243: Dave Feldman’s “The Cholesterol Code" Kicks the Lipid Hornet's Nest

Dave Feldman, a software engineer turned citizen scientist, went on a ketogenic diet and watched his LDL cholesterol spike dramatically while his family members on the same diet saw no such change. Unable to get cardiologists or lipidologists to study the phenomenon, he built a public charity from scratch, crowdfunded the research, and designed the Keto CTA study — enrolling 100 metabolically healthy people with very high LDL and scanning their coronary arteries with CT angiograms at baseline and one year later. He filmed the entire process. The result is the documentary The Cholesterol Code.

The findings were notable, but the story became more complicated when the AI-assisted plaque analysis from the company Cleerly produced results Feldman considered statistically anomalous — including zero regressors across 99 participants, and greater plaque progression in people with a calcium score of zero than in those with a positive score, which contradicts established research. 

Dave Feldman Contact Info
The Cholesterol Code Movie: https://cholesterolcodemovie.com/
Citizen Science Foundation Website: https://CitizenScienceFoundation.org: 
Cholesterol Code Website: https://cholesterolcode.com/

Pre-Order Stay Off My Kitchen Table at Amazon.

Like what you hear? Head over to IFixHearts.com/book to grab a copy of my book, Stay Off My Operating Table. 

Ready to go deeper? Talk to someone from my team at IFixHearts.com/talk.

Stay Off My Operating Table on X:   

• Dr. Ovadia: @iFixHearts
• Jack Heald: @JackHeald5

Learn more:

• Stay Off My Operating Table on Amazon
• Take Dr. Ovadia's metabolic health quiz: iFixHearts 
• Dr. Ovadia's website: Ovadia Heart Health
• Jack Heald's website: CultYourBrand.com

Theme Song : Rage Against
Written & Performed by Logan Gritton & Colin Gailey
(c) 2016 Mercury Retro Recordings

Any use of this intellectual property for text and data mining or computational analysis including as training material for artificial intelligence systems is strictly prohibited without express written consent from Dr. Philip Ovadia.

Transcript

0:59

Welcome back, everybody. This is a podcast that you all want to hear the Stay Off My Operating Table podcast with Dr. Philip Ovadia and a guest. I know I want to hear Dave Feldman. Dave was one of our, the earliest guests on the show back right after fire was invented. But we hadn't had him back yet. And Phil, you guys I really just want to sit and listen because you bring the scientific medical jargon stuff, but Dave is kind of, comes from my world sorta, he translates this stuff. So I'm just gonna shut up and listen. And I'm not really promising that. Yeah, I'm really looking forward to this conversation. People may be wondering what, why has it been so long? Why didn't we have Dave back on sooner? And the reality is 'cause he's been too damn busy doing lots of cool stuff. And that's where I. We wanted to bring him on 'cause it's now really gotten to the point where it's out there. And lots of exciting things to talk about. Most notably the movie about what Dave's been up to basically the last three years since we spoke to him. And that is something I'm particularly excited about, as well as, of course, the study that led to the movie and everything else. So Dave, maybe just for people who have been under that rock and don't know who Dave Feldman is, give the brief version of how you got into, what we're gonna talk about with the study, with the movie. You were an engineer computer guy, and somehow you stumbled into becoming a citizen scientist. Let's talk just a little bit about how that happened. Yeah, so first of all, thanks for having me. There's just no getting around it. It gets tougher and tougher to give a brief version because it's been an action adventure with multiple sagas. But yeah, here's my best attempt. I had no interest in nutrition or medicine or any of that sort of stuff 11 years ago. But I go on a ketogenic diet to dodge type one, type two diabetes, and my dad and my sister get inspired to do it at the same time with me. We're all feeling the awesome effects of it. I'm excited. They're excited. They get their blood work six months later. They're level look great. They helped all kinds of issues that they were dealing with. Their cholesterol hardly changes, but then I get mine, my cholesterol shot through the roof. Why I become quite obsessed with this and I start learning everything I can, not just about biochemistry, but also about lipidology, the study of lipids, of cholesterol and so forth. And then I'm really under, I'm really trying to understand why mine doubled in my dad and my sister, who were first degree relatives didn't. And that starts leading me down these roads with self experiments where I start experimenting with myself, gathering a lot of data, and I start working on something that's now known as the lipid energy model, which has now been published. But then in addition to that, I realized that while we're all interested in the mechanisms, the bigger question that everyone's most interested in is the risk. Are people who go on a ketogenic diet like me, who see their cholesterol go up because of the ketogenic diet. Are we at a higher risk? So I then start a a public charity, public scientific charity called the Citizen Science Foundation. And I crowdfund a study. I go out and literally raise the money, a lot of it from the low-carb community to make a study happen where we take a lot of people like me who are metabolically healthy, but with high levels of LDL. And then we studied them with high resolution heart scans called CT angiograms. CT angiograms get a high resolution scan of the heart. We get them a baseline, so they go there in a day zero, they get one scan in this roughly 365 days later, a year later, we get a second scan. That's the follow-up scan for 100 of these participants. And the a hundred participants were healthy. People like me who again, metabolically healthy high HDL, low triglycerides, but sky high levels of LDL, that study as it's happening is getting filmed. So we have a documentary that was at the time that we were doing the study, and I'm getting the data, is capturing it all on camera and then ultimately cuts it together into a documentary which we're seeing in your neck of the woods. I believe it's St. Pete, right? Yeah. So this ties it all up in a bow for you. What's the date on St. Pete? When am I coming out there and joining you and all of these other superstars for your show. You're coming out next week as this drops. So April 6th is the date that you're coming to St. Pete. And the movie is in the process of being rolled out widely. And I believe we can now announce, right? It's gonna be on Amazon in what, about three weeks or so? A little less than on, on the 17th. Yeah. So if you can't, if you can't join us in St. Pete or one of these other various locations, and by the way, you can just go to cholesterol code movie.com right now and look to see a nearby showing for you. You could watch it in the theater with likely a whole bunch of people like us, which is a great experience. Or if you need to feel free to wait until April 17th and then you can rent it and watch it on Amazon there. And it's exciting. It's exciting we're, it's happening. It's right now, we're right about at the cusp of everybody not only seeing the study, but seeing the story around it. Now what what kind of gave you the idea, right? We'll say this, right? It is one thing to be a citizen scientist, right? And say, I'm gonna, get this study going, right? Because the background is you approached a number of cardiologists, a number of, people from the traditional medical world and say, Hey, I've noticed this interesting phenomena in myself and other people. We have this whole community. This sounds like something that you should be interested in studying, right? And you got crickets, right? To the point that you had to say, okay, I guess I'm just gonna have to figure out how to do this myself. So that's amazing enough. But then you had the foresight to say, we should capture all this on film. And, this could probably lead to a good movie. And it turns out that it's gonna lead to a maybe even more interesting movie than I imagined you anticipated because of some of the things that happened around the study. But what gave you the initial concept of, I'm gonna do this study that's hard enough, but let me bring it one level more and let's get this all on film. And oh, by the way, you kind of crowdfunded this, funded it yourself, right? You really weren't it is not like you had some, big company sponsoring this or anything like that. Yeah. I mean, put yourself in my shoes. Imagine you have spent two years, and I can tell you the years it was from 2017 to 2019, and you're knocking on doors, you're trying to get to. To the cardiologist and the lipidologists who have research budgets and you're like, Hey, it's, look at this. There's this group that's over here. They have sky high levels of LDL, but they don't have the other things that cluster with high levels of LDL that are independent risk factors on their own. This is exactly what you want in science. You want the variable of interest all by itself, right? But you're not getting enough interest from them. And you're starting to get a sense of just how strong the lipid hypothesis is, not just with regard to the papers that have been published, but with regard to the level of expectation with everyone around us, right? Meanwhile, the Facebook groups of lean mass hypers responders, it's growing and growing. Right now, by the way, it's up at around, I wanna say 14,000 members. Like it's a lot of people. And it starts dawning on me more and more that, okay, if we do go so far as to do a study, we need to be sure that as many things we can make happen to ensure its integrity are in place. And therefore, I could have gone with a GoFundMe, and a GoFundMe would've been way cheaper. It would've taken a bigger margin on raising of the money. But then there's this other issue with GoFundMe, which is that it's not as clear exactly where the money goes. It's not quite as accountable as say, a public charity would go, right? So I'm like, ah, screw it. I might as well go so far as to actually get a public charity registered. And by the way, it's huge amount true 5 0 1 C3 public charity. You have to go through a lot of hoops, especially these days. Fine, we do that. But on top of that, I have a 0% admin overhead. So everybody who contributes. Outside of third party credit card processing fees, we have something like a 96% throughput. It's like ridiculous. We're like a plus, plus. Good lord. Yeah it's because You could consult with nonprofits just about how to do that. Part of it is we backstop it. We have own Your labs, our company over here. That covers a lot of the hard costs, and it's also that I can turn back around to the people and I can say, Hey, every dollar you give us, almost all of it is gonna go towards the science. Then of course, we needed to work with a a bonafide research center, which is what we did with Lundquist and Yes. Then here's where the filming comes in. Hey, wait a sec. What if we actually film things as it happens? Would you trust a documentary that was capturing things in real time? Or would you trust one that after it was all done, then they determine what it was that was going to make it onto film? Or even if there was gonna be a film? So no, we've been filming for four years. That hats off to Jen Eisenhardt and Wide Eye. They're the ones who did all the work. They did absolutely phenomenal job. But it was a lot. It was a lot to put all this together. And again, all of this comes back to how can we bring more transparency and and accountability to out there, to the public, as well as wrapping it into a film. But lastly, and I think Jack will appreciate this one. Lastly, there's this one other factor, which is how much can we anticipate from those two years, from 2017 to 2019, how much. Can we anticipate that if the data goes in the way, we hope it will, that the cardiologist, the lipidologist, the journals, the institutions will carry that through to the general public I would offer. It still has to get done. You still have to do the good research, but the real message carrying potential is gonna be outside of that. It's gonna be in things like having a documentary on real research that can reach the lay audience, that can reach the public. And that's the story of low carb, right? Low carb wasn't implemented through the medical zeitgeist. There wasn't like some landmark low carb studies. And then all of a sudden, doctors, around the world were like, I think I need to put my patients on a high saturated fat, low carbohydrate, low fiber diet. No we know that's not how it happened. So I think that this is gonna be no different, and that's why I'm glad we've got a film to help deliver the message. Yeah, I mean, we've seen quite the opposite, right? There actually have been landmark studies on low carb diets. For instance, Virta Health, right? Showing that you can get 60% reversal of type two diabetes at two years on a low carb diet. And again, crickets, right? What should have a major impact on the practice of the everyday physician has not because of, it just doesn't get out there. So we're working against that headwind. And quite frankly, the cholesterol hypothesis, the diet heart hypothesis is probably the most firmly entrenched belief in medicine at this point. And so getting doctors to think that, and again, you're not out there trying to say the diet heart hypothesis is totally invalid. You, the purpose of this study was to say, here's a small group. Admittedly, this is a, relatively small phenomena, right? So it's a probably a small percentage of people that would see this, and then they have to do something unusual, go on a low carb diet to get this hyper response. But, it provides some interesting context as to where the diet heart hypothesis may be incomplete. And but you're right, getting, getting that then to change actual practice is a monumental task. And I do believe you'd believe that this is gonna be a ground up movement ultimately. So maybe let's talk about some of the, unexpected things that came up with the trial. One of the first things I want to touch on is you were able. Finally to get one cardiologist who is credit interested and to be a co-investigator Dr. Matt Budoff, who's a prominent cardiologist, very well known, in that space. And thankfully, maybe go a little bit into a, what it took to get him involved and get him as a co-investigator on the study. Yeah. Of course, first things first, just to put credit where it's due. The biggest thing to get the interest is the money. So once you raise a little pot of gold, then you can start taking that around. And now the door is open that you were knocking on before. I. In my earliest days of this research journey, I thought, oh, I hear about these millions and millions of dollars of NIH grant funds and so forth. Maybe it's just me doing the right pitch and then that gets in. No, once I have the money, then I can go to people who might be interested. And Dr. Budoff has lots of studies under his belt. I think as of this moment. He has something in the neighborhood of 2000 papers with his name on it, 2000, which is kind of, that's not even possible. Yeah, it's, he is a professor. He's a chair. He's, if you look at his resume it's a phone book. That's for those of us old enough to know phone books. Yeah. But anyway we got connected to him. We pitched what we were interested in. And I thought that we would need a longer span of time. Like I thought we'd need something more like say, three to five years. But with CT angiography there's a, there's already lots of studies that are a year or even less than a year for which they're tracking changes in plaque. Now, there's more that I would unpack with that topic at a later point that gets a bit more esoteric. But the most important question was the one we all had on our mind, which is that if you go by the present day lipid hypothesis as it stands right now, and you've got people like we end up having in our cohort who have not just high LDL. Levels of LDL comparable to those who have homozygous familial hypercholesterolemia. Now it's a mouthful, but what it means is those people who have LDL of 400 or higher, like the ones Brown and Goldstein, were looking at, the expectation is that even in the very first baseline scan, if our eligibility criteria was two years or more, if a number of them showed up with say half a decade on them at those levels, that there would be a clear signal of plaque at those levels. And so we might not even, okay, real, real quick. I have been doing this with Phil for going on four years and I'm still, I still have trouble keeping in my mind the LDL and the HDL and what's good and the what's bad, and I can't be the only one. So real quick, just describe what an LDL of 400 or higher. What that with the standard belief system that's out there in the zeitgeist, what that should mean. Yeah. And to, to the average person like me who can't remember what's good, bad. No that's, I'm glad you asked that. So if you, right now are John Q Public, you go get blood work and it includes a lipid panel. You're gonna get four numbers, total cholesterol, LDL, cholesterol, HDL, cholesterol and triglycerides. Total cholesterol is the total. It's LDL, cholesterol, it's HDL, cholesterol. And it's also another kind of cholesterol we won't get into in the moment. It's the higher that is, the more bad it should be, but more in, I wanna say 50, 60 years ago, they focused in a bit more on making the distinction between LDL cholesterol and a lot of cardiologists like to say, as a pneumonic, remember, L for lousy as in it's the bad cholesterol. LDL is bad. HDL, the pneumonic is happy. H for happy is the good cholesterol happy and lousy? Okay. Yes. So the lousy cholesterol, the higher it is, the more there's the expectation that you are at a higher risk. The HDL cholesterol, the happy cholesterol, the higher it is, the lower you are at risk. You usually want your LDL low. You want your HDL high and then there's this, That's the current lipid hypothesis. That is correct. Okay. Now, a lot of modern lipidologists would say, actually just don't even look at HDL weed. We put too much time focusing on it. But now with CETP inhibitor trials, we've determined that HDL is not that relevant. It's not that important. That fourth marker triglycerides, that's a measure of fat in the blood. If somebody has very high levels of triglycerides they have something called hypertriglyceridemia, you can actually even see it in the blood. They can spin the blood and there's even a foamy, cap above it. I'm not gonna get into the geeky side with the lipin model, but there's an interplay between all of those that become relevant. And the study that we had didn't just have a requirement that you had to have high LDL cholesterol, the lousy cholesterol, you needed to also have high HDL and low triglycerides, those three together. The reason for that is because that's a common pattern that I and many others have. And that leads into this phenotype, you'll hear me use a lot called a lean mass hyper responder. I've seen Nick Noritz talk about that as well. And he's very much a lean mass hyper responder. He has his LDL is in the mid five hundreds. So while we're here, while we're in the lay person version, let's really put this in perspective. What's recommended is for you to have an LDL, even if you're just healthy, an LDL of one of under 100. That optimal would be under 100 would, unless you have existing heart disease. If you have existing heart disease that was detected in some fashion. I believe it is normally, or I think until recently, Philip, you correct me. It's, you're trying to get your LDL below 70, and I think more recently there's a push to get it under 55. Am I remembering this correctly? Yeah. As we're recording this, there was a study released a couple of weeks ago suggesting 55 should be the goal. And there are some guideline changes that have occurred that have moved that direction. The one thing I wanna point out about what you were talking about, that triad, right? The high HDL, low triglycerides, high HDL those are very rare outside of a low carb diet, right? Seeing that triad together is actually pretty unusual. The most patients who come into my office or come into the standard doctor's office with a high LDL level they're gonna trend towards. Higher triglycerides and lower HDL. And I think that's a very important context, right? Because one of the, one of the discussion points around all of this is, okay, is all high LDL the same? Essentially? That's the question that really you're trying to answer. And so many of us have been trying to answer within the low carb space is, most of the time high LDL is in the context of low HDL high triglycerides, which is an indicator of metabolic disease and this whole host of things that goes with it. So in this fairly unique situation, high LDL, but high HDL and low triglycerides, does the LDL mean something different? And so Jack, here's my lay person version. I want you to grade me on this one. We have we population for which we know. That there are fat that's moved around in the blood is in these vehicles called lipoproteins, a mix of lip lipids and proteins. So it's a vehicle, it starts with cargo. That's how it leaves. It comes out with cargo, goes in through the freeways of our arteries, and then drops off that cargo and gets smaller and becomes LDL. Okay. Now, if you're metabolically unhealthy, do you think it would make sense that there would be a problem with those vehicles finding spots to drop off their cargo? That question is over my head. That should, I'll help you out. The answer is best. The mechanism is not clear to me. So There's less parking availability for the fat that they have on board because there's already so much fat. In the other fat cells. Okay, so somebody who's metabolically unhealthy, they're overweight and they're past their personal fat threshold. There's difficulty dropping the fuel off to the tissues that normally are reticent to, except fuel. They're like no, we're full. Right? Therefore, you have a traffic jam. You have more of these lipoproteins that started off full but aren't turning over as quickly. Okay. And that so far, yeah. That in turn means that you have higher, what's that cargo that they start out with? Triglycerides. So there's higher triglycerides, and also particle turnover creates higher levels of HDL, so that's why you also have higher triglycerides and lower HDL, just like Philip described, right? Yeah. You also have this other thing when you do get LDL, it's this weird small dense particle version. So you hear about small dense LDL. That's also correlated with these other two markers. Now I'm an engineer, so I go, oh, okay. So you have all of these three things that seem to be downstream of a metabolic problem. Seems pretty straightforward. But then there's this other group over here that's going, or maybe triglycerides themselves, all by themselves are atherogenic in some way and maybe are atherogenic, they cause atherosclerosis. Okay. And then separately, low HDL. Also causes problems.'cause it picks up plaque from cells. We learn that from the arteries. Oh. And small dense LDL particles independently cause atherosclerosis. So I'm looking over there and I'm going, I don't know that I agree with you guys. I think these three things, because they all come together and they're all downstream of this problem. I could take you Jack, I could take you to Dave's island where I feed you cake and cookies and all pizza and change your lipid profile. The lipids inside your body change to the thing I'm describing. And so with that, it's called an, it's a profile called atherogenic dyslipidemia. Okay. I'm sure Philip knows it really well. And I'll bet a lot of your patients Philip, have atherogenic dyslipidemia. It's this lipid profile. That tends to be with people, and again, I'm thinking o okay, but this is a reflection. It's a consequence seemingly more than it is a cause. Like they're hyperinsulinemic, they have all sorts of other, okay, so the reason in that case for why there's marginally higher LDL for those same folks along with those other three, I believe that reason is important. Now we have this other population over here, lean mass hyper responders instead of a higher LDL because of the traffic jam. They have a higher LDL because they're actually powered more by fat and they're successful at dropping off the cargo. So you have more traffic on the road at the vasculature. Oh. Because they're dropping off more cargo. That's why the lower triglycerides matched with the higher HDL. Makes sense. There's turnover and it's much faster. Yeah. A good model, by the way. I really like that analog. Exactly. So you've got high LDL over here, but you've got really high LDL over here, but all of our research is over here on this population. That's all they research. That's why when I go to the lipidologist and the cardiologist, I'm like, this is great though. If you guys think it's the LDL in the A OB, because of all the research over here, don't you want to get to this one over here that doesn't have high triglycerides? It doesn't have low HDL. It doesn't have a preponderance of small dense LDL particles. They have big, fluffy particles. Is it? Is it because I think big, fluffy particles are benign and small, dense are. I don't, as much as I think it's just a reflection of your metabolic state, it's the reflection of the city, your biological city on the inside. And you're also describing a research situation where we're extrapolating across the entire population from a narrow slice of the population. That's right. That's right. And that ain't that's not good research. Yes. So when you talk about something like smoking, do I think it matters what the reason is for why you inhaled three packs a day of smoke, as to whether or not it's gonna make your health worse? I can't imagine the reason why. No, it would matter. That's what gets to causality is if something is causal, then it doesn't matter what the reason is for the higher levels. I do believe if you have three packs a day of smoke going through your lungs, it doesn't matter if it's because secondhand smoke, it doesn't matter if it's because you work in a coal mine. It doesn't matter if it's your own cigarettes filtered or unfiltered, it's causal. That's my expectation, given what I've seen. Now, to be fair, it's still Bradford Hill. There's not an RCT, yada, but do I feel there's the evidence meets with my expectation? Sure. But now let's take it to LDL. If there's a reason for high LDL that would not associate with greater development of atherosclerosis, that gets to the heart of independent causality, so to, to what Philip was talking about earlier. Are we saying our study, if it does indeed show that there's not a high LDL association with plaque progression or vice versa, that we believe we knock out the entire lipid hypothesis? We're not making that claim. We're saying we're looking at this particular group and we're trying to fi, we're trying to answer that question with this particular group. But that said, if you expect the lipid hypothesis to apply to everybody in the same way that you would expect smoking to apply to everyone, it should show up. In our study, we should see people with sky high levels of LDL, those six people I mentioned who have an LDL above 400. They should show a higher rate of progression in plaque than everybody else in the cohort or for that matter in the general population. And so do they do. You may not know the answer to this one. If our study showed that Philip does Well, one of the, one of the questions that occurs to me Yeah. Is this lead mass hyper responder. Has that been identified relatively recent, recently as a phenotype or has that particular phenotype been around or known to be around for a long time? The first time I ever ran into it was after we had Nick Noritz on the show. And that's just been in the last three years, I would guess. Nope. You ran into it earlier. Oh, yeah. When we had Dave on the show the first time. Okay. But yeah, this is but it's been recently, this is something that Dave you know, coined, right? Dave correct. Kind of came up with this term. You know what's interesting is, Dave mentioned earlier right? The atherogenic dyslipidemia, which is basically the opposite of. Lean mass, hyper responder, essentially. And, that's been recognized for a long time. Now, the interesting thing in clinical practice, right? What I see, in day-to-day practice is the atherogenic part of that atherogenic dyslipidemia that's been so well known and described, gets forgotten about, right? Lipidemia, dyslipidemia is dyslipidemia, right? And we forget about it, it had this atherogenic part attached to it that turned out to probably be pretty important. The lean mass hyper responder phenotype that Dave described and coined the term for, quite frankly, with rare exception, you can go back in the medical literature, you can kind of find some things that looked at it in the past, but it had gotten forgotten about, right? It was just assumed basically that dyslipidemia was dyslipidemia. We don't really need to pay attention to this atherogenic part of it that's sort of attached to it. And that's, that's what medical practice has been now for the past 30 plus years until, Dave Pain in the Ass came around, asking some of these questions that as we alluded to, it turns out that most lipidologists who claim to, this is the everything they think about their whole career. Don't really want that question asked and may not be particularly interested in the answer. Which is a pretty interesting thing to see in real time. Let's get, why don't we talk about what has been released from the study results. Kind of what is known out there. People may not be aware of it, but it's now out there and then, I think that will lead into maybe what is so interesting about, and what gets revealed more in the movie. So there have been two significant, two major publications now from the trial. The first one looked at. Like you said, the initial expectation is we have people who middle age, important to say, right? The average age was somewhere in the mid fifties had been on low carb ketogenic diets for at least two years was the criteria. But on average it was closer to five years, I believe. And had documented, this elevated LDL and the low h the low triglycerides with the high HDL again for a number of years. And so by the diet heart hypothesis the standard lipid hypothesis, we would expect these people to have a fair amount of plaque in their arteries, a fair amount of atherosclerosis. So the first publication that came outta the trial was looking at their baseline level of plaque. You found a group to compare it against that wasn't on this diet and didn't have the high LDL, but was matched for other criteria. So what did we learn from that to start with? Yeah, so this was the match analysis. We published it in 2024 and what was neat was we had this other cohort, the only other one I'm aware of thank you, principal Investigator Karam Nasser called Miami Heart. And Miami Heart had a total of 2,400 in its population pool. It was cross-sectional. So unlike us, they were only getting baseline scans and then doing follow up but not follow up scans. Just actually following the population. I think they followed'em for eight years or something like that. And what's neat though is in that population 2,400, there was enough people that we were able to find, or rather I should say Quist was able to find 80 that could be matched with 80 of our a hundred in our cohort for a nice match analysis. So while we don't have a control group, this is what's called a match control. So we grabbed the 80, or rather they grabbed the 80 that were within their age range of Miami Heart took that 80 and then went and found on the other side an 80 that matched in ethnicity, age, sex the like biomarkers such as A1C fasting glucose. C-reactive protein was, it was actually an extremely close match between the two. And when they did, when they had this match, it was really tight. One very exciting thing was that the, there was no statistically significant difference in total plaque score between both groups. This group over here, that was ours, had an average LDL of 272. In that 80, the Miami Heart Group had an average LDL of 1 23, so less than half, right? But our group had an average time on the diet of, I wanna say 4.7 years for that 80, right? And then it makes it even more interesting. Ours trended better, even though it was not a statistically significant difference. There was less plaque in our group, in our cohort compared to the Miami heart matched group. So both groups, mi metabolically healthy, matched across all of these different demographic and metabolic metrics. But. Our group actually had less plaque based on the based on the total plaque score. That was pretty exciting. Even though the LDL was double. More than double or double. Yes. Over double. Now the second. So good. Yeah, no. So then I was gonna say, okay, we fast forward a year and the movie goes into some of the interesting things oh, by the way, COVID happened during that year. So getting people traveling and, coming back for follow-up scans and all of that became a little bit more complicated. But, you go a year. All a hundred people, right? If I'm correct, came for their follow-up scan. Which is pretty extraordinary to have a hundred percent retention in a study like this. But all a hundred people come back, they get their follow-up scan and we actually fast forward about two years, I guess, and now we have the paper that comes out, the second paper looking at what happened to plaque levels during that year. Yeah. So importantly that first analysis, it's called a semi-quantitative analysis, and that's where cardiac readers actually are going through with software. It doesn't actually take that long. It can be done pretty quickly. I lovingly call it the horseshoes and hand grenades version of the scan. The second analysis was done by a company called Clearly, and that's an AI guided reading analysis. So it goes through an algorithm and then they, through their algorithm, come back with data and then we get that data. Now about a year and a half ago, as of right now, about a year and a half ago, I saw the first the first preview of the clearly data behind the scenes. And I've since been able to be public about it up until this point after we had published. But I'd said then, now this is exciting that two of the major findings we're interested in are that with both the semi-quantitative and clearly baseline plaque predicts future plaque progression. That's builds on existing research that says that, and also that A POB does not have any association with plaque progression in both those data sets. But there is a third thing, which is that actually the semi-quantitative and the clearly dataset were somewhat different in the magnitude of plaque change. And what I was seeing, me personally, what I was seeing is I was like, actually, this doesn't seem to make sense to me. This, the clearly data set seems too distinctively different from what my expectations were. And so I was kind of pushing it back a little bit to the research team and said, I'm not sure if this actually completely makes sense to me. There's this other factor that there's only progression in the clearly data set, and that doesn't make sense to me. I would expect noise below the noise floor. Things along those lines. But there was an issue, which is that I am also the funder, as I just mentioned. I created the Citizen Science Foundation, so forth. So I'm the funder and I'm not an expert. I'm not an imaging expert, and I'm not a statistician. So all of those, plus my wearing the funder hat I couldn't say, Hey, keep reworking the data until I like it. I appropriately said I guess we need to get to where it's published. But I knew once it would get published, I would get the raw, anonymized data. So fast forward to April 7th, the team, the research team is thinking this is great though. We've got two major findings. The plaque be getting plaque, the a OB, not associating with it. That's what everyone's gonna be focusing on. And I was a little skeptical. I thought maybe the internet might be focusing more. On the clearly data in particular, spoiler alert, turned out that I was right, that actually the internet cared a lot more, I should say more of our critics did. I vaguely remember this now. Yeah. And yeah. And Dave, let me just jump in and unpack this a little bit.'Cause this is complex to understand. So when the study was originally designed right, you the plan was to do what's called the semi-quantitative analysis. And I've heard you describe this and I'm gonna try and describe it and jump in if I get anything wrong, right? But when we get these images CT angiogram images, right? They come at, they come to me, the clinician, and it's a bunch of pictures and typically what's called a radiologist who's a, specialist in reading x-rays will look at it and they need to translate the pictures into usable information for the doctor, right? And what the doctor cares about is roughly how much plaque is there and how narrow the blood vessels are. So one way to do this is with the semi-quantitative analysis we have picked. In advance of doing a study. These are the key areas in the blood vessels that we wanna look at. And we're going to grade each one of these areas from no plaque to zero to four, essentially, right? Four being there's a lot of plaque, zero being, there's neuro plaque, and then these levels in between. And then we're gonna add up all of these areas that we have, and that's gonna give us a score, right? And if someone has had two scans, we can look at the score, you're, on one scan and we can look at the score and the next scan and we can figure out did their plaque go up, down or stay the same. Clearly came along and did, is able to because they were able to leverage AI and, better computing power and everything that happened technologically over the past few years. And they say we can do a much more detailed analysis. And when people get a clearly analysis of their CT angiogram, we get a report that says there's this many cubic millimeters of plaque, and it gets broken down into how many different types of plaque there is. And then we can go through the blood vessel, like millimeter by millimeter, and we can see how narrowed it is at each point. And we can report, what is the most narrowed point of the blood vessel. So it's a much more detailed analysis. And if I understand everything correctly, it wasn't initially intended to be the analysis that was going to be used in the study, but because it became sort of a dominant technology while you were doing the study, it made sense to try and have this analysis of the study. Yeah, it was a why not, right? It, it could be accomplished faster than the pre-specified analysis that we need to do anyway, which was likely going to be the third analysis. At this point in time. But this is the key thing to bear in mind, Jack. And it's something I say over and over again. The scans are the scans. The CT angiograms are the source of truth. So the semi-quantitative E was just talking about with scores. That's n analysis. One analysis of those same CT angiograms clearly is an another analysis of the ct. So it's almost like you have multiple cameras looking at the same scene. And they're making an interpretation for us. So that's why I was going, wait a sec. I was, we shot the whole movie against the semi-quantitative data, fully confident that the quant, any quantitative analysis was gonna look close. To the semi-quantitative data and Philip, you probably more than most people could speak to this, you probably have the same expectation. It's not that they're not gonna be identical it's not that's an expectation, but you do expect, if you were to do, let's say a semi-quantitative read yourself right where you went through all and you scored it all, you don't expect a clearly analysis to come back radically different where all of a sudden there's just all kinds of plaque or something like that. And so that's why I was surprised. I, and again, it'd be different if it was an outlier here and there, but it was population level and that's what was strange. So the entire population was radically different. It, yeah, it was. And again, to be sure the team was a lot more excited about the two major findings for which both data sets agreed with each other. The plaque predicting future plaque progression, the apo be not predicting it. But I was very curious and I wanted to get the raw data, but I couldn't get the raw data until we were published. So we get published on April 7th of last year, and then 11 days later, April 18th is when I get the raw anonymized data. And that's when it gets really interesting because I found within 24 hours, I found I had I have a local LLM, so I had a secure way to work with AI to, to analyze the data. And very rapidly, I was like whoa. There are a bunch of different issues with this data set. And I immediately alerted Lundquist and I immediately alerted clearly and said, guys, there's, and I presented too clearly. I said, there's these issues. And I made a report of 24 of the different problems that were in there. I already mentioned one of them, which is that there's no noise below the noise floor. The fact that there's no regressors. It is just super strange because it's, it not impossible, but highly unlikely. It's nearly impossible. Yeah. In a cohort of 100. Like it, it's here's an analogy I use. You're familiar with a bathroom scale? Yes. So if you step on and off the bathroom scale 10 times, is it gonna show you exactly the same weight every time? Should. Okay. It should, if it's really accurate and does not let you, does not let you see where its margin of error is. But a lot of scales will show you the margin of error of where it's at, such that when you step on and off it, it can give you a different number each time by a little bit. So that's the noise. That's the error. Yeah. Yeah, that's in it. Okay. So if you stepped onto it with a grain of sand in your hand, you're objectively heavier. But you may not have the expectation that it's gonna pick up that grain of sand. You are absolutely heavier for sure. Yeah. But just by one grain of sand. So the resolution of the bathroom scale can't catch that grain of sand. Yeah. So for people at very low levels of plaque, and if you see these CT angiograms, you'll see what I'm talking about when you're trying to move through it, picking up something like a tiny pixel artifact and accurately determining whether or not that's plaque is it's variable. It's hard, it's just hard at those very minute resolutions. So that's why I expect resol regressors that some scans will show less plaque if the plaque levels are so low that it's hard to measure because there should be what's known as bidirectional scatter. It's a wobble. Yeah. Makes sense. Makes sense. Yeah. Okay. That's why. And the problem is the lipid hypothesis is so strong that people were like the people behind the scenes, the research team were like, oh, but this is an untreated population. So it's not surprising that everybody would increase in plaque. And I, as the engineer would be going, okay, but that still doesn't explain measurement error. You still expect measurement. You s you, you could believe everybody was increasing in plaque. It still doesn't change that. We should still have that lack of resolution capability. Anyway, that was an example of many of the different, this I'm just going to connect what I think are dots that connect, there's something squirrelly going on behind the scenes with clearly measurement or at least their analysis of their measurement. Is that where we're going? Let me tell you the next part of the story because if that's the case. Then this seems pretty simple, which is I alerted them to it. I said let's just rerun it. Okay. Rerun it. Sure. I mean, it's just, yeah, just take it through what's known as a quality control check. I just, I really wanna be sure it's fully blinded because the fact that they're all positive when it's supposed to be unknown, what the pairings are, I think that's kind of something I just want to rule out. I just wanna be sure that there's no relevance to the AI picking up the metadata or anything like that. And although it's supposed to be fully blinded, I didn't know for sure if it was, so I said let's do this fully blinded now. The lawyers will tell me that, what I'm supposed to say here is it's my opinion. It's my opinion. Given the communications that I had an understanding, they were in fact going to do that. They were going to rerun the blinded analysis, but then things shifted per my understanding of events. And they said, actually no we're not. We're going to stand by the original analysis and we're not gonna do a quality control check. And I said that, but wait a sec. Why wouldn't you? I'm showing you this report. I'm showing you all of these issues with this analysis. I mean, more than what I just mentioned. For example, you're familiar with the power of zero. If you have a CAC of zero, your likelihood of plaque progression or even just events is much, much lower in the clearly analysis of our scans, those people with a CAC of zero actually had the lowest progression relative to the sorry. Those people with a CAC of zero had three and a half times more plaque progression than those people who had a positive. Calcium score. Let me, lemme repeat that for emphasis. Yeah. If you had a CCC of zero, I think I understood what you just said, but you had greater soft plaque progression than those who had a positive CAC. Again, I feel like Philip knows exactly why this stuff is like alarm a little, but I understand. Tells me that's doesn't make any sense. Yeah, it doesn't make any sense. And again, today's credit, no one at the time was really asking these questions. I'm unaware of any of the, and understand So the data gets released. Dave said, the conclusions of the paper were, that the more plaque you had to start with, the greater the risk that your plaque progressed over time, and that what your LDL or a p oob level was, had no relation to whether or not your plaque progressed over time, which are two pretty interesting conclusions. The first one, we kind of had a knew about, right? But this was the first time the second one was shown. What happened in the social media world? Was everyone zeroed in on the people that had plaque progression and said, look at this group. They have more plaque progression than we expect. There were some, I would say in there were some comparisons to some historical data that may not have been really valid because it wasn't really the same data that we were looking at. But anyway, everyone zeroed in on. Look exactly what we said. Your LDL is high and the group LDL is high and the group had more progression than we expect by the clearly analysis. Even though this other analysis. Didn't show the same amount of progression. But no one was really interested in asking that question. And Dave being the engineer and the computer scientist that he was, was really, able to go through this data and say, something's not quite right here. The follow up then became that, there were, so there were two competitors to clearly, essentially that, do similar type analyses and this, the data was able to be run. The studies were able to be run through both of those analyses as well. So now we had four analyses and the other two agreed more with the semi-quantitative data than they did with the clearly data. But the clearly data was still the only published data. So that's what everyone was focusing on. This sounds would make a great movie plot. Yeah, it would be. It would be the, so yes. After clearly basically kept saying no, we went to a major competitor of theirs, which was HeartFlow. HeartFlow also does AI guided reading, and I made sure to do a speech in advance of getting heart flow and Q angio the upcoming ones before they were in hand. And I did this speech at PHC, which is a public health collaborative in London. And I said what those patterns were that I had concerns about with clearly, so that when I did have the new data sets in hand, either they matched those patterns or not. Once again, if I was post hoc saying, oh good, these, this is where it's different. And that's what I was expecting, that I would be a harsh judge of that. So sure enough, the heart flow data comes. It's all the stuff I just talked to you about. It had 33 regressors out of 94 that had valid scans. And they were predominantly at the lower levels of plaque. So some amount of them are just part of the noise. But then again, some amount of the progressors are part of the noise. But that's what I would expect. I would expect below a certain level, you'd get almost 50 50. For the same reason the bathroom scale can't pick up that extra grain of sand. That's what I would expect. But then on top of that same thing with the CAC, the people with the positive CAC had more progression. The people with the zero CAC had much less progression as expected, as matches all the data out there that supports this whole power of zero. So I bring this back to clearly, and I say given heart flows data, given the anomalous nature of the patterns, would you guys consider. They said no, we moved on to Q Angio. The fourth and final analysis that Philip mentioned, same thing. The plaque levels matched the average plaque levels matched heart flow. They matched the total plaque score of the semi-quantitative. So you have these four analyses, but one of them is sticking out like a sore thumb. It's two and a half or three and a half times the average of the other three that all agree with each other. And I still couldn't talk them do into doing a blinded reanalysis at that point. But one last one, I'm baffled here. I know it's the one last one. It's the most recent news that I broke, which