#085 Dr. Peter Attia on Mastering Longevity – Insights on Cancer Prevention, Heart Disease, and Aging

Peter Attia on ApoB, lipid management, and why diet alone rarely fixes cholesterol

Peter Attia explains why ApoB is a superior cardiovascular risk predictor and covers the limits of using diet alone to manage lipid levels.

"But the majority of what's happening is a disease that leads to plaque formation inside of coronary arteries. And we can go as deep or as shallow as you want into that and why that happens and how that's a function of endothelial injury, lipoprotein burden, and inflammation. But this leads to a reduction in blood flow to key parts of the heart muscle. And when that happens, the heart undergoes an ischemic event. Now, sometimes that can be chronic and sometimes that can be acute. And if an acute event occurs in a region where enough muscle of the heart is compromised, that's going to result in sudden death. That's a heart attack. And it's important to understand that a little, when I was in medical school, it was more than 50%. It's now a little less than 50%, but it's still a very high number. A little less than 50% of people's first brush with a symptom of coronary artery disease is sudden death. That's worth repeating because we couldn't, I remember, I still remember being asked this question in medical school. You know, you're sitting there as a first year medical student in cardiovascular pathology class. And the pathologist said, what's the single most common presenting feature for someone having cardiovascular disease the first time? And everyone was like, chest pain, shortness of breath, you know, rattling off all the usual stuff. He goes, no, sudden death. Again, today it's not quite 50%, but that's a very sobering statistic. Absolutely. I do want to dive into some of the, you know, major causes of the atherosclerosis and the atherosclerotic cardiovascular disease that you're talking about. So lipoproteins, you mentioned, and most people, they hear about lipoproteins, they hear about LDL or HDL. But ApoB, why should people know about ApoB? Well, again, I think it's worth maybe just getting everybody on the same page with cholesterol. Let's start with that, right? So everybody's heard of cholesterol. And I think most people would probably even have kind of a negative valence when they think about it's like cholesterol is a bad thing. So it's worth explaining that that's not really true, right? Cholesterol is an essential thing, right? So without cholesterol, we wouldn't be alive. And there are really rare, fortunately, genetic conditions in which cholesterol synthesis is compromised, and those tend to be fatal in utero. So if an organism can't make enough cholesterol, it ceases to exist because cholesterol is the thing that gives every cell fluidity, the membrane of every cell fluidity. And it's the precursor to some of the most important hormones we make. So in the case of us as humans, right, testosterone, estrogen, progesterone, cortisol, these essential hormones are all made from cholesterol. So every cell in the body, with the exception of red blood cells, makes plenty of cholesterol. all. The lion's share of it is probably done by the liver and the steroidal tissues. And we have to figure out a way to move this stuff around the body. And the highway system of the body is the blood. And the blood, of course, is water. So if we want to move things that are water-soluble throughout the body, like proteins and ions, it's easy because they dissolve freely in water and they move around. But when you want to move something around water that is not water soluble, such as cholesterol as a lipid, you have to wrap it in something that is water-soluble. And that something is the lipoprotein. And the big protein on the surface of that sphere is called an apolipoprotein. And there are, broadly speaking, two classes of apolipoproteins. There are the A class and the B class. So some of the lipoproteins are wrapped in an apolipoprotein called ApoB100. And we just abbreviate that to ApoB. But I'll just say it this one time and we'll never talk about it again. There's also an ApoB48 that wraps another type of lipoprotein called a chylomicron. We won't talk about that again because it doesn't really factor into cardiovascular disease. So ApoB is short for ApoLipoprotein B100, which is the structural apoprotein that sits on low-density lipoproteins, abbreviated LDLs, intermediate-density lipoproteins, abbreviated IDLs, very low-density lipoproteins, abbreviated VLDLs. The APOAs, and this is big A, never to be confused with APO little a, which we may talk about, those wrap the family of high-density lipoproteins. They're much more complicated