Best Tools for Gut Health & Weight Loss | Dr. Chris Thompson | Andrew Huberman Transcript

Andrew Huberman interviews gastroenterologist Dr Chris Thompson on gut health, metabolism, and weight loss

Andrew Huberman speaks with Dr Chris Thompson, professor of medicine at Harvard Medical School and chief of interventional gastroenterology at Mass General Brigham in Boston.

Summary

Andrew Huberman interviews Dr Chris Thompson, a gastroenterologist and obesity medicine specialist who is credited with developing new endoscopic procedures for treating metabolic disease. The conversation covers the full anatomy and function of the GI tract, the gut microbiome's role in metabolic health, and the mechanisms behind hunger and satiety hormones including ghrelin, GLP-1, CCK, and peptide YY. Dr. Thompson explains the limitations and side effects of GLP-1 drugs such as semaglutide, including muscle loss and the dangers of cycling on and off them, and describes how micro-dosing can serve as a maintenance strategy. He also details the sequence of metabolic dysregulation — from glucose spikes to ectopic fat to insulin resistance to metabolic inflexibility — and argues that tools like continuous glucose monitors and fasting insulin tests can detect problems years before standard markers like HbA1c. The episode concludes with a discussion of cutting-edge treatments including duodenal ablation, magnetic anastomosis to boost GLP-1 naturally, and a gene therapy approach that would make GLP-1 production nutrient-responsive and permanent.

Key Takeaways

FULL TRANSCRIPT

Overview of the GI Tract

Andrew Huberman: A lot of us hear these days about the gut microbiome, the gut-brain axis. We hear about GLP drugs that help people lose immense amounts of weight and stop feeling this food noise thing, and on and on. But could we start by just having a conversation about this tube that is the digestive tract — get real basic and just educate people a bit on what happens that stimulates them to want to eat, why perhaps for certain periods of day or night they don't want to eat, and then what the passage of food through us looks like as a series of steps.

Dr Chris Thompson: This is such a critical part of our biology and our lives. It's becoming more and more complex all the time. The gut does a lot of things. It's obviously involved in digestion, but it's also an endocrine organ. You can hear it called the second brain. There are a lot of different ways we think about the gut, and it is compartmentalized — each area has a different job.

First you have the esophagus, and its job is to just move the food into the stomach safely. It's thick and has a different lining so it can handle things that might be a little rougher, and it pushes food sequentially down into the stomach. It's taking that food bolus and driving it into the stomach. You can have all sorts of problems in your esophagus. Each one of these organs has things it's supposed to do and things that it doesn't do well. Sometimes people don't swallow well. It gets too tight at the bottom. There's a condition called achalasia where the bottom of the esophagus doesn't relax. We have procedures we can do in my line of work where you can tunnel down in between the layers of that esophagus — it's very thin, just a few millimeters — and cut that muscle to relieve the obstruction.

Andrew Huberman: What are the symptoms of that?

Dr Chris Thompson: Inability to swallow. They'd feel like they're choking. They'll swallow food, it'll get down and stop, and then they'll feel pressure. They'll feel really uncomfortable. If they drank some fluid with it, it might start coming back up. It'll just stay there, and then sometimes they'll have to induce vomiting to remove it. It's very uncomfortable for them. It's not a terribly common condition, but it's becoming more and more frequent. I see it every week.

That inability to swallow can also occur for other reasons that are far more common. Chronic heartburn — if someone has reflux, that burning sensation can damage the lining of the esophagus and lead to a precancerous condition called Barrett's esophagus, which needs to be treated, looked at, and followed. With time, it can actually cause scarring, so you get a stricture — very fibrotic tissue — and that's another reason why people might have difficulty swallowing. So that's the job of the esophagus: just to move the food down safely. And a lot of times it doesn't work.

Then you have the stomach. First, what the stomach does is stretch to accommodate and accept a meal. Normally it's like a tube in your abdomen, but when you start to smell food, it starts stretching and becoming more like a bag.

Andrew Huberman: Really — just the odor of food?

Dr Chris Thompson: Yeah, it can stretch and relax to accept that meal. And if it doesn't do that properly, it causes symptoms like nausea. So then it accepts the meal and has to do its job, which is to break it down and pass it on. The stomach isn't just transporting — it's breaking food down. It does that mechanically. The fundus, the top of the stomach, is holding that meal. That's what stretched up to hold it. Then the rest of the stomach is working on it. The body of the stomach, the next segment, is breaking it down, grinding the food into smaller bits. Acid is part of this as well. The stomach secretes acid, and then the bottom of the stomach, called the antrum, will push the food out slowly into the duodenum — that's the first part of the small bowel.

Satiety and satiation all become part of this because the stomach is what secretes ghrelin — we'll talk about that probably more later. This is part of your satiety signaling. All sorts of problems with the stomach, right? Similar to the esophagus, food might not leave as it should in the right timing. That can happen due to ulceration in the stomach, scarring, or something called gastroparesis, where for a variety of reasons — it might be post-viral, it might be due to diabetes, neural or hormonal origins — the stomach just doesn't empty as it should. People have nausea and vomiting with that.

So then you get into the small bowel. The small bowel's job is typically to absorb calories. You do a little digestion early on because you have pancreatic and biliary secretions going in there, but its main job is going to be to absorb calories. It's very thin — it's one cell thick — and has about the surface area of a pickleball court.

Andrew Huberman: One cell thick?

Dr Chris Thompson: Yeah, one cell thick. The lining is one cell thick.

Andrew Huberman: That's where the cell type — enterocytes?

Dr Chris Thompson: Yeah. Columnar epithelium. So they're pretty sturdy. They rely on more than just the cell itself to maintain that barrier. There are certain cells called goblet cells that produce mucin, and that creates a nice thick layer there that helps as another part of the barrier. They have something called tight junctions right between the cells, which are complex little structures that are part of that barrier as well. And there are immune cells in there. There are other elements to that barrier, but it is one cell layer thick. That's why the esophagus and stomach do a good job of processing that food so that it's safe to go down through the small bowel and be absorbed.

All sorts of issues with the small bowel. Similarly, you can have different diseases that affect it — celiac disease, Crohn's disease, different inflammatory conditions. And what we're learning now, and hopefully will get into, is that it plays a central role in metabolic disease. That's what's very exciting — its role in obesity, diabetes, and other similar conditions.

And then eventually you have the colon, and that's where your microbiome is the star. The colon's job is to usually just absorb water. Most of the nutrients are gone by then. But it does play an important role, working hand-in-hand with your microbiome. The microbiome is producing short-chain fatty acids, and one of them — probably the most important — is butyrate, which has a lot to say about your metabolism. It's involved in satiety signaling, and you have a lot of GLP-1 produced in the colon as well. So you're getting these endocrine functions of your colon that are very involved.

And again, diseases in the colon — colon cancer is a big one. Colon cancer screening is important. They moved the age back to 45, so everyone should start getting screened. You can do it different ways. There are genetic tests you can do like Cologuard, and if you do that you have to do it every few years. You can do screening colonoscopy every 10 years if it's normal. CT colonography is not as common. Those are the two most common approaches. It's important to do that.

Andrew Huberman: How common is colon cancer?

Dr Chris Thompson: In our line of work, it's the most frequent cause of cancer, unfortunately.

Andrew Huberman: Are more people being diagnosed because of more diagnostic procedures, and are more people surviving colon cancer?

Dr Chris Thompson: The survival rates are definitely improving due to screening programs. So it is definitely important to get screened. Starting at 45 is better than many people who started at 50 and wouldn't get screened until 55 or 60. But also, if you have a family member who has had cancer, you want to start at 40, or if they were younger, you want to start 10 years younger than when they were diagnosed. So you want to start that screening process early. It's very effective.

Things like Cologuard and other genetic tests are going to keep getting better and help, because you don't have to have that uncomfortable screening procedure. Colonoscopy is not a great way to do screening — you shouldn't have to have a relatively invasive procedure to be screened for something. It should be something you just do with a blood test or a stool study. I think we're getting there with technology, and that will definitely show dividends. You can have the colonoscopy to remove the lesion, which is something we can do now. It's a newer technique where we can actually go in and remove these very early cancers endoscopically — we call it organ-sparing surgery. You don't have to actually remove a piece of the colon anymore. You can just take the lining where that precancer is residing. It's a complicated procedure, but it's easy for the patient. They keep their colon, they go home the same day. And these tests are easy ways to diagnose those patients and get them in for proper care.

So that's pretty much the quick overview of the gut. One thing we haven't touched on too much yet is its role in satiety and how it's involved in the processing of food in detail, and the endocrine system that's involved.

Bowel Movements as Health Indicators

Andrew Huberman: There's a weird thing about the GI tract and bowel movements in particular. With babies, with puppies, and to some extent with ourselves — but especially with babies and puppies — because they can't speak, we have a couple of key readouts that we intuitively understand reflect their health. One is the color of their skin and eyes, and the other is the quality or lack thereof of their bowel movements — quality, frequency, etc. But then something happens where speech comes online and we get toilet trained, and then everyone's responsible for understanding their own bowel movements. And we're never really told what healthy bowel movements look like, but we all kind of know what's normal for us or not normal. I'd be lying if I didn't say these are important metrics of health.

Dr Chris Thompson: Yeah. Well, there's so much you can tell from bowel movements. The rule of thumb is you don't want to have more than three a day and you don't want to go longer than three days without having a bowel movement. That's the general rule, and you want it to be one formed bowel movement, or a couple. You don't want little tiny pebbles — that's called scybala stool, and that's a sign something's going on.

There's a lot you can tell about how much fiber someone is taking in. The World Health Organization published something in the Lancet years ago on fiber synthesis, and they found that the vast majority of the population — especially in western countries — just are not getting enough fiber, which is obviously concerning because that causes a lot of issues long term. If you're having scybala stools, if you are constipated — meaning you're having a bowel movement less than once every three days and they're hard stools — that's a sign you're not getting enough fiber. That's one thing you really have to think about.

Additionally, there are other things you can tell. If a bowel movement is very dark, tarry, and shiny, that's a sign you have blood in your GI tract. So there are different things you can tell from the stool that are important to keep an eye on. But in general, that's the rule of thumb — more than three a day is probably leaning towards being too loose, and if you're not having one every three days, you're probably bound up and you really have to think about fiber.

Andrew Huberman: As I recall, the recommendations were 35 grams of fiber per day for adult men and 25 for women, obviously not accounting for variations in body weight and height. Does that sound about right?

Dr Chris Thompson: Yeah, that's about right. And it kind of depends on the quality too. There are a couple of other really interesting studies that came out recently, just within the last few years, looking at the importance of fiber related to certain conditions. One was fatty liver. It was an interesting study — they were using resistant starch, level two, so basically raw powdered potatoes or something like that. They were supplementing it at 40 grams and found that when they did that, they actually saw a significant improvement in fatty liver, which is phenomenal, and it was relatively weight stable. So it has important treatment effects.

Another group studied insulin sensitivity. They did clamp studies where they would really be able to detect insulin resistance and look at glucose utilization and clearance. They found that with this RS2 type resistant starch, they were able to improve insulin resistance and insulin sensitivity as well. So fiber is very important. It's not just about the bowel movements — it's also about really just having health.

The Microbiome, Fiber, and Fermented Foods

Dr Chris Thompson: It probably helps the microbiome. There's all sorts of evidence that if you don't have a lot of fiber, your microbiome is not healthy. You get less diversity in your microbiome. It is important to have that fiber, and constipation is an early window into it — an early sign that maybe you're not getting enough fiber.

Andrew Huberman: Yeah. I make it a point to eat fruits and vegetables because I like them, but recently I started supplementing with a powdered psyllium husk. Some of them actually taste pretty good. My expectation was that I was going to feel really bloated. It was quite the opposite. Not that I had gut issues before — if anything, it was normally kind of hypernormalized — but it actually made the post-meal sensation subjectively feel really good. I didn't expect that. I thought, okay, more fiber. I think the perception people have is like more fiber, more regularity, and more bloat.

