|
|
In the Matter of:
Dietary Guidelines Advisory Committee Meeting Transcript
Wednesday, June 16, 1999
Economic Research Service
1800 M Street, N.W.
Waugh Auditorium
Washington, D.C.
Pages: 1 through 317
HERITAGE REPORTING CORPORATION
Official Reporters
1220 L. Street, NW, Suite 600
Washington, D.C.
(202) 628-4888
The meeting in the above-entitled matter was convened, pursuant to Notice, at 9:11 a.m.
IN ATTENDANCE:
CUTBERTO GARZA, M.D., Ph.D.
CHAIRMAN
Vice Provost and Professor, Cornell University
Associate Director, Food and Nutrition
Programme, United Nations University
EILEEN KENNEDY, D.S.C.
SUZANNE P. MURPHY, Ph.D., R.D.
Researcher, Cancer Research Center of Hawaii
University of Hawaii
LINDA MEYERS, Ph.D.
HHS/OPHS
CAROL W. SUITOR
SCOTT M. GRUNDY, M.D., Ph.D.
Chair, Department of Clinical Nutrition
Director, Center for Human Nutrition
University of Texas Southwestern Medical Center at Dallas
SHIRIKI K. KUMANYIKA, Ph.D., M.P.H., R.D.
Associate Dean for Health Promotion and Disease Prevention
University of Pennsylvania School of Medicine
Center for Clinical Epidemiology and Biostatistics
ROLAND L. WEINSIER, M.D., Dr.P.H.
Chair and Professor, Departments of Nutrition Sciences and Medicine
School of Medicine
University of Alabama at Birmingham
LESLEY FELS TINKER, Ph.D., R.D.
Assistant Member, Fred Hutchinson Cancer Research Center
Affiliate Assistant Professor
Department of Health Sciences
University of Washington
JOAN LYON, M.S., R.D., L.D.
HHS/OPHS
ALYSON ESCOBAR, M.S., R.D.
USDA/CNPP
SHANTHY BOWMAN, Ph.D.
USDA, Agricultural Research Service
KATHRYN McMURRY, M.S.
HHS/OPHS
RACHEL K. JOHNSON, Ph.D., M.P.H., R.D.
Interim Associate Dean, College of Agriculture and Life Sciences
Associate Professor, Nutrition Food Sciences
University of Vermont
JOHANNA DWYER, D.Sc., R.D.
Director, Frances Stern Nutrition Center
New England Medical Center
Professor of Medicine (Nutrition) and Community Health
Tufts University School of Nutrition
RICHARD J. DECKELBAUM, M.D.
Director, Institute of Human Nutrition
Columbia University College of Physicians and Surgeons
ALICE H. LICHTENSTEIN, D.Sc.
Professor, Tufts University School of Nutrition Science and Policy
Senior Scientist, Jean Mayer USDA
Human Nutrition Research Center on Aging
Tufts University
MEIR J. STAMPFER, M.D., Dr.P.H.
Professor of Epidemiology and Nutrition
Harvard School of Public Health
Associate Professor of Medicine
Harvard Medical School
CAROLE DAVIS, M.S., R.D.
USDA/CNPP
DR. STEVE BLAIR
Cooper Institute for Aerobic Research
Dallas, Texas
DR. RUSSELL PATE
University of South Carolina
Columbia, South Carolina
DR. DAVID LUDWIG
DR. ANN SHAW
DR. GARZA: Good morning. I want to welcome everyone, especially the committee members who have a busy three days ahead of them. And also, Dr. Lesley Tinker, welcome to the committee. Dr. Tinker was appointed at the beginning of the process, but was on sabbatical leave and chose to respect that. I don't understand why. But obviously she is bright and we're -- we're always happy to see bright people in this group. So welcome.
We also have three guests that I want to welcome, Dr. Blair, Dr. Pate and Dr. Ludwig today. And they are going to be presenting some information for the committee in terms -- that -- that should help our deliberations. I also want to thank both Dr. Meyers and Kennedy for taking time out of their busy schedules to join the committee.
And with those very brief introductions and welcomes, I would like to -- to get started because we do have a busy three days ahead of us. And judging from our last committee meeting, we likely will be working until the very last moment. So that -- to help assure ourselves that we're going to give as much attention to the beginning items as to those at the end of the agenda, I would like to get started.
And so with that, let me welcome Dr. Steve Blair from the Cooper Institute for Aerobic Research from Dallas. Dr. Blair was asked to come and discuss the issue of physical activity and weight loss and weight maintenance for adults as the committee considers separating the weight gain or weight maintenance guideline from its physical activity components and developing two guidelines instead of one.
Dr. Blair, thank you for joining us.
DR. BLAIR: Thank you, Dr. Garza. Well, Dr. Garza, committee members, colleagues, it is a pleasure to come to the committee this morning. Obviously, I'm very excited about the possibility that you will consider having a separate guideline for physical activity. It obviously is the single most important risk factor for all kinds of diseases and the biggest public health problem in the United States. No reaction.
(Laughter.)
Either stunned, disbelief or complete agreement, I don't know which.
DR. GARZA: We are very well trained not to be responsive.
(Laughter.)
DR. BLAIR: I would like to discuss the four points on the slide with you this morning. First of all, I will review some of the evidence actually on cardiorespiratory fitness and health. And as you will see as I get into it, the large cohort that I follow, we have the I think good fortune of having data from a maximal exercise test on a treadmill as part of the baseline examination which I think the value of that is that it is an objective laboratory measure albeit of a physiologic parameter, cardiorespiratory fitness, the way you get to be fit or to be unfit by and large is to be active or sedentary.
So I think it is a marker for habitual physical activity. And, of course, habitual physical activity, just as habitual diet, is very difficult to assess with a lot of misclassification and so forth.
But as I present these data on fitness and health, I want to underscore or make clear that I think this is an indication of the association of -- of physical activity patterns to health factors. I will focus this specifically on how fitness might fit into the picture of overweight or obesity and health; give a very brief summary of what I think are the current consensus public health recommendations for physical activity; and conclude very briefly with a discussion of physical activity interventions.
This last point, kind of added after talking with Dr. Garza, the physical activity intervention area is a very new one. And I find health professionals and members of the public tend to have the view, you can't get people to be physical active. This is a research area that is really less than ten years old. So we haven't solved all the problems yet. But a lot of progress has been made. And I will review just a little bit of that for you.
I did send along several reprints that are included in the orange folder that the committee members have before them.
As I said, I'm going to start reviewing the association of cardiorespiratory fitness to various health outcomes. And these data do come from our own follow-up studies of a large cohort of men and women who have been examined at the Cooper Clinic in Dallas at least once during 1970 to 1989. These are men and women on the average in what I would call early middle age, although the age range spans essentially the adult years from about 20 to about 90.
I make no claim that this is a representative population sample. It's an opportunistic sample of people who have elected to come to this clinic for an examination. They are predominantly white. They come from mid -- middle to upper socioeconomic strata. Seventy-five or 80 percent of them are college graduates.
However, on some key risk factor indicators such as triceps skin fold, BMI, cholesterol, triglyceride levels, even physical activity patterns, they are very similar to other large North American cohorts such as in Hanes Surveys, the Canada Fitness Survey and the like. But certainly I recognize we cannot generalize from this one sample.
The first bit of data I will show you come from a paper published in the JAMA in 1996 which is in your folder. It's for mortality follow-up in this cohort through December 31st, 1989. Note that we have a reasonably large number of deaths in the men. Of course, we have a smaller number of women in the cohort and death rates in any cohort are always lower in women. So we only have 89 deaths in the women. And some of the analyses that we've done, some of the papers we've published, at this point we've had to limit to men simply because of not having enough power to look at the data in the women.
But as I will show you and in other analyses, whenever we have done similar analyses in men and women, we see virtually identical patterns, dose response pattern, between cardiorespiratory fitness and health outcomes.
As I said, all of these individuals had a maximal exercise test at baseline. We take the treadmill time from this standard protocol, put that in age and sex-specific distributions. And we have typically called the least fit 20 percent of the population as the low-fit or unfit group. When we look at three categories, the next 40 percent we've called moderately fit.
And I often wonder if high fit is the correct adjective to use for this top group because it's 40 percent of the distribution. So high fit doesn't mean three percent who are marathon runners or are triathletes. And as you'll see for some analyses, we simply compare the unfit or the low fit with the fit which is everybody else.
Here are the cardiovascular disease death rates by these three fitness groups for men and women. And as you can see, there is a fairly steep inverse gradient across fitness categories in both women and men. And one of the striking things that we first observed, oh, nearly a decade ago is the very large difference in death rates that we see between the low and moderately fit groups which has led us to believe that people at the lower end of the activity or fitness spectrum can actually get a fair bit of health benefit, a good bit of protection from going from doing nothing to doing a little bit. And you'll see some of that thinking reflected later in the consensus public health guidelines.
Now, this paper that Ming Wei published in January of this year in the Annals of Internal Medicine was dealing with the issue of physical fitness as it might relate to the development of Type II diabetes. In the last five or six years, it has been concluded -- was concluded in the Surgeon General's report on physical activity and health that a sedentary way of life does lead to the development of Type II diabetes.
The studies available at that time had used self- report of physical activity as the exposure and self-report of a clinical diagnosis of diabetes as the outcome. Since we know that it's very difficult to measure activity by a questionnaire, we also know that about half the prevalent cases of Type II diabetes are unknown. That leads to tremendous rates of misclassification for both exposure and outcome, and led us to believe that the current studies, although showing an association, probably were under- estimating the magnitude of the association.
In this report, we have about 8,600 men with two examinations. So we have objective measures of fitness at both exams. And for the definition of diabetes, we use the American Diabetes WHO criteria that you see on the slide so that we could screen out all people at the beginning who had diabetes and then at follow-up use fasting plasma glucose criteria as the indicator of diabetes which at least is an objective marker, whether or not you agree specifically with these cut points.
During the six-year follow-up, 149 new cases of diabetes developed in these men. Again, we saw a very steep inverse gradient across low, moderate to high fitness groups. The low-fit men had relative risk for developing diabetes, about 3.2 times higher than the fit men. And this is a much stronger association than had been reported in the physical activity diabetes studies. And that's after adjustment for kind of all of the confounders that you think might influence the relationship.
And I show you just one slide from this study. Again, you do have the full paper in your handout, but showing that this association is similar across overweight men. At the time we wrote this before the NIH treatment guidelines came out, we used the BMI cut point of 27.
But you see a similar pattern of results in the lighter men and in the heavier men. And in fact, the lighter men who were -- or the heavy men who were high fit actually had about the same if not a little lower rate of diabetes development than the unfit men who were in fact normal weight.
I want to follow up then on that suggestion of the data from the last slide and go a little more into the fitness, overweight or obesity, and health area. Certainly, we have enormous -- an enormous database indicating that as one moves up the BMI scale, we see higher rates of many different chronic diseases, higher death rates, especially from cardiovascular disease.
But much of that work has not really taken into account the physical activity status of individuals in the studies. And I don't think any of them had objective measures of fitness. So this is a line of work we've been pursuing in recent years. This reprint, published in the AJCN in March of this year, is also included in your packet.
In a cohort of about 22,000 of our men, we have measure or estimates of body composition and fat distribution from hydrostatic weighings on the seven skin folds and waste circumference and, again, cardiorespiratory fitness as assessed by a maximal exercise test. In this subgroup of the total cohort, 428 deaths.
Now, what we did in this analysis or the analysis shown on this slide is to sort the men into three categories of percent body fat as you see on the slide: less than 16, 16 to 25, and more than 25 percent.
And in each body fat in a stratum, further divided them into the fit and unfit men. And as you can see for all cause mortality, we really don't see higher mortality rates in the men who were fit, whether they were lean and a normal body fat or in fact obese. It's the unfit individuals in each stratum who had the higher mortality rates. In fact, the fit obese men had much lower mortality during this follow-up than did the lean men who were unfit.
