October 8, 2018


Transcript — The Drive — Dave Feldman

Transcript of episode 19 of The Drive with Peter Attia, released October 8, 2018.

Read Time 129 minutes

Note: This is the transcript of episode 19 of The Drive with Peter Attia, released October 8, 2018.

Included in this transcript are comments from Tom Dayspring, M.D.

Please visit this page to access the show notes from this episode.


[5:00] (Intro notes)

Ultimately Peter found Dave’s model unconvincing for three reasons, and provided them in the intro of the podcast:

  1. “Dave was unable to explain the mass balance, meaning how does one account for the greater amount of cholesterol in, and the greater number of, LDL particles. No one, including Dave, is disputing that the phenotype of interest has more LDL-C and more LDL-P. There are only 3 ways this can happen (these are [collectively exhaustive, but not [mutually exclusive]): make more, clear less, transfer from other pools that we can’t see (e.g., cell membranes). I think the data make the first of these by far the most likely driver, but Dave seemed unable to address this and could not explain, to me at least, what could account for this increase in LDL-P/C. So on first principles, my doubt of this model has gone up from the start of this discussion, as the person who developed the model could not actually explain the mass balance. This is one of the most fundamental requirements of any model. And to be clear, even if this fundamental condition were met, it would not be sufficient to make the case that [lean mass hyper-responders or LMHRs] are not at risk. It’s a [necessary but not sufficient] criteria that, in my mind, has failed.”
  2. “Dave argues that VLDL production is driving the LDL concentration, but the fact remains that in insulin-sensitive people (which presumably the LMHRs are), the opposite is true: there are fewer, not more, TG being exported from the liver and there is less, not more, apoC-III on the VLDL, thereby reducing, not increasing, their residence time. In other words, LMHR would have less VLDL to LDL conversion than, say, someone with T2D. So again, I can see no evidence whatsoever that his energy model, which can’t be explained on mass balance, and can’t be explained on what is known about the physiology of VLDL and LDL, is plausible.”
  3. “Even if you ignore the points above—which you can’t—I am more unconvinced than ever at the notion that we should exclude the roughly 2,000 genetic mutations known to produce a phenotype of high LDL-C, high HDL-C, and low TG. We have 2,000 natural experiments. Surely at least some of these cases (e.g., PCSK9 gain of function) are excellent proxies for the key features of LMHR. Yet to ignore them for imaginary reasons (e.g., having gain of function PCSK9 is somehow toxic to endothelial cells because it impairs their ability to take up cholesterol despite there being no evidence that endothelial cells require PCSK9 to uptake LDL in a receptor-mediated fashion) is to say, in my opinion, one does not want to know the answer to this question.”

Peter emphasizes that probabilities play a very important role in diseases like atherosclerosis, and this nuance is often missing when discussing this disease.

Atherosclerosis is impacted by many things beyond the lipoproteins, but that doesn’t diminish their role in the causality of atherosclerosis, Peter argues.

Note from Peter: Some low carb enthusiasts argue that as long as they are insulin sensitive, have high HDL-C and low TG, their LDL-C (or LDL-P or apoB) is irrelevant. Further, many confuse imaging tests like calcium scans (CACs) as biomarkers and argue that as long as CAC = 0, there is no need to treat, despite the risk predicted by biomarkers. If you are confused by all of the noise on this topic, consider this example: A biomarker like LDL-P or apoB is predictive. It’s like saying you live in a neighborhood with a lot of break-ins. A CAC is a backward-looking assessment of damage that has already taken place. So it’s more like an investigation into a break-in that already happened. In my opinion, waiting until there is grossly visible (i.e., no longer just microscopic) evidence of disease in the artery to decide to treat for risk already predicted by biomarkers is like saying you won’t get a lock on your door—even if you live in a high-risk neighborhood—until you’ve suffered a break-in. This is bad risk management. As the saying goes, “When did Noah build the ark?”

Ultimately, it’s up to the individual, who’s LDL-P and LDL-C are very high while consuming a low-carb high-fat (LCHF) or ketogenic diet (KD), to make a decision: the hope is that the following discussion (and related references and material in the show notes) can help people think through the issues and make a more informed decision.


Peter Attia: Hey, Dave. Welcome to San Diego.

Dave Feldman: Thanks for having me, Peter.

Peter Attia: You’re here for a conference, is that correct?

Dave Feldman: That’s right, Low Carb USA.

Peter Attia: Got it. That would explain all of the low-carb folks that seem to be in town this weekend.

Dave Feldman: How many are going to end up showing up here?

Peter Attia: None. I’m actually leaving very early tomorrow to go to New York-

Dave Feldman: Wow.

Peter Attia: … so I will be missing this.

Dave Feldman: I’m either deeply honored or very scared right now.

Peter Attia: No, I actually was going to try to talk to one other person, too, but they bailed on me and took a better offer, so … Anyway, this is going to be a bit of a different episode, Dave, in the sense that I think this will be, even by the standards of this podcast, perhaps a bit more technical at times. I think this will probably be a podcast where … You and I have already spoken offline. I suspect a lot of what we discussed will need to be included in show notes because there’s just going to be so much visual stuff.

There’s so much data we’re going to be talking about, and some of it’s just quite graphical in nature, which is not to be confused with graphic. So, we will apologize in advance that this might be one of those shows when you’re going to probably get the most out of it sitting in front of your computer, but nevertheless, hopefully, we can certainly get some interesting stuff out of the way.

I’ve been trying to think about how to set the stage for this, because I think it’s safe to say many people listening to this don’t know who you are and don’t know what we’re going to be talking about or why we’re even having this discussion. So, I’m going to take a small liberty of trying to synthesize some of what I’ve heard you say in the past, but then turn it over to you to clarify it and kind of put it in context.

From my standpoint, I think you’re one of the more thoughtful people on what I would call the LDL-is-not-necessarily-causal-in-heart-disease camp. Certainly, there’s a number of people out there for various different reasons who have argued that the causality of low-density lipoprotein and atherosclerosis is not a foregone conclusion, and, in fact, there may be a subset of people in whom it’s not quite relevant.

What you and I have done over the years is had email exchanges and things like that, and you’ve been very curious. You’ve done a lot of self-experimentation, which you won’t find a more sympathetic audience for self-experimentation than me. I think, in large part, we want to kind of explore some of the deep lipidology around these ideas, but ultimately, it comes back to a question.

This is, I think, the question that, at least, if I’m going to be selfish, is the question I care about, which is today, I have to make a decision. I mean that literally, so meaning at 5:00 this morning, I had a call with a patient. Luckily, he was in a different time zone, but we had to make a decision about his lipids. I will have three more interactions with patients today, of which two will center around the same discussion. So, ultimately, decisions have to be made about how to manage dyslipidemia, and most decisions have to be made with incomplete information. I certainly don’t have any expectation that we will emerge from this discussion knowing an answer, but nevertheless, hopefully, we’ll have clarified a few things. Before I go any further, Dave, if someone were asking you, “What are you known for?” With respect to this. You can probably juggle five tennis balls simultaneously or something else, but with respect to this discussion, how does the low-carb community kind of describe you?


Dave Feldman: Well, many of them right now would describe me as a lipid expert and kind of one … The irony is that I actually kind of push off that reputation to some degree. I am not a formally trained biochemist. I’m not a medical professional. I regularly feel like I need to emphasize that. In fact, I think that your series, Straight dope on cholesterol, was probably a way for a lot of people, myself included, to kind of short-circuit a lot of the formal education that typically one would have to go through a university to get to really the general concepts of lipoproteins and how they work within the system.

What got me to this place, going a little bit backward, is that I went on a low-carb diet, and now this story’s fairly ubiquitous. You hear it a lot. A number of people, myself included, saw their cholesterol rise substantially. After that happened, yes, per what you just talked about, I started doing enormous amounts of self-experimentation, and I started elucidating a pattern. Part of what motivated me to do that was that even though I had very little training on the medical side, I did have a lot on the network side as a software engineer. I saw a pattern that looked very familiar to me. Without getting into a lot of geeky terminology, if there happen to be any software engineers listening, you’ll probably be familiar with this term. It’s called dependency injection.

It’s something that gets involved with networks of distributed objects. I saw that with lipoproteins, which you talked about at length in The Straight Dope on Cholesterol. If I can, by the way, interject just this one thing. Thank you for making that series. I know I’m not the only person listening right now who would say that it really was kind of a light during a very dark time. From that, kind of this whole journey began, and weirdly, I went from being a fairly well-paid software engineer to kind of a poor, n-of-1 scientist, definitely tackling just exactly how far I could take moving around my cholesterol.

I’m going to make an outrageous claim. I couldn’t do it this week because I have too many things going on, but I had always fantasized about the possibility of having a conversation with you, and giving you where I’m at on the research now, saying, “I would like you to write down a number between 100 and 350. Then, once you do, in about a week’s time, I’m going to move my LDL-C cholesterol to that number, plus or minus 20 milligrams per deciliter.”

I think that would have been a lot of fun. Maybe we’ll get a chance to do that in the future, but the bottom line is I feel as if I’ve come across enough at least with my own mapping of my own metabolism that I found how I can move LDL cholesterol and LDL particle count up and down without medication or supplements by finding what I believe to be the primary influencer, which is the energy metabolism, especially that of fatty acid utilization for energy.

Peter Attia: Okay. I think the other thing we’ll want to make sure listeners have done by this point if they want to get really deep on the understanding of this is probably going back and listen to at least one but potentially two or three of the other podcasts you’ve been on. You’ve been interviewed a number of times. I’ve had the privilege of listening to several of them, which is what kind of helped me get more up to speed on some of your arguments.

I think rather than just spend an hour going over those again here, I’d rather we sort of get to it more quickly, which we will, and then let the listener go back and get that way of background. With that said, let’s talk a little bit about your story. Before you went on a low-carbohydrate diet, you probably had a standard lipid panel. It probably showed what?

Dave Feldman: It actually showed … I believe the very last one that I had was typical of what I’d usually had gotten, which is my total cholesterol, I believe, was 186. My LDL-C was 131. My HDL was 40, and my triglycerides were 80.

Peter Attia: Okay. Obviously, people who have heard me talk about this before will know that that tells us virtually nothing. Your total cholesterol is of no interest. Your LDL cholesterol at 131 by the Framingham puts you a little over the 50th percentile…

Note from Tom Dayspring: 66th percentile in MESA.

…Your HDL cholesterol of 40 puts you quite a bit below that, actually…

Note from Tom Dayspring: Not for a man.

Your trigs [note: trigs, triglycerides, and TGs are used interchangeably] of 80 in someone who’s Caucasian doesn’t give us a great insight

Note from Tom Dayspring: 25th percentile.

Gives us even less insight if you’re African-American. But your trig-to-HDL-C ratio of about two would be considered acceptable by most. Most people would even consider up to three acceptable. So, we would have no way of knowing from this what your LDL particle number would be, or your apoB, which would be better predictors of your risk than any of these numbers here. Nevertheless, this is what everybody gets. Right? This is sort of the standard test.


Dave Feldman: Yes. In fact, anybody who’s ever considering going on a low-carb diet, it’s one of the first things they jump on. As I say, do me a favor and take a particle count test before you start the diet. Because I think a lot of us would be very interested to know what particle counts are before people start the diets. It may actually hold keys to understanding what’s going on with what … I’m not sure if it was you or Tom Dayspring. I think one of the two of you first started using the term hyper-responder to elucidate those people who, going on a low-carb diet, see their cholesterol go high, not just their LDL-C but their particle count.

Peter Attia: Yeah. I think Tom would deserve the credit for that. In fact, it’ll probably come up many times throughout this discussion, and we’ll certainly link to it. Tom wrote a really fantastic piece on this in his Lipidaholics series. I think it was in 2013. Might have been 2014. It was following a number of cases that he and I had shared back and forth about this phenomenon. I would add something else, Dave, if you’re going to make a request that people draw the advanced lipid panels beforehand. The other thing that is essential is that they get a Sterol panel. That’s not to be confused with sterile like I-L. It’s Sterol, O-L.

The reason for that is that there are basically four things that are moving LDL particle number…

Note from Tom Dayspring: Only two: 1) Production of apoB particles which may relate to hepatic pools of TG and cholesterol, 2) Clearance of particles.

…When I sit down with patients and talk about this, we always start from this place, which is what moves the LDL-P? Well, three things that move it are generally cargo related, and one is generally clearance related. The two things that move it, the macro level, the cargo, is the amount of triglyceride it’s carrying and the amount of cholesterol it’s carrying, or to be more specific, cholesteryl ester.

To get a sense of what its triglyceride burden is, you can get a crude sense from looking at the serum triglyceride level, though there are no commercially available tests to my knowledge that actually measure the triglyceride content of an LDL particle. At the research level, that’s been done. I have some data on a self-experiment I did in 2012 where we tracked the movement of cholesteryl ester and triglyceride through all of my lipoproteins, including chylomicrons [also referred to as chylos, and CMs). If it becomes relevant, we’ll certainly go over that. I think you’ll find it super interesting.

Dave Feldman: Yeah, and I think it’s worth stratifying that just for a moment for the audience. What you’re referring to is if we do a blood test for triglycerides, that’s inventory of all lipoproteins, so we’re not actually gathering it on a per-lipoprotein basis. This is what you’re saying, is this test will help at least what we’re trying to elucidate is exactly what it is on total triglycerides.

Peter Attia: Yeah, the blood test tells us nothing about the triglyceride burden within the lipoprotein directly, but we know that that’s one of the cargos. As a general rule, the more we see the triglyceride go up, the more we know we need particles disproportionate to their cholesterol content to traffic them. The next two things we look at that are also, quote-unquote, cargo-related is the synthesis of cholesterol.

Turns out we can measure that quite well. We measure that by using a number of molecules, but most commonly a molecule called desmosterol, which is the penultimate molecule in one of the cholesterol-synthetic pathways…

Note from Tom Dayspring: Since there are two main cholesterol synthesis pathways you should be in the habit of always saying lathosterol and/or desmosterol.

…Cholesterol synthesis begins, as you know, with the creation of a molecule from two molecules of acetyl CoA. Many, many steps later, I believe it’s north of 30-

Dave Feldman: Yeah, it’s 30.

Peter Attia: Yeah. You’ll have this pathway where you’ll go from desmosterol into cholesterol.

Note from Tom Dayspring: The other would go through lathosterol.

When we measure the desmosterol level, especially when we measure changes in it absent any other drugs, because there are some drugs that can interfere with the conversion of desmosterol to cholesterol, but we get a sense of what the synthetic function looks like. There are some people that are hyper-synthesizers. There are some people that have normal degrees of synthesis, and there are some people that actually synthesize a relatively low amount.

The third thing we look at are phytosterols. These are plant-based sterols, which means we can’t make them. But by measuring them, we get a clever insight into how cholesterol may be recirculated in the body. Again, I know you know all this stuff, Dave, but I think for the listener, it’s important for them to get a little brush-up on this stuff. Most of the cholesterol in our body is endogenous, meaning we made it, and then we recirculate it. Maybe about 15% is exogenous…

Note from Tom Dayspring: 15% of the potentially absorbable cholesterol pool in the gut.

…maybe less. It would depend on a number of other factors, but the majority of the cholesterol that you eat.

Every once in a while, you see a funny case study. There was one this week about, “Guy eats 30 eggs a day and has low cholesterol. How is this possible?” It’s sort of an idiotic discussion that I can’t believe we’re still having. Even Ancel Keys (24:45) noted this a million years ago. Dietary cholesterol plays a very trivial role in the circulating cholesterol pool because it has esterified side chains that can’t be absorbed…

Note from Tom Dayspring: But cholesteryl ester can be de-esterified by pancreatic cholesteryl esterases (a specific type of lipase) – once CE becomes free cholesterol it can be absorbed.

…Nevertheless, you make all this cholesterol. We’ll talk about in detail, I’m sure, how it’s trafficked.


It [i.e., cholesterol] comes back to the liver. A portion of that is secreted through biliary means. Now, that biliary cholesterol, along with phytosterols…

Note from Tom Dayspring: With a minuscule amount: Niemann-Pick C1-Like 1 (NPC1L1) has a major affinity preference for cholesterol – far less for PS and very, very little for stanols.

…are brought in through this [membrane influx transporter] Niemann-Pick C1-Like 1 (NPC1L1) transporter into the enterocyte where the LXR gene basically tries to regulate how much of this stuff do you need. If it’s doing its job correctly, it jettisons out anything [i.e., sterols] excess through this ATP-binding cassette G5/G8. Of course, there are people that have deficiencies [i.e., gain or loss of function] and all of these things that can lead to hypercholesterolemia…

Note from Tom Dayspring: Hyper- or hypocholesterolemia, or phytosterolemia.

…In theory, the system should balance itself out. People who are very high synthesizers tend to compensate by being low absorbers, and vice versa.

The long-winded soliloquy was for a reason. It’s not only important to get the lipid and lipoprotein numbers, but it also helps to know those three things at baseline, one of which you get for free. You’re going to at least get a benchmark of our triglyceride, but to also know your levels of desmosterol[/lathosterol], which would be your proxy for synthesis, your levels of sitosterol, cholestanol, sitosterol…

Note from Tom Dayspring: Humans cannot synthesize sitosterol, campesterol, or any phytosterol: cholestanol is not a plant sterol – it is a zoostanol – converted from cholesterol by liver or gut microbes.

…These things are phytosterols, meaning we can’t make them. So, the higher they are, the more we know we’re absorbing sterol.

The fourth thing that regulates LDL-P … Again, it’s triglyceride burden, cholesterol synthesis, cholesterol reabsorption. The fourth thing is LDL clearance. Now, that is not as static as people would like to believe. It’s probably not even as static as I used to believe. I used to believe that it was sort genetically determined what your LDL clearance would be, and obviously, there’s a great variability there. We see it all over, but it turns out that that is highly regulated at the level of the liver.

So, even though we can use a drug to demonstrate the variability of it, a statin being the most obvious example … Statins are specifically designed to increase LDL clearance from the liver by decreasing liver synthesis of cholesterol [which in turn increases the expression of LDL receptors]. Other changes in cholesterol concentration throughout the body, probably the burden of reverse cholesterol transport and other things, will also impact that clearance, and the majority of LDL, of course, is cleared hepatically.

So, we don’t have an assay for that. This is the one where I always have to say to patients, “The only way I can really figure out if your LDL-P skyrocketed because of defective clearance,” which would, by the way, have to be a new onset of defective clearance, “is if the other three things don’t change.” Or if they get better, meaning they all … I hate to use the term better or worse, actually, because this is really … They’re neither better nor worse. They just are what they are. Right?

Dave Feldman: Right.

Peter Attia: But if the synthesis goes down, the absorption goes down. The trigs are largely unchanged, and the LDL-C goes up, the LDL-P goes up, then you know clearance has gone down. Most of the time, you can’t actually measure that unless you get lucky, and by measuring, I mean, sort of impute. Anyway, this is helpful, and I suspect this will offer an alternative hypothesis to sort of what we’re seeing. Anyway, I apologize. I’m talking more than I should be.

Dave Feldman: Not at all. For what it’s worth, what you just mentioned … I myself have not gotten a sterol test. I haven’t actually broken down these, but I have been particularly interested in this. For what it’s worth, Peter, I’ve been looking forward to this because I think I may actually be just the stealth interviewer in the room because I think it’s just as possible I may be asking you more questions than you’re asking me.

I do want to add one thing to what you were just talking about … A lot of people go on a low-carb diet, and they know they’re bringing up their total amount of dietary cholesterol, who then see a likewise increase in their serum cholesterol, the cholesterol in the blood, naturally, because it’s intuitive, come to the conclusion, “That must be because I am a hyper-absorber. I must be absorbing more cholesterol.”

I think in the course of this conversation, this will help to elucidate another particular … I’m really using elucidate a lot … will help to illustrate another reason why that may be the case, because you may, in fact, be trafficking more fat as energy, and therefore it may be ridesharing with cholesterol in these lipoproteins…

Note from Tom Dayspring: But since fat is trafficked as TG, why does TG drop in such persons?

Peter Attia: Yeah. I think the terminology is going to be confusing, so when we talk about hyper-absorbers, we’re referring very specifically….

Note from Tom Dayspring: Very specifically to cholesterol, not fat, absorption, and, of course, TG cannot be absorbed without being de-esterified to FA by intestinal lipases.

…to this mechanism about this Niemann-Pick C1-Like 1 transporter in the ATP-binding cassette, which is called … It’s usually referred to as ABCG5/G8. But as you said, that person who says, “Hey, Dave, I just went on a low-carb diet. I’m eating more eggs and more this and more that, and my cholesterol has gone up”…

Note from Tom Dayspring: Note that if this person is among the 20%, or so, of people who are cholesterol hyperabsorbers, indeed his cholesterol would increase – in most who are not hyperabsorbers, it would not. So there is a minority of people who do indeed have to limit cholesterol intake – monitoring sterols would help identify them.

…Well, the problem with that is it’s wrong on many levels. Right? We should never be talking broadly or vaguely about cholesterol. What went up specifically? Did LDL cholesterol go up? Did LDL particle number go up? Did total cholesterol go up? Etcetera. But it’s quite likely that those two things are not causal

Note from Tom Dayspring: But interestingly, genetic loss of function of NPC1L1 (hypoabsorbers) have significantly less CAD.

…meaning the person who’s increasing their consumption of dietary or exogenous cholesterol is also usually increasing their consumption of dietary fats.

Dave Feldman: Absolutely.


Peter Attia: As we’ll talk about later, I’m sure, one of those sub-types of fats in a subset of susceptible individuals seem to set off a hyper-synthetic pathway for cholesterol, which … When we get to it, I want to share with you my dataset on hyper-responders, meaning these patients, which is, what is the pattern of hyper-response?

Because not everyone has this experience where they go on a ketogenic diet, and their LDL metrics skyrocket. I didn’t have that experience, but I have the privilege of getting to see the blood of 10s if not 100s of people over the past few years, so I get to see, “Oh, sometimes this happens. Sometimes it doesn’t. What else is going on here?” What year did all this start for you?

Dave Feldman: In April of 2015 was where I got my second A1c of 6.1, which is the hemoglobin test, which suggests that I’m pre-diabetic. That-

Peter Attia: And that was with a triglyceride of 80?

Dave Feldman: Yes. Actually, that’s correct. That was the very last time. That was that same test.

Peter Attia: Okay.

Dave Feldman: When I saw that, I then immediately felt compelled to go and learn everything I could on how to change my blood glucose levels, because also I had a fasting glucose of, I want to say, like, 103. At the doctor’s office, they said, “Well, yeah, this is the second year in a row, but we’ll keep monitoring it.” I said, “Well, no thanks. I’m going to start trying to figure out what it is that I can do to dodge type 2 diabetes because it’s rampant on my Dad’s side of the family.”

I started to go to diabetic forums. Diabetic forums, they were talking about this LCHF diet, which I would then find out is a low-carb, high-fat diet. I would then look a little bit further. I found out about the ketogenic diet. It all sounded very interesting. I remember at that time asking on the forum, saying, “Okay, now wait a sec. How can I be sure my cholesterol won’t go up?” At the time, the common answer was, “Well, it only happens for a few people…

Note from Tom Dayspring: Probably 20-30%. Even then, it’s complicated, but it’s really probably not a problem.”

There wasn’t really a very solid answer to it, but I felt at least confident enough that it was unlikely to happen to me. Therefore, I would go ahead and take the shot because my cholesterol numbers were generally pretty good. I did like hearing that it typically raised HDL-C, and that was the one thing my doctor would occasionally ping me for. He’d say, “I’d like it if your HDL-C was a little bit higher.”…

Note from Tom Dayspring: Well, at one time we all said that, but certainly not in the last 5 years, or in my case 10 years.

…After I started, both my dad and my sister got enticed to do it as well. My dad is type 2 diabetic. His last A1c was like 8.3. My sister is not diabetic, but was hypertensive

Note from Tom Dayspring: And thus likely at least also insulin resistant.

…They both got inspired. They end up going on the diet. To this day, now my dad’s in the, like, 5.9s, something along those lines, and my sister’s no longer hypertensive when she’s staying on the diet fairly well.

The two of them get their cholesterol test before I do, after they started the diet around the same time, and both of them … I warned them in advance that their LDL cholesterol might bump a little. Sure enough, that is what happened to both of them, but it wasn’t that concerning. I get mine a little bit later, about seven and a half months later, and mine skyrockets.

Peter Attia: This is, like, late 2015?

Dave Feldman: Correct. November 2015, I get … I believe my LDL-C was 329. I believe my … I’m sorry. My total cholesterol was 329. My LDL-C was 200 and … I want to say 250. I can’t remember. Somewhere around there. That was a very cathartic moment for me, at least as far as … I’m looking at the lab work. I’m going, “What the heck happened? How did I get to this place?”

For two very miserable weeks, I found this guy, Peter Attia, who happened to already have a lot of data and a great series. I engulfed your series, but I could hardly understand it at the time. I started reading up on Thomas Dayspring. I found Tara Dall. I started looking at just anybody and everybody who could say anything about lipids. In the course of doing it, that’s when I started seeing this pattern. As I started tweaking a little bit into Clinical Lipidology, the book, and-

Peter Attia: When you say, “the pattern”, just to be clear, the increase in LDL cholesterol in the presence of a low-carbohydrate, high-fat diet?

Dave Feldman: No, I actually mean it being a network. That-

Peter Attia: Okay, so say more about that.

Dave Feldman: Okay. Lipoproteins are a boat. This is going to be 101, but let’s kind of do that for the listener for a second. There are lipid-carrying proteins. Your body makes them, and they make them at numbers we can’t even imagine. They’re measured in quintillions, which is like in million-trillions. It’s doing this both in the gut and in the liver. When they make it, they’re basically packing in lipids…

Note from Tom Dayspring: TG that certain cells need and PL that all cells need and cholesterol that few cells need.

…that these cells need. In particular, they pack just about every kind of lipid, not just triglycerides, which your body uses for energy, but also cholesterol, which we’re going to be talking about a lot, but also fat-soluble vitamins, like A, D, E and K. It packs all of these in the same container. But what’s neat about it is this boat, in order for it to get to the cells that want to make use of it, need to have a complex system in place so that they can kind of special order what it is that they want to take off of these lipoproteins.

That’s where it gets interesting because the lipoproteins have these kinds of sneaky bumps on the top of them that are proteins that are the apolipoproteins. Hopefully, we don’t have to get too technical for the audience on that, but of course, I’ll appreciate it if we can.

Peter Attia: I think we’re going to have to, but yes.


Dave Feldman: Yeah, we’re probably going to have to, but it’s those apolipoproteins that you could relate on a computer side to, like, metadata, to, like, headers, for example, as far as where it is that they’re going to go and why. What excited me as I was learning about this … I was like, “I don’t know how much of this I’m projecting my own experience as a software developer, but this looks like a very complex series of distributed objects that clearly, if I’m looking at it from a payload perspective, is primarily an energy-distribution network.”

It appears as if its primary job, more than any other job, particularly chylomicrons and VLDLs, are to deliver this fat-based energy to tissues. That’s what they go out with, and that’s what they come back typically not with coming back to the liver. Because almost everything ends up coming back to the liver…

Note from Tom Dayspring: Not true.

…this looks like it’s a central regulator. In that sense, this looks … I mean, in many respects, it has many attributes common to what we would call in software, as a cloud network, something we use a lot.

It’s very important to be able to do things at scale, but particularly with the level of interaction that goes on with the lipoproteins between each other. It’s not like these boats go out autonomously and never have any interaction beyond just going to dock and dropping off their different cargo. They actually constantly connect with each other through things like cholesteryl ester transfer protein, phospholipid transfer protein, and so forth.

These are different ways in which they, in the process of moving through the bloodstream, actually have further interactions that move around the total pool that’s being used by the entire system. Hopefully, I didn’t get too technical there, but you get the sense that in many ways, this has a lot of overlap with techniques that we use right now in how we build networks out of servers.

Peter Attia: I think we do need to get pretty technical on this because I suspect that you and I will draw different conclusions from the data. In my experience, the easiest way to understand where those differences lie is to sort of start to get into some of the things that we would view differently. I’ll start with one thing that you said … I like to be, I think, maybe clear on where I believe the chylomicron, the VLDL, the IDL, and the LDL are coming from, going, and what they’re doing.

Now, we can’t actually know for certain any of these things…

Note from Tom Dayspring: We can from kinetic studies.

…I’ve had some very interesting discussions with people about this over the years. I mean, I’ve had … One of the most brilliant lipidologists I’ve ever spoken to said, “Why do we have LDL?” to which the answer is, “It’s just God’s cruel trick on our species,” because most other species don’t have the LDL burden we have. I think a more thoughtful answer, though, is the overwhelming burden of evidence is that the purpose of LDL is to carry out reverse cholesterol transport. You alluded to this already.

You basically have these three different lineages. This is a bit of an oversimplification of lipoproteins. You have these chylomicrons which, as you said, are primarily getting fat from the gut and very rapidly undergoing a process of hydrolyzing themselves [lipolysis] and releasing…

Note from Tom Dayspring: mediated by — “release” — is the wrong word.

…through an apolipoprotein called apoC-II, all other triglyceride, through interaction with something called lipoprotein lipase. So, we have this rapid chemical reaction that very quickly gets rid of these incoming dietary and sometimes non-dietary …

Because I want to be clear that we can’t really distinguish exogenous and endogenous fat in that pool because you’re going through that same recirculating process. But at the risk of oversimplifying, fat comes in the body. If you’re on a high-fat diet, you’re eating fat. It’s coming in the gut. The chylomicron, which is its own little lineage because it has a different apolipoprotein … It has this thing called apoB48, as you know. That comes in, and then the [multiple copies of] apoC-II interacting with the LPL is what’s extracting that…

Note from Tom Dayspring: i.e., promoting hydrolysis of chylo core TG (into FA and MAG).

…Oh, yeah. Great. Dave just whipped up a great picture, so-

Dave Feldman: We’ll have this in the-

Peter Attia: We’re going to link to all this stuff. Then, that way when we’re making the show notes, we’ll link all that stuff. Okay. Then, you have another path we’re not going to talk about much today, which is the HDL path. Totally different. These particles last much longer. They [were once thought to be solely and] primarily responsible for reverse cholesterol transport, of which there are two types.

Note from Tom Dayspring: We now know is a very complex and dynamic process.

…Reverse cholesterol transport means taking cholesterol from the periphery back to the liver [or intestine]. That can occur directly, which is when the HDL brings cholesterol itself back from another tissue

Note from Tom Dayspring: I divert, but to make this even more complex much if the cholesterol in HDLs has a hepatic origin.

…to the liver.

The most important place it does this is from the subendothelial space. If you have oxidized sterols waiting to cause atherosclerosis, the HDL can actually go and, through another one of those ATP-binding cassettes, it can delipidate that HDL,…

Note from Tom Dayspring: A specific process specifically termed macrophage RCT, which is mediated thru several membrane receptors and free diffusion

…the sterol, take it right back to the liver. That’s direct RCT. But there’s also indirect RCT, which is the LDL can bring cholesterol back and give cholesterol to the HDL through CETP.

Note from Tom Dayspring: misspoken – The LDL accepts CE from the HDL — it does not give the HDL any CE, but rather, TG.


Dave Feldman: The other way. Right? You mean HDL actually taking the cholesterol and giving it to the LDL to take back to the liver?

Peter Attia: No, the … Well, actually both. LDL gives you HDL to go back to the liver.

Note from Tom Dayspring: No: it gives the HDL TG, which renders the HDL more susceptible to the lipolytic actions of hepatic and endothelial lipase. Also understand that the HDL actually transfers some CE to VLDLs, which can return it to the liver as part of indirect RCT.

Dave Feldman: I see what you’re saying.

Peter Attia: Yeah, yeah, through CETP.

Dave Feldman: Right.

Peter Attia: That’s indirect RCT. But it’s that VLDL to IDL to LDL path that is governed by these apoB … Their lineage is described by apoB100, which differentiates them. Now, one of the things I didn’t learn until recently, and I don’t know if this is accounted for in the model because the figure that we’re going to link to is not actually showing that, is that 40% of LDL comes directly from the liver. It’s de novo created.

Dave Feldman: Right. Tom Dayspring linked me to the study that I had since read on this one. Really-

Peter Attia: Which one? Is this the Frank Sacks?

Note from Tom Dayspring: Yes.

Dave Feldman: I don’t remember. I don’t remember the authors. They get into the different subspecies of apoC-IIIs, and apoE, and how they counterbalance each other as far as degree of affinity for clearance, and so forth.

Note from Tom Dayspring: Some truth to that: apoC-III increases apoB particle hepatic synthesis and decreases LDLR-mediated clearance of apoB particles, whereas apoE enhances LDLR-mediated clearance.

Peter Attia: Yeah, that’s sort of … I mean, I shouldn’t say it’s unrelated. This stuff’s all related, so we’ll park this topic because it’s super interesting. But when these lipoproteins [i.e., LDL] have apoE on them, which is pretty unusual … It’s about 2 to 5% if my memory serves me correctly. We’ll fact check that. It might be different, but a very small number of these apoB100s are carrying apoE…

Note from Tom Dayspring: In fact, apoB100 VLDLs and IDL can carry numerous copies of apoE.

…When they do, they actually have much more rapid clearance. So-

Dave Feldman: As in having apoE only and not having apoC-III?

Peter Attia: Well, that’s actually a good question. apoC-III is clearly the worst actor you could have here. There’s nothing worse than having an apoC-III sitting on your apoB100 cell. In fact, some would argue that may be the single worst thing you could ever have happen because it increases the residence time of them.

Note from Tom Dayspring: Both apoB particle synthesis in the liver, delays their lipolysis in plasma and severely hinders with LDLR mediated clearance. In addition apoC-III is a very pro-inflammatory protein.

We’re going to come to this, I know, because you’ve written about remnants. I’m going to argue that we have no way of knowing what a remnant is without being able to measure apoC-III because when we look at VLDL cholesterol …

I apologize to the listener. I swear we’ll get back to our main point here. Unless we actually know something more than just how much cholesterol is in VLDL, we have no way of knowing whether it’s an appropriate remnant, what we call a physiologic remnant, or a pathologic remnant. The pathologic remnants disproportionately carry apoC-III, which increases their residence time. The same is true on LDL, but the point here is-

Dave Feldman: I do want to follow up on that point, though.

Peter Attia: Yes, yes, yes. We absolutely will because it’s such an important point. To me, it’s one of the two most interesting clinical assays I’d like to see developed. I would love to see a clinical assay for apoC-III-

Note from Tom Dayspring: Irrelevant for this discussion, but there are three different isoforms of apoC-III with different degrees of risk.

Dave Feldman: I would, too.

Peter Attia: …and for LDL triglyceride concentration, which goes back to a point we had a few minutes ago.

Dave Feldman: And for anybody listening who’s developing that assay, please reach out to me, because for all the self-experimentation I do … I mean, one of the things I didn’t mention is that I literally just did my 100th blood draw last Tuesday. I have obviously done enormous amounts of self-testing to do this, and that’s exactly one of the things I want to check is how dynamic, or not dynamic, the distribution of things like apoC-III are based on,…

Note from Tom Dayspring: Theoretically – but apoC-III is transferable and a large part of it is on HDL – perhaps at a certain level an identifier of dysfunctional HDL.

…for example, existing illness or the energy distributions and so forth.

So anyway, I realize we’re kinda getting in the weeds here, but it’s-

Peter Attia: Yeah, I mean you should have a low level of C-III because your insulin levels are quite low. So C-III tends to move with insulin. So this may be one of the things that explains why someone with Type 2 diabetes, who is hyperinsulinemic will, on a particle for particle basis, maybe even have a greater burden of the lipoprotein because the actual residence time of each of their particles, both the VLDL and LDL, is longer than someone with lower insulin.

So, you have this de novo creation of VLDLs, and you have this de novo creation of LDLs, and they form this circulating pool, but to my knowledge, we can’t really differentiate those when you look at that snapshot. I can’t tell, is that an LDL that came from a VLDL, or is that an LDL that came straight from the liver in that form?

Dave Feldman: And that was actually one of the questions I had for you was how, with a kinetic study, can you actually determine if an apoB100 lipoprotein that was secreted by the liver ever has, say an apoC-II on it?

Note from Tom Dayspring: We do know that every VLDL and certainly every chylo from the intestine is loaded with multiple copies of C-II.

Peter Attia: We don’t have any clinical way to measure that.

Dave Feldman: Right. And in that sense, I fully can see that I can’t be sure, even with the energy model, that the LDL particles that I’m seeing, the LDL-P, that I can say with any level of real confidence how many of the total proportion of those were truly for energy delivery.

Note from Tom Dayspring: The answer is zero.

Peter Attia: This is where we get into the semantics. I would argue none of them are for energy delivery, because that’s not what LDL does. But I think what you mean is, how many of them came from VLDLs that were trying to deliver energy?

Dave Feldman: Right, originated as VLDLs, for the purpose of doing it. In the morning, let’s say that your job is to deliver pizzas.

Peter Attia: Right.

Dave Feldman: And that’s your job, and you know that it only takes you about an hour to do. And then, guess what, the rest of the next two to four days you’re actually going to be patrolling the neighborhood. You’re the neighborhood watch. And you’re going around, and you’re also helping to fix up…


Note from Tom Dayspring: LDL does not fix anything.

…people’s houses, or something along those lines. Somebody who comes into the neighborhood and sees a whole bunch of these cars patrolling, they don’t know how many of those people actually delivered pizzas before they got started on that part of the shift.

And that’s basically what we’re both coming to, right? We don’t actually know how many people left the liver, how many VLDLs-

Note from Tom Dayspring: Well we could if NMRs still reported total VLDL-P.

Peter Attia: Well, we sort of know. I mean, what Frank Sacks’ paper showed is if you have a triglyceride … If you take patients with low triglycerides, and I believe he used a cutoff of 130 mg per deciliter, 38 percent were de novo secreted by the liver, 62 percent came from either IDL or VLDL, where you had de novo … I don’t think the paper differentiated between which ones went IDL to LDL versus VLDL to IDL to LDL…

Note from Tom Dayspring: But there is a paper by Krauss which did.

…So that’s an important point.

The second thing is the half-life … I actually had to go back and look at these kinetics because, I did a podcast with Ron Krauss, as you know. I mean, I don’t remember when we recorded it. I think it came out kind of recently. But he mentioned that LDL half-life was a day. And I was like, “I always thought it was longer than that.”

Dave Feldman: The literature says two to four days.

Peter Attia: No, actually if you go back and look at the kinetic study, people are confusing half-life with residence time. The half-life of an LDL particle is about a day. Now it can be longer, but that’s a pathologic state, which gets back to this apoC-III thing.

Dave Feldman: That’s interesting.

Peter Attia: Yeah, you can have a pathologic state where LDLs will hang around longer, but if you look at the actual kinetic studies, and Brown and Goldstein did this work. This is part of the work, I believe, that they won a Nobel Prize for. The kinetics of LDL are pretty well understood, using very elegant tracers. And we’ll link to that paper, because I actually had to go back and look at it, because I was surprised by Ron’s answer, when he said that the LDL particle half-life is really only a day, under nonpathologic states.

But, you’re right. If I do that snapshot of, if I look at your LDL particles and I see 3000 nanomoles per liter, I do not know with absolute certainty how many of those your liver made directly, versus not. But again, assuming you’re insulin sensitive, assuming you fit Sacks’ model of patient, that would suggest that roughly 40 percent of those were just de novo created. And of the remaining 60, some of those were from de novo IDL and some came all the way through the VLDL pathway.

Dave Feldman: With that in mind, here’s what I would speculate. This is purely hypothetical, but I would speculate if you were to grab a whole bunch of people who are … We’ll hopefully get into this model that I’m talking about, that I call lean mass hyper-responders. People who are at the far end of the spectrum, they are athletic, they are thin, and they are very, very low carb, and therefore see very high levels of LDL-C and LDL-P, but they also have very high levels of HDL-C and low levels of triglycerides.

I suspect that they would show a very high rate, proportionally, of VLDL secretion, that they actually are trafficking a lot more, for their energy, triglycerides in VLDL particles…

Note from Tom Dayspring: If they were, their serum TG would have to be high, as the preponderance of serum TG are from VLDLs. They turn into LDLs via lipolysis – meaning they lose their Core TG and surface PL.

…and therefore have succeeding LDL particles as to the explanation as to why their LDL-C and LDL-P would be higher.

Peter Attia: So let’s use this as a moment to, because I want to get into that in greater detail, but let’s take that step back and have you maybe just put a little bit more color on what you mean by your lean mass hyper-responder phenotype.

Dave Feldman: Sure. I even have a graphic we’ll be sure that we link. For those people who go on a low carb diet, some subset … Nobody seems to agree on this because there’s really not been any large study done on it…

Note from Tom Dayspring: Pretty sure Sarah Hallberg has emerging data on this from her recent study.

…Some people will say it’s five percent, some will say it’s 30 percent … will, like me, see that their LDL, their total cholesterol, they will see both of those rise-

Peter Attia: Can I interrupt you for one sec?

Dave Feldman: Yeah.

Peter Attia: Have the people at Virta Health released any of these data? Because they would probably have the most rigorous database on this.

Dave Feldman: Here’s a little bit of my qualification here. The problem that I had is, at least with what we see at cholesterolcode.com, the blog that I have, we have lots and lots of hyper-responders that sent it in, there is seeming to be a higher proportionality of people who are lean and/or fit, who seem to be metabolically flexible. Virta, of course, its pool of participants, they had to start — I hope I don’t get this wrong, but I believe that their BMI is at a much higher level when it’s-

Peter Attia: Well, it’s a company, obviously, that is dealing with patients with Type 2 diabetes. So yes, they’re not going to be disproportionately lean and fit to begin with.

Dave Feldman: Right.

Peter Attia: So you’re saying that that basically wouldn’t be the ideal pool to observe this phenomenon.

Dave Feldman: I would prefer a broader base.

Peter Attia: Okay. All right, sorry.

Dave Feldman: With that said, it’s absolutely true they’ve got some of most pristine data. They also have that data that I would be looking for, where they get NMRs…

Note from Tom Dayspring: I think they did apoB.

…for example, before the participants start. That was nice, too. They’re going to have so much great data that comes out of that. But getting back to hyper-responders, per what you were talking about before, this was what we call people, and this predates me, who would go on a low carb diet, would see their LDL cholesterol, their LDL particle count climb.


Then, there seemed to be a subset, and I wrote about this about a year ago last month, of people who are on the furthest end of the spectrum actually tend to have the highest levels overall of LDL cholesterol, but also have other things in common. This pattern is very distinctive. They would have, say, an LDL of 200 or higher.

Peter Attia: You mean LDL cholesterol?

Dave Feldman: Oh, sorry. LDL cholesterol of 200 or higher, HDL-C cholesterol of 80 or higher, and triglycerides of 70 or lower. And this is so prominent, to the extent to where I even did this recently at another conference. I called on a speaker, where I said, “I’m very interested in your lipid numbers, because I think you might be a lean mass hyper-responders,” and she said, “Well, I hadn’t actually taken it in years.” We tested it on the spot and it hit all of those points. Her LDL-C was 189, her HDL-C was 80, and her triglycerides were 70.

This seems to, so far, span across all sorts of, for example, apoE types, apoE 3/3s and 2/3s, as well as the 3/4s and 4/4s. We had not been able to identify any other SNP, or anything that’s clearly associating this type.

Peter Attia: Including the PPAR-alpha and PPAR-gamma?

Note from Tom Dayspring: How would one measure that?

Dave Feldman: That’s one of the ones I wanted to follow up on, particularly since I heard your podcast with Rhonda. And I had a bunch of people on Twitter just send me full body pictures to use in this most recent speech I did on lean mass hyper-responders.

Peter Attia: You have to be careful with that. You can get into trouble that way.

Dave Feldman: I did say, “Be sure to respond to this Tweet, where I am specifying exactly what I had to use them for.” Anyway, generally speaking, they tend to be very fit, they tend to be very thin. And oftentimes, and you kind of had a story of your own before I even got into this, they’ll say, “I really don’t want to stop. I really love this way of life. I feel better than I ever have in my life.” I give the same answer, not too far different from yours, in that I say, “Okay, generally speaking, I feel like all of your markers look great, low inflammation across the board.”

Of course, I give the standard, “I’m not a medical doctor and this isn’t medical advice, et cetera. But all of that said, it’s hard for me to come to the conclusion that you’re in trouble unless we can likewise see further markers, such as expanding CIMT and CAC and so forth, that’s going to show that you actually are developing higher rates of atherosclerosis. Now, I’ll qualify in advance,” because I’m sure you’d want to say this as well, and I’ve mentioned this to several people, “Things like atherosclerosis can develop without any sign for a fairly long period of time.”

For example, I think it’s, what, 60 percent before you even see occlusion in the lumina, for example.

Peter Attia: Well let’s, yeah, let’s back up a little bit. So I always want to be careful when people are talking about CIMT and CAC that we’re never using those in the same terms as we would think of biomarkers. Right? So a biomarker, what you’ve described is, “Let’s look at your LDL cholesterol, or better yet your LDL particle number.” But remember something. A CAC, which is a calcium score, so it’s a dry CT scan that very quickly scans over the heart and just picks up calcification, no anatomic detail, where a CIMT, which is a type of ultrasound that looks at the intimal thickness, so that’s one of the walls of the arteries thickness in the carotid arteries in the neck, these are both tests that are used to try to gauge advanced disease.

