The Peter Attia Drive - #395 - Brain lipidology： understanding APOE, cholesterol homeostasis, Alzheimer's disease risk, and the effects of lipid-lowering therapies on brain health ｜ Tom Dayspring, M.D.

00:00:00.000 Hey, everyone. Welcome to The Drive Podcast. I'm your host, Peter Atiyah. This podcast,

00:00:16.540 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:21.520 into something accessible for everyone. Our goal is to provide the best content in health and

00:00:26.720 wellness, and we've established a great team of analysts to make this happen. It is extremely

00:00:31.660 important to me to provide all of this content without relying on paid ads. To do this, our work

00:00:36.960 is made entirely possible by our members, and in return, we offer exclusive member-only content and

00:00:42.980 benefits above and beyond what is available for free. If you want to take your knowledge of this

00:00:48.100 space to the next level, it's our goal to ensure members get back much more than the price of the

00:00:53.420 subscription. If you want to learn more about the benefits of our premium membership,

00:00:58.040 head over to peteratiamd.com forward slash subscribe.

00:01:04.220 My guest this week is Dr. Tom Dayspring, who returns to the drive for another deep dive

00:01:08.940 into lipidology, but this time through the lens of the brain. Tom's been a frequent guest on the

00:01:13.720 podcast and has had an extraordinary career. He's an extraordinary teacher, a mentor to me personally,

00:01:18.900 along with many others, and of course, a colleague of mine for many years now in the practice. He's

00:01:23.040 one of the most thoughtful lipidologists I know with a very remarkable ability to take

00:01:27.280 complex physiology and make it not only clinically relevant, but understandable.

00:01:31.760 In this conversation with Tom, we cover the fundamentals of cholesterol transport in the

00:01:35.220 body, mostly just so that those who are coming to this for the first time or frankly don't

00:01:39.420 remember our earlier discussions on this have the baseline. But then we really focus on the brain.

00:01:44.180 We talk about why the brain's cholesterol system is almost entirely separate from the peripheral

00:01:48.640 system, that is the rest of the body. And we talk about the role of APOB, which I've talked

00:01:52.740 about a lot in APOA1, and specifically APOE as it pertains to cholesterol. So we talk about how

00:01:58.880 the APOE genotype relates to Alzheimer's disease risk, which is something we preferred to a lot,

00:02:03.460 but then the link between APOE, cholesterol, homeostasis, amyloid, and tau. What we know

00:02:09.260 and what we don't know about the effects of statins, ezetimibe, omega-3 fatty acids,

00:02:13.500 and then the emerging CETP inhibitors on brain health. This is a technical conversation. I won't

00:02:19.940 shield us from that, but it is an important one, especially for anyone trying to understand the

00:02:24.900 relationship between lipid-lowering therapy, cardiovascular disease risk, and neurodegenerative

00:02:30.700 disease. There's a lot of misinformation around this, and so unfortunately, you have to kind of

00:02:36.740 get into the details if you want to understand these complex relationships. So without further

00:02:40.800 delay. Please enjoy my conversation with Dr. Tom Dayspring. Hey Tom, great to be with you again as

00:02:52.120 always. For sure, Peter. This has become a bit of a routine for us. We've done it, but I love the way

00:02:58.900 we interact on this topic. Today we're going to talk about some different things. We're going to

00:03:03.240 really focus on a topic that's really become an enormous passion of yours and your curiosity

00:03:08.820 drives so much of your learning and then by extension our learning in the practice. So I

00:03:14.860 want to kind of go on a journey with you into this idea of cholesterol in the brain. It's obviously

00:03:20.500 a very important topic for reasons that we'll get into, but I think before we do, it is worth making

00:03:27.020 sure that everybody's starting from the same sort of knowledge base or singing from the same sheet

00:03:33.820 of music, as some might say, as it pertains to lipids. So I know that you and I have discussed

00:03:39.200 this in great detail elsewhere, and I realize that not everyone will have seen that, and even

00:03:43.840 if they have, they might not recall. So let's start at the very beginning in a very short

00:03:48.460 sort of five-minute version. Let's talk through the idea of cells in the body making cholesterol

00:03:54.980 and how they have to move that cholesterol around the body in the periphery, just the sort of the

00:03:59.540 nuts and bolts of it. Yeah. As you've stated many times, cholesterol is essential for human life

00:04:05.840 because it's used for making some critical things, but its most important function is it

00:04:10.980 positions itself in the cell membranes and every cell in our body and cell membranes regulate the

00:04:17.420 integrity, what gets in, what gets out of cell. So evolution has given every cell in the body

00:04:23.640 the power to de novo synthesize cholesterol? A little bit. Now, each cell needs a minor number

00:04:30.520 of molecules or so. But if it does that, we got great cell membranes and those cells are

00:04:36.200 functioning happily or so. But we also know, and people sometimes don't understand this,

00:04:42.360 like so many things, in excess of anything can be harmful. So if any cell somehow has

00:04:48.400 oversynthesized cholesterol, accumulated cholesterol, and has excess molecules. Cholesterol

00:04:54.300 has the ability to crystallize, which is toxic to a cell. It will kill the cell.

00:04:59.980 So evolution has also given cells the ability to export cholesterol out of its cytosol into

00:05:07.040 the plasma. But you know, you've talked many times, lipids are hydrophobic. They cannot

00:05:13.260 circulate in plasma, which is an aqueous or water solution. So again, evolution said no problem.

00:05:20.020 Evolution has given us proteins that can bind and adhere to lipids and then wrap them into particles

00:05:27.200 that are the lipoproteins. And that's how lipids, cholesterol, triglycerides, and numerous other

00:05:33.360 lipids that we don't have to mention circulate in our bloodstream. So if a cell E fluxes cholesterol

00:05:39.060 out, it joins on a protein. The protein happens to be called APOA1, which is sort of the structural

00:05:46.720 protein of our high-density lipoproteins. So that's how HDLs are created. They accept cholesterol

00:05:53.300 from whatever cell in the body is effluxing it or so. We have another family of lipoproteins

00:06:00.200 that are much bigger than the HDLs, and those are produced in the liver. One type is produced in the

00:06:07.820 small intestine, and they belong to the APO-B family of lipoproteins. And the difference

00:06:13.660 between them and HGLs is their structural protein is this very large peptide called

00:06:18.920 apolipoprotein B. The intestine makes a full-size APO-B. The intestine makes a truncated one.

00:06:25.920 We call the, excuse me, the hepatic APO-B, APO-B-100. And the intestinal produced one,

00:06:32.940 because it has 48% of the molecular weight of 100 is ApoB48. So when the intestine makes a

00:06:39.620 chylomicron, which traffics absorbed fatty acids, which become triglycerides, absorb cholesterol

00:06:47.180 into the bloodstream, it's in an ApoB48 particle, a very transient postprandial particle.

00:06:54.000 The liver manufactures ApoB particles. One is a very low density lipoprotein. It's quite big

00:07:01.580 because it's packing the triglycerides, which like the chylomicrons, it transports to muscles

00:07:07.060 and fat cells primarily, and then returns to the liver. Some of the VLDLs, as they lose the

00:07:12.780 triglycerides, they shrink and they become something called either a VLDL remnant,

00:07:17.980 a very transient particle called an intermediate density lipoprotein, which rapidly becomes a

00:07:23.900 low density lipoprotein or LDL. But the liver also has the ability to de novo manufacture and

00:07:30.660 secrete LDLs also. So our LDLs that are floating around have two sources. They're sort of like

00:07:36.420 the son of VLDL or they're, hey, a liver produced one. Now the ApoB particles carry a lot of lipids,

00:07:44.920 triglycerides primarily in the VLDL. The LDL is very interesting. It's pretty much a cholesterol

00:07:50.420 carrying particle, X amount of triglycerides, but it has the longest plasma residence time of

00:07:56.740 anything in the ApoB family. It can last three to four, even five days in some circumstances.

00:08:03.520 Ultimately, just like the VLDL, it gets cleared by the liver expressing and sticking into the

00:08:09.960 plasma, something called an LDL receptor, which binds to these ApoB particles and pulls them

00:08:15.860 into the liver. And then the liver digests them and does whatever it wants with the component

00:08:21.000 parts of the lipoprotein. The LDLs hang around for that amount of time. And this is not well

00:08:28.000 recognized because they interact with the HDLs, something totally not well known. If we look at

00:08:34.580 all the lipoproteins in the body, 90% of them are HDLs and the rest of the ApoB family. Now,

00:08:42.140 the ApoB family traffics far more lipids because of their size. You know, the volume of the spear

00:08:47.860 is a third power of the radius. So a couple of nanometer increase in diameter, boy, a lot more

00:08:53.580 lipids can be carried. But after the HDL has sucked out all of the cholesterol from whoever it has,

00:09:00.160 it becomes a big fat mature HDL. Now it has to do things with that cholesterol.

00:09:05.560 It has the option of delivering it to steroidogenic tissue that make cortisone or gonadal

00:09:11.120 hormones. It can bring it to the adipocytes, the cholesterol storage organ, or of course,

00:09:16.000 it can return it to the liver and even now the small intestine. But a lot of what an HDL does

00:09:22.460 is it transfers its cholesterol mass into the ApoB particles, the majority of which are LDLs

00:09:29.420 because of its long plasma residence time. So if an HDL, we've always been taught they do reverse

00:09:35.280 cholesterol transport and they can, they can bring it back to the liver or the gut. But interesting,

00:09:40.780 if they send their cholesterol to an LDL, the HDL becomes very small and it starts its journey all

00:09:47.620 over again. And then the LDL says, thank you, HDL. I'll take your cholesterol and I'll return

00:09:52.940 it to the liver. So what we used to think was a very simple reverse cholesterol transport system

00:09:59.500 becomes an indirect RCT, meaning an LDL is bringing it back to the liver or direct where

00:10:07.320 the HDL will bring it back. Total RCT is the sum of both. Most people are not aware that the

00:10:13.840 primary function, why we have LDLs, is to return cholesterol to the liver. Everybody thinks it's

00:10:19.320 delivering cholesterol to cells. Almost never, because every cell can make all the cholesterol

00:10:25.380 it needs. Now, in an emergency, any cell can upregulate an LDL receptor and pull in the

00:10:33.380 and LDL if it needs it, but just doesn't happen for the most part. And this is one reason,

00:10:39.240 and we're going to talk about it because it's pertinent to the brain. If LDLs are bringing

00:10:43.740 cholesterol back to the liver, if we can induce that with some of the drugs that we have that

00:10:48.820 make LDL receptors express and stay expressed longer, we will drop LDL cholesterol levels in

00:10:55.680 the plasma extremely low. And as we get deeper into the brain, unfortunately, a prevalent

00:11:02.760 belief out there in the real world is I don't ever want to lower LDL cholesterol too much

00:11:08.600 because I'll deprive the brain and I'll injure the brain. And soon we'll talk about why that

00:11:14.120 is not true. So that is what you said. This is the peripheral way that our body handles

00:11:20.600 cholesterol. By peripheral, anytime we say peripheral, we mean anything that's not in the

00:11:25.080 brain. So the brain lipid and lipoprotein system that we're going to talk about has almost nothing

00:11:31.660 to do with the plasma, a transportation of lipids and lipoproteins. And that is such a

00:11:37.700 crucial concept that must be understood. So what didn't I explain, Peter? I hope I've

00:11:43.980 touched on that in a rapid fashion. Yep. Let me just maybe synthesize some of those

00:11:49.560 points. So first off, maybe just even adding a little bit more context, the body does shuttle

00:11:55.500 a lot of things around plasma. Plasma is kind of the highway of the body, or at least the major

00:12:00.260 highway of the body. Obviously, there's the lymphatic system. And plasma is, as you said,

00:12:04.800 it's water. It has proteins in it, like hemoglobin and things like that within red blood cells,

00:12:09.820 but it's essentially water. And therefore, things that are water-soluble can transport easily. So

00:12:15.860 glucose doesn't need a transporter. We just have glucose floating around our bloodstream.