Cardiovascular disease mortality, we do now see that there is a progression even in the fit men across fat in this category. So I'm not here to claim that body fatness makes no difference in terms of health. But certainly, these data do suggest that at least activity and fitness seem to ameliorate the effects of -- although these data aren't on BMI, we've looked at that, as well -- but BMI body composition and mortality.
And in this analysis, we actually tried to focus a little more on kind of a pure fitness measure. So we look at the bottom 25 percent of the estimated VO2 max. per -- in ml per kg of fat-free mass. And here you see that in the fit men, there is quite a strong trend upward with the fit obese men having nearly three-times higher risk of dying from cardiovascular disease.
But again, in every stratum, the fit men do much better than the unfit men. So I think these data strongly suggest -- and actually, I think we're the only group that has looked at this specific issue in this way, although you can ferret out from other reports analyses that looked at state of BMI by levels of physical activity, and those results are consistent with what I am showing you here.
The differences in most of those other reports is they didn't set out to do this kind of analysis. The top stratum of BMI in virtually all the reports tended to be 25 or greater, 26 or greater. So overweight, mild overweight perhaps and low levels of obesity. Whereas we go on up. In fact, the BMIs in our group range on up to about 35 or so.
But let me say also, I don't think the pattern of results that I've shown you are likely to apply in Class II or Class III obesity. They do apply in Class I obesity, at least in this set of observations.
And now some new work that is not past peer review. We have a paper currently out for review. We've extended the mortality surveillance in our cohort through 1994. And I'll show you in this analysis data from 25,000 men that have been followed for about ten years used and about 1,000 deaths in the cohort, more than 400 from cardiovascular disease using now the NIH guidelines BMI categories for those definitions, and has -- no surprise to anyone in the room, of course -- as we look across the BMI categories with the normal weight as the reference category for cardiovascular death or for all-cause deaths, we see the sort of progression that many others have reported.
So I show you this just to indicate we see the same pattern in our men as the Physicians Health Study or the Nurses Health Study or in any number of other recent reports.
Now, what we were trying to look at in this set of analyses was whether or not inactivity should be considered as a more important co-morbidities perhaps of obesity. The treatment guidelines of course say that you stratify a person by BMI, by waste circumference, and then you look for co-morbidities.
And diabetes is usually the first thing that the obesity experts mention, that obese -- overweight or obesity individuals with diabetes have very high rates of cardiovascular and all-cause mortality, and of course hypertension, cholesterol, smoking and so forth -- high cholesterol, smoking and the like are other risk factors. Physical activity is only mentioned kind of as an after- thought later. So that prompted me to take a look at this issue, again, in our cohort with the objective measure of fitness.
So what I've done in this analysis for these five risk predictors, high cholesterol, diabetes, hypertension, smoking and low fitness, we show the -- in this slide, the relative risk for all-cause mortality with the orange bars, the reference category, being the normal weight men in that analysis who do not have that specific risk factor, so these normal weight men who have cholesterol 239 or less.
This then is the relative risk for men who are normal weight, but have high cholesterol, maybe a little increase in mortality. The overweight men with high cholesterol and the obese men with high cholesterol. So then the same for each of the other analyses.
Normal weight men with diabetes, normal weight men -- or overweight men with diabetes, obese men with diabetes, hypertension, cigarette smoking, low fitness. Now, even with my biases that I bring to this from studying physical activity, frankly I was a little surprised that in obese men, low fitness is just as important a predictor of mortality as is actually having diabetes.
We see a very similar pattern of results, only with higher relative risk, when we look at cardiovascular disease mortality. So, again, keep in mind, these are preliminary. They've not undergone peer review. I'm surprised the journal didn't write back immediately and accept this paper. But they've had it now for a couple of months.
But it seems to me that for overweight and obese individuals, low fitness which you get by being sedentary is an important determinant of mortality risk. And therefore, it deserves to be considered anytime one is interested in the public health problem of overweight and obesity.
Now, if one takes the view of the Surgeon General or the State Health Officer -- of course, epidemiologists, as you know, like to go on and look at population attributable risk -- let's just look at the population attributable risk for all cause mortality. And this is in the 3,300 men in our cohort who were obese, BMI of 30 or greater.
Granted the relative risk for fitness is -- it's one of the higher ones. It's almost as high as actually having cardiovascular disease at baseline. And all of these relative risks are adjusted for all the items in the table, age, examination, year and family history of CVD. The relative risks I suppose don't differ all that much, except maybe for hypertension. But the prevalence of the condition does.
Diabetes is a bad thing to have. And obese people are more likely to have diabetes than those who are not obese. But still, only ten -- ten percent of our obese men have diabetes. Fifty percent of them are low fit. So from a population attributable risk perspective, I would put forth the argument that low fitness is a very important item to consider in the context of the public health problem of overweight and obesity.
We've just recently -- and these -- this isn't even out for review yet. So it's even more preliminary. We're still working on this paper and I'll just show you one data slide from it. But it's a group of 1,200 of our men who had diabetes at baseline, either by a clinical history - - a physician diagnosed diabetes -- or elevated fasting plasma glucose or they're taking anti-diabetic medicine.
We looked at the low fit group. Again, that's the least fit, 20 percent of the total population. But in these diabetic men, 42 percent of them actually were low fit. Followed them for about 12 years, 180 deaths --heavy touch on the button -- with 180 deaths in the cohort.
And what I'm showing in this slide is the relative risk for cardiovascular disease mortality adjusted for age and examination year across the low, moderate and high fit categories. And again, you see this very steep inverse gradient in this group of men who already had diabetes at the start of follow-up. So high fitness appears to provide some protection against mortality and in men who have diabetes.
Here is the relative risk for cardiovascular mortality across the BMI categories, normal weight, overweight and obese; the sort of upward progression that you would expect. But again, I would argue it seems to me that low fitness is as important a predictor of mortality in diabetic men as is obesity.
And this is not to diminish the importance of obesity, but I am trying to elevate the importance of physical fitness and physical activity in healthy people, in unhealthy people.
We've just been looking at this same kind of analysis in hypertensive men. We see the same thing. The hypertensive men who are low fit, much more likely to die of cardiovascular disease and all cause mortality than those who are fit.
So there is a large database. And I didn't come preparing to do a meta analysis and report all the data. The Surgeon General's report on physical activity and health published in 1996 did that. And certainly conclusions, conservative conclusions drawn from that were that sedentary living habits and low fitness are associated with the kinds of disease outcomes that I have been talking about this morning.
Further, since the prevalence of inactivity and, therefore, I suspect the prevalence of low fitness, although we don't know that yet, is on the order of 20 to 25 percent of the population. So we have a huge group of people exposed to this hazard of being sedentary and unfit.
We do have pretty good consensus on the public health recommendations for physical activity. The first kind of major report on this -- the next speaker, Russ Pate, led a group; Centers for Disease Control and the American College of Sports Medicine published a report in JAMA in '95. And the main recommendation for sedentary individuals -- this is not the recommendation for marathon runners. This is the recommendation for the 25 percent who are totally sedentary. That if you get up and move around a little bit and accumulate 30 minutes of at least moderate intensity physical activity on most, preferably all days of the week, you will be getting enough exercise to get some health benefits.
Now, the two relatively new elements of this recommendation was the word, "accumulate". I know exercise scientists for years preached that exercise needed to be aerobic and large muscle and continuous to the point that people were afraid to stop and tie their shoe. And I think we now know that you can accumulate the dose. It's really -- and those of you in the dietary community can certainly think of, you know, it's the total calories spent over the day, over the week that is probably important.
I don't think we can make a strong case that intensity of exercise matters very much. There may be some minimal intensity below which there is no benefit. But if there is, no one has yet, at least to me, has presented compelling evidence that that's true. It's kind of how much do you do, keeping in mind the high death rates in our low fit men and the much lower death rates in the moderately fit men.
You get up and move around a little bit, take three ten-minute walks a day meeting this criterion, you will move into our moderate fitness category.
Now, this same recommendation was repeated almost word-for-word in the -- also in 1996, the -- first of all, the National Institutes of Health Consensus Development Conference on Physical Activity and Cardiovascular Health, and was essentially repeated, not word-for-word, but conceptually repeated by the Surgeon General's report also in 1996.
The American Heart statements on exercise that in '92 labeled inactivity as the fourth risk factor and then their revised statement in '96 also embraced this concept. So the important public health advice is you don't have to go through marathon training to get benefit from activity.
And finally, can we get people to follow this recommendation? I'll show you data from one of our clinical trials recently published. And again, you have the reprint of that published in January in the JAMA. It was a two-year randomized trial which we recruited 235 very sedentary and unfit middle-aged men and women in -- in the Dallas community.
Now, this clinical trials cohort was very diverse. We had 25 to 30 percent minorities across a fairly wide age range. We were interested in comparing in this study the established exercise treatment advice, that is the structured advice -- go to the gym; join a fitness class; run on the treadmill; run around the block, etcetera -- with what we've called lifestyle physical activity.
It was a two-year trial, but there was six months of intensive intervention and then 18 months of follow-up with fairly minor intensity of intervention.
The goal was to increase energy expenditure in these sedentary individuals on the order of three kilocalories per kilogram per day which if accomplished would achieve comfortably, more than achieve the CDC/ACSM public health recommendation that I mentioned a moment ago.
The structured intervention group in this study was given a free membership to a state of the art fitness center for the six months of intensive intervention. And they followed the traditional kind of American College of Sports Medicine exercise prescription model that you see on the slide.
The lifestyle intervention group came to the institute for an hour a week for the first four months, and then every other week for the next two months. But they came not to exercise. They came to sit in the room like this around a table and talk about exercise. So it was pretty non-threatening. You know, don't even have to come and exercise. We come and talk about it.
Now, obviously, it was a little more complicated than that. The goal of the intervention was to help people develop the behavioral lifestyle skills to integrate more physical activity into their daily lives. We tried to take them where they were.
We tried to move them along a motivational readiness spectrum. We tried to get them to use the behavioral skills of goals setting -- or self-monitoring, goals setting, evaluating progress, finding and using social support, positive reinforcement, etcetera, etcetera.
Here are the results in terms of their energy expenditure obtained from a carefully structured and validated seven-day physical activity recall. You can see both groups increased their physical activity at six months during the intensive intervention. There was some slight decay over the 18 months following, but no difference between groups at either time point, and both groups being more active at 24 months.
I'm not taking the time to show the data. Both groups were also more fit at 24 months by VO2 max. criteria. They lowered their blood pressure -- I didn't put in -- they also lost body fat. And again, no difference between the two groups. So these two approaches appear to work which give us encouragement that this lifestyle recommendations, lifestyle advice can appeal to a segment of the population that will never join a fitness center, take up jogging, join aerobics class and -- and so forth.
We have a paper coming out in preventive medicine on the cost effectiveness of these analyses. And you can see kind of for any of the outcomes we looked at per unit change in these outcomes, the lifestyle intervention was a good deal less expensive.
So we think we're accumulating evidence that these sort of behavioral skills training sort of intervention does work for physical activity just as it has worked in smoking cessation and diet intervention programs. Don't have all the answers but, as I said, this is a very new area of research. I think we have made some progress.
So just to summarize, I think we can certainly say that sedentary habits are an important determinant of chronic disease. I didn't show any data on loss of function, but that, too. Decreased longevity.
I think physical activity and certainly fitness in our analyses seems to attenuate the force of the other well established risk factors including overweight and obesity on mortality, both all cause and cardiovascular disease mortality.
Now, physical activity doesn't contribute a lot to weight loss. I mean, just the thermodynamics of it make the situation -- you can't lose a lot of weight by activity. You certainly can't do it quickly. I mean, it takes a long time. You just can't spend that many calories, and especially if you take an overweight, sedentary, unfit person. They don't have the engine to spin the calories. So you can't lose a lot of weight certainly quickly by physical activity.
It does, however, in meta analyses appear to add a couple of kilograms of weight loss in dietary intervention programs or even pharmacologic interventions over weeks or months. But it does appear to be critical, maybe almost essential to maintain weight loss. We don't have all the randomized trials and maintenance that we would like. But it appears that people who lose weight by whatever means and then manage to keep it off are the people who take up physical activity.
So I think inactivity is a major public health problem. We have some good consensus recommendations that I've reviewed. And I would hope that your committee would -- if you decide to issue guidelines on physical activity, I hope you will consider it carefully and make it consistent with the NIH/CDC/ACSM/Surgeon General's recommendation so we don't confuse the public.
I think -- I think we are learning how to amount the effect of physical activity interventions. And we've never had a large-scale, nationwide coordinated physical activity intervention ala National Cholesterol Education Campaign and the like. It is time that we begin to organize ourselves for such an effort.
Thank you. And I intend to address questions, Dr. Garza, if you have time and if -- if the committee has questions.
DR. GARZA: We saved some time for questions. Roland?
DR. WEINSIER: Steve, I really appreciate your overview and an excellent summary in such a short period of time. A couple of questions, if I may. Why do you think there is a difference in the effectiveness of --
DR. GARZA: Roland, can you please use your mike? Otherwise, they're going to --
DR. WEINSIER: Yes. Why do you think there is a difference between the effectiveness of physical activity or exercise to sustain weight loss, but not to induce weight loss. What are the proposed different mechanisms by which this occurs?
DR. BLAIR: Well, I'm not sure it is different mechanisms. I think it's largely -- and it's the same kind of dose of exercise. It's just that in the shorter term weight loss studies that people -- I mean, some of them short, ten, 12, 14 weeks. And again, with unfit, sedentary, obese individuals, they just can't do enough exercise.
But if they begin to exercise, begin to develop a little bit of fitness and they lose some weight which makes it easier to do exercise, the if they can sustain according to the data 45 to 60 minutes a day of just moderate intensity, walking activity, that that does of exercise then seems to prevent the weight regain. And some data would even suggest slight further weight loss.
I don't know that it's a different mechanism. It's just a different set of circumstances I would say. Now, there are -- if I can interrupt -- some of the psychologists like Kelly Brownell would say that actually exercise increases self-efficacy and has psychological benefits and makes it easier for the person to deal with staying with a diet and so forth. It's a nice hypothesis and I hope it's true. It could be true. But I don't think there is a lot of data to support it at this point.
DR. WEINSIER: Yes, I guess it's an unfair question because I don't think there are any data to indicate that there is a mechanism the same or different. But it's an interesting combination.
Let me ask you, in -- in the beginning you probably defined physical fitness, but I missed it. I heard the VO2 max. But it's a VO2 max. relative to what? Is it - - is it --
DR. BLAIR: Well, the definition --
DR. WEINSIER: -- divided by weight?
DR. BLAIR: Well, the definition of fitness in our studies is the bottom 20 percent in each age-sex group --
DR. WEINSIER: Yes, but the --
DR. BLAIR: -- on this maximal exercise test. So it's time on treadmill in this protocol which correlates 0.92 with measured VO2 max. ml per kg per minute. So this really would be analogous to VO2 max. and ml per kg per minute.
DR. WEINSIER: So it's -- basically, is it function then as a ratio of VO2 max. relative to body weight?
DR. BLAIR: Yes.
DR. WEINSIER: If -- let me see if this makes sense. I'm thinking back to the decades of time when we looked at resting metabolic rate relative to fat-free mass in obese versus lean people. And we were convinced the obese people have a lower resting metabolic rate relative to body size. But it was an artifact of the ratio effect.
If -- if I have -- if I want to change my fitness, it looks like I can do it in two ways. I can either keep my weight the same and be more physically active and increase my VO2 and, therefore, keep the denominator the same and I increase my fitness. Alternatively, could I not decrease my weight with keeping the VO2 max. the same and look like I'm becoming more fit?
DR. BLAIR: Yes, you can at least theoretically. But you have to lose an awful lot of weight to kind of get any kind of clinically significant --
DR. WEINSIER: Right.
DR. BLAIR: -- change in fitness. It's just going through the math --
DR. WEINSIER: Right.
DR. BLAIR: -- through the mathematics of it. Theoretically, that is possible. We have done some analyses, particularly in the -- in the area of fitness, fatness and mortality in which we estimated VO2 in liters where we take out body weight, carry out the analyses. And we see about the same kind of thing, trying to get at a pure and a cardiovascular power function, if you will, by looking at it in liters.
DR. WEINSIER: Within an individual -- yes, your point is well taken. I may not change my weight that much. But here we're crossing over groups of BMI -- I mean, percent body fat. It's gone from less than 16 and 25 percent. So we're getting pretty significant comparisons. And I wonder if this has been looked at in an analysis of variance rather than a ratio effect to make sure that --
DR. BLAIR: Yes.
DR. WEINSIER: -- this is not an appearance of decreased fitness and obese people --
DR. BLAIR: Yes, we've --
DR. WEINSIER: -- rather than the truth.
DR. BLAIR: -- we've done a lot of that over the years, just looking -- not looking at BMI categories, but looking at fitness in these kinds of outcomes and in multi- variant models including BMI, including percent fat. And it changes -- it rarely changes the relative risk much at all.
DR. WEINSIER: And can I ask one last question, or am I taking up too much of my time?
DR. GARZA: Well, let me go to others and then we'll come back to you, Roland.
DR. WEINSIER: Okay.
DR. GARZA: Scott?
DR. GRUNDY: There were questions that I had in these impressive results. One is it always helps your case if you can tell me why it provides benefit, what is the mechanisms. And, you know, even if you had a list of three or four most important, it would be more convincing to me that, you know, what does exercise do.
The other question I have is how do you know that the patients or people who entered this study weren't already ill when they entered this study and that impaired their ability to exercise or be in shape?
DR. BLAIR: I'll take the second question first. The epidemiologists, the ones on the committee, now we always worry about subclinical disease at baseline in these populations. However, in our study, these patients have -- have gone through a five or six-hour intensive physical examination. Each physician at the Cooper Clinic sees four or five patients a day.
Now, that's not to say that people couldn't get through that screening and still have subclinical disease. I think it's much less likely that they do than if we did a questionnaire survey and asked them about their health status. So we certainly try to take that into account.
And further, when we look at patients that we know have disease, hypertension, diabetes and the like, we see the same pattern. Now, you could say, "Well, yes, but the most sick diabetics are the ones who don't exercise as much." And, you know, we try to control for that by the biochemical measures we have and blood pressure and other risk factors. And we never make this association go away. We attenuate it a little bit.
What are the mechanisms? Certainly, the conventional risk factors explain a little bit of this, but not a whole lot of this because when we adjust for lipids and blood pressure and smoking status, it only attenuates the association between fitness and mortality by some relatively small amount. But there are newer mechanisms under investigation in clinical studies for which we and other large cohort studies just don't have the data such as immune function.
Now, I don't know how those of you on the committee feel about is atherosclerosis partly an infectious disease. I mean, there are people who believe that it is and there are some data that support that.
We do have good data from clinical studies that the proper dose of exercise seems to enhance immune function. You can do too much and drive it down. But that takes an awful lot of exercise. And that's -- an awful lot of exercise is not a public health problem in the U.S. So immune function could be one.
Exercise affects clotting mechanisms, lipolytic mechanisms. And it's certainly possible that that is part of the overall picture. We just don't have those data in the epidemiological studies.
And finally I'm just a -- you know, a simple Kansas farm boy epidemiologist. It seems to me that -- we know that exercise increases the strength of the heart. It gives you bigger LV mass and increases cardiac output and stroke volume. If you've got a bigger, stronger pump and then even if you have an infarct and you lose X grams of tissue, the fit person has more tissue left over. And we do have data from epidemiology that fit people -- active people I should say are less likely to die if they have a heart attack.
So, you know, maybe you've just got a better pump, Scott, is part of it. But certainly we don't know all we would like to know about mechanisms.
DR. GARZA: Shiriki?
DR. KUMANYIKA: Steve, this may not be a fair question for you. But let's get it on the table. Can you help us relate physical activity to dietary intake?
DR. BLAIR: There is not much of an association between activity and dietary intake as -- as I understand it. We've looked at this in our randomized clinical trials. Where we're intervening on physical activity, we're not intervening on diet. We measure diet, three-day diet records, and we give the participants feedback kind of as a customer service. And they're very interested in it. They want to know about a healthful diet.
But intervening on physical activity does not in our studies cause a spontaneous change in diet. We haven't been able to find it. And some of the dieticians working with us have looked for it. And we just can't find any evidence that your diet changes.
At a consensus conference on physical activity and obesity where Scott chaired the consensus panel, we heard a report on this topic by Dr. Blundale from the U.K. And Scott, if you can help me remember, I -- I think the bottom line of his presentation was really not a lot of evidence that people have been able to identify direct links between activity and diet pattern.
Now, obviously total calories -- I mean, over time if you exercise a lot, your total caloric intake is going to go up. And to some extent, if you become more and more sedentary, it's going to go down. It probably doesn't keep up. It probably doesn't go down enough which frankly may be a major cause of the overweight and obesity. But there doesn't appear to be a very tight coupling as far as I understand it.
DR. GARZA: Dr. Tinker?
DR. TINKER: I have two questions about the CDC recommendation. And the first one is, let me just make sure I understand, the 30 minutes of physical activity, that is not in any way related to anything related to weight loss. So that's just the physical activity without any --
DR. BLAIR: For health.
DR. TINKER: Basically for health.
DR. BLAIR: Physical activity for health in general. It's a broad stroke recommendation. You can ask Russ when he gets up here. I certainly believe and think many people on the committee believe that if you want to focus on weight loss, that's probably not enough. You probably ought to be recommending more. And the follow-up studies on weight maintenance suggest 45 minutes to moderate -- moderately vigorous activity is probably a better target to aim for.
So if the sole focus is on weight loss, weight control, I'm not sure that's the optimal. But it's enough to get health benefits. And if our data are correct, even if you stay fat, stay overweight, if you're exercising at that level, you're probably better off than if you weren't exercising.
DR. TINKER: And then have you considered moderate intensive exercise?
DR. BLAIR: Yes. Moderate intensity exercise in these various guidelines has been defined somewhat differently. But as I boil it down, what it -- what it amounts to is brisk walking which three to four miles an hour; not race walking, just striding along three to four miles an hour, 15 to 20 minutes per mile.
DR. GARZA: Meir?
DR. BLAIR: Oh, can I say just one more thing about that? We received a lot of criticism from some quarters when the Surgeon General's report came out. And I think the thing you have in your packet talks about washing the car and washing windows. And I can't tell you how many drive-time VJs would call and say, "Oh, Dr. Blair, should I wash my car every day for 30 minutes?". And I would say, "Well, no, you dummy, you wash your car today and rake the leaves tomorrow."
(Laughter.)
But actually just in the last year, there have been three laboratories looking at -- with measured oxygen uptake with portable metabolic cards, measuring the energy costs of these kinds of activities, of weeding and raking and vacuuming and washing windows and dusting. And all of those fall in the three to four met. category. And there is remarkable consistence across laboratories.
I was a little surprised. Dusting I think in the one laboratory that studied dusting was 3.1 mets., in that range. Patty Feregson says she now has 15 coats of wax on her tables, but --
DR. GARZA: We've just discovered why men die so young.
(Laughter.)
DR. GARZA: On that happy note, Meir?
DR. STAMPFER: Just to follow up on that, is there -- is there -- what's your take on any further benefit of more vigorous activity?
DR. BLAIR: Thank you, Meir. Good question. I think there is more benefit. Certainly our data suggests additional benefit in the high fit group. And as I look across all of our studies and I look at Pattenbarter studies and so forth, if you go from -- say in our studies, from moderate fit to high fit, the risk drops another ten to 15 percent. We've actually published on fitness change, also. I don't think I included that paper, 1995 in the JAMA.
The big drop going from low fit to moderate fit, the risk dropped by about 50 percent. Going from moderate to high, it dropped another ten to 15 percent. So some of you know, I'm a runner of 30 years. I ran this morning. I run every day. I'm proof positive. Running 100,000 miles does not make you skinny. But I hope that there is benefit to more vigorous exercise. And the data certainly suggests that.
So I think we need to always mention that so as not to discourage the people who are runners and cyclists and go to aerobics class every week. I think they probably do get a little additional benefit.
DR. GARZA: Johanna?
DR. DWYER: Thank you for a wonderful presentation, Steve. One question about sources and the second one about the message we might want to convey. On sources, it would seem to me that one of the biggest couch potato processing companies in the United States is the U.S. military because they're taking fairly sedentary people and putting them through basic training. And I wondered if there were any data that you were aware of that might -- might be useful for us in our deliberations.
The second is the focus. And should we be focusing on an active and healthy lifestyle to avoid this inactivity crisis that Surgeon General Coop referred to a couple of weeks ago rather than achieving and maintaining healthy weight? What is your take on that or are both messages -- certainly, the latter is important. It was in the guidelines. But what about the first?
DR. BLAIR: There certainly are some military studies not with health outcomes. There are studies with what happens to injury rates as these recruits go through basic training; some studies on changes and fitness. But another nature of the military is that basically young people and basically leave early. So you can't study the chronic diseases in that population.
DR. DWYER: Fair enough.
DR. BLAIR: I would say one of the people who knows the most about this if you want to contact is Colonel Bruce Jones, retired, now at CDC. Bruce Jones at CDC in the injury branch has over a career of 15 or 20 years conducted studies of that type. So he would be a good source of information.
And your second -- what should the focus be? Well, I'm certainly not going to come here and say let's ignore overweight and obesity. I do think that the pendulum has maybe swung too far to the side of focusing on overweight and obesity as just the problem.
And I know the old guidelines mention physical activity. And to me, we -- I think we need to come to an energy balance discussion. I think it makes no sense to talk about energy intake and not talk about energy expenditure. It makes no sense to talk about diet and not talk about physical activity.
And I say that to people in both camps. I do not say that if you're active, you can ignore diet and forget about your weight. I only say that when I'm trying to annoy the speaker you have tomorrow, Claude Bouchard, or someone of that sort. But, no, I think we need to focus on both. But certainly, you can see a passionate believer. We need to have more emphasis in this country on organized programs to promote activity. And it's going down. I mean, the estimates show the U.S. and the U.K. -- of course, a lot of assumptions go into these -- but the average daily energy expenditure of Americans has probably declined 200 or 300 kilocalories a day, not since 1900; probably since 1980, 1975.
Why is that? Just think of your own life. How many of you order Christmas presents now over the internet or Father's Day present which we -- some of us just did this week over the internet? Now, shopping is not your basic aerobic training activity. But it does spend more calories than sitting at your computer shopping over the internet. So 10,000 examples we can all come up with of how we've engineered activity out of daily life. It would be difficult to overcome that in this larger context of an evermore sedentary kind of energy expenditure-free society.
DR. GARZA: Rachel?
DR. JOHNSON: Thank you, as well, Steve. As you probably know, the dietary guidelines are designed to apply to all Americans over the age of two. And all of your data that you presented on was adults. I wanted to know if you think there is any special caveats for the pediatric population, if the CDC guidelines of 30 minutes per day are applicable to children, and if you have any comments achieving fitness -- you know, appropriate fitness levels in children and prevention of obesity, the role of physical activity.
DR. BLAIR: My work, as you point out, is with adults, both in epidemiological studies and clinical trials. And I might have some opinions. But your next speaker is an authority --
DR. JOHNSON: Okay.
DR. BLAIR: -- on that issue. So I'm sure he'll be addressing it.
DR. JOHNSON: Thanks.
DR. GARZA: Steve, I, too, want to thank you for an excellent presentation. I have three -- three brief questions. On the slides where you had body fat, how was that mentioned?
DR. BLAIR: In -- in those 22,000 men, it was either by hydrostatic weighing or by sum of skin folds, or both. There was some -- some group that had only the sum of skin folds. In that group, we used the data on the men that had both to develop a regression equation to standardize an estimate of percent fat from the skin fold measures. And then actually, it appeared to work pretty well. So it -- there were different measures, but we tried to put them together in a coherent way and didn't -- well, I have confidence that it's a reasonable estimate.
DR. GARZA: Okay. Is there a -- is there a benefit from the cumulative duration of physical fitness, or is -- is -- are the benefits available to the individual the moment he or she becomes more physically fit? Do you see a risk -- an immediate risk reduction or does it take time?
DR. BLAIR: There is an immediate effect. There is an acute effect of physical activity that has been overlooked for too long. We are accustomed to thinking of exercise as something you train and weeks and months later, you've had some changes and you get benefit.
But actually, we know one bout of exercise improves insulin sensitivity. One bout of exercise effects fibrinolytic activity. One bout of exercise typically makes you feel better. And if you doubt that, when it comes time for your break, get up and walk five minutes down the hall. You'll feel better.
So there are acute benefits of exercise as well as chronic benefits from training. And Bill Haskell has suggested there is kind of an augmented acute effect that as you progress along over weeks and months being more active, improving fitness, you get a little bit of an augmentation to some of these acute effects. But there are some immediate effects.
DR. GARZA: So you would expect that the risk to chronic disease would fall in someone over 50 that decided to -- to adopt either a lifestyle or a structure type of program and, therefore, become more physically fit?
DR. BLAIR: Well, our data show that. We looked at men who were unfit at the first exam, fit at the second exam, and just moderately fit, just getting out of the low fit category. Their risk of all cause mortality was 50 percent lower than the unfit men who stayed unfit.
Now, I don't think I can answer the question of acute benefits and reduction in mortality. I mean, nobody has any data that would really address that. So we're kind of looking at habitual activity or an attained level of fitness which get by habitual activity.
So how quickly the results would come, you know, we said over the years for smoking cessation, within a year risk begins to return. We don't have the data to really know how quickly that risk returns.
DR. GARZA: You don't have any data that suggests it's too late to begin exercising at any particular age? I mean --
DR. BLAIR: It's never too late according to our data and according to other date; for example, on musculoskeletal fitness and function. We see this 50 percent reduction in mortality in men 40 to 49 who improved from low fit to moderately fit. We also see it in the men 60 and older who improved from low fit to moderately fit.
And with our new mortality data -- I haven't yet looked at change in that group, but we see this same inverse gradient of fitness, much higher death rates, but same inverse gradient in men and women 80 -- 80 years of age and older at baseline. Never too late to get the benefits of exercise according to our data.
DR. GARZA: And lastly, what do the -- what is the body of evidence that you find most convincing that goes counter to what you've told us for the last 45 minutes?
DR. BLAIR: There is essentially no reasonable study over the last ten or 15 years, epidemiological study as far as I know, that has shown sizeable increases in mortality, for example, or increases in some of these other end points in people who are active or fit compared with not.
So there is -- I don't think there is any reasonable evidence on the other side of that issue. Another issue, however, is what about the costs of exercise; what about the sudden deaths. Well, these epidemiological studies, when we look at death rates at the end, we're really looking at a net effect.
So if the exercisers -- if some of them are more likely to drop dead while they're out there exercising, I mean, that -- their death gets counted in the active and fit group. So we're looking at net effects.
Another issue of course is injury, musculoskeletal injury. I attended last weekend an NIAMS Academy of Orthopedic Surgeons meeting. And they're very interested in the injury issue. Certainly, if you go out and move around or if you take up competitive sports, you're more likely to develop a musculoskeletal injury than if you stay home in bed all day. So that there is that. And frankly, that hasn't been studied enough.
But I think we could expect that if all the sedentary Americans started exercising, there would be some increase in, you know, musculoskeletal injuries of one type or another. However, we do have data to suggest that the injury rates in people who are walking is just not very high. It's if you take up running, aerobic dance or more vigorous activities and sports where the injury rates will go up.
But the ACSM/CDC recommendation has been looked at in some controlled -- or activity of that type looked at in controlled studies. Injuries don't seem to go up very much. There will be some. I think the benefits outweigh the risks clearly.
DR. GARZA: We're going to have to move on to our next speaker unless your questions are very brief and Steve promises to answer them within the ten -- a
ten-second --
DR. BLAIR: All right. I'll promise.
DR. GARZA: Roland?
DR. WEINSIER: Yes. Real quick. What was the data on anaerobic fitness? Are there -- is it absence of data or that the data are negative regarding anaerobic resistance fitness to increase strength, therefore, being able to climb stairs, stand from a chair, carry a load, etcetera?
DR. BLAIR: We have data from both epidemiologic studies -- we've been looking at this -- and from training studies, especially in elderly individuals, that musculoskeletal fitness delays the development of functional limitations. And that's so logical, it has to be true. I mean, the 80-year-old who can't get out of the chair can't get out of the chair because they have no muscle left in their quadriceps and other places.
So, I mean, it's clear that there are those benefits. There is a suggestion that musculoskeletal fitness-type training, strength training benefits insulin sensitivity. But we have no studies of disease outcome with musculoskeletal fitness that I am aware of.
DR. GARZA: Okay. Alice?
DR. LICHTENSTEIN: Just there are some data on the very old and increased physical activity and being able to maintain activities of daily life. Fiaderoni has published that.
DR. BLAIR: Yes.
DR. LICHTENSTEIN: But very briefly, do you collect any data on prior participation or exposure to activity, let's say, during childhood in your subjects and is there any relationship?
DR. BLAIR: We looked at that in a paper ten years ago or so. And just a bottom line, I can't find evidence that participation in activity or sports, I should say, in that study early in life has anything to do with health and function later on life. And Pattenbarter's data tends to show the same thing. It's the activity you're getting in middle age and older that gives you the benefit.
DR. LICHTENSTEIN: But would a relationship with whether when they come in to your study they are active or inactive?
DR. BLAIR: I included that in my comment actually, not specifically. But we saw that -- again, these data were limited to men. Men who were athletes in high school or college were no more likely to be physically active and fit when they came in average age 45 to the clinic. And, yes, I was actually referring to Fiaderoni's data and our epidemiological data that suggests that strength training, musculoskeletal fitness does indeed benefit function.
DR. GARZA: Thank you very much. We're going to move on then to our -- out next speaker, Dr. Russ Pate from the University of South Carolina in Columbia. He is going to address the appropriate physical activity level for children. Dr. Pate, thank you very much for joining us this morning.
DR. PATE: I apologize for dashing out. I had a -- have a committing British Columbia night.
DR. GARZA: The door is locked. Don't worry about it.
(Laughter.)
DR. PATE: Well, I'm relieved. Well, I have followed Steve many times over the years including on to the faculty of the University of South Carolina where he used to work. And it's never an easy assignment. But I'm very pleased to have the opportunity, glad to be here, and I am, too, very pleased that consideration is being given to including an independent guideline on physical activity in -- in the next set of dietary guidelines.
So that there will be no confusion about where I'm heading, I do intend to encourage you to proceed with -- with a development of an independent guideline on physical activity.
But I will also be cautioning you that I think if that guideline is to apply to young people, children and adolescents, that consideration should, in fact, must be given to the fact that children and youths are not the same as adults when it comes to physical activity. Their physical activity behaviors, as I will be showing you, are -- are different.
And, unfortunately, I will have to admit to you right up front that the knowledge base that links physical activity to chronic disease health outcomes is orders of magnitude less voluminous in young people than is the case with -- with adults. And Steve, of course, has done a very nice job here this morning, of over-viewing that evidence in -- in adults.
My primary objective, as per the title of my presentation, is going to be to overview for you the current thinking on guidelines for physical activity in children and youths. I am going to begin by addressing a couple of secondary objectives which get at why I think it would be appropriate to be concerned about physical activity in young people in the context of the dietary guidelines that are being developed.