than APO-Bs, believe it or not, and there are many of them. But nevertheless, broadly speaking, that's what's going on. So why do we care about all this stuff? Well, in the 1950s, when it became clear that cholesterol was playing a role in cardiovascular disease, the first observation was people with very, very, very high total cholesterol, because at the time that was all that could be measured, was total cholesterol. By the way, what that meant was the total amount of cholesterol in all of your lipoproteins, in your HDLs, in your LDLs, and in your VLDLs. Those three lipoproteins constitute the amount of total cholesterol you have in the lipoproteins. We can come back to this idea because it's important. That represents about 10% of the total cholesterol in your body. The total cholesterol concentration was loosely correlated with cardiovascular outcomes, but only at extremes. meaning if you took people whose total cholesterol was in the top 5% and compared them to people whose total cholesterol was in the bottom 5%, there was a clear association with cardiovascular disease. March forward many, many decades, we came to realize that actually this low-density lipoprotein, which is a subset of your total cholesterol, but it's the cholesterol contained within the low-density lipoproteins, that's much more strongly associated. And what we now know is the case is there's an even better way to predict risk than just saying how much cholesterol is contained within the low-density lipoproteins. A better way to predict risk is to add up the concentration of all the ApoB particles. So that number, ApoB, measured in milligrams per deciliter, is the concentration of the entire burden of particles that are capable of undergoing something that I'm sure we'll talk about, which is the initiation and progression of atherosclerosis. So how the ApoB number, can you talk about how that, so you mentioned LDL, total LDL cholesterol. That number is like determined by some equation, right? It can be, but it can also be measured directly. Would that be particle number though? No. So there's two ways to go about doing this. So in the olden days, and unfortunately many labs still do this, they rely on an equation called the Friederwald equation. So total cholesterol is relatively easy to measure. So you draw the plasma, you spin it down, and you basically lyse all of the lipoproteins, and you can measure total cholesterol. So if you just basically apply something to lyse all of the proteins, you'll say, all the lipoproteins, you'll say total cholesterol is 200 milligrams per deciliter. Then they directly also measure two other things. They can directly measure total triglyceride concentration. And using a separate assay, they can measure the total concentration of cholesterol within the HDL particles. So now you've measured total cholesterol, HDL cholesterol, and triglyceride. The Frieder-Wald equation stems from an observation that kind of, sort of, on average sometimes, VLDL cholesterol is approximately one-fifth the triglyceride concentration. So the Frieder-Wed equation is quite literally used to estimate LDL as follows. LDL cholesterol is estimated as total cholesterol less HDL cholesterol less triglyceride concentration divided by five if you're doing everything in milligrams per deciliter. And unfortunately, most labs still do that. So when you look at your cholesterol report, it'll say LDL-C, it'll give a number, and unless it says direct, you can assume they've done the Frieder-Wald equation, which is, I've seen that wrong more often than I've seen it right. A good lab will do a direct assay. They will actually measure LDL concentration, and they will give you, in milligrams per deciliter, the total concentration of LDL-C. That is still an inferior predictor of risk relative to ApoB. Yes. Okay. So the reason I wanted to mention that LDL-C is because, as you mentioned, many labs do measure it indirectly. And there are many types of LDL, right? So there are different densities and sizes. So I'm curious about what your thoughts are on the different sizes of like more atherogenic sizes of LDL, such as the smaller dense particles. And, um, you know, like, like how you view that, like this, the, the different particle sizes and the particle number. And then of course, ApoB. So like the whole. I mean, there's been a big evolution in the way we've practiced medicine in our practice with respect to this. So 10 years ago, we were looking at LDL particle number, which both the MESA population, so the multi-ethnic study of atherosclerosis, and the Framingham offspring population have both demonstrated unequivocally that when you compared LDL particle number to LDL cholesterol, LDL particle number always predicted risk better than LDL cholesterol. So how would you do this? You would follow people longitudinally for cardiovascular events, and you would do this in sort of like a cumulative incidence graph. So on the x-axis, you have time. On the y-axis, you have incidence of cardiovascular disease. And you plot out everybody as a function of whether LDL-C was higher or lower as a percentile than LDL-P. So LDL-P stands for the number of particles. LDL-LC is the concentration of cholesterol. And this was, again, unequivocally the case. Particle number always predicted better. So how do you count the number of particles? Well, it turns out there are different ways to do this. You can do this using NMR. So nuclear magnetic resonance is like how an MRI works. So it's applying a magnetic field. It's basically doing, I mean, this is being a little cheeky, but it's sort of like doing an MRI on the blood and you can count the number of particles that way. That's not actually the gold standard, but that's the way it's most commonly done in clinical practice. It can also be done with ion motility. We switched from NMR to ion motility for LDLP because it was more accurate. But ultimately, and this is now about five years ago, we actually switched to APO-B, which was superior on all fronts. And here's the reason why. First of all, there are different ways in labs to do this. So LabCorp, for example, and Boston Heart have different magnets and different algorithms for how they run their LDLP. So if you run an LDLP on each of those labs, you'll get a different number. That's a bit disturbing to me. I want to know that the APOB that I get at one lab is the same as the APOB I get at another lab, and it's standardized across all fronts. But there's a more important reason why I favor APOB over LDLP, and that is it encompasses the total atherogenic burden. And you can get burned and fooled by patients who have very high VLDL, meaning they have a high burden of very low density lipoproteins, even if their LDL burden is low. So I won't go into it because it's so nerdy. It's not worth getting this deep in the weeds. But there are certain genetic conditions where people have completely normal LDL, but very elevated VLDL. And they have a very high atherogenic risk. And you will miss that if you're looking at LDLP or LDLC. You will not miss that if you're looking at ApoB. What about the fact that it's small dense LDL, which has been shown to be more atherogenic. So ApoB does become... So you mentioned the structural role of APOob in the lipoproteins it's very important it also plays a role as you mentioned in allowing the lipids to be soluble in in in the plasma right um but it becomes it plays a role also in recycling so it gets you know it interacts with the ldl receptor and can be taken back up into the liver. The small dense LDL particles, ApoB is somewhat obscured as the LDL particle gets smaller in size and more dense. Therefore, it's not- Harder to clear. Harder to clear. Exactly. So what about in the case, and the reason I'm asking is because as you mentioned, ApoB is on VLDL, IDL, LDL, right? But there's different sizes of these LDL and the larger, more buoyant LDL is better than having a higher proportion of the smaller, dense LDL. Right. But ApoB captures that risk, right? So in other words, this is another reason why I think that ApoB is the great equalizer, because once you have the ApoB concentration, you're accounting for the fact that clearance is going down. I mean, the one way to think about this is anytime you see an elevated ApoB, it always comes back to something on the clearance side is not working. Now, there are really, broadly speaking, when I talk about this with patients, I go through the four sort of pillars of what elevates ApoB. So it can be driven by cholesterol synthesis. And we can talk about that because it's going to factor into dietary choices, for example. So how certain dietary patterns will lead to higher LDL than others. it's impacted by cholesterol reabsorption so the we can talk about what the life cycle of cholesterol is but again it's um you know we make it and we reabsorb it and it gets circulated. It can have to do with triglyceride burden. So this is where insulin resistance really factors in to how ApoB can go up. And ultimately it comes down to clearance. And clearance has everything to do with the presentation of the LDL receptor on the liver, the confirmation of it, the number of them, and how long they survive on the liver. And all of these things have an enormous effect, some of which we can manipulate with drugs. So for example, all drugs that are used to treat LDL in some way or another, indirectly