00:12:20.940 Ions, sodium, potassium, chloride, they don't need to be bound to anything to move around.

00:12:27.320 Conversely, steroidal hormones like testosterone or cortisol, they actually are virtually all

00:12:34.140 bound.

00:12:34.620 There's a slight amount that's free, but they're bound to albumin or sex hormone binding

00:12:38.000 globulin or things like that.

00:12:40.200 And of course, to your point, cholesterol, given how important it is that we can transport

00:12:45.380 this thing, we had to come up with a carrier.

00:12:47.480 These carriers are called lipoproteins, which gives rise to the name lipid protein.

00:12:53.360 lipid on the inside where it repels water, protein on the outside where it dissolves or is soluble

00:12:59.800 within water. And then again, you mentioned the two families, the APOA family, the APOB family.

00:13:06.760 We always want to make sure people know that when we're talking about the APOA family, it has nothing 0.88

00:13:10.320 to do with LP little a. That's a totally different APO lipoprotein, which we're not going to talk

00:13:15.420 about today, although we've got lots of content on that. You also mentioned how much the APOAs

00:13:20.640 outnumber the ApoBs in absolute numbers, but because they're so much smaller, the total

00:13:27.560 cholesterol carrying capacity is much greater in the ApoB family. And a way for a person to

00:13:33.200 appreciate that is to look at their lipid panel. If you see that your total cholesterol is 200

00:13:38.320 milligrams per deciliter, you'll easily notice that the sum of your LDL and VLDL cholesterol,

00:13:45.500 it could easily be 140 of that 200 milligrams per deciliter, whereas the HDL cholesterol might

00:13:52.220 only be 60 of that. So again, many more in number, but much less in cholesterol carrying capacity.

00:13:58.640 And then of course, you talked about this idea of reverse cholesterol transport. We have the

00:14:02.120 indirect and the direct. We've talked about those in the past. But again, I think the most important

00:14:07.060 takeaway that I get from what you said is the old version of that, which is that it's HDLs that do

00:14:12.840 at all is untrue. The LDLs do more by volume. I guess one question I would have for follow-up

00:14:19.540 is for the person who says, but Tom, I understand everything you're saying, but if LDL is so

00:14:27.780 important for reverse cholesterol transport in the periphery, what happens as LDL goes down?

00:14:34.300 Is that a bad thing? Am I losing the ability to return cholesterol to the liver?

00:14:40.380 No, that just means that if LDL cholesterol goes down, if you really even look at your

00:14:45.380 total cholesterol, it's going to go down too.

00:14:48.480 So LDL's function is to bring cholesterol back to the liver.

00:14:52.040 So if your LDL cholesterol is, there's just not a need to get cholesterol back to the

00:14:57.820 liver.

00:14:58.100 The cells are not effluxing as much cholesterol because they're in cholesterol balance.

00:15:02.900 The HDLs are sending, transferring less cholesterol to the LDLs.

00:15:08.840 And so the system is in, it's a very operational system and they all talk to one another.

00:15:16.000 The nuclear transcription factors that regulate all of the, some of the mechanisms I spoke

00:15:21.840 to are in balance.

00:15:23.380 So low LDL cholesterol, yes, there would be less cholesterol going back to the liver,

00:15:29.000 but there's no need for cholesterol to go back to the liver because it's in balance

00:15:33.420 in all the other cells.

00:15:34.580 Tom, another question that might be worth addressing here is what is the amount of total

00:15:41.720 cholesterol in the body that is in the plasma, i.e. that which we measure versus not in the

00:15:49.180 plasma? I mean, cholesterol is this essential molecule for life. We've talked about how it

00:15:52.960 makes up cell membranes. It forms the basis of producing many hormones. But if I were to measure

00:16:00.100 somebody's serum cholesterol and I measured, you know, again, total cholesterol of 200 milligrams

00:16:05.000 per deciliter, I could calculate how much cholesterol is in their blood because I know

00:16:10.160 what their circulating blood volume is. And, you know, presumably I could do the math and it would

00:16:13.840 be a few grams of cholesterol. How does that compare to the total body store of cholesterol?

00:16:19.840 Well, it's much smaller. I mean, most of the cholesterol in the body is within the cells

00:16:24.680 of our body. And we've already divided it. Hey, the body is the peripheral system and the brain

00:16:31.820 system. And if you look at total amount of cholesterol, most of it in some is in all of

00:16:38.740 the periphery. That means your liver, every organ you got, your skin, every membrane in every cell

00:16:44.940 in our body. So that's the mass of total cholesterol. And the brain has its own component

00:16:50.480 because they don't interact. But in the plasma, it's interesting. Most of the, of course,

00:16:57.300 you would think all of the circulating plasmas within lipoproteins, but it's not.

00:17:03.140 This comes to a surprise to many people too. The biggest carrier of cholesterol in our bloodstream

00:17:09.760 is in our red blood cells. Yeah. Because they're cells, they have cell membranes and they're big.

00:17:15.400 they're vastly larger than a lipoprotein. So they actually carry more milligrams of cholesterol than

00:17:21.980 do our, everybody thinks it's the lipoproteins. It's not. So that's how it's distributed in the

00:17:28.580 blood. There is no free cholesterol. I mean, there's a minuscule amount on albumin, not much,

00:17:33.520 but that's it. It's in a lipoprotein or it's in a red blood cell membrane. Then we have the organs

00:17:39.080 of the body and it's another question that you can trick up people because if you tell ask the

00:17:45.780 average person or even physician even lipidologist where's most of the cholesterol in the body or

00:17:52.020 what organ has the most cholesterol and everybody says the liver and wrong it's not even close the

00:17:59.140 brain of all the organs in the body has 20 times more cholesterol than does the liver the liver

00:18:06.160 I've read the brain has like 20 to 25 grams. There's like 140 grams total in the body of

00:18:12.720 cholesterol where the liver would have three to five grams. Now, one reason is, and we're going

00:18:18.560 to get into this, the brain holds onto cholesterol like the bank holds onto its gold in the vault and

00:18:25.580 everything. It doesn't, but whereas cholesterol, the liver is just sort of a handling station.

00:18:30.400 whatever cholesterol a liver has it's sent out or it's effluxed into the bile through bile acids or

00:18:36.660 free cholesterol so the liver is like a transfer station so it stores a little bit of liver because

00:18:42.980 it always has to have a pool of cholesterol to do what it does whereas the brain holds on to

00:18:48.140 its cholesterol and this is another physiologic point we'll have to get into yeah the liver is

00:18:53.460 more like yeah sorry the the brain is sort of like pardon me the liver is more like a bank with money

00:18:58.260 which is it's got a high flux. It takes a lot of deposits in, but then of course the only way it

00:19:05.380 makes money is by loaning out or distributing that capital and putting it to work. So yeah,

00:19:11.100 it's a good point, which is, as we'll talk about, it's the storage of cholesterol within an organ

00:19:16.260 versus the transfer through the organ. So going back to finish the swing on that point, of course,

00:19:22.380 I just want the listener to be cognizant of the idea that if your peripheral cholesterol

00:19:28.340 goes down by 50%, 75%, right? If your total cholesterol falls from 200 to 100 milligrams

00:19:36.960 per deciliter, it's tempting to think, oh my gosh, my total body cholesterol has fallen by half.

00:19:45.800 In reality, it's fallen by a couple of percent because it's a tiny, tiny amount.

00:19:51.380 Yeah. Between cellular cholesterol and circulating cholesterol.

00:19:55.780 Yeah. So this is definitely one of the misconceptions people deal with. And again,

00:20:00.240 although we're not going to focus on it, we'd be remiss to be sitting at this point in the game and

00:20:05.580 not mention why one might want to have a total plasma cholesterol of 100 milligrams per deciliter

00:20:12.340 as opposed to 200 cholesterol. He's like, why are we in the business of lipid lowering if we're

00:20:17.300 trying to help people avoid certain diseases. And how does just lowering that tiny fraction

00:20:24.280 of the total body's pool have such an outsized effect on atherosclerosis?

00:20:30.380 Yes. So now we're into the pathology associated with cholesterol. And we know the leading global

00:20:37.660 killer is atherosclerotic disease. It's not, hey, the industrialized countries, it's all over.

00:20:44.120 people are dying of heart attacks for a variety of reasons. And I always like to say, if you have

00:20:49.300 atherosclerosis, there's one sine qua non, you have cholesterol buildup in your artery wall.

00:20:55.000 If we do not have cholesterol buildup in our artery wall, you do not have the disease called

00:21:00.540 atherosclerosis and you can't suffer the consequences thereof. So the next question is,

00:21:06.140 all right, Tom, well, how in the world does cholesterol get into that artery wall?

00:21:10.060 It's not like the artery is oversynthesizing cholesterol and building it up.

00:21:14.500 That is not happening.

00:21:16.060 So that means we've already described the cholesterol is floating in our highways in

00:21:20.840 the plasma.

00:21:21.720 So how does cholesterol get from the dump trucks, the lipoproteins that are carrying

00:21:26.820 it into the artery wall?

00:21:29.360 And this is why one of the reasons we talked about the APO-B containing lipoproteins.

00:21:35.040 By the way, henceforth, we may refer to beta lipoproteins.

00:21:38.660 That's the ApoB family. The HDL family, sometimes we call them the alpha lipoproteins.

00:21:45.080 But we now know, and this is really not even up for discussion. You've done podcasts on this and

00:21:51.540 the references on it. You have that great slide, the Ferrin slide, where every single trial that's

00:21:58.280 ever been done, every Mendelian analysis of lipids and lipoproteins shows the more you lower

00:22:04.800 cholesterol, the less atherosclerotic events happen. So we now know, we've already told you,

00:22:11.820 it's the beta lipoproteins that are carrying most of the cholesterol in the bloodstream.

