324‒Metabolism, energy balance, and aging: the impact of diet, calorie restriction & macronutrients

1. 0:00 – 1:34

   Intro

   1. PAPeter Attia0:00

      I think people want to believe that there is one perfect diet, one perfect way, but I just don't buy it.

   2. EREric Ravussin0:08

      No, I, I'm, I'm with you, and I think that we are now in the era of precision medicine or personalized medicine, and I think that this is what the NIH is embarking on when they do this Nutrition for Precision Health, when they do this study that I'm also a PI of clinical site, which is molecular transducers of physical activity, to look at in lieu of your genetic background, your environment, your socioeconomical status, one size does not fit all. We have the dietary guidelines. They are applied for the entire nation, and they tell you, you know, every time it's the Med- Med or the Mediterranean diet or the DASH diet which are the best, and all these kind of things, but not for all. And I think that this is where we are going to have a huge development. It's to go... I don't think it's going to be individualized, but at least for groups of people having the same, you know, you know, different strategies, and restriction has to be one of that.

   3. PAPeter Attia1:23

      [upbeat music] Hey, everyone. Welcome to "The Drive" podcast. I'm your host, Peter Attia. Well,

2. 1:34 – 7:18

   Eric’s background and current work metabolism and measuring energy expenditure

   1. PAPeter Attia1:34

      Eric, thank you so much for joining me. Um, I'm sorry that we can't be in person, but given that you're in the south of France and I'm over here in Texas, that would have made it pretty difficult, um, 'cause normally we're quite a bit closer. You're obviously in, uh, in Louisiana. Um, but anyway, it's wonderful to see you again. It has been close to 10 years, right?

   2. EREric Ravussin1:54

      I think so. I think so.

   3. PAPeter Attia1:57

      Yeah, too long. [laughs]

   4. EREric Ravussin1:58

      But I have, I have, uh, found memories of our interactions.

   5. PAPeter Attia2:01

      Yes, as, as do I. Um, I guess by way of background, maybe we can even just give people a bit of a sense of how closely we did work together for, for four years. Um, at the time, um, I was part of an organization called NUSI, and you were one of the six principal investigators for, um, what was a, a very audacious experiment, um, and, and part of something called the Energy Balance Consortium, along with, uh, Kevin Hall, Steve Smith, Rudy Leibel, Mike Rosenbaum, and, um, [clicks tongue] God, I'm... Who am I missing? Uh, Mark, Mark, um... Sorry, I'm blanking on Mark's last name at the NIH. Uh, M- Mark Reitman.

   6. EREric Ravussin2:41

      Mark Reitman.

   7. PAPeter Attia2:42

      Yeah.

   8. EREric Ravussin2:42

      Yeah.

   9. PAPeter Attia2:42

      How could I... Sorry. Apologies to Mark. Um, and nevertheless, uh, so this, this, this was kind of a, a, a once in a lifetime opportunity to do something that had never been done before, which was a multicenter i-, uh, uh, indirect calorimetry inpatient study that was really aimed at trying to at least pilot what would be necessary to do the definitive experiment to answer a question that I'm sure we're going to talk about a lot today, which has to do with the, with the nature of calories and weight gain. Um, so I, I will leave it at that only by way of background to say that we had the, the incredible pleasure, uh, of working together and, uh, and really getting into the weeds of, of how one technically tries to measure these things and how difficult it is, um, and, and, and all of the, the challenges and potential pitfalls of this. But let's just take a step back and, and maybe give folks a sense of your, your life's work and your life's passion. Um, you, you've been the, the head of, of Pennington Biomedical Research for how many years now? Is it 20 years?

   10. EREric Ravussin3:40

       I have been, uh, for 24 years at Pennington. I was associate executive d- uh, director for clinical science. I'm not the head of the Pennington. Uh, but, uh, yeah, I joined, uh, after spending two and a half years at Eli Lilly, and before I was 15 years intramural NIH. And the reason I came to this country was to build the first metabolic chamber or indirect calorimetry chamber to measure energy expenditure in people over periods of hours and days.

   11. PAPeter Attia4:14

       Tell folks a little bit about what it means to be intramural versus extramural NIH. I think when people hear NIH, they sort of have a vague sense of what it is, um, but maybe that distinction isn't clear to most people.

   12. EREric Ravussin4:26

       Yeah. The, the NIH budget is about $33 to $35 billion, uh, but there is a group which is intramural and, uh, but most of the money, 80 to 85%, goes to all the academic institution doing biomedical research in the country. Now, intramural, uh, I, I like to be intramural.

   13. PAPeter Attia4:49

       And, and just so folks understand what that means, that means you are an employee of the NIH. You work at the NIH campus, and your funding comes from the organization that you work for, unlike someone who's at Stanford or Harvard who's receiving extramural NIH funds.

   14. EREric Ravussin5:04

       Correctly. And I was intramural, but based in Phoenix, Arizona, because I worked a lot with the Pima Indians, who have, as you know, the highest prevalence of type 2 diabetes and the second highest prevalence of obesity in the world. But, uh, intramural, what I like, you are judged after the facts. You have a budget, and you can do whatever you want, but every three years you are judged, whereas extramural, as you know, you, we chase these grants, and you have to basically provide preliminary data, hypotheses, and you are judged before doing the study. And I think you can be much more creative being intramural than extramural. I'm sure that my colleagues intramural are going to hate me that I say that, but it's true. They should enjoy to be judged after the facts rather than before. [laughs]

   15. PAPeter Attia5:56

       Tell us a little bit about Pennington. It's obviously a pretty unique institution, um, and, and you've been there, as you say, for 24 years. But yet I suspect it's a, it's a place that most people listening to us won't necessarily be familiar with.

   16. EREric Ravussin6:09

       Yeah. It's becoming, I think, the largest institution for research in nutrition and obesity. Uh, we concentrate mostly on nutrition and obesity. Uh, we have about 500 people working at Pennington, and everybody is doing research. We have no teaching. I mean, yes, I have post-docs and things like that, but no formal teaching. Uh, and we do research, and we chase, uh, we have a maybe a budget of, uh, functional budget of 80 million per year. Uh, 50% or 45% of that is NIH extramural money coming to Pennington. Uh, we do some sponsored project for pharma or biotech. Uh, we have grants from association like the American Diabetes Association or American Heart Association and so on, and we do, uh, basic science, clinical science, and population science, basically, which is to reach the community and implement, uh, you know, some of the discovery into the community.

3. 7:18 – 17:47

   The science behind metabolic chambers for measuring energy expenditure, and the complexities of indirect and direct calorimetry

   1. PAPeter Attia7:18

      The other thing you mentioned a few minutes ago, um, Eric, was the words metabolic chamber and indirect calorimetry. And I think at some point, because this is going to come up in our discussion as we get into the weeds on metabolism, uh, it's gonna be necessary, I think, for the listener to kind of understand how those techniques work. Uh, and I've-- I know I've discussed this on previous podcasts. There's at least probably half a dozen episodes where I've explained how one uses the ventilation of oxygen and CO2 to quite accurately estimate energy expenditure and f- uh, substrate utilization. But it can never be a bad thing to explain it one more time. So-

   2. EREric Ravussin7:59

      [laughs]

   3. PAPeter Attia8:00

      ... why don't you explain how it works, and, uh, you can even use the example of when I was in the chamber for a couple of days, you know, 10 years ago when, when I was out at Pennington as a subject or-

   4. EREric Ravussin8:11

      Yeah

   5. PAPeter Attia8:11

      ... you know, pretend subject.

   6. EREric Ravussin8:13

      Yeah. It-- There is two, two things, indirect calorimetry and direct calorimetry. You generate your energy metabolism or ATP by oxidative process, and you oxidize carbohydrate, fat, and a little protein. You don't want to oxidize too many proteins. And this, uh, basically build up these ATPs, uh, which are used online for generation, uh, or, or maintenance of the cells or activity and so on. Now, as a, you know, when you have this metabolism, you produce heat. If you have no exercise, no external work, all the energy which is, uh, provided in generating this ATP is lost as heat, and this is direct calorimetry. And I was fortunate when I did my PhD back in Lausanne, Switzerland, to have access to both indirect and direct calorimetry. And you, you can really do fantastic studies. We did a study of measuring in vivo what we call the PO ratio. You know, how many oxygen do you nu-- do you need to generate one ATP and all these kind of things by the combination of these two techniques. Now, going to indirect calorimetry, this is an easier one than direct. I mean, direct, you have really-

   7. PAPeter Attia9:43

      And, and sorry, just to interrupt you for a second, Eric. Um, when you're doing direct calorimetry, obviously, when it comes to something like food, if you want to know the energy content of food, you can burn the food and measure the heat. When you're doing it otherwise on a living organism, is it possible to do direct calorimetry on something as large as a human, or is that something you're typically doing on, you know, a mouse or something that is small enough that the chamber in which you can measure changes in heat is much smaller and easier to manage?

   8. EREric Ravussin10:11

      Yeah. Direct calorimetry is really measuring heat losses, and we lose heat by convection, evaporation, and we measure these two kinds of heat, and this is basically equal to your metabolic rate. You generate energy by oxidation of substrate, and the byproduct is heat. And without exercise or external work, energy in is equal energy out, and this is what we measured, and I did a lot of these studies during my PhD. But then at that time, we decided to, uh, build a metabolic chamber, which is like an hotel room, uh, in which you can live one day or two days, and I think you stayed for two days in one of these rooms at, uh, Pennington in Baton Rouge. And here, we measure oxygen consumption and CO2 production. Oxygen is used to oxidize the substrate. It produces CO2 and water. And knowing oxygen consumption, CO2 production, you can calculate the energy expenditure or the energy generated by this oxidative process, and you can also calculate the substrate that you oxidize. If the ratio between the VCO2, the CO2 production, and the oxygen consumption is one, you oxidize carbohydrate. If it's 0.7, you oxidize fat. And protein is something in between 0.82, depending on the protein.

   9. PAPeter Attia11:52

      Yeah. And so basically, by measuring just these two things, um, the actual concentration of oxygen that's being consumed and the rate at which it's being consumed, and the same with production of CO2, you can really infer two things. One, the total energy consumption, total calories consumed, kcals. And by looking at the ratio of the, uh, VCO2 and the VO2, you can calculate what fraction of that is coming from oxidizing carbohydrate versus fat. And, and, and, and, you know, having spent probably a total of eight days of my life inside metabolic chambers across a period of a couple of years, it never ceases to amaze me, never, just how accurately these rooms work. Because I use an indirect calorimeter all the time measuring VO2 max, which is another type of... It's, it's basically the exact same technology, but it's easier because you're strapping the mask to your face, so you have perfect access to exactly the, uh, O2 consumption and the CO2 production. But it's doing the exact same calculation. It's going to tell you exactly how much energy you're utilizing and, um, where the substrate comes from. And of course, in that situation, you're interested in knowing the maximum amount of VO2 or O2 consumption. Um, so, uh, what do you... what, what would you estimate is the error on, uh, the, the hotel room sized indirect, uh, calorimeter where you don't have the luxury of just slapping a mask to the person's face for a few days?

   10. EREric Ravussin13:29

       Yeah. When I was at NIH, we did a lot of reliability, uh, testing in the same person, or you can calibrate the system. You calibrate the system with, uh, standard gas that you know the exact concentration to calibrate the analyzers.

   11. PAPeter Attia13:47

       Mm-hmm.

   12. EREric Ravussin13:48

       But you can also mimic someone by burning alcohol or propane, and you can vary the rate of burning. And this is what we did when we did this study with four different chambers. We, uh, validated the chambers once against the other one, and we accepted to have 3% deviation for CO2 or oxygen based on stoichiometry. You know exactly-

   13. PAPeter Attia14:16

       Yep

   14. EREric Ravussin14:16

       ... how much alcohol or propane you burn. You know how much CO2 should be produced and m- how much oxygen has been consumed.

   15. PAPeter Attia14:24

       Yeah.

   16. EREric Ravussin14:24

       And this is what we did. And you are right. They are very, very precise, and you don't have the inconvenience of the mask. I mean, you couldn't measure with a mask for 24 hours or two days.

   17. PAPeter Attia14:35

       Right. Because when I was in that chamber for a couple of days on separate occasions, we were trying to replicate as much as possible my life. So I had three meals prepared. I had an exercise bike in there. I had, uh, um, I had weights in there, equipment to l- to do everything, and we were very interested in seeing how that would compare with, for example, what we knew I was experiencing in the outside world based on, um, when we could put a mask on me. And also separately, if you may recall, Eric, we also spent, uh, about 10 days using doubly labeled water, which I'm sure we'll talk about at some point as well as a, um, a more real world example of this. So all of that is to say that, um, scientists such as yourself have some pretty high fidelity tools to measure energy expenditure. Um, but it comes with a caveat, which is it has to be done under this very controlled setting. Um, these people need to be in the hospital, and not just in a hospital, but in a metabolic ward and inside a chamber, which is, you know, basically a NASA grade gas chamber with, uh... H- how many sensors, by the way, just out of curiosity, how many O2 and CO2 sensors would be in a particular chamber?

   18. EREric Ravussin15:55

       Uh, we don't need many because we need to homogenize the air in the room.

   19. PAPeter Attia16:00

       I see. So that's the bigger effect.

   20. EREric Ravussin16:01

       And what you sample in one corner of the room is the same as the other corners.

   21. PAPeter Attia16:05

       Mm-hmm.

   22. EREric Ravussin16:05

       Uh, and, uh, that's why. But we have to measure humidity. You have to measure, uh, temperature because you have to do this correction, what we call STPD, which is standard pressure, et cetera, to calculate the energy generated by this oxygen. But, uh, yeah, I mean, it's, it's quite phenomenal the accuracy of those. And to go back to your question, the accuracy is, is, uh, or the precision is about two and a half percent. Now, if you repeat yourself five times or... and myself, there's a little bit more variability because we vary from one day to another one and so on. But I mean, this is quite amazing that we can calculate so precisely your energy expenditure during a day or two days and so on. I wish we had the same system to calculate the energy that we put in.

   23. PAPeter Attia17:01

       Yeah, that's a great point. We... Outside of having food prepared in a metabolic kitchen, which of course was also done for, for, for this experience, um, we're-- I think at best we're probably plus or minus 10%, right? I mean, it's very difficult to estimate [chuckles] energy intake.

   24. EREric Ravussin17:19

       And if, when it's in free living, it, it's, it's worse.

   25. PAPeter Attia17:22

       Yeah.

   26. EREric Ravussin17:23

       I mean, you know, when you do that under supervision and you provide all the meals like Kevin Hall is doing, like we're doing, then you have a good estimate, and you subtract what is returned on the plate or this kind of things. But it's not, it's not normal life. And, uh, as soon as you go to normal life and free living conditions, things are falling apart.

   27. PAPeter Attia17:46

       Well, I

4. 17:47 – 25:00

   The body's regulatory systems for maintaining energy balance and the primary influence of energy intake on body weight

   1. PAPeter Attia17:47

      think, I think that is a, a really nice, uh, intro into kind of a philosophical question that will take us towards some of the experimental work. So, um, I was actually gonna launch into the explanation of the energy balance model and, and talk about the, the work that you and, and, and the other folks that we just talked about did a decade ago. But, but let's pause for a moment on the statement you just, you just said, which was, look In the free-living environment, we cannot estimate energy intake worth a damn, and, and m- which is to say that 99% of people out there who aren't weighing every single morsel of food they eat every single time they open their mouth, which again, is virtually nobody, um, are going to have enormous variability in what they're eating. And so I put myself in that category, right? I pay no attention to how much I'm eating in terms of weight. I'm just sort of eating by kind of my appetite. Uh, sometimes I overeat too much. My appetite gets the best of me. I'm surrounded by, you know, highly palatable foods that I overeat, and other times, you know, I'm just really busy. Like yesterday, I just didn't get around to eating that much. But here's what's interesting, Eric. My weight lives within about a two-pound, uh, variation for years, and I don't think I'm unique in that. I would guess that you're pretty similar as well. I think many of us find ourselves in kind of a couple of pounds of variation, most of which is probably water weight varying over years, and yet we're not really paying attention in excruciating detail to how much we eat. So first of all, do you agree with the assessment that that's not an uncommon phenomenon?

   2. EREric Ravussin19:31

      Yes, but on the other hand, if you take your lifetime history-

   3. PAPeter Attia19:36

      Mm-hmm

   4. EREric Ravussin19:37

      ... maybe you are an exception, but, uh, I have gained some weight. When I got married, uh, in 1975, uh, I weighed 69 kilos. Now I, I weigh close to 77. Uh, now it's 50 years later, of course, but [laughs] uh, you know, you, you have this change. You have also a change in your body composition.

   5. PAPeter Attia20:00

      Yeah.

   6. EREric Ravussin20:01

      Uh, and it's very rare that people are staying the same, uh, constantly. I mean, you may be an exception.

   7. PAPeter Attia20:09

      No, no. I, yeah, I, I, I would not suggest that I'm an exception. I've had probably more variation in my weight than even what you're describing over the period of time. But when... I guess the question I'm getting at, Eric, is how does the body regulate energy storage, which is effectively what is determining our weight? It's how much are we storing as fat? How much are we storing as lean tissue? Um, that seems very, very regulated despite very little control over the input and the output. We're free-living animals who, you know, live in an environment with endless inputs, so we have to somehow regulate that, um, and frankly, very little reason to do deliberate outputs, right? Like, obviously, the majority of our energy expenditure is based on essential function of life, um, you know, just the, the, the necessary ATP production, as you described it, for cellular function. Um, but it's not like most of us are out there expending energy deliberately, i.e. exercising and moving, you know, uh, nec- necessarily to, to expend energy. So, uh, as we, as we st- enter this discussion of the energy balance model, let's just talk about energy balance. Why are we so largely able to conserve mass despite very noisy inputs and outputs into the black box?

   8. EREric Ravussin21:39

      Yeah, I, I think I had an answer, but, uh, I hope that I had an answer, but I don't, and I think that we are still... You know, when, uh, leptin was discovered, it was a haha moment. We have a signal which comes from the energy storage, and therefore, you know, it seems to regulate food intake, also energy expenditure and all that. I was fortunate to have access to, uh, the three Turkish leptin-deficient people that we measured in these chambers in, in, in, uh, Baton Rouge, and, uh, we thought that we would have a clue, and it was not the case. We know that-

   9. PAPeter Attia22:20

      [laughs]

   10. EREric Ravussin22:20

       ... leptin is very good to defend against the loss of weight when it becomes extreme, but it doesn't work on the other side of the equation when you creep up, you know, with your weight because you become leptin resistant and, you know, your leptin is not telling you stop to eat. And I, I think that, uh, this, this was, uh... We learn a lot about the biology of energy balance thanks to leptin, but then we need to think that we have also signals coming from the fat-free mass. Now, what is the fat-free mass? You know, we measure it as everything but the fat, and we call that fat-free mass. But, you know, it's liver. It's, uh, it's muscle, skeletal muscle. It's everything but the fat, and I think that there are signals that we are still chasing.

   11. PAPeter Attia23:16

       Mm.

   12. EREric Ravussin23:16

       You know, John Blundell showed that we eat according to our fat-free mass, which is basically we eat according to our resting metabolic rate because fat-free mass is the major determinant of your resting metabolic rate. But what are the signals?

   13. PAPeter Attia23:33

       Right.

   14. EREric Ravussin23:33

       I don't know. I mean, I, I still don't know.

   15. PAPeter Attia23:36

       [laughs]

   16. EREric Ravussin23:36

       And I, uh, we have to be, uh, you, you know, modest here and say there are still things that we don't understand when it comes to this energy balance. Now, I can understand that why, why do you, you don't weigh, uh, gain five kilograms over this coming year. Because, you know, in response to creeping up your weight, you burn more energy. You know, if you do the same physical activity, this extra five kilos or 10 pounds is gonna cost you more energy. Your resting metabolic rate is creeping up, and basically you offset that. But why are we so good at maintaining, like you said, within a couple of pound over a year or two years or three years or five years? And I think that we are still missing some signals, but I think that there are signals coming from, you know, you, you... I-- When FGF21, which is a signal-

   17. PAPeter Attia24:36

       Yes

   18. EREric Ravussin24:36

       ... coming from the liver was discovered, I said, "Oh, maybe that it." Or when you have some of the, uh, myokines, you know, coming from the skeletal muscle, I said, "Maybe this is it." And it's, it's, it's not been it, and I think that we have to be modest and say, "Hey, we still have things to discover when it comes to this regulation of energy balance."

5. 25:00 – 29:15

   The epidemic of obesity and a discussion of resting metabolic rate

   1. EREric Ravussin25:00

      But as a population, we have not been very good. I mean, from the 1980s to 2010, the American population has gained 10 kilograms. Uh, you know, it's 22 pounds in, uh, 30 years. It means that there is still the major driver is this environment, and the change in the environment has been the trigger of this weight gain. But of course, people, uh, some people are successful, some are less, and, uh, this is a problem. And of course, now we have an epidemic of obesity and, uh, you know, it has replaced the contagious, the disease of the past, and now we have all this constellation of conditions which are associated with obesity.

   2. PAPeter Attia25:50

      So Eric, if you were to guess, which I think is the best we can do, although I think it's informed guessing, I don't think we're just guessing in the dark. How much of the, uh, regulation do you think is occurring on the appetite side, meaning feedback from, uh, feedback from whatever is happening in the system, and let's just posit, as you said, that it's coming from the lean part of the system, so the organs, the skeletal muscles, all but the adipose tissue, and that that is working to feed into the appetitive system? Uh, or do you think the balance is more on the expenditure system, where those signals are feeding to make us move a little bit more? Uh, or simply, as you say, just it's the extra weight you have to move around. If you, if you're gonna carry... You know, people should pick up what 10 kilos is. That's an awful lot of weight. If you have to carry an extra 10 kilos up a flight of stairs, and every time you get up and move around, you're going to expend more energy. Um, so, so, so do you have a sense of the balance of that, uh, uh, of, of that regulatory pressure?

   3. EREric Ravussin27:02

      Yeah. Uh, Peter, when I first joined the NIH, I, I did a study in Pima Indians, who are very prone to weight gain, as you know. And we measure in 150 people the 24-hour energy expenditure as well as the resting metabolic rate. And we found that there was large variability between people after adjusting for their body size, fat free mass and fat mass. Now we do better because we have organ size and we do MRI and all these kind of things. But there is variability, and this variability is associated with family membership. In other words, we can say that there is a genetic background to that. But what we found was that low-- those people in the lower tertile of this metabolic rate adjusted for body size were at much higher risk of gaining weight. But when we try to attribute the weight gain either to energy expenditure being low or the other-

   4. PAPeter Attia28:07

      Intake being high

   5. EREric Ravussin28:08

      ... side of the equation-

   6. PAPeter Attia28:09

      Yep

   7. EREric Ravussin28:09

      ... intake was 80%. It means, uh, you, you know, I have always this, this, uh... It reminds me there are so many system which are, uh, basically regulating energy expenditure and food intake. Take nicotine. Nicotine is a stimulant of energy expenditure. Smokers have higher metabolic rate, but they have less appetite. Take the activity of your sympathetic nervous system. It is also thermogenic, but it ca- it costs, uh, it cut appetite. And I think that we have to understand better these two system. But to answer your question, I would say 80% is on the side of the energy intake.

   8. PAPeter Attia28:56

      Intake. Yeah. Okay. That's, that's sort of been my intuition as well. Um, but because I don't keep up with this literature nearly as much as I used to, I, I, I wanted to make sure that that was still l- largely the, the, the view of-

   9. EREric Ravussin29:10

      Yeah

   10. PAPeter Attia29:10

       ... of the people at the forefront of this field.

   11. EREric Ravussin29:12

       And, and you know, for example, uh, sorry

6. 29:15 – 40:11

   The impact of exercise, appetite, gut hormones, and eating patterns on weight regulation

   1. EREric Ravussin29:15

      to interrupt, but-

   2. PAPeter Attia29:15

      Yeah

   3. EREric Ravussin29:15

      ... uh, we know that, uh, exercise is pretty bad for weight loss.

   4. PAPeter Attia29:21

      Yeah.

   5. EREric Ravussin29:22

      I- if you just tell people, "Okay, go to the gym, like, three times," or you even do under supervision, people don't lose weight. And Donnelly at University of Kansas has done ton of study like that, and the, uh, physical activity for weight loss is a B-minus at best in term of evidence that it plays a role.

   6. PAPeter Attia29:44

      And let's talk about why, because again, it is counterintuitive. I mean, again, I, I, I know this and I preach this to my patients left, right, and center, which is, you know, you've probably heard me say, Eric, I, I will, I will go to my grave maintaining, I suspect, uh, until new evidence emerges, that, that exercise is the single most important thing we need to do for health. But I'm always quick to follow that up, if asked, with, "But it's not the tool you're going to use to regulate body weight," right? Energy intake, nutrition is the more important tool to regulate body weight. Um, and so W- w- explain to people why perhaps... Like, if you could provide a teleologic explanation for why, because it is counterintuitive, right? If you just look at the human body as a machine, it... You should have no preference for expenditure of energy versus input of energy. That you, you should say they're equally efficacious. But, but again, any clinician will tell you that's not true, and any scientist will tell you that's not true. But do we have a reason for why it's not true?

   7. EREric Ravussin30:47

      Yeah, I, I, I think, uh, you, you know, first of all, if you, uh, try to put calories on your exercise, it's easy on a bicycle. You know the efficiency, 25%. You know the workload and so on, and you can calculate, or you can use your mask and measure oxygen-

   8. PAPeter Attia31:06

      Yes

   9. EREric Ravussin31:06

      ... consumption. It's very easy like that, and I think that there is a compensation. Now, there is probably a higher... Y- you know, exercise, very, uh, vigorous exercise is anorectic. You don't want to eat after you finish a marathon, or, or at least I don't.

   10. PAPeter Attia31:26

       No, and, and by the way, just to interject, I interviewed George Brooks from Berkeley recently, and he shared with me on the podcast something I didn't know before, which was that lactate specifically is probably partially driving the anorexia we experience following intense exercise that might not be present with low-intensity exercise. So I thought that was an interesting way to connect a fact that I think most people appreciate, which is go out and do, you know, interval training. The last thing you want to do for the next hour or so is eat.

   11. EREric Ravussin31:57

       But light exercise has the opposite effect. I think-

   12. PAPeter Attia32:02

       Yeah

   13. EREric Ravussin32:02

       ... it drives a little bit appetite, and I think, uh, Tim Church, who was at Pennington, did a study called eMechanic, and basically they measured every calorie spent on exercise, on a treadmill, on a bicycle ergometer, and all that, and look at the weight of the people with these different doses of exercise, and there was a compensation. Now, was that... What, what it didn't measure, uh, was to use doubly labeled water. Did they compensate by being less active out of their periods of exercise? You know, you, you have-

   14. PAPeter Attia32:42

       Mm-hmm

   15. EREric Ravussin32:42

       ... 90 minutes of exercise, but then, you know, you relax a little bit more rather than fidgeting or moving and so on, or you sleep a little bit longer. Uh, this he didn't measure, but there was this compensation with food intake, and I think, uh, first of all, I didn't say, say that exercise is not good for health.

   16. PAPeter Attia33:04

       Right. No, no, I didn't, I didn't, I didn't suggest you were. I was just making sure people-

   17. EREric Ravussin33:07

       Okay. [laughs]

   18. PAPeter Attia33:08

       ... knew the distinction between-

   19. EREric Ravussin33:08

       Okay. 'Cause I want to make sure-

   20. PAPeter Attia33:09

       Yes

   21. EREric Ravussin33:09

       ... because I'm, I'm an aficionado of exercise-

   22. PAPeter Attia33:12

       I know you are. Yeah, yeah

   23. EREric Ravussin33:13

       ... and I know how good it is for health.

   24. PAPeter Attia33:15

       Yes, yes.

   25. EREric Ravussin33:15

       But it's not as good for weight control in general, and I think that, uh, you, you know, you, you have to do the calculation of how many calories are burned, you know, when you run a K or a mile and so on versus how quickly you can ingest the calories, the, the s- [laughs] the same amount or much more. And I think that there is a difference here that people understand. But exercise and physical activity is the key for weight loss maintenance. And, you know, the, the weight registry, uh, from Rena Wing and Jim Hill, they showed that those who were successful at maintaining the weight loss at five years after the intervention were those who engaged in more physical activity. And I still believe that, you know, a nutritional approach to, uh, weight loss without physical activity or exercise prescribed is, is not a good strategy. We should implement both and try to have people enjoying the exercise. Most people... I remember one of my colleague, Richard Bergmann, said, "I would never exercise. I'd rather be sitting in a cold bathtub-"

   26. PAPeter Attia34:37

       [laughs]

   27. EREric Ravussin34:37

       "... to generate some heat rather than exercise." And there are some people who cannot exercise, but, you know, I, I, I think, again, this is from childhood, that you take the taste of exercise, that you become more physically active in the rest of your life. It's not at the age of 40 because you have too much weight that they can tell you, "Oh, you have to do your 150 minutes per week of moderate to vigorous ex- uh, activity," and all that. It doesn't work in general, but I think physical activity is a key, uh, point in metabolic health in general.

   28. PAPeter Attia35:18

       Well, I agree with that completely. When I look at the, the data from Jim Hill that you've just cited, um, I've always found it difficult to infer the direction of causality, and it... So, so you're, you're correct that the people who are most successful in maintaining weight loss are going to be, on balance, far more likely to be people who do a lot of exercise than those who don't. Um, and so of course it could be that doing a lot of exercise is a valuable tool to drive weight maintenance, i.e. maintenance of energy balance. Alternatively, it could be the people who metabolically get fixed through weight loss are the ones that have an easier time exercising. But I will say this, and I'm curious how robust the data are. I do think one thing that exercise can do for you, and I feel this personally, and I know a lot of my patients do as well, is yes, you know, you could go ride your bike for an hour, and you could burn 600 kilocalories, and you can eat 600 kilocalories [laughs] in way less than an hour. But exercise seems to sharpen your appetite ... to a point that makes you a more responsible eater, if that makes sense. In other words, I think it hones your appetitive signals, and the one thing you as the individual have to do is just be mindful of the speed at which you eat so that you give the brain a moment to recognize what is happening. So let me give you an example of how that works tangibly. If I do, like, a zone two bike ride for an hour, again, that's gonna be, for me, I know because I measure it, I can measure it exactly. That's gonna be about 600... Uh, pardon me, 750 to 800 kcal is how much energy I will expend in s-

   29. EREric Ravussin37:05

       Above your resting?

   30. PAPeter Attia37:07

       Uh, no. Total. Total energy-
7. 40:11 – 54:05

   Experiments looking at how macronutrient composition affects energy expenditure

   1. PAPeter Attia40:11

      Yeah, and, and, and again, to make it an even more interesting experiment but now much more complicated, you introduce a fourth variable, which would be the macronutrient composition of the meal itself at isocaloric substitutions. You know, and, and maybe that's something we'll talk about now as we go back to the, the idea of energy balance. So, so, w- you know, going back to how you and I met and, um, the, the kind of work that, that we were interested in, we, we, we were interested in a question, a very specific question. And I like specific questions, uh, because, you know, the, the more, the more specific, uh, the question, the easier it is to design an experiment to test it, and the more likely you are to get an answer that you can interpret as highly probable. And, uh, feel free to modify this, Eric, but I think the question we were trying to ask was, under isocaloric conditions, does varying the macronutrient composition, specifically between carbohydrate and fat while holding protein constant, have any bearing on non-deliberate energy expenditure? S- would you... First of all, would you agree with that statement? And then we can turn it into English. But just scientifically-

   2. EREric Ravussin41:31

      [laughs]

   3. PAPeter Attia41:31

      ... would you agree that that was the question we, we were trying to answer?

   4. EREric Ravussin41:33

      Yes, it was the question.

   5. PAPeter Attia41:35

      Okay.

   6. EREric Ravussin41:35

      But the, the u- isocaloric is the important word.

   7. PAPeter Attia41:39

      That's right. That's right.

   8. EREric Ravussin41:41

      You know, because now can you do that in the everyday life under, you know-

   9. PAPeter Attia41:48

      That's right. Yeah

   10. EREric Ravussin41:48

       ... not isocaloric condition, but what you want to eat?

   11. PAPeter Attia41:52

       Yep. And w- which of course, we'll, we'll, we'll talk about that in spades. But this was a very important question, although truthfully, I think most people felt the answer was known, right? I think most people... Well, I guess I should translate what I just said into English. W- w- w- [laughs] feel free to interject and help me. Uh, what, what we basically said was if you give people, a bunch of people, meals... Let's just not even use a bunch of people. One individual. You take one individual and you feed them meals on subsequent days or weeks or whatever, where you don't change the total number of calories one iota. You don't even change the amount of protein. You simply manipulate the ratio of carbohydrate and fat Such that obviously when one goes up, the other has to go down, such that you can preserve the total number of calories. And when you feed people meals under those conditions, the null hypothesis should be that their energy expenditure doesn't change. Why would holding a fixed number of calories in to this system change the number of calories that the system expends? Again, it's important that we held protein constant because we know about the thermogenic effect of protein, so that can actually sway things if you change protein a lot. Um, but the alternative hypothesis that was being tested was no, actually, if you really swing fat and, uh, carbohydrate a lot, you could indeed change energy expenditure. Would you, would you add anything to the English version of the question?

   12. EREric Ravussin43:25

       No, I think, I think it was a question.

   13. PAPeter Attia43:27

       Yeah.

   14. EREric Ravussin43:27

       And we were a little bit skeptical. Uh, it was the start of the carbohydrate insulin model. And, uh, you know, we were used to the energy balance model. And, uh, but I think it was a good question, and I like the way we kind of debated that, uh, with the red team, the blue team, if I [chuckles] recall correctly. And we were arguing about the best design. And at the end, you know, now if I had to redesign this study, I, I would do it differently, but, uh-

   15. PAPeter Attia44:00

       Well, let... Tell, tell people how it was done, and then tell people what you would do different, because this is probably an experiment that, that might be worth redoing under improved conditions. But I'm ver- I'm very curious 'cause I haven't given this much thought, but I'm very curious as to, uh, what you would do different. But let... Explain to people how this experiment was done.

   16. EREric Ravussin44:20

       Uh, the experiment was done to do a high, isocaloric, uh, intake over four weeks, I think. We had the baseline diet, which was the SAD diet, the standard American diet, and we had after a low carbohydrate diet or ketogenic diet for four weeks. And the hypothesis was that, and this was led by the people who have been bringing up the, uh, carbohydrate insulin model, uh, saying that you create a basically uptake of the substrate from the blood into the storage, mostly the adipose tissue and maybe some of the, in the liver, and you kind of deplete from, uh, substrate or energy substrate your circulation, and it puts you in a state of semi-starvation. What is happening when it goes to the brain when you have semi-starvation, you shut down your energy expenditure, and you increase your appetite. And this was, uh, playing with that, saying that now if we switch to a diet which is gonna be less conducive to storage, i.e. less insulin secretion, because it was less, less than 10% carbohydrate, between 5 and 10%, if I recall correctly, uh, we would cause this, uh, you know, basically this semi-starvation condition in the systemic circulation, and this would basically increase your energy expenditure. And I think that David Ludwig at the time had some data already showing by doubly labeled water that there was an increase. But we were arguing that doubly labeled water is not precise enough. It goes back to our discussion about indirect calorimetry. And we said, "Why don't we do it in the confinement of a metabolic chamber?" And this is what we did. And to my surprise, there was an increase, at least early, 'cause we had couples of... I, I didn't go back to the papers, but we had measurements, uh, you know, every week for two days, if I recall correctly, for the four weeks. And, uh, the first week and maybe the second week, there was an increase which was significantly higher in sleeping metabolic rate as well as 24-hour energy expenditure under the ketogenic diet. And boom, it, it was, yes, there is an increase. And, uh, now was it a significant physiological increase? I mean, we can argue about that. You know, when, when we talk about metabolic adaptation, because people are talking about that now, uh, we always say we need at least 150 calories. In the study, we achieved statistical significance, but it was just above 100 calories per day. Uh, but you know, it, it was there, and I was surprised myself of that. Now, we didn't measure the appetite of the people. I mean, you know, were they less hungry under the ketogenic diet? Now, it brings us to all the weight loss studies which have been done either with a low-fat diet or low-carbohydrate diet. If you do a meta-analysis, it seems that there is a slight advantage to the ketogenic diet for the weight loss period. And, uh, you know, I mean, basically I was surprised, and I was, as a Swiss, you know, I was kind of a little bit more neutral sometime [chuckles] rather than the two camps, the red and the blue. I, I, I was surprised that there was this effect, but it disappeared at the fourth week, uh, or the third and the fourth week. But anyway-

   17. PAPeter Attia48:22

       Now, Eric, if it... So, so again, just to make sure people understand that finding. So the, the finding, and I'm glad you remembered it 'cause I forgot, I, I forgot the numbers. I didn't think it was 100. For some reason, I thought it was 90. But again, it, who cares at that point?

   18. EREric Ravussin48:36

       Yeah, it was-

   19. PAPeter Attia48:36

       It, it-

   20. EREric Ravussin48:36

       It depends. We had-

   21. PAPeter Attia48:37

       Yeah, yeah

   22. EREric Ravussin48:37

       ... sleeping metabolic rate or 24 hours.

   23. PAPeter Attia48:39

       That's right. That's right.

   24. EREric Ravussin48:39

       That's why.

   25. PAPeter Attia48:40

       Yeah, yeah. That's, that's right. Now, so, um- What the study basically found, and the study was, again, it was a very well-designed study in that, you know, there were maybe 16 subjects, which is obviously sounds like a very small number, and it is. But given the complexity and cost of keeping 16 subjects housed in a hospital for a month, and then also putting them in metabolic chambers every few days, uh, y- you know, this isn't a stu- study that you're gonna do with, you know, 500 patients. Um, but to increase the statistical power of the study, it was a crossover study, which meant that every patient was their own control, and that's what allows you to get away with having so few subjects. Um, so interestingly, I think it's safe to say that the effect size was large enough that it suggested a signal, um, but not so large that it dispositive answered the question. And in many ways, it, it served as, uh, fodder for more questions, uh, which I guess is not uncommon in research, right? It's very rare that one experiment gives you the definitive answer, and rather it just points you in a slightly different direction and gives you more questions. Which then begs the, the question, uh, which is, you know, if, if, if, if resources were not a limit, what would be the follow-up experiment to that?

   26. EREric Ravussin50:03

       Yeah. I, I wish I had an answer. I mean, I, I'm, I'm tortured between very well-controlled studies like we do, like Kevin Hall does, where you basically domicile the people, you feed them, uh, whatever you want. You know exactly what you feed them, and you look at outcomes. I mean, I like that. But then how do you transpose that in real life? And one of my mentor was JP Flatt, and JP Flatt says, "You know, obesity scientists, they have a tendency to either look at the expenditure side of the equation or the energy intake, but they never put the two together." And he, he was right because [laughs] you know, a, a lot of studies you have an outcome which is on this side or on this side, and you maintain the other one. And I think that, uh, to me, uh, designing a study, uh, would be, first of all, in free-living condition, but it would need a lot of people, uh, because you know you can prescribe all what you want to people. They're gonna do what- whatever they want at the end of the day. Some of them, if you screen them very carefully, are gonna be much more compliant, adherent to the instruction. But this is what needs to be done. And by changing here, we change not the calories, just the composition of the two diet, 10% carbohydrate versus 45% or 50% carbohydrate in the o- other diet. And that's all what we did, and I think we concentrated only on the energy expenditure. But we even didn't ask very much, the visual analog scale, if they were more hungry or all these kind of things. But we kind of ignored the food intake that we were clamping.

   27. PAPeter Attia52:01

       Yeah.

   28. EREric Ravussin52:01

       And this is not real life. You, you, you know, if you do something and, you know, you, you talk about engaging on physical activity regimen or, or, you know, exercise and all that, I mean, you have to look at the impact on the other side, and that's, this is the same question that I have. Now, I'm not helping you in designing the perfect study [laughs] by saying that, but on the other hand, the, again, we have good tools to measure energy expenditure. We have reasonably good tools to measure where do the calories come from, but we have no tools to measure energy intake. But it's gonna come. I bet you that, you know, within... I may not be here, but [laughs] within a f- a couple of decades, we'll have a collar here which is gonna measure your calories coming from fat, carbohydrate, or protein.

   29. PAPeter Attia53:00

       Yeah. And I think, I think it should be sooner than that, Eric. I mean, I really do think as image recognition, uh, gets better and better with AI, to me, I, I would actually hope that within a decade, if not less, we are at the point where if you can weigh something and take a photograph of it, you know, we should have enough training data that you should be able to know exactly what is in it. Now, that doesn't account for how much of that thing you eat. But assuming you, you have something that you weigh and you can photograph and you say, "I ate all of it," or, "I ate half of it," that should... We should be able to do better than 10%. We should be able to do within 5% what is, what is the, um, uh, what is the caloric, uh, density of that food and the, the macronutrient breakdown. Again, when I... When... And I, and I wouldn't have said that five years ago. Like, five years ago I would say that's impossible. But given what I'm seeing with image recognition in AI, um, that to me has to be the future for nutrition research, um,

8. 54:05 – 1:13:27

   The challenges of studying diet in real-life settings, the potential of personalized nutrition, and how public health policy could play a role in guiding nutritional habits

   1. PAPeter Attia54:05

      in, in a free-living environment.

   2. EREric Ravussin54:07

      I think you put your finger on the, uh, the exact point. And now I'm the PI of one of the six clinical sites for nutrition for Precision Health.

   3. PAPeter Attia54:17

      Mm.

   4. EREric Ravussin54:18

      And y- you know, this is basically a, uh, ancillary study of All of Us. All of Us is a million Americans, uh, who are basically providing biosamples, access to their, uh, health electronic records and all that. And then this subsi- this sub-study is really to look at the intersection between their health and their nutrition. And, uh, there is three modules. One is on, uh, 10,000 people, and one of the way to measure is exactly this little camera sitting in your glass- on your glasses.

   5. PAPeter Attia55:00

      Mm.

   6. EREric Ravussin55:00

      And also a system which is measuring if you are chewing or not.

   7. PAPeter Attia55:05

      [laughs]

   8. EREric Ravussin55:05

      You know, it, it's, it's not enough to see the food and going, but do they... Is it chewed? And I, I, I agree with you. I mean, I was not thinking about that. I was thinking about something much more li- like a CGM, you know?

   9. PAPeter Attia55:21

      Mm-hmm.

   10. EREric Ravussin55:22

       Who was dreaming of CGM 30 years ago when I was working with the Pima Indians? We were not thinking about that. Uh, now you have CGM and, uh, you, you, you can measure your, probably your insulin from contact lenses and things like that. And I think these progress are gonna help us. Now, are we gonna be smarter at designing the study? I'm not so sure, but we'll have the tools to be a little bit more real life rather than incarceration in a metabolic ward.

   11. PAPeter Attia55:57

       Yeah. I mean, I, I think that, um... And, and again, it's, you know, you're, you're so much more thoughtful on this, Eric, 'cause it is your world. It's not my world, but I occasionally will think about it. Um, but as I'm sitting here now reflecting on it with you, I... My, my intuition is that the questions are complicated enough that the difference between efficacy and effectiveness have to be separated in studies. And I, I think if this question of macronutrient composition, isocaloric macronutrient manipulation impacting energy expenditure is to be put to rest, it, it c- it can only answer that question. It cannot attempt to answer the impact of appetite, uh, as a movable variable. It can do it as a swing variable, meaning, as you said, you can force people to eat a clamped amount of food and then measure subjectively, or even using, you know, PPY and ghrelin and, and other hormones, um, so somewhat objectively, somewhat subjectively measure appetitive response to that. But I think the, I think the biggest mistake of that study was actually not creating a big enough divide in the macronutrients. I think as an efficacy study that was testing a theory, just a theory, with... Again, you start with a very theoretical response, uh, and then you build from there to say, "Okay, is this theory applicable?" In retrospect, I think it should have been more extreme in the carbohydrate and fat differences. I think one should have been, if the carbohydrate insulin model was being tested, one should have been a very, very, very high insulin diet and a very, very low insulin diet that were isocaloric.

   12. EREric Ravussin57:45

       I'm with you-

   13. PAPeter Attia57:46

       Yeah

   14. EREric Ravussin57:46

       ... to take the extremes to-

   15. PAPeter Attia57:48

       To really see the system

   16. EREric Ravussin57:49

       ... see the effects of.

   17. PAPeter Attia57:50

       Yeah, yeah, yeah.

   18. EREric Ravussin57:50

       Yeah.

   19. PAPeter Attia57:50

       And by the way, let's assume you do that experiment and you get a difference of 250 kcal a day, well, you still don't know if a ketogenic diet is a better diet in a free living environment because you're probably not comparing it to somebody who's eating 80% carbohydrates. So then you still have to do the next experiment, which is maybe the one we did, and then ultimately you have to be able to do the free living experiment where people make their own choices based on appetite. Um, but that's, you know, that has to be sequenced, I think, to answer the questions. Um, uh, so anyway-

   20. EREric Ravussin58:32

       Yep

   21. PAPeter Attia58:32

       ... these are difficult-

   22. EREric Ravussin58:33

       I, I agree with you

   23. PAPeter Attia58:33

       ... these are difficult and costly experiments to do.

   24. EREric Ravussin58:36

       But al- also one thing that y- you know, I still have in my mind all the studies which came from Europe about modulating the composition of the diet and look at the impact on the matching of oxidation to the intake.

   25. PAPeter Attia58:57

       Mm.

   26. EREric Ravussin58:58

       And y- you know, even Steve Smith that we, we were working with did this study called ADAPT. He was isocaloric all across, but all of a sudden you continue with more fat. The FQ of the diet, the food quotient-

   27. PAPeter Attia59:14

       Goes down

   28. EREric Ravussin59:15

       ... was decreased.

   29. PAPeter Attia59:16

       Yeah.

   30. EREric Ravussin59:16

       And then it takes days to basically have a matching of your RQ to the FQ, which means you oxidize what you eat. Whereas if you do the contrary, you increase carbohydrate, it takes one day. And that's why I still believe, and back in Switzerland we did this study where we were giving extra fat as LCT, long chain, uh, triglyceride, or MCT, because the MCTs are oxidized quicker.
9. 1:13:27 – 1:21:30

   The importance of protein in the diet, the limitations of dietary data collection, and how AI could potentially transform nutrition science

   1. EREric Ravussin1:13:27

      factor.

   2. PAPeter Attia1:13:27

      Tell, tell folks what that is. I, I know that, uh, I've talked about it a little bit on the podcast, but let's, let's remind people what that theory is. I, I actually think... I don't think there's any one theory. I think it's a, an amalgamation of theories, but I actually think there's some validity to this theory. What-- T- tell folks how that works.

   3. EREric Ravussin1:13:44

      I mean, in simple words is that we eat for a given amount of protein, which is proportional to our body size and all that. And, and by the way, I, I remember when I was studying, we were saying, you know, during a low-calorie diet, you need 1.2 gram of protein per kilogram of body weight or these kind of things. But they, uh, ex- the, the experiment started with insects, and then they went in rodents, and they manipulated the content of protein of the diet. And they found out that basically the intake was all to gravitate around a protein content which was sufficient for the weight of the animal and so on. And this is called the protein leverage, uh, theory. And, uh, you know, now I'm a little bit more skeptical when, you know, uh, Simpson and, uh, Rob, um, the, the two guys in Australia are saying, you know, the, the pandemic of obesity has been paralleled by a decrease in the protein content of the diet, and this is what has triggered the increased caloric intake to get the same amount of protein.

   4. PAPeter Attia1:15:04

      Yeah, it's basically they're arguing that it's, it's not that total protein has gone down, it's that protein density within food, and especially processed food, has gone down, and so people are seeking more calories to subconsciously-

   5. EREric Ravussin1:15:17

      For the same amount of protein

   6. PAPeter Attia1:15:17

      ... get more protein. Yeah.

   7. EREric Ravussin1:15:18

      Yep.

   8. PAPeter Attia1:15:18

      And, and the data, I mean, I reviewed these data in huge detail for a previous podcast. Um, the data in the rodents is staggeringly compelling. Um-

   9. EREric Ravussin1:15:28

      Yeah, it is

   10. PAPeter Attia1:15:29

       ... uh, it's, it's unambiguous. I mean, it's re- it's-

   11. EREric Ravussin1:15:30

       Yep

   12. PAPeter Attia1:15:31

       ... a, a rodent will consume to its level of protein in the studies where this has been tested, um, even if it means eating way more calories. Um, I do think humans must be more complicated because I don't think the human literature are quite as clear. Do you?

   13. EREric Ravussin1:15:47

       No. O- once again, you know, uh, with rodents, you can feed them whatever you want.

   14. PAPeter Attia1:15:52

       Yeah. [laughs]

   15. EREric Ravussin1:15:53

       With people, you don't. [laughs]

   16. PAPeter Attia1:15:54

       Mm-hmm.

   17. EREric Ravussin1:15:55

       And, uh, you know, I remember Dr. George Bray saying, you know, "These, these dietary recall are not worth the paper on which you write the data."

   18. PAPeter Attia1:16:03

       Yeah. [laughs] I c- I couldn't agree more-

   19. EREric Ravussin1:16:05

       [laughs]

   20. PAPeter Attia1:16:05

       ... with Dr. Bray on that.

   21. EREric Ravussin1:16:06

       [laughs] And, uh, you know, thi- this is where we lack tools to know exactly, you know, and, you know, your glasses with the camera or whatever is gonna be useful, but, uh, I think we need better ways of, uh, measuring what people eat and what is the content of their meal and all these kind of things. And, and that's why it's so easy to do that in rodents because you feed them with this or these three diets or five diets, and you look at w- how much they eat. You weigh the food, and you know the calories. That's why I'm a little bit more pessimistic.

   22. PAPeter Attia1:16:47

       Yeah. Yeah. Y- y- you know, you mentioned that George Bray made this funny comment offhand about the role of food frequency questionnaires, which is they're not worth the paper they're written on. Uh, and of course, John Ioannidis has famously said that, um, the food frequency questionnaire belongs in the wastebasket. I mean, that's basically the only place-

   23. EREric Ravussin1:17:06

       Oh

   24. PAPeter Attia1:17:06

       ... it belongs. And yet the food frequency questionnaire is the backbone, is the scaffolding of nutritional epidemiology. And so do you as an empirical scientist, an experimental scientist, have concern at how much food policy is being driven by nutritional epidemiology rather than experimental science? When we understand why you have to, you know, why, why there's an effort around nutritional epidemiology because these questions are otherwise difficult to answer. But given the fidelity of the data, i.e. the thing you put in the system to calculate isn't worth the paper it's written on according to basically anybody who understands how it works, so how do we, how do we reconcile this problem, right? Which is... I mean, even things that we're talking about now, which is the role of ultra-processed food, well, that's, those are determinations from epidemiology. It's, it's epidemiology that's at least telling us or hinting to us that ultra-processed foods on balance are bad. That confirms what I think most people would intuit, but where do we draw the line between what we are letting nutritional epidemiology tell us from a health policy perspective, uh, to, to where maybe it's overstepping and getting things wrong because of the data integrity problem?

   25. EREric Ravussin1:18:32

       I, I think the problem is really jumping from, uh, y- you know, basically nutritional, uh, epidemiology to policies or labeling or dietary guidelines. And I think to me, now we are at a point that the epidemiology should basically provide us with hypotheses to be tested in better controlled situation and maybe in domiciled with full feeding of people. I know these, these studies are expensive, but we are missing a step, and I think once again, uh, these new studies, uh, from NIH, this consortia of nutrition for precision health, are going towards this direction of basically, you are right, the food frequency questionnaire is here every day for 10 days in this module one of the study. But then there are these other ways. There is the remote photography system that you-

   26. PAPeter Attia1:19:37

       Yeah

   27. EREric Ravussin1:19:38

       ... take a picture with your phone of the plates. You have these cameras and all that. And I think that now we I hope that we're not gonna make policy only or policy or guidelines only based on the nutritional, uh, epidemiology, but also on, uh, diet, uh, on, uh, studies basically testing some hypothesis related to what the nutrition, uh, uh, epidemiology has shown.

   28. PAPeter Attia1:20:07

       Yeah. Again, I, I think all roads, both from a personal health perspective and from a nutrition science perspective, need to point towards AI. And, and I... It's a... You know, it's such a cliche thing to say right now, where everybody-- basically everybody's saying AI is everything. Um, but when people ask me, "How is AI-- how could AI change medicine?" Um, it's not by being a better doctor and being better at diagnosing if you have syphilis or not. You know, sure, that's valuable. Um, a lot of it's in the very unsexy stuff, image recognition and radiology, um, insurance billing reconciliation, um, and this. You know, when you want to talk about biomedical research, the thing that gets all the attention is protein folding, and that is truly magnificent. The protein folding predictions from the amino acid sequence is mind-boggling. Um, and that will absolutely shave some time and money off drug discovery. But if, if AI could solve this, quote-unquote, "simple problem," I say simple in conceptual terms, not technical terms, you change nutrition science. You really start to answer questions that have vexed us for hundreds of years. And, uh, so anyway, I, I hope somebody out there is-- who's got serious AI chops is listening to this and thinking this is an area to pursue.

10. 1:21:30 – 1:35:20

    How Eric’s interest in caloric restriction (CR) began with Biosphere 2, metabolic efficiency's role in aging, and goals of the CALERIE study

    1. PAPeter Attia1:21:30

       Um, let's, let's pivot a little bit, Eric, to talk about o- one of the, the most important parts of your career. Um, and I, I believe you are, if not one of the senior PIs, you might be the single most senior PI on the CALERIE study. Is that, is that correct?

    2. EREric Ravussin1:21:45

       Yeah, I was, uh, yeah, I was one of the four PIs, but I was the one who kind of drove the write-up and, and the design of the study and all that. Uh, CALERIE was an important study. It was funded by the National Institute on Aging, and it was really the first attempt to look at the impact of caloric restriction on biomarkers of aging. Now, don't ask me what are the biomarkers of aging, because there's still a lot of discussion around that. But, you know, I mean, if I tell you it's your fasting insulin going up with life, it's your VO2 max going down. Uh, it's not your gray hair or your lack of hair or these kind of things. But anyway, you have some more sophisticated protein glycation and production of isoprostane and all that.

    3. PAPeter Attia1:22:42

       And how much, uh-- I want to talk about the details of the study in a moment, but just before I forget, how much banked serum do you still have that is available for analysis in five years or 10 years when better and better tools or biomarkers become available? Is that, is it all been spoken for and has it all been tested, or are there still some banked samples?

    4. EREric Ravussin1:23:04

       It's great that you ask the question because it's, uh, we are still, uh... Now there is a CALERIE legacy study, which is to follow up these people, but also there is a biorepository of all the plasma samples, muscle biopsies-

    5. PAPeter Attia1:23:22

       Mm-hmm

    6. EREric Ravussin1:23:22

       ... fat biopsies, uh, at Duke, uh, which was the coordinating center. And these samples are available, of course, with a request, uh, to, uh... And the PI of that is Bill Kraus.

    7. PAPeter Attia1:23:39

       Mm-hmm.

    8. EREric Ravussin1:23:40

       You may know the name. Uh, he's a cardiologist, but, uh, he's at, uh, at, uh, Duke. And, uh, y- it's interesting that you ask that because there's some, uh, you know, one of my colleague just published a paper in Science on a, uh, postdoc analysis of adipose tissue in these people before and after caloric restriction. He found a gene of interest. He's an immunologist. Deep Dixit is at, uh, Yale. And he really mined these, uh, transcriptomes from adipose tissue and found a gene which is related to the immune function and found that if you knock out this gene in mice, they are resistant to weight gain. And th- this is like a calorie restriction mimetic, and they improve the immune function and all that. And, uh, to, yeah, your question is very appropriate. There are still, uh, samples available. Of course, they become less and less, uh, available or more and more difficult. I'm just, uh, digging in some of the samples that we had to send to somebody at UT S- Southwest, uh, Southwestern, uh, because he has a new molecule that he would like to test before and after weight loss in non-obese people. But yeah, uh, the goldmine of these studies is really to be able to bank bio samples.

    9. PAPeter Attia1:25:16

       Yeah.

    10. EREric Ravussin1:25:17

        You know, we all bank the data [laughs] and the results, but the bio samples-

    11. PAPeter Attia1:25:22

        Yeah, that's, that's the treasure trove

    12. EREric Ravussin1:25:24

        ... it's very, very important.

    13. PAPeter Attia1:25:25

        So, so tell folks about the study. So how many subjects? What was the intervention? How were they monitored and tracked? How long were they followed? Uh, let's just start with the basics.

    14. EREric Ravussin1:25:35

        Yeah. The CALERIE-- First of all, I became interested in caloric restriction, uh, because of, um... I don't know if you remember, Peter, Biosphere 2.

    15. PAPeter Attia1:25:46

        I sure do.

    16. EREric Ravussin1:25:47

        It, it was a, uh, glass and steel structure southeast of Phoenix, in between Phoenix and Tucson. And, uh, eight people went into this, uh, biosphere. They had seven different biomes. Uh, there was desert, marsh, ocean-

    17. PAPeter Attia1:26:09

        How big was it?

    18. EREric Ravussin1:26:10

        ... agriculture, habitat. It was about three, three acres.

    19. PAPeter Attia1:26:15

        Okay.

    20. EREric Ravussin1:26:15

        And, and they, uh, this was a rich Texan, uh, donor who wanted to do that, uh, basically for the sake of NASA to know how people can live in autarchy, autarchy. Anyway, uh, eight people entered this, uh, Biosphere 2. It's called 2 because Biosphere 1 was the Earth, and they decided to call it 2 [laughs] . And, uh, among these eight people, there was a, uh, faculty from UCLA, Roy Walford-

    21. PAPeter Attia1:26:51

        Mm-hmm

    22. EREric Ravussin1:26:51

        ... who wrote the textbook with Rick Weindruch on caloric restriction. And while they were in the biosphere and I was in Phoenix, Roy called me and said, "We would like to do measurements of energy expenditure. Can you, can we sneak in a delta track or metabolic cart to measure our energy expenditure?" And I said, "Oh, absolutely, and we can also measure your free-living energy expenditure using doubly labeled water." And we did that, and, uh, anyway-

    23. PAPeter Attia1:27:24

        Remind me how long they stayed in the Biosphere 2.

    24. EREric Ravussin1:27:27

        Two years.

    25. PAPeter Attia1:27:28

        Yeah. God Lord.

    26. EREric Ravussin1:27:28

        Two years. In, from 1991 to 1993. And, uh, Roy Walford was a physician of the group, but he was very, very we- interested in caloric restriction. What he didn't know, things went south. The agriculture, you know, they had pests they couldn't control, some of the insect. They had, uh, goats dying. And very quickly they didn't have enough food, and they lost on average 15% of their weight. None of them except one was maybe a BMI of 26 or 7. They were all between 20 and 25. And, uh, they became calorie restricted. And you had the guy who was writing the textbook on calorie restriction being the physician in the Biosphere 2. And, uh, we started to collaborate. And then when this RFA came out in the early 2000, a repla- uh, a request for application grant from the NIA, immediately I called Roy and I said, "Roy, you, we have din- done this study of energy expenditure," and they came and they s- they stayed in the chamber, five of them, for one day, uh, right when they came out of the biosphere. "And, uh, we need you as a consultant, and we want to write a grant which is going to be competitive. I'm not, I'm not known in aging research." And he, he accepted.

    27. PAPeter Attia1:29:01

        What... By the way, what did you, what did you find of their energy expenditure when they were in the chambers?

    28. EREric Ravussin1:29:07

        It was low. I mean, very low because of their caloric restriction. They were about... We, we had to compare to a group of 72 people. I mean, it's all published, and I can send you the, some of the papers. Uh, they, they were about 200 calories below what you would expect for their weight and body composition.

    29. PAPeter Attia1:29:28

        Interesting. Even correct- even correcting for their highly-

    30. EREric Ravussin1:29:31

        Yep
11. 1:35:20 – 1:49:16

    The CALERIE study: exploring the real-world impact of caloric restriction

    1. EREric Ravussin1:35:20

       and therefore a higher degree of oxidative stress.

    2. PAPeter Attia1:35:25

       Yeah.

    3. EREric Ravussin1:35:26

       And when we designed the first part of our calorie study, it was a six-month intervention. We had four groups: a, uh, ad libitum group, 25% caloric restriction; another group being 25% energy deficient, but half by caloric restriction, 12 and a half percent; and half by increasing energy expenditure by exercise. And the last group was weight loss 10% with very, a, a low-calorie diet and maintenance of this weight loss. It means you are in calorie restriction compared to that baseline.

    4. PAPeter Attia1:36:06

       So just to be clear, the last group, the target was the weight reduction.

    5. EREric Ravussin1:36:12

       Yeah.

    6. PAPeter Attia1:36:12

       Whereas in the other groups, the target was the energy intake and/or energy expenditure.

    7. EREric Ravussin1:36:17

       Correct.

    8. PAPeter Attia1:36:18

       Now, how did you even do this, Eric? Because in theory, this sounds like a brilliant experiment, right? I mean, that, those are big enough, those are seismic changes, right? You take a 100-kilo person down to 90 kilos. I mean, that's a big loss of weight, right? You know, a, a-

    9. EREric Ravussin1:36:33

       Yeah

    10. PAPeter Attia1:36:33

        ... 190-pound person is gonna be 170 pounds when that experiment is done. That's meaningful. Uh, the caloric restriction of 25% is enormous one way or the other. But now the question is, you're doing this in the real world over, in one case, I think months, ultimately over years. How did you think about tools for compliance to see that you could hit even close to those targets?

    11. EREric Ravussin1:37:01

        I think it was a lot of work with our, uh, psychology group who said, "You know, we have to screen the people. We have to screen them for barriers to the intervention, for adherence," and so on. And there was a lot, a lot of screening of our volunteers.

    12. PAPeter Attia1:37:22

        Do you remember what your, what your sort of, uh, uh, rejection rate was? You know, for every 10 people you screened, how many would pass the, the psychology test?

    13. EREric Ravussin1:37:32

        It, it, it is in the, in the console diagram of the publication.

    14. PAPeter Attia1:37:36

        Mm-hmm.

    15. EREric Ravussin1:37:36

        But it... I, I, I th- I know, if I recall correctly, for the f- two-year study, we screened more than 5,000 people to en- enroll 225.

    16. PAPeter Attia1:37:51

        Wow. Wow. So basically only 5% of people.

    17. EREric Ravussin1:37:57

        There was telephones, web screening-

    18. PAPeter Attia1:37:59

        Yeah, yeah, yeah

    19. EREric Ravussin1:37:59

        ... telephone screening. uh, in-person screening visit. And there were five screening visits, you know, to make sure that they-

    20. PAPeter Attia1:38:07

        Sure

    21. EREric Ravussin1:38:07

        ... show up. If they don't show up for the screening visit, forget them.

    22. PAPeter Attia1:38:11

        Yeah.

    23. EREric Ravussin1:38:11

        Uh, and the first study we published that in JAMA, and we did find that there was metabolic adaptation or you become more efficient. And we had also some indication of less oxidative stress. We did what we call a comet assay. You look at damage of DNA, uh, in nucleated blood cells. Uh, and we measure isoprostane and all that. After that, the three sites were Washington University with John Holloszy, was, uh, Tufts in Boston with Susan Roberts, and us in Baton Rouge. And we, uh, you know, after our... Our study was six months, our preliminary study. Theirs were, uh, one year. But after two years when we analyzed the data and all that, we started to design the study which was gonna be the same for the three sites.

    24. PAPeter Attia1:39:10

        And was the only difference in the short study, Eric, the, uh, duration of the study and the sample size to just confirm the technical doings of the study?

    25. EREric Ravussin1:39:22

        No, they, they were all different.

    26. PAPeter Attia1:39:24

        Hmm.

    27. EREric Ravussin1:39:24

        Uh, you know, we insisted to have these four groups ourself in the preliminary study. Uh, Boston had only, uh, two groups, low glycemic index versus higher glycemic index f- in the diet of the caloric restriction. And Wash U, I don't remember the details, but there was a component of physical activity in their study. It was John Holloszy was a, you know, a, uh, biochemist of exercise and all that.

    28. PAPeter Attia1:39:56

        Mm-hmm.

    29. EREric Ravussin1:39:57

        But after that, we, uh, you know, uh, compared our endpoints and our, uh, basically intervention, and we designed this study. Now, the RFA was specifying that they have to be non-obese. Uh, in our preliminary study, we didn't want to go because of this 10% weight loss. You know, if you start with a BMI of 26 and you have t- you are 70 kilos and you have to s- lose seven kilos, you become thin. And, uh, we anyway, uh, we decided, uh, about the range of BMI, and we decided for the m- the, the study to go from 22 to 27.9 of BMI for admission. Means, you know, normal weight on the upper range.

    30. PAPeter Attia1:40:53

        Yeah. Mor- mostly normal weight, you know, on the plus or minus side of normal weight.
12. 1:49:16 – 1:57:21

    Notable findings from the CALERIE study after two years: sustained weight loss, participant retention, and more

    1. PAPeter Attia1:49:16

       now. Um, okay. So let's, let's talk about the findings at the end of two years. What were the most notable findings, aside from the obvious, which is weight loss?

    2. EREric Ravussin1:49:25

       One amazing was the retention rate.

    3. PAPeter Attia1:49:28

       Yeah.

    4. EREric Ravussin1:49:28

       I thought that it was exceptional that between the three sites, and it was much more difficult in Boston, to be honest, you know, uh, uh, than it was in Baton Rouge or even in St. Louis. But anyway, uh, retention was exceptional. Uh, the camaraderie-

    5. PAPeter Attia1:49:46

       Sorry to keep interrupting you, Eric, but what did the subjects tell you? Like, you got to know these people really well. Why did they stay in the subject? Uh, why did they stay in this study? I mean, it can't be pleasant to be that calorically restricted for, uh, two years. Or did they say, "Actually, you know, guys, it really felt lousy for six months, and then it didn't feel lousy at all, and I really became, you know, I, I came to really enjoy this"? Like, I, I'm very curious as to what their subjective experience was.

    6. EREric Ravussin1:50:14

       It, it's interesting that you are asking because we, we have been interested in that, and we are even now following up these people we studied from 2005 to two thou- 2008, and they are part of this legacy study-

    7. PAPeter Attia1:50:30

       Yeah

    8. EREric Ravussin1:50:30

       ... CALERIE legacy study. Uh, but it, it was really a kind of building a team between the investigators and the study coordinators and, you know, the, the s- uh, study managers and all that, and the people. It was also providing them with some of the results, and we had to fight for that because, uh, y- you know, normally you don't want to have people having their results. But we said, "You know, it's not gonna change their behavior-"

    9. PAPeter Attia1:51:01

       Right. It's not blinded. Yeah

    10. EREric Ravussin1:51:02

        ... if we tell them about their total cholesterol, HDL, L- LDL, and it's not gonna change anything, or their body composition. They love to see this DEXA and, you know, uh, my percent body fat, it went from, uh, you know, 15 to 12 in a guy and so on. And, uh, it was kind of building a... Uh, we had Mardi Gras with the, uh, with the volunteers. We had... Every two months we had a reception and all that. Sending postcards for birthdays and all these kind of things are part of the retention and building a team between the investigators and the study volunteers. And I think, uh, we succeeded in that. But again, this was a screening which was es- essential. Because you are right. I, I remember the first person that we randomized, she ended up in the ad lib group. She came to my office and said, "You know, I want to do this study because I have read, uh, about what it does to your metabolic health, to your cholesterol, to your body composition, and I'm in the control group." She was crying and s- uh, "But we're gonna give you data, and after we're gonna offer you maybe a weight loss or help you for caloric restriction and all these kind of things." Of course, two years later, we, we still see them [laughs] but not on a regular basis like-

    11. PAPeter Attia1:52:39

        Hmm

    12. EREric Ravussin1:52:39

        ... during the study. But, uh, it's a lot of investment from the invest- the investigative group and also for the study participants because they had so many visits.

    13. PAPeter Attia1:52:52

        Yeah. So, um, talk to me about the lipid changes, the markers of insulin sensitivity, the body composition changes. Let's just kind of run through the list of, of, of what, what improved. Again, I, I, I, I, I don't think it's a surprise that these things improved, but, but the magnitude in some cases is interesting. Um, and again, this question of I'm very curious as to subjective markers of satisfaction in that group because most people would say, "Look, if you told me I have to restrict my calories by 20 or 25%, there's no amount of benefit to my health that could justify the unhappiness and unpleasantness of that experience." So wh- where-

    14. EREric Ravussin1:53:37

        Yeah

    15. PAPeter Attia1:53:37

        ... where were they psychologically as, as that, as that trial progressed into its second year?

    16. EREric Ravussin1:53:43

        I, I think first of all, uh, there was quite a lot of press around the study at the time done at a national level as well as local levels. And they like to, to know and to know that they are part of an important study. It's like this Nutrition for Precision Health now or MoDePaC are, are things which get some press and all that, and they... Most of the people like to be part of a kind of key study. Uh, the, the second thing was really building this relationship between the investigators, uh, and the study volunteers, and also the return of results and the sharing of experience with other people. Uh, you were asking, you know, how long does it last to have this problem with hunger and, you know, suffering and all that, and it seems that for most people, after one month it was totally, uh, manageable.

    17. PAPeter Attia1:54:45

        Hmm.

    18. EREric Ravussin1:54:45

        We taught them also to increase the volume and decrease the fat content to satisfy, you know, on a volumetric basis how much food they were eating with less fat and less calorie density in the diet against the ketogenic diet [laughs] here.

    19. PAPeter Attia1:55:03

        And do you have a sense of how their macronutrients shifted? So did the ad lib people effectively mirror the standard American diet of, you know, 50 to 55% carbohydrate? And did the CR group end up at, you know, a higher fraction of carbohydrates because of this attempt to lower caloric density?

    20. EREric Ravussin1:55:25

        I... We, we have published. I, I, I don't remember exactly if there was a clear-cut impact of being in the caloric-

    21. PAPeter Attia1:55:35

        Hmm

    22. EREric Ravussin1:55:35

        ... caloric restriction group versus the ad lib group. Uh, this I don't know, but I know that quite a few people wanted to make a difference in their diet, and they wanted to know more about the MED diet, the DASH diet, the low-fat diet and all that. And, uh, surprisingly, uh, there are some people, and like I said in this legacy study, and Dr. Redmond is the PI now, uh, people are still, uh, you know, a lot of them, uh, did not regain all the weight, and now it's more than 15 years after. And, uh, they are still kind of using what they have learned during these two years because there was a lot of learning

    23. PAPeter Attia1:56:25

        Do you... When was the last time you took blood, uh, or any, gathered any sort of data on those people besides weight? Like in other words, is there any way to infer whether two years of caloric restriction 20 years ago had any lasting difference relative to the controls?

    24. EREric Ravussin1:56:42

        Like, like I said, this legacy study is, is recruiting, uh, targeting all the people. Now, I don't know how many-

    25. PAPeter Attia1:56:49

        Oh, we, but we don't have the results yet

    26. EREric Ravussin1:56:50

        ... but I'm pretty sure we're gonna get.

    27. PAPeter Attia1:56:52

        Yeah.

    28. EREric Ravussin1:56:53

        No.

    29. PAPeter Attia1:56:53

        Okay. Got it.

    30. EREric Ravussin1:56:54

        No, it's ongoing now.
13. 1:57:21 – 2:07:07

    The effect of caloric restriction on the hallmarks of aging

    1. PAPeter Attia1:57:21

       So what do you think from a, from an anti-aging perspective were the most important, uh, biomarkers that changed? Like for example, did you do OGTTs or euglycemic clamps? I mean, how much did you, uh, uh, scrutinize, uh, glucose disposal and insulin sensitivity in these people?

    2. EREric Ravussin1:57:43

       Yeah. We, we didn't do clamps. Uh, let, let me tell you, there are two kind of aging. There's primary aging, which is more like senescence-

    3. PAPeter Attia1:57:53

       Mm

    4. EREric Ravussin1:57:53

       ... or mitochondrial dysfunction or the, uh, you, you know, leaky membrane. And there is secondary aging, which is basically the impact of your environment and your lifestyle. When you ask about insulin sensitivity or cardiovascular factor, it's more secondary aging. And here we did a lot of measures, and, uh, Bill Kraus published a, a quite cited paper, uh, like five years after the end of the study in JAMA on all the cardiovascular, cardiometabolic risk factors. And everything was tremendously improved despite the fact that these people were healthy to begin with. BMI 22 to 27.9 and so on, and remarkable improvement in all these markers of secondary aging. Now, when it comes to primary aging, that's a different story because, you know, it's much more difficult to measure autophagy, to measure mitochondrial function, to measure leakiness of a cell membrane in these people. And what we ended up to measure was really, uh, some of these, uh, biomark- or y- you know, the hallmarks of aging. You have probably seen these papers.

    5. PAPeter Attia1:59:17

       Of course. Yeah.

    6. EREric Ravussin1:59:17

       There was one in 2012 and one 10 years later. Uh, you know, we, we don't have as many data that we would like, but for example, we did measure mitochondrial biogenesis by looking at the relationship between nuclear, uh, um, markers and mitochondrial DNA. And we, uh, observed in humans that there was an increased mitochondrial biogenesis. Despite the fact that you use less energy, you become more efficient. You impro- you increase your biogenesis of, of mitochondria, which is quite spectacular because the ROS production or these reacts- reactive oxygen species are produced by older mitochondria. The new mitochondria are much more efficient. They, they, they don't produce as many ROS.

    7. PAPeter Attia2:00:22

       So let me make sure I understand that, Eric. Sorry. You're saying-

    8. EREric Ravussin2:00:25

       Yep

    9. PAPeter Attia2:00:25

       ... that the people in the CR group had a greater turnover of their mitochondria.

    10. EREric Ravussin2:00:31

        Yep.

    11. PAPeter Attia2:00:32

        And it's interesting-

    12. EREric Ravussin2:00:34

        In the muscle, skeletal muscle. It was-

    13. PAPeter Attia2:00:36

        Yeah. And, and, and so we would expect that people who are calorically restricted would have lower ROS generation, but we would expect that the reason for that is lower substrate utilization. You're saying, yeah, they had lower ROS generation, but it might actually be just as much from the fact that they had more mitochondrial biogenesis, and they were using newer machinery for the substrate utilization. So it could be both of those things are reducing ROS. Is that what you're saying?

    14. EREric Ravussin2:01:08

        You are right. Absolutely right. I think it's both. Uh, y- you know, when it comes to mitophagy, you know, these old mitochondria, it's been shown in insect, in rodents that this is improved with caloric restriction. Autophagy, of course, all these organelles in the cells.

    15. PAPeter Attia2:01:28

        Did you measure any of the tubules or anything for autophagy?

    16. EREric Ravussin2:01:31

        No.

    17. PAPeter Attia2:01:32

        No LC2.

    18. EREric Ravussin2:01:32

        No.

    19. PAPeter Attia2:01:32

        Okay.

    20. EREric Ravussin2:01:34

        But we measured mitochondrial turnover by, by this method, and we found that it was increased with, uh, with caloric restriction. And I think you, you are right on the, on the point here. There's both mechanism, less energy requirement, but also more efficient mitochondria.

    21. PAPeter Attia2:01:55

        Today, if you were to do this study, for example... I mean, I know the legacy study isn't the same because not as many of the people are going to still be carrying out. But if you could go back and look at the banked samples, are there any other biomarkers you would be able to look at today vis-à-vis the hallmarks of aging? Is there anything we could be doing to better understand senescence? Um, h- what would you look at in terms of inflammatory markers? I would certainly put that down as primary aging. How much of a, a, a suite of the interleukins, uh, TNF, and things of that nature did you look at?

    22. EREric Ravussin2:02:28

        Yeah, I mean, we did. We did it at, uh, six, 12, and 24 months. We didn't do 18 months for that, but we had the whole panel of, uh, inflammation and high sensitivity CRP, and then all the interleukin, TNF alpha and all that and-

    23. PAPeter Attia2:02:46

        Everything improved or just some things?

    24. EREric Ravussin2:02:50

        No, everything improved. One thing which was spectacular when it comes to immune function, we did some imaging of the thymus.

    25. PAPeter Attia2:02:59

        Mm.

    26. EREric Ravussin2:02:59

        And there was loss y- y- of fat.

    27. PAPeter Attia2:03:03

        Involution. Oh, oh, involution. Okay, fat reduction.

    28. EREric Ravussin2:03:06

        Yeah.

    29. PAPeter Attia2:03:07

        Yeah.

    30. EREric Ravussin2:03:07

        Uh, there was a reduction of, uh, fat in the thymus, and our friend, uh, Deep Dixit, was all excited by that, and this is why he, he pursue all the, uh, immune, uh, you know, responses from transcriptomics in different tissues and all that. But, you know, we, we... Chronic inflammation, first of all, it was not abnormal to begin with, but there was-
14. 2:07:07 – 2:15:08

    The challenge of applying CR to the general population, the potential of drugs and exercise to mimic the effects of CR

    1. PAPeter Attia2:07:07

       I mean, to me, Eric, the experiment that I wanna see done is not even an experiment at this point, frankly. It's literally... Uh, I mean, it's, it's, it's a much easier experiment, which is your, your tissue and serum samples provide a very compelling case of you can take healthy people, CR them for a couple of years, and make them better across the board by every objective measure of both primary and secondary aging. The secondary stuff's so obvious it's not even worth talking about, so people that are interested in that can go back and look at the papers and see the lipid profiles and see the insulin sensitivity. But, um, but it's really these primary markers of aging. I didn't-- I, I had forgotten or didn't know about the thymic fat reduction. That's very interesting to me. Um, so now the question is this. Well, you, you were dealing with a very narrow population, right? You, you started with 2,500 people, and through intensive psychological profiling, you winnowed it down to the 5% of people who were gonna be able to most adhere to this. And then on top of that, you provided arguably the most robust support for those people during this journey, the kind of support they couldn't buy outside. Like, you-- Those people, as, you know, regular people in the outside world, couldn't have purchased the level of care and support and encouragement they got for that two-year study. So all of that is not to be a criti- a, a critic of the study. It's to simply say those were the necessary steps to get such exceptional compliance, which is what was necessary to get such a meaningful answer. But if we want to ask the question: how would you help the 99% of people who on their own could not adhere to that? The answer, of course, becomes CR mimetic drugs, right? This is the question that is on everybody's mind, which is, "Eric, you've convinced me that CR, to this extent, will absolutely help me live a longer, better life, but I just can't do it. You know, I'm not in that percent of the population. But will a GLP-1 agonist give me the same benefit? Will rapamycin give me the same benefit? Will acarbose give me the same benefit? Will an SGLT2 inhibitor give me the same benefit? Will metformin give me the same benefit?" And those studies are much easier to do. We can easily put people on those drugs for two years and do the exact same analysis. You've given us the gold standard. You've given us the template. Now we just need to compare every single marker you've measured, plus some you haven't. I really wanna see the epigenome in these subjects pre and post. Um, and so I don't know. I mean, i- is it, is it just me that feels this way, or is that also screaming out to you as the most interesting question to come out of this?

    2. EREric Ravussin2:10:11

       Uh, it's been a, a little bit of a roller coaster, uh, the CR mimetics. You, you, you remember s- uh, s- all the sirtuin stories.

    3. PAPeter Attia2:10:22

       Yes, but, but notice I picked something very carefully. I only picked drugs that, with the exception of GLP-1 agonists, I only picked winners of the ITP.

    4. EREric Ravussin2:10:34

       Yeah.

    5. PAPeter Attia2:10:34

       And sirtuins are garbage. Everybody knows they're garbage. I'm sorry to offend people who work on them now, although I don't think anybody's still working on them. But the sirtuins were a scam story that never materialized, and it never, ever, ever was the mechanism by which CR worked, ever.

    6. EREric Ravussin2:10:52

       Yeah.

    7. PAPeter Attia2:10:52

       I've gone back. I have talked to Matt Kaeberlein, who did the first study.

    8. EREric Ravussin2:10:55

       Yep.

    9. PAPeter Attia2:10:56

       I have gone through every experiment that was done in that lab. There is zero evidence that sirtuin's supposed efficacy, which I think is true in yeast, by the way, was mediated through CR.

    10. EREric Ravussin2:11:10

        Yeah.

    11. PAPeter Attia2:11:10

        And of course, CR a- a- and sirtuins have had no efficacy post yeast. So anyway, just... I'm just do- I'm just being a little pointed in my language to make the point that I'm, I'm being s- I'm being particular about which CR mimetics I discuss.

    12. EREric Ravussin2:11:24

        Yeah, okay. And you picked metformin, and of course, I'm sure you, you know Nir Barzilai and-

    13. PAPeter Attia2:11:30

        And my hyp- my hypothesis, by the way, is that metformin is not a CR mimetic, but I just think we should include it. Like, it's, it's, it's, it's... It, it... There's a reasonable enough probability that all of those molecules I suggested, I think should be tested.

    14. EREric Ravussin2:11:43

        I mean, the GLP-1 are so potent when it comes to weight loss, metabolic health, and all that. I mean, there's not a week without a new paper on GLP-1 doing-

    15. PAPeter Attia2:11:54

        I know. [laughs] I just read a blood pressure-

    16. EREric Ravussin2:11:55

        Please

    17. PAPeter Attia2:11:55

        ... one today on, uh, tirzepatide.

    18. EREric Ravussin2:11:57

        [laughs] And, uh, and, uh, you know, why, why would that be? Because when you calorie restrict yourself, you eat less, and therefore you stimulate less these GI peptides. And why, [laughs] why would... You, you know, I, I, I don't see a relationship.

    19. PAPeter Attia2:12:18

        None of it makes sense. None of it makes sense.

    20. EREric Ravussin2:12:22

        But, you know, if you tell me, I mean, are they true caloric restriction mimetics and, and do they do, uh, decrease oxidative stress? Do they do improve, uh, insulin sensitivity and all that? Among the one you cited, there's a lot of those and, uh, now the, the question is, is how do you go about to test that? And, uh, of course, there is all these failed drugs and things which have been tested for toxicity and safety and all that, that some people want to recycle in, you know, in aging, and, uh, I'm not sure where it's gonna be. To me, the secondary aging, which is the impact of your lifestyle and your environment, is what we should target first. The primary aging, you know, it... I, I, I like to hear, uh, David Clarke talking about autophagy or the people on the East Coast and all that, but what do you do to have more autophagy or more mitophagy and all that? I don't know.

    21. PAPeter Attia2:13:41

        I mean, we know exercise does it, right? I mean, we know that autophagy is not just fueled by caloric restriction, but exercise is an incredibly potent driver-

    22. EREric Ravussin2:13:50

        Yeah

    23. PAPeter Attia2:13:51

        ... of, of that for, for similar mechanisms, right? You could argue that it's basically a substrate utilization problem. It's the input, it's the output, and when you transiently deprive the output, right, by dr- if you drive more output and you transiently create a deficit of energy, I don't think at the cellular level the body is particularly concerned with, "Am I short on what's coming in or am I short because too much is going out?" Um, and, and again, this doesn't have to be an either/or. Like, that, that's the beauty of biology is we can, we can look for accretive solutions. So, um, I, I, I think it's... I, I, I think that, you know, those particular geroprotective agents, Eric, are probably not going to be as impressive on the secondary markers of aging with the exception of the GLP-1 agonist because of the obvious weight loss. I don't think metformin, RAPA, SGLT2s, acarbose are gonna result in weight loss, and even in the, uh, in the ITPs they did not result in weight loss. Remember the hypothesis of David Allison-

    24. EREric Ravussin2:14:51

        Yep

    25. PAPeter Attia2:14:51

        ... who proposed acarbose is this is gonna be a CR mimetic that induces weight loss because of bad absorption. Well, guess what? Those mice weighed the exact same amount. They just lived a heck of a lot longer, suggesting it was actually the glucose metabolism that, that gave the benefit.

15. 2:15:08 – 2:18:51

    Upcoming study comparing caloric restriction to time-restricting eating, and Peter’s takeaways from the discussion

    1. PAPeter Attia2:15:08

       But I don't know. Anyway, I just, I just think, Eric, what you've done is amazing. I think that... And I know it just wasn't you. Of course it wasn't you. I mean, you had an incredible team. I've been to Pennington on many occasions. I, I think it's a, it's a wonderful place, and I, I, I, I think it's dedication. It's just, it's the little stuff that matters, Eric. It's like, I, I know what your staff is like. I remember working with Courtney and the entire team of nutritionists there. I mean, that's the stuff that's not sexy, but that's what's necessary to do this kind of work. And I, I, I hope that through this podcast we're giving people an appreciation for how hard it is to do, uh, nutrition science if you're an experimentalist. Anybody can do epidemiology, and I kinda share the view of George and John and the people who think that it is a very, very limited use. Not of no use. There is a case. But we over-index on that stuff when what we really, in my mind, need to be doing is doubling down on the kind of work you do, um, because that's the only place I think we're going to get causal information, um, and information that we could make people's lives better with. So thank you for what you're doing.

    2. EREric Ravussin2:16:22

       My pleasure, and, uh, stay tuned because you, you mentioned time-restricted eating, and now there is a study which is gonna be implemented in a year or two, a five-year intervention calorie restriction versus time-restricted eating.

    3. PAPeter Attia2:16:39

       What's the time window of restriction gonna be?

    4. EREric Ravussin2:16:42

       I, I, I don't know the details. The protocol is not finalized.

    5. PAPeter Attia2:16:46

       Ah, okay.

    6. EREric Ravussin2:16:46

       It's one of these exercise where, you know, we were part of one of the study-

    7. PAPeter Attia2:16:50

       Yeah, yeah

    8. EREric Ravussin2:16:51

       ... and now you put all the brain together and you... But it's, it's gonna be probably eight hours a day.

    9. PAPeter Attia2:16:58

       All right. My, my, my, my vote, if you, if you give me a vote, Eric, make it more restrictive. I, I just think cl-

    10. EREric Ravussin2:17:04

        Six? We, we, we did the... The first study we did was six.

    11. PAPeter Attia2:17:08

        Yeah.

    12. EREric Ravussin2:17:08

        Six hours.

    13. PAPeter Attia2:17:09

        Yeah, I would go, I would go... If you really wanna see if there's a signal independent of calorie restriction, 'cause I think the goal has to be you're gonna have to match their calories, right? So it's, it's-

    14. EREric Ravussin2:17:19

        Yeah

    15. PAPeter Attia2:17:19

        ... this is the isocaloric difference between CR and time-restricted feeding. Uh, in my experience, eight hours will give you a null answer. I, I would bet, I'd bet a million dollars you'll get a null answer if you compare eight hours. But four to six is interesting, and I'd be very curious.

    16. EREric Ravussin2:17:37

        Yep. I, I mean, it's interesting you bring that up, but our study wa- was the first in humans, uh, and this was Courtney Peterson who did that.

    17. PAPeter Attia2:17:48

        Yeah.

    18. EREric Ravussin2:17:49

        And we had six hours, and people said it would be easier with eight hours and all that. But after, if I recall correctly, after two weeks anyway, they were perfectly fine with eight hours.

    19. PAPeter Attia2:18:02

        With six hours.

    20. EREric Ravussin2:18:02

        It means it's a question... It, uh, with six hours.

    21. PAPeter Attia2:18:05

        Yeah.

    22. EREric Ravussin2:18:05

        Sorry.

    23. PAPeter Attia2:18:06

        Yeah.

    24. EREric Ravussin2:18:06

        Uh, even if at the end, the exit interview, they said it would be easier for eight hours.

    25. PAPeter Attia2:18:11

        Sure.

    26. EREric Ravussin2:18:11

        But I think it was at the start.

    27. PAPeter Attia2:18:13

        Yeah.

    28. EREric Ravussin2:18:13

        And you can maybe ramp it down, you know, from, from eight to six hours.

    29. PAPeter Attia2:18:19

        Yeah, yeah.

    30. EREric Ravussin2:18:20

        Yep.

Episode duration: 2:18:51