or directly, impact the LDL receptor. Some do it really directly, like a PCSK9 inhibitor directly does that by targeting a protein that breaks down LDL receptors. Um, so, um, anyway, a long-winded way of saying the, and this is another advantage of ApoB is it allows you to, in one measurement, capture all of that risk. Because if you have two individuals, if you're just using LDLP as your risk, you might miss some of the elevated VLDLs. If you're looking at LDLC, you'll clearly miss some of the size issues. That should be captured in LDLP. But again, I guess maybe what you're asking is if you have a low ApoB but they're all small, is that worse than having a low ApoB where they're all big? And the answer is probably, but you'll also see that in the – like there are other metrics that are kind of coming on board now which are looking at LDL triglyceride levels. so you can look at the degree to which the LDLs are cholesterol depleted. And that can also give you a sense of risk. The question is, is that a first or second order term? And I think the first order term is still going to be the number of particles. That's the biggest driver of risk. And everything else factors into it. In other words, that's not an independent risk because it's driven by the resident's time of the LDL, which is driven by the clearance rate. So let's talk about the number, so the LDL, sorry, the ApoB number, because if most people go to a standard lab and they measure their ApoB, there's a reference range. And it says, okay, if you're less than 80 milligrams per deciliter, then you're okay. Where does that number come from? And has anyone measured ApoB levels across the lifespan? Do we know, is there a correlation with ApoB levels and the beginnings of atherosclerosis? Has someone done those studies? that sort of thing. Yeah. So the reference ranges are purely population-based distribution questions. So every lab will have a different way of doing this, but a general, you know, sort of philosophy for labs is, you know, let, you know, so for the lab we use, and by the way, we completely ignore these reference ranges, but they're there. We can't avoid them. They're there. And we explained to our patients that we're going to editorialize on top of them. But, you know, the reference lab we use will say APOB below 80 is wonderful. Well, 80 just happens to be the 20th percentile of the population. It will say 80 to 100 is intermediate, or 80 to 120, it says is intermediate risk, and above 120 is very high risk. So for the lab we use, we know that 80 is the 20th percentile, 120 is the 80th percentile or the 60th percentile, I can't remember. So it's literally just putting you up against a population distribution, and that's it. Now, our philosophy on ApoB is completely different. And as you may recall, I devote actually quite a bit of real estate to this in the book, because I think it is such Thank you. on ApoB is completely different. And as you may recall, I devote actually quite a bit of real estate to this in the book because I think it is such an important concept. And it is, in my opinion, certainly top three failures of Medicine 2.0 is in failing to appreciate the point I'm about to make, which is that once you understand the causality of ApoB, meaning once you understand that ApoB is not just associated with cardiovascular disease, but it's causally linked to it, meaning it causes ASCVD, to get into this discussion about managing 10-year risk, thinking about being in this percent versus this percent, makes no sense. When you have causal things that cause disease, you eliminate them. And the analogy I use is cigarettes with lung cancer. So nobody disputes that cigarettes are causally linked to lung cancer. They are. It's as clear as, you know, Tuesday follows Monday. But people forget that, you know, causality doesn't mean everybody who smokes will get lung cancer, and it doesn't mean that every person with lung cancer smoked. So you don't need to be necessary and sufficient, necessary or sufficient to still be causal. But our approach to patients who smoke is very clear, which is never smoke. And if you do smoke, stop immediately. Do we look at people who smoke and say, well, once your 10-year risk of lung cancer reaches this threshold, we're going to tell you to stop smoking. Or once your pack year smoking is above the 50th percentile or the 80th percentile, we're going to tell you to stop. Absolutely not. You immediately eliminate smoking. And so similarly, it makes no sense that we would look at a causal driver of ASCVD in the case of ApoB and kind of take an approach of, well, being at the 20th percentile or the 30th percentile or the 40th percentile is acceptable."

Why restricting fat to lower cholesterol often backfires metabolically

Extreme dietary fat restriction can cause insulin resistance and poor carb quality. Attia recommends solving lipids with pharmacology and using nutrition for energy and protein.

"A lot of times when I see people on these excessively restrictive fat lowering diets, they actually become insulin resistant. A lot of them because they're really over consuming a lot of poor quality carbs."

HRT timing for Alzheimer's risk: early initiation may protect APOE4 carriers

Late HRT initiation may increase Alzheimer's risk, while early initiation at menopause appears beneficial specifically for women carrying the APOE4 allele.

"Late initiation of HRT may be counterproductive for AD risk, may actually increase AD risk. Early initiation appears to potentially only be beneficial in E4 women, but not E3 women. So for you, I would say doubly beneficial to initiate at the time of menopause because of your E4. Got it. That's really good to know. And also defining what is early and late, like, you know, so that, you know, there was a study, there's a couple of studies. One was the elite study and one was the DOPS. So like, I don't know, it was like early intervention for estradiol or something. And then there was another one that was the Danish osteoporosis prevention study. In both of those studies, in the DOPS one, the initiation of the, they did estradiol, I think they did like triphasic or something, but also they did the progesterone. So it had progesterone as well. It was like the cardiovascular disease risk or mortality went down the b venous thromboembulism like the things that happen um that can increase with uh menopause went down over the follow-up which was 16 years or something but these women only took it for 11 years and they started at age either between the age of 45 and 58. So perimenopause was in there and also just a few years after. Yeah, this is to me the biggest unknown question, Rhonda, and we don't know the answer. And what I find very frustrating is we're not going to know the answer because nobody's going to do the study. I am as comfortable with anything in medicine as I am that initiating HRT at the time of menopause does not increase a woman's risk of heart disease, breast cancer, or anything else. In fact, it reduces her risk. It clearly reduces her risk of heart disease, dementia, and BMD. And it's either protective or neutral on cancer. It was neutral on the top side. and it's either protective or neutral on cancer it was like it was neutral yeah it's protective or neutral on cancer i am very confident of all of those things here's the thing we don't know what do you do 10 years later what do you do when she's 60 right and again if you look at the hrt data from the women's Health Initiative with all of its flaws, the answer would be you should probably stop. But again, that study is so flawed on so many levels that I'm not sure. And here's where I would argue. There's one area where you absolutely know things will get worse when you stop the estrogen and that's bone density. So the other things are a little less clear to me. There's some, there's some, you know, opacity around what will happen to cardiovascular disease risk, dementia disease risk, and cancer disease risk if you start appropriately initiated HRT after 10 or 15 years. But what is unambiguously clear is her bones are going to get brittle again. Because the moment you take the estrogen away, bone density goes down. Estrogen is the most important hormone in men and women for the regulation of BMD. It is the chemical transduction system that turns force into bone building. so so have basically strain gauges in our bones that are sensing forces on the bones, and that force is being turned via estrogen into a chemical signal to osteoblast and osteoclast to promote bone building. And once estrogen goes down, that goes away. So if you take the estrogen off a woman 10 years post-menopause, she will once again go into a rapid state of decline. Now, she's still better off because she'll still be further ahead than where she is if you put her in decline 10 years sooner. So, you know, I've had arguments with people on the anti-HRT side, and they say, you should never use estrogen for treating BMD because we have bisphosphonates. And I say, first of all, you only use bisphosphonates for three to five years. Two, they suck, meaning they're not as good as estrogen. And third, you can use estrogen for longer. And they say, well, yeah, but once you take it off, it still goes down. And it's like, yeah, but it's a new baseline. And it's a higher baseline. It's like waiting to retire. You're going to have more in your retirement fund if you retire at 70 versus 60 but so this is the big question because because i mean and my and and again i we've done a back of the envelope calculation that would suggest even if the risk of alzheimer's disease or heart disease or cancer even if you lost any protection from HRT and maybe had a slight increase in risk, given how big the risk of falling is, you might still end up being neutral risk carrying out HRT indefinitely. So this is where the lifestyle factors probably- Which is quality of life. Right. But this is probably where lifestyle factors do play somewhat of a role as well, because if you have, obviously, if you've been doing resistance training up until that point and continuing it, you're building not only bone, you know, certainly built up a lot of bone density reserve earlier in life, but muscle mass helps, right? And then let me throw this at you. Because I've, you know, I've thought a lot about the nuclear hormone, nuclear hormones, basically, like, so vitamin D is the one that I've really focused on. And when I was doing a lot of research on it, so nuclear hormones, you know, we have steroid hormones, nuclear steroid hormones, I should say, sorry. So we have estrogen, testosterone, vitamin D is one. So these are binding to a receptor that, you know, in some cases, the receptor complexes with other ones, it goes into the nucleus of a cell, which is where all the DNA is. And it goes down to the level of genes and it recognizes a little sequence of genes. So in the case of estrogen, it's called an estrogen response element, an ERE. In the case of vitamin D, it's called a VDRE, vitamin D response element. There's a lot of overlap between vitamin D and estrogen in terms of the genes they're regulating. And so I'm wondering if avoiding vitamin D deficiency also becomes one of those important lifestyle factors because, you know, in some cases, obviously, vitamin D also plays a role in bone metabolism, right? But independent of that, also just looking at the crosstalk of the genes that vitamin D and estrogen regulate, and they're like, they're both, and the response elements are, they're different, but they're somewhat, I'm looking at that and it's like, oh, I wonder if there's like, that seems like you might be able to compensate a little."

Danish DOPS study: 11 years of early HRT reduced cardiovascular mortality

The DOPS study showed that women starting HRT between ages 45-58 had reduced cardiovascular disease risk and mortality over 16 years of follow-up.

"This is to me the biggest unknown question, and we don't know the answer. And what I find very frustrating is we're not going to know the answer because nobody's going to do the study."

Testosterone replacement: early cardiac risk spike, then benefits at 2-3 years

A large TRT study showed slightly increased cardiac events at year one that vanished by year two, likely from blood pressure effects in the highest-risk men.

"We're very aggressive. If you look at the sprint trial, I think it's very clear that 120 over 80 or better is the place to be."