00:22:16.460 So if a beta lipoprotein, an LDL or a VLDL, and because of its residence time, the vast majority

00:22:23.480 of those are LDLs, exceed a certain threshold number, they will enter the artery wall. It's

00:22:30.880 a simple diffusion process, you could have endothelial dysfunction and they get pulled

00:22:35.800 in, they get in a little easier, but they get in, even in healthy artery walls, once you exceed a

00:22:41.900 certain concentration of ApoB particles. So, and once they enter the artery wall, and this would

00:22:47.740 be another whole podcast, all sorts of things, they get trapped, they get aggregated, they get

00:22:53.120 oxidized, and the immune system sends in white blood cells that engulfs them, and that creates

00:22:59.120 a cholesterol-laden macrophage, the foam cells, they stick together creating plaque.

00:23:04.960 So it's the particle number. And there are assays that we can get LDL particle numbers or VLDL

00:23:12.640 particle numbers if you want them. But since there is one ApoB on every one of those particles,

00:23:17.700 we simply measure ApoB, one ApoB per particle. Once your ApoB level starts to exceed certain

00:23:25.500 thresholds, atherosclerosis is very likely to occur. The main driver of your ApoB concentration

00:23:32.860 is two things, of course, a little bit of production out of the liver, but most of the

00:23:39.000 escalation becomes is due to defective clearance of the ApoB particles from the plasma, meaning

00:23:45.420 those LDL receptors, the liver, for whatever reason, is not expressing enough of them to clear,

00:23:51.600 to keep the ApoB concentration physiologic in the bloodstream. So once ApoB particles are not

00:23:58.560 cleared, then there's only one other option for them. They have to invade an artery wall.

00:24:03.880 So it's the ApoB concentration and they deliver cholesterol and that explains atherogenesis.

00:24:11.100 And in the old days we used to, and still do, what are our ways of estimating ApoB concentration?

00:24:18.080 most of it is LDL particles. We look at LDL cholesterol. And for decades, that has been

00:24:24.020 the poor man's surrogate that you have too many ApoB LDL particles floating around.

00:24:30.280 We use VLDL cholesterol. Triglycerides divided by five is sort of an estimate. Is there too much

00:24:36.180 cholesterol in the VLDL particles? We don't have as great a test on that. But the vast majority of

00:24:41.920 these dump trucks entering your artery wall are LDLs. So ultimately, this podcast is not directed

00:24:48.840 at it, but if we can make the liver express more LDL receptors or let the LDL receptors recycle

00:24:54.460 more, you will have increased clearance, you will lower the ApoB, and every ApoB particle that goes

00:25:01.800 into the liver is one less that's going into your artery wall. And that's basically the pathophysiology

00:25:08.680 of atherogenesis. It's those ApoB dump trucks. Follow up on that point. Again, let's take two

00:25:14.880 individuals whose ApoB concentration and documented LDL cholesterol level is above that physiologic

00:25:23.500 threshold, such that diffusion is going to favor entry of the LDL, the low-density lipoprotein,

00:25:31.700 into that subendothelial space to begin that cascade that we talked about. Everybody has the

00:25:37.440 story of, you know, my grandmother is 90 years old. She's got an LDL cholesterol of 160 milligrams

00:25:45.100 per deciliter. Her total cholesterol is over 200. I mean, she probably smokes and she hasn't had a 0.90

00:25:52.480 heart attack. Whereas you can see another person with that same lipid profile that's having their

00:25:57.120 first heart attack at 51. I don't expect you to have an answer for this because I just think there

00:26:01.800 are certain things we can't understand. We don't understand why not all smokers get lung cancer.

00:26:05.120 We just don't understand a lot of things, but what do you think are the most compelling

00:26:09.900 explanations for why we don't have complete and total homogeneity of risk factor and disease

00:26:17.040 when we confine it to this disease?

00:26:19.260 I mean, we don't have it for any disease, but what do you think is the best explanation

00:26:22.640 for a disease in which we so well understand the physiologic steps?

00:26:28.220 Sure.

00:26:28.600 Well, as I mentioned, if cholesterol gets in your artery wall, you have the disease

00:26:32.260 and it's the ApoB particles bringing them in.

00:26:34.620 but that is not the only etiologic reason why one would have atherosclerosis. There are a number of

00:26:41.600 other factors that go into play, and it's the rest of your health. Your metabolic health is a major

00:26:47.720 concern. If you are insulin resistant, up to type 2 diabetes, you have chronic inflammation in the

00:26:54.140 body, you have endothelial cell damage in the body, so it's easier in those people for these

00:27:00.180 particles to get in earlier in life and generate in plaque. We should make the point that this

00:27:07.080 ApoB entry into the artery wall is an incredibly slow process. It takes decades to develop.

00:27:13.320 And this is why the concept now is not only lower is better, but the longer you keep things low

00:27:18.860 with ApoB is better. So your blood pressure would be a factor. Smoking, as you said,

00:27:24.460 if you have some autoimmune disease that's contributing to inflammation. We know people

00:27:29.280 who have chronic inflammation have increased atherosclerosis, collagen diseases, rheumatoid

00:27:36.300 arthritis. They have lifelong inflammatory factors going on and other abnormalities that

00:27:42.760 weaken the arterial defense against atherosclerosis. Oxidative processes is a big part of atherogenesis.

00:27:50.500 So if that is going on in the body, but sometimes we do see, like you said, great grandma who smoked

00:27:57.940 all her life and has high LDL cholesterol and why no plaque? And there are forces at play that we

00:28:04.800 just do not understand. There's other protective whatever going on in their body that we have not

00:28:11.420 been able to identify even genetically, or we're testing this test. Oh, they got some elevation of

00:28:18.440 molecule Z. It's protecting them. Something's going on, and one day we'll ascertain that.

00:28:25.320 You know, as the polygenic risk scores come into, if we do them early in life, it can sort of predict who is going to make it to 80 and never have a heart attack and who is not, because they're looking at a multitude of genetic things that you are never looking at one at a time in an individual patient.

00:28:42.820 Look, genes control everything. They are genetically blessed. But important thing to make is don't ever think because your LDL cholesterol is 200 that I'm one of them because there's no way to know that. Why play Russian roulette and think it's not going to bother me when for the vast majority of people it does create havoc and pathology.

00:29:04.020 Yeah. So let's now talk about the brain. So we've got these ApoB and ApoA lipoproteins,

00:29:12.500 the HDLs and the LDLs predominantly. You mentioned though that the brain has the greatest

00:29:18.040 source of cholesterol in the body, greatest storage source of cholesterol in the body.

00:29:23.040 Does the brain need to rely on any of the periphery's cholesterol? And if so, can ApoB

00:29:30.980 and ApoA lipoproteins get in there and deliver cholesterol as needed?

00:29:35.380 Well, the quick answer to that, and then I'm going to elaborate, is what's going on with

00:29:40.240 cholesterol in the brain, how much cholesterol is stored in the brain has zero to do with what

00:29:46.340 is floating in the plasma. So there are certain lipoproteins that we'll talk about that can work

00:29:52.160 their way into the brain, but the ApoB-containing particles, which carry the vast majority of

00:29:58.280 cholesterol cannot. They're much too big to pass through that, what we call the blood-brain barrier,

00:30:03.600 which is actually a barrier that separates the brain from the periphery as we've talked.

00:30:09.760 But I like to start to give you an idea about why has the brain got so much more cholesterol? Why is

00:30:15.140 it storing it so much more than, say, the liver or any other organ in the body? Well, as we are

00:30:22.400 in utero with mom. And the second and the third trimester, the fetal brain is already starting to

00:30:28.340 de novo synthesize its cholesterol because evolution knows it's going to need cholesterol

00:30:33.360 because the brain probably has more cell membranes than any other tissue put together,

00:30:38.740 especially our neurons. Those cell membranes are kind of critical on do our neurons work or not,

00:30:44.060 whether the neurons work or not is do we work or not normally or so. So every brain cell starts

00:30:51.420 producing cholesterol in utero. Very quickly, brain cells, it's very easy. You have neurons,

00:30:59.800 the ones I allude, but any cell that is not a neuron in the brain is called the glial cell.

00:31:05.060 And there's only three of them. You have astrocytes, which in the adults produce a lot

00:31:11.620 of the cholesterol. We have oligodendrocytes. It's a big word. And they produce about 70%

00:31:17.980 of the brain cholesterol, because one of the mega things the brain does with cholesterol is create

00:31:23.880 myelin, which sheaths every axon and dendrite, the nerve endings that are in our body. So that is a

00:31:31.540 big, big reason why the brain stores and has so much cholesterol. It's in myelin. The other glial

00:31:37.960 cell in the brain is a microgliocyte, and they are the brain immune cells. So they are the last

00:31:46.540 remaining cell. So in the utero, the day we're born, there's no more mom contributing cholesterol 0.84

00:31:54.380 to the brain. It's the brain making it itself. And every cell I just mentioned is overproducing

00:32:00.580 cholesterol because the brain knows as it grows and grows, it's going to need more and more

00:32:06.740 cholesterol for all these cell membranes. So everybody that can produce cholesterol has to do

00:32:11.900 it, knowing that the brain cannot extract any cholesterol from what's circulating in the plasma.

00:32:18.400 You've mentioned it many times on your podcast. If you take a two-year-old and measure their LDL

00:32:24.180 cholesterol, it might be 30 milligrams per deciliter. Yet that is the time when the brain

00:32:29.980 is growing more than it ever will. Between birth and age of 10, the brain is expanding to its adult

00:32:36.900 size. And it can't do that without cholesterol. So it's super manufacturing cholesterol. But it's

00:32:44.640 doing it in people who, little children who have very low detectable LDL cholesterol. So that tells

00:32:51.420 you basically physiological levels of circulating cholesterol have nothing to do with a growing or

00:32:57.740 a normal brain. At around the age of 10, really much the adult brain size is foreign. So at that

00:33:05.620 point. There's a readjustment of cholesterol synthesis in the brain. Oligodendrocytes keep

00:33:11.280 making it. They always will. Microgliocytes, they don't have to make that much. Astrocytes continue

00:33:18.160 to produce it at a high form, but there's one cell that stops producing cholesterol and it's the

00:33:23.700 neurons. When the brain is full adult size, the neuron says, no, no, no, I'm not going to make

00:33:30.520 any more. I want the astrocytes to make it and send it to me. And there's a simple reason it

00:33:35.260 does that. We've in our earlier podcast discussed the very complex cholesterol synthesis pathways.

00:33:42.280 It's actually 37 steps. Every step is a different enzyme. Every step requires ATP. So any cell to

00:33:51.640 synthesize one molecule of cholesterol consumes over 30 molecules of ATP. The neuron, of course,

00:33:59.680 is the most active cell in the brain because it's firing off all these action potential in their

00:34:04.560 synapses all day long. And that requires ATP. So the neuron does not want to waste ATPs making

00:34:13.400 cholesterol. If it can get it elsewhere, the neuron starts using ATP for its functioning.

00:34:20.160 So, and then it falls on the astrocyte. So that's a little bit about cholesterol production

00:34:26.200 in the brain. All of the cells can do it, but at a certain point, the neurons say,

00:34:31.080 I don't want to do it anymore. Astrocytes, can you please make cholesterol and get it over to me?