The real way to think about this is to … And I think that Ron and I talked about this at length in the podcast … is to look at a pathology textbook. When you look at the autopsies you’ll get a sense of what’s going on. Long before you have luminal narrowing, which may or may not accompany a problem, you have a very clear documented path of what this disease does. So again, I think you know this Dave, but I think for the listener it is worth repeating this, if they don’t want to go back and read some of the posts I’ve written on the progression of atherosclerosis.

When we’re born we have these beautiful arteries. The arteries have this endothelial lining, so this very thin type of cell that coats the luminal, meaning the part that’s closest to where the blood is flowing. There are spaces between these and, via diffusion, lipoproteins get in there and out of there all the time. This is relatively well understood to be a gradient phenomenon, so the more lipoproteins you have the more of them that are going to go in.

But, as we talked about earlier, other things will influence it, the residence time for example, which might be why this apoC-III thing is such a pain in the butt, because if it allows these cells [lipoproteins] to stick around longer, bad things happen. Now, what most people don’t know, I think, is that an LDL particle is more likely to come back out when it goes back in there, than it is to stay in there…

Note from Tom Dayspring: Impossible – The apoB on LDLs immediately bind to proteoglycans and are trapped forever.

That’s good news. HDL particles always come out of there. There are other types of particles, Lp(a) and pathologic VLDL remnants can do the same thing as well.

Note from Tom Dayspring: Likewise once in artery VLDLs cannot escape.


The problem occurs when proteoglycans bind to, and let’s just make math easy and not get into the Lp-little-a’s and the VLDLs at the moment. Let’s just talk about the LDL, since that’s the largest burden of this. But when these proteoglycans bind to the LDL, and it gets retained, all of a sudden now you have something that’s where it’s not supposed to be. That’s not where we want that thing.

It’s obviously in a high oxygen environment, so it’s going to undergo a chemical reaction called oxidation, because it’s carrying a cargo, a sterol that … And by the way, they can be carrying phytosterols…

Note from Tom Dayspring: Yes, but PL are the major oxidation target.

…and other things like that, but these things have ample opportunity to undergo an oxidative reaction. It’s that oxidative reaction that then kicks off an inflammatory response in the endothelium. Now the good news is, today we at least have one laboratory test that can measure that burden of oxidation…

Note from Tom Dayspring: Weak estimation at best.

…It’s called the OxLDL assay.

Now, this has been around for a while, but clinically we’ve only been using it for a couple of years because it turns out some very small percentage of those LDLs, once they are oxidized, escape back into the circulation…

Note from Tom Dayspring: Some might, but the test measures circulating MINIMALLY OXIDIZED LDLs. Any LDLs escaping would be, way more than minimally.

…By sampling those we can track, indirectly, “Hey, what’s the likelihood that oxidative damage is happening?” For me, this is one of the most important metrics…

Note from Tom Dayspring: I would not call it the most important metric – it is one of several (MPO, Lp-PLA2, F2-isoprostanes, thromboxane catabolic molecules (Aspirin works test).

…I look at is, because I want to spend some time later on going over some clinical cases. I want to see some of the data on yours. I want to show you some of the data that will explain maybe how I’m thinking about this. But this oxidized LDL, which is well documented and described in different quintiles, right, is giving you a small sample of what’s going on. But for the listener, it’s important to understand that when you get a blood test, that’s not telling you what’s happening in your artery. It’s giving you probabilities of things that are largely stochastically governed, that are going on in your artery. And the OxLDL is no exception. Even though it’s a beautiful marker, it’s still dependent on the idea that a subset of those oxidized sterols…

Note from Tom Dayspring: Particles – the sterols or PL do not escape as free molecules- but again mostly the assay measures LDLs that have started to be minimally oxidized in plasma.

…are now escaping.

Dave Feldman: Can I actually ask a little more on that one? We already know that LDL particles, specifically apoB100 at the LDL stage, have alpha-tocopherol I think is how I’m saying, basically it’s vitamin E, right, as part of the antioxidant defense system. So they’re actually, I mean, part of the purpose of an LDL particle is to actually provide that as a means to battle reactive oxygen species. Right?

Note from Tom Dayspring: Vitamin E is absorbed in the intestine, and the vast majority gets to tissues in chylos – but some are on VLDL, which completes the delivery, but some are on LDL, but their function is not to deliver Vit E. Type IIIs have no LDL particles, yet have no vitamin E deficiency. VLDLs and chylos transporting Vitamin E in plasma (meaning plasma is rich antioxidants) is part of the reason why there really are very few fully oxidized LDLs in plasma – only minimally oxidized.

Peter Attia: I don’t know about that. If that were true, if it were solely true, it would make me wonder why people with LDL deficiencies wouldn’t have deficiencies of those processes as well, which to the best of my knowledge they don’t.

Note from Tom Dayspring: People with abetalipoproteinemia (who have NO apoB particles) and some with severe hypobeta do have vitamin E issues (But mostly it is because they do not have VLDLs) Many do supplement severe hypobeta patients with vitamin E.

Dave Feldman: Because I’ve actually been wanting to get into this a lot more, recently. Correct me if I’m wrong, but basically, there’s a certain degree to which you’ve got vitamin E on board. On top of that, you’ve got the potential of the phospholipid shell to become oxidized. So if you get oxidized phospholipids that also can bring about the role Lp(a) that can cleave off the oxidized phospholipids.

Note from Tom Dayspring: The lysine binding domains on 1 or 2 specific kringles traffic any oxidized lipid moiety – but most are oxPL.

That’s ultimately what Lp-PLA2 is, right?

Peter Attia: Correct.

Dave Feldman: It’s the enzyme [i.e., Lp-PLA2] that’s ultimately involved in helping to-

Peter Attia: Yes.

Dave Feldman: And this is also, I don’t know how much of this is actually demonstrated, but is ultimately where a lot of the concept behind why it is you would have a higher detection of small lipoproteins, particularly small LDLs, can come around to, is if you’re getting them constantly oxidized and having to constantly cleave them down to much smaller amounts.

Note from Tom Dayspring: Virtually all of the oxidation occurs within the arterial wall from which they really cannot escape.

And then they constantly remodel.

Note from Tom Dayspring: They do not.

Peter Attia: Yeah, but we’re getting off into two different things here. So let’s come back to this. It’s not clear to me that there is sufficient evidence to suggest that part of the role of LDL is to combat the oxidative stress.

Note from Tom Dayspring: Other than it has some vitamin E – that is not the primary role.

Dave Feldman: Okay. Then let’s put that as homework that we’ll catch up on after this.

Peter Attia: Yeah.

Dave Feldman: But this is relevant for whether or not we’re detecting oxidized LDLs that had never entered the intima. Right?

Peter Attia: No. The oxidized LDLs that we’re detecting have escaped the intima.

Note from Tom Dayspring: Maybe a few – Once again the assay measures circulating minimally oxLDL particles.

Dave Feldman: Interesting.

Peter Attia: There’s a very small subset that are getting out.

Dave Feldman: Okay. That’s definitely something I would like to follow up, because I’m genuinely curious about this stuff, as to whether or not they can be oxidized sufficiently that they’d get big…

Note from Tom Dayspring: Oxidation does not change particle size.

…because it also may be something that is part of the test or isn’t a part of the test, but I’d be curious as to how it can actually determine that.

Peter Attia: Yeah. What you’re basically asking is, how do we know they weren’t oxidized never inside a subepithelial space, and that’s a fair question. I don’t know the answer. I know very little about this assay. I mean, I know the technical stuff of how the assay works, like it’s an ELISA assay. I know what enzymes it’s looking at, but the broader question is, without a tracer, do we know if that LDL has actually been in the subepithelial space where it was bound, oxidized, and then escaped or liberated? So, fair question.

Dave Feldman: It’s certainly relevant to this larger question of the value of LDL particles, as to whether they play an important part of the immunological role.

Peter Attia: Well, they do, probably nowhere near as important as the HDL particle, which is probably why the HDL particle has such a long residence time…

Note from Tom Dayspring: Correct – HDL is the immunomodulatory lipoprotein.

…And the HDL particle, as important as it is for reverse cholesterol transport, both direct where it’s taking the lipid back to the liver directly, or indirect, which we talked about and you corrected me — thank you — where it takes it to the LDL and the LDL takes it back to the liver.


Certainly, some have argued that an even more important property of the HDL is the proteins that it carries, the immunoglobulins, all of the other things that it carries that play this important role in immune function. So it really seems that the overwhelming body of evidence is that the purpose of the LDL particle is to carry the cholesterol back to the liver.

Dave Feldman: Interesting, but this gets back to the multipurpose value of a vehicle. Is it doing things other than that, that also turn out to be relevant. And I think this kind of gets to the larger and more important question overall. The question that I started with, going back to my November 2015 days, was, I thought, very naively, that in a few days I would learn all I would need to about cholesterol and lipoproteins, find the landmark study that had a gazillion people, and it would just show that if you had lower LDL cholesterol you just died less. That was it. End of story.

At first, I thought that I had found that because I had found plenty that pointed to events and pointed to a lower cardiovascular risk, but then wouldn’t necessarily talk as much about all-cause mortality. I then had to learn about all-cause mortality. And then more and more I felt like I couldn’t get to something that really emphasized … I thought for sure at least I would see, for example, an elderly population. Generally speaking, the lower your natural LDL … We can get into SNPs, for example, on this … the more likely it is that you would just live longer, period.

Peter Attia: But you have to remember how these studies are powered. The challenge with ACM is, I don’t think any study in the history of civilization is going to be powered to detect that. It’s hard enough to detect cardiac mortality in a study. So I think we need to be more clear in what our concern is. If the concern is, if you are less likely to die of heart disease, you are more likely to die of something else, then we should state that explicitly and say, “Hey, low LDL, while may be protective of cardiovascular disease,” I will argue that is unambiguously clear and we can discuss that. But the bigger question is, are you concerned that, well, it’s increasing the risk of cancer or a neurodegenerative disease?

Dave Feldman: A trade-off.

Peter Attia: Yes. So the question there, that’s a question of power. It’s not uncommon in cardiovascular studies to see a reduction in coronary mortality with no change in all-cause mortality, or a non-statistical change. You know, most of the time you just don’t see a change, or it’s a change that’s very slight. And then you have to ask yourself the question, even if it looks, hey, death went up or down of other causes, you have to go back and ask yourself, “Was the study actually able to detect that?” That’s a very hard thing to detect.

Dave Feldman: Absolutely. And in fact, there’s even a paper that I pointed to recently that says, “Is this even worth chasing after?” Because it takes so much expense and time in order to get to a level in which it would be powered to detect for all-cause mortality, should we even make that part of the criteria that are required?

Peter Attia: Well, there’s a broader issue here, which is the lifetime exposure problem.

Dave Feldman: Exactly. And this is, of course, just the problem with atherosclerosis in general, is, you do a drug study that’s two, three years, but atherosclerosis doesn’t take two to three years from zero to-

Peter Attia: Yeah. I mean, I try to not get into any wars on Twitter, but once in a while, I’m just … I don’t know. I’ve had one too few Topo Chicos and I’ll let it rip. But if I have to see one more person try to tell me why FOURIER and ODYSSEY aren’t interesting trials because they didn’t show a big enough benefit, I might scream. So, just for the listener, FOURIER and ODYSSEY were trials that looked at two PCSK9 inhibitors. And I want to also be clear before I get into my rant. I am never having an economic discussion about this.

I’m saying that because people often confuse efficacy and effectiveness, and cost and value and benefit. I am not for a moment suggesting those things don’t matter. I am not going to argue one way or the other that the cost of a PCSK9 inhibitor is worth it. That’s an individual decision. Unfortunately, that decision is, for most people, made by their insurance company and that’s totally reasonable. I’m only interested in this as a conceptual tool, which is: does inhibiting PCSK9 make a difference?

And if you had told me — I remembered knowing this — that FOURIER and ODYSSEY had such short time horizons, I thought there’s no way they’d find a benefit, in particular, FOURIER. FOURIER took patients with an average LDL cholesterol of something like 90 or 92 mg per deciliter. These patients were already on the maximum tolerated dose of a statin. Okay, so they were at the 10th percentile in terms of their LDL-C, and in 2.2 years showed a reduction in events?

The null hypothesis should be it should have never worked. You should have needed 20 years to show any benefit when you understand Allan Sniderman’s lifetime exposure model. So a lot of people are critical and say it didn’t show a mortality benefit, it just reduced revascularizations and events in 2.2 years, to which I say, “You’re looking at that incorrectly.” The fact that it showed anything, to your point Dave, lifetime exposure is staggering, also on patients who were already maximally statinized.


So, coming back to this thing about lifetime exposure, this is where the Mendelian randomization becomes a very important tool in understanding LDL’s causality. What you alluded to at the outset, you are correct in noting is deficient. There is no lifetime study where, without a drug, you can prospectively manipulate LDL and follow people for 100 years and determine outcomes. That would be the ideal study. Right?

Dave Feldman: Yeah, that would be. You actually, you talked about this in the Straight Dope, with the wand.

Peter Attia: I know, I remember sort of going on … That’s right, the magic wand test.

Dave Feldman: Right. And I think it’s worth plugging into things, and important here. You’ve already said this on a prior podcast, and I don’t want anyone to misunderstand that you don’t necessarily buy into the zero LDL hypothesis. You don’t know that you would have LDL as zero. So you do believe there’s some kind of trade-off. You mentioned a few, for example, commonly known diseases for when your LDL actually gets to such a low level, like cognitive diseases and-

Peter Attia: The way I would say it is this. I’m glad you brought this up because it is a very important distinction. I absolutely believe that the lower the LDL, the lower the risk of cardiovascular disease, all other things equal. Why? Because LDL is necessary, but not sufficient for atherosclerosis, and I say that full stop. Now it’s important to understand what necessary but not sufficient means, because there’s going to be some people listening to this who are getting all phosphorylated now, and they’re just getting super pissed off. My advice is sit down, shut up for a minute, and pay attention.

Dave Feldman: Extra points for the phosphorylated insert, by the way.

Peter Attia: Yeah, yeah. So necessary but not sufficient is the relationship between oxygen and fire. Oxygen is necessary, but not sufficient for fire. Can you have oxygen and not fire? Yes. Can you have fire without oxygen? No. The lower the concentration of oxygen, the less likely you are to spontaneously get a fire. Then again that’s a bit of an oversimplification because there are so many other factors, endothelial health and oxidation, and inflammation, of course, are so important here.

But the problem I had with the zero LDL model, which again I think I don’t want to speak for people who … But I think what people are basically saying is, there’s a subset of people in the medical community who are saying, “We should just be driving LDL to zero because that’s the best thing.”

Note from Tom Dayspring: No one has been able to do that (make LDL or apoB zero) even with apheresis Statins or statins plus ezetimibe do not make LDL-P or LDL-C zero in anyone. Maybe 10 mg/dL with PCSK9 + statin and no short-term issues have been seen. Regardless, measuring cholesterol in circulating lipoproteins tells us nothing about cellular cholesterol. There are other sources of cholesterol in the blood, namely RBC – and cholesterol can be delivered to tissues, or extracted from tissues, by RBC. The drugs which lower lipoprotein cholesterol do not lower RBC cholesterol, where the majority of plasma cholesterol resides.

Well, my view is, as you said, no, that’s not necessarily the best thing. That might be the best thing for the heart. In other words, it might be the best thing to lower the risk of atherosclerosis, but it’s irrelevant because it might come at other costs. And so it depends on your point of view.

I do, from time to time, get into arguments with other physicians who take care of my patients as well, because I’m not a primary care physician, so I have to share my responsibility with other physicians. And about twice or three times a year I do have to sort of go to war with one of these docs, and it’s usually over one extreme or the other. And the most recent example of this was a patient of mine who came to me on 80 mg of Lipitor, which is the maximum dose of Lipitor. He had a very high calcium score and a very bad CTA, but he had not had an event that we knew of.

But for all intents and purposes, this is a secondary prevention patient, meaning, we define secondary as, “Has he had an event or not?” Well, he has had an event. His event is, look at his coronary arteries. Right? But nevertheless, he came in on 80 mg of Lipitor. His LDL cholesterol was very low, but his particle number was not quite at goal. The goal for a patient like that would be 10th percentile or lower, given how aggressively you’re managing. He also had a slight elevation of Lp(a).

So I added Zetia because he had very high levels of absorption, not uncommon given how much his cholesterol synthesis was being hit, and that brought him into goal. So now he was totally at goal. His cardiologist was happy. I was marginally happy, but what I didn’t like was his desmosterol level was now unmeasurable. Now, it turned out it was unmeasurable before, but I was so fixated on just trying to get him in the right zone, but now we had some breathing room and I said, “You know, now that I’m thinking about it …”

Oh, and by the way, in the interim Repatha and Praluent had been approved. These are PCSK9 inhibitors. And I thought, “This guy has no measurable cholesterol [in plasma],” in terms of synthetic function, so it’s very, very low. Now, a lot of people are right now going, “Aha, aha. That’s the problem with those statins. They inhibit cholesterol synthesis.” Well, careful. That’s true, but every cell makes more than enough cholesterol for its own use, with maybe a couple of exceptions. Gonadal tissues, steroidal tissues, during periods of high stress need to borrow [receive] cholesterol from other tissues

Note from Tom Dayspring: HDL.

But for the most part, every cell can sufficiently produce its own cholesterol, but …


Dave Feldman: I think you’re right on the most part. I’m not sure if I am convinced that … I mean, your body’s running a buffet. That’s the bloodstream. The bloodstream is this buffet of things that the body anticipates it wants to make available on demand to cells. And I believe, I mean again, this is kind of just the engineering approach.

Peter Attia: But we don’t know that’s true. That’s a hypothesis, with respect to the lipoproteins at least.

Dave Feldman: Absolutely true, hypothesis, but it’s now been proven to the other side as well. Right? We can see that the synthesis can happen within most cells, to be able to make their own cholesterol. We don’t know. I mean, do we have even in vivo studies, where we can actually observe that every amount of cholesterol that they would need would ultimately be synthesized, even under periods of stress, for example, muscle repair and growth?

Peter Attia: Well, we have natural experiments, right? We can look at the abetalipoproteinemia patients who can’t traffic cholesterol, therefore they would be entirely dependent on their own cellular endogenous production and they seem completely fine. So that’s not proof, because we don’t have proof to your point, but it’s certainly evidence to suggest … I mean, we also know when that’s off. Right?

One of the first things we used to see in the ICU, though at the time I didn’t pay any attention to it, was anytime a patient came in and they were septic or under great stress, so they had what’s called systemic inflammatory response syndrome, SIRS, so you could be a car accident, you were shot, you have a horrible infection, their HDL cholesterol would transiently take a huge bump [drop]. And I didn’t think anything of it at the time, other than it was neat. It was like, “Wow, 2X bump in HDL cholesterol, overnight.”

I think I would now look back and interpret those data as huge reverse cholesterol transport. Now the HDL is going out of its way to deliver cholesterol to, probably the adrenal glands first and foremost, because the enormous uptake of glucocorticoid, even epinephrine, norepinephrine are needed. So clearly there are examples of when this is not in a homeostatic balance. So I’ll take your point that …

Dave Feldman: Because the abetalipoproteinemia [hypobetalipoproteinemia] patients, in theory, should be the ones who are outliving us all. Right? They can take out the whole component of heart disease, of atherosclerotic plaque, everything. They should have massive longevity, relatively speaking, to everybody else.

Peter Attia: So the ones who get the longevity, because there’s only about 12 genes that are well enough studied, we have enough patients that we think we know something. And the most important of the longevity genes in cardiac is the hypo-functioning apoC-IIIs

Note from Tom Dayspring: And loss function NPC1L1 and hypobeta.

Dave Feldman: And that actually shows a net …

Peter Attia: A net longevity benefit.

Dave Feldman: Interesting.

Peter Attia: Yeah. Work out of Albert Einstein has identified these, roughly a dozen, genes. And the hypo-functioning apoC-IIIs are, I mean most of those genes are, you know, like GHR, IGF. apoE would be one, right, so apoE2 would carry with it protective benefits in terms of longevity, both cardiac and, but more that is neurodegenerative. But it’s those C3s. In fact, as we have kind of alluded to a couple of times, I believe there’s an antisense oligonucleotide (ASO) in clinical trials now, trying to impair apoC-III, so now it’s becoming a therapeutic target.

Dave Feldman: Dayspring alluded to that one, actually.

Peter Attia: Yeah. Then again, this is one of those drugs that might not have much of a benefit in an insulin sensitive person. They may have already captured that benefit by lowering insulin levels.

Dave Feldman: Well, and that’s actually part of what this energy model, and particularly that with hyper-responders, specifically lean mass app responders comes back to. I know you don’t necessarily hang out on Twitter too much, but you know that I’ve had-

Peter Attia: More than I would like.

Dave Feldman: I have this pinned Tweet, I’ve been pinging lots of lipid-lowering experts on this, I’ve said, “Look, I’m looking for any studies that show people with high LDL will have high cardiovascular disease, if they likewise have high HDL and low triglycerides.” But there’s one qualification, it can’t be a gene or drug study. It’s got to just be-

Peter Attia: That’s two qualifications.

Dave Feldman: What’s that?

Peter Attia: That’s two qualifications.

Dave Feldman: Oh, fair enough. Two qualifications. But, the point is-

Peter Attia: No, I’m seeing that. Here’s my concern with that, Dave. I have no doubt in my mind that you are a truth seeker. I don’t think that’s true of, necessarily, some of your peers. I do think a number of your peers are deluded and so filled with their own confirmation bias, and so unwilling to acknowledge that their precious low carbohydrate diets could be hurting them. That not with malicious intent, but with blind carelessness, they are absolutely ambivalent to anything.

I don’t put you in that category, so I will challenge you in the following way.

Dave Feldman: Great.

Peter Attia: When you say, “Show me an example of something that is not a genetic study, that can point to that phenotype.” The reason I would call issue with that is, why would you limit yourself from genetic studies? That’s sort of like me saying, “I want to know if there are people who are six feet tall. I think they might be, but I’ve never seen one. So, if you can go into a kindergarten class and find me one, I’ll believe it. But, you must limit yourself to the kindergarten class.” In other words, that’s an obscure example. What I’m basically saying is, you’re excluding so much potential data by excluding all of the genetics.


Because when people talk about genetic studies, we have to remember something, most of the genes, most of the SNPs that lead to alterations in lipids and lipid metabolism, are completely unidentified. FH, for example, familial hypercholesterolemia, which would be the most obvious example to counter that point, you’re excluding because it’s a genetic condition. What the listener might not know is that FH is a phenotypic diagnosis, not a genotypic diagnosis. FH is arguably the most heterogeneous collection of genes you can imagine.

So, why would we exclude looking at those people, when that’s, in many ways, one of the richest bodies of evidence for a natural experiment in … To answer the question, can you have LDL-C, high HDL-C, low triglyceride, and still get atherosclerosis? That’s the question you’re asking, right?

Dave Feldman: Yes, yes. Well, so, we’ll double back to that in a sec, but basically, you’re taking us back to genes, and this is why … This is another hypothesis, fully untested, I’m in the process of trying to collect on it, but I call this loosely, cellular lipid malabsorption. Right, just generally shorten it to lipid malabsorption.

Basically, here’s the issue that I have with the existing mendelian randomizations. For that matter, almost all of the gene-based studies, is what we’re trying to get, what we’re trying to get, is as much as we can, the isolation of just a higher gradient of LDL particle count. That’s what we all secretly … We want your wand. You’re talking about where we can wave it, and then there’s just magically more LDL particles in some people, or for that matter, fewer LDL particles. Without touching any other parts of the process.

The problem is, that I believe … I’m keeping a list of my own SNPs of those genes that are either resulting in higher or lower LDL-C. Unfortunately, of the ones that I find in the mendelian randomizations, they don’t just result in the higher LDL-C and LDL-P, they also come to be that way because there’s a lack of lipids or lipoprotein uptake by the cells.

Therefore, particularly with endothelial cells, you’ve got to be concerned that that could cause dysfunction. And therefore could be a reason for why you’d have higher levels of atherosclerosis. This is why-

Peter Attia: Wait, wait. So, explain that part again. The last part.

Dave Feldman: Endothelial cells being dysfunctional.

Peter Attia: Yes.

Dave Feldman: Would that be potentially problematic for atherosclerosis?

Peter Attia: Yes.

Dave Feldman: Okay, then why would we want to look at any SNP that would in any way impair, inhibit them relative to a normal person endothelial cell?

Peter Attia: Why do we believe patients, or a subset of patients with FH, as a result of their FH have defective endothelial cells?

Dave Feldman: Well, if you’ve got defective LDL receptors.

Peter Attia: There are not receptors on the endothelial cell. It’s diffusion mediated.

Dave Feldman: Okay. Well, yes, but you’ve got the receptors with the adipocytes, right?

Peter Attia: Yes, but, at least 20, if not 40% of LDL uptake, is not even receptor bound in the body.

Dave Feldman: Okay. What about which is-

Peter Attia: And not all cases of FH have receptor deficiencies. So, there are at least 2,000 vaguely identified genetic causes of familial hypercholesterolemia. They have fewer receptors. So, the PCSK9s are a subset of FH, right? About three to 5% of patients with FH have-

Dave Feldman: Overexpression of PCSK9.

Peter Attia: Overexpression of PCSK9.

Dave Feldman: Gotcha.

Peter Attia: But that’s how PCSK9 was discovered.

Dave Feldman: Okay, but in that case, you’re impacting a cell’s capability of uptake for lipids, or for lipoproteins, right?

Peter Attia: Yes, you are in that situation. Those patients’ livers, will take up less LDL, because PCSK9 is a protein that does, among other things, degrades the LDL receptors because they have hyper-functioning PCSK9, they are more rapidly degrading their LDL receptors on the livers, so they’re taking up fewer LDL particles. Which explains why they have higher LDL.

Dave Feldman: But this, again, introduces a dysfunction on the lipid metabolism itself.

Peter Attia: But that has nothing to do with the endothelium. That has nothing to do where atherosclerosis occurs. All that’s doing is giving you more LDL and circulation.

Dave Feldman: Let me put it this way. Why not take anything that results in a higher level of LDL-C or LDL-P that doesn’t impact any lipid absorption from any tissue at all?

Peter Attia: Right. But that might be a bit of an artificial constraint right? As you pointed out yourself and I think anybody listening to this will appreciate, this is a complicated dynamic system, so it is going to be difficult to have some perturbation in a system that will lower or raise LDL that won’t have some other effect.

The question is, how do we, with some reasonable degree of certainty, look at those other effects, and ask whether or not they’re germane to the question of atherosclerosis and the causality of LDL to atherosclerosis. So, I think the PCSK9 example is not an unreasonable one, because we have a pretty clear understanding of what that gene does. We have a very clear understanding of where that protein lives, and what it’s doing-

Dave Feldman: But if anything, that’s resulting in the other direction. Where, if you have lower LDL-C or LDL-P from underexpression PCSK9. That actually results in a hyper-absorption of lipids…

Note from Tom Dayspring: May be confusing absorption (which occurs in the gut) with cellular uptake of lipids.

…for example.


Peter Attia: In the liver, yeah. They have enhanced hepatic clearance. So both ends of that though, right? So, if you have hyper-functioning and hypo-functioning…

Note from Tom Dayspring: Should use loss or gain of function terminology not hypo or hyper.

…PCSK9 patients out there, both of whom exist, I believe the hyper-functionings were discovered first. But the hypo-functionings are kind of the ones that gave the drug companies the desire to go in … Not the desire, the idea to go and create a drug to mimic that phenotype. But these patients walk around with LDL cholesterol of 10 to 20 milligrams per deciliter, and as far as anybody can tell, there’s no other side effect of that.

Dave Feldman: This is the thing I want to zero in on. Let’s say that we do that. Let’s that we go “Okay, never mind this side part of the lipid hypothesis end of it. I’m sorry, the lipid metabolism end of it.” We should then be able to look back at these people with the more novel versions of SNPs and assuming there’s at least a large enough population, we should see that longevity. Your mentioning of the apoC-III from earlier, is the first that I’ve been able to find of that one. I’m interested to see if we would see that across the board with these people who have these SNPs.

Peter Attia: Yeah. I mean, I suspect it will have to do with how many of them there are, and how long they’re being tracked.

Dave Feldman: Because, the all-cause mortality, you’d have to understand, I sympathize with your concern, as it’s absolutely the case, nutrition medicine, there’s certainly a lot of personalities that are out there. But, I can understand at least for me, on my end, I like hard endpoints over soft endpoints. Maybe it’s just the engineer in me. I like ones and zeros. Death is pretty easy to diagnose. Whereas, soft endpoints, the downside is there can be arbitrary decision making on the part of the patient and the doctor.

Peter Attia: Yeah, you know, I heard you mention that on one of the podcasts. I got to tell you. I disagree with that. Having seen more patients in an ER when I was in residency with MIs, I can honestly tell you, Dave, never once knew what their cholesterol levels were. When someone comes in the ER with chest pain, I care about the advanced cardiac life support algorithm, which involves oxygen, which involves an EKG, which involves troponin, which involves morphine, aspirin, and potentially a trip to the cath lab. But, we are, and nowhere in that algorithm, are we asking, “What’s their LDL?” And letting that help us think, is this indigestion versus other things. So, I do take issue with calling MI a soft outcome.

Dave Feldman: It’s not so much whether it’s a soft outcome. It’s whether or not there are things like, say revascularization…

Note from Tom Dayspring: Trials have strict rules on who may or may not get revascularization It is just not the whim of the card.

…That can be determined based on the decision on the part of the doctor and the patient, that may or may not have to do with their knowledge of the lipids, right?

Peter Attia: Agree. These are different things. But, I also think we should be careful not to take mortality as the only outcome. I will say this, and I hate putting on the stupid doctor hat, because it sounds ridiculous in this context, but unfortunately, I feel like I have to go back into and out of that world here. I would say at least half the patients that come to me, do not actually find themselves asking for an extension in lifespan. Right? My interest is longevity. But longevity has two components. How do you increase lifespan? Meaning, how do you delay death? And how do you improve health span? I won’t go into what that means, but, the bottom line is, there are many people who say, “I honestly have no interest in living one day long then I might otherwise live. But I want that quality to be much higher.” So, if we’re going to say … And, again, I don’t necessarily agree with that, I think the bigger issue is a statistical one with all-cause mortality, but, nevertheless-

Dave Feldman: But you go into modality. Like, if somebody has an MI, and it actually impacts their quality of life afterward.

Peter Attia: Right. What if your quality of life is decreasing as a result of a procedure, necessary or otherwise, or an MI, or a decrease in ejection fraction? Because remember, about half the people who first present with atherosclerotic coronary disease present with sudden death. But, half the people don’t, right? Half the people go through MI, stroke, God knows what else that follows. So, you know, again, I see it as my chief responsibility to delay the onset of death. If a patient decides that that comes at too great a cost, that’s great, that’s their decision. In the end, the patients decide everything. But, going back to what got us here, I am convinced that if patients didn’t have LDL, there would be little to no atherosclerosis, if you could give them no LDL over the duration of their life. If a patient comes to you and they’ve already got disease, and you lower LDL…

Note from Tom Dayspring: Cholesterol in their arteries. Type IIIs have no LDLs but plenty of atherosclerosis – but it is because in that rare case the remnants delivered the sine qua non – cholesterol.

…I don’t think that that gets them out of the woods. I think that that’s, sort of, just stochastically moving them in the right direction.

What we have to be careful of, kind of going back to that patient I was talking about, is we have to be able to identify the patients in whom the risk of LDL lowering is starting to cause a problem elsewhere. Meaning, they’re incurring an unacceptable risk elsewhere. In the case of this particular patient, who’s desmosterol levels had now become unmeasurable, my concern was we have overdone it with him on the lipid side, there’s a safe, easier way we can lower his LDL without impairing his cholesterol synthesis, because of the limited, but, to me, quite convincing data on the plausibility of Alzheimer’s disease in patients with overly suppressed cholesterol synthesis.


So, I want to be really clear when I repeat that. I am not suggesting that statins cause Alzheimer’s disease, which I know, the blogosphere loves to talk about. If anything, statins slightly increase the risk of diabetes in susceptible people, over a great period of time. But, at the population level, there’s actually no evidence that statins are causing Alzheimer’s disease. However, I think there are a subset of patients who are susceptible, and you have to be able to identify those patients. That’s the problem with population data, you can lose the nuance. The nuance is you, right? What matters to you, Dave?

In the end, I don’t really care what your LDL is, I care about you not getting atherosclerosis.

Dave Feldman: Right.

Peter Attia: If there is indeed someone walking around out there with an LDL cholesterol of 300, who’s not getting atherosclerosis, and there are indeed examples of that, then that’s great news.

But we have to, sort of, use this heterogenous population-based data to then try to probabilistically figure out, what do you want to do with somebody at the individual level? With this patient in the end, after a lot of fighting, did, decision-wise, we’re going to put him on Repatha, we’re going to start cutting down the Lipitor. Until we get that Desmosterol to bump. We’re still in the process of doing that actually. But, that’s the goal.

That, to me, is like an example of what I would think of is hopefully where precision medicine would be going. Which is, you’re now well outside of a clinical trial, right? There’s never going to be a clinical trial that’s going to ask the question, if you take a bunch of patients and statinize them ad nauseum, and you drive their cholesterol synthesis very low, and follow them for 30 years, do a subset of those people get it? No, you have to be able to look at retrospective data where those things were gathered. And we’ll link to, what I consider, one of the best papers on this topic.

Dave Feldman: But, getting back to the challenge. In a sense, you’re saying by ignoring the genetic data, that the genetic data basically answers the question, to your satisfaction. To where you don’t need to look at non-genetic-

Peter Attia: Not alone. I think of it as the genetic data, coupled with the pharmacologic data, coupled with the mechanistic data, give me a high enough degree of certainty, that I am willing to act in a certain direction. Remember, everybody, me, you, whoever’s listening to this, they have to make a decision.

Dave Feldman: Sure.

Peter Attia: Indecision is a decision. So, when you showed up with the hemoglobin A1C of 6.1, did you have type two diabetes? Nope. Your doctor said, “Hey, I’m cool just waiting.” But you said, “No. Indecision is not a decision anymore, I’m going to do something about it.” Because, presumably, you said, “Look, I have a family history of this. I think I have a sense of what the progression of it is. Quite frankly, I don’t want to wait until I have this disease to do something about it.” So, you decided indecision was not a viable decision.

Sometimes indecision is a reasonable decision. But, the point is, people have to understand they are making a decision whatever they decide to do.

Dave Feldman: Absolutely. Well, and for what it’s worth, as I say outside of here, and as I’ll say on this podcast, as I actually just said it, the speech, I don’t know if you saw the one that I did from last month. I told people, I prefer they not be echo chambering. I prefer they find everything that challenges from every side. So, with that said, going back to the lean mass hyper-responder, you would say, given what you know right now, given everything we’ve just talked about, that they are at high risk of cardiovascular disease. Would that be correct?

Peter Attia: I’d want to know more data. But, yes. If everything … If I didn’t know anything else other than-

Dave Feldman: Let’s say all cardiovascular risk markers say, LDL of 200 or higher, LDL-P of typically 2,000 or higher, everything else is just pristine perfect, like CRPs at the floor, their Lp-PLA2…

Note from Tom Dayspring: Not causal factors and not goals of therapy.

…may be-

Peter Attia: Let’s look at this patient here. So, we’ll link to these labs. I asked this patient, this is a patient I saw last week. So that the only reason I printed this up, because I see this so often, but I’m like, “Let’s just get the last one.”

This is a gentleman who’s been on a low carb diet for a couple of years, is achieving amazing success with it. He’s a new patient to me, but he’s been around the block on this stuff before, and he’s got an amazing history of his labs going back many years. I’ve seen what he looks like on and off all of these therapies. On and off drugs, et cetera. He’s one of these guys, where, across the board, looks fantastic, right? His glucose disposal is remarkable, his insulin levels are very low, his c-reactive protein is 0.3. Everything looks good. Read off some of his numbers, just for the folks, Dave. He doesn’t quite meet your lean mass, because his trigs might be a bit higher, but, talk to me about this guy’s numbers.

Dave Feldman: So, total cholesterol is 504.

Peter Attia: Is that high?

Dave Feldman: I know what you’re doing there.

Peter Attia: No, I’m just kidding.

Dave Feldman: I get this all the time where somebody sends me just that number.

Peter Attia: No, no. Okay. Go ahead.

Dave Feldman: Total cholesterol 504, LDL-C direct, and it’s worth emphasizing just real quick for the listener, when they say direct, it’s very important to notice that, because usually, LDL-C on a typical lab is actually calculated through the Friedewald Equation. So, when it’s direct, that actually is a direct measurement. And that matters for remnants. Hope we’ll get a chance to talk about that.

Peter Attia: We will talk remnants for sure.

Dave Feldman: So, LDL-C at 362, HDL-C at 94. Triglycerides at 125. The very first question I would ask if somebody was sending this to me is whether it was fasted or not?


Peter Attia: Yeah, this was. But, I’ve gone back and looked at all of his other trigs, and he actually, normally, does reside below about 70.

Dave Feldman: Oh, he does. Okay. So he would be typical for a lean mass.

Peter Attia: Yeah, he might have just eaten dinner a little too late, or something. I’m not sure what was going on.

Dave Feldman: Do you want me to keep going on the particles?

Peter Attia: Yeah, hit the particles.

Dave Feldman: So, apoB is 283?

Note from Tom Dayspring: 99.9th percentile.

…That actually is a little higher than I’m used to seeing. LDL-P is about 3,500. Small LDL-P is at 1,483. Small dense LDL-C is at 47.

Peter Attia: All right, so we’ll just stop there and come back to it. I’ve told you that everything else on this guy looks pretty good. Is this guy at risk?

Dave Feldman: Well, I’m actually looking ahead, because I would have cared about these other markers. That could indicate inflammation, so, for example, the fibrinogen is very high. Lp-PLA2 is above 600. I don’t … In fact, I think I just Tweeted about this recently. I don’t know that I’ve seen an Lp-PLA2 above 300 or 400, of the labs that have been sent to me. I don’t get a chance to interpret oxidized LDL, but you have the oxLDL as above 135. So, I would say by this lab, as it looks, I would be concerned about the triglycerides, I would ideally want the triglycerides to go down.

Peter Attia: But, is it your impression that if his trigs were normal, he would be okay?

Dave Feldman: I would be interested to see if the other inflammatory markers linked back to the reason as to why the trigs would be a bit higher. That would be something I would be very curious about. But, yes. If you were to say, I think what you’re trying to drive to, is if I had only the information of the lipid panel itself, and it did say, not 125 on the triglycerides-

Peter Attia: Let’s make it even easier. Let’s pretend that this patient had a zero calcium score.

Dave Feldman: Well, zero calcium score, I’m not entirely on the calcium score, but I do care about the calcium score. The thing is I would say if you were to give me the same numbers … Let me make this easier. I have Craig Moffitt’s, who’s very close to this, except his triglycerides are much lower. If it was the same one, then I would wonder if there really was a risk, yeah.

Peter Attia: Okay.

Again, that is a decision that every patient is going to have to make in that situation. In the case of this patient, I feel very strongly that he is at increased risk, though I think on many other metrics, I think his risk of cancer, he’s actually an apoE 2/3, this patient, so, his risk of dementia is going to be a bit lower, he’s metabolically quite flexible.

Dave Feldman: But, his other inflammation markers were low. Like, the Lp-PLA2, the ox-LDL, all this stuff that was below. Would you feel he was at risk. If the only thing that was different was-

Peter Attia: I would still feel he is at risk, because again, this is still one of the three legs of the stool. It’s the burden of lipoprotein, it’s the endothelial function or health, and it’s the inflammatory response to it. And I can’t measure number two very well. And even number three is pretty kludgy. Meaning, all of these things, like fibrinogen and CRP are not very specific. You have to, sort of, at the individual level be very careful you don’t draw too much of a false sense of confidence.

But, looking at him, he has among the lowest asymmetric, and symmetric dimethylarginine levels I’ve ever seen. These are staggeringly low. I would have expected those to be through the roof. So, ADMA and SDMA are things that we use to look at endothelial health. They inhibit nitric oxide synthase. When ADMA and SDMA are elevated, you’re inhibiting nitric oxide synthase, you have less nitric oxide produced in the endothelium, you’re more prone, obviously, to constriction.

Dave Feldman: I’ve never taken that test. I’d be interested in trying it.

Peter Attia: Again, just because I’m cheap and I didn’t want to print up a bunch of paper, I only printed two pages of this guy’s labs because those are the two that are most relevant. But, you can take my word for it, the others were exceptional. Right? This is a guy who looked really good across the board. The question is, should anything be done?

Now, if this patient had a negative calcium score, which he did not. But, if he did, I would have still recommended lipid-lowering therapy, and, or modification of diet. Why modification of diet? Because I’ve now seen more of these patients than I can count, and there is a pattern that is emerging. This is, I think I wrote about this in one of the cholesterol things. But, the pattern that always occurs in these folks, and I say always with a relatively small end. Maybe there are 30 of these cases I’ve seen, is this exact pattern. This one, he’s like a perfect example of. His desmosterol is very high.

Remember what we talked about at the outset?

Dave Feldman: Yes.

Peter Attia: When a patients LDL particle number is through the roof, you go through the checklist. Are their triglycerides high? No. There’s no way in hell a trig of 120 accounts for someone being above the 99.9th percentile of LDL particle. Does he have an LDL receptor defect? Turns out, I don’t think so, because I’ve seen his … Even though I haven’t seen his LDL-P, I’ve seen his LDL-C, of a ketogenic diet, and it was 125.

Dave Feldman: Right. Is he lean and or fit?

Peter Attia: He is.

Dave Feldman: Okay.

Peter Attia: Yup. So, his cholesterol synthesis is through the roof, and his cholesterol absorption is quite high as well.

Dave Feldman: Are these affordable tests? Because I would definitely want to turn these around to the existing group of lean mass-


Peter Attia: I’m sure, the cash cost on these is not onerous.

But, my point is, I think that the explanation for this phenotype is the up-regulation cholesterol synthesis from the saturated fat. I don’t think this is an energy issue per se. I think this is a sterol-regulated binding protein issue, or some sort of regulatory path around what the body is doing with ketones and/or saturated fat.

Tom Dayspring told me this a long time ago, and I totally forgot about it, and then the other day I went and looked up a case, because I never paid attention to this. We used to see patients all the time with diabetic ketoacidosis. These are usually patients with type one diabetes that come in the E.R. and, usually, it’s precipitated by some acute illness. But, basically, what happens is their glucose level becomes very high, they don’t have enough insulin, of course, they get a bunch of electrolyte abnormalities but they present with very high levels of ketone. This actually, is an emergency. All the talk about ketosis being dangerous, this is the example of where it is very dangerous, it’s life-threatening.

What I didn’t realize, is, I went back and looked, and sure, it turns out a lot of these people have very elevated levels of LDL cholesterol with total cholesterol. They also have elevated levels of triglyceride. But, of course, it’s hard to know exactly which one, or what’s driving that. But, once you correct this metabolic deficit, which is quite easy to correct. It’s basically potassium, IV fluids, glucose, and insulin, and you normalize their glucose levels, and their fluid balance, and their electrolytes, the cholesterol returns to normal.

It might be that the ketones themselves are a substrate to make more cholesterol. There are very … And again, we’ll link to a great paper that Tom wrote on lipidaholics several years ago, where he goes through the biochemistry of how saturated fats specifically, and ketones could in a susceptible individual produce this phenotype.

Bringing it back to this idea of genes, we might really be dealing with a subset of people, these hyper-responders, whoever, whatever percentage of the population they are, who are the people who are susceptible to this. Because you are not going to find a leaner person, exercising harder than I was when I went on a ketogenic diet. I never had this response.

Dave Feldman: But there is a distinction that I tend to find and this is Occam’s razor. Again, more theory. I’m actually going to be testing this myself in the next series of experiments that I’m doing. There is a difference between those people who are doing things like say endurance running, and weight lifting, or resistance training. In that, I think there is a greater overall gradient of receptor mediated endocytosis for muscle repair and growth. I could be wrong about that, but I’ll be very curious to see if that turns out to be the case when I’m doing it myself.

Peter Attia: Sorry, a greater amount of endocytosis, of which lipoprotein, and for which product?

Dave Feldman: Of LDL-P in particular.

Peter Attia: Into muscles?

Dave Feldman: Yup.

Peter Attia: For what product?

Dave Feldman: For repair and growth.

Peter Attia: You’re saying that in these people, they’re relying on their LDL for cholesterol delivery to the muscle?

Dave Feldman: Well, and phospholipids and just about anything else that would be inside of an LDL particle…

Note from Tom Dayspring: Chylos and VLDLs and phospholipid transfer proteins deliver the PL, not LDLs.

…There are existing studies that are out there, as far as those people who do a lot of weight training will also see LDL-C and, this is why I’m saying it’s completely theoretical. I’ll actually be testing this myself over the next few weeks…

Note from Tom Dayspring: How is he going to measure muscle uptake of PL or LDL?

…Because I’m actually going to be eating a very fixed diet, fixed sleep schedule, fixed everything. Then I’ll actually be introducing basically, any way in which I can get my muscles sore in a very fixed fashion, I can then turn around this data. If the hypothesis is true, I would expect that my LDL-C, my LDL-P might change.

Peter Attia: But, I’m confused. Why is the runner’s muscle more demanding than the weight lifter’s muscle? Or vice versa.

Dave Feldman: The other way around. That I would see the weightlifter actually seeing a difference. Because I think there’s more use of the product of LDL-P directly by the cells. I may be wrong about that.

Peter Attia: But, what’s the evidence that that’s happening?

Dave Feldman: The evidence as far as … The Ketogains groups. I’m sure you’ve heard of them?

Peter Attia: No.

Dave Feldman: There’s a ketogenic group, that’s ketogains, there are not as many lean mass hyper-responders that come out of that group. They’ll tend to see their LDL-C go up, but not as pronounced as those people who are, say, runner-types, or aerobic-types, or even people who are doing yoga. There seems to be actually, a more pronounced difference of higher LDL-C, depending on how much you’re doing resistance training, or anaerobic training.

Peter Attia: Yeah, again, I’m not aware of any evidence to suggest that the muscle is relying on LDL for delivery of anything. Including energy.

Dave Feldman: I’m not so sure about it on energy. What I’m thinking about is in terms of just raw material. I mean, as far as damage that can happen to, for example, the membrane of a cell…

Note from Tom Dayspring: Muscles synthesize all the cholesterol they need.

…I realize this is kind of a key difference between us, is that your sense is that effectively, anything that the cell is going to need, it can basically synthesize on its own, right?

Peter Attia: No, I think my sense is that … Occam’s razor would at least have me start from a place of plausibility. And I’m just not aware of any data that suggests that LDL is functioning to do this.


Dave Feldman: What’s the value of non-hepatic receptor-mediated endocytosis from your perspective?

Peter Attia: So, you’re talking about very specifically, the little bit of LDL that gets out of circulation, either with or without a receptor to non-hepatic tissue?

Dave Feldman: Yes.

Peter Attia: My sense is the most important value of that would be to tissues that need more cholesterol to synthesize hormones.

Dave Feldman: But specifically cholesterol and not the phospholipids or anything else that-

Peter Attia: I think the phospholipids probably may be more delivered through others. Certainly, the VLDL delivers far more phospholipid than LDL. But, LDL is really a custom built package for cholesterol. If you look at how many cholesterol molecules fit inside and LDL particle, versus even an HDL particle … Remember, the HDL is the general of RCT. Yet, it can still only carry about 50 molecules of cholesterol. The LDL particle can carry 1,500 molecules of cholesterol. That’s staggering, again, when you consider the size of these things, right? It’s tailor-made for that. That is largely conserved.

I don’t want to get us too far in the weeds, but I actually did a very interesting kinetic experiment many years ago. I did three blood tests every day for three days. Like the full NMR panel, but this is with kinetics. So, this is not commercially available. So, what you’re looking at is my ability to track, you’ll have to lay it down, because I barely remember what I did, but this is pre-workout, immediately post-workout, four hours later, looking at my LDL particles, my VLDL particles, my HDL particles, both in terms of their cholesterol and triglyceride content.

Dave Feldman: So, you see them going down yourself?

Peter Attia: I don’t see any change in the cholesterol content. It’s minimal change in cholesterol content, right? What I think you see here is, “Yeah, wow. Under really periods of super high-intensity exercise, I actually did take some triglycerides out of this.”

Dave Feldman: Right.

Peter Attia: Minimal out of here. By the way, this backs up Garvey’s data, which is, there’s virtually no way to distinguish what’s going on at the VLDL level. We can’t tell what a remnant here, or what’s not a remnant. I apologize for the listener, we’re looking at a chart, but we’re going to link to it, so you’ll see it. We’re basically talking about this idea of how much movement of cholesterol is going into, and out of, the LDL particle under these extreme conditions. I just did different types of workouts. So, on this day I did a crazy high-intensity interval training. On this day, I did a crazy intense swim. And I think on this day was the hardest workout of them all, was a crazy intense bike ride.

Dave Feldman: And the listener can’t see this, but I’m smiling ear to ear. It’s almost as if you knew I was going to-

Peter Attia: Well, I thought you would appreciate … I forgot I did this. I did this six, seven years ago.

Dave Feldman: That’s fantastic.

Peter Attia: Again, it’s not commercial, these are not assays that can be reproduced. Again, I suspect you and I will look at these differently, right? I’m looking at these to say the VLDL is definitely moving its triglyceride, the LDL, a little bit. The cholesterol is barely moving. This is different from a broader issue, which is how much did my cholesterol actually change in those nine blood draws? That was the very ability of my LDL-P, so we’ll link to a graph that shows LDL-P versus LDL-C. These are nine points, across three days.

Note from Tom Dayspring: If one does not do radioactive kinetic studies one has no idea what various pools of lipids were doing. Lipid or lipoprotein measures cannot analyze that.

Dave Feldman: I wish I could go back in time, find you, what is this 2012? 2013? Something like that?

Peter Attia: Yeah, probably 2011, or 2012.

Dave Feldman: I’d be like, “Peter, this is Dave from the future. I need you to eat exactly the same thing you do on all these days.”

Peter Attia: I pretty much did.

Dave Feldman: Oh, you did?

Peter Attia: Yeah, yeah, yeah. I controlled for that.

Dave Feldman: Okay, good.

Peter Attia: And I was timing exactly when I would eat it. This was back when I was very strictly in ketosis.

Dave Feldman: Fantastic. That’s effectively what I’ll be doing. So, I’ll actually have … I’ll add a two to your N, but I’m super excited to be seeing this now.

Peter Attia: Yeah, but you’re not going to have, unfortunately, this. I mean, this-

Peter Attia: Yeah, but you’re not going to have, unfortunately, this. I mean, this to me is the interesting part, this is the part that surprised me the most, was how little the cholesterol was actually moving out of the LDL. Even when the particle was going down, and remember the height of this bar is artificial. It doesn’t mean anything. This is not the LDL-P. This is milligrams per deciliter of cholesterol of triglyceride.

Dave Feldman: But how do you know about the … maybe I’m getting a little confused here. How do you know about the cholesterol moving out of the LDL-P?

Peter Attia: Well, what I can say is before the workout, there are 116 milligrams per deciliter of cholesterol in the LDL, and after the workout that’s what was there. This is how much triglyceride was in the LDL, that’s what was there.

Dave Feldman: Okay.

Peter Attia: It’s sort of a flux.

Dave Feldman: Right. Here’s the thing, I mean in a sense, and I know you can acknowledge this as well. We don’t exactly know what the true circulatory level of recycling is. You’re just saying on a per particle basis for testing it afterward. For example, we’re talking about the pool, not the individual particles in that sense.

Peter Attia: Yeah, it’s an average. It’s basically saying if you took four snapshots in a room and how many people had blue shirts on and how many people had red shirts, and you saw the deltas, you could not infer the actual numbers that went in and out, but just the net delta. There is either a net influx or efflux of people in blue shirts or red shirts.


Dave Feldman: Right.

Peter Attia: That’s the best we can do. Now, there may be some kinetic studies that could do even one better, but that’s, to me, pretty interesting stuff.

Dave Feldman: Let’s circle back to remnants real quick.

Peter Attia: Okay.

Dave Feldman: This is why I pause a little bit on the case study that you showed me where you had the triglycerides a bit higher. Now, as you know, what is the poor man’s version of remnants is you basically can just take your triglycerides divided by five. You probably actually …

Peter Attia: No, no, no. We got to be very clear on this stuff. We’re going to confuse the hell out of people.

Dave Feldman: Okay.

Peter Attia: That’s the poor man’s version for VLDL cholesterol.

Dave Feldman: Correct.

Peter Attia: Very important distinction.

Dave Feldman: Fair point. Yes.

Peter Attia: The poor man’s version, which should never be done because it is such an abomination is to take the triglyceride level divided by five, and that number would be an estimate of your VLDL cholesterol.

Dave Feldman: Now, what would you, Peter Attia recommend as the most effective means by which somebody could determine their remnant cholesterol?

Peter Attia: We can’t. It’s impossible. We have no way of knowing remnant cholesterol. Let me be clear. I think I know what you mean by remnant, which is why I’m asking that question. You’re asking pathologic remnant. You’re asking VLDLs that have shed their triglyceride and are now basically pathologic…

Note from Tom Dayspring: Basically small and cholesterol-rich.

…small. Started out as big, triglyceride-rich, and now have shed that through their apoC-II-to-LPL pathway, and now have the potential for atherosclerosis, is that what you’re meaning about remnant?

Dave Feldman: No, I now think we may think of it differently. Straight up Wikipedia right now would define remnant cholesterol as basically all cholesterol that’s not in either an LDL particle or in an HDL particle. If you were to just subtract HDL cholesterol and LDL cholesterol.

Peter Attia: If you directly measure total cholesterol, which you can and you can directly measure LDL cholesterol and you can directly measure HDL cholesterol, you subtract those two and you have the amount of cholesterol that is virtually all in a VLDL, and presumably some IDL, if it sticks around.

Dave Feldman: Some IDL, possibly chylomicrons remnants if you ate recently, but you shouldn’t have any chylomicrons remnants.

Note from Tom Dayspring: True pathological chylomicron remnants have significantly increased plasma residence time.

Peter Attia: Yeah, that’s very easy to exclude.

Dave Feldman: Right, but effectively, if you’ve had a fasted cholesterol test, pretty much all your remnant cholesterol, pretty much, will be in VLDL has longest residence time relative to the IDLs.

Note from Tom Dayspring: True – but we have no test to identify them.

Peter Attia: That’s correct, but that’s like telling me outside a very few pathologic states, like Fredrickson type-III Bs, that’s as interesting to me as your eye color.

Dave Feldman: The remnant cholesterol?

Peter Attia: Yeah.

Dave Feldman: Really?

Peter Attia: Yeah. I mean it’s generally going to be very low. It tracks quite well with triglycerides, though there are lots of examples where it’s been … Actually, I think I brought a copy of one of my other goofy experiments. I don’t know. I probably won’t find it anytime soon, but there is an example of how mine was so far off. It was like a 700% delta between actual versus predicted in one of these studies, but now, the point is yeah, directionally speaking, I would love to see a VLDL cholesterol below 15 milligrams per deciliter as calculated by taking non-HDL cholesterol subtracting the LDL cholesterol, but if you have a direct LDL, cholesterol is your best measurement of that.

Whether it’s 10 milligrams per deciliter, 15 milligrams per deciliter, 20 milligrams per deciliter, that just tells me the sum of cholesterol in all of my VLDL remnants. It tells me nothing about the pathology of them. It tells me nothing about what they’ve done or where they’re going.

Dave Feldman: But, that said, the question then becomes even for as much as what you just qualified, does that become a more powerful predictor relative to something like, say, LDL cholesterol?

Peter Attia: Maybe, but that’s like saying is rubbing two stones together better than rubbing two logs together to start a fire? It’s like why not just use a Zippo lighter? It’s like we could split hairs on whether non-HDL cholesterol or remnant cholesterol is a better predictor of cardiovascular disease than LDL cholesterol, but again, given that LDL cholesterol is such a crappy predictor of cardiovascular disease, I prefer not to really even think about that.

Dave Feldman: We do have now a situation where this particular phenotype where lean mass hyper-responders will have very low levels of LDL or I’m sorry, of …

Peter Attia: VLDL cholesterol.

Dave Feldman: VLDL cholesterol, have very high levels of LDL cholesterol, will have very low levels of remnant lipoproteins.

Peter Attia: But we don’t know that. There’s no way you’ve measured that. I’ve never measured that. That’s not measurable. In a commercial, I say that’s worth its salt. I think VAP does a vague ass version of that, but it’s bunk, but look at this. Look at Garvey’s studies. We’ll link to this as well. This is actually measuring the number of particles. This is an insulin-sensitive person. Their total LDL-P is about 1,200. Their VLDL-P is about 80. You go to someone whose insulin resistant, but not diabetic. Their LDL-P goes up to 1,435 on average. Their VLDL goes to 84. Their IDL is counted. It’s around an area, and then you take the population with type 2 diabetes. Their LDL cholesterol is up to 1,600 nanomoles per liter and their VLDL goes up to 100.


You’re right, the VLDL is going up as you get more insulin resistant, but it does not appear very clinically relevant, because remember this is all the burden of disease, is from these apoB bearing particles, and so the increase in VLDL particle number is not what’s driving the risk of the disease. You actually had a really nice graph in one of your figures that you titled, that is remnant that was going up. You know the figure I’m talking about?

Dave Feldman: Yeah.

Peter Attia: We’ll link to that as well, but you had a graph that showed I think as people were becoming more insulin sensitive, their pool of remnants was growing. What you perceived as remnants.

Dave Feldman: Well, I actually want to qualify something real quick. If I think it’s the graph that you’re talking about, it was the one downside to that is it was non-fasted remnants, which I’ve been trying to find fasted remnants.

Peter Attia: Oh, got it. Got it. Okay, okay.

Dave Feldman: Which I have a problem with non-fasted remnants.

Peter Attia: It was in one of your talks.

Dave Feldman: It was in one of my talks.

Peter Attia: But anyway, my point is what’s missing from that analysis is apoB or LDL-P. In other words, the expansion of that … See this study, which is I mean the most elegant study of this ever done shows that if you only saw the top line, this to this to this, everyone would be like wow. The more insulin resistant you get, the more your total burden of apoB goes up. What this is showing is where is that burden coming from. The VLDL only increased by 20 nanomoles per liter, but the LDL-P increased by 400 nanomoles per liter.

Dave Feldman: Right, and the point I’m coming to is this is where we’re in uncharted territory, because I believe will apply to people who are ketogenic fat adapted. I don’t believe this will apply to people who are very ketogenic fat adapted, and particularly who have these phenotype.

So, let me see if I can come at it this way, right now, if I were to be able to get the data set for Framingham Offspring, because that is one of the studies that I was showing out for before, and I could basically just do this basic calculation of remnant cholesterol the way we were talking about, and conceding the point of your time that you said earlier that’s there’s no way to truly know. Would I still come up with a more valuable metric for subtracting HDL-C, particularly when associating it to all-cause mortality? Relative to say LDL-C-

Peter Attia: Let me make sure I understand what you’re saying. You’re saying if you could develop an assay to distinguish between the pathologic remnants of VLDL versus the physiologic remnants …

Dave Feldman: I’m not even developing that. Let’s say I’m not even developing. I’m just taking the existing HDL-C and LDL-C metrics as they’re recorded right now. I’m just grabbing the data says it is right now.

Peter Attia: Yeah.

Dave Feldman: Will the remnant cholesterol that I get from that subtraction, will that actually be more relevant to all-cause mortality than say LDL-C?

Peter Attia: I don’t know, but it would certainly rival it. I mean, again, the data are probably clearer on non-HDL cholesterol versus LDL cholesterol, that’s typically what the literature talks about, but as you can tell that non-HDL cholesterol and LDL cholesterol are two of the three variables you would need. There’s a strong correlation between non-HDL cholesterol and remnant cholesterol, and yes, I believe non-HDL cholesterol is more predictive than LDL cholesterol.

Dave Feldman: But there’s one problem with non-HDL cholesterol I definitely want to bring up for people who are on a low carb diet. If you’re going to be powered much more by triglycerides directly, literally triglycerides being brought to you in VLDLs, then that is going to be relevant. Now, again conceding Peter, what we talked about before, I can’t know if I can stake my flag on it just yet, but I’d be willing to bet if I’m right on the energy model that if fact you are being powered more by triglycerides found in VLDLs, then you can have a higher resulting LDL particles.

Note from Tom Dayspring: Yes, folks with TG-rich LP usually have sky high LDL-P.

Peter Attia: But, wait. That’s a … sorry, I’ll let you finish, but I think your missing something. Go ahead.

Dave Feldman: No, no. Go ahead.

Peter Attia: No, no. I’ll remember it. I feel bad I interrupted you.

Dave Feldman: Okay. No worries. Effectively, the bigger question is, are particular lean mass hyper-responders showcasing directly that they’re being powered much more by triglycerides brought on these VLDL boats, if you will, and therefore having more subsequent LDL particles.

Note from Tom Dayspring: Yes, folks with TG-rich LP usually have sky high LDL-P. But usually, their TG levels are high.

Peter Attia: Okay. My hypothesis is that that is not the case.

Dave Feldman: That it’s the higher degree of synthesis, right?

Peter Attia: Yes.

Dave Feldman: Going on with the cholesterol.

Peter Attia: That’s correct. That it is the higher degree of synthesis, which may or may not also be matched by a higher degree of absorption.

Dave Feldman: How would, if you were wanting to, if you’re going to suggest a way that we could test this, how would you suggest that?

Peter Attia: Before I do that, let me unpack where I think energy is moving from. I think we all agree that someone who is very insulin sensitive on a low carbohydrate high fat diet is utilizing a lot of triglycerides. We agree on that, right?

Dave Feldman: Yes.

Peter Attia: Okay. Let’s take an artificial construct and separate endogenous from exogenous triglyceride. Meaning someone on a low carbohydrate diet is eating themselves, and they’re eating triglycerides from the outside world.

Dave Feldman: Right.

Peter Attia: Assuming they’re in the phase of getting leaner. They’re losing weight or even someone who’s weight’s static, they’re utilizing their own internal stores of triglyceride that they’re replenishing. If they’re staying weight stable, right?

Dave Feldman: 100%.


Peter Attia: Okay. The exogenous triglycerides enter the body through chylomicrons, that’s a pure lymphatic play, through CETP that’s rapid hydrolysis. I think we all get that. I think it’s this other endogenous pool that’s interesting.

Dave Feldman: And if I could just interject this one thing, because this is one thing we’re dancing around that we both know that probably somebody’s who’s not familiar with my work should be aware of. Chylomicrons, they drop off these triglycerides and they’re just gone.

Note from Tom Dayspring: They are cleared via apoE receptor paths.

Almost like depending on who reading within minutes to hours, at the most. If you’re taking a fasted cholesterol test, Peter, we were talking about it earlier, you shouldn’t see any chylomicrons or chylomicron remnants…

Note from Tom Dayspring: Like all remnants, chylo remnants have increased plasma residence time.

…they should be gone, and the cholesterol payload on those chylomicrons should be gone.

Peter Attia: That’s correct.

Dave Feldman: With that in mind, go back to the endogenous triglycerides.

Peter Attia: Endogenous means we’re dealing with the pool of triglycerides that are coming out of you, as the person. You have adipocytes, adipocytes stored triglycerides, and those triglycerides are hydrolyzed such that you have free fatty acids that will be transported. Where did they go? When they come out of the adipocyte, who picks them up? Albumin. Albumin then does two things. It can take it directly to the muscle, so that the muscle can use it in the highly fat adapted athlete, or it can take it back to the liver, and it can be repackaged in VLDL, or it could be turned into a ketone, if we’re getting into an extreme state of someone who’s ketotic.

Let’s talk about Craig Moffitt who looks like a super fit dude, who’s running around, has a total cholesterol of 457 milligrams per deciliter. His LDL cholesterol is 335. His HDL cholesterol 109, and his trig is 67. I’m assuming you did the math correctly. I’m not going to check it, but if that’s presumably the LDLs direct, his remnant cholesterol is 13 milligrams per deciliter, and by the way, that’s not terribly far off from what you would get by the trig [divided] by five formula. He’s not that far off. Okay, fine.

Where is he getting his energy? Let’s say he’s out for a run. He’s not eating anything and he’s fat adapted, and he’ll say he’s fasted, let’s make it even easier. He’s fasted going out for a run. His adipocytes are releasing free fatty acids to albumin. The albumin is taking some fraction of that to the muscle directly, and they’re undergoing beta-oxidation there. The albumin’s also going back to the liver, and some amount of that is being converted into beta-hydroxybutyrate, which goes down its own metabolic pathway, and some amount of that is being packaged in either a VLDL or an IDL because remember there’s still de novo IDL production in the liver, just as there’s the de novo LDL production.

Those VLDLs and IDLs are leaving the liver and dropping off their payload of lipid to that tissue. The tissue’s basically getting ketones from the liver, triglyceride from albumin directly to the muscle and triglyceride through the VLDL and IDL directly to the muscle. Do we agree on that?

Dave Feldman: We do. I’m just going to expand a little bit on what you just said. So, yes, it’s full body lipolysis, and that he’s releasing the free fatty acids. In the literature, they’re usually calling them NEFA, non-esterified fatty acids…

Note from Tom Dayspring: NEFA and FFA are identical terms.

…but we’ll just keep it to free fatty acids. It’s getting released from all over-

Peter Attia: We usually measure them by the way as one.

Dave Feldman: Yeah. It’s a little frustrating because a lot of these terminologies gets interchanged, but the free fatty acids are ultimately making it back to its liver getting packaged into the VLDLs…

Note from Tom Dayspring: Albumin trafficked FFA can be brought to any cell that needs it for energy.

…What we’re talking about the target sites of the muscles for which are making use of triglycerides. I should emphasize that I believe that the primary purpose of the creation of those is to replete everything. It’s not just to fuel the muscles, it’s also to put it back into the adipocytes that just now released it as well.

In other words, Craig Moffitt like many people who are lean mass hyper-responders, if we could install a little turnstile into their adipocytes. We would see that turnstile just spinning like crazy. They barely park the triglycerides there before it’s heading right back out…

Note from Tom Dayspring: TG do not head back out- they are hydrolyzed to GFFA which exit and join albumin.

…and that’s because there’s less total adipose mass overall in Craig Moffitt compared to somebody who’s a lot heavier, and therefore there needs to be more global supply of VLDLs relative to somebody else who has a lot more fat mass.

Peter Attia: But this depends on his energy requirement?

Dave Feldman: Absolutely.

Peter Attia: I mean, of course, while he’s running, and again, I wrote a blog post on this a long time ago, which I guess we ought to link to. It’s something about fat flux. I don’t remember the name of it exactly, but the gist of it was oversimplifying a fat cell as having two input doors and one output door. The two input doors as being the de novo lipogenesis door, which is still an esterified entry door, but I separated as a different storage. Meaning it’s coming from a carbohydrate, not from a fat, and then you have the re-esterification door, which has the turnstile that allows fat to go right back in, and then you have the lipolysis door, which allows the fat to exist. A person who is in fat balance has a situation where L, lipolysis equals the sum of the esterified de novo lipogenesis plus the re-esterified fatty acids. Agreed? That’s just straight up mass balance.


Dave Feldman: Yes, yeah.

Peter Attia: When Craig’s running, he’s in negative fat flux. Make no mistake about it. His de novo lipogenesis is zero at that moment. His esterification is something, and his lipolysis has to be something bigger.

Dave Feldman: Right.

Peter Attia: If he’s not depleting glycogen, which if he’s highly fat adapted, he’s not.

Dave Feldman: And maybe it’s worth putting on a distinction. I’m not talking about whether they were successfully, at the moment that he’s running — successfully re-esterifying these fatty acids back into the adipocytes.

Peter Attia: And they might be though. That’s my point, because we don’t know. What we know is that in, which goes back to the question …

Dave Feldman: But do we know that it’s the job of the liver to keep maintaining that buffet? To keep putting that energy back out there, and generally speaking, we do know that. It’s just to what degree?