The major reasons are, again, so that you will know where I'm heading, that obesity is an increasing problem in young Americans. And it is my view and I think a view that is increasingly accepted that lower levels and/or decreasing levels of physical activity is probably a major, if not the exclusive, reason for that trend.
And I, like Steve, understand that I am amongst a group of people who focus largely on diet. So that you will know where my biases are, I am a traditionally trained exercise physiologist. I work in a department of exercise science which is in a school of public health.
And I'll -- I'll tell you right now that I don't think we're going to effectively address the growing obesity problem that we confront in this country if we don't deal with physical activity.
And that certainly is not to say that dietary considerations are -- are unimportant or less important. But I think there are some fairly profound physiologic reasons why diet won't be the whole answer to the problem and that physical activity will have to be addressed.
So with that as introduction, many of you in the room are as or more familiar with these data than I. This is -- is simply to make the point that the obesity rates in young people in the United States are clearly growing. And Rick Troiano's paper from a year or so ago based on the National Health Survey data collected since the '60s indicates this.
In this analysis, they looked at obesity rates from the '60s to the '90s based on the National Health Examination Survey data and then subsequent NHANES data, and defined overweight in these analyses as the age and sex- specific ninety-fifth percentile for -- for BMI from the first two surveys in the '60s.
And I suspect you are well aware, these are data for kids in the six to 11-year age range working from left to right, black males, black females, white males, white females. Again, the bars go from the initial National Health Examination Survey data on up to NHANES III.
And as I'm sure you all know and appreciate, the prevalence rates in all groups -- prevalence of obesity rates in all groups have gone up and are very alarming in particularly black females. Similar pattern in older kids. These are the data for ages 12 to 17. And, again, the prevalence rates are very alarming, particularly in black females, but apparently have increased in -- in all of the demographic groups, and perhaps most rapidly in the most recent surveys.
And this is just a capsulization of that trend across the various groups showing perhaps a disproportionate and alarming great increase in the prevalence of obesity in young people just in -- in recent decades.
Now, the issue is what, if anything, does this have to do with physical activity and what might physical activity do in -- in addressing this -- this alarming trend. I think there are, at least in my mind, four areas in which we would like to know a lot. And frankly in some of these areas, we know something, probably not as much as we would like to. And in a couple of areas, I think the science is - - is really just beginning.
The four areas that come to my mind that would potentially link physical activity to this obesity problem are, first of all, cross-sectional epidemiologic studies looking at associations between activity and body fatness. These studies exist. There are a fairly large number of them. I don't intend to talk about them today. I have reviewed this literature in the past. I can tell you the conclusion that I draw.
I think that those studies, if you look across them, generally show that the more active kids tend to be leaner or vice versa, less active kids tend to be fatter. Now, we all understand the limitations of these cross- sectional surveys and they certainly do not establish causality.
And you can make the argument that the causality could be and may in fact even operate in both directions: Fatter kids less likely to opt for activity and, therefore, are found to be less active; and, of course, inactivity also could be causally related to the development of obesity.
Now, where it gets a little murkier though is in the context that we're all here considering today which is the body composition fatness issue. And based on my review of those cross-sectional studies, it's not nearly as clear that heavier kids or fatter kids expend less energy in activity. They be less active in terms of gross movement, but, of course, they're heavier. And so when they move, they expend more energy in the process.
And that's where the whole relationship becomes a little less clear. And I think that at this point, or at least based on my last review of that literature, it is not as clear as we might like that -- that heavier kids expend less energy and activity, though they are probably less active.
Now, second issue is probably the one that we most likely -- would most like to be able to present compelling data concerning. And that is that we would like to know based on longitudinal -- prospective longitudinal observational studies, that kids that are more active are perhaps less likely to develop obesity in the future; or conversely, that -- that -- that may not happen, I mean, not to bias ourselves of what the outcome of those observations would be.
There are very few studies, unfortunately, in the literature at this point that address this critical issue. And I'm going to overview a couple of them for you because I think they really are most germane. So that the consideration here is if kids are more or less active, are they more or less likely to develop obesity in the future.
And with the credit to Ivo Brisanic at NHLDI who kind of lives in this world and helped me identify a few of these papers, I'll mention a couple of them.
One is Bob Clujis' paper from '95 in Pediatrics, presenting a longitudinal analysis here. They looked about 150 healthy pre-school kids; followed them from age three to five up to the point where they were in the first grade. And they did have an over-sampling of kids that were considered to be obese. Physical activity was rated by both parents. They looked at it at baseline and with two years of follow-up.
And the significant predictors of change in BMI over the two-year follow-up period in addition to age included baseline aerobic activity and change in what was called leisure time activity between years two and -- two and three. So this study did provide some evidence that activity-related variables measured in a limited way, which is often true in these studies, that those activity markers were associated with -- with weight gain in pre-school kids.
Another study in -- in young children based on the Framingham offspring study and again looked at pre-schoolers physical activity level and change in body fatness. They had better measures of activity in this study. And here they had about 100 kids, initially three to five years old; followed them through the first grade and assessed activity in this case by Caltrack.
Many of you know, but for those that don't, the -- the technology in this area is developing rapidly. And many of us are now using objective motion sensors, accelerometers, which can be attached to the belt and if appropriately calibrated are indicative of movement and -- and energy expenditure.
So they used one of the earlier versions of these accelerometers. The Caltrack in this particular study observed these kids twice a year for five-day periods. They also had some diet data. And they looked at triceps and BMI.
There was evidence in this study that activity associated with weight gain and that the active kids were observed to gain less weight. These are the data for triceps skin fold, less fat than was the case in the inactive kids. These differences were significant. And it was interesting that in -- after controlling for age, TV viewing, energy expenditure, baseline fatness and parental BMI, the inactive kids were almost four-times as likely as the -- as the active kids to show increasing triceps slope over the period of observation.
One more, and this is a study that I like a lot. And I wish there were more studies like this in the literature. This is called the Young Finns Study by Dakari and others, have published several papers based on this study. And what they did was, again, a longitudinal observation of -- of activity and looked at changes in fatness and other chronic -- physiological risk factors for chronic disease in Finland who were --
DR. WEINSIER: Excuse me. What was the journal on that, if I may ask?
DR. PATE: I think it's American Journal of Epidemiology, but I'll have to check. Methods in this study were that they drew the sample from a national population registry in Finland. They had about 1,000 kids who were 12, 15 and 18 years of age at baseline, and then they followed them up about -- over a period of three years. Activity was assessed by questionnaire.
Two groups were formed at baseline, based on their activity status. At that time, they were labeled active and sedentary, though I'll say now and have more to say about it in a minute, sedentary is rarely an accurate label in -- in kids. It's often an accurate label in adults, but rarely is that literally an appropriate label in young people. But that was the label that was used here.
And they did look at BMI and subscapular skin fold thickness as well as some other physiologic markers. And to sort of cut to the main conclusion here that bears on the issues today, data for males and females, sedentary versus active groups, and controlling for lots of things. Basically what they see is lower subscapular skin fold, lower triglycerides and a higher HDL to try to total cholesterol ratio.
In the males who tracked in the active group versus those that tracked in the sedentary group, females, some of the same relationships were observed, lower subscapular skin folds and lower triglycerides, although the HDL to total cholesterol ratio was not different across those two.
Now, I've just shown you what to my knowledge are most of the papers -- not all of them; there are a handful more -- that really have taken this prospective longitudinal approach to examining the issue of whether or not higher levels of activity prevent the development of obesity and some of the physiologic risk factors for chronic disease that are often associated with obesity.
Clearly, the available data are at least suggestive and encouraging, but the number of studies is clearly very limited.
Now, another area in which the literature though is a bit deeper has to do with small group experimental observations of the effects of exercise training on fatness and physiologic risk factors for chronic disease in overweight kids. Again, I'm not going to show data on this. I'll simply tell you what I think that literature shows.
It is that if you take kids who are overweight and systematically expose them to substantial increases in activity, there will be the expected changes in body fatness. And based on Bob Gootin's fairly recent studies in -- at Augusta Medical College of Georgia, I think there is reasonable expectation that physiologic risk factors, some of them, will get better when that happens.
Now, again, we don't have nearly as much evidence on all of this as we would like. Lynn Epstein has shown on multiple occasions that you can work closely with kids in a counseling session -- setting that includes their parents and produce increases in activity that appear to last for a period of some years.
But I think based on what we do know in this area, there is some experimental evidence that -- that in overweight kids, activity can -- can be somewhat effective.
Now, what I guess we would like to know, again, most from the standpoint of interventions is from the perspective public health interventions, can we be successful. And here I think, as Steve indicated, not only are we -- are we really beginning a new line of work -- he said it's about ten years old. It's probably ten years old in adults. It's from a public health intervention standpoint maybe not even that old in kids. So our knowledge is pretty limited.
But I'll mention one paper which suggests that maybe there is -- there is hope. And this is the Planet Health Project. Steve Gortmaker and colleagues published it just earlier this year. Looking at a school-based intervention, they had five intervention and five control schools. And it was a sizeable and ethnically diverse group of sixth and seventh graders.
And it was -- it was intervention that was implemented in the school setting, but it was focused on making changes in kids' behavior outside the school setting. And there was a substantial emphasis on decreasing TV viewing, as well as modifying dietary behaviors and generally increasing moderate and vigorous physical activity.
And the findings were encouraging in that the -- and they -- the outcome measure here was prevalence of obesity and -- and then remission of obesity. And essentially what you see is adjusted odds ratios indicating that in girls in particular, but also boys, to a not significant extent in the boys, that the intervention -- kids in the intervention group were less likely to be obese at the end of the project and that they were more likely to move from an obese to a non-obese status.
Now, why? And this is probably the most interesting finding in this study. And it really came down to reductions in TV viewing appeared to be the most salient intermediate change that was associated with change in body fatness.
Now, there are other school-based studies looking at promotion of activity in kids. I work in that area and I can tell you that the success stories are few. Probably the one area where we -- where we know we can be successful is in increasing physical activity in physical education classes.
And even getting that done is not easy. It takes, you know, a couple of years of working with school personnel and -- and teachers to make changes in the curriculum. But from an intervention standpoint, this -- this literature is pretty young.
Now, what do the experts think about physical activity behavior in kids; how active do we think kids should be; and what forms of activity should they engage in? Well, there are a number of documents which have addressed this issue. One of them would be Healthy People 2000 where objectives 1.3 and 1.4 both include -- both include young people. And 1.3 is the sort of moderate intensity guideline which says increase to 30 percent for the young people at age six and older; engaging regularly, preferably daily, in light and moderate physical activity for 30 minutes.
So if you took the word, "light", out of there, you would have something that's very similar to the CDC/ACSM recommendation which I'll have more to say about in a minute.
Now, the second guideline here that -- that also applies to younger people is the more traditional vigorous physical activity guideline calling for performance of activities that would develop and/or maintain cardiorespiratory fitness, three or more days per week for 20 or more minutes. Healthy People 2000 included two guidelines which have some considerable currency, though tweaking has gone on here in the last decade.
Steve mentioned the CDC/ACSM guideline which has been essentially gratified by the NIH consensus conference and the -- and the Surgeon General's report, as well as other documents, not only in our country, but in a number of countries around the world.
I would -- I put this up this morning only to point out that when -- when we developed this guideline, it -- it was very intentional that the word, "adult", was included in it because that project really was not intended to consider what may be the special needs of young people.
But there have bene productive studies that have been designed to address what that -- those specific needs of young people. One was organized by Jim Salis and colleagues in San Diego, and was focused on adolescents. It produced two guidelines or recommendations.
The first one was rather nonspecific and it essentially said that adolescents should be active daily or nearly every day as a part of their lifestyle. And then there was sort of along trail of settings in which kids could be active: transportation, sport, PE classes, after- school jobs and so on. But no specific number of minutes was included.
Now, when this guideline has been operationalized in subsequent studies, it has often been operationalized as a 30-minute guideline. But the consensus group didn't actually specify that.
The second guideline is a lot like Healthy People 2000, objective 1.4. And it says that adolescents should engage in three or more sessions of exercise per week that lasts for 20 or more minutes and that require moderate to vigorous levels of exertion. That's pretty much the traditional exercise prescription recommendation.
Now, a fair amount of though has gone on though in the -- in the interim. And the Health Education Authority in England convened about two years ago a conference. And a number of us from the U.S. were involved in this. And I'm going to tell you that I think this is the approach that lot of consideration should be given to.
It is that the guideline might reference the fact that all young people should participate in physical activity of at least moderate intensity of one hour per day. Now, let me tell you were that came from. There is -- there is a concern that 30 minutes is just not enough in kids. And I'll show you part of the reason why in just a minute.
Not that 30 minutes wouldn't be fine, and the next guideline addresses that. It basically says for kids that are currently doing little, 30 minutes is recommended; but that it would preferable that a standard of one hour per day be met.
And then the third guideline is intended to address the fact that cardiorespiratory or aerobic-type activity, as important as it is, is likely not the only form of activity that is important for young people. And of course, there are increasing concerns about bone health and the fact that we have an enormous osteoporosis problem on our hands, and that -- that early in life is when we lay down bone; so that maybe we need to be focusing on musculoskeletal health in our guidelines for young people.
And so this -- this additional component was added. And it says that at least twice a week, some of the activities performed that meet these standards should help to enhance and maintain muscular strength and flexibility.
Now, where are our kids? Well, I wish -- I wish we knew more about this than we do. But we do know something. The YRBS, the Youth Risk Behavior Surveillance System, which involves completion of a self-report survey by high school kids, a national probability sample of high school kids in the United States on odd-numbered years does survey physical activity.
Frankly, the moderate -- the question that is intended to get at moderate intensity activity I think does not do that very well. And so I am not going to show you those data. It's not an easy thing to do. I'm not being critical of the folks that put these surveys together. And I've participated in some of those conversations.
But it probably is a little easier to get at vigorous activity. And so I'll show you what these data show. And these are data for ninth through twelfth graders. The darker bars are -- are for the girls and the lighter bars are for boys. And this is the percent needing a vigorous physical activity, 20 minutes or more, three or more days per week standard.
And what you can see is there is a rather consistent gender difference and decline in the percentage meeting the standard across those high school grades; but that quite a sizeable percentage of particularly boys and in the younger ages, girls report meeting that standard.
Do they really? I don't think so. And it's not that I think they're lying to us. It's that I think that the way maybe people in general, but kids specifically process questions of this sort, that I think there is reason to question whether literally the self-reported data that we get in surveys like YRBS is accurate.
And I have been involved the last few years with a collaborative group that has been using objective activity monitors on -- on larger groups of kids than have been studied previously. And in this study where -- well, it's a hundred kids in each of these four grade categories, going from first in the primary grades on up to the high school grades, equally split between boys and girls in Amherst, Massachusetts were activity monitors for a week.
And what's reported here is the median number of activity minutes. And this is anything above three mets., moderate intensity activity, on up looking at minutes per day. And what you see is a precipitous decline across those -- across that age.
A lot of kids, a lot of young kids would probably meet about any reasonable standard. But when you get out here to the high school age and middle school and high school age, very sizeable percentages of them do not meet it, certainly in our standard if we went with the 60 minute standard and significant percentages of them don't meet a 30-minute standard either.
I'm not going to show you this morning, but we also have -- have examined those data for bouts of activity. And all I can tell you is it's a myth. Kids, particularly in the older age groups, just do not spontaneously perform 20 minutes of vigorous activity. I mean, we've scanned these one-week accelerometer data every way there is to scan them. And it's -- it's just not there. I mean, the -- the percentage of kids that do that sort of thing with any frequency is -- is just dramatically low.
Now, I'll finish by saying that I have on more than one occasion tried to pull all this together and in the context of that San Diego conference that I mentioned a bit ago did do that, trying to look at those two guidelines from that meeting, one which was operationalized as 30 minutes of moderate to vigorous physical activity; went back to the National Children and Young Fitness Study from the mid-'80s which is a study that, since I have an important group of people here today, I'll say I think we need to replicate about once a decade. And we missed it in the '90s. So I hope we'll -- I hope we'll come back and try to get this done again before too long.
But based on the activity reports in the National Children and Youth Fitness Study from the -- from the mid- '80s, about 84 percent would have met a 30-minute standard, but it would be a considerably smaller percentage that would meet a 60-minute standard.
And Guideline 2 from the San Diego recommendation was the more structured, vigorous approach. And there, depending on age, about 62 to 70 percent of the males based on YRBS reports would have made it. A considerably smaller percentage, probably less than 50 percent of girls would meet that standard.
So where does that bring us? Well, I think that with all acknowledgement of the fact that our knowledge base concerning physical activity and its relationship to obesity and associated physiologic risk factors for chronic disease being less extensive than we would like it to be, I would nonetheless encourage you to -- to proceed with the development of a physical activity guideline, to include young people in that guideline, and to consider an application of that guideline to young people that would acknowledge the fact that they probably need more activity than is -- is currently being recommended in public health guidelines for adults in the United States. And I will be happy to try and respond to questions.
DR. GARZA: Any questions? Rachel?
DR. JOHNSON: Thank you very much for a great presentation. One quick question. Could you make sure we get the citation for those recommendations from the England --
DR. PATE: Yes, I will.
DR. JOHNSON: -- conference that you talked about?
DR. PATE: Yes.
DR. JOHNSON: And my other question is, as -- as you well know and showed, most of the research when you look at sedentary activity emphasizes television watching. And there is some supposition that it's more than just the TV watching. It's the ads for various types of food products. It's the fact that kids might be eating while they're watching television.
DR. PATE: Yes.
DR. JOHNSON: That even worsens that effect of television watching. We've sort of had the issue about whether we can extend the findings from television watching to time that children are spending in other very sedentary activities like computer games or video games. Are you comfortable with us making that connection because I'm not sure that the research has really kept up with the technology of how kids are spending their time?
DR. PATE: Yes, it hasn't. Yes, I have tried to keep up with that literature concerning TV watching, physical activity, and -- and body fatness measures. And my reading of that literature is that the associations between TV watching and fatness are tighter than the associations between physical activity and fatness. And of course the point you raise is often drawn in, and that is that TV watching may not only be an inactive pursuit, but one which either directly or indirectly impacts eating behavior, as well. And that's -- strikes me as a rational thought.
However, I think part of the problem with this may be that our measures of physical activity have traditionally been quite limited. And I'm not quite ready to conclude that in fact the TV watching fatness association is more powerful than the physical activity fatness association. I think we've got to do the studies using the more objective measures of activity before we firm that conclusion up.
Now --
DR. JOHNSON: I guess, you know, to ask you to bottom line it, would you feel comfortable with the recommendations that said limit the amount of time you spend watching TV, playing computer games, video games? I mean, would you feel comfortable --
DR. PATE: Yes.
DR. JOHNSON: -- linking those?
DR. PATE: Yes, I would.
DR. JOHNSON: Okay.
DR. PATE: And I would suggest that we do that. The rationale for that is probably not as -- something that we can document as extensively as we would like to. But, yes. And I think that the literature is growing. And of course the study that I cited here, the Gortmaker paper, is -- is the most encouraging one from an intervention standpoint that -- that's -- that's come into the literature so far.
DR. JOHNSON: Thank you.
DR. GARZA: Alice?
DR. LICHTENSTEIN: I'm interested in what your thoughts are on the determinants of activity. And is inactivity sort of the default mechanism? Because, you know, you showed a relationship between activity, let's say, and age, that it goes down with age. Well, if you think of an elementary school, recess time comes and there is suddenly a burst of activity when those kids are out. There is an opportunity for them to engage in active play.
You know, you presented data on level of activity among children, but we know that there are a lot of subgroups. If you live in a community where there is a girls soccer program starting in first grade, there is going to be a lot more physical activity associated with that group.
So, a) do you think that inactivity is sort of the default mechanism when there aren't other opportunities? And what are the major determinants that we should really be focusing on or maybe encouraging the development of in the guideline?
DR. PATE: Yes. The strongest predictors of activity in kids are -- are the demographic variables, age and gender, ethnicity, at least in some data sets. There has been a lot of focus over the last decade on psychosocial variables. Sort of health behavior models have been applied to these studies.
The -- I think the bottom line is there are a number of these variables that with some considerable consistency across studies do either cross-sectionally associate with or predict future physical activity variable -- behavior. Of all of those variables, the one that I believe most consistently and powerfully associates with and/or predicts activity is something that we call physical activity self-efficacy which basically comes down to a person's sense of confidence or competence in their ability to enjoyably engage in physical activity.
And a number of the intervention studies including those that we're working on now are focusing on trying to enhance physical activity self-efficacy on the premise that that's an important intermediate to behavior.
Now, having said that, I believe that -- that this work is now moving on from largely a focus on -- on this set of psychosocial variables that we think are important to social and environmental interventions. My personal belief is that kids -- not only kids probably -- but kids in particular are very reactive, responsive to the environment in which they find themselves, both from a social and physical environmental perspective.
Tom Verinowski's studies on young kids show, from the SCAN study some years, show that the strongest predictor of activity in that group of young kids was whether they were inside or outside. If they were outside, they were much more likely to be active than if they were inside.
So I think we -- we -- we now need a round of studies that really are carefully looking at social and physical environmental factors. I'm interested in day care centers and after-care centers; and is there a TV; is it turned on, you know, that -- that sort of thing as -- as potentially associated with activity. Not many of those studies have been done yet.
Jim Salis in San Diego has done some of that showing that the proximity of green spaces and parks and so on does -- does seem to have an impact on activity.
DR. LICHTENSTEIN: But is there enough data in the guideline to say that there should be more opportunity for children to engage in physical activity because that is a determinant of physical activity, I mean, you know, for the guideline now as opposed to what data is going to be collected in the future?
DR. PATE: I think there is sufficient data in the literature to -- to support the -- a comment on the self- efficacy piece. I think that's reasonably consistent and I think most that work in this area would be comfortable with that.
I don't think so yet on the -- on the physical and social environmental factors. I think that that's where this work is heading. But I wish we --
DR. GARZA: It's 10:45. I'm going to ask for you to please focus your questions. Otherwise, we're going to fall behind as we did at our last committee meeting. I - - I don't know whether it was -- Johanna, do you have a question before I move?
DR. DWYER: Yes.
DR. DECKELBAUM: Just one quick question. The issue about whether this is -- we're putting this in because it's linked to nutrition and weight reduction versus the benefits of exercise as Steve pointed out in his previous presentation. What is your view on that? And then maybe I could comment on my view just briefly? Well, maybe I'll just say that real quickly.
DR. PATE: I'm not sure I understand the question.
DR. DECKELBAUM: Okay. Well, whether we would put in a recommendation about exercise because it is a way to control obesity or whether there is a health benefit from it that's for all people unrelated to obesity. I guess that's --
DR. PATE: And, of course, in my case here, the issue is kids. And, of course, it's even more complicated. I -- I agree with something that Steve said. And that is the long-term tracking data are not terribly impressive; meaning if you look at activity behavior performed early in life, to the extent that this has been studied, it's not yet very clear that that impacts health outcomes later in life.
And the chronic disease outcomes that we're perhaps most concerned about don't manifest in childhood from a clinical perspective. And so it's difficult to -- to address these issues. We don't have the long-term prospective studies that would allow you to follow this from childhood on into adulthood to really nail that question down.
However, I will say that those of us that work with -- with kids and think about these kinds of things a lot -- this may -- this may be circular reasoning, but it goes like this. If -- if kids are this active -- and I've shown you some of the data that we have on how active they tend to be. And it's highly variable, a lot of -- a lot of variability.
And if, however, the -- the obesity rates are on the increase, and if you believe that activity almost has to be a factor in this -- in this increasing prevalence of obesity, then it would appear important to -- to consider activity in young people in the context of developing guidelines that would be intended to address this obesity problem.
DR. GARZA: Roland?
DR. WEINSIER: Yes, in trying to develop this guideline and think through the description of the type of activity, going back to the resistance/restraint training versus the aerobic, it seems to me that since 1965, there are about 33 national recommendations of which less than about 16 percent even address, you know, us making a recommendation with regard to strength or resistance training.
You're saying here that Jim Salis made a recommendation that it should include at least twice a week musculoskeletal strength and flexibility activities. So is this important that it should be included or we don't have enough data and that's why other recommendations don't include it?
DR. PATE: The experimental literature in young people looking at the effects of controlled exercise do provide some support for that recommendation. But I cannot tell you that at this point, we have compelling evidence to link performance of resistance exercise during childhood or adolescents to, let's say, bone health, you know, decades later. I'm not aware of evidence to support that.
Now, at this point, given the limitations in the literature, the thought process has been we certainly know from other studies that the development of bone and bone density and so on occurs early in life and that because bone density declines more or less steadily beyond early -- the ages of early adulthood, and since there is some evidence from the experimental literature that activity performed early in life can impact skeletal development, that it's early but may not too early to be saying something about the performance of resistance activity and the kinds of activity which are most likely to impact the development of bone.
DR. GARZA: Please, Dr. Pate, setting aside then the osteoporosis and obesity issues --
DR. PATE: Yes.
DR. GARZA: -- the other health benefits that we learned about in terms of physical activity in adults suggests that physical activity may also be important in children, but only if there is a significant behavioral carry-over.
Am I correct in concluding that in fact there is preliminary data that relates to self-efficacy, but no strong data that links physical activity patterns in children to sustained behaviors in adulthood that relate to physical activity?
DR. PATE: Yes.
DR. GARZA: But that's in absence of data. There's not -- or is it actually data that supports the idea that there is no connection? I mean, that was the sense that I got from Dr. Blair; that when people had looked at this, they had not found it.
DR. PATE: No. The -- the data sets that are available in which you might look at this are very limited. And, no, it's -- it's not that we know that those associations don't exist. It's that they have not been demonstrated at this point.
Now, I'm not -- I'm not here to argue that I'm absolutely positive that there is a powerful tracking of activity behavior early in life to activity behavior later in life. But I think it's too early to reject the hypothesis that that tracking may exist.
DR. GARZA: Okay. Then we're going to take a break. We'll be back by 11:00. How is that for optimism, all right?
(Whereupon, a brief recess was taken.)
DR. GARZA: Okay. We have a few committee members missing, but we're going to get started without them. Otherwise, we won't -- we won't get through the material we need to get through by 12:00 on Friday.
We're very pleased to have Dr. Ludwig, our next speaker, from Harvard University in Cambridge. He is going to help the committee wrestle with some of the issues on glycemic indexes and health disease relationships related to the glycemic index. We have another speaker tomorrow who will be addressing similar problems. We couldn't schedule them both on the same day. So this is the first half of a two-part presentation, one by Dr. Ludwig and the second by Dr. Sunyer from Colombia.
Dr. Ludwig, thank you very much for joining us.
DR. LUDWIG: Thank you, Dr. Garza and committee members, for inviting me to speak about glycemic index and human nutrition. It's a -- a great pleasure to be here today.
I would like to begin by discussing issues regarding the background of glycemic index, what is it, how is it defined, what studies have been done in general; and then focus on how glycemic index may affect obesity, cardiovascular disease and diabetes mellitus; and then conclude by a consideration of the significance of this to current nutritional recommendations.
As far back as the early 1900s, carbohydrate was thought to differ primarily according to chain length based on some classic studies of Allen involving dogs. Official recognition for this, the distinction between simple sugars and complex carbohydrate, was given in 1977. And current nutritional guidelines provide at least tacit endorsement for this distinction between simple sugars and complex carbohydrates.
Recently, however, the biologic significance of saccharide chain length has been questioned. Wahlquist and colleagues showed that consumption of glucose as a monosaccharide, oligosaccharide or polysaccharide -- in other words, starch -- produced similar changes in blood glucose and insulin levels.
Bantle in the New England Journal of Medicine in a classic paper 15 years ago showed no significant differences in blood glucose response to meals with sucrose, compared to meals with wheat among normal and diabetic subjects.
And in a recent rather provocative study, Rickard and colleagues demonstrated actually improved glycemic control in subjects with Type I diabetes after iso-caloric substitution of sucrose for starch.
A potentially more physiologic basis for characterizing carbohydrate is the glycemic index. Glycemic index was proposed by Jenkins and colleagues in 1981 as a measure of the rate of absorption of carbohydrate after a meal. It's affected by a variety of dietary factors that in any way alter gastric emptying, intestinal tract activity, or the rates of nutrient digestion and absorption.
It's affected by, for example, macronutrient composition in that protein and fat delay gastric emptying. Fiber content, fiber has a very important role in forming an emulsion in the small intestines, particularly soluble fiber, that serves as a mechanical barrier for the diffusion of glucose through the intestinal wall, and therefore delaying rise in blood sugar. And food form and method of food preparation also affect digestion, primarily through mechanical properties.
There are currently over 200 scientific articles on the subject in the world's literature. And the glycemic index of most commonly consumed carbohydrate-containing foods have been determined and reported.
Glycemic index is actually defined as the area under the glucose response curve after 50 grams carbohydrate consumption from a test food, divided by the area under the curve after 50 grams carbohydrate from a controlled food, generally white bread or glucose.
We find grain products in potato are very rapidly hydrolyzed into glucose in the digestive tract. In fact, in the refined form, there is essentially no rate limitation to this process. And, therefore, these substances have a very high glycemic index.
As suggested by the study by Rickard, many refined grain products, especially white bread and others, have a higher glycemic index than table sugar because sucrose is -- 50 percent of its constitute of saccharide structure is fructose. Fructose needs to go to the liver and -- to be metabolically transformed into glucose through a rather slow process. Fruits, vegetables and legumes, by contrast, have a low glycemic index.
Glycemic index has previously been shown to affect appetite in a number of single-meal studies. For example, addition in this Appetite 1988 paper -- addition of a low glycemic index, starch, bean flakes to a test meal delayed return of hunger and request for food in comparison to addition to -- addition of a glycemic index starch, potato, to the same meal.
In fact, 15 of 16 studies published to date show beneficial effects of glycemic index on either appetite, hunger or voluntary food intake, though it should be pointed out -- and I would like to emphasize -- that some of these studies do not control for macronutrient ratio, energy density, palatability or other potentially confounding dietary properties.
To further explore the issue of glycemic index and appetite regulation, we conducted a cross-over study looking at the effects of three different meals containing identical energy, but differing in glycemic index. The low glycemic index meal was a vegetable omelette with fruit. The high glycemic index meal, instant oatmeal with milk and sugar. And for the medium glycemic index meal, we used steel cut oats, a -- an old-fashion preparatory method that maintains the structure of the oat kernel intact, thereby lowering the -- slowing the rates of digestion and lowering glycemic index.
I would like to point out that the medium and high glycemic index meals were controlled for macronutrient composition, energy density, and in fact contained virtually identical foods, whereas the low glycemic index meal differed in many ways and was there to demonstrate the maximum effect that could be attained by using all dietary properties available to modify glycemic index.
And all meals -- just to emphasize, all meals for all subjects were isocaloric and determined as 18.5 percent of resting metabolic rate.
The subjects for the study were obese teenage boys, at least 120 percent of ideal body weight, but otherwise in good health. Each subject was studied three times in this cross-over design in the Clinical Research Center according to this protocol, admitted to the CRC in the evening, given a standard dinner, bedtime snack.
And then in the morning, an intravenous line was placed and one of these three meals was administered. For the next five hours, blood was obtained every 30 minutes for determination of hormones and metabolic fuels.
At lunch, the intravenous line was removed. A second meal identical to the breakfast meal was given. And then the subjects were told to request a large ad-lib test meal platter if and when they got very hungry.
They were encouraged to eat as much or as little as they wanted from these platters to feel satisfied. At the end of the afternoon, the amount of food remaining on the platters were quantitated for assessment of voluntary energy intake.
This slide depicts the change in blood glucose and fatty acids after the three test breakfasts. As expected, blood glucose was highest after the high glycemic index meal compared to the medium or the low. The ratio of areas under the curve here was four to two to one.
Note that at four and five hours after the high glycemic index meal, there was a relative hypoglycemia which amounted to ten milligrams per deciliter and significant at the 0.02 level in comparison with the other two meals.
Fatty acids were suppressed to a greater degree after the high glycemic index meal than after the other two meals. And thus the concentration of the two major metabolic fuels, glucose and fatty acids, were lowest after the high glycemic index meal, even though all three meals provided identical amounts of energy at T equals zero.
These changes in blood glucose and fatty acids can be understood by considering the secretion dynamics of insulin and glucagon. Insulin levels were highest after the high glycemic index meal owing to the rapid influx of glucose. And glucagon was suppressed after this meal.
The combination of high insulin and low glucagon would together promote uptake of glucose in muscle and in liver, suppress or inhibit release of glucose from the liver, and also inhibit the release of fatty acids from fat cells.
So after the nutrients of a high glycemic index meal have been absorbed from the gastrointestinal tract, glucose and fatty acids are low because the body has difficulty accessing its stored fuels under these hormonal influences.
The physiologic significance of these biochemical changes can be seen by examining the counter-regulatory hormones, epinephrine and growth hormone. After the low and medium glycemic index meal, epinephrine remained essentially flat for five hours, but surged after the high glycemic index meal, indicating that the body is truly under a metabolic stress. And similar changes were found with growth hormone.
This slide depicts the voluntary food intake after the three test meals in the afternoon. Subjects ate 53 percent more after the high compared to the medium glycemic index meal, and 81 percent more after the high compared to the low glycemic index meal, both comparisons statistically significant.
So in summary, from this first study, high glycemic index meals are associated with elevated insulin and suppressed glucagon levels, lower post-absorptive blood glucose and fatty acids; and a stress response in the post- absorptive period. There was also increased hunger. I didn't show you the hunger scale, but I did show you increased voluntary food intake.
Now, this study and others that I described at the beginning of the talk examine the short-term effects of glycemic index on energy balance. However, in the field of obesity, a central concept is that body weight may be regulated around a -- at least a set point or if not a set point, a set range, a concept that is based on several observations.
First and importantly, the very poor long-term outcome of conventional energy-restricted diets according to this NIH consensus statement several years ago, and also that under and over-feeding induce a sequence of metabolic adaptations that defend against further weight loss.
In a classic article by Leibel in the New England Journal of Medicine a few years ago, the team underfed either lean or obese subjects to obtain a ten percent decrease in baseline weight. And they found that as a result of that underfeeding, resting metabolic rate was suppressed by ten to 15 percent.
This body weight set point or set range is believed to have substantial genetic contributions, as evidenced by concordance in BMI among identical twins raised apart, and also the fact that many genes have recently been identified that affect body weight on a chronic basis.
Nevertheless,l environmental factors must also affect body weight regulation as simply demonstrated by the rising prevalence rates of obesity amongst genetically stable populations. This line of argument leads to the hypothesis that dietary composition influences the physiologic adaptations to energy restriction.
To test the study, we conducted a randomized cross-over study looking at the effects of high and low glycemic index energy-restricted diets. The diets were 50 percent of total -- predicted total energy expenditure, controlled for energy density and fiber content. However, they did differ in macronutrient composition -- we'll come back to this point -- in order to achieve maximal differences in glycemic index. And that will affect to some degree the interpretations we can make from this.
Outcomes were arresting energy expenditure, fasting serum leptin levels, nitrogen balance and voluntary food intake. And this is just to summarize the protocol. The subjects were admitted for a brief baseline evaluation and then randomized to receive one of the two energy- restricted diets for six days; and then were allowed to -- to eat ad-lib and at the end of the energy restriction, but re-examined on the high or low glycemic index diets.
As expected from the fact that the two diets provided identical amounts of energy, weight loss did not differ significantly over the six days of energy deprivation between the dietary treatments. However, serum leptin level was significantly different between the two treatments from day two onward.
Resting energy expenditure decreased by the expected ten percent on the high glycemic index diet, but decreased by less than half as much, 4.5 percent, on the low glycemic index diet, a difference which achieved statistical significance at the end of the evaluation.
Nitrogen balance was negative on the high glycemic index diet, but positive on the low glycemic index diet, approaching statistical significance. And this should actually read, "Nitrogen balance in milligrams per kilogram per day." Consistent with the single day studies, voluntary energy intake after energy restriction was higher after the high compared to the low glycemic index diets.
And in conclusion from this study, diets with identical energy content may affect leptin levels differently compared to a high glycemic index diet in accord with current nutritional guidelines in that it was a low fat, grain-based diet. The low glycemic index diet employed here showed beneficial effects on energy expenditure, nitrogen balance and voluntary food intake.
Now, because the diets differed in macronutrient composition in order to achieve substantial differences in glycemic index, the effects seen cannot be definitively attributed to glycemic index alone.
Nevertheless, the study I believe argues that the physiologic adaptations to energy restriction can be influenced by dietary composition. And these results are potentially also attributable to glycemic index, as well require further investigation.
This brings us to I think the central question which is does glycemic index affect body weight regulation over the long-term. To date there are, unfortunately, no randomized controlled perspective clinical trials of a low glycemic index diet and the treatment of or prevention of obesity.
Nevertheless, I believe that there is substantial theoretic basis for believing such an effect may exist. High glycemic index diets, as has been demonstrated here and is quite clear that this happens, elicits higher insulin levels than low glycemic index diets. This relative hyperinsulinemia on a high glycemic index diet may predispose to weight gain. This contention is supported by a variety of lines of investigation.
Amongst animal studies, prior insulin treatment of normal rats stimulates glucose uptake in the fat, but not i the muscle; increases food intake; and promotes weight gain. Turning now to human studies, individuals with the greatest intravenous -- with the greatest insulin response to IV glucose prospectively gain the most weight.
Fasting hyperinsulinemia also predicts weight gain in Pima Indian children. And as is becoming I think abundantly clear in the literature in the last few years, excessive weight gain is a complication of insulin treatment of Type II diabetes and also, for that matter, intensive insulin treatment for Type I diabetes as seen in the DCCT.
And then finally, returning to animal studies and looking specifically at glycemic index -- high glycemic index starch consumption compared to low glycemic index starch consumption resulted in increased fatty acid synthase activity, increased adipocyte size, and increased glucose incorporation in total lipids in rats.
Several studies suggest that glycemic index -- low glycemic index diet may be beneficial for certain cardiovascular disease risk factors and perhaps myocardial infarction rate itself.
The first study in this area was published by Jenkins and colleagues in 1987. They conducted a two-week cross-over study involving six healthy men. The diets were higher low glycemic index, identical in energy, macronutrients and fiber. The meals were prepared in a dietary kitchen. And importantly, there was no change in body weight during either treatment.
They found that C-peptide excretion, a measure of integrated insulin levels, was lower on the low glycemic index diet than on the high glycemic index diet, quite a -- I think an impressive difference.
Total cholesterol and fructosamine -- fructosamine being an integrated measure of blood glucose levels -- decreased significantly only on the low glycemic index diet, not on the high glycemic index diet.
And the effects of glycemic index in subjects with hyperlipidemia were examined in a one-month uncontrolled study with 30 subjects. The intervention consisted of substitution of low glycemic index for high glycemic index foods without change in energy content, macronutrients or fiber. The meals were prepared at home according to standard recipes. And I think it would be very important to note that in this uncontrolled study, body weight change during treatment was minimal -- was minimal.
They found that total cholesterol decreased by 7.7 percent, LDL cholesterol by 8.5 percent and triglycerides by 15.3 percent, whereas there was no deterioration in HDL cholesterol.
Recently, Jarvy examined 20 patients with Type II diabetes in a one-month cross-over study. The diets were similar in energy, macronutrients and fiber, but differed in glycemic index. The meals were prepared in a diet kitchen and there was some weight change on treatments, about 1.5 kilograms over the month. But the change happened to be identical between treatment groups allowing for a fair comparison.
They found that plasminogen activator inhibitor 1 activity decreased by over 50 percent on the low glycemic index diet, but did not change on the high glycemic index diet. PI-1 is a newly identified cardiovascular disease risk factor that may relate to abnormal fibrinolysis and atheroma formation.
They also note that total cholesterol, HDL cholesterol and fructosamine were significantly lower on the low compared to the high glycemic index diet.
Two large epidemiologic studies have examined the relationship between glycemic index and cardiovascular disease. In a survey of British adults initially performed in '86 and '87, 1,420 adults were examined in a cross- sectional analysis. Their diets were determined by a seven- day food diary. The data were controlled for age, sex, BMI and a variety of other cardiovascular disease risk factors.
They found that glycemic index was the only dietary component associated with HDL cholesterol in both men and in women. No other dietary factor including total or saturated fat was seen in this epidemiologic analysis to have any predictive effect.
In the Nurses Health Study, 75,000 women were examined prospectively over ten years. Diet was assessed by a food frequency questionnaire. Data again were controlled for age, smoking and other coronary risk factors. They found that individuals in the highest quintile of glycemic load, here defined as glycemic index times the amount of carbohydrate in the diet in percent -- individuals in this highest quintile of glycemic load had a two-fold higher greater relative risk of myocardial infarction compared to those in the lowest quintile.
Now, this result certainly seems quite intriguing. But I will point out that it's present only in abstract form. The full report has yet to be published and is subject to the attendant peer review process.
Two epidemiologic studies suggest a connection between glycemic index and risk for Type II diabetes mellitus. In the Nurses Health Study, 65,000 women were followed prospectively over six years; diet again assessed by a food frequency questionnaire. And in light of earlier presentations, it's important to note that the data were controlled not only for BMI, but also for physical activity and other risk factors for diabetes.
They found a 37 percent increased risk of Type II diabetes in the highest quintile of glycemic index, here glycemic index rather than glycemic load, compared to those in the lowest quintile. And virtually identical results were found in the Health Professionals Follow-up Study.
Use of glycemic index in the treatment of diabetes has been very controversial. A number of studies over the past two decades have shown improvements in blood glucose, glycosylated hemoglobin, fructosamine and C-peptide excretion on low compared to high glycemic index diets.
Nevertheless, the clinical significance and practical applicability of these findings have been hotly contended by other authors. And I just point out the titles of some of these articles to indicate the ardor with which this argument has been waged in the literature.
Currently, the American Diabetes Association does not recognize a role for glycemic index in the treatment of diabetes mellitus. However, a number of diabetes organizations outside the United States do recognize a role for this factor.
In summary, a substantial body of research suggests that glycemic index may affect several chronic diseases in the United States, though I think I would have to say that the data to date falls short of proof. Regarding obesity and body weight regulation, possible mechanisms may include decrease in the insulin to glucagon ratio, increase in the concentration of metabolic fuels, and thus better control of hunger and appetite.
Fifteen of 16 short-term studies show beneficial effects of glycemic index in this regard. Nevertheless, these studies can have difficulty in controlling for other dietary factors such as macronutrients, fiber and -- and food form and the like because many of these factors are intimately related to the determination of glycemic index.
And most importantly, there are currently no long- term randomized control prospective clinical trials of a low glycemic index diet and the treatment or prevention of obesity.
Regarding cardiovascular disease, possible mechanisms I think importantly include reductions in hyperinsulinemia which I think is becoming increasingly well documented to be an independent risk factor for cardiovascular disease thought central in the etiology of the metabolic syndrome, or syndrome X.
A handful of studies have demonstrated beneficial effects on lipids, fibrinolysis and possibly risk factor for myocardial infarction itself -- possibly for myocardial infarction itself, though the total study number is small. Some of these studies have few subjects. And again, there were no long-term clinical intervention studies.
Diabetes prevention, glycemic -- low glycemic index diets may operate by a -- in effect, a beta cell rest mechanism and increased insulin sensitivity. Blood sugar -- blood sugar rises markedly less high after a low glycemic index diet. Therefore, there is less secretory demand placed on the beta cells. And there -- there are in fact some clinical trials designed to prevent the progression of Type I diabetes through this similar sort of mechanism, by providing insulin and helping to allow the beta cells to rest.
Two major epidemiologic analyses show protection by low glycemic index diets. These studies being epidemiologic in nature may be confounded by uncontrolled variables and, importantly again, no long-term interventional studies have bene conducted.
Finally, regarding diabetes treatment, protective mechanisms may involve decreased glycemic excursion around the meals and increased insulin sensitivity. A majority of the studies show some sort of clinical benefit. But the area is hotly contended.
So finally, I would like to consider the implications of glycemic index to nutritional policy. Current dietary guidelines recommend the consumption of a grain-based, low fat diet. And in fact, dietary fat has decreased, at least in percentage of total energy, from 42 percent to 34 percent since the 1960s. And because these two macronutrients are so intimately linked, there has been a well documented compensatory increase in dietary carbohydrate during this time.
Unfortunately, most of the carbohydrate that is being consumed in increasing amounts are very high in glycemic index. According to data from the continuing survey of food intake by individuals, these are the top 20 sources of dietary carbohydrate for children ages two to 18 in the United States.
Actually, perhaps ice cream here should be in -- in red here. But most of these items in white have a glycemic index equal to or greater than table sugar. Perhaps an over-emphasis on dietary fat without sufficient attention to the quality of carbohydrate in the American diet may have paradoxically contributed to the rising prevalence rates of obesity, Type II diabetes, and certain cardiovascular disease risk factors in the United States.
And in my last slide, I would like to present what might be an alternative low glycemic index pyramid that would include fruits and vegetables at the base, consumed in abundant quantities. These foods are not only low glycemic index, with a few exceptions. Banana is a notable exception here. Corn, beets and carrots are exceptions here; and of course, potatoes.
But most of these are quite -- not only quite low in glycemic index, but they also have the added benefit of being very low in energy density compared to the current pyramid base.
There would also be moderate intake of protein and healthful oils. These items would help lower the glycemic index of the diet. There would be moderate -- modest intake of unrefined grain products and pasta. Pasta is unique amongst the refined grain products in having a -- a relatively lower reduced glycemic index. And there would be also substantial reductions in the intake of refined grains, potato and concentrated sugars. Thank you very much for your attention.
DR. GARZA: Thank you very much. I have difficulty getting this thing on.
DR. MURPHY: Thank you for a very interesting presentation. Can you help me understand a little bit better the role of macronutrients on glycemic index? For example, are -- if all else is equal, are higher fat foods lower glycemic index than low fat foods?
DR. LUDWIG: I want to preface my remark by saying I'm not encouraging a very high fat diet. But, yes, glycemic index is -- or the glycemic load, which I think is a useful concept that actually Walter Roet proposes, really represents the primary stimulus for raising insulin levels. And insulin is this extraordinarily potent substance that can probably shift the metabolism toward anabolism.
So glycemic load is a product of the glycemic index of the individual foods times the total amount of carbohydrate. So as dietary fat -- dietary fat decreases, all things being equal, glycemic load will increase. And as is not necessarily the case, but unfortunately the case, as dietary fat intake is reduced, those calories are replaced not with fruits and vegetables, but refined starchy foods and sugars. And so there is a pretty good relationship between -- an inverse relationship between dietary fat and glycemic load.
DR. MURPHY: And so, for example, are