00:34:37.580 And this is where we get into the brain lipid transportation system. Because in the blood,

00:34:43.720 as we've enumerated, lipids travel within the lipoproteins in the plasma. Well, in the brain,

00:34:51.060 the cholesterol that's going back and forth between cells doesn't use the blood.

00:34:55.140 it uses the brain interstitial tissue, which is called the matrosome. So if we take the brain,

00:35:02.740 it's this connective tissue and there are zillions of these cells in them, the glial cells in the

00:35:07.680 neurons. Now they're very close together, but they're not contiguous. They're not binding to

00:35:13.280 each other. So if an astrocyte produces cholesterol molecules in the neurons over there saying,

00:35:19.140 hey, I need that, send it to me. We have to have a brain cholesterol transportation system

00:35:25.000 or a brain lipid transportation system. And so what do the astrocytes do? Same thing that happens

00:35:31.480 in the periphery. It makes a lipoprotein, but there's going to be a big difference here.

00:35:36.960 So the first thing the astrocytes are going to have to do is synthesize cholesterol.

00:35:41.780 Very quickly, we won't elaborate in depth, but we've had podcasts on this before.

00:35:46.820 One of the cholesterol synthesis pathways goes through the next to last sterol, penultimate

00:35:53.400 cholesterol. And in the brain astrocytes, it's called desmostrol. And then desmostrol becomes

00:35:58.680 cholesterol. So we'll probably talk about this is one way why we can measure desmostrol in the

00:36:04.900 cerebral spinal fluid. Nah, that's kind of hard to do. But in the plasma, it correlates with

00:36:10.660 brain cholesterol production. So the astrocyte makes cholesterol. It's now going to obviously

00:36:17.160 have to wrap it with a protein, an apoprotein, so we can shoot it out into the matrosome,

00:36:22.800 where it can travel, swim over and get to the neuron. And here's the difference. In the

00:36:29.280 periphery, we said, hey, the structural proteins are ApoB and ApoA. In the brain, it's the famous

00:36:37.300 apolipoprotein E. And ApoE, many people know that has something to do with the brain because we know

00:36:44.680 there are types of ApoE that are associated with cognitive disorders and Alzheimer's disease.

00:36:50.280 But let's just stick to the ApoE protein. So the astrocyte synthesizes, it binds the cholesterol

00:36:57.240 and it becomes a little lipoprotein, which it secretes into the matrosome. But it's an ApoE

00:37:05.480 containing lipoprotein. Now, if we could take out those ApoE containing lipoproteins and put them

00:37:12.560 in a centrifuge, they would sink right to the bottom of the centrifuge. But what else would

00:37:18.260 be sinking to the bottom of the centrifuge, high-density lipoproteins in the plasma.

00:37:23.120 So the brain lipoproteins are referred to as HDLs because they have the buoyancy and density of a

00:37:30.020 plasma HDL. But they're very different because the plasma HDL will have two, three, four copies

00:37:37.600 of ApoA1. The brain HDL will have a couple of three copies of ApoE. And that is the big difference.

00:37:45.400 Now, once it's in the matrosome, this particle, it continues to mature. Cholesterol becomes

00:37:52.100 cholesterol ester, goes to the center of the particle, and it becomes a big fat particle.

00:37:57.060 But remember, its mission is to deliver cholesterol to the neuron. So the neuron's

00:38:02.580 going to have to grab that ApoB-containing particle and internalize it or grab it and

00:38:07.500 delipidate it. So guess what the neuron expresses? Low-density lipoprotein receptors.

00:38:15.400 And that creates confusion because if somebody says, oh, I know the brain, the neurons have LDL

00:38:20.980 receptors, so there have to be LDLs in the brain. No, because the LDL receptor has affinities for

00:38:29.580 just a couple of apoproteins. In the periphery, the LDL receptor is looking for ApoB100.

00:38:36.800 But in the periphery, even ApoE can bind to an LDL receptor. But in the brain, the LDL receptor

00:38:44.340 only binds to ApoE containing lipoproteins because there are no ApoB containing lipoproteins.

00:38:50.360 So it's the same bond receptor. And this is why I think we should stop calling it the LDL

00:38:55.900 receptor. We should call it the ApoB, ApoE receptor because that's what it recognizes.

00:39:01.880 So Tom, I'm actually quite confused by this. So there's a lot I want to back up on. I'll just

00:39:07.160 start with that point. So let's back up to the liver for a moment. The liver has got this receptor,

00:39:13.040 which we will continue to refer to as an LDL receptor. When an ApoB particle, an LDL, a garden

00:39:20.260 variety LDL makes its way to the liver, it has one and only one ApoB around it. Can you briefly

00:39:29.940 explain confirmationally how that LDL interacts with the LDL receptor? What is it about the

00:39:39.040 ApoB protein that enables the key to fit into the lock? There's a very small segment of the ApoB

00:39:47.740 receptor that's called the LDL receptor binding domain. Excuse me, on the ApoB, there are certain

00:39:53.780 amino acids that line up and they have a surface charge. And here's the LDL receptor. Now the LDL

00:40:02.380 receptor has a certain segment that is called the ApoB recognition domain. There are certain amino

00:40:08.260 acids there that create certain electrostatic forces. And if the domain on ApoB and what

00:40:15.980 determines is that sticking out properly is the confirmation of ApoB, that explains the difference

00:40:21.740 clearance rates between big LDLs and small LDLs as opposed to normally constructed in size LDLs

00:40:29.000 that have a normal ApoB confirmation. They have much higher clearance. The small LDL where that

00:40:35.940 domain may not be exposed as readily or the big LDL where it's not where it should be. The LDL

00:40:43.040 receptors don't as easily recognize big LDLs or small LDLs. And that's why people with small LDLs

00:40:51.180 or even big LDLs often have very high LDL particle counts because clearances decrease.

00:40:56.820 So there are certain just small areas on the LDL receptor and the ApoB that if they align properly,

00:41:03.180 you have great clearance. New news is just discovered and published last year from our

00:41:09.820 friends at the NIH is LDL receptors act as a dimer. There's actually two of them that express

00:41:16.300 at the same time. It's like two lobster claws and they grab two LDL particles at the same time. So

00:41:22.380 that's sort of irrelevant to just understanding the LDL receptor clearance process. So that

00:41:28.820 explains part of the extended plasma resonance times of LDLs. How is the APO-B conformed?

00:41:35.660 So Tom, given that the size of the LDL within a variation of normal can impact clearance,

00:41:44.060 it really surprises me that that same LDL receptor can easily find somewhere on the APO-E

00:41:52.980 wrapping a very, very, very small lipoprotein in the brain, enough of a conformational match

00:42:02.120 to make that work. So that is not only news to me, but very difficult to wrap my little

00:42:08.640 cholesterol-rich brain around because I would think that the ApoE lipoprotein being so much

00:42:15.920 smaller than an LDL and being much closer to an HDL, would never be able to find it,

00:42:23.140 even with complete homology between that section of APOE and APOB, which presumably must be the

00:42:29.260 case or you wouldn't have to match. Yeah. The primary reason where APOE gets

00:42:34.800 involved with clearance of lipoproteins is on chylomicrons and VLDLs. They carry several copies

00:42:40.820 of ApoE per particle, unlike the ApoB, which is one copy per particle. So when they are fully

00:42:47.160 full of triglycerides, they're very big. The ApoB is distorted in a certain way.

00:42:53.340 Now, the receptor in the liver that's going to clear VLDLs and chalomicrons is called the LDL

00:42:59.060 related receptor one. So it only has an affinity for ApoE. So it's the LRP that clears most of the

00:43:09.440 ApoE containing particles, the chylos and the VLDLs. And that's why they have such short plasma

00:43:15.920 residence time. I'm going to mention it now. Oh, sorry. But Tom, I was asking a different

00:43:20.840 question, which maybe I misunderstood something you said. I was asking about the neuron with its

00:43:25.980 LDL receptor. How does the neuron with an LDL receptor tag and pull a tiny, tiny, tiny lipoprotein

00:43:34.520 with an ApoE on it out of- And the real reason, and this is why I'm explaining to you how the

00:43:39.320 liver clears, VLDLs, and chylomicrons, because the LRP only recognizes APOE. And the brain not

00:43:46.560 only expresses LDL receptors, but they express a lot of the LDL receptor-related protein one,

00:43:52.500 which is an APOE affinity clearing. So, yes, the LDL receptor can clear some of the

00:43:59.120 APOE part, but it's the LRP that's doing most of it. The last receptor that the neuron expresses,

00:44:05.660 and we've talked about this is called the scavenger receptor B1 that binds to the HDL

00:44:12.400 and it delipidates it, but it's an ApoE recognizing scavenger receptor also. So everything in the

00:44:19.360 neuron is basically looking for ApoE and it gets it through, especially the LRP, which is only an

00:44:25.580 ApoE recognizing receptor and the scavenger receptor recognizes ApoA1, typically not in the

00:44:33.860 brain, but it can be, and we'll get to that also. But ApoE works well with the scavenger receptor

00:44:39.100 too. So very few LDLs in the periphery. I mean, maybe 2% of your LDLs have an ApoE on,

00:44:47.960 and mostly there's no ApoE. Although the LDL receptor can recognize it, it's a minor

00:44:54.420 clearance pathway, ApoE on an LDL. I want to go back to the synthesis. You alluded to this

00:45:00.520 briefly, we have two cholesterol synthetic pathways. I mean, one pathway that branches

00:45:04.640 and bifurcates into two pathways. And in each of those pathways, they make cholesterol,

00:45:10.440 but the intermediaries are quite different. So different enzymes and different intermediaries.

00:45:15.220 And we often refer to them thinking of what their penultimate molecule is. So you already referred

00:45:19.720 to one, which is the path that turns desmostrol into cholesterol. And then the other one, of

00:45:26.320 course, turns lithosterol into cholesterol. What is the relative balance of cholesterol

00:45:32.060 synthesis in the brain between those two pathways? Very interesting. In the periphery,

00:45:37.420 the vast majority goes through the lithosterol pathway. Very little goes through the desmostral

00:45:43.380 pathway. In fact, the primary cells that use the desmostral pathway in the periphery

00:45:48.200 are our steroidogenic tissues. All of our other cells, I mean, a little bit will go through the

00:45:53.720 desmosteral pathway, but most is lithosterol. So if you are measuring sterols in the blood,

00:45:58.900 lithosterol is up, you know, it's the peripheral cells that are overproducing cholesterol.

00:46:03.360 Very interesting in the brain, when I told you up to the age of 10, all of the cells are producing

00:46:09.220 cholesterol, including the neurons, the neuron synthesis pathway actually does go through

00:46:15.560 lithosterol. But at the age of 10, when the neuron decides I don't want to make cholesterol anymore,

00:46:21.320 there's no lithosterol being produced by the neurons. It's all desmostrol that's winding up.