Peter Attia: No, I mean my hypothesis is yes. My hypothesis is, which is not, by the way, I don’t think is a commonly held view. I think a lot of people would disagree with me, but my view is that the liver is the energostat of the body. I borrowed that term from Mark Friedman, who wrote an amazing chapter on this in 2008. That has been one of the most influential things in my thinking on appetite, but he described the liver as the energostat. It was, an engineer will appreciate the nomenclature. It looks kludge to someone who’s not an engineer to say why would he call it an energostat, but in engineering speak, that makes perfect sense, but that the liver is probably most susceptible to detecting some currency of circulating energy and circulating metabolites.

ATP would be the most logical thing for it to be sensing. Probably ratio the ATP to ADP, or ATP to AMP, or ADP to AMP, something like that, but yeah. I think the liver … I mean, I think most people appreciate how impressive the liver is in general. I was just talking to a patient this morning, and I said, “Look, man, here’s the deal. Anything that goes wrong with you can be supported extracorporeally. You get into a coma, no problem. You need to go on a left ventricular assist device, okay it sucks, but it’s there. You need dialysis, you need a ventilator, all that stuff, we do not have extracorporeal support for the liver. It is too complicated.

Dave Feldman: Anybody who follows me knows just how much I’m loving this because you’re the preacher’s preaching to the choir by far. I’ve often referred to the liver as the straight-laced partner, who always puts up with your crap. Whatever you’re giving it, it’s having to pour us out and figure out, and balance the ledgers, and get everything in.

Peter Attia: No, no. That’s a great point, and that was part of the other point I made to this patient, who was not in any way opposing that view. He was just asking if the elevation we’d seen in his liver function tests, which was mild could explain a synthetic issue to which the answer is not a chance in hell under normal circumstances because the liver has an enormous capacity to do its job under even the most ridiculous stress.

Yeah, but going back to Craig, at the moment that he is running, which is the same as saying if someone’s losing weight while they’re on a low carb diet, they are in negative energy balance. They are in negative fat flux, and again, when I say losing weight, let’s ignore the water weight and stuff. I’m talking about legitimate weight loss or shedding of …

Dave Feldman: Metabolism exceeds.

Peter Attia: Yeah. Yeah, but that’s what it means, right? On a practical level, it means lipolysis, the amount of fat that is leaving the fat cell has to exceed that which is re-entering it, and again, I don’t know that this is entirely relevant, but you’ve eluded to it, so worth reiterating. Not all of that is oxidized. Some of that free fatty acid leaves doesn’t get oxidized, and guess what? It’s mopped back up provided the hormonal milieu still permits it.

Dave Feldman: Yes.

Peter Attia: We agree on that completely. I look at his remnant cholesterol of 13 milligrams per deciliter, and say okay, it doesn’t tell me anything. I apologize. Your question was looking at that, what could we infer?

Dave Feldman: Yes, actually, and this probably gets to just a larger problem that I feel like remnant cholesterol is helping us to address is why is it that anybody would have high triglycerides at all? Why aren’t all triglycerides making their way to either the tissue that’s using it immediately, the skeletal muscle or the cardiac tissue, or to the adipocytes if the body means to not have it sitting inside of lipoproteins part in your bloodstream?

Peter Attia: That’s a totally separate question. I want to come back to this remnant cholesterol question, though.

Dave Feldman: But this is why the original graft that we talked to, this is why I draw that dotted line. I draw a lot of people’s attention to it is …

Peter Attia: Can you go back to that?

Dave Feldman: I’m sorry. I’m trying to find another one. Okay, yeah. This is the core emphasis.

Peter Attia: This is the energy delivery support diagram for the person who’s going to be looking at this later?

Dave Feldman: Right.

Peter Attia: Okay.

Dave Feldman: I mean, if I’m going to way oversimplify it, but not by too much chylomicrons’ job deliver fat-based energy.

Peter Attia: Yeah, and is so far gone that it’s not really entering the discussion we’re talking about outside very, very rare diseases.

Dave Feldman: HDL, not to deliver energy will just say, things not related to delivering energy, but I just call it support, generally speaking. Operations not related to delivering energy.


Peter Attia: I agree with that.

Dave Feldman: Okay. Last line. Liver, this being the apoB containing lipoprotein is the one lipoprotein that clearly is pulling double duty.

Peter Attia: No, no, but hang on. Remember this is where your diagram at which you acknowledge is oversimplified. The oversimplification is hurting you. The liver has three purple arrows coming out of it, the VLDL, IDL, LDL.

Dave Feldman: Agreed, which is why earlier I was emphasizing that I believe there’s a higher secretion of VLDLs overall for those people with lean mass hyper-responders, which is very relevant to our discussion.

Peter Attia: How do we know that?

Dave Feldman: It’s a theory. I’m not saying that. What I’m saying is as if this area was fatter, if the VLDL secretion is at a greater degree, it would make sense why there would be more remodeled final LDL particles remaining, and why would we see the inversion pattern in the first place? Because it originated in order to deliver more of those triglycerides, which was brought about.

Peter Attia: Why the excess cholesterol? If that were the case, Dave, wouldn’t you hypothesize that the LDL particle would be very high because you have more VLDL particles, but they’re shedding all of their triglycerides in an effort to deliver their energy payload, but you shouldn’t have an increase LDL-cholesterol. You should actually have a reverse discordance from what we see in the insulin resistant patient where we typically see the LDL particle number, being this proportionately higher than the LDL cholesterol by percentiles, of course, and absolutely members, that’s always the case. In other words, if I was buying your hypothesis I would say the LDL cholesterol should be very low. You should have very cholesterol-depleted skeleton particles that were mostly used to shed triglyceride as VLDL.

Dave Feldman: But instead, we see the concordance. We see the LDL-C and the LDL-P very concordant in people who don’t appear to have other types of diseases.

Peter Attia: Well, at that point they’re so high. It’s hard to know, but clearly not discordant in the direction that would make sense given your hypothesis. In other words, what I’m getting at is, why is there so much cholesterol in those LDLs.

Dave Feldman: Correct me if I’m wrong, but as far as the actual drop off rate, the LDL-C is still going be relatively standard on a per particle basis in a healthy subject. How much variability is there typically in LDL-C per LDL particle because and again, correct me if I’m wrong, I thought that the secretion level tends to be fairly standard…

Note from Tom Dayspring: It’s not.

…like a spare tire is standard for a car.

Peter Attia: Yeah, it’s …

Dave Feldman: The triglyceride levels can be very variable.

Peter Attia: Yeah, but the cholesterol levels. I mean they have the capacity to carry a lot, but think about it. If you have large and small particles that even for the same amount of triglyceride have different amounts of cholesterol, but the bigger point is, where is the cholesterol coming from? If we go back and look at my guy, or look Moffitt. Moffitt’s LDL cholesterol was 335 milligrams per deciliter.

Dave Feldman: Right, but that’s what’s in circulation in the blood at that time.

Peter Attia: Hang on. Just to be clear, Dave, is there any point during his 24 hour day when that number is 30 milligrams per deciliter of LDL cholesterol in his bloodstream.

Dave Feldman: I don’t believe so.

Peter Attia: In other words, if you take the area under the curve, if we could get real-time LDL cholesterol number on Moffitt, and integrate him over 24 hours, we could argue for argument’s sake, he’s going to be always over 300 milligrams per deciliter. Hus AUC would be very high.

Dave Feldman: For the existing traffic of the LDL particle, I mean basically what we’re talking about is what’s, I almost want to say it in terms of birds in the air. You have so many ships that have left the dock, that are continuing in circulation, but I guess here’s a different question that I’ll pose back to you.

Peter Attia: But, wait. I’m actually asking this because I’m trying to understand it.

Dave Feldman: How much of hits cholesterol has already made a lap? Are you thinking of this in terms of it all getting synthesized and then reabsorbed, and then recreated it again.

Peter Attia: Let’s go back and make sure we’re agreeing on the same conditions here. Notwithstanding the experiment where they eat a ton of fat and they go from having incredibly high LDL to very high LDL, but let’s just take Moffitt, over the course of a week. Assume you could do real-time LDL cholesterol sampling on him.

Dave Feldman: And specifically LDL cholesterol?

Peter Attia: Yes.

Dave Feldman: Okay.

Peter Attia: Or everything. Everything that’s on your page. You could sample his total cholesterol, his HDL cholesterol, his LDL cholesterol, his VLDL cholesterol, and his LDL particle number. You could sample all of these things in every second for a week. I think we’re agreeing that he will always have a very high LDL-P and a very high LDL-C and a low VLDL-C. Correct.

Dave Feldman: Unless he eats a lot of fat.

Peter Attia: Okay. Yeah. Let’s, we’ll come back to that after, but yes.

Dave Feldman: Sure.

Peter Attia: Okay. So, given that the half-life of his LDL is a day, where is that extra LDL cholesterol coming from?

Dave Feldman: I believe it’s being recycled.

Peter Attia: Well, it’s always being recycled. How is his being recycled? So, so where is his, where is he deficient in cholesterol that a person who has an LDL cholesterol of 100 is not?

Dave Feldman: I guess I don’t understand the question. Where is he deficient?

Peter Attia: Yeah, if he’s got 335 milligrams per deciliter of cholesterol in his LDL particles, are you telling me that he has less cholesterol in his cell membranes or less of it somewhere else?


Dave Feldman: No.

Peter Attia: So he has more cholesterol in his body.

Note from Tom Dayspring: Wrong: serum LDL-C has no relationship to body cellular cholesterol.

Dave Feldman: Correct.

Peter Attia: Why?

Dave Feldman: For the same reason that we would have, say, life rafts on a boat and once we have more boats, we have more life rafts. So, if we had a harbor just outside this window, right? And you had 100 boats and on those hundred boats, their main job is to deliver something unrelated to life rafts. They’re delivering cargo to the other island. Right? And they deliver, and 100 of them go out, 100 of them come back. And then demand on that island has changed. Now they need to deliver five times as many things as they were delivering before.

Peter Attia: But the problem with that analogy is it assumes a completely fixed number of life rafts per boat.

Dave Feldman: It definitely does.

Peter Attia: But that’s not how the lipids work.

Dave Feldman: How does it work?

Peter Attia: There’s much more flex in the system. And furthermore, you could ask the question in reverse, why isn’t it higher?

Dave Feldman: Why wouldn’t LDL-C be higher per boat?

Peter Attia: No, why wouldn’t LDL-C be higher in that patient? In other words, what’s regulating it? What’s regulating how much LDL cholesterol he has?

Dave Feldman: The demand for the boats themselves for LDL particles.

Peter Attia: But what’s driving that demand? This is, I think where we differ, right?

Dave Feldman: The demand of the delivery for the triglycerides. The demand is for the cargo, the originating cargo that is clearly getting used.

Peter Attia: Right, but we agree that the VLDL vanishes very quickly.

Dave Feldman: The VLDL remodels to LDL very quickly.

Peter Attia: Yes, but if you go through the kinetics of this, I can’t follow why he should still have that much LDL cholesterol unless he is making more cholesterol. In other words, I’ve tried to think of this 10 ways to Sunday, the only way on a mass balance that I can explain this hypothesis is if he’s making more cholesterol.

Dave Feldman: And not if he’s recycling the same cholesterol? Certainly, he’s making more relative to someone who doesn’t.

Peter Attia: Only if he were depleting it in some other store. So in other words, I’m making this up, but if you could say, well, all of us … I mean you probably know this. Everyone loves to quote this fact, right? But red blood cells have more cholesterol in them than LDL particles, right?

Dave Feldman: Right, right.

Peter Attia: So, the LDL denier loves to say, “Well, we don’t think red blood cells are causing atherosclerosis and yet they have more LDL.” Whatever. But my point is, unless you’ve depleted a pool of cholesterol elsewhere in his body, just on mass balance, you had to make more of it.

Dave Feldman: As far as I understand, the liver can recycle cholesterol as many times as it wants to.

Note from Tom Dayspring: Liver has a lot of options – but physiology demands when the liver has excess cholesterol, it must excrete it in the bile as FC or bile acid.

Peter Attia: Again, that’s true, and the liver and the gut have a very clear pathway, which he described.

Dave Feldman: Correct me if I’m wrong, the liver is the only organ that can actually degrade cholesterol, right? Nonhepatic tissues can’t degrade cholesterol.

Note from Tom Dayspring: No organ can degrade cholesterol — liver can change it to a bile acid. Steroidogenic tissue can change it to a hormone — that is it.

Peter Attia: Well, again, it depends what you mean by degrade. Remember, cholesterol has no caloric value. It’s not a metabolite. It’s not something that we metabolize, right? We turn it into-

Dave Feldman: If you were synthesizing more, you’re saying it’s going to go somewhere if it needs to be synthesized more.

Peter Attia: Right. If you’re synthesizing more, presumably you have a higher growth of the organism. You have more cells, you need more cells because obviously probably the highest demand for cholesterol is for cell membranes. So that’s what I’m trying to figure out. It’s like, where is all this extra cholesterol coming from? If it’s not being synthesized de novo?

Dave Feldman: And that’s my larger point. The inversion pattern is part of what should bring this to light. Is, why then, when I have an enormous amount of fat over three days, I would see my LDL-C dropped by 73. Why would I see my LDL-P drop by 1,115 in three days from eating huge amounts of dietary fat? Why would that happen?

Peter Attia: I mean, it’s an interesting question. I’m just not sure. All it’s basically saying is you have a way to perturb those levels. I mean, I’ll give you another anecdote here. So, you know, I did a one week fast, I went keto for a week and then ate nothing for a week and now I’m actually back on Keto for a week and I’m checking my blood every seven days. So that was my LDL-P before I … That was just me. That’s my normal LDL-P. So I’m walking around at 920 nanomoles per liter. This is after months and months of time restricted feeding with virtually no carbohydrate restriction other than just I don’t eat crappy carbohydrates.

Dave Feldman: Right.

Peter Attia: So then I went keto for a week and look, my LDL-P actually went up. Now I don’t think I would meet criteria for being a “hyper responder” because it went up to 1380, which is not that high, but you know, that’s still a significant jump for me. Right? Okay. What should it have done when I fasted for a week? Shouldn’t it have gone up according to this model?

Dave Feldman: Well, here’s the catch. The catch is, I only know the three-day window. I don’t have a lot of data from people who fasted for a week, as in just water fasting.

Peter Attia: I do.

Dave Feldman: Okay. You do? Good.

Peter Attia: Yeah. I’ve done this on multiple patients who have done three, five and seven-day fasts.

Dave Feldman: Okay. Who are fully ketogenic and you’re saying it typically goes-

Peter Attia: Not always ketogenic. No. Sometimes they’re just, you know, fat adapted. Sometimes they’re not. Sometimes they’re actually insulin resistant and we use the fast to kick them into a state of Ketosis to make it easier. But, and again, I want to be very careful. This is simply just anecdote because I’ve only done this on maybe 30 people, but this is not uncommon. I mean, look, my LDL cholesterol went from 64 to 37 after the fast. I mean it went way down.

Dave Feldman: Wow. Fantastic.


Peter Attia: And that’s consistent with what I see. Now, that’s not the reason I’m fasting. To be clear. I’m not using a fast to manipulate lipid proteins. I’m doing it for a completely unrelated reason, but my point is I always see, and again I can say always because it’s a relatively small end. Obviously, at a large enough and you’re going to see counterexamples to anything and everything, but the general principle seems to be under caloric deprivation, LDL goes down and under fat deprivation, LDL goes down.

Dave Feldman: I’ve got to put in the one footnote and it’s an annoying footnote that I keep putting in. I really need to. And this is why I’m dying to get these parts of the exercise. I need to look at a population that also is not getting any, particularly weight training or resistance training and so forth, because that seems to impact it. Whether it’s my theory or not, that seems to impact overall lowering LDL-C numbers. I’d be very interested in this in fasting.

Peter Attia: I’ve got that in my patients. Not every one of my patients lifts weights despite my best efforts. So, again, I think the more common thing that we see is that when you put people on a high-fat diet, the ones that go on to have this hyper response, as you note, their trigs usually go down, their cholesterol goes through the roof and it’s driven by a doubling, tripling, or even greater output of synthetic biomarkers like desmosterol. Now, for reasons I don’t understand, you also tend to see at least two of their three phytosterols go up. Now, one thing we didn’t talk about though, I wrote about it, so I’m sure you know about this, is this seems reversible. If you eliminate the saturated fat.

Dave Feldman: It does seem to matter in this case, and you’ll like this, at least in our own data with apoE 4s, seems more likely if you’re an apoE4, you will see a drop somewhat in saturated fats.

Peter Attia: So I don’t think I have a large enough sample size because I’ve only put seven patients through that protocol. Which is, and I wrote about the very, very first one. So that’s probably the only one that you’ve seen that I’ve talked about, but this was a young guy who went on a ketogenic Diet, was crushing it, meaning like everything was going well. I mean he’d gone through the adoption period. His performance was exceptional, mentally never felt better. He was not an overweight, or metabolically ill guy to begin with. He was just kind of a normal software guy who just decided he wanted to take it to the next level. But then he showed up with labs, not unlike these. He was the first guy I ever saw where there was a greater than sign on the LDL-P.

Dave Feldman: Right. It maxes out at 3500.

Peter Attia: I was like, “Oh, I didn’t actually realize the assay stopped at 3,500. This guy is my guy.” He looked better on some of the other metrics. He didn’t actually have a lot of the inflammatory stuff. This was before the Ox LDL assay was commercially available. So I didn’t have that. But you know, his CRP and is Lp-PLA2 we’re okay. But we had the discussion, right, which is the discussion that at the end of the day I’m accountable for having, which is what are we going to do about it? And he was young. I mean the guy was 30, so it wasn’t like we had to do something tomorrow. This is not a guy who’s going to have a heart attack in a week.

Dave Feldman: Because I think when you originally writing about it, he actually pushed back a little bit. Like he wanted to come up with a-

Peter Attia: Well, what he pushed back on was, because I basically said, “Look, man, I don’t think the ketogenic diet is right for you. These numbers are crazy.” And he was like, “Yeah, I don’t ever want to go back to what I was doing before.” I mean he basically said in not so many words, I’d rather die of a heart attack and feel this good. And he’s not saying like, “I’m going to die of a heart attack tomorrow”, but he’s like, “Look, I’d rather live to 60 and feel this way, than live to 70 and not.” Which I say totally fair by the way. That’s a very reasonable tradeoff to make, but let’s also think about this a little more logically. So that was actually probably the first case I ever discussed with Tom and this was I think back in 2011 and it was actually this discussion I think that lead Tom to go on to write the lipidaholics case, even though it wasn’t the patient that he used in that case because he then went out and found others like them.

What we just decided was on biochemistry first principles, our hypothesis was it’s the saturated fat, more than the ketones. Because that was the other thing, this guy didn’t want to leave ketosis. My thought was, let’s just dial this back, get you out of Ketosis. So our hypothesis was both ketones and saturated fat can be readily converted into cholesterol, but if he’s adamant on staying in Ketosis, let’s at least get his SFA down to 25 grams per day. Which was hard. This is a guy that was eating about 100 grams of SFA a day, maybe 80 grams. I mean, it was a lot. We basically just made most of it MUFA. So I said to the patient, I said, “Look, the way we could do this just for no other reason as the purpose of a thought experiment is, you’re going to basically have to become a nonstop olive oil, Macadamia nut eater. Even the avocado you can’t go hog wild on because eventually, you’ll get too much carbohydrate out of it for this purpose.”

And sure enough, I think he came back at like 1300, after eight weeks or something like that. So six more patients have been so adamant about staying in Ketosis and not taking any medication but wanting to go through this experiment and all of them have had the same response. Which is, if you can get them to mainline MUFA, you fix the whole problem. Without reducing or increasing to any measurable effect, how much fat they’re consuming.


Dave Feldman: Have you been keeping track of their PUFA levels too? Because it’s hard to add a lot of MUFAs-

Peter Attia: Their PUFA was going up.

Dave Feldman: And you know, the issue with that, with adding more PUFA, is the downside is there’s the potential that you’re actually adding more peroxidation on the particle basis.

Peter Attia: Yeah. Yeah. I mean, I think it depends. This is one of those areas where I’m trying to get a lot smarter and I want to make sure I’m not sitting in the echo chamber. I’ve historically, you know, measured RBC levels of arachidonic acid and all of those things and tried to keep track of the ratio of that to the EPA and the DHA, but ultimately I don’t really think I know the answer to this yet. And I also don’t think … My guess is PUFAs are not as bad as I have historically thought them to be, but they’re probably not as good as MUFA. So notwithstanding that, I just want to talk about this from the level of the thought experiment, so to speak. My sample size of those people is too small to know if there’s a relationship to their apoE gene. To your point, what I’m hearing you say is maybe that’s more … That’s something that you’re more likely to do in someone who’s an apoE4 carrier.

Dave Feldman: It’s been a proportional kind of thing that we’ve sort of noticed.

Peter Attia: I don’t know, but what I do know is it seems to point back to this idea that the hyper-responders … That the Occam’s razor here is that they’re making more cholesterol. Because that makes sense from a mass balance standpoint. Again, I’m still-

Dave Feldman: And I would have agreed with you were it not for this energy inversion that I see. If it were not … The very thing that I said from earlier that I could move my LDL-C to where I want to move it. That would be based on me basically arranging for a few days to eat to a certain level and I’d be pushing down my LDL-C by eating up to a certain-

Peter Attia: Notice you’re pushing up and down on LDL-C within supraphysiologic levels. Meaning, if I recall seeing your data, do you mind showing me that again?

Dave Feldman: Unfortunately the computer died. That’s why I’ve had it closed, but yes, what you’d see as an inversion graph and-

Peter Attia: I’m very familiar with it and we’ll obviously link to it, but it was, as I described it, you’re showing that you can move your LDL between the ranges of very high and stupidly high.

Dave Feldman: No, I’ve moved my-

Peter Attia: What’s the lowest you’ve ever got your LDL?

Dave Feldman: 98, and I suspect-

Peter Attia: Okay.

Dave Feldman: I suspect I could get my LDL down to 70 if I was willing to go through with it. There’s another part of this conversation we haven’t had a chance to touch on, but that you might find very interesting since we talked about the liver and again, more theory, so about 80 percent of what I’ve talked about is an explanation that I’m trying to fit onto what I know and this is definitely one of those, but I suspect given the data that I have for the whole second phase of my research that actually part of how I’m reducing my triglycerides in the blood through a certain series of experiments I call carb swap experiments, is that I’m just trying to get a certain threshold of glycogen stores up in my liver, which seemed to be at a certain point, meaning that there will be less VLDL secreted and therefore less LDL. That’s the theory.

Peter Attia: I’m sorry to say again. By moving glycogen, you’re doing what?

Dave Feldman: So again, I’m trying to think about it more like as an engineer, whether right or wrong. I think, okay, if I were trying to engineer this body and I were caring a lot about the long-term tank of storage, which is your adipose sites, but it was also to care about the short-term tank. What is the reason why it would make sense from a mechanistic standpoint as to why the body would want to be so adamant about mobilizing these triglycerides for fuel in a low carb high fat athlete, especially somebody is very lean? The short Occam’s razor from my perspective is it comes down to, well, there’s very little glycogen stores relative to somebody who’s on a high carb diet. This isn’t to say that they have bottomed out glycogen stores, but relatively speaking, there’s less play and because of that, it makes sense as to why the body would activate more lipolysis, have more of the free fatty acids moving through, keep making it more available, etc. Okay, so then how could I tweak that?

