perceived as added sugar, and whether there are any suggestions on alternate terminology that would really capture that issue.

CHAIRMAN GARZA: You want to respond now? 

DR. JOHNSON: Well, I think that's a good point, and I don't have the answer. I know in the focus group study they said they're confused how the general public would perceive the term "added sugar." 

DR. LICHTENSTEIN: I just think we need to get the information. 

DR. JOHNSON: I think that's a good question, yeah. 

DR. LICHTENSTEIN: Yeah. 

DR. JOHNSON: I think I'll sit down if that's all right. 

CHAIRMAN GARZA: Meir. 

DR. STAMPFER: I think it's clear that the added sugar is having a big impact on diet even though we can't pin much in the way of specific diseases to this, but its adverse effect in -- main adverse effect is that it's displacing foods that do provide nutrients. So I think this might make a good model case to consider as one of the sort of second tier guidelines, if we got to that proposal that several people had made of distinguishing sort of top tier and second tier guidelines, that this, I think, should remain as a guideline, and perhaps the text could be shrunk a little bit, and it could go into the second tier where maybe some other guidelines might go. 

CHAIRMAN GARZA: Any other? Johanna? 

DR. DWYER: Just a -- thank you for an interesting presentation.

One thing it seems we need a little more work on is cariogenecity, and it seemed to me in a brief review of the literature a couple of months ago that it might be helpful to think of all of these variables, these various diseases as which ones seem to be most associated with whatever this thing is, extrinsic or whatever you want to call it, and what isn't.

Certainly with cariogenecity there are two issues. One is something to do with the composition of food with respect to sugars. Another is the carbohydrate, the cooking of the carbohydrate so that the starch is also associated with cariogenecity. And then there are all of these other things about when you eat it, what you do afterward and so forth. Do you have a toothbrush afterward?

DR. DWYER: It would seem to me it would be helpful if we could just array them as we ponder this issue.

The other thing, Rachel, was I wasn't sure what you were -- the group was suggesting. Is it change the existing guidelines to choose a diet moderate in added sugars?

DR. DWYER: And what are the others? 

DR. JOHNSON: We're looking for help. That was our primary option that we considered, was whether or not we wanted to use that term "added sugar" in the guideline itself.

I mean, I suppose another option is do we need a sugar guideline.

DR. KUMANYIKA: I think that the presentation is very convincing that a guideline is needed on something like food and beverages with added sugars, which is different from what it says now, which is "moderate in sugar." So you get into the "avoid" issues. And if you said then you're back to the limit, foods and beverage with added sugar, but something in that spirit is much clearer this round than I've remembered in the past because it's very clear that it's a displacement issue and it's not that you're trying to link sugar in the diet itself to the health problems, but it's part of the pattern, that it's not -- that it's replacing things that are needed, so maybe that word could give someone an inspiration for how to -- 

DR. JOHNSON: Say that again, Shiriki? Food is? 

DR. KUMANYIKA: Foods and beverages with added sugars, because then it doesn't matter who adds them. People know it's been added. 

CHAIRMAN GARZA: Any other comments or suggestions?

Dr. Dwyer?

And the second things is I'm having trouble remembering all the things you said and which ones are related to "total." For instance, cariogenecity, I know from our own work that vegetarian children who eat a lot of raisins get just as high carries as kids who were non-vegetarians who ate a lot of added sugar. So that it would help me to array those things and think of them before I made a decision.

DR. DECKELBAUM: I guess added sugars was a major part of our discussions and how to define it. And I guess in the simplest way, because it's added sugars by food producers or its added sugars in the home. But I think that it's where you add simple sugars, as a first step add simple sugars to food, to sort of natural food either in home or industrially. So I guess if you were a soft drink manufacturer and you put it in the mix, that would be an added sugar. Similarly at home by adding it to tea or coffee or whatever would be an added sugar.

We didn't really discuss too much about corn syrup and corn syrup solids, but, again, those added sugars, and we had some discussion on that yesterday, are in large part small glucose polymers. Corn syrup, I think the mean sort of size of a glucose polymer is about 15 glucose molecules together and they're not very sweet, so they're not added -- they're not added --

DR. DECKELBAUM: They're not added for sweetness. 

DR. DWYER: Okay. I take issue with the importance, I take issue with that statement, but I also take issue with the question that they are not important. I think they are very important and we need to see some breakout data because it strikes me that a lot of -- 

DR. DECKELBAUM: It's not important. 

DR. DWYER: A lot of the corn syrup solids, sweeteners, I think a lot of what's added to the soft drinks that Dr. Rachel mentioned are corn syrup, it's not -- isn't it? 

DR. JOHNSON: Well, remember that soda is the number one source. 

DR. DWYER: So they're not added to the make up --

(Laughter.)

DR. KUMANYIKA: I just wanted to pose the question of if the goal of this guideline would shift to being foods and beverages with added sugar or sweeteners or whatever, then it might not be the place to address some of the issues of total sugar or total carbohydrates. It's just -- I mean, we think about some carbohydrate issues as part of another guideline. If this one could be clearer, focusing on this added sugar displacement problem. 

CHAIRMAN GARZA: Carole, is there either plans within the department to do some focus groups -- 

MS. DAVIS: Yes. 

CHAIRMAN GARZA: -- before, that one could explore at least the meaning to consumers of added sweeteners, added sugars, a variety of various messages such as the one that Shiriki suggested? 

MS. DAVIS: Yes, we have plans -- I don't know if this is on or not. We have plans to do that. it's going to be very limited, and we're using this to get other things that you want to -- to have us study, and we just hope the timing will be right. We're in the process now of going through all of our clearances that we have to do. 

CHAIRMAN GARZA: Because what the group might want to do is to focus its attention on the rationale for modifying the current guideline either as a first or second tier, and then coming to some judgment later on, based on that evidence and the added input of the focus group, as to what would be the best wording, and then making a final recommendation based on the basis of both types of evidence. Is that -- 

DR. LICHTENSTEIN: I think that's a very good idea. I think when the subcommittee was deliberating, you know, we came up with the word "added," and we all -- we're just abnormal, but it all seemed real clear to us what we were, you know, talking about, and it wasn't until I actually heard the presentation and then relooked at the contribution of sugar that I realized that it probably wouldn't be perceived as added sugar, and I also think we need to find out more about corn syrup because if you taste corn syrup it's sweet. If the mean polymer size is 15, one wouldn't predict it to be -- 

CHAIRMAN GARZA: That's a smaller size though. It's not -- 

DR. LICHTENSTEIN: Right. One wouldn't predict thought that it would be sweet because even the monosaccharides would be glucose, which have less sweetening, relative sweetening that fructose. I just think we need to find out, but that's something that's sort of a factual thing. But it seems in a lot of the food labels that you look at, a lot of it is corn syrup and not sucrose. 

CHAIRMAN GARZA: Okay. Can we move on to the next one? 

DR. DWYER: It seems to me it depends on what the meaning of "is" is.

(Laughter.)

(Laughter.)

All right, let's move on then.

Kathryn?

CHAIRMAN GARZA: You're saying that the term "added sugar" is on the food label? 

MS. MCMURRY: No. Total sugars. 

CHAIRMAN GARZA: Total sugars. 

MS. MCMURRY: Total sugars is what's on the label, although there is -- 

CHAIRMAN GARZA: We'll get somebody to clarify.

All right then let's move on then to another noncontroversial guideline. Dr. Kumanyika is going to provide us with untold wisdom in about 30 minutes. In 30 minutes we will have this guideline resolved.

CHAIRMAN GARZA: At the present time there is no lunch. 

DR. KUMANYIKA: Okay. 

CHAIRMAN GARZA: But we'll see what we can do. 

DR. KUMANYIKA: All right. Okay. 

DR. DECKELBAUM: It's the hour and a half that we went over. 

DR. KUMANYIKA: The sodium subcommittee consists of myself and Drs. Dwyer and Stampfer, and Joan Lyon, who is giving us staff support and very helpful in keeping us on track.

What I'm going to do is just highlight what we've been doing and some of the issues that we think are important for sodium this time. If we leave that one for awhile, the key issues in sodium are whether to keep the guideline at all, and then if we decide to keep it, what should it say. And the reason I say whether to it at all is because it has been suggested, for example, yesterday that we drop the guideline, so I think the committee needs to consider that since the challenge is put before us of whether this guideline should continue.

What we did after the last meeting was to go through and identify all of the issues that had anything to do with either the validity of the guideline or the reasons for supporting the guideline, the evidence. And we had the sort of unusual situation of having the National Hot Line of Blood Institute convene a conference on this topic between the two meetings. So rather than -- and I certainly am not going to try to summarize the proceedings of the conference, because I think we needed to have a two-day conference where, fortunately, our entire subcommittee was present for the whole time, so we've heard a review of the evidence on the topic. I'm going to go through what was covered at the conference first. Then as Drs. Dwyer and Stampfer to give their impressions of whether -- the question I posed to them is whether anything that they heard in the presentations at the conference alters their understanding of the supporting evidence.

Before I do that, I want to read off a list of questions that I raised about this guideline because there is a sense that this case may not be strong or may not be existent. So one of the questions is if the cases for sodium reduction as a guideline is not strong, why is it not strong? And here is the list. I don't have this on the side.

One is, is it not strong compared to that for other guidelines, which would mean it's a second tier type of guideline or something? Is it not strong compared to the case for other dietary factors that might influence blood pressure, which is a different issue? And some of the arguments that are made are that it's not strong because by itself it won't do as much as other factors and therefore we don't need it.

Thirdly, is it not strong simply because the methodology to address the issue is limited, which means that people think it's there but because urinary sodium, 24-hour urinary sodium is hard to collect and dietary doesn't get sodium? It just not strong because the type of evidence we have is by its very nature inconclusive. Another reason it might not be strong is because the studies that would tell you if the case is good haven't been done yet, and I'm going to mention a little bit later one of the studies that's currently in the field. Or is it not strong because the evidence is actually equivocal and, you know, the case isn't there?

And, finally, is it not a strong case because there's a lot of noise in the system, which I take to mean to mean that it is actually a strong case, but there are -- people put noise in the system by trying to cite odd and invalid evidence to confuse whether there is a strong case or not.

So why don't we put up the next overhead.

I think this actually summarizes the sense that we had at one of our conference calls, but this is still open to discussion; that the guidelines says there is a role for sodium reduction in the general population, and that there is a defensible case for that based on the evidence.

At the NHLB conference, which was chaired by Drs. Martha Hill and Aram Chobanian, we had an overview of the relationship between sodium and blood pressure by Dr. Paul Whelton; differences in the blood pressure responsiveness to sodium intake, which was the salt-sensitivity issue, by Dr. Myron Weinberger, and that was discussed -- there were discussants: Drs. Margo Denke, John Flack and Steven Hunt; and Dr. Frank Sacks talked about the interactions between sodium, potassium, magnesium and calcium, and gave a little DASH advocacy at the end of his presentation. He's one of the main players in the DASH study and we had that discussed by Drs. Ted Kotchen, David McCaron and Laura Svetkey.

As you might know if you know this literature from some of the names, we had almost all of the antagonists and protagonists there except that, as Dr. Lenfant mentioned in his opening remarks, Dr. Mickey Alderman chose to be out of the country during the days that we had the conference. That's the way that Dr. Lenfant put it. And so he was not there. Someone else presented some of his data that he's published.

We went to sodium and blood pressure in the young. Dr. Ron Prineas talked about the effects of neonatal sodium intake. Dr. Bonita Falkner, Sodium and blood pressure in children, which was discussed by Dr. Gerry Berenson, Clarence Grim and Julie Ingelfinger.

We looked at the clinical trials and studies. Dr. Graudal presented his meta-analysis of trials of sodium reduction. Richard Grimm talked about subpopulations by age, race and gender. Janice Douglas, prevalence of sodium sensitivity in postmenopausal women, and I talked about sodium reduction and quality of life issues, and we had discussants including Larry Appel, David Freedman, Diana Petitti, James Robins and Michael Stoto. And by this time it was really a very lively meeting with, I think, people who hadn't all been in the same room together to discuss these issues actually having a chance to hear each other's arguments and agree with them or rebut them at that time.

Paul Elliott presented by INTERSALT data, and then Dr. Chris Sempos talked about the cohort studies that have some long-term data and mortality data, so he presented Alderman's Work Site Cohort Study and the NHANES I analysis that Dr. Alderman has published, and the Scottish Heart Health Study data, so those were the three -- three of the studies that have relationship, some relationship between sodium and mortality.

And then Dr. Jerome Cohen presented the mortality data from the MRFIT-Follow-up. So that's a fourth source of mortality data; and we had Dr. Nancy Cook, Kesteloot, Graham MacGregor and Louis Tobian to discuss that.

The role of sodium in non-cardiovascular conditions was a topic on the second day. Dr. Dan Jones addressed that. We talked about sodium in left ventricular mass, Richard Devereaux, and Jay Cohen, Ed Fruhlich and Lew Kuller discussed that.

Dr. Suzanne Oparil was at the -- gave a presentation on the renin-angiotensin system. The sympathetic nervous system was discussed by Allyn Mark, and then Plasma insulin, cholesterol and coagulation factors. So this was really an exhaustive discussion not designed for the Dietary Guidelines Committee, but certainly more than we ever wanted to know about the details of the topic.

The basic research in the area was discussed and the future studies, clinic and epidemiological research and the question was posed as a discussion topic of whether there should be a randomized clinical trial on sodium reduction and blood pressure with morbidity or mortality as an outcome.

At the end we had a presentation by Dr. Michael McGinnis on dietary guidance public policy, and Drs. Dwyer Stampfer both were discussants, Dr. Bill Harlan and Marion Nestle all talked about some of the policy considerations.

In the hallways, some of the people who were present actually revealed that they thought it was a consensus conference, and they thought we were going to take a vote at the end of the conference, and some of the NHLBI people, and Dr. Hill, certainly made it clear that the purpose of this meeting was not to have a consensus, but it was just to really describe the evidence and have people have a chance to hear the arguments and hear the arguments about the arguments so that we could make up our minds.

So I think I'll stop there and just ask for whatever comments that the other two subcommittee members from the DJAC want to make, and then I'll go into the talk about what we would have in the guideline if there will still be a guideline.

Do you want to come up here or use your mike there? As long as people can hear you.

The first was the level of heat was considerable, the level of vipe was also considerable in this conference, but I thought that the data were very well reviewed. Some impressions I had was that one of the problems is the data aren't as strong as some of the other things we've been talking about, like saturated fat. Nevertheless, I think what your side says, Dr. Kumanyika, is correct; there is a role. The problem is not exaggerating on either end of the debate.

That slide from DASH is interesting, and certainly it will be of great interest to follow the community-based intervention, but a feeding study with the facts is not the same as what we're after, which is more what happens if you give dietary advice or recommendations to federal officials or other officials. Nevertheless, it does seem to be safe. Any concerns I had on that were dispelled by what I heard.

The question is how effective it is and how big an effect, what people can do, can make.

Meir?

I think the evidence is actually pretty good for modest effect, and I think the evidence for big effect is weak. So my take on it is that I agree there is a role for sodium reduction in the general population, and where to rank it in the hierarchy of advice is open to question, you know. If we want to tell Americans some number of things that they can do in their diet to improve their health, I think sodium fits in there, but it's not in the top tier of recommendations.

So my bottom line take would be to keep the guideline, perhaps shrink the text, and like with sugar, to put it in the second tier. But I wouldn't weaken it in the sense of relaxing the recommendation at all. I think the evidence is good for a modest but important effect on a population level.

So if there will be a sense that there is no scientific evidence that's been put forth since 1995 that warrants dropping this guideline, then the question is what should we say in the supporting evidence, and is there something we can put in the text that would make it clearer to people.

Put up the next.

So we went through the points that are made in the current guideline, and have just summarized the things that the committee has discussed and will be working on according to those points.

These three bullets, "Sodium and salt are found mainly in proc