00:46:27.340 If there's cholesterol molecules winding up in the neurons, it's through the astrocyte,

00:46:33.340 the block pathway going through desmostrol. Now in a pinch, if there's a cholesterol deficiency

00:46:39.520 in the brain, the neurons can start synthesizing cholesterol again, but in normal brain physiology,

00:46:45.840 that doesn't happen. So lethosterol is not used as a marker of brain cholesterol synthesis for

00:46:53.340 the big reason, even though there is some lethosterol pathway going on in the brain,

00:46:57.940 if you measured it in the blood, 95% of it is your other cells making it. Whereas if you measure

00:47:04.380 desmostrol in the blood, the majority of it reflects, correlates extremely high with cerebral

00:47:10.720 spinal fluid, desmostrol, and brain tissue, desmostrol. So that becomes a very cool marker

00:47:17.140 that we can actually measure in the bloodstream because desmostrol in the plasma correlates very

00:47:23.840 highly with cerebral spinal fluid and brain cholesterol. And why is that, Tom? That's

00:47:29.100 counterintuitive to me because they seem like completely independent pathways. Why should

00:47:34.080 the desmostrol you measure in the blood tell us anything about the cholesterol synthesis of the

00:47:39.360 I think, and you're better at figuring out these teleologic reasons than I, that evolution decided

00:47:45.860 there's one pathway that we're going to do in very critical areas. The brain, which only makes

00:47:52.280 its own cholesterol in stores and in the steroidogenic tissue, we want them to be dependent

00:47:57.280 on that pathway. Why? I don't have an answer for you on that, but that's what that pathway reflects.

00:48:05.340 All right. We'll come back to that because I know there's a clinically relevant reason

00:48:08.980 that we might want to think about that. Okay, so we've established that the neurons,

00:48:14.300 once they reach a certain age, want to start optimizing less around being general contractors

00:48:20.940 and construction workers and more around being architects because of the energy cost.

00:48:26.700 And we've also established that you have a different lipoprotein that is transporting

00:48:33.540 cholesterol in the brain so that the neurons can still acquire plenty of it from their neighboring

00:48:38.840 oligodendrocytes and presumably to some extent astrocytes. I do want to just make one point

00:48:44.620 clear for the listener, which we haven't really explicitly stated, but the astute listener of

00:48:50.100 course has already picked up on the fact that we've talked about APOE. And as you said, APOE

00:48:54.240 has a relationship to Alzheimer's disease. I just want to make sure people understand the difference

00:48:58.820 between APOE genes and APOE the protein. Because to date, through this discussion, we have only

00:49:07.000 spoken about apolipoprotein E, a protein. And this is denoted with a small a, small a, small p,

00:49:14.900 small o, big E. And that's when we're talking about APOE, the protein. But if you were to write

00:49:20.660 all caps, A-P-O-E, you'd be referring to the genotype. And of course you have two of these,

00:49:26.340 so you could be a 3-3 or 3-4-4-4-2-3, et cetera. You want to just explain the relationship between

00:49:32.280 those different six combinations of genotypes, everything from a 2.2 to a 4.4, and how the

00:49:38.740 different genes make different proteins. And then we should talk about why that's relevant.

00:49:46.140 Yeah. And it's a big part of this discussion. So the APOE protein comes in different shapes.

00:49:52.260 They're called isoforms. Peter has explained this many times. It's really only one different

00:49:58.020 amino acid in the darn protein that separates these, but just removing or replacing or putting

00:50:04.440 the wrong amino acid in the entire peptide changes its ability to bend in shape. And that will affect 0.95

00:50:10.780 what it can bind to, which is the crucial function of apoproteins. So there are the,

00:50:17.840 you inherit the genes from mom and dad, and that means one gene from mom, one from dad.

00:50:23.040 So you get one allele in your gene and the other allele from each. So was your mom an APO E2,

00:50:29.240 three or four? And likewise with dad, and you're going to inherit, there's several potentials.

00:50:34.940 You can be an E2, E2, E2, E3, E3, E4, E3, E2, E2, E4, E4, or E double homozygote for E4.

00:50:50.740 So depending which of those genes you attack, your ApoE protein is going to be constructed

00:50:56.840 a little bit differently, which is going to affect its ability to function whatever ApoE

00:51:02.620 is doing when it's stuck to a lipoprotein.

00:51:05.560 And the main thing it's doing, it's serving as a ligand to what things are going to bind

00:51:09.980 to or even what the ApoE will bind to other than the lipoprotein.

00:51:13.840 So, the type of ApoE you manufacture is critical to certain disease pathologies.

00:51:21.340 Peter can give you the exact indices.

00:51:23.420 The average person has an ApoE3, E3 genotype.

00:51:27.580 I believe it's about two-thirds of people that have that.

00:51:31.020 Far less people carry the ApoE2 gene, and especially ApoE2 homozygosity.

00:51:36.700 Peter, why don't you tell how many carry the E4 heterozygotes and the E4 homozygotes?

00:51:42.340 You have those percentages.

00:51:43.260 Yeah. I mean, again, it depends on the series you look at, but it seems about 55% of the

00:51:48.320 population are E3, E3, the so-called wild type. 20% to 25% might be E3, E4, and 1% to 2% would

00:51:57.820 be E4, E4. As you pointed out, E2, E2 is the most rare phenotype by far. That's significantly less

00:52:05.680 than half a percent. I think E2, E3 is probably on the order of two to three percent. E2, E4 is also

00:52:13.740 quite rare. So the two most common by far are E3, E3, and E3, E4. And as we've talked about many

00:52:22.860 times on the podcast, the risk associated with Alzheimer's disease between 3-3, which is always

00:52:28.660 the reference case, and 3-4 and 4-4, those go up non-linearly. So the 3-4 individuals,

00:52:36.900 the people that have one copy of 3, one copy of 4, they make a version of ApoE, the protein,

00:52:41.520 that's not as good as the wild type. And their risk of Alzheimer's disease is about two times

00:52:46.660 higher than someone who has a 3-3. And again, it depends on the series. Sometimes you'll see that

00:52:52.040 at three times higher, but directionally, that's about the level. Conversely, if you have two

00:52:56.000 copies of the four, that risk is significantly higher. There was a day, Tom, 15 years ago,

00:53:01.120 the literature was calling that 20 to 25 times higher. That number has come down considerably.

00:53:06.300 And I think most series would talk about that as being an eight to 12 fold increase. So it's,

00:53:11.940 you know, it's a full log increase in risk for sure to have two copies of the E4 gene,

00:53:18.580 which means you're making an APOE protein that is far less effective.

00:53:22.540 Yes. And this is going to have ramifications. We've done podcasts and Peter's had Dan Rader on here. The most important thing about the peripheral HDLs is not the amount of cholesterol they traffic. It's kind of trivial and it gets transferred here and there. And it almost tells us nothing. If you're measuring HDL cholesterol, it tells us nothing about what the HDL particles. Remember, they're 90% of your lipoproteins out there.

00:53:48.900 So clearly, what they're doing to cholesterol is not their major function.

00:53:54.820 So that means HDLs do other things.

00:53:57.140 And as we're learning, they do innumerable other things that regulate all aspects of

00:54:02.520 human health.

00:54:03.520 They're actually a part of the innate immune system.

00:54:06.380 So they're involved with fighting inflammatory diseases, infectious diseases, chronic diseases,

00:54:13.400 cancer.

00:54:13.900 So what we wish we had is not HDL cholesterol, which tells us very little. We wish we had tests

00:54:22.680 that would tell us, are the HDLs in a given patient's body doing what they're supposed to

00:54:27.980 be doing? Are they functional or not? But there's so many different functions that HDLs perform.

00:54:35.400 It has to do not with the cholesterol they're carrying, but yet the types of proteins they

00:54:40.640 might be carrying. Well over 200 proteins have been described in the periphery as being found

00:54:46.800 on HDL particles. Now, that doesn't mean there's an HDL particle carrying 100 peptides on it.

00:54:52.980 Impossible. They're too small. But each HDL might carry one or two peptides. And each of those

00:54:59.160 peptides might have some function that it's hard to even know what they are. Are they helping the

00:55:04.080 immune system or are they involved with coagulation or what? So we have actually

00:55:11.220 armies of HDLs, each constructed with one or two of those peptides in addition to APOA1

00:55:19.680 and APOA2, some of the lipolipid-related apoproteins. And so there's no way to know

00:55:27.680 for us to measure these HDL subpopulations. Now, all of the HDLs that are carrying these proteins,

00:55:35.640 they're not carrying cholesterol. So they are the really small HDL particles. They have the

00:55:41.340 highest density because really what determines the density of a particle in the centrifuge is

00:55:46.560 its lipid content. The more lipids, the more buoyant they flow. Now, HDLs carries the least

00:55:53.040 amount of lipids compared to the ApoB particles. That's why it sinks in the centrifuge tube.

00:55:57.680 But the tiniest HDLs, the discoidal HDLs, APOA1 by itself, they're sitting right at the bottom

00:56:05.540 because there's zero buoyancy to them. So if we have this whole army of very tiny,

00:56:13.660 high-density HDL particles that are packing probably critical proteins,

00:56:19.680 geez, don't you wish we could measure them? But here's where it gets interesting.

00:56:23.960 We've hinted earlier in this podcast that there is a lipoprotein that can traverse that

00:56:29.100 blood-brain barrier and get into the brain.

00:56:31.320 And it's these extremely small HDL particles, either free ApoA1 or an ApoA1 that's bound

00:56:40.060 to a couple of these other proteins.

00:56:42.000 And maybe some of these proteins are very important, antioxidative proteins, anti-inflammatory

00:56:47.840 proteins. So if those tiny HDLs that we cannot measure jump into the blood brain barrier or

00:56:55.940 through it, and they're now in the matrosome, where do they go? They immediately bind to the 0.99

00:57:02.080 first ApoE containing brain HDL that they bump into. So all of a sudden this astrocyte ApoE

00:57:09.440 constructed brain HDL particle is also carrying a copy or two of APOA1 that actually originated

00:57:18.520 from the plasma. The brains can't synthesize APOA1, the brain cells. So if it's in the brain,

00:57:24.540 and we know it is, they do pass the blood-brain barrier. It is believed that is receptor-mediated

00:57:30.720 and might be this good old scavenger receptor, again, expressed at the blood-brain barrier that

00:57:35.620 facilitates entry of APOA1 or the really small, dense HDL, APOA1s carrying accessory proteins.

00:57:44.460 And the hope is, hey, number one, if they get in, great. So now the brain HDLs, you have different

00:57:50.360 subpopulation of brain HDLs. You might have only APOE containing brain HDLs. You might even have

00:57:56.740 some APOA1 brain HDLs, but most of them are going to be APOE plus APOA1 brain HDLs. And those

00:58:05.520 other proteins that came with the APOA1 maybe can do, start doing some good things in the brain.

00:58:10.580 And where this might be really good. So in the periphery, we have brain functionality.