Dr Chris Thompson: That might be true for some people, but it certainly wasn't your experience. I think the messaging on fiber to the general public is pretty lousy — meaning people are told to take it, great, they're told all the time, but I think people think, "Oh, if I have a lot of fiber, I'm going to be really gassy, I'm going to be really bloated." But as you point out, it's not just about regularity and speed of digestion. It's about creating a healthy milieu for the gut. I think if more people knew that, they'd probably make a move to consume more fiber.

Totally. It's like feed your microbes, or they're going to eat you. And it's kind of true, because they need to be fed. And what they eat is fiber. That's what you want them eating. And if you're not feeding them fiber, they'll eat your mucous layer. We already talked about how thin that barrier is, and all of a sudden they start eating your mucous layer. They're not producing the butyrate you need. And the butyrate is needed to maintain the tight junctions. So there are layers to this. It's like a snowball effect — if you're not feeding the microbiome and keeping it healthy, you're going to run into all sorts of trouble.

Andrew Huberman: That raises some interesting questions about intermittent fasting. I think very few people are doing long-term fasts of more than a day or so. Most people do some sort of time-restricted feeding or they'll skip breakfast. I'm one of those people most days just by default. I had a colleague and friend at Yale who studied the microbiome, and I said, "Oh, so does fasting improve the gut microbiome?" And he said, "No, actually during the fasting period, your microbiome starts eating up your digestive tract." But then he said the rebound often puts you at a slightly better place afterward. So it's tricky. Should people avoid intermittent fasting if they're having gut issues?

Dr Chris Thompson: I don't see it as being a major issue. I think the benefits of intermittent or time-restricted eating probably outweigh that risk. You need to give your pancreas time to relax. You need to have insulin come down. If you're eating frequently, your insulin levels are always up, and that causes problems. So I do think the benefit of time-restricted eating definitely outweighs that potential risk.

Andrew Huberman: Yeah, great to hear — especially as somebody who just by default doesn't eat breakfast and doesn't get hungry till 11.

Dr Chris Thompson: I skip breakfast as well. But there are studies — because initially they actually used to say, well, you have a cortisol spike in the morning and you're more likely to store the food you take in if you eat in the morning. Turns out maybe that's not so true. It might be better to actually eat earlier and then have your fasting window start in the afternoon. I think doing it is better than not doing it. I still skip breakfast.

Andrew Huberman: The topic of fermented foods — low-sugar fermented foods — as a possible benefit for gut health has come up since Justin Sonnenberg and colleagues published that study. It's a small number of people in that study admittedly, but taking in some low-sugar fermented foods really helped lower inflammation. They didn't look specifically at symptoms of gut irritation, but what are your thoughts on low-sugar fermented foods? We're not talking beer — we're talking kimchi, sauerkraut brine.

Dr Chris Thompson: I think they're important. And they're missing in the western diet, which is an issue. The study you referred to — I think they compared it to fiber, right? With the fermented foods, you reclaimed some diversity in the microbiome, which was great, as well as reduced inflammatory markers, where the fiber didn't seem to do that. And you saw all these benefits in the fiber trials we're talking about with the resistant starches. So it stands to reason that we'd probably see that as well as we do more research into fermented foods.

They're beneficial for a variety of reasons. One, they're prebiotic — you're feeding your microbiome things they want, which is phenomenal, and it's already kind of started, a little partially digested, which is really helpful. And they're also a little bit of a probiotic as well, because you do have some live cultures in there. They usually have bifidobacteria, lactobacillus, or something like that in them, and a variety of other things as well. So it gets the ball rolling. It's sort of like when you're trying to grow something, you want to plant the seeds but also have the fertilizer. This is what fermented foods do for you.

It's all about maintaining this healthy microbiome that can produce things like butyrate, which has a lot of benefits. You can't just take butyrate — it's not going to make it to the colon. It needs to be in the colon to have its effect. What these bacteria do is they cross-feed in a sense. You have that first layer of bacteria that will take the fiber and break it down, creating acetate and lactate and whatnot. That can then be used by other bacteria, which can turn it into butyrate. The butyrate is magical. It will feed your colon cells. Your colon cells live on butyrate. Butyrate is needed for those tight junctions. Butyrate does all sorts of things via GLP-1 pathways and satiety pathways. So it has a lot of different roles.

Additionally, it keeps your bowel acidic. These short-chain fatty acids and acetate and whatnot are great to make sure you are protected from certain pathologic organisms that might want to take root. The aerobic organisms and the other organisms you don't want don't survive as well in an acidic environment. So it's really important to take these fermented foods in addition to fiber.

Andrew Huberman: Do you make it a point to consume them?

Dr Chris Thompson: I do. Yeah. I like kefir — everyone says that, right? I think it's phenomenal. Kimchi is good, sauerkraut. There are all different types. Yogurt. Kombucha. There are different types I think that everyone should be able to find. And it's certainly missing in our diets. So I think it's important to recommend that folks do.

Ulcers, H. Pylori, and Stress

Andrew Huberman: Canker sores and ulcers — my understanding for a long time was they were caused by stress or wounds to the mouth. And then a couple of folks won a Nobel Prize for identifying a soil-based bacterium that causes ulcers. I loved that Nobel Prize year. As a scientist, some people watch the Super Bowl — I'm like, who won the Nobel Prize? It's never surprising who wins, at least for the scientists. It's often surprising who doesn't. But that was a very surprising set of findings. A gut bacterium is causing ulcers. I love the findings, but at the same time I think many millions of people across history would say stress gives people ulcers. So there's something going on there that's more than a soil-based bacterium, right?

Dr Chris Thompson: Yeah, definitely. That's the problem with the way the media covers these findings. Stress can give you ulcers, right? Or am I missing something?

Andrew Huberman: Stress can play a role?

Dr Chris Thompson: Stress can play a role. It is certainly complicated. Barry Marshall was phenomenal in Australia — he found H. pylori could cause gastric ulcers. No one believed him, so he had to consume it himself and then he had gastric ulcers.

Andrew Huberman: I love it when scientists do self-experimentation.

Dr Chris Thompson: That's crazy, right? But that was phenomenal. He proved H. pylori, and we need to treat that. And H. pylori was actually found even in Ötzi the Iceman — I don't know if you remember him. He was this 5,000-year-old Homo sapiens found frozen in the Italian Alps. You could actually get into his stomach and see what's in there. He died stressed. He had H. pylori in his stomach. That thing's been around a long time.

There are other lessons there, like a loss of diversity of the microbiome. With industrialization, we have far fewer species and less genetic diversity in our microbiome. But regarding ulcers — I actually did study this in gastric bypass patients quite a bit. And it was not a bacterium that was causing it. Sometimes it was relative ischemia. Type two diabetes causes microvascular ischemia. Smoking can cause microvascular ischemia. In gastric bypass patients, the distal part of the small bowel that's connected to the stomach doesn't have bicarbonate being secreted from the pancreas in the area. So there's no way of neutralizing acid. If the pouch is too large and makes acid, the jejunum has no natural defense against that. So acid clearly plays a role, and if you're stressed it can produce more acid. Generally there's probably multiple hits. We don't fully understand things, but clearly it's not just an infectious organism. It depends on individual circumstances and susceptibilities, but ulcers can certainly occur short of a bacteria.

Andrew Huberman: So important for people to hear that — just because one thing can cause something, it doesn't mean it's always the case.

GLP-1 Drugs: Benefits, Limitations, and Micro-Dosing

Andrew Huberman: So let's talk about metabolic health, hunger, obesity, weight loss. These are areas squarely in your wheelhouse. Can't have this conversation without talking about the GLPs. Most everyone has heard of these things nowadays. Millions and millions of people — I've heard as many as 20% of people 18 and older have taken or are currently taking a GLP, either semaglutide, tirzepatide, and soon retatrutide will be out to market. What's your thought on these compounds? Are they the perfect solution to weight loss?

Dr Chris Thompson: Well, I'm grateful we have them. Obesity is a serious problem, and all the metabolic issues that come with obesity need to be addressed, and we weren't doing much with it unfortunately until the GLP-1s came around. So GLP-1s are fantastic from that standpoint. They're not perfect — there are limitations — but it's much better to have them than not have them, for sure.

There are issues with adherence, unfortunately. Over a million people a month are coming off GLP-1s. About 30% come off GLP-1s in the first month, and then about 50% by the end of the year. And it's not specific to GLP-1s — you see that with any medicine. You see that with blood pressure medicines, with cholesterol medicine.

Andrew Huberman: What's the primary driving force in the case of GLP-1? Is it the side effects? Is it they don't like having to inject themselves?

Dr Chris Thompson: I think because the numbers are so curiously similar to all the other medicines, maybe there's some underlying thing where people just don't like taking medicines frequently. Sticking yourself is probably for some people something they don't want to do once a week — needle fatigue. That's probably there. I think when you take a medicine orally every day, it gets hard to remember to take it. And then I think there are issues with how you ramp them up to the effective dose and side effects. Nausea is an issue with some of these. Muscle loss is an issue. And then additionally, long-term, you're taking a super-physiologic dose of something and we don't know what the long-term ramifications could be. So even though I believe the benefits outweigh the costs — you're treating obesity, we know obesity is a problem, we don't know GLP-1 to be a problem long term — that does weigh heavy on some people's minds and that might be why they stop as well.

In my practice where we do endoscopic therapies, over 85% of people have already been on a GLP-1 and either they're struggling on it or they've come off.

Andrew Huberman: 85%. Wow. There are a number of people now who are micro-dosing the GLPs and finding that they're getting some benefits without taking the prescribed amount. I'm not recommending people do that. I know people are doing that — initially it was because of cost and pen sharing, but also people feeling like they get the same effect. Do you have any knowledge of whether lower dosing takes people away from side effects?

Dr Chris Thompson: I think it's actually very useful. The approved dosages are kind of just an effect of our regulatory system, as you've alluded to. It's too expensive to do different doses. Plus it takes away personalization. We're all trying to get to precision medicine and personalized medicine, and that's what micro-dosing allows you to do.

The first time I heard about micro-dosing was from one of my patients, who was a physician. He came in and said, "It's too expensive, I don't feel great on it, and I'm doing this thing where I take the pen and inject it into a sterile vial, then use an insulin syringe and take a small amount out and give it to myself." He said, "I'm doing great. I don't feel nauseous. My weight is staying off." He was a physician so he was familiar with the equipment. That was the first time I came across it, and I thought, wow, that's actually a great idea.

A lot of my patients actually do micro-dose these things. Generally, you get up to the point where you want to lose weight, you get to that dose, you're losing weight, and then for maintenance — rather than just stopping it, because if you stop the GLP-1s there are problems; these are not meant to be stopped, they're kind of lifelong medicines — instead of stopping it, just go to micro-dosing. You'll find a spot, hopefully, where you keep the weight off, you feel good, and you're not taking as much of the medication.

Now, the problem with coming off them — especially with the original drugs like semaglutide — is that when you lose weight, about a third of the weight you lose would be lean mass. Mostly muscle, maybe some bone. The problem is when you cycle on and off. Say you come off it and you put your weight back on. You're not putting the lean mass back on — you're putting the fat back on. So now you've shifted your body composition to be less favorable than before you ran the GLP-1. Then you go on it again and you lose weight again and you lose a little more muscle. Then you go off and you put more fat on, not more muscle. So basically, you're taking your body composition and shifting it worse every cycle. There has to be a game plan. If you're coming off the GLP-1, you need either to micro-dose it or have a bridging plan to a procedure or something else that will keep the weight off for you.

Andrew Huberman: Are there any good studies showing that resistance training can offset the muscle loss from a standard or micro-dose of one of these GLP drugs?

Dr Chris Thompson: Yeah, resistance training definitely plays a major role in maintaining muscle, and that's with anything — not just GLP-1 medications. It's with the first-generation medications, with any surgical procedure or endoscopic weight loss procedure. If you're doing resistance training, you tend to maintain your muscle because the body realizes, "Hey, I need this muscle. I'm not going to get rid of it as the person's losing weight." When there's a caloric deficit, the body is looking for what it can do to maintain energy levels, and you don't want it chewing up the muscle to do that.

Andrew Huberman: The side effects that I see getting the most coverage are increased feelings of apathy, general food noise being down, alcohol appetite being down, appetite for life being down. You hear this, but I don't know how accurate that is. Social media is a weird place — certain things get amplified out of proportion to the real data. The other one is that GLPs can cause blindness. Turns out that's in a very rare set of individuals who have this ischemic optic nerve head condition. So yes, the GLPs can make certain people blind, but it's a very small number of people. You want to get screened for this structural thing in the eye, but it's not true that GLPs are making people go blind all over the place. When patients come to you and say, "I didn't like the GLP or it wasn't working for me," are there some resounding themes?

Dr Chris Thompson: There are. I think muscle loss honestly is one of the bigger ones. It might just be subtle — Ozempic face, Ozempic butt — you're losing some muscle in places where it's noticeable. Other people actually truly develop sarcopenia, where you have significant loss of muscle. It's rare, but those are people who aren't really exercising a whole lot when they take it and might have had a predisposition to it to begin with.

So in people I'm concerned about that, it's good to get a DEXA scan beforehand. Make sure you have adequate muscle mass. And if you don't, you really have to think twice about whether you want to do the GLP-1, or if you want another avenue to try to lose the weight, and you definitely have to start hitting the gym. I think that's the most common thing — muscle loss in obvious places or sarcopenia developing. The other one's nausea. A lot of folks do get nausea on the higher doses, and they will not lose weight on the low dose. If they go on the high dose, they feel nauseous. So that's another issue. Some people say it stops working, and that might be because they don't tolerate the higher doses. Those are the primary reasons I hear.

Surgical and Endoscopic Approaches to Weight Loss

Andrew Huberman: Maybe we can move a bit towards some of the surgical procedures. I always like to remind people there are basically two ways you can affect your brain and body — chemical methods and mechanical methods. When I think of quote-unquote stomach stapling, I think of that purely as a mechanical thing. You're making the stomach smaller, making people feel full earlier in the meal. But of course it stands to reason that you're also removing tissue and changing the chemical milieu of the environment. We had stomach stapling for a long time. Why did we need the GLPs?

Dr Chris Thompson: Surgery really started back in the 1950s. University of Minnesota, I think, was the first place they did it. The first procedures were focusing on malabsorption — the idea was to bypass a portion of the small bowel so that you don't absorb your calories. It was called a jejunoileal bypass. But this procedure was awful. People did lose weight, but the problem was they created a long blind limb — no actual food was going through the limb. You connected the jejunum, the early small bowel, to the very bottom of the small bowel, and the rest of the small bowel was still in there but no food was going into it. So you had bacterial overgrowth in there. You had all sorts of problems. The fat that was being malabsorbed was binding calcium, so the oxalate that normally binds calcium got absorbed and then bound calcium in the body in the kidneys. You were having all sorts of renal failure issues. It was a disaster that went on for years because people were desperate, but it was a very bad procedure.

It was replaced by something called gastric bypass, which I think came about in the mid to late 1960s. Mason, I think, was the surgeon who came up with this. His goal was to avoid the problems with the jejunoileal bypass and still get a treatment effect. He thought of this as restriction — when you eat, the stomach is smaller so you'd have some element of restriction, and then also an element of bypass where you're not absorbing all your calories. Turns out that's not really how this thing works, but that's what he thought was going on.

From there you keep moving forward. You have all these other procedures — lap bands, adjustable gastric bands, which were just purely restrictive. True stomach stapling, which was the VBG, and now the sleeve gastrectomy. These are the real surgeries, and they were created just conceptually thinking about either restriction or malabsorption, but they work entirely differently than what was thought.

Andrew Huberman: I have a question about your profession generally. I'm guessing there are not large-scale clinical trials of each of these surgeries. So how much license do surgeons have to say, "You know what, I'm very familiar with this tissue — maybe I just graft these two, cut out the middle"? And there's got to be a huge incentive for physicians to have a procedure named after them that saves lives. But there's another facet to this — you can read about the brain, but if you get your hands on brains, record from them, slice them up, look at them under a microscope, you develop a familiarity with the tissue of interest, especially in the context of the whole person who's coming back and saying, "I don't know, I'm still hungry, less hungry, but I got this pain on my left side." Are there any procedures that you would love to be able to do because you have the sense it could really help people, but the red tape is just too thick?

Dr Chris Thompson: I don't think so. I think the proper channels are workable. I do think that there are compassionate use cases where you need to make exceptions, and they have expedited protocols for that. I remember one time I had a person who was bleeding chronically and it couldn't be stopped, and we needed something that was not yet approved in the United States but was approved in Canada. The person had no other option, and we were actually able to get approval within 12 hours to use it as compassionate use, and it worked for the patient. So there are even pathways for that.

It does slow things down — yeah, you're excited to do something and it slows things down — but I think it's always workable. Now, there are other examples where you have a device that's approved for one thing and the company doesn't want to get it approved for everything because they have no money to do that. So you use it off label. That happens every day in every hospital.

Andrew Huberman: Just like drugs are used off label.

Dr Chris Thompson: Yeah. Same thing. Like we use wires when we're accessing a bile duct to remove a stone — that wire has not been approved for that. It was approved for some vascular indication, and we've been using it that way forever because no company ever went through and got it approved. So the whole field is based on this. But if you're developing something truly new, generally the proper channels are very workable. And actually a lot of times they give you even better ideas — like, "Oh, why don't you think about checking these studies? If you're doing this, check this gut hormone." So they have a lot of times they give good feedback that helps improve the study.

Hunger and Satiety Hormones

Andrew Huberman: When it comes to the chemicals associated with hunger and satiety, could we just kind of list out the big players? We're talking about the GLPs, which obviously play a role in satiety and other things. What are some of the big ones that we don't hear about so much anymore because of the GLPs?

Dr Chris Thompson: Well, ghrelin's a big one. That's the hunger hormone. It goes up and it is produced in the fundus of your stomach.

Andrew Huberman: The fundus is the —

Dr Chris Thompson: The very top of the stomach, the top thin part, where the esophagus comes in. A lot of the ghrelin is produced there. And that's how you feel hungry.

Andrew Huberman: So it leaves the gut, travels to the brain?

Dr Chris Thompson: Yeah, and stimulates hunger. What a beautiful mechanism. You're the top of the gut and you're like, "I haven't seen food in a while." I'm still, I have to say, marveling in the background about this thing that the gut expands in anticipation of food and that it's odor-based. So does that mean that the olfactory neurons are communicating with the gut directly? Are they talking to insulin, which goes up, and then the gut expands?

Andrew Huberman: Does insulin actually go up?

Dr Chris Thompson: Insulin does go up too. So before you eat, you'll have a little spike in insulin. I don't know if they ever figured out exactly the mechanism by which smell — or tasting food early on, or seeing food potentially — triggers this whole process to start. But before you swallow any food, you already have insulin coming up a little, your stomach's already starting to stretch to accommodate the meal. So maybe some of it's learned as well. I'm not sure of those mechanisms, but it's very interesting. It certainly plays a critical role.

So after ghrelin — that's your hunger hormone — when you eat it drops, and then it comes back again sometime after the meal. Ghrelin is one to watch because we actually use ghrelin. It's one of the mechanisms we use to get our treatment effects with endoscopic procedures and with surgical procedures too.

So then after the food leaves the stomach, you have your CCK, which goes up. CCK will cause the gallbladder to dump bile, but it actually also will be a satiety signal as well. That's secreted from the first part of the duodenum there. You also have peptide YY and GLP-1, of course — they're big ones. Before you get there, you have GIP from the K cells proximally too. GIP is like GLP-1 — it's kind of similar, not quite as potent. People think of it as Batman and Robin with GLP-1 and GIP.

Andrew Huberman: People might be curious to know that the drug retatrutide — which the more cavalier peptide-curious folks are already getting off from compounding and gray market — as I understand it, promotes GLP-1, GIP, and glucagon. I think the clinical trial Lilly ran showed a 30% reduction in body weight, which is really striking. So it's kind of curious that GIP never really took off as a druggable thing, but now by combining with other things, maybe you actually get some synergistic effects.

Dr Chris Thompson: It does help. I think it helps with nausea, so it allows you to have higher doses potentially of GLP-1 with less nausea. GIP plays that role. It has a role in insulin sensitivity as well, and it does some of the same stuff GLP-1 does, and it's synergistic. But what's interesting about the glucagon is the potential for muscle sparing. Glucagon is usually up when your insulin is down and vice versa. Its job is to say burn fat — that's its main job. It also causes you to dump your glycogen out of your liver a little bit, but the main job with glucagon being up is to say burn fat. It's kind of nice that they're adding that as a muscle preservation mechanism as well as a way of helping to burn some of the fat potentially.

Andrew Huberman: These pharmaceutical companies — however a bunch of people might hate quote-unquote big pharma — they're putting hundreds of millions of dollars into the research. It's kind of an amazing case. Like 20 years ago there was nothing druggable for obesity, as I understand it. What was there was mainly stimulant-based.

Dr Chris Thompson: Like phentermine, which was a sympathomimetic really —

Andrew Huberman: Speed, basically.

Dr Chris Thompson: Yeah, basically. And nicotine — there's a whole set of conversations there. Some people think that when we basically abolished smoking, people started eating more and then America got fat. Then snack foods and highly palatable foods, less fiber, less movement — a lot of things, right? But now nicotine's back in oral forms in a big way, mostly with men but also with women. A lot of people like it because it's an appetite suppressant. I'm not a fan for a bunch of reasons — raises blood pressure, highly addictive, and so on. But it's interesting. People have struggled forever with how can I enjoy food but not eat too much, whether it's a compound that increases something we already make or a drug that makes us move around more.

Andrew Huberman: I think it's metabolic dysregulation, and there are a lot of causes of it. The processed foods, which you touched on, certainly is an element. There was a study done — I believe it was an NIH study in Bethesda — where they had about 20 subjects and randomized them and crossed it over. They could either have whole foods or they could have processed foods, and the people eating the processed foods were eating like 500 calories more a day. So it is something that in your normal environment, if you're eating stuff in a wrapper, you're inclined to eat more of it. And not only are you eating more of it, it's easier to digest — you're getting bigger glucose spikes and you have a lower thermogenic effect of food. It's probably also not doing great for your microbiome because there's less fiber in it.

Dr Chris Thompson: And then there are different layers to it. Then you have your peptide YY and your GLP-1. The GLP-1 is triggered by anything, but glucose tends to trigger more of it. The peptide YY — that's more your proteins and your fat. It does something similar. You stay full longer, I think, with a big heavy fat and protein meal, probably because of the peptide YY. That's something that's been very hard to drug. They didn't have a Gila monster to solve the problem the way GLP-1 did, but it's also very potent. They both come from the L cells in the distal small bowel and the colon.

Those are kind of all the major players. You also have leptin in the background. That's more of a thermostat, if you will. It gets involved in set point and things like that. It's secreted from your fat cells, almost proportionate to fat. So if it's high, generally you're going to probably eat less. If it's low, you're going to eat more. But there's all sorts of problems with leptin resistance and other things like that that complicate it.

Andrew Huberman: I remember coming up through science — leptin was all the rage. Its discovery, its cloning, and everyone thought, okay, drugs are going to come along to mimic or stimulate leptin and we're going to solve the obesity or overweight issue. But it didn't really pan out. Why was that?

Dr Chris Thompson: I think leptin never panned out in large part because of leptin resistance. The hypothalamus and the brain itself just become resistant to it because it's so high in people with obesity for so long. The receptors are saturated. There's low-grade inflammation in those tissues and eventually you just don't respond to it anymore. So the drugs just didn't pan out.

The History of Incretins and GLP-1

Dr Chris Thompson: I think with the GLP-1s it's another story. Incretins in general — we've been talking about these hormones that are produced in the gut, go into the blood, and do something. The concept first came about in the 1930s in London. They were basically grinding up animal duodenums, emulsifying it, and injecting it back into the animal in the vascular system.

Andrew Huberman: Science in its not-crudest form. This is the 1930s, folks.

Dr Chris Thompson: Right, nitty-gritty. The idea for that was secretin. Someone had found secretin — a hormone produced in the duodenum that goes to the pancreas and says secrete fluids for digestion. So this person thought, well, if the duodenum secretes secretin, maybe it secretes something else. They did this study, and in the animal the blood glucose fell. Like, holy cow — something in the duodenum is causing glucose to fall. Phenomenal. "I'll call it incretin because we have secretin." That was where it started, in the 1930s.

Then there was another lab — Sheila Sherlock's lab in London. She was famous for being one of the founding physicians who started the field of hepatology. They had this concept, but what they had access to was a new tool — they had access to a way of actually detecting and measuring insulin. They did a very interesting study where they gave subjects a set amount of glucose intravenously and measured the amount of insulin produced. Then they gave them the exact same amount of glucose orally and found they produced much more insulin. They coined this the incretin effect. They had no idea if it was based on taste, but they knew the insulin was going up, and they thought it was probably coming from the duodenum based on that old 1930s study.

Then other studies came after that, getting closer and closer. Eventually, at Lilly Labs I believe, there was a physician named Bell who cloned the pre-pro-glucagon gene, and from that you get GLP-1 and GLP-2. So now we had GLP-1 identified. Then there was a physician — Bloom, I believe was the name, in London again — who did some phenomenal work. He found GLP-1 was in the bowel where we thought it was. He also found that when you actually gave glucose, GLP-1 increased in the blood. Then he actually infused GLP-1 and found that when he infused it, insulin went up and glucose went down. So now all of a sudden we had a real sign of what this incretin was, and it was GLP-1.

The problem was you had to infuse it for it to work because it gets chewed up really quickly by dipeptidyl peptidase. There's something on the end terminus of it that is susceptible to that, and that's the part that binds the receptor, so you can't really get rid of it.

Then it was in the Bronx in the 1990s when there was a Dr. Eng who was studying Gila monsters. In the Gila monster, he finds this thing in the venom that looks very much like GLP-1. It has one substitution — the second amino acid in from the inside — otherwise it looks just like it and will bind the receptor. The C-terminus is a little longer and different, but this is exendin-4, and basically this is the molecule he discovers. This is what ends up becoming all the GLP-1 drugs.

Andrew Huberman: These Gila monsters don't have to eat very often, so it makes a good candidate to stay.

Dr. Thompson's Path to Developing New Procedures

Andrew Huberman: So you do surgeries of various kinds. People are coming to you — they've all tried GLPs and they don't like them or they're not working, or they'll micro-dose but it's not solving the problem. What sorts of surgeries were you trained to do, and at what point did you become the doctor who seeks out IRB approval to build something better? Tell us that story. How did it start and where are you at now?

Dr Chris Thompson: Really for me it started in fellowship. I moved to Boston to learn interventional gastroenterology — not colonoscopy and whatnot, but doing procedures mostly focusing on pancreatobiliary conditions. The big problem at the time was really pancreatic cancer diagnosis. I was moving there to learn a new procedure called endoscopic ultrasound. You'd be able to put a scope in the mouth into the stomach and small bowel, and then use the ultrasound probe embedded in its tip to see the structures just outside the lumen and gain access to them. You could put a needle in them, and that held a lot of promise. You could maybe ablate lesions with it.

Andrew Huberman: So you're feeding a needle through a tube, watching it on a screen. You're not opening up the abdominal cavity.

Dr Chris Thompson: Yeah. You can do it through the mouth — a natural orifice. You're going through the mouth rather than opening up, which for pancreatic cancer a lot of times was how they would do it. They would go to surgery, open the belly up, and get the biopsy to see what it was, because it's really hard to make the diagnosis. I wanted to learn this new technique where the patient goes home the same day and doesn't feel anything. I thought it was phenomenal.

When I got there, I'd done a master's in health evaluation science at Penn State before going, and I thought I would be doing epidemiologic research. When I got there, my mentor Bill Brugge at the time was a pioneer in this ultrasound. He gave me a needle and said, "Hey, this thing doesn't work to make the diagnosis of pancreatic cancer. I need you to try to fix this." And he was right. The thing didn't work. We had about a 50/50 chance of getting a diagnosis with the needle, because it was designed like a hypodermic needle — like when you get an IV placed. The IV is not taking chunks of tissue out of you. It's designed to atraumatically split the tissue. And those were the needles we were using.

Andrew Huberman: To deliver stuff, not take stuff.

Dr Chris Thompson: Yeah. So I kind of figured out what the problem was. They didn't know the solution honestly, but I gave him my report and the company thought it'd be too expensive to fix. But it still went on. When I was a couple of years into practice on faculty there, we still had the problem — you'd take these FNAs, fine needle aspirations, and you wouldn't have an answer. You'd have people who wouldn't want to have a major surgery having their pancreas taken out without an answer, and then they'd have worsening cancer, and by the time you'd be able to make the diagnosis it'd be too late to treat them.

So that's where I started the entrepreneurial stuff. My first company was based on that. I needed a team — one of the engineers had the brilliant idea of how to change that bevel design. I knew what the clinical problem was, but you needed a team to fix it. We hired engineers, got together, and came up with a needle that could biopsy the pancreas without causing pancreatitis or any problems. That really became very instrumental in helping a lot of people get the diagnosis earlier. We're saving lives with that. And now we have preserved cellular architecture, so you can do precision medicine — you can actually test different drugs on the tissue and see what it's going to respond to, do immunostaining. It's a lot better than just having a few shaved cells.

That was the first time I really got involved in trying to solve a problem like that, and that was before I started diving into metabolic disease, where I've spent really a large part of my time.

Detecting Metabolic Dysfunction Early

Andrew Huberman: What I hear when you say biopsy — people hear biopsy and they go, "Oh boy, you're getting poked with a needle." But if you told me I could come into the clinic and spend one long day under anesthesia and get completely non-damaging biopsies of every single one of my major organs through the mouth, I'd be like, why wouldn't you do that? We take blood tests now — people go, "What's my testosterone, my estrogen, my luteinizing hormone, my lipids, my small LDL, my ApoB." 20 years ago, if you wanted to get a blood test and you didn't have a problem to motivate that, it was thousands of dollars at best. Now it's trivially inexpensive for most people. So how soon are we going to just be doing biopsies with non-damaging procedures?

Dr Chris Thompson: So I think a lot of times with biopsies, you have to be very targeted to get the tissue of interest. Even in the pancreas, you could be even in the area that looks like a lesion and not get cancer cells out. So you have to be very targeted. But once you do get the tissue, you can do all these stains and really figure out what's going on — is there a genetic predisposition? Is there some way it'll respond to one drug over another? I think that's phenomenal.

But I would like to see the diagnostic studies become less and less invasive so they can scale easily. The one problem with procedure-based diagnosis — I like procedure-based treatment, I love it, it's better than surgery — but when you get diagnostic studies similar to colonoscopy, there's a scaling problem. When a patient has to come in and spend an hour or two hours with a doctor, that doctor is taking care of one patient for two hours and is outnumbered. Everyone needs screening and it becomes very complicated. So I would love to see innovation in technology go where we have minimally invasive ways of diagnosing things, whether it's via your smartphone and AI or via minimally invasive scans. Blood tests are great because they're quick and easy to do. And we're not even talking about metabolic health — there are several things we could be doing non-invasively at home right now that we're not doing that would catch it much earlier.

An example is most of the time we're waiting for hemoglobin A1C, and that's the marker of diabetes. Once you have a high A1C or a high ApoB — which they're probably not checking, maybe just an LDL-C or something — once those are high, we know there's a problem. However, there are signs much much earlier than that.

Metabolic dysregulation follows a fairly predictable sequence. First it's calorie excess. In the western diet it's usually glucose — you have too much glucose around, too much saturated fat, but too much glucose. You could catch that by doing a CGM, a continuous glucose monitor. You can see if you have particularly glucose spikes to certain foods. If your glucose is shooting up to 200 with certain meals, you know you're sensitive to that and maybe you should change how you're eating it — try to eat it after having something fatty, maybe avoid it. Because we know this is part of a sequence that's going to lead to problems.

This goes back to the Whitehall 2 study, which was on British civil servants. It was a prospective longitudinal study, and they found that if someone had high fasting insulin, they were more likely to get diabetes long term. They could detect this thing 15 years earlier. They could do something about it. But no one does, because no one looks for fasting insulin.

The other thing that's very relevant is the NHANES study — another large database, more cross-sectional. What they found was that less than a third of people who are lean are metabolically healthy. That's crazy. 12% of the whole population — less than a third of lean people are metabolically healthy based on their parameters. The parameters looked at waist circumference, glucose, blood pressure, and whatnot. The word there is: start looking early, and don't look with the traditional things. We have to look at other things.

Getting back to metabolic syndrome ideas — first you could check for glucose. Glucose spikes — a CGM can do that. I wouldn't say wear it all the time. Get one for a month or two, learn what spikes your glucose, see if something spikes your glucose and adjust it. Next you have fasting insulin. The next thing that happens is, in anybody, they have the insult of excess calories. Insulin's job is to take that sugar and push it into the cells because glucose is really bad for the body. We know this — if you look at end-stage diabetes where they can't control their glucose anymore, they go blind, they have kidney failure. It's killing the vasculature. Glucose is sticky. It glycates things. It causes problems. So the insulin's got to get it out of the bloodstream.

Next in the sequence of metabolic dysregulation is high insulin levels. Fasting insulin goes up. You can get a fasting insulin level — that's the next thing you check. It's not a lot to ask for. It's an inexpensive test and you can see if you've evolved into that problem where now you have chronically high insulin levels. Part of that is due to eating too frequently and eating things that have a high glycemic index or load that's going to cause your sugar to spike. If you're eating every few hours, insulin goes up and spikes. It drives the glucose out of your blood, but then the insulin stays high — it doesn't go right back down. It stays high for a few hours. So if you're eating every few hours, you always have this high insulin. That's going to lead to other problems.

The next thing that happens is ectopic fat. Your fat exists in different areas. You have subcutaneous fat — that's where it's supposed to be, your depot for energy. Then you have visceral fat, which is really in your omentum, in the abdomen, in your mesentery around the bowel. Then you have organ-associated fat — some fat around the heart, some fat around the kidneys. They're all adipocytes, all fat cells, and their job is to store fat and release it. Then the last bucket is ectopic fat, where you have fat in cells that is not their job to store fat — like liver cells, or muscle cells, or pancreas cells. And that becomes a problem.

Andrew Huberman: It's like Wagyu beef.

Dr Chris Thompson: Yeah. It's like Wagyu beef. Those cows are overfed and don't move.

So that's the next phase of metabolic dysregulation. They've done all sorts of great studies showing exactly from each step what happens and how you get there. That's when you get fat in your muscle and fat in your liver, and that's bad. Fatty liver is very bad. And then that is what goes on to insulin resistance.

For the fat, how can you look for that? You can do a waist circumference measurement, waist-to-height ratio. You can get a DEXA scan that'll tell you if you have visceral fat. A CT scan, MRIs, other things will do it too. Or an ALT — look at a liver test measurement, an aminotransferase in your liver. That'll signify some inflammation.

Then you have insulin resistance, which is a little harder to check. There's a formula — it's a fasting blood glucose and a fasting insulin level. You multiply those and divide by a constant. If it's greater than two, you have insulin resistance. That's the next phase.

And then finally, you have metabolic inflexibility. Your body is supposed to change between what it's burning. If you're fasting, it's supposed to be burning fat. If you're eating, it's supposed to be burning some element of carbs depending on what you eat. You can develop metabolic inflexibility once you have insulin resistance — when you're fasting, you're not really accessing your fat anymore. Your fat's still there, but you're burning more your glycogen stores, and god forbid you're chewing up muscle. And when you eat, it doesn't shift over to burn the carbs well either. It kind of just doesn't know what to do. That's a loss of metabolic flexibility.

By then you're getting near the time when something's going to happen, because once you have a loss of metabolic flexibility, studies have shown you're more likely to gain weight and develop obesity. You're more likely to start losing beta cells — they become apoptotic and you lose beta cell mass — and you start having all sorts of other problems. This is a very typical sequence that's backed by science and different clinical trials. And at each step of the way, you have a study you could do to find out about it.