Peter Attia: But wouldn’t that also just as easily be explained by the fact that when someone’s walking around with 60 percent of the glycogen in their muscle, there’s by definition, assuming they don’t have, you know, a pathologic condition, there’s not much circulating glucose and therefore it’s probably, they’re also in a low insulin environment which is fostering lipolysis.

Dave Feldman: Right, but it’s gradiated, right? It’s going to have to do with … There are threshold points which are part of what I’m trying to isolate out and there seems to be a threshold point with me. For example, if I have around 90, it seems to be at the last time I tested it. It’s at around 90 carbs per day, net carbs per day. That seems to be the magic threshold that if I do that for about three days, my LDL-C will just drop substantially. Now, let’s see, I did 70 net carbs for three days. I’ve already done that, doesn’t do it. I have to get up to a certain threshold and once I get up to that certain threshold, which seems to be somewhere around 90, all other things being equal, like I have to structure my life around this where like sleep the same amount and so forth.


That seems to be the point at which there’s an actual drop and I actually wrap that around one of my presentations is I actually had a fat shake, a ketogenic shake for, I want to say three days for a washout period, and then for just two days I added what I thought would be the calculation to demonstrate what I was looking to do. So I swapped out the fat for carbs, kept it isocaloric for two days. I had one that was a lower amount and one of them was a higher amount.

I can’t remember the numbers right off hand, but then to further emphasize this, I switched back to the high-fat diet for the next five days and we saw my LDL-C drop almost immediately and continued to remain low for the next five days. And in fact, it got to the lowest point, the end, right?

Peter Attia: So it seems to be at least a sensitive on your carbohydrate intake is your fat intake, right?

Dave Feldman: Correct. And this is very relevant because again, I’m still thinking about this from the energy model-

Peter Attia: But hang on. For this to be … I’m not saying this isn’t interesting, but for this to be relevant, it has to be reproducing something that’s going to last over much longer than five days. How do we know if these effects aren’t transient? And also how do we know they’re relevant if they require an extreme condition such as, one of them, I know you talked about, how you just have to force feed yourself a ton of fat.

Dave Feldman: Yeah, that was the very first presentation that I did, where I was having just a very high pro-

Peter Attia: That also dropped your LDL-C paradoxically. Correct?

Dave Feldman: Correct.

Peter Attia: Right. But the point is … We have to back up for a moment. Why does any of this stuff matter in the long run? If you’re listening to this podcast or if you’re sitting in front of me as a patient, you have a very important question you have to ask yourself if your LDL-P is through the roof as a result of what seems to be your diet. Does it matter and do you want to do anything about it?

Dave Feldman: Right. And this is why this is the most relevant question is if you’re right, and I’m not saying that I know for sure either way. Let’s say that you’re right, you, Peter, you’re right that this, regardless of how we got here, if you have a high LDL-P because of being on a ketogenic diet, then there was a lot of people who need to know that if they’re at higher risk and I definitely want to be one of the people that brings that to their attention. I have emphasized this several times before and I’ll say once more, I am on a journey of science, not of advocacy. I’m going to be quite a skeptic, but all of that said, if I come to a point in which I can feel convinced the LDL-P is, in fact, atherogenic, absent remnant lipoproteins, absent having a high HDL, low triglycerides-

Note from Tom Dayspring: That is called FH phenotype.

Peter Attia: But we already know it’s absent remnant lipoproteins…

Note from Tom Dayspring: Lots of CAD patients do not have atherogenic remnants.

…I mean there’s a thousand studies including this to demonstrate that the atherogenicity between these people and these people, I’m pointing to the Garvey study, has nothing to do with what may or may not be remnants. The LDL-P alone here … I think one thing that sort of makes sense-

Dave Feldman: I’ve not seen a study yet, where remnant lipoprotein … I’ll send you the ones I have. If you have somewhere LDL particle is more relevant than remnant lipoprotein, I would be very interested.

Peter Attia: Again, we have to be very careful of what we mean by remnant. So there are clearly going to be a subset of remnants that are potentially the most pathologic on a per particle basis. But I think the body of evidence implicating the causal role of apoB and LDL-P is so overwhelming. Is it perfect? Of course not. But I think what concerns me with this culture, when I say this culture, I mean this sort of low carb culture of LDL doesn’t matter, is if I had a dollar for every time I had to see some low carb enthusiast basically dismissing the idea that LDL is relevant and touting the idea that statins are a big conspiracy theory, that’s a really dangerous problem.

Dave Feldman: It, it’s dangerous if we know … Just generally speaking, I think, dismissiveness in general, of evidence, is bad, but we do have to care about the quality of evidence regardless.

Peter Attia: We sure do. So, let me offer a very controversial viewpoint that I can’t believe I’m about to voice publicly. I think one of the challenges in the low carb community is you have a group of people who have become very used to rejecting mainstream information because they did so with nutrition, right? If you are on a low carbohydrate diet, on some level you have decided that the ADA, the AHA, the USDA, the NIH, and the CDC are full of shit.

Dave Feldman: I think that’s a fair point.

Peter Attia: Okay. So, do you know what I think that when it comes to nutrition, that’s largely true. Okay. Now they think they’re coming around, but I think it’s largely true and I think the body of evidence, the body of literature that pointed towards the Food Pyramid was quite shotty. I don’t think it was nefarious. I don’t think this was as much of a conspiracy as people want to make it out to be. I don’t think Ancel keys with some evil dude who was like scheming a way to … I just don’t buy that. I think there were strong personalities and lousy science. Those are very different things.

Dave Feldman: And let me go on the record with saying I basically agree. I feel like-

Peter Attia: This is the easy part to agree with. This is the hard part that’s coming.

Dave Feldman: Okay, good.

Peter Attia: Okay. So what happens is a lot of people get in this mindset of, “Well, look over here, I saw an entire body of evidence that was very easy to dismiss, and by the way, look at the results.


You don’t have to be a rocket scientist. If you’re sitting there following the food pyramid, getting fatter and sicker and you abandon it and get better. There’s the proof.” The real problem, and, I again, I apologize for getting on a soapbox. The real problem is when people try to look at the last 50 years of lipid literature through that same lens. Nobel prizes have been won in this field. Now I know that somebody is going to start screaming, “Oh, that’s an appeal to authority. That’s a logical fallacy.” I got it. To you, whoever said that, I acknowledge it, but unless you’re willing to go back and read every one of those papers, something even I have not done, though, I’ve probably read more of them than most. We’re talking about apples and oranges with respect to a body of literature here.

The body of literature implicating LDL as having a causal role, a necessary but not sufficient role in the pathogenesis of atherosclerosis is on a different level from the body of literature that gave us the food pyramid and the real challenge I think in this low carb community, this LDL-denying community, is they’re throwing the baby out with the bathwater. Now, part of that is because I think there are too many doctors who are too lazy at the other end of the spectrum. They just assume, “Well, statins or what we give everybody. You know, anybody who’s LDL-C is about 100 gets to be on a statin”, and these doctors are equally guilty in my mind of being ignorant and not thoughtful and not understanding the pathophysiology of the disease, but somewhere between these, is a measured space that requires a very careful consideration of the literature.

Dave Feldman: With that in mind, to kind of broad base this pretty well, this whole energy model that we’ve been talking about that I’ve been kind of walking you through, and the audience to some degree, it’s the lens by which I came up with the challenge. I didn’t come up with a challenge because I was just like, “I just, today I just want to tweet out something that I think we’ll get a lot of people annoyed.” I believed that. I looked at those two sides of the dotted line and I said, it looks as if you’re insulin resistant, we tend to see that there are high levels of triglycerides. We tend to see the high levels of VLDL, forget even remnant. We can detect VLDL to a certain extent. We can see that people who are down the range of metabolic derangement, they’ve got that and that tends to be highly associated with cardiovascular disease.

This is all on the energy delivery side. On the other side of the LDL, we tend to see on the support side that there can be further problems in the immunological role and oftentimes that can be induced by likewise, lipolysis. That can actually be something that can … you’ll see on the vitamin E studies, for example, this gets brought up quite a bit. For example, if you inject lipopolysaccharides (LPS) into a body, you’ll actually have a higher fatty acid synthesis that’s going on in the liver along with lipolysis, in order to induce that higher response. So again, we see on that side we see higher levels of triglycerides and I kept coming back to, I kept-

Peter Attia: But you’re also, as we talked about, when you inject LPS in somebody, you’re going to see a higher HDL cholesterol too. I mean, everything at that moment. LPS is a terrible toxin. It’s going to kick the body into a four-alarm fire. Of course, it’s going to want all of the energy substrate it can muster and all of the hormonal precursors it can muster.

Dave Feldman: Agreed, agreed. So when we’re looking at an association between LDL particles and a bad outcome, we want to absolutely confirm it was the cause and not the association, right? But the same reason that if we were to say, I want to confirm ambulances aren’t the cause of death for people who are dying inside of ambulances.

Peter Attia: Yeah, I mean we have to be a little careful with that analogy. So let’s, let’s be clear. Can you get atherosclerosis without having an oxidized sterile taken up by a macrophage?

Dave Feldman: No. We agree on that. If I could take all of the LDL particles out of your body right now, I could feel totally confident that you will not die of atherosclerosis.

Peter Attia: Okay. So does that not imply that LDL is necessary but not sufficient for atherosclerosis?

Dave Feldman: I don’t disagree with that. Without question, LDL particles are-

Peter Attia: It’s important for some people to understand that because I do think, I mean, put it this way, I’ve certainly heard people in this community argue the following, that the burden of proof should be on the lipidology community to demonstrate that LDL is causal rather than the reverse, and I find that comical, if it weren’t for tragic. Right?

Dave Feldman: Let’s go back to the analogy for a second. Are ambulances causal for ambulance related deaths? Absolutely. They’re a part of the-

Peter Attia: That’s not the same thing, because what you’re basically saying is-

Dave Feldman: I’m emphasizing an association over causation. We both realize that, right? So, okay, it could be in this town, you actually are in worse shape if an ambulance picks you up. There’s very incompetent EMTs, and their life-saving measures are poorly done and therefore, if you could ban all ambulances in this town, you’d find that actually-

Peter Attia: A Mendelian randomization of ambulances would ferret that out.

Dave Feldman: I would agree if there wasn’t anything associated with the ambulance in the Mendelian randomization with the response by ambulances in the first.


Peter Attia: The only way I think you can discount the Mendelian randomization is if you believe that the mutations that you are measuring … So you’re looking at a series of mutations that are affecting a phenotype. In this case cholesterol level. You’d have to convince yourself that each and every one of those is also affecting something else that’s driving the underlying cardiovascular process. But we’ve already went through this, right? It can’t be the LDL receptor because that’s not even ubiquitous and there aren’t LDL receptors on your epithelial cells.

Dave Feldman: I want this tested on a healthy vascular system, however that’s occurring. I want every cell to have-

Peter Attia: Why does a person with FH not have a healthy vascular system when they’re born?

Dave Feldman: When they were born?

Peter Attia: Yeah. Meaning they inherit a clean slate, right? Someone who’s born with FH has a normal, beautiful vascular system, that over time in most of them becomes destroyed.

Note from Tom Dayspring: Actually, in many, atherosclerosis starts in utero.

Dave Feldman: Just answer this question, is there any cell in somebody who has FH that would function like a normal cell and somebody who doesn’t have FH in order to be able to acquire the lipids or lipoproteins it wants to take?

Peter Attia: Yes. There are plenty of patients with FH who do not have defective, completely defective LDL receptors and therefore are not impeded. Put it this way-

Dave Feldman: Their metabolism is not impeded is what you’re saying.

Peter Attia: Not all of them. Again, we have to be very careful when you talk about FH because there are at least 2000 known versions of that disease. It’s very cumbersome, so that’s why I think FH gets talked about like it’s one disease, it’s a phenotype that has all of these things that can cause it. So, the broader question is, is everyone with FH struggling to make steroid hormones?

Dave Feldman: I don’t know the answer to that question.

Peter Attia: No. In fact, FH maybe slightly protective in the case of an infection and in the case of diabetes and one argument for that … The diabetes one is a little hard to explain. The infection one, because … I mean, FH has stuck around for a long time, so there may have been a time when having the ability to mount an incredible immune response would have proved to have a survival advantage, and if you have four times the cholesterol of somebody else, that’s one moment in when that could come at a huge advantage.

Dave Feldman: Right, which gets back to the immunological response, but to get back to the larger point, I wouldn’t blame somebody who has a poor digestive system for being malnourished. So, as long as we can count and everybody’s tissues to be properly nourished and-

Peter Attia: I don’t understand what you mean by blaming them, help me understand what you mean by that.

Dave Feldman: What I mean by that is, if there was a problem with absorption of liquids or lipoproteins unrelated to total quantity of LDL particles, that’s what I’m going to care about and I hopefully will have an answer on this soon. I’m actually working with a couple of researchers who I’m trying to get an SNP list together that doesn’t include livid metabolism issues. So Ronald Krauss, who you had on from earlier, he was talking about this, they are looking right now on, for example, the genetic studies and he was explaining the receptor issues associated with … and this is how you end up with higher levels of LDL-C or LDL-P, right? Is that you end up having less absorption [uptake], particularly on the liver side, but I’m especially interested in non-hepatic tissues-

Peter Attia: But there are, there are people with Niemann-Pick C1-Like 1 transporter deficiencies. There are people with ATP binding cassette deficiencies who have a huge increase in cholesterol. It has nothing to do with an LDL receptor. It’s a transporter.

Dave Feldman: I’m not pointing just to the LDL receptor, I’m pointing to just the health of the cell. If the health of the cell is not compromised than I’m interested. If the Lipid metabolism difference-

Peter Attia: But why would someone who’s ATP binding cassette in their enterocyte, that is not appropriately excreting cholesterol, therefore driving up the recirculated cholesterol pool. Why? Why does that mean that their endothelium is somehow defective?

Dave Feldman: I would have to follow what the path is that we’re talking about. I don’t know that I could give an answer to that until I can actually see the study that’s associated. If you can take a biopsy of anybody who’s going to have this issue and you can basically effectively see that the cells for which would be targeted, there was not going to be any problems, I wouldn’t have any problem with using it. I mean basically what we really want, is just the means of just an overproduction on the part of the liver without it touching any other part of the liquid system. And your point, I’ll make your point for you. It’s hard to get an SNP that doesn’t in some way touch other parts of the liquid system, but that’s also the point against it. You see what I’m saying?

Peter Attia: So, so let me ask you this. You’re, you’re saying, “Look, I want more evidence.” And I think science is based on skepticism. I completely respect that, but I think we also have to temper that with some modicum of understanding probability theory and saying, look, at some point.

Peter Attia: Modicum of understanding probability theory and saying, “Look, at some point the probability looks disproportionately one way versus the other. So right now, what would your confidence be in the idea that LDL is playing a causal role in atherosclerosis, just as endothelial dysfunction and inflammation play a causal role in atherosclerosis?


Dave Feldman: Let’s make a distinction, the distinction is, if you’re saying, “Is it part of the development of an atherosclerotic plaque?” It’s nearly 100%. If you’re saying, ” Is the total quantity of LDL particles absent any inflammation or anything else along those lines?”

Peter Attia: Nobody is saying that. Nobody reasonable is saying that. So again, listen to what I said, right? So you’ve got three things that we can sort of use a metaphor and say they’ve formed the three legs of the stool. Three things have to happen for someone to get atherosclerosis. Each of them is necessary, none of them alone are sufficient. That’s just the nature of complicated biology.

Dave Feldman: Let me help you with the question. I think this would be a better way of asking it. If given the same quantity of oxidative stress, whether it’s low or high, would you rather have 1,000 nanomoles of LDL particles, or would you rather have 2,000 nanomoles of LDL particles? I think-

Peter Attia: Yeah, well, you don’t need to ask me that question.

Dave Feldman: Right.

Peter Attia: I think the question is, what would you rather have?

Dave Feldman: I used to think that I would say the first. That I would rather have 1,000. I would say last year it was probably more like, it could be about the same difference. Learning what I’ve learned, especially with the antioxidative defense system and so forth, and particularly given my own data, especially with the CIMT data that I’ve presented recently. I don’t know if you’ve seen that one as well. I was getting a carotid intima-media test every six months. And during those six months, the beginning of this diet, and through the experimentation, I was running at LDL-C levels of 200 or higher, LDL-P levels of 2,000 or higher. For four tests in a row, you can actually see the regression that’s happening on both the left and right side of the carotid arteries. Now-

Peter Attia: Now, again, I don’t want to get started on CIMT, which is hopefully it’s the same tech doing it the exact same way. I mean, I’m guessing your CIMT initially was pretty good and it may have gotten a little bit better. But, I don’t know. CIMT is even worse than calcium scoring, frankly.

Dave Feldman: Sure enough. But, what-

Peter Attia: But again, Dave, we’re putting a couple of N of how many’s, we’re saying, “Look, these three little interesting anecdotes are basically calling us to suggest that the null hypothesis around this topic should be what you’re discussing, rather than what I think is a remarkable body of scientific literature, that is not without its problems, and that is not absolute in its inference.” But-

Dave Feldman: That’s not what I’m saying. What I’m saying-

Peter Attia: That might not be what you’re saying, but it’s certainly what a lot of people are using your words to say.

Dave Feldman: I have an energy model that a lot of people are utilizing probably over-simplistically. But, if my energy model is right, it would suggest as to why, the answer of why-

Note from Tom Dayspring: Get it published.

Peter Attia: But, David, you haven’t even described it correctly to me today. I mean, you’re … I mean, I guess it depends how liberal we want to be with the term model. But, there is no evidence that the LDL is there to carry cholesterol. You have yet to explain to me where Moffitt got his cholesterol.

Dave Feldman: You’re talking about to the quantity that he has it at?

Peter Attia: Yes. You guys got three times the amount of LDL cholesterol.

Dave Feldman: I think it typically tracks with the total particle count. The people-

Peter Attia: You have to give me the mass balance. You’re an engineer, you know this stuff just as well as I do.

Dave Feldman: If you are a hyper responder coming to cholesterolcode.com right now, and you turn over your lab, I can look at your LDL-C before you-

Peter Attia: No, no, no. That’s fine. That’s just pattern recognition. That’s not the interesting thing to me. I’m asking a very important physiologic question, which you have yet to provide an answer to, and it seems to be the central tenet of your belief system. Where did Moffitt get his cholesterol? Why does he have three times more than he had before?

Dave Feldman: The short answer to that is, he synthesized it and he’s recycling it. Now, there’s some degree with which he’s synthesizing-

Peter Attia: Okay, so this is a totally different answer than before. He has now increased his synthesis of cholesterol. He doesn’t have the same circulating pool. This is not a shell game with boats. Right?

Dave Feldman: I was talking about circulating before.

Peter Attia: No, but I don’t think, what I certainly didn’t hear you say before was that he has actually increased his own endogenous production of cholesterol.

Dave Feldman: There’s some amount where you’re increasing it in order to meet that existing demand. I don’t know how much that is.

Peter Attia: But, this is different from what I understood you to say earlier, which is the reason he has more cholesterol is it’s just along for the ride with the boats, and he has to have more boats, which defies-

Dave Feldman: No, that is correct.

Peter Attia: But, that defies the principle of mass balance. You can’t create matter out of nothing.

Dave Feldman: I’m not saying he’s creating matter, okay.

Peter Attia: So he had to make more cholesterol. I don’t see a way around that.

Dave Feldman: I’m not disagreeing with him making more cholesterol. I think where we’re disagreeing is, I think you’re saying in total, he’s making three times more every day. Am I wrong on that?

Peter Attia: On average, he is making three times more, or reabsorbing three times more. But, just based on what I’m seeing-

Dave Feldman: Reabsorbing at the liver, or reabsorbing in non-hepatic tissues?

Peter Attia: Probably in the gut…

Note from Tom Dayspring: The only place where absorption occurs.

…That’s where the majority of the reabsorption is taking place.

Dave Feldman: Okay, in other words, he is sending it back out the other side.

Peter Attia: Well, again, this is what we look at these sterile numbers for. When the desmosterol goes through the roof, plus or minus the phytosterols, that tells you these patients are making more cholesterol.

Dave Feldman: Okay.


Peter Attia: But, here’s the question. If this were purely about energy, he shouldn’t be making any more cholesterol. He should have more particles perhaps, but they should be cholesterol-depleted.

Dave Feldman: I’m interested in this, I have no … You answer this question for me. When does somebody make more cholesterol-depleted … because everything that I’ve read in Clinical Lipidology and so forth, is it’s like a standard quantity on the non-triglyceride side of the ledger. If you’re making cholesterol on a per particle basis, it can vary on a per particle level, but generally speaking, it tends to hit averages that are fairly consistent.

Peter Attia: But, this is an unusual circumstance you’re describing. Right? This is the whole purpose of this experiment, is you’re describing people who-

Dave Feldman: It’s not unusual-

Peter Attia: … who’s demand you’re saying is so great for triglycerides, but they’re doing-

Dave Feldman: That you make more boats. But, the boats, if they already have a standard composition, why would they change that standard composition per boat?

Peter Attia: Well, think about … So are you telling me you’re saying that the large LDL particle and the small LDL particle in the insulin resistant versus the insulin-sensitive patients have the same cholesterol composition?

Dave Feldman: No, that’s my point. My point is getting back to remnant cholesterol. Why is it that I think there would be something that would happen on that dotted line, something before and after. Right? Why would there be … Why would there be a problem with somebody who’s metabolically deranged with their cholesterol relative to one of these people that are theoretically metabolically flexible? Why would there be a difference? And the short answer to that, the short answer is, I don’t know fully all of the aspects to it. I do know though, there seems to be a longer residence time with VLDLs, and we see that because that’s the fasted blood test-

Peter Attia: Yeah, we know that that’s explained very clearly by apoC-III, the residence time on the LDL for that matter as well in pathologic states.

Dave Feldman: So if I became more insulin resistant, and therefore, ended up with higher VLDLs I couldn’t, say two years later, have healed that and then have fewer VLDLs.

Peter Attia: No. I mean, look back to the Garvey study. There’s a reason I printed this up, because I knew we’d be talking about this over and over again. There’s very little difference. To try to impute or infer something about remnant cholesterol from VLDL is as complicated as trying to assess Lp(a) by looking at LDL cholesterol. Think about that for a moment. When you look at LDL cholesterol, if it’s directly measured, do you agree that it’s the sum total of LDL cholesterol plus Lp(a) cholesterol?

Dave Feldman: Yes, I believe that’s how it does work.

Peter Attia: And it excludes VLDL cholesterol and IDL cholesterol because they contain apoE, whereas the Lp(a) and the LDL-P do not. So if you have a direct cholesterol measurement, the LDL-C is technically LDL-C plus Lp(a)C. But, there is no way on God’s green Earth that you can look at that and infer what the Lp(a) is.

Dave Feldman: Right, without testing directly.

Peter Attia: Yeah, and similarly, we don’t know what’s going on with these VLDLs, meaning in Moffitt, because we haven’t measured it. But, we’ve measured this in patients that span the spectrum of insulin sensitive to diabetic, and that doesn’t appear to be the answer. The difference in the atherogenicity, the difference in the residence time, and the difference in the total apoB load appear all to be driven through the LDL particle, not the VLDL particle. So something else explains why they have more LDL.

Dave Feldman: That’s what I want to find out. Again, I’m very upfront about what it is that’s theoretical and what isn’t.

Peter Attia: But, we kind of already know the answer to that question. It’s the triglyceride content.

Dave Feldman: But, until we can actually test it on people who are fat adapted, or ketogenic, we can’t say that we do. When we can do a kinetics study … When we can do a kinetics study on VLDL secretion with people who are particularly likely mass hyper-responders, then we’ll have some idea.

Peter Attia: But, Dave, that will only offer you an explanation. It will not change the question.

Dave Feldman: Of risk?

Peter Attia: Yes.

Dave Feldman: Right.

Peter Attia: Let’s say you can do a kinetic study, and hopefully someone wants to fund this, because it is an interesting question. Again, I’ve done the kinetic study on myself, you’ve seen my data.

Dave Feldman: Right.

Peter Attia: I lose triglyceride, not cholesterol.

Dave Feldman: Right.

Peter Attia: Which, I would expect. Right? I am seeing cholesterol basically stay the same in those cells, and it’s during the periods of extensive exercise and fasting, we’re seeing triglyceride movement within the cell. But, the point is, even if this theory turns out to be correct, it’s an explanation, not a reason. It’s an explanation for something, but it’s not a reason to ignore it, is it?

Dave Feldman: This is where I think we’re getting circular. It’s an explanation as to why it could be benign or even beneficial, and that’s where we’re disagreeing ultimately, which I figured we would be. Why would you have high LDL for a good reason? And you’re answer would be, “There wouldn’t be one.”

Peter Attia: No, no, that’s not true. There wouldn’t be a good reason with respect to cardiovascular disease, there are plenty of good reasons to have high LDL, we just talked about them. The FH patients obviously get some benefits from their high LDL. But, from a cardiovascular standpoint, I don’t think there is a single good reason to have high LDL, and I am not aware of a single card-carrying lipidologist or member of the community that spends a lot of time in this literature that could come up with one, and I’ve been asking. I mean, it’s something I’ve been very interested in.


Give me a teleological reason to have high LDL from a cardioprotection standpoint. I mean, I was asking this question seven or eight years ago, and there is no answer.

So again, doesn’t mean that having high LDL is always bad, but it’s really important to understand this distinction. The other thing to keep in mind is, lots of things in biology are not linear. So look at Gilbert’s syndrome, Gilbert’s syndrome is a very common condition. 10%, maybe not 10%, 2% or 3% of people listening to this have it and probably don’t even know it. But, they have elevated unconjugated bilirubin, but very slightly elevated. So if you’ve had a blood test done, you probably know down at the bottom it says, ALT AST, bilirubin. And normal bilirubin would be less than one.

But, these patients with Gilbert typically get to about two. Well, in half a dozen studies, these patients have an enormous risk reduction in cardiovascular disease. Why? Why would having a slight doubling of bilirubin, which by the way, at high levels, is toxic. So if you walk around with a bilirubin of 10, you’re not going to be around very long. And those patients present, they get sick, they have obvious symptoms, they’re jaundiced, and they usually have some pathology that’s leading to it. But, these patients can walk around with a bilirubin of 1.6 to 2, and they seem to be getting a benefit from it. And they also seem to have lower LDL, and even if they don’t have lower LDL, because the literature is mixed on this, they always have lower Ox-LDL. And it may be that the best explanation is that bilirubin has anti-oxidative properties. So they get this protection from cardiovascular disease, but it’s a U shaped curve, or an inverted U shape curve rather. Meaning, as that bilirubin gets higher and higher, they start to lose any of that benefit, meaning, whatever oxidative benefit they get, it’s more than being outweighed by the damage that comes from that elevated bilirubin. So I guess my point here is, even if there’s an explanation for why this is happening, from an energy trafficking standpoint, which again, I really want to be clear, I do not think there is. I do not think that energy trafficking explains this phenotype. I think that is not the Occam’s Razor answer, I think the Occam’s Razor answer is, they’re making a boatload more cholesterol, because I think we have pretty good data to suggest that.

Dave Feldman: Which I’m dying to test, by the way.

Peter Attia: Yeah, yeah. We should make sure that you can, and that other people can do this. But, of course, the point here is, it still won’t actually answer the question, what should you do about it? Just because there’s a reason for something doesn’t mean that it’s a benign condition, or that it should be ignored.

Dave Feldman: I agree, and not only that, separate subjects. I’ll even go a step further and say, it could be a U shaped curve on this end as well. It could be that you could have an LDL-P of say, 1800, and it turns out that’s actually the bottom of the curve, and people at 1800 don’t turn out to be as high a risk as people that are like the one you just showed me, above 3500. I not only grant that, I also further tell lean mass hyper-responders, “I may turn out to be right on my cautious optimism, as far as the risk of cardiovascular disease, but it could turn out that there’s something else we haven’t yet determined.” That’s a problem with this phenotype, which is another reason why we should be sharing all of the symptoms that may be coming along with it as well. But all of that said, all of that said, the larger question is, why then, would I be able to identify a certain set of parameters that, when studied, seem to suggest that high levels of LDL-C, I want some with high LDL-P, doesn’t prove to be problematic. And that’s why I want to get a hold of something, I want to get a hold of really large-

Peter Attia: But, how will you demonstrate that?

Dave Feldman: By stratifying. Stratifying for high HDL, stratifying for low triglycerides.

Peter Attia: No, no, no. How-

Dave Feldman: Stratifying for high level-

Peter Attia: How will you know if they do or do not have, if they’re not at increased risk for cardiovascular disease? How long will you need to follow them to know that?

Dave Feldman: Well, that depends on the data set I can get a hold of. I’m not in your space so I have to work with other people who are researching-

Peter Attia: Oh, you’re saying you want to do this with retrospective data.

Dave Feldman: Correct.

Peter Attia: Okay, so meaning this is your challenge to say, “Don’t give me genetic data, don’t give me drug data.”

Dave Feldman: I like just normal non-drug, non-genetic stratified people preferably. And I like to stratify just on those three, just on HDL, LDL-

Peter Attia: Even though your patients probably have some genetic SNPs that are explaining their phenotype?

Dave Feldman: Oh, I would definitely want to know that as well. That’s why I’m trying to actively get the 23andMe, I would love to send it your way for obvious-

Peter Attia: But, the point is, you’re excluding anybody who has anything that could be called the genetic alteration, even though the patient population you’re trying to understand this is, almost assuredly has a genetic alteration that’s rendering them susceptible.

Dave Feldman: I’m not trying to exclude that.

Peter Attia: But, you just said you don’t want to consider any of the genetic drivers of FH.

Dave Feldman: Let me emphasize. If you’re making a study that is gene specific, then it’s the gene that drives the detection, the discovery of those people. Right? I don’t want to do that. I want to actually see if I can get a broad base study of people who happen to already have high HDL, low triglycerides, and high LDL, and see if they have high rates of not only cardiovascular disease, but all-

Peter Attia: But again, I come back to the FH patients. You can’t find a more broader demographic of people in terms of variable genetic inputs that produce a phenotype similar to what you’re looking at.


Dave Feldman: I think we’re just going to end up in one of these got to agree to disagree moments. Until I can-

Peter Attia: Which is fine. I totally respect that. But, I just want you to understand what it sounds like from over here is, you’re looking for six-footers, you’re looking for a and you’re not going to be happy until you see a six-footer, and gosh doggone it, you’re not going to leave the kindergarten classroom until you find one.

Dave Feldman: I would argue the opposite. I would say, “Look, why shouldn’t I be able to grab …” If I could right now just grab a million people in the United States, just absolutely randomly determine. Why would that not be significant data if I found that there was this stratification for which high LDL did not result in high levels of cardiovascular disease or all-cause mortality?

Note from Tom Dayspring: That’s what Mendelian randomization does.

Peter Attia: Well, because if you’re going to do that honestly, you’re going to say, “Well, they can’t have a single genetic mutation, they can’t be taking a single drug, and they can’t be on any funky diet.”

Dave Feldman: Let’s say all of that turned out to be true.

Peter Attia: And what if they don’t exist?

Dave Feldman: If I found that out, that would be definitely something I think would be very interesting to my followers. I would turn that back around.

Peter Attia: But, you’ll never know if you found that out or not, Dave.

Dave Feldman: Well, I’ve already found two studies that do stratify for those three. And of the two that do, high LDL does not result in high rates of cardiovascular disease.

Peter Attia: Wait, wait, you’re talking about these glycogen storage disease cases?

Dave Feldman: No, no, no, Framingham Offspring has one study where they stratified by three, and unfortunately, my computer is dead or I’d show you the other one. There’s another one that stratified discreetly between below 170 LDL-C and above 170 LDL-C. And the high HDL, low triglyceride group, when compared to above and below, were nearly identical, both on the high side and on the low side.

Peter Attia: Yeah but, this study didn’t stratify by apoB.

Dave Feldman: Right. I would love to have apoB.

Peter Attia: Okay, but, that’s the Quebec Heart Study for you right there.

Dave Feldman: The Quebec Heart Study has the stratification by all three of those metrics?

Peter Attia: The Quebec Heart Study, here I printed it up here. I mean, basically, it’s showing it has nothing to do with the LDL-C once you know the apoB

Note from Tom Dayspring: P nonsignificant.

Look at the risk.

Dave Feldman: Okay. You have to understand, what I’m trying to get is … Sorry. I’m trying to get those three in conjunction. I want to specifically stratify those three. And in software, this is where I get a bit frustrated, because I feel like there’s such a cultural difference between medicine and software. We’re used to having just loads and loads of free data, just we’re awash in free data. Google can’t wait to give me everything that I want to see. I requested, I’ve actually applied-

Peter Attia: No, I’ve heard you talk about it. Are you being denied that, or did they not have the data?

Dave Feldman: They just don’t return my emails. I mean, there’s even people that I would think would be sympathetic inside the low carb community, and I’m not going to try to call them out, who I’ve also tried to get this information from. And I just can’t get it, and I want just a nice, clean regression on three axes, that’s all I want. That’s nice and fat.

Peter Attia: So the three axes being triglyceride-

Dave Feldman: Triglycerides, HDL, and preferably LDL-P. Now, there is an important distinction we’ve got to make with apoB, because apoB can, in theory, also include remnant lipoproteins.

Peter Attia: Yeah, LDL-P is more accurate than apoB.

Dave Feldman: Right. And LDL-P would be extremely fantastic. If you could help me get in touch with that data set I would be very interested, and not with any major adjustments, I mean, whatever Cox proportional might be fine. But, just generally speaking, if I could get a big fat dataset and stratify on those three axes, I think that would say a lot as to whether there’s any validity to the energy model overall.

Peter Attia: So when you look at the MESA data, which stratify on a Kaplan-Meier curve, the difference between LDL-C and LDL-P. You’re saying that that’s not relevant because why?

Dave Feldman: The thing we’re dancing around here is, obviously when you have high HDL-C and you have low triglycerides…

Note from Tom Dayspring: MESA adjusted their LDL-P data for TG and HDL-C.

…it suggests a number of different things. But, more broadly, it’s suggesting a properly functioning lipid energy system, and probably not being in a state of challenge-

Peter Attia: HDL-C tells us absolutely nothing. If we’ve seen enough from Mendelian randomizations and in other, how many more CTEP failures do we need to see? HDL cholesterol tells us nothing about HDL function. In fact, any time you increase HDL cholesterol pharmacologically, you seem to make patients worse.

Dave Feldman: I know, but these are modifications to the existing lipid system. If you block cholesterol-

Peter Attia: I get that, Dave. But, boy, if you’re going to hang your hat on it’s all about HDL-C triglyceride, I mean, wow, we are so far beyond that in the lipid world at this point. If you’re going to go through this brain damage, make it for something worthwhile.

Dave Feldman: But, wouldn’t you predict right now, that if I did hang my hat on those two things, on those two markers, against LDL-C or apoB or LDL-P, that it would fail. That if I were to say, “Hey, I want to get a stratification just of high HDL-C and low triglycerides.” That you’d say, “Sure, Dave.” I’ll bet you $10,000, I’ll give you 100 to 1 odds, those people with high LDL, even if you stratify for those two, will still have high rates of cardiovascular disease.

Peter Attia: Again, I’d have to completely see the patient population before I could even hazard a guess.

Dave Feldman: But, right now, you would assume that. Right?

Peter Attia: I am going to assume that LDL-P is going to be a stronger maker of prediction than HDL-C.

Note from Tom Dayspring: The NLA statement points out that virtually all trials that show HDL-C is a risk factor were not adjusted for apoB or LDL-P.

Dave Feldman: And that’s not what I’m making the case on. What I’m making the case on is, whether or not there’s a properly functioning lipid metabolism, which would be indicated by all three of those.

Peter Attia: No, you have absolutely no understanding of the lipid metabolism by looking at HDL-C and triglyceride.


Peter Attia: Not even close. No, no, no, we can disagree on things that are nebulous. This is not nebulous, Dave. I mean, you do need … Again, I hate that I’m saying this, because I sound like a jerk and I don’t mean to. You’ve got to spend more time with lipid people, you really do. You are not dealing with your peers at this. You have to go and figure out … I mean, HDL-C is just categorically not a useful metric. It is like a first-order term on a … No, it’s not even that. In engineer-speak, it’s the fourth order term on a fifth order polynomial.

Dave Feldman: That hurts, Peter, that hurts.

Peter Attia: No, I mean, come on. It’s just not that.

Dave Feldman: I’m just kidding, I’m just kidding. Look, I-

Peter Attia: It’s super crude. And don’t confuse the ubiquity with its utility. The ubiquity of it is, “Yeah, it’s cheap. It’s easy. Everybody’s got it.” But, let’s not let people listen to this and get lulled into a false sense of, “Hey, if my HDL is high and my trigs are low, who cares what my LDL is.” And unfortunately, that-

Dave Feldman: I want to prove that right or wrong.

Peter Attia: Well, first of all, you’ll never prove anything in science, so let’s be really clear on our lingo.

Dave Feldman: Okay.

Peter Attia: No, no, no, it’s very important. It’s important for your listeners to understand that.

Dave Feldman: Fair enough, fair enough. But likewise-

Peter Attia: But, nothing is proved, it’s about probability.

Dave Feldman: Sure, but likewise, would you say the lipid hypothesis is proved?

Peter Attia: Absolutely not. I just said, “There’s nothing outside of mathematics that exists in a proof.” Nothing. And I have the basis of the luxury of having been a mathematician once, so I get it. There’s a luxury of being able to write QED at the bottom. We will never do this here, and if people are sitting there saying, “Well, I’m going to keep eating my bacon and eggs like it’s mainlining, and ignoring my LDL-C because my HDL-C is high and my trigs are low, because I’m on a low carb diet, and somehow that makes me special, because no one’s proved that this is wrong.” Wow. That’s not the legacy I want.

Dave Feldman: So what if I continue to find more data sets that actually support that? What do I do?

Peter Attia: I mean, I don’t know. I don’t know what that means. What do you mean by more datasets? Meaning more anecdotes?

Dave Feldman: No, I’m talking like, say I do actually get a hold of Framingham Offspring, let’s say I get a hold of … I forget what some of these larger data sets are-

Peter Attia: MESA.

Dave Feldman: MESA, sure. Let’s say I can get MESA and I can stratify for those three and it’s showing the same thing, without doing a lot of adjustments or anything along those lines. What would I tell my followers? I would say, “No, it looks as if, still there’s further evidence that’s showing high LDL-C, in this case, is not problematic.” MESA did actually stratify for LDL-P, didn’t it?

Peter Attia: Yes.

Dave Feldman: Yeah. So that’d be a great example. MESA would be fantastic data to get a hold of. Is that something that you think I would ever actually be able to see or be able to run regressions against?

Peter Attia: I mean, I’ve never thought of it, but I agree with you. That would be great. I don’t know who owns the data.

Note from Tom Dayspring: The US government.

Dave Feldman: But would that be compelling to you? If it turned out that we could run regressions on it, let’s say that it was in the next room right now and we worked it up on the computer, and sure enough, I went by the stratification I was looking for that are identical to somebody who’d be … And they were typical to somebody who is already a lean mass hyper responder, and it would show that they didn’t have high rates of cardiovascular disease. Would that be compelling data to you?

Peter Attia: I think compelling is the wrong word. The question is, how would it add to the existing body of literature that informs a decision we have to make every day with a patient. And the answer is, “I’d have to see the strength of it and decide, how does this fit into the existing body of literature?” I mean, that’s the only way I can imagine thinking about this. But, everybody listening to this, and you and I, all have to put our heads on a pillow at night with a null hypothesis, against which we have to challenge existing data. I’m not convinced that the null hypothesis here should be anything other than the lipid hypothesis. Now, the lipid hypothesis gets bastardized all the time, it gets misstated all the time, it gets based on LDL-C and a whole bunch of other stuff.

But, I’m talking about the real honest to goodness, no bullshit, LDL hypothesis, which again, I’ve written about eloquently, and people have written about it far more eloquently. I should say I’ve written about it in a kluge way, others have written about it eloquently. The lipoprotein, the endothelial damage, the inflammatory changes, all of these things cascading, that’s my null hypothesis. And in the end, if there’s data to counter that, I’m all for it.

For example, even when you look at the IL-1, IL-6 agonists, the low-dose methotrexate studies that showed you could delay or reduce cardiac events without changing lipoproteins, I don’t think I’m being delusional when I say that doesn’t change the model. It actually feeds into the model. The model is, there are three things that are driving this pathology, if you reduce one of them, things get better. All things equal, if blood pressure goes down, do outcomes get better?

Dave Feldman: I believe they do.

Peter Attia: Absolutely they do. Very potent. Why? Endothelial function. All things equal, if you stop smoking, do outcomes get better? Absolutely. So when you start to look at all these things, and again, with those-

Dave Feldman: But by outcomes get better, you’re specifying-


Peter Attia: Cardiac outcomes.

Dave Feldman: Right, cardiac outcomes.

Peter Attia: The all-cause outcome is a much more complicated question that probably is a podcast in and of itself. So to your question, yeah, Dave, of course. I’d be incredibly curious to see this. Who wouldn’t be? But, don’t think that one regression analysis on MESA is going to turn over 50 years of data…

Note from Tom Dayspring: Less than 20 years old.

…regardless of what it shows. The question is, how does it alter our understanding and thinking of the problem?

Dave Feldman: Sure. Look, the whole reason I’m even pursuing this particular strata is because of the model in the first place. I had to have something that I could conceive of, that would inform the decision by which I would be looking for what the data is that would disprove it. That’s why I’m in pursuit of disproving it. At the end of the day, Peter, I can’t emphasize this enough. I’m not looking to talk to the echo chamber or looking to just maneuver around inside a number of people that are going to congratulate me. I specifically-

Peter Attia: But, I think you’re better off going to an NLA (National Lipid Association) meeting than a low-carb meeting.

Dave Feldman: Sure, but they’re freaking expensive. I’ve looked at all of them.

Peter Attia: Get the low-carb community to fund you. If they want to know the answer, because I don’t think they do, if I’m going to be brutally honest. I think the worst of that crowd just want their confirmation bias. They have seen these incredible benefits of low carbohydrate diets, and their belief is, nothing can be wrong with this. Like we somehow live in a mono-dimensional, monochromatic world, where it’s that black and white. And if the diet is good for this, it can’t be bad for anything. And they are so wed to that they construct these crazy arguments. But, if they share your passion for truth, then they should happily fund you to go to an NLA meeting and spend a week there, and actually, start hanging with these guys who are way smarter than me. I’m a knucklehead, I know a lot about lipids for a knucklehead. But, I’m talking about, the smartest people in the world are the ones you need to be talking to on this topic. And they’re not at low-carb conferences, I promise you that. They’re not on Twitter, and they’re not playing patty cakes on their high-carb, whatever, low-carb blogs. It’s just not about that stuff, man. And again, I think what you’re doing is really interesting. I don’t agree with the model, but I’m glad that you’re pursuing it. I wish you the best in pursuing it with the right people.

Dave Feldman: Absolutely. Perhaps you’ll be able to help me set up with the right people. I would definitely be more interested in finding those voices that can help tear at this model.

Peter Attia: I would be more than happy to help in any way I can.

Dave Feldman: Great. Well, I can’t emphasize enough as I anticipate, I was going to ask you more questions than you asked me. I’m really appreciative that you took the time to chat with me about this.

Peter Attia: No, my pleasure, Dave. Thank you very much. I apologize if this just took longer than we thought it might’ve, and I know we went off on tangents all over the place. I guess this will be one where the show notes are probably quite helpful. But, nevertheless, it was great meeting you in person. I didn’t realize it’s only been three years since you’ve been at this. It feels like a lot longer actually.

Dave Feldman: It certainly does for me. And my wife would say its felt twice as long for her. You have to realize something, almost nobody knew about me a year and a half ago, and I knew almost nothing about cholesterol three years before that. So this is absolutely been a fresh journey, and that’s why I have to oftentimes emphasize that I’m not a formally trained biochemist, and I really have a lot of gaps, I’m sure, in my knowledge that I’m looking to fill, and find as fast as I can.

Peter Attia: All right. Well, it was great to meet you. Enjoy your time in San Diego. Oh, by the way, for the listener, this is being recorded on July 26th. It will be a long time before this goes up, Dave, so hopefully the listeners will understand that whatever has transpired since then is just … We pre-record these things many months in advance. We may have to bump it up a little bit, depending on, maybe we can reshuffle it and get it out before the end of the year, which is probably right now where it sits in the pipeline. But-

Dave Feldman: You’re going to subject me to quite a hell, because I guarantee every single follower I have is going to be knocking on my door until this thing is opened up, so that will be pretty funny.

Peter Attia: Well, we’ll do what we can. Anyway, man. All right. Thanks so much.

Dave Feldman: Absolutely.

Peter Attia: Awesome to finally meet you.

Disclaimer: This blog is for general informational purposes only and does not constitute the practice of medicine, nursing or other professional health care services, including the giving of medical advice, and no doctor/patient relationship is formed. The use of information on this blog or materials linked from this blog is at the user's own risk. The content of this blog is not intended to be a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard, or delay in obtaining, medical advice for any medical condition they may have, and should seek the assistance of their health care professionals for any such conditions.
Facebook icon Twitter icon Instagram icon Pinterest icon Google+ icon YouTube icon LinkedIn icon Contact icon