00:58:17.220 I did not introduce it, but it's easily, you can deduce that. Wait a minute. If there are

00:58:24.020 functional HDLs in the periphery, I'll bet there are circumstances where there are dysfunctional

00:58:28.760 HDLs in the periphery that are not equipped with the proper protein or they're carrying proteins

00:58:34.000 they shouldn't be carrying proteins that can do harmful things to tissues. They would be

00:58:39.340 dysfunctional HDLs. Don't I wish we had a blood test for that? And we do not. So in the brain,

00:58:47.120 now you have these ApoE particles, maybe can carrying ApoA1. But now if you're an ApoE4

00:58:54.880 producer, when your astrocyte produces ApoE, it's going to be an ApoE4 type of... And that

00:59:02.760 Just like in the periphery, it's a dysfunctional type of ApoE. So if you have the ApoE4 genotype

00:59:10.520 and your astrocyte is producing ApoE4 peptides and they're what's constructed on the HDL,

00:59:18.340 that's likely to be a dysfunctional HDL in the brain. And what would that mean? It means that

00:59:24.180 it doesn't bind to the neuron receptors as well as an E3 or an E2 might to those receptors.

00:59:30.420 and therefore you have disrupted cholesterol transport into the neuron. Now, all of a sudden,

00:59:37.160 the neuron is not getting the cholesterol it needs, and that will create havoc because the neuron

00:59:43.100 puts it right in the cell membranes. If you don't have the proper amount of cell membrane cholesterol,

00:59:50.020 this is where something called amyloid precursor protein sits. And if you don't have the right

00:59:56.180 cholesterol balance, it's acted upon by a certain enzyme called secretases. That's where you start

01:00:02.740 producing beta amyloid and even tau because you don't have the right amount of cholesterol in

01:00:09.240 your neuron cell membranes. And this is how E4, one of the many reasons why it's associated with

01:00:16.540 Alzheimer's. I'll stop there perhaps for you to jump in before I maybe describe some of the other

01:00:22.580 things that APOE4 brain HDLs don't do that an APOE3 or an APOE2 HDL would. Well, I actually want

01:00:30.680 to take us backwards for a second, Tom, because one thing that we've danced around, but I don't

01:00:35.960 think we've explicitly addressed is what is the relationship between brain cholesterol movement

01:00:41.800 and something that people are very familiar with if they've listened to this podcast, which is

01:00:47.080 amyloid. So people are obviously familiar with the accumulation of beta amyloid and P-tau in the

01:00:52.660 brain, and people are now really starting to understand that we actually have great biomarkers

01:00:57.200 where we can start to track those things. Is there any relationship between those? In other words,

01:01:01.540 as you talk about all of this dysfunctional movement of cholesterol in the brain, and we

01:01:07.180 know that that is highly associated with your ApoE genotype, and we also know that your ApoE

01:01:13.580 genotype is highly associated with Alzheimer's disease. So the one thing we haven't put together

01:01:16.800 is what's the relationship between amyloid tau and cholesterol? There must be a link, right?

01:01:21.540 Definitely. We go way back. You can read the studies of autopsies on patients with Alzheimer's

01:01:28.420 disease and they're really cholesterol overloaded tissues, especially the neurons. So remember the

01:01:34.120 neuron, the main thing determined in its function is its cell membrane integrity. And if you have

01:01:40.640 the proper cell main construction, signaling occurs properly, the synapses fire properly or so.

01:01:47.560 Now, what will happen if you have too much cholesterol in that cell membrane, and this is

01:01:54.680 what happens in the Alzheimer's patients, what I just alluded to a few seconds ago,

01:02:01.800 also located in the cell membrane is amyloid precursor protein. That's a protein that is

01:02:07.300 going to evolve into the production of beta amyloid. And whether it produces, there's two

01:02:15.560 types of that amyloid, 40 and 42, with 42 being the more injurious type of amyloid beta and the

01:02:23.560 40 being a less toxic type of amyloid beta. So when there's too much cholesterol in the cell

01:02:30.300 membrane of a neuron, there's something called beta and gamma secretase. They're enzymes that

01:02:36.920 make the amyloid precursor protein cleave into the production of amyloid 42. If there is the

01:02:44.600 proper amount of cholesterol in the neuron cell membrane, it's a secretase alpha secretase that

01:02:53.180 sort of slows the cleavage of amyloid precursor protein into apobine. You wind up producing more

01:03:00.600 the beta amyloid 40, which is the less toxic form. So obviously it's the cholesterol content

01:03:07.780 in your cell membrane that is a major, major factor. There's one other aspect of cholesterol

01:03:15.080 homeostasis that we might as well introduce now because too much cholesterol in the cell membranes

01:03:21.700 is a danger to the neuron because the membrane isn't going to function. The neuron is the one

01:03:29.460 cell in the brain that has the ability to get rid of cholesterol. We've spent a lot of time saying

01:03:36.520 the brain makes cholesterol and it retains it. In fact, the half-life of cholesterol in the brain

01:03:41.080 is five years as opposed to a few days in the periphery. So that tells you the brain is

01:03:46.540 reserving cholesterol. But early, early, I told you too much cholesterol in any cell is toxic

01:03:52.300 in the neurons. Not only will it disrupt membrane function, but it crystallizes in the cytosol of

01:03:58.280 the neuron and it kills neurons. You don't want to kill neurons. You're going to have some sort of

01:04:03.780 chronic brain disease if that happens over time. So evolution has given the neurons the ability

01:04:10.220 to change cholesterol into something called an oxysterol. And the one it produces is called

01:04:17.660 24S hydroxycholesterol. People who know what cholesterol looks like biochemistry wise,

01:04:25.420 it has one oxygen molecule at the third position of the first ring. 24S hydroxycholesterol not only

01:04:34.420 has that one cholesterol molecule, it has a second one at carbon 24. So now you have a hydroxy group

01:04:41.420 on both ends of the cholesterol molecule. That makes it a little bit more water soluble. So when

01:04:47.780 the neuron says, I got to get rid of cholesterol, it has an enzyme, 24S hydroxycholesterolase, that

01:04:55.100 will make cholesterol change into 24S hydroxycholesterol, which is water soluble. It

01:05:01.660 comes out of the neuron. It floats right through the matrosome to the blood brain barrier where 0.98

01:05:07.420 it can pass right through it because it's sort of a hydrophilic lipid with an oxygen hydroxy group

01:05:13.520 on each end. When it hits the blood brain barrier, the fatty acids and the phospholipids hate the

01:05:20.240 hydroxy groups. So they separate and it just creates a little tunnel through which the 24S

01:05:26.200 hydroxycholesterol can jump into plasma. Now, wait a minute, it's a lipid. It can't jump into free

01:05:31.360 plasma, but what's floating in the plasma that rapidly binds to the excreted 24S hydroxycholesterol,

01:05:39.720 either albumin or any brain lipoprotein that floats by. Now it's part of a protein. It's

01:05:45.940 it on allumin or it's on a lipoprotein, they bring it back to the liver. Now, here's the cool thing.

01:05:52.280 What's the only other tissue in the body beside the brain that can produce an oxysterol? It's the

01:05:57.520 liver. And what does the liver do with oxysterols? Well, the liver has cholesterol. You know, Pete,

01:06:05.840 that the liver is our major, our only manufacturer of bile acids, which are oxysterols. So cholesterol

01:06:13.220 gets transformed into an oxysterol in the liver, same enzyme that the neurons express,

01:06:19.400 and the oxysterols, they go through several steps, but they become your bile acids down

01:06:23.840 to your gut, goodbye fecally. So the brain actually has this cool way of getting rid

01:06:29.340 of excess cholesterol by that transformation and send it to the liver where it could be

01:06:34.620 fecal excluded or so. So this 24S hydroxycholesterol gets very important. But if, again, you start to

01:06:45.380 build up too much cholesterol in your neuron cell membrane, it's in the cell membrane now. So there's

01:06:53.760 less cholesterol in the cytosol of the neuron. The neuron stops making 24S hydroxycholesterol.

01:07:01.020 brain is not escaping into the plasma anymore this is why researchers use 24s hydroxycholesterol

01:07:09.400 in the plasma as a biomarker of brain health it shouldn't be there because the brain is

01:07:15.460 retaining all its cholesterol the neurons not trying to excrete any cholesterol but if it does

01:07:21.500 the concentration of that in the plasma goes up the liver doesn't secrete its oxysterols into the

01:07:28.400 plasma, but the brain does. So it's a great biomarker on brain health. So too much tells

01:07:34.360 you the brain is in danger. This is why people who are developing drugs for the brain to try

01:07:40.260 and prevent dementia, they monitor 24S hydroxycholesterol because they think if their

01:07:45.480 drug is helping the brain prevent Alzheimer's disease, you won't find 24S hydroxycholesterol

01:07:52.120 in the bloodstream. And that's one of the sterol biomarkers, as is the desmostrol that we alluded.

01:07:59.300 So we actually have two things that we can measure. Here's the bad thing. In the real world,

01:08:04.820 we can get desmostrol measurements fairly easily. There's no commercial laboratory that 24S hydroxy

01:08:11.720 cholesterol has become available outside of research studies. I wish we could measure that

01:08:17.500 in our patients because it would just be another of the many biomarkers that are starting to emerge

01:08:22.580 on brain health. So finally, back up, it's disruptive, this ApoE4 that is going to the

01:08:32.460 receptors that should be internalizing the ApoE HDL in the brain into the lysosomes in the neuron,

01:08:40.320 which will generate cholesterol for the neuron to use. But since there is markedly decreased

01:08:46.800 clearance of the E4 brain HDL, just when it touches the membrane, the cholesterol can jump

01:08:54.380 into the cell membrane of the neuron, but it doesn't get to the cytosol. So it's very complex,

01:09:00.940 these lipid mechanics that are going on in the E4 patient. And I'll let you ask about that before

01:09:07.740 we get into other attributes of what APOE4 might not be doing well in the brain.

01:09:13.140 Well, I kind of want to ask a question that brings it even further to something clinical,

01:09:18.080 which is we've come this far in the discussion without really talking about the impact of

01:09:23.040 pharmacology. So I want to sort of make that bridge now. Obviously, we're not going to get

01:09:29.060 into all the reasons why one would lower ApoB pharmacologically. It's implied in so much of

01:09:35.160 what we already discussed in the periphery. And when you talk about that, the thing that

01:09:38.980 comes to most people's minds. I mean, most people aren't thinking of bempedoic acid and azetamide

01:09:43.500 and PCSK9 inhibitors or bile acid sequesterants or CPET inhibitors or CTEP inhibitors. When you say

01:09:50.640 lipid lowering therapy, everybody defaults into one class of drug and that class of drug is called

01:09:55.560 the statin. So let's talk for a moment about what statins do, if anything, in the brain. And I'll

01:10:03.920 bracket the discussion by saying maybe we can formulate it through the lens of the two types

01:10:09.780 of statins, those that tend to be more hydrophobic and those that tend to be more hydrophilic.