The last one, metabolic flexibility, is a little harder because you have to do a breath study where you're looking at gas exchange. Athletes do this to optimize performance. You can do this where you get a DEXA scan — a lot of places will tell you how many calories you're burning or what you're burning. It's based on the first law of thermodynamics. It's a ratio of a volume of carbon dioxide divided by volume of oxygen. When you eat carbohydrates, carbohydrates have an equal number of carbon and oxygen, so it doesn't require much oxygen to burn them. But when you burn fat, it requires more oxygen. So if that ratio is around 0.7, you're using more oxygen — that means you're burning fat. If it's one, you're burning carbs. And there's an in between. You sit in a chair and breathe for half an hour. There are companies developing at-home methods for this too, and one of them recently has one that does both the oxygen and the carbon dioxide.

Andrew Huberman: I've seen this — like a little box that you breathe into.

Dr Chris Thompson: So you can do everything. And then, or you can just wait until you have diabetes and your A1C goes up. There are a lot of things we should be doing before we do the standard test of looking at your fasting glucose and looking at your hemoglobin A1C.

The Case for Early Metabolic Testing

Andrew Huberman: It's gratifying to hear that you put the CGM pretty early on that list. I want to be clear — I'm neither complaining about nor trying to turn you against your colleagues. But I'll just be really blunt. I've gone public many times saying, as the cost of blood testing comes down, this is awesome — you get a window into lipids, hormones, things that can be very informative whether you have issues or not. And the pushback on that from the medical community — not all, because I have friends in the medical community who will quietly say, "Yeah, I totally do that test, I take that test" — but many of them just say, "Oh great, now patients are going to be coming to me saying, do I need to be worried about this?" I actually put a post out recently about whole-body MRI as the cost comes down. It generated some press this week. There's a celebrity who took one of these types of scans and identified a malignant issue that could be cut out and very likely saved their life or at least extended it.

I get it on one hand why a lot of physicians are worried about people walking around with a lot of data. I heard the same about CGMs — "Oh gosh, who needs to know their glucose goes up and comes down in the absence of diabetes or pre-diabetes? You can have a glucose spike. We don't want people walking around neurotically worried about eating a grape." I totally understand. Your body can manage these things. But now that CGMs have been out for a couple of years, I don't hear much pushback. Somebody wants to use a CGM for a couple of weeks and see how they react to different foods post-exercise, post-poor sleep — cool. So it's kind of wild to me that physicians don't want patients to have data. But here I'm hearing something very different — you're saying yeah, I think people should pay attention to how they're regulating their blood glucose.

Dr Chris Thompson: The problem with medicine is it moves very slow. A lot of people are going to want a randomized control trial, another randomized control trial, maybe a few more, and then a meta-analysis. Maybe understandably. But from a patient perspective, people want data now if they can get it inexpensively. And these are elective procedures — no one's saying you have to get this done.

It's unfortunate. There's a reason for it — do no harm. But by the same token, it does not necessarily do the patient any favors by waiting for something that is logical and makes sense. There's enough evidence for this sequence of events for metabolic illness all the way back to Syndrome X in the 1980s. We know there's this constellation of things. And we also know that if you don't act early, you're much less likely to have a good treatment effect. If you start treating someone once they have diabetes, it's much harder to get them back to healthy and normal. They've already lost beta cell mass.

Andrew Huberman: Or peripheral nerve — they could have loss of sensation in their fingertips and toes.

Dr Chris Thompson: And now, before, we might not have had access to things. We had access to fasting insulin. But think about this — say you don't get the CGM and you have fasting insulin, and the fasting insulin is normal. You don't know if the step before that is a problem and you're going to have problems with fasting insulin. Or if it's abnormal, you need the CGM to learn how to eat, because that fasting insulin is high for a reason. You're spiking your insulin — you've got to figure out why. So then you go back to the CGM and you learn. I don't have any stock in the CGM company either. You've got to learn how to eat to not spike your insulin in that case.

Most of the time in western countries it's the glucose and insulin that are the problem, and it's the very beginning of it. So why not learn about yourself? Take some responsibility. Learn and prevent these diseases from going on. I do hope physicians are more open to encouraging this as well, where we start acting earlier. It's going to be better for the population in general.

Andrew Huberman: Because I could see why physicians would be very reluctant towards self-directed interventions. I get it. But here we're just talking about getting data, and it's voluntary. The costs are coming down. And as you point out, having data early is better than having data late when it comes to physician-guided interventions. What I've heard in conversations is, "Well, the problem is a patient's going to figure out they can eat junk food that's really bad for them, and they'll think it's okay because it's not showing up in their CGM." I think that's not giving patients enough credit, because generally if you're getting a CGM, you probably are trying to do the right thing. More information is better almost in all circumstances with diagnostic stuff. Now, I agree with this MRI thing being concerning because you might have a lot of little lumps and bumps that now you're doing diagnostic studies for that could increase risk. But more information I still think is better. We just have to find better ways of doing the follow-up confirmatory diagnosis.

Dr Chris Thompson: I agree. And actually, a lot of times they give you even better ideas. Even when you have procedures that have gone through rigorous evaluations and have FDA approval and they're ready, you have so many people that are reluctant to send patients for them.

Andrew Huberman: Why is that?

Dr Chris Thompson: I think it's in the culture. Some of these procedures, you can't learn them in a weekend course. You have to spend a year or longer learning some of these things. So for a physician who wants to add something to their practice, they're not going to dedicate a year to it. They might do a weekend course and realize it's too hard and then they don't adopt it. The problem is the people that do adopt it and they're not ready. So then doctors that are referring go, "This procedure's been around for six months or a year, it really has great data in the clinical trials, but does the guy down the street know how to do this after doing a weekend course?" And so they're reluctant maybe for that reason. Just give it time. Wait to see if the insurance companies think it's a good idea, then maybe we'll start sending patients. It just moves slowly.

Andrew Huberman: I have a friend who's really into cars, and he told me that in the mechanic and automobile community, a similar thing happened as things became more and more computerized — there was a lot of pushback because it makes it hard for auto shops to do their work. It changes the field, you need more tools, sometimes those tools are expensive, and people like to hold on to the way they were trained. This is absolutely true of most every field. Adapt or die — or your patients die. I have to imagine there are good surgeons, mediocre surgeons, and exceptional surgeons. Are there places where you've brought in devices or machines that could offset the mediocre surgeons?

Dr Chris Thompson: Most devices we see are kind of incremental improvements — a little bit of better wire, devices a little more ergonomic. But what I see happening more recently is AI starting to have an impact where it can actually coach you through procedures. You have a heads-up display on the screen. It will actually give you information so you're not just seeing the images you're working on. It can actually highlight certain structures you want to work on. It can actually point to something where you want to put your stitch. It can count the stitches as you're placing them and tell you if they're close enough together. It can change the shape of the stomach as you're working on the stomach to let you know if you're having a good treatment effect. This is something we never could have done before in real time.

Right now it's not widely available yet — this is in research centers. But you can actually see this happening in real time, and it's phenomenal. You see that in more and more surgical procedures where AI is real-time coaching you and in endoscopic procedures. Additionally there's the hope for robotics to help as well. We've done a lot of research in our lab on robotics and how it can take trainees learning a very complicated procedure and shorten the learning curve dramatically. We'll randomize the trainees and have them do the traditional way — this is usually resecting a tumor from the colon leaving the colon in place, a very complicated procedure — and the fellows will learn. They'll spend a couple of weeks training in both modalities. They'll struggle horribly with the original way. That's why it takes two or three years to learn how to do it. They'll sit down with the robot and be almost as good as an expert. So robotics are very interesting.

And now in the future, when you start layering on AI and automation with the robots, you might have a big win. We've seen this before with different surgeries — with intuitive surgical's robots when they first came out years ago. It democratized the field. It took mediocre surgeons and made them excellent, and the excellent surgeons were still excellent. But it really helped the ones that were struggling.

Andrew Huberman: How did the excellent surgeons feel about it?

Dr Chris Thompson: I think they're supportive of robots, but I think that a truly exceptional surgeon is probably just going to be better without the robot. The robot is just going to make them worse.

Andrew Huberman: How do I know if I'm getting a truly exceptional surgeon?

Dr Chris Thompson: That's a good question. Understanding whether you're getting the best physician for something is really hard to determine. Even as I'm looking for a doctor for something, it's hard to find the right person. I'm in a massive medical center and have great connectivity, but knowing who truly is the best is complicated.

Some things we rely on are volume — case volume and historic case volume. How many procedures do they do? That's important. And probably more important, how many have they done over the course of their career. Volume is important. It's not the whole story, but volume is important.

We need in medicine to move more towards objective metrics. This is one thing AI can do for us. I'm involved in a healthcare delivery platform called Everself. What it does basically is the doctors doing these procedures are held to a certain metric. It starts with just collecting the data — finding out what their weight loss outcomes are, finding out how many stitches they place per procedure, looking at their procedure time, looking at their complications. You're grading all that. But the next layer is putting AI on top of it where the AI, not only can it coach you through the procedure, it can give you a grade at the end. It can be very specific — it can tell you you placed this many full-thickness sutures versus this many. You want 100% of your stitches to be full thickness. Maybe the doctor's putting in 70% full thickness. That's not good. This number of sutures were close enough together, some were too far apart. It will give you a grade. This is the pattern used. This is the volume of stomach you reduced it by. It'll give you a grade at the end of that procedure.

And that grade is incredibly important. The idea next would be to share that data so people know what grade you're getting. It'd be great to share that with governing bodies that do credentialing, so people who are truly underperforming maybe get a refresher. It would be nice for patients to be able to select who they're going to go to based on objective metrics. AI can do this probably across the board with other things as well.

We've seen this a little bit with adenoma detection rates in colonoscopy, where they used to publish that. Doctors were expected to have a certain number of polyps they'd see per colonoscopy and they'd report that. But then the problem was all the patients wanted to go to the one or two doctors that had the highest rates, and their wait times became enormous. That's a problem as well. But there should be a reasonable cutoff where a certain level of expertise is required, and I think AI hopefully will help us get there.

Andrew Huberman: I'm excited by what you told me about how AI can provide real-time data and perspective about how the stomach will change shape, with the opportunity to make adjustments as you go as opposed to having the patient heal up and have to come back in for another surgery. Years ago, I saw something amazing. A neuro-ophthalmologist friend allowed me to sit in on something. He said, "People forget that surgeons wear microscopes on their eyes." They wear these optics that allow them to see things bigger, and then there are all these new tools — like a little drop of fluorescein, a little bit of innocuous liquid that creates a contrast for the surgeon to see what is what and not cut the wrong tissue. It seems like such an obvious thing, but I was told that for a hundred years the same procedure had been done without that. Eye surgeons had to essentially guess based on their intuition and training what was tissue to preserve, what was healthy, what was unhealthy. These what seem like simple technologies have improved the margins of safety and outcomes tremendously. So the idea that you would have AI combined with really good microscopes and better surgical tools — to me it just seems obvious.

But a lot of people hear AI and robot and surgery and they go, "Oh my goodness." I think with AI people think it can go rogue, has a mind of its own. When you sit down to do a procedure and you're getting information from AI, where does your trust come from that it's giving you good information as opposed to faulty information?

Dr Chris Thompson: The AI is trained on thousands and thousands of procedures — more than I've done. So it recognizes patterns. You have to use your clinical judgment, and you're not using this AI blindly. You might ignore it sometimes. You don't have to follow it. Now, if it becomes the time where you're automating operative robots using AI — like suture placement — that'd be different. If it's doing it itself, that's different. But for this, where it's just suggesting where you put a stitch or showing where a blood vessel is, I think it's a huge advantage.

We do these procedures that are very technical where you tunnel — you're creating a potential space in the esophagus. Back to that earlier person that couldn't swallow because they had achalasia. The procedure is: we go in through the mouth, inject a little fluid under the mucosal layer to lift it with a pocket of fluid, make an incision in that, take the endoscope, slide under in between the mucosal layer and the muscle, dig all the way down to the bottom of the esophagus, and then cut through the muscle. When you're doing that, there are vessels in there and they're hard to see. AI can actually see those vessels because it's got pattern recognition, and color them for you so you don't hit the vessels as you go, reducing your chances of hitting a blood vessel. Beautiful. That's just one example of something that's a very complicated procedure where you're making certain aspects of it a little easier.

Physicians are looking at the equivalent of a medical textbook, but it's not in color — it's black and white and gray and beige. And with endoscopic ultrasound, it's not even that. It's all gray. When we're doing endoscopic ultrasound looking for a pancreas tumor, it's all different shades of gray. You can turn on a button to see if there's blood flow, but it's all gray.

Years ago in my lab, I was trying to use image registration. I could take a CT-PET scan and link it to the angle of the probe, so you could see a CT scan fluctuating in the probe of the ultrasound and lay the ultrasound over it, and then you'd get an idea of the tumor you're looking for, the lesion you want to biopsy. It was too hard to do — it would take three hours of preparation to set that up. You could never scale that. Now with AI, other groups are doing similar work and it's almost automated. So I'm hoping we'll see image registration with these very advanced imaging tools help us with diagnosis and hopefully even with therapy too.

They're doing something now called hyperspectral imaging, and they're doing it in surgery as well. There are several groups doing this. One group in London is doing phenomenal work. They're using all these narrow bands of wavelengths — just tons of wavelengths — and they're finding out that each tissue actually has a fingerprint. You can actually use this hyperspectral imaging to fingerprint tissue, and you can actually see margins of tumors with this, without giving a dye anymore. So that's what LEDs are doing in different kinds of cameras. Instead of CCD chips, you have CMOS. With newer technology, even though it seems incremental, with LEDs being able to fluctuate the wavelengths of light and your chips being able to read it faster and better, we're able to make better diagnoses.

Understanding the Sick Duodenum and Endoscopic Procedures

Andrew Huberman: I feel like one conceptual structure we could put on things is that medicine has a couple of different ways to determine what's going on. One is the stuff that comes out of the body. And then we have the surface of the body — the power of skin, how the eyes look. And then as we go in, we're still trying to do this. X-rays let us see fractures. The goal has always been to get more information with less invasive procedures. And I feel like now we have blood tests so you can pull stuff out of the body. In 2026, this is where we're at. I think it's super exciting, but the tools might still be kind of crude now compared to where they'll be in 10 years. Are you hopeful that in 10 years you can go into a tube, 20 minutes later walk out, and we might be able to scan with good enough resolution to detect any tumors anywhere?

Dr Chris Thompson: I don't know if we'll get there. I'm hoping, because we are seeing capsule technology improve, imaging technology improve, and blood tests improve. There are all sorts of things you can do with genotyping and whatnot. I think that's exciting. But where we are seeing improvements, I think, are in learning about physiology and how things work, and then being able to do a targeted approach. So you're not just doing that with drugs. GLP-1 isn't really targeting a deficit necessarily. There are different ways you can treat things — you can either find a pathology and treat the pathology, like cutting a tumor out. Or you can take normal physiology and augment it, and that's what they're doing with GLP-1.

Andrew Huberman: But they're ramping things up thousandfold over what they would be even in the healthiest person. Most people don't know that, by the way. They think the GLPs are bumping things up like two or four fold. It's like no — never before in human history, at least to my knowledge, have people walked around with this level of GLP-1 circulating in their blood.

Dr Chris Thompson: Yeah. You're supposed to have a little tiny amount that's produced in response to a meal, and then it goes away. Our GLP-1 is secreted — it's nutrient responsive. It's secreted from L cells, and then it does its job. It goes to the pancreas and says produce insulin. It goes to the stomach and says slow emptying. It goes to the brain and says you're full. And it does it in response to a meal, in much smaller doses. So medicines have done this for a while where they kind of see something as a mechanism they can augment. But that augmenting is very interesting.

Surgery for a long time wasn't doing that. They were just thinking, "Oh, I'm going to make you malabsorb calories. I'm going to make this tight so you feel full quicker." But now that we're understanding mechanisms, and there's some great research that has gone into this, we can actually develop targeted therapies. And I think that's what's very exciting — more so even than a new device. It's being able to do targeted therapies and get better outcomes.

Where I started with this is in fellowship. I saw a patient with a gastric bypass anatomy — they have a small gastric pouch and a bigger stomach. A patient was sent to me who had bad reflux, weight gain after the gastric bypass, and their diabetes had come back. The surgeon basically said, "Hey, take a look at this patient. See if they have an ulcer. Find out what's going on." So I went and looked, and there was this little hole between the pouch, the new stomach, and the old stomach. I thought, well, maybe the acid is produced in the other side. Maybe the acid's coming up through that fistula. We had a new device — a suturing device that you could actually put in through the mouth and put stitches in. I thought maybe I could use it to close that hole.

I waited until I was on faculty a few months and talked to the surgeon. He was supportive. The procedure was FDA approved. The device was FDA approved. The procedure for this specific use was not — no one had closed a fistula with it. But we talked to the patient, told them we weren't sure if it was going to help or not, they were willing, and we did the procedure and closed the fistula. I was hoping the reflux would stop. The reflux stopped, but the person started losing weight and their diabetes went away almost immediately again. And that was for me — this is 2003, 2004 — I was flabbergasted. Was it a coincidence? What the heck was that? Why is closing that little hole so important?

So that's what got me involved in understanding the gut hormones, because if I learned about the gut hormones and why we saw this treatment effect, we could potentially manipulate them to get better results. That was the beginning of it for me.

Shortly after that, one of my friends and colleagues actually did some animal work. He had a rat model — GK rats, which are rats with diabetes that don't have obesity. They were a great model because you didn't want weight loss to confound things. He did two surgeries — a foregut and hindgut method. One surgery basically excluded the foregut. He excluded the duodenum and the very first part of the jejunum. He did a little bypass surgery there so no food could get into the duodenum. The other one he did a gastroduodenal anastomosis — stomach to small bowel but left the rest open, so food could go either way. What he found was that the ones that had the exclusion, their diabetes got much much better. The ones that didn't have exclusion didn't get better at all, even though you were dumping stuff into the distal gut. Very interesting. So he thought there was something very important about foregut exclusion, and he hypothesized there was something called an anti-incretin in that bowel that, if you exclude it, you got a better treatment effect.

That was my fish work, and then his very interesting animal work got us going down that path. And it kind of fed well into something that was done in the 1980s — a continuation of that work that Sherlock had done looking at incretins. This famous publication looked at the same study that Sherlock had done in London, where they were giving glucose to look at the insulin response, but they did it in diabetics and in a normal healthy population. The normal population had that exact same incretin response — you gave a certain amount of glucose intravenously, little spike; same amount of glucose orally, big spike. Diabetics didn't do that, and they had already tied it maybe to GLP-1 and maybe in the bowel. So very exciting. Maybe by excluding this foregut you're playing a role, or maybe not. That was the beginning of trying to understand the procedures for me.

With that, I then did another study where I closed those fistulas, and 60% of people had resolution of their diabetes. If we didn't close it, no one got resolution of diabetes. So we learned there's some important element to foregut exclusion.

Then various device companies started getting involved in the space because of this information, and a company came up with the idea of putting a liner in endoscopically — like a little sleeve. You anchor it, it has a little stent that springs open, you anchor it in the first part of the small bowel. It covers the duodenum and protects it. It's an implant so it has to come out at some point, maybe a year later. But it was very interesting because I was part of those clinical trials, and we found you had a one-point drop in A1C in diabetics — that's fantastic — and you lose weight, about 7% total weight loss. So clearly it's doing something and it's important. The problem is it's an implant and it's got to come out. But it's exploiting this mechanism potentially.

Andrew Huberman: So you're essentially creating more compartmentalization along the tube.

Dr Chris Thompson: Yeah. The duodenum, and then there's a liner you place in it. This is like a stent — it springs open and holds its form inside it. And then it's a sleeve that kind of goes down. You can still get all your secretions that go on the outside of the sleeve and track down, but it's right after the pylorus, the outlet of the stomach. So all the food is going in the tube. The food is inside the sleeve. The digestive enzymes are outside the sleeve and they don't mix for a few feet down. So that's very interesting, and it worked. The problem is it's still in clinical trials, been around for a while, but it's an implant. So it's just like taking a drug — eventually it has to come out.

But then there was a brilliant idea that came about from one of my colleagues at the Brigham. He's a cardiologist, and he knew I was trying to do something. I had traveled to Brazil and was doing endoscopic procedures there, and there was a doctor in a room nearby doing a very novel experimental surgery called ileal interposition. This was a lean diabetic — not suffering from obesity, just type two diabetes. He was taking the distal small bowel, kept it on its mesenteric blood flow, resected it out of the distal small bowel near the colon, moved it up, and put it near the duodenum. His idea was that GLP-1 was denser in that part of the bowel and also denser down lower, and if you moved that up higher, you'd get a more immediate incretin effect from GLP-1. You'd hit GIP and then immediately GLP-1 and have this amazing effect. And he did. It was incredible. These people's diabetes went away and they didn't lose any weight, because he didn't actually have any blind areas. The food — he didn't change anything. There's no restriction, no absorptive change. He just moved that part of the bowel up. That was phenomenal.

So I was trying to do that endoscopically by harvesting tissue from the ileum via colonoscopy, creating stem cells, and then injecting it in the foregut and getting it to take — that wasn't successful. But one of my colleagues, the cardiologist, said, "Why don't you just burn the duodenum? Ablate the duodenum. There are different ways you could do it — steam, hot water, etc. Just ablate it and see if you can reset those stem cells, because the duodenum is sick."

This is very interesting research — the duodenum is sick. If you can reset the duodenum, it might work. And he did. He started a company and it's been great. Now you don't have a sleeve in place. You don't reroute any bowel. You just ablate the duodenum. And what happens is your A1C drops by over a point. You don't lose a lot of weight by just ablating the duodenum, but your A1C corrects. That's a potential treatment for diabetes. They've also done some studies — I don't believe these are published yet — showing that when someone comes off a GLP-1, if you use this treatment, it keeps them from regaining their weight. So you can take a GLP-1 and then have your duodenum kind of reset, the stem cells come back, and you've maybe healed those tight junctions and other problems that you're having.

Andrew Huberman: So it regenerates?

Dr Chris Thompson: It regenerates. Yeah. It comes back more healthy and more normal.

And the rationale for that comes from a lot of very good research. There were studies that showed in mice — overfeeding studies with a control group — when you take them to necropsy and look at their bowels, the bowels in the overfed mice are longer, heavier, the villi are longer. They've adapted. They've upregulated the ability to absorb calories. And then these studies have been repeated in humans where people getting gastric bypass are already going to be doing surgery, so they resect part of their small bowel, and someone getting cancer surgery is a control patient. They look at the differences, and there are extreme differences. The villi are longer, it's thicker, there's more inflammation in people with obesity or type two diabetes. A lot more inflammatory cells. The natural killer cells are up eightfold. Macrophages up 1.5-fold in these studies. So you have more inflammatory activity going on in these patients. The only thing that's different is really obesity.

Additionally, if you look at those patients and you do immunostaining for tight junction proteins — scaffolding proteins — you'll see that those are much lower and they're disorganized.

Andrew Huberman: Two questions. So if I understand correctly, if people overeat, the villi — little finger-like protrusions inside the gut that can sense things but also collect nutrients — they're growing to adapt to the elevated levels of calories. And so essentially you've changed the digestive tract in a way that yes, they can make more use of those calories, but that also creates a more pro-inflammatory environment. Do I have that correct?

Dr Chris Thompson: That's absolutely correct. Also, because they're changing in configuration and using that energy, the cells are using energy to do other things, and your tight junctions are deprioritized.

Andrew Huberman: So then there's this secondary or parallel effect on the tight junctions. We haven't really talked too much about tight junctions here, but I'm familiar with them from the blood-brain barrier. Some tissues you want things sticky but not too sticky. Some tissues you want them really sticky. My understanding is that the tight junctions — as the name suggests — the goal is to keep stuff inside the gut, not let bacteria out. Is leaky gut a real thing? Because I've heard it's sort of like chronic fatigue syndrome — a lot of the standard medical community hears "chronic fatigue syndrome" and they go, "Okay, that was made up by people in the Bay Area." I'm only half kidding. But that's how a lot of physicians react online to this phrase "leaky gut." But we've had a fair number of people come on here and talk about tight junction deficits, bacteria getting out of the gut — this isn't good for the body, inflammation going up, bacteria circulating places they shouldn't be. So is leaky gut real?

Dr Chris Thompson: Well, increased gut permeability is 100% real.

Andrew Huberman: But that sounds like a different language for leaky gut.

Dr Chris Thompson: Yeah. So it is.

Andrew Huberman: So why is this phrase "leaky gut" so irritating to the medical community?

Dr Chris Thompson: I think if you say leaky gut, it could have other connotations. Someone might think that leaky gut means it's responsible for a certain constellation of symptoms — like irritable bowel or Alzheimer's. Because you see in lay literature they'll say leaky gut is associated with XYZ, and it's not clear that that phrase is really talking about the same thing I'm talking about. The danger is calling something leaky gut when people already might have a definition for leaky gut in mind — like it's responsible for all these other problems.

But let me tell you what leaky gut is to me, or what increased gut permeability is. And I'll tell you that it's very real and it is actually tied to metabolic illness.

We can start with a study that used small bowel biopsies — this was recent, just last year. They did small bowel biopsies and from the stem cells grew little organoids — three-dimensional cultures that behave as they should. They had a control group and a group with MASH — metabolically associated steatohepatitis — and obesity. They looked at the organoids and found that the tight junctions were far less well-developed and more disorganized in the MASH patients compared to the control patients. Additionally, they did transcriptomics on it and found that they weren't even producing the RNA to produce the tight junction proteins. So clearly at the transcriptional level, they were downregulating the tight junction proteins. With the immunohistochemical staining and then transcriptomics, they found that the tight junctions just weren't functioning as they should in people with MASH.

So if you don't have tight junctions, it stands to reason you might have quote-unquote leaky gut. Another group actually looked at something similar — same population, patients with MASH. They studied it using different tests you can do to look for a leaky gut. You can give something that's very small but should not get through those tight junctions. Chromium-51 EDTA is one that was used in this study. You give it — it's not supposed to get into the bloodstream. In patients with MASH, it zipped right in, much higher levels than there should be. In patients without MASH, it wasn't getting in. Additionally, in patients with celiac disease that was treated, it wasn't getting in. But in patients with fatty liver disease, it was getting in, and it's probably playing a role.

If you think about it, the gut — the first place it goes is the liver. There's a portal circulation and the gut goes to that portal circulation. Everything that goes through there has to stop by the liver, with the exception of fat. Fat gets into the lymphatics and dumps out of the thoracic duct — it doesn't have to actually go to the liver. So if you have bacterial products — LPS, lipopolysaccharide, a portion of gram-negative bacterial cell membrane — if that gets through these tight junctions, it causes all sorts of problems. It is going directly to the liver. They're inflammatory. They interact with toll-like receptor 4, and that starts all sorts of inflammatory cascades via NF-κB signaling, etc. That can be problematic.

Another group proved that was problematic — this was done at Duke. They actually took LPS and injected it into healthy people and found that their inflammatory markers went through the roof. They also did clamp studies in these patients and found it induced insulin resistance. So yes, I think leaky gut can be involved in all this stuff. And that gets back to our very early discussion about fiber and fermented beverages and how important it is to keep your microbiome healthy, because that microbiome and that butyrate is critical to producing healthy enterocytes. As well as healthy tight junctions, a healthy mucin layer. And actually, butyrate and the microbes and the byproducts of the microbes work with your immune system, your innate immune system, and it tells them what to recognize and what not to recognize — which is just as important, because your bowel is full of bacteria. So absolutely very important, and you do see where this increased gut permeability is associated — hardcore, good science evidence — with real illness. So absolutely it's a problem. I just don't want to blame it for everything.

Artificial Sweeteners, Fats, and Dietary Recommendations

Andrew Huberman: I get a lot of questions about artificial sweeteners and negative effects on the gut microbiome. It seems like there's marginal to zero effect on insulin and resting blood glucose from artificial low-calorie sweeteners in a way that would lead people to say these are bad. There's no reason to run out and use them if you don't want to. But the weight loss data say people who drink diet sodas instead of water actually lose more weight. I've seen those data, but this is not an incentive for people to start drinking diet sodas. It sounds like saccharine and Splenda are probably worse for you than stevia and aspartame. Where are you at with these things in terms of their potential negative effects?

Dr Chris Thompson: I think they're better than high fructose corn syrup, for sure. I think we should be treating fructose like alcohol. Fructose in fruit is fine — I'm not worried about fructose in fruit because it comes with a matrix around it. It's not a rush of fructose into your liver. But fructose can only be processed by the liver, and so it's busy as it is. It's got to take the burden of a beverage, which is absorbed very rapidly, goes directly to the liver, and has to be dealt with. It gets trapped in the liver very quickly and that's the only place it can really process it. So I think fructose is something to watch. Again, not if it's in fruit — even in juices, it can be kind of processed fruit, so it's similar. Minimally processed stuff is better.

I think the problem with artificial sweeteners is they come in foods that are highly processed as it is, and you can't separate the two. I think that's why for a while people were so down on polyunsaturated fats — because they'll come in a bar full of a bunch of other stuff that's not good for you.

Andrew Huberman: The polyunsaturated fat is also bad for you in some way — that was like a recent phenomenon.

Dr Chris Thompson: Well, no — the food that it's in is bad for you, but the polyunsaturated fat has been shown to reduce LDL and has health benefits.

Andrew Huberman: Essentially, translation for people — seed oils, basically.

Dr Chris Thompson: Yeah. There's some still debate about whether or not the processing of them can make them worse, but it's hard to say. Did you see this recent avocado oil thing out of UC Davis? UC Davis went and analyzed all these avocado oil-containing products that are supposed to be healthier. You know how much avocado oil these products contain?

Andrew Huberman: Zero?

Dr Chris Thompson: Oh no. And the pushback has been that maybe they're looking at the wrong metabolites of avocados. I don't know how this is going to play out, but it could potentially do more damage to the non-olive-oil community. In my mind, the safest thing is you just use olive oil, a little bit of butter here and there. No one debates olive oil. Everyone knows it's good for you. No one thinks it's bad for you. But the seed oil and lard thing — they go back and forth and it's kind of like professional wrestling. Olive oil, butter — or you're the physician. Tell me, am I thinking about this wrong?

Andrew Huberman: I do the same.

Dr Chris Thompson: No, olive oil is the best, obviously. Then small amounts of butter. I think the problem is it's the overall amount of saturated fat. But polyunsaturated fats have a lot of proof that they're very safe. Where you get into problems is if you have a big container of it, you're not going to use it in a reasonable amount of time, it's sitting in the sun or something, and it oxidizes. That's a problem. You don't want to take an oxidized oil into your body. Or you're deep frying with it over and over again and you start generating trans fats. That's a different story. But in general, I think they're fine.

And I don't even think you need this omega-3 to omega-6 ratio people used to worry about. I think you need a certain minimal amount of omega-3s. If you eat fish once in a while, you're getting all you need.

Andrew Huberman: Do you strive to get some fatty fish in your diet?

Dr Chris Thompson: I do. Yeah. I love fatty fish. It's good for you. Try to do it a couple times a week.

Andrew Huberman: I take Lovaza, the high-dose omega-3 pharmaceutical, because I don't want the mercury. It's cleaned of mercury. My blood markers are where I want them to be, but I'm curious — what's your read of the data on omega-3s for metabolic health and cardiovascular health?

Dr Chris Thompson: It's mixed. I mean, it's probably better for Alzheimer's — if someone's starting to show signs of Alzheimer's, I think it's better for that. The data — I think the problem is universally supplementing is not necessarily the way to go. You want to find a deficit and then supplement. Even with vitamin D — most people probably are deficient, so they benefit from it, but there's no point in really doing it unless you're deficient for most things. And I think with omega-3s it's similar. You want to get your daily allowance. And if you're a vegetarian, you can do it from algae — the original source. The fish are just consolidating, right? So you don't have to eat the fish, but you can get that in some kind of supplement form.

Exercise, Set Point Theory, and the ESG Procedure

Andrew Huberman: What else do you recommend to your patients as they start to move away from obesity? So obviously fiber, some fermented foods. It sounds like resistance training might be in the list given that they're at risk of becoming thin but with more jelly tissue than lean mass. Do you prescribe resistance training?

Dr Chris Thompson: Absolutely. All my patients I ask them to do resistance training, even before they start losing weight, before they go through a procedure. It's essential. Zone 2 cardio is great — it's good for fat burning. You're in that zone where you're burning fat and not carbs as much. HIIT is great. High-intensity interval training is great for mobilizing visceral fat because your visceral fat has beta-adrenergic receptors on it. It also has gonadotropic hormone receptors on it as well. So it's responsive to acute stress. It will mobilize when you're going through the stress of high-intensity interval training. It won't be burned right away because you're burning carbs at the time — you're burning your liver glycogen and your muscle glycogen. But you mobilize the fat at least. And that's kind of what it's designed for. That's why you have some visceral fat there. So I try to have them do HIIT, a little Zone 2, and then resistance training. I think those are the most important things long term.

Andrew Huberman: Do they do it?

Dr Chris Thompson: This is very interesting. I think they try, and depending on how they lost the weight determines if it's effective. It's the theory of set point — or really a defended range, if you will. You have this defended range of what you think your weight's supposed to be. That's set by a variety of things — leptin is part of it, your thyroid hormones and whatnot. And you think you're supposed to be a certain weight. Then what you do is you do a crash diet. You lose a bunch of weight. The Biggest Loser was a great example of this. You lose a bunch of weight, and now you're fighting several factors.

One factor is your body is smaller, so it burns less. You have to eat less to just maintain the same weight you're at now, this lower weight. That's a bit of a problem. You downregulate your gut hormones — you're producing less GLP-1, less peptide YY, a little less CCK. So your satiety hormones are being produced less. Your ghrelin goes through the roof if you do this with diet and exercise. In addition to that, your muscles become more efficient — I think they become 25% more efficient in doing a similar task. They're going to burn less fuel to do the same task. Your non-exercise energy expenditure, your basal metabolic rate — they all kind of go down. So you're burning less calories at rest. Your whole body is fighting you. It wants to go back to that weight, whatever it thought it was supposed to be at.

The Biggest Loser — there was an NIH follow-up study to that, and they found that they were burning 500 fewer calories per day after that. There are other studies that have shown this as well if you lose weight that way. So that's why it's so important. GLP-1s help fight part of that — you're replacing the GLP-1 that you're not producing. But you're not addressing the ghrelin or other things.

So with our procedures, we're addressing these very targeted. One bridge into that is the ESG procedure — the endoscopic sleeve gastroplasty. This is the procedure I developed in 2012. You go in through the mouth, someone's sleeping obviously, with a little scope, and you fold the stomach on itself. The goal of that was to do two things. One was to augment the stretch receptor — it's a smaller pocket, so when food hits it, the stomach stretches quicker and you have the vagal afferents that go up to the nodose ganglia and then into the hypothalamic area.

Andrew Huberman: You tell the brain we're full.

Dr Chris Thompson: Yeah, exactly. We're full. Stretch fast. Boom. So when you stretch, you get that signal. That's part of it. The other part is you suppress ghrelin because food stays in the stomach longer. So it's doing two different things. When those people lose weight, they don't have to worry about their ghrelin going up because it's been suppressed. So it's easier to keep the weight off for 10 years or longer because you're not fighting that part of the countermeasures that the body will do to defend its potential range.

We're not doing anything necessarily with GLP-1 and other duodenal hormones, but you'll see it actually does have ways of dealing with this. So how do you augment weight loss? You have all these different targets. One thing we're doing now — we talked about how ghrelin resides in the fundus. In addition to that ESG where we tighten the stomach, someone developed an idea — I think they were in Germany — where you can actually ablate those fundal ghrelin cells because they live in the mucosal layers. You can get to them. They use argon plasma coagulation — there are different ways to ablate it. You just kind of spray this over the fundus and it kills off the ghrelin-producing cells. They grow back and there's not much of them. So now all of a sudden you can suppress ghrelin as well. The weight loss goes from about 18% with ESG alone in a top center, up to way over 20%, maybe 25%, if you start ablating the ghrelin-inhibiting cells.

And then you add to it. Now if you've delayed gastric emptying, your CCK is not spiking as much as it was. But GLP-1 is an issue. So now what if you combine that with a small bowel procedure? There are different small bowel procedures that we've come up with. Magnetic anastomosis is one we published about 10 years ago. We did it in the Czech Republic where we used endoscopes — it was a hard way to do it. My partner went from below via colonoscopy. I went from above. We released these two magnets and connected the jejunum, the first part of the jejunum, to the lower part of the ileum.

Andrew Huberman: We should probably tell people what anastomosis is — basically when you connect two tubes?

Dr Chris Thompson: Exactly. You're bridging two tubes. They did it originally with sutures — you cut a hole and suture the tubes together. Then they did staplers, but they're big and bulky and hard to position. So our lab developed magnets. These are ring magnets, encased in nitinol so they can take a certain shape. You put them through a tube — in this case, an endoscope. They come out and form a ring. We went from the top endoscope and formed a ring in the jejunum, and from the bottom formed one in the ileum way downstream. Then we had an anastomosis that would allow the food to directly pass there. And what we found is you get these big spikes in GLP-1.

So now what people are doing — I'm conflicted and can't do this part of the procedure — but what they're doing is they're doing that anastomosis and they're doing a suturing procedure endoscopically, and together you're really replicating a full gastric bypass. You're having GLP-1 hindgut spikes. You're getting that sense of restriction and the vagal afferent signaling. You're getting ghrelin suppressed. And you're getting really amazing weight loss.

So what we can do now is take a procedure that was really big — it started off as a big open procedure that had certain risks to it, we didn't know how it was working, and it did a bunch of different things — and we're targeting different aspects of it. The goal moving forward is to even be more precise and find out what someone is going to be more responsive to, and then just do the least you need to do. Maybe they just have ghrelin that's driving them — just ablate the ghrelin. Maybe they need something more. And people are actively studying that. They're studying the phenotyping of obesity. It's quite exciting.

Andrew Huberman: I'm sensing another theme here. This procedure that your lab developed — it's increasing GLP-1, but I'm guessing it's not increasing it thousandsfold like a GLP drug would. It's got some other positive consequences that help address the obesity. I'm kind of sensing a theme — we have these drugs like Ozempic, Mounjaro, etc. that blasted GLP-1 through the roof, helped a lot of people who needed help, but there were a lot of side effect issues. Then along comes retatrutide, which is like, okay, let's increase GLP-1 but let's also kind of bump up the GIP system a nudge or two, let's also bump up the glucagon system, and lo and behold we get a much better effect — muscle sparing and actually better weight loss. So kind of a lesson that you don't really want to push really hard on one lever in biology. Maybe the more combinatorial approach is the better approach.

Dr Chris Thompson: Oh definitely. I think you can mitigate risk by doing that — by not giving too much of one thing. Hitting multiple levers is definitely a way to get a treatment effect without exposing the body to potentially the harms of going too big on one thing. So that'd be the argument for these multimodal approaches. Then you can also combine these procedures with the drugs. You do an endoscopic procedure like tightening the stomach and then give a drug and see if you get much more weight loss, or at a lower dose.

Andrew Huberman: This was what years ago on this podcast we looked at with the whole ADHD thing and the effects of these drugs. Parents who get a great effect of an Adderall or a Vyvanse for their kid who couldn't focus — it's remarkable. But they're worried about the reduced growth effects, the sleep effects, and so they're in this trade-off. And that's where I think it doesn't have to be either-or. Maybe this child could get by with a lower dose of medicine if they're also doing some things behaviorally, some things with nutrition, etc. We don't often think like that. Americans want the drug that fixes the problem. And then we get all pissed off when we had a generation of kids raised on amphetamines. Well, maybe kids just needed a little more exercise.

So it's gratifying to hear that you're doing these multi-pronged approaches and that you do recommend exercise including resistance training. And you know, they do all these studies that show diet and exercise alone don't work because of the set point. Look AHEAD was a great study — that was a lot of treadmill and heavy diet, and they found they got like 6% total weight loss or something like that at 10 years and no improvement in heart disease. There are several other versions of that where it's hard to do it alone because you're not addressing the countermeasures the body throws at you. But that doesn't mean it's irrelevant. When you do a procedure like a gastric bypass surgery, or you go on a GLP-1, you still have to fix the fundamentals that got you into the problem to begin with. You need to start getting more fiber, have a better diet, try to avoid the insulin spikes, start moving, start exercising. Otherwise the treatments will fail — the endoscopic procedures, the surgeries, the medicines will fail unless you really address those underlying problems.

Gene Therapy and the Future of Metabolic Treatment

Andrew Huberman: So given where things are at now, where are things headed next? You mentioned AI. What's the potential role of other technologies to improve health and outcomes?

Dr Chris Thompson: Well, I think one thing that's very exciting is gene therapy. We talked about GLP-1s and how it's mega-dosing, super-physiologic, not nutrient responsive. There's a new company working on a new approach — a gene therapy — and I was involved in the very early work for this. Basically what they're doing is they've developed a viral vector that has the gene for GLP-1 in it, and they're using the promoter for the beta cell insulin gene. So basically when a patient would secrete insulin in a nutrient-responsive way, they would simultaneously be secreting GLP-1.

These viral vectors are a beautiful tool of biology where you can put some genetic cargo into a virus that doesn't cause any problems but allows for stable expression and the production of certain proteins in a cell. How are you getting into the pancreas? We're actually using endoscopic ultrasound — that same device we developed to biopsy the pancreas. We're now using something similar to actually treat. You can ablate tumors with energy as well. People are using electroporation to cause apoptosis, thermal means, but you can also do a fine needle injection. And we're injecting the viruses basically into the tail of the pancreas. You wouldn't want to just take this intravenously because they end up in other tissue. We've done a lot of work to make sure those things stay in the tail of the pancreas. We've done a lot of animal studies where we've injected it and used green fluorescent protein to make sure it doesn't end up in areas it's not supposed to be.

Andrew Huberman: Is there a pancreas-specific promoter — translation: this would allow even if some got out, it could only get expressed in pancreatic islets?

Dr Chris Thompson: Very very close. But you still just don't want it getting anywhere else. The only place it becomes active is in the beta cells. It doesn't become active in the alpha cells. So you're really just active in beta cells. And you secrete insulin into these little vesicles, so you're secreting GLP-1 into those same vesicles. So then when you have your meal, the vesicles release GLP-1 and insulin together.

Andrew Huberman: Oh, that's clever.

Dr Chris Thompson: Nutrient responsive. You're making the drug. We were already making the drug, but now you're making it at an elevated rate. Not only that, you're not making it in the L cells where it has to go all the way up through, go to the liver, go around, do its thing. You're making it at the place where it's needed, right at the pancreas. So it has a paracrine function, and it's much faster.

Andrew Huberman: How often are these cells turned over? Because if it were brain, no problem — brain cells don't turn over. But how often does the pancreas turn over?

Dr Chris Thompson: Very important. Why you can't do it in the bowel is because they're not terminally differentiated — you're turning over your whole bowel every five days or whatever. Pancreas cells are terminally differentiated. So they're not going to be changing. It's a permanent thing — the episomal DNA stays in there, doesn't integrate into the host DNA, stays next to it and transcribes with it.

Andrew Huberman: I feel like there's another theme emerging. We're hearing about drugs that you can get one injection to permanently lower your LDL. We're now hearing about gene therapy to chronically elevate GLP-1 at exactly the place and time that you want in order to offset excess calorie consumption and obesity. Is this what we're going to see — instead of people taking drugs, they're going to take a one-time injection?

Dr Chris Thompson: That's what I'm hoping. It's very exciting. They actually just entered clinical trials in the Netherlands. So we'll see how that goes. But it looks very promising. One-time GLP-1 injection — that could be nice. Additionally, you could use it to augment other therapies. You can use it to augment the gastric procedure or the small bowel procedures. It might be another tool in your armamentarium. It might be more useful for diabetes than for weight loss. We don't know — it's so early right now, but it's certainly very encouraging.

Andrew Huberman: Really glad you're doing this work, because I'm aware of a few conditions — rare, fortunately, but not exceedingly rare — where hyperphagia is an issue. Prader-Willi syndrome and other syndromes where these kids just can't stop eating because they lack hypothalamic signals. My read is that the traditional GLP drugs are not really working there. This would be amazing.

Dr Chris Thompson: And in rodent models it's phenomenal. We did these trials where you randomize mice to get semaglutide at high dose — much higher than you'd get for a human — and then the transgene. Both groups lose weight. The transgene group loses a little more. Then they stop losing, so they don't keep losing weight forever, which is good. Then you took the group that's on semaglutide and randomized them further to get nothing or to get the transgene. They get the transgene and they go right back down to the same settling point, which is great. And the ones that were randomized to nothing put all the weight back on. Phenomenal. So it seems to be getting really good results from a weight loss standpoint as well.

Dr. Thompson's Background and Philosophy

Andrew Huberman: I want to take too much more of your time, but if you're willing, we could just briefly talk about you for a second. You're an interesting person. It occurs to me that you had certain solutions in hand but you decided to search for better solutions. Were you always a tool builder? In medical school and residency, or even prior — in high school — are you the person who sees that the reason you have to keep fixing the antenna on the TV is because the antenna sucks? Were you that kid?

Dr Chris Thompson: Yeah. My mother would attest to that, unfortunately. I took my motorcycle apart in high school. Couldn't get it back together. Had to have it flatbedded away and fixed. And I fixed some parts in my car that ended up bursting into flames. So I'm much better at dealing with patients than with machines.

Andrew Huberman: Could have been the other direction.

Dr Chris Thompson: Exactly. So no, I always would tinker with things for sure, and I needed to do things with my hands. That's why in medical school I couldn't be a general internist. I needed to solve problems with my hands, and I think that's fulfilling to me. I don't like managing a slow demise. I felt like internal medicine — we were giving people a reason to continue with their current life rather than addressing problems. Like, your blood pressure is high — instead of finding a way to really help them address that, you give them a medicine. Your LDL is high — instead of finding a way to address it, you give them a medicine. And you see what that gets us into. We treat high LDL and ApoB really effectively and we reduce mortality from that specific thing, but we took our eye off the ball. Fatty liver is up, diabetes is up, and people are still dying in greater numbers. So I feel like we need to address the underlying problem. And that's very important. But aside from that, I just like doing things with my hands, and I think that was a large part of why I was going to go into cardiology or interventional gastroenterology.

Andrew Huberman: So grateful that you're a tinkerer. It's a unique thing to find these qualities and expertise woven into the same person. The fact that you clearly have immense compassion for your patients and you're willing to come here and share information publicly — you have many many important roles in your daily life. So the fact that you take the time out of your schedule to educate the public is something I and everyone listening have immense gratitude for. And that you're thinking about what could be done better — that's the ultimate quality in my opinion of an excellent physician or scientist or engineer. When it comes to physicians and the general public, we need people who are thinking about how things can be improved. Yes, there are some solutions for some people, but we need to broaden the treatments to help many more people. So I'm just very grateful to you, and thanks for coming here today and sharing this info.

Dr Chris Thompson: Well, thanks so much for having me. Very kind. It's definitely always a team effort, as you know as well as anyone. Everything is a team effort, and I think innovation is never the result of one person's work. It's a whole group, and I've been very fortunate to be surrounded by a bunch of amazing people that help us move things forward.

Andrew Huberman: Well, throughout today's discussion your reflex to give proper attribution is more a testament to what you just said. It's not lost on me — and nor on the people listening — that people who give credit where credit's due, it says a lot about them. So thank you. Come back again maybe in a couple of years when you've solved everything, or close to it. Just joking. I'm sure you guys are making tremendous strides, but these things take time. Once again, thank you very much. This was very very informative and has enriched my thinking a tremendous amount. I'm sure everyone listening as well.

Dr Chris Thompson: Thank you.