is that we also had a handful, right? More than a handful. We had eight participants request their study, CT angiograms took it to their cardiologist. Resubmitted them through their cardiologist back to clearly. So through a valid, clearly account of their cardiologist, they effectively made a way of getting their scans to basically do a blinded reanalysis. Now, what's your guess, do you think that it would turn out to be the same as the data that clearly provided us? My, the, my brain that has gone through way too much of this is thinking completely different results in this completely different results? Yes. Which implies nasty things. Of those eight, half of them had less plaque on their follow-up scans. So four of them became regressors. Let me repeat once more that the clearly dataset they gave us for the study. All of them, at least 99 of the a hundred.'cause one had to get removed for procedure between 99 of the a hundred had no regressors. There were no regressors In the 99 they gave us, of the eight samples, so far, half of them. So that's already not just half, but it's also four more than what we got in the original data set that was provided us. Needless to say we went to the journal and we said we can no longer support the paper that we published. We the authors published, because it relies on this clearly data set and did the right thing. We asked for it to be withdrawn, which we knew was likely gonna lead to retraction. And even though retractions often count against the researchers, that's the data that we had. So that was just the right thing to do. Incidentally, was there an objection by the journal, given the information we provided, there was not, was there an objection by clearly themselves as to our retracting a paper based on their data? The answer to that question is also, no, there was no objection. And there's, as far as I know, there's been no objection by any of the critics either. It's a real life. Emperor's new close. Everyone can see the dataset. Everyone has access to ai, so they can take the dataset and throw it into AI and ask these same questions. But yeah, again, to your credit, Dave, right? You said, here's the data. Right after the initial publication, the data became public, got released to the Citizen Science Foundation and the Citizen Science Foundation put it out there. Said here the link. Do whatever analysis you want, use all of your tools, please. Show us where, our thinking is wrong. And to my knowledge, no one really was able to show, where your thinking was wrong. And yet we're in this weird situation now where, people still wanna hold onto the con, what they drew from the original paper of, oh, but look how much their plaque progressed. And this story is still being told. Obviously there's lots more to come, but needless to say this is all captured in the documentary and I think that's one of the reasons that people are gonna be so interested in seeing it. And the other thing and I haven't seen the full documentary yet. I'm waiting till next Monday when I get to see it with Dave and with everyone that's coming to St. Pete. But what I have seen of it, some clips in the trailer and all that you go into what's really important, right? Because we gotta keep in mind these scans are real people. Yes. These are people who are trying to figure out, do I need to be worried about heart disease? I've gotten all of these great effects from going on this diet. And some of them just have absolutely amazing stories to tell, and I know a lot of them personally. And, all of these great effects, all of these things changed in their health. But I've been having this constant messaging from my doctors, from my family that, this is dangerous because my cholesterol has gone up. And is it really dangerous? This is a real answer, a real question that people need the answers to. And I think that's what, really makes all of this so fascinating in the movie. Goes into, I think it, I don't think I've ever seen Jack's face like I have this time around, When I found out, okay, we're having Dave on. He is got a new movie. It's called The Cholesterol Code. I will admit I instantly just assumed. This is basically gonna be laying out what we've been talking about with carbohydrates and saturated fats and statins for the last, the entire life of this show and years before that. And that's lovely. I'm sure there's lots of people who need to, I, it sounds like we're dealing with a crime mystery. It sounds like there's it sounds like there's an antagonist who does not want the truth to come out. That's what it sounds like to me. It sounds like there may actually be commercial interests who really want to torpedo the scientific findings. I'm guessing I haven't seen it. I don't know. I am way more interested now. The, so at least. Before Dave even starts to answer this actually leads me to a question that I've never asked Dave, but I've been curious about. Your original website, right? It probably goes back a decade or more now, right? Called the cholesterol code, right? That's where the name of the movie comes from. Did you name that after the Da Vinci Code, or was that coincidental? No. Now, really got me thinking like what this sort of meant to be. But yeah, that was just an interesting side question that came to mind. When I was actually, I was in Italy a couple of weeks ago and, da Vinci and all of that, and I was thinking, did Dave know? No, I I write code. That's what I do normally for a living. Before I got into this. Wild business. I, it's sort of funny because coming from engineering, I mean, I've seen whole industries, come up and go down overnight. And while it can be very cutthroat at the same time I often tout how much more transparent the world, especially of software engineering is. There's more open source, there's more benchmark testing. There's more, and there's a lot of hard scrutiny for, proprietary software. And it's a lot more intense, I'll just say. Whereas in the world of research I am, I'm genuinely awestruck at how much the, there's a bit more of an interconnectivity of the system and a degree with which everybody's friendly with each other. I think too much. Because I, I think that science should have a healthy level of antagonism and things like discovery are themselves disruptive. So if there's a, if there's a paradigm built up around a given thought, and if you have a bunch of like-minded people in the field and they have the capability to keep out the new ideas, that's a problem.'cause science is supposed to be truth seeking. So if there's something that's not quite the truth, you should be all the more interested in finding it out to correct the problem. So John, to your point from earlier, I, the Citizen Science Foundation has some people who are paid a lot of money by the hour. Thank you. To help from donors who are very interested in us. Being represented properly. And they'll, they're very there's been several meanings for what are the things I can say, the things I can't say. So my broader answer is, we don't know what happened with that dataset, but am I comfortable saying things like it's absolutely anomalous. Am I comfortable pointing out facts like both Philip and I were describing, which is that literally nobody's come to the defense of this thing, but yet at the same time, is there a parade, bringing around oh, all of this content that was made on top of the clearly data let's, rewrite that, or something like that. No, that's what's, that's what's also, I would argue kind of the silver lining to all of this is it kind of created a natural social experiment. So how many people really did. Want this research to fail and got a glimpse of what that might look like. And then shifted gears. Because before if you were watching social media before that April 7th paper, a lot of people were saying, actually, imaging isn't gonna give us enough information. It's gonna take decades of time where we could see any plaque change. The April 7th paper comes out and all of a sudden a whole bunch of people are like oh, actually imaging is great. In fact, here's a whole bunch of other story studies on imaging and how this one compares to those and so forth. As the clearly as the dataset it, as it appeared more and more to more people, that its credibility was getting lost. The chatter around it was dissipating to the point where I pretty much, after Q angio, it just disappeared altogether. I mean, there's quite literally all the people who couldn't talk enough about it in the April 7th may period of time. I, I still think even in the wake of the of the retraction, which you would think would get a lot of chatter, you would assume there, there's still a lack of interest in talking about the plausibility of the clearly dataset. And I've invited it. I've actually asked people if they'd like to come. I've offered many of the major content creators to come and sit with me for the Feldman protocol where we could talk about the plausibility of the clearly data set. Not one single person has taken me up on that offer. Yeah, you may not be able to speculate as to the why for various reasons, but Phil, can you speculate why? I could speculate why, and it will maybe get me into less trouble than Dave. I mean, it does strike me. If I'm being charitable right, I can say, okay, there was something that happened with the clearly algorithm, okay? In between the time that the first scans and the second scans got done that may have led to this anomalous data and clearly doesn't want to do a re-analysis and show themselves to be wrong because it's going to perk the trust in clearly as a company, right? And okay might be bad for business. So I guess that would be my most charitable explanation of what may have happened here. I can also look at, the environment around this study. And, an important fact that Dave maybe glanced over a little bit in this discussion but he's discussed a little bit more, is around the blinding, which may not have been done exactly correctly. And again, not blaming anyone, not saying that was an intentional thing, but just 'cause of the way that this data gets transmitted and handled and what's within the data that process may not have been done right. And that's a very important process to data integrity, and to research integrity. And it looks like there were some things that didn't get handled right there. I guess that's where that's what that implies. I don't know, Didley about research. And that implies to me that there was a serious problem with the blinding it. Sorry. And Phil Philip correctly pointed out, I didn't actually include that as part of the story, but that did come out during the summer of last year. It was revealed that the data set that clearly got was not blinded. And to be fair to clearly it actually was Dr. Budoff who conceded this was a mistake that happened with their lab. But it's all the more reason why, regardless of what the reason was, it's all the more reason why I would want even more for it to be a fully blinded analysis. It's also worth mentioning this one other piece, which is that we offered to pay for in case cost was a consideration in case that was what was relevant. Add a donor on hand, ready to cover any costs involved. So it was such a difference of experience when we engaged heart. Because when I engaged HeartFlow, it was after we'd been getting a no over and over again from clearly to redo this. And I said, before starting anything with HeartFlow, I said, now wait a sec, before we engage you, we want to never again run into this issue. Would there be any problem if when you run your analysis, we request further reanalysis so that we confirm your work? It doesn't mean that we expect radically different results, just that you could do a blinded reanalysis as needed. And they said of course for them it was like yeah, we sell measurements, we sell accuracy. So test retests are just like standard. And they said, you can run as many of these as you want, as long as you pay us for it. And I said, okay, that's great. That's all we've ever wanted. And there, there may be an even bigger story here, right? Beyond, lean, mass, hyper responders, cholesterol, all of that. The whole process that Dave has now gone through around this, and the integrity of the data. And again, like Dave mentioned this is not the standard in medical research. I think about, I've thought even more about since all of this has occurred, right? The fact that the original data that led to the approval of statins, for instance has never been made public, has never been made available for a independent third party to verify the findings of the study. It sounds insane, right? Jack's laughing. But it's the reality, right? And why isn't this the standard? And those questions that you know, have now been raised around this study really bring us greater questions about all of the medical literature that's out there, right? And all of the studies, and we're in this, we're again, we've been pushed as physicians now, right? Towards trust the science evidence-based medicine. And, we have fundamental questions that we should be asking of, can we trust this evidence? Can we trust any of this science? I think that's an even bigger question that's going to, they, this whole study saga is part of that. And I hope that more of my colleagues start asking those types of questions. I'll be honest. I'll be honest that my own You like that? Be honest. My, my own trust levels have been modified being up close and personal to this. Here's the thing, I think errors are human. I think we should expect people to make mistakes and if mistakes are made, that's what happens. I'm a lot more ornery when it's a lack of course correction, but I'm especially a lot more mindful of how these things are treated. So the me of before this research, if you were to pose all of the same things back to me up until the point of the request, if you were to say, Hey, what would happen if you got a data set that turned out not only to seem anomalous, but on closer examination really had a lot of signs that would draw you to believe it was, and then you went back to who provided the data set and said, Hey. Can you do a quality control check and just be sure that it's, fully blinded. I, I would've bet everything in my bank account. Of course. I mean, and that just, that makes total sense. But that's me thinking in terms of what I would expect of the field of science and of their interest in maintaining things like the apparatus of the measurement for the study to come back and to feel that level of confidence in the integrity of it. And that gets back to the lack of course correction and how it's getting treated right. To, to this day. Like I said there's been no objection to our retracting the paper based on that dataset, but I don't know if there's been, I don't think there's been a single public statement made or anything along those lines, and I'm not sure if there will be. Okay. I'm not gonna ask for the punchline, but I am going to ask, is this. Is there more information in the movie? You're gonna, you're gonna see it anyway, right? I mean, we shouldn't spoil the movie. You're gonna see it in a week. Yeah. So you'll we'll talk about it afterward and then see what you think. This is fascinating. I just this morning I was listening to a podcast Brett Weinstein's Dark Horse podcast. I've been following him for probably 10 years. He's one of my favorite theoretical evolutionary biologists. Possibly the only theoretical, evolutionary by, I was gonna say Not a big field, but, But he talks about the a study he attempted to publish back in 2000 that ran into what appears to be. Not very good scientific reasons for not being published. And in that study that did eventually get published and was just recently validated with other research, he postulated that part of the reason that pharmacology phar pharmaceutical companies are reporting no adverse outcomes from the use of their drugs is because the mice that they're testing them on have been evolutionarily evolved to not have any problems. And he goes into the entire biology of it all. In other words it was not in their financial interest to find problems and the mice that they tested them on. Don't actually occur in the wild. They are lab, bread, mice. It's a wild story. For my listeners, do a re do a search for all our mice are broken, Brett Weinstein, and you'll get the details. I'm hearing echoes of the same thing you hear, Dave. Yeah, and otherwise I'm a conspiracy nut. I have my tinfoil hat firmly placed on my head. I know that, but come on people. Okay, I'm gonna end my rant here. I mean, it does, bring me back. Gary Todds was a major reason, that I went on low carb and I think a lot of people in the low carb space right. Owe a lot to Gary. And, Gary started with exactly this question, bad science versus good science, and it led 'em to, oh, nutrition happens to be pretty bad science most of the time. And oh, here we are, right? So it is just very interesting and if this all ends up bringing to light how bad so much of the science that we have really is I think that will be a kind of fascinating outcome from all of this. I can't wait to see the movie again. Cholesterol code. The movie is the name of the movie, the website you can go to. We're doing a screening in St. Pete. There are screenings happening all across the country. And and then it's gonna be released on Amazon pretty soon as well. Dave, any other links or any other places you wanna direct people who are interested in it? I think actually, I think it is important for us to mention you are actively working on expanding the study and extending the study and there are some ways that people can support that as well. So we should probably mention that. Yeah. Thanks for the opening for the plug. If you want to help out the science, you can go to citizen science foundation.org and donate directly. As I mentioned earlier in the podcast, for real, we have a 0% admin overhead. We the only charge through third party like credit card processing and donor box. It's quite clearly, I think almost all of it. And the two things are that we're getting an extension to the current keto CTA study, so that we'll have a third set of scans. A number of participants have already shown their willingness and interest in being able to do this. That would probably start later this year. And then the other thing is that we have a companion study that will have a built-in control group. The companion study we call the Triad Study that's in the works. Right now we have just a little more funding to complete. I wanna say we're three fourths of the way there. And so if you can contribute, that's great. We would get there for both of these here pretty soon. And yeah, thank you guys both again for all the support. Wow. Our guest has been the original citizen scientist himself, Dave Feldman. All the contact information will be available here in the show notes. The movie is the cholesterol code. Thanks for joining us folks, and spread This one around The world needs to know what's going on. This is literally a matter of life and death. Phil, wanna say anything else before we say goodbye? No, I think that's all. We're good folks. Thanks for joining us. We'll talk to y'all next time.