01:10:15.540 So maybe talk a little bit about that class of drugs. I don't think we have the time to go into

01:10:20.080 the entire history of them. So we can even do it through the lens of the modern versions of those

01:10:24.780 drugs as opposed to going back in time. But talk a little bit about how those drugs work

01:10:29.660 in the brain specifically. And of course, statins are the number one drug to lower APO-B in the

01:10:35.300 periphery because that, no doubt about it, reduces atheroscronic heart disease. But of all, and

01:10:40.540 people rattled off the classes of APO-B lowering drugs that are primarily used nowadays. Of all

01:10:47.460 of those, there's only one that can penetrate the blood-brain barrier and get into the brain,

01:10:51.620 and it is the statin class. All of those other drugs mentioned either work solely in the liver

01:10:57.040 or no way they could penetrate the blood-brain barrier. It didn't do anything to brain cholesterol

01:11:02.440 homeostasis. So if a statin gets into the brain, now a little bit, Peter mentioned what we call

01:11:08.940 hydrophilic, hydrophobic statins, lipophilic, lipophobic. A hydrophilic loves water. Lipophilic

01:11:17.800 loves lipids, hates water. And early on, there was lots of data showing just traversing a cell

01:11:25.300 membrane border, the lipophilic statins get through easier. The border itself has got a lot

01:11:32.840 of lipids in, so they, all right, come right in. So it's a little, for the hydrophilic statins

01:11:39.600 to penetrate a barrier, pretty much there are receptors that pull them into, even the liver,

01:11:46.900 the hydrophilic statins, there are receptors that pull them into the liver and they get in quickly.

01:11:54.480 So tactically, the lipophilic statin should get into the brain a little easier than the

01:12:00.720 hydrophilic statins.

01:12:01.840 But more modern studies have shown that really doesn't matter as much because once you're

01:12:07.280 in a steady state, meaning you're on a statin, you have your blood level of the statin, ultimately

01:12:13.220 they're all in the brain.

01:12:15.120 Yes, the lipophilic ones may get in a little easier, but the hydrophilic ones get in also.

01:12:20.580 And they all have the ability, therefore, to various degrees, inhibit cholesterol synthesis

01:12:27.020 in the brain. So I don't think you necessarily have to pick a statin based on its lipophilicity

01:12:33.180 or hydrophilicity worrying about the brain. I think because in real world practice,

01:12:40.160 resuvastatin, a hydrophilic statin is used more commonly. It certainly can get into the brain,

01:12:47.520 maybe a little less slowly than lipitor, lipophilic statin can, but if you're in a steady

01:12:52.480 state, they're all in, they all have the ability to reduce cholesterol synthesis in the brain.

01:12:58.080 So is that, is that size driven, Tom? Is it just that the size of a statin is such that it can get

01:13:02.440 across the blood brain barrier, whereas the other classes can't?

01:13:05.300 It's just the construction of the, uh, the statin drug on how, you know, what converts a hydrophilic

01:13:13.620 or a lipophilic property to that given statin.

01:13:16.880 You know, if you look at, we put up a slide here showing all the different statins, they're

01:13:20.640 all a little bit different.

01:13:21.700 And there are certain aspects of that construction that gives them hydrophilicity and other aspects

01:13:28.000 of that alignment or construction of their molecules gives them the lipophilicity or

01:13:33.860 lipophobicity.

01:13:35.860 So anyway, since we earlier, we just said, hey, Alzheimer's disease is too much cholesterol

01:13:41.580 in the brain, too much cholesterol in the neurons, you could hypothesize that if statins did get into

01:13:47.920 the brain, which they do, and all of them do, and in a steady state, they all have the potential to

01:13:53.800 affect cholesterol synthesis in the brain, it might actually be good in a lot of people to slow down

01:13:59.620 a little bit of the cholesterol synthesis in the brain because too much cholesterol results in

01:14:06.020 pathology of the neurons and tissues. And this is why, and we're not going to review them here,

01:14:12.460 if you look at all the statin trials, the meta-analyses, most of them show statins

01:14:17.180 really have no harm to the brain, but there are a few that do show statins seem to reduce the

01:14:23.540 incidence of Alzheimer's disease or cognitive impairment in the brain. None have shown that

01:14:29.180 statins injured the brain. Yeah. Well, just for the listeners, we'll link to that in the show

01:14:33.580 notes, we did an AMA on this a few years ago where I went through all of the meta-analyses.

01:14:40.060 And yeah, the TLDR is that every study we looked at for either MCI or Alzheimer's disease or

01:14:48.300 dementia, otherwise not specified, showed either neutrality or improvement. And these are all RCTs,

01:14:56.900 of course, though these are not studies that used dementia as a primary outcome. These are

01:15:02.780 studies that are using dementia as a secondary outcome. And I always find this to be interesting,

01:15:08.320 Tom, because it's both intuitive and counterintuitive, right? It's intuitive in the

01:15:13.780 sense that you just laid it out, which is, look, if cholesterol accumulation is highly toxic to

01:15:19.820 the neurons, then a drug that reduces cholesterol synthesis in the brain should be beneficial.

01:15:25.680 But at the same time, cholesterol is essential to the brain. So if we overcook it and we reduce

01:15:30.760 cholesterol synthesis too much in the brain, could that also be problematic? Yes, and this is more in

01:15:36.880 the hypothetical range right now because nobody's going to do these studies to prove it one way or

01:15:41.760 the other. But because, as Peter just says, cholesterol is so important, you would never

01:15:47.620 want to over-suppress cholesterol synthesis in the brain. That would not be good. So can that happen?

01:15:54.280 And I think intuitively we know anybody who's prescribed a bunch of statins to people have known a few of them get brain fog.

01:16:04.160 Hey, I'm on the stat and I'm not thinking right.

01:16:07.540 Nothing, you know, my addition isn't as good as it used to be.

01:16:10.680 And we stopped the statin and rather quickly that goes away.

01:16:14.600 So one hypothesis would be that is the person that the statin is over suppressing cholesterol synthesis rather rapidly.

01:16:23.160 and severely, and that's why they got neurologic symptoms. And they stop it. Obviously, you've

01:16:30.080 stopped the statin. You're restoring whatever synthesis was going on in the brain or so.

01:16:35.920 So that becomes a plausible hypothesis. And I said, nobody's ever going to do a study to prove

01:16:41.940 that or disprove it. But you could also say Alzheimer's disease takes decades to develop.

01:16:48.560 So again, if I give you a statin and you don't get that acute brain fog, it's probably safe

01:16:56.600 to over-suppress cholesterol a little bit over time, and maybe especially so if you're

01:17:03.040 an E4 or you have a family history putting you at risk for Alzheimer's disease.

01:17:07.800 Again, a theory, but it would be supported by the trials you just said that tend to show

01:17:13.240 you're not much going on or benefit, and that could be the plausible reason.

01:17:18.560 Now we go back to, you've went through the desmostrol and the thosthrol pathways.

01:17:25.960 There's a nice study published almost a decade ago where they were doing cerebral spinal fluid

01:17:32.200 desmostrol levels and plasma desmostrol levels and measuring it by mass spec.

01:17:38.180 And there was high correlation between the CSF desmostrol and the plasma desmostrol saying

01:17:44.680 that, wow, desmostrol in the plasma, it reflects desmostrol in the central nervous system.

01:17:51.580 And even more interesting, that study showed that the people with low desmostrol have the

01:17:58.800 higher incidence of cognitive impairment in Alzheimer's disease.

01:18:02.880 So if, and you've talked about this many times on the podcast too, if we are administering

01:18:09.100 statins to our patients, even the E4 patients, I'm a hope that we are going to help the

01:18:14.280 lessen their incidence of Alzheimer's disease. Maybe keeping an eye on plasma desmostrol

01:18:19.900 sort of makes sense. And if you do over-suppress it with your statin therapy, maybe you can change

01:18:26.480 the dose of that statin therapy, or maybe you can just abandon statin therapy and lower ApoB to

01:18:31.500 reduce heart attacks with the several other drugs that you ran through very quickly there. So it

01:18:37.080 gets very interesting. And last thing to tie it into that 24-S-hydroxycholesterol. Remember,

01:18:42.940 if you're on the way to Alzheimer's disease, that's increased in the plasma. There are studies

01:18:47.460 showing that if you prescribe a statin, the 24-S-hydroxycholesterol disappears. That would

01:18:53.900 again be proof that the statins are lessening cholesterol synthesis in the brain and maybe to

01:19:00.560 a level that's really desirable because you don't want to see that. But then you would back it up

01:19:05.600 with the desmosterol because if that was low, ooh, I've maybe suppressed it a little bit too much.

01:19:10.620 All wonderful hypotheses. It has a lot of data. I can easily provide 20 references on

01:19:17.340 desmostrol in the brain, how critical it is. So this is a very plausible theory right now,

01:19:23.880 and don't expect a clinical trial to prove or disprove this hypothesis right now.

01:19:28.880 We'll link to those sources, Tom, in the show notes. One other drug I just want to talk about

01:19:33.260 really quickly is ezetimibe. Again, ezetimibe is a drug that really works outside the body,

01:19:40.200 so to speak, right? It works in the gut. It blocks the Neiman-Pixi one like one transporter.

01:19:45.720 And in people who are not hyper absorbers, it's not even a particularly effective drug.

01:19:50.840 Yet, there is a kind of a suggestion that it might have some benefits in the brain,

01:19:55.660 which is the furthest place from where we think of it working. What can you say about that?

01:20:00.380 You know, it's kind of amazing. Like you said, who would ever even hypothesize that,

01:20:04.740 that this drug that acts in the intestine might have beneficial effects in the brain or so.

01:20:10.820 And I think, you know, we have a couple of neurologic colleagues, Richard Isaacson and

01:20:15.260 Kellyanne Otis, who are very involved with these diseases. And it is their anecdotal belief that

01:20:21.060 azetimibe, in addition to helping them control their APLB in their patients, there seems to be

01:20:27.580 some cognitive benefit in the people they deal with or so. So now there's some actual plausible

01:20:33.880 reason. Now, zetimibe is one of those drugs that just cannot cross the blood-brain barrier. So

01:20:39.500 how in the world could it be helping dementia or so? But it has a metabolite called the zetimibe

01:20:45.180 glucuronide that actually can pass through the blood-brain barrier in small amounts,

01:20:51.120 but unlike a zetimibe, it gets in. And there are animal studies showing that it interferes with

01:20:57.780 hexokinase and the glycosylation of brain proteins. If you reduce that, there's some benefit,

01:21:05.000 less inflammation in the brain or so. So there is that. And again, it's a study I will definitely

01:21:10.880 give you the reference to that people can read that there's some plausibility to it. And there's

01:21:16.360 an anecdotal belief among neurologists who live in this field that it's a helper. So wouldn't that

01:21:22.760 be cool. You know, Peter, as we control ApoB aggressively in your patients, we use a lot of

01:21:30.500 ezetimibe because we prefer to use low-dose statins. And if we don't get to the ApoB goal,

01:21:37.380 we're adding ezetimibe. We also, day one, check synthesis and absorption. So there are patients

01:21:43.180 where we use ezetimibe day one because they're hyperabsorbers, and that's where you get the most

01:21:48.240 efficacious APO-B lowering. So in the future, as we have people who carry the APO-E4 alleles

01:21:56.000 and they have APO-B issues, we might pick a statin first, we might pick a zetimibe,

01:22:02.160 but I think they might be a patient where you need a little bit of a statin and a little bit

01:22:05.900 of a zetimibe until somebody proves this. And I would not hold your breath waiting for a randomized

01:22:11.140 control trial that is zetimibe, what it's doing to even some of the Alzheimer's biomarkers in the

01:22:17.840 blood, because only emerging drugs are they starting to do those type of studies on.

01:22:23.120 Nobody's going to go back and look what ezetimibe does, the P-TOW or the amyloid ratios or so.

01:22:28.460 I wish somebody should.

01:22:29.740 You'd only need maybe a small study.

01:22:31.580 Well, I'm surprised.

01:22:32.720 You could probably pull it out of a biobank for an existing study that was already done

01:22:36.940 on ezetimibe.

01:22:37.820 So we could at least get the suggestion of that from such a study, because we do have

01:22:42.720 at least one.

01:22:44.280 Well, I know we have statin versus statin plus ezetimibe trials.

01:22:47.440 Don't we also have a monotherapy SETIA trial? 0.97

01:22:50.500 Only in Japanese elderly people. 0.72

01:22:53.460 And I don't think cognition was one of the things.

01:22:55.760 It was just, and it was not a blinded trial.

01:22:59.040 So it was an open-label trial.

01:23:00.560 Open-label trial, yeah.

01:23:00.820 Just to show it was efficacious in lowering ApoB in a primary prevention setting.

01:23:05.440 But they certainly didn't look at cognition or biomarkers and that stuff.

01:23:09.280 But if they still have serum banked, you could at least look at pre and post levels of P-Tau.

01:23:15.340 You definitely could.

01:23:16.320 Yeah, AB4240.

01:23:18.740 Be a great research project for some young PhD or a budding lipidologist.

01:23:24.580 Hopefully listening right now.

01:23:26.600 Let's talk about something else that is half drug, half supplement that gets talked about

01:23:30.800 a lot for brain health, which is the role of EPA and DHA.

01:23:34.240 Again, they're readily available as supplements over the counter.

01:23:37.660 And there are certainly some brands out there that are legitimate, which is to say you're

01:23:41.700 getting what the label says you're getting and they're free of contaminants.

01:23:44.780 but they also make pharmacologic variants of both of these fatty acids. So take that in whichever

01:23:50.620 way you'd like, but what do we know about EPA and DHA and brain health?

01:23:54.480 Without talking about specific products, let's talk about if an EPA and DHA are both important

01:24:01.720 to the brain. There's far more DHA, but we're finding out that even EPA is important for the

01:24:06.920 brain now also. So since we can't produce omega-3 fatty acids, we have to eat them.

01:24:13.580 And when you eat them, they're mostly in the form of a triglyceride carrier or a phospholipid carrier.

01:24:20.220 And that's exactly how the supplements deliver omega-3s to us.

01:24:23.660 They're packaged in it as a triglyceride.

01:24:25.800 Typically, one of the fatty acids on a synthesized triglyceride would be an omega-3.

01:24:31.280 And there is a product that delivers omega-3s as a phospholipid from krill oil.

01:24:37.880 So now once you ingest a triglyceride or a phospholipid,

01:24:42.040 Remember, the only thing that can really be absorbed is free fatty acids. So pancreatic

01:24:47.320 enzymes, lipases, leave off the fatty acids from the triglyceride or phospholipid vehicle,

01:24:53.760 and then the free EPA or free DHA, which joins with other lipids and the biliary micelles,

01:25:00.460 gets absorbed by a fatty acid absorber, CD36. Interestingly, not only can a free fatty acid

01:25:08.740 be absorbed, but a lysophospholipid can be absorbed. A phospholipid has a phosphorus

01:25:16.220 moiety in a head group and two fatty acids attached to it. That's called the dyradal

01:25:21.700 phospholipid, two fatty acids. But if I took one fatty acid off of a phospholipid,

01:25:28.360 it's called a lysophospholipid. It's actually a smaller molecule and they're easily absorbed.

01:25:34.260 So the lipases either makes free fatty acids or it could make lysophospholipids.

01:25:40.000 But hey, if the remaining fatty acid on that lysophospholipid is an omega-3, it gets in.

01:25:46.200 So once they're in the enterocyte, what happens to them?

01:25:49.340 The enterocyte immediately resynthesizes them to a full phospholipid or attaches them to

01:25:55.580 a triglyceride molecule, which goes in the core of the chylomicron.

01:25:59.100 The phospholipid goes on the surface of the chylomicron.

01:26:02.080 It shoots them into the lymphatics, and they rapidly get into the plasma.

01:26:06.940 They undergo rapid hydrolysis at the muscles and fat cells by lipases, and that frees up

01:26:14.060 these phospholipids.

01:26:16.280 Aha.

01:26:17.140 Now, phospholipids are a lipid.

01:26:18.720 They can't circulate in the bloodstream.

01:26:20.700 They immediately bind to something called a phospholipid transfer protein, and the phospholipid

01:26:26.060 transfer protein will bind to either a full phospholipid or a lysophospholipid, and it

01:26:32.060 that little delivery truck of an omega-3 phospholipid transfer protein, which goes up,

01:26:39.200 butts into the blood-brain barrier, and there's a specific receptor in the blood-brain barrier

01:26:44.420 that will internalize the lysophospholipid form of DHA or EPA. And once it gets into the brain,

01:26:53.420 it's in the brain. It can be trafficked in these brain HDL particles, and it's part of the things

01:26:58.640 they do too, or it can jump right into the cytosol of the first cell that it bumps into

01:27:03.480 while it's in the matrosome. So that's the journey. So look, it almost doesn't matter

01:27:09.200 the vehicle you're ingesting an omega-3 with. We would prefer that you've established that

01:27:15.240 a supplement is actually has the amount of omega-3s they say they do. Don't trust the

01:27:21.580 labels of every supplement you may buy, and they get in. Now, in the periphery, there's a lot on,

01:27:29.240 and there's a big trial it shows, perhaps for preventing or lessening residual risk in people

01:27:34.760 who have ApoB controlled, the EPA is a little bit more important. And the DHA plus the EPA,

01:27:42.680 there's a trial where that didn't work as well as the EPA. But once they get into the brain,

01:27:48.300 the brain has its omega-3 fatty acids. And it used to be all DHA, but I know the thought on

01:27:55.960 that is changing. EPA is required there too. Some people can convert EPA to DHA, but not everybody

01:28:03.040 can. And that's how they get up into the brain. And obviously, since their concentrations in the

01:28:09.740 brain are so high compared to other tissues, it's an integral part. And why wouldn't it be? Because

01:28:14.740 where do omega-3s go in the cell membranes? And that's everything, cell membrane health in the

01:28:20.600 brain cells. And I guess, where do you stack this in terms of evidence, right? Like in the hierarchy

01:28:28.920 of things that we really know are as close to capital T true as possible when it comes to brain

01:28:36.240 health, right? Which is lipid homeostasis, good blood pressure, sleep, exercise. I mean,

01:28:43.920 things that just demonstrably matter when it comes to brain health, where in that pantheon would you

01:28:50.740 sort of put having a serum or EPA DHA level in the RBC membrane of 10% versus 6%? What's your

01:29:00.840 level of confidence? Well, the data is all going to come from observational trials for the most

01:29:06.260 part. And in those trials, there are ones that specifically looked at certain brain functions

01:29:13.000 and correlated omega-3 index with the observational outcomes related to neurological issues,

01:29:20.980 and they seem to be positive. An academician would tell you, Tom, don't even talk to me.

01:29:25.700 There's no level one randomized blinded controlled study doing what you say, so it's irrelevant to

01:29:31.380 me. But if you look at all the observational data, just like we did with the statins where

01:29:36.040 it's in general pretty good, I think if you went through all of that data with omega-3s in the

01:29:41.060 brain, you would find it. Bill Harris has studies relating it to brain size or at least certain

01:29:46.260 sections of the brain size in omega-3 context, I believe in the hypothalamus or other areas of the

01:29:53.200 brain. So again, it's this plausible stuff, but there's no level one evidence. We certainly have

01:29:59.340 evidence in the blood that low levels of omega-3 index are certainly associated with sudden death

01:30:05.300 and increase atherosclerotic heart disease. Again, we lack the randomized controlled trials

01:30:10.940 that changing that will reduce events other than that one trial of EPA and insulin-resistant

01:30:17.260 high-risk people who had APL-B well-controlled. So you're in that gray zone area with the clinical

01:30:24.100 trials on this, and they're not the type of trials that any guideline is going to tell you this is

01:30:28.820 what you have to do. But again, it's just like our desmostral hypothesis. There's a plausibility

01:30:34.540 there. I see little downside to using omega-3s. Bill Harris, who you've interviewed, has looked

01:30:41.960 at his trials, and he's pretty content that when you hit the 8 to 9% omega-3 index, you pretty much

01:30:49.400 have the proper amount of omega-3s in the scenario cell membranes of your body. There is no study

01:30:57.340 you can allude to that, hey, therefore, so it's so important in the brain, let's make it 10%

01:31:01.260 percent other than if there's no harm to it, why not try for it? So you're going by that type.

01:31:08.640 That's all a little bit of guesswork right there. Yeah. Well, Tom, I want to close with something

01:31:14.280 that's you and I are very excited about. I did a brief podcast on it a little while ago, which is

01:31:19.980 a new drug in a new class. I alluded to this class very briefly a few moments ago, the CTEP

01:31:27.100 inhibitors. You and I have spoken about these drugs in the past on a podcast. We've got several

01:31:31.500 podcasts on this topic, including most recently one with John Kasterlin, probably, God, about

01:31:37.540 four years ago. But since that time, we've had some exciting data, which I talked about in my

01:31:42.980 podcast. But maybe just we could remind people about that drug, Obacetrapib, and the Broadway

01:31:48.080 trial specifically, and how we tie it into what we've talked about today.

01:31:53.020 Yes. Basically, CETP inhibitors, of which Obisetrapib is the latest, have been investigated

01:32:01.700 to see what they do to atherosclerotic heart disease and LP little a, and maybe brain

01:32:08.880 functioners. There's a signal that if you have CETP loss of function genetically, those people

01:32:15.440 have less Alzheimer's disease or cognitive impairment. So that makes it plausible. Well,

01:32:20.600 if we inhibited CETP pharmacologically, we almost convert you into the genetic status,

01:32:26.440 maybe there would be less Alzheimer's disease. And now in that Broadway trial, look, the people

01:32:32.620 at New Amsterdam Pharma recognize this. So they're actually putting a little money into clinical

01:32:37.600 trials, perhaps investigating this hypothesis. And in that Broadway trial where you administered

01:32:43.740 Obisetrapin, remember it was given to them primarily to reduce ApoB and ultimately reduce

01:32:49.320 mace in those people, but they actually looked at some of the biomarkers of Alzheimer's disease,

01:32:56.540 the phosphorylated P-tau, the amyloid 40-42 ratio, the other various ratio of these markers,

01:33:04.080 the fibrillatory markers, all things you can measure. And they saw some very interesting

01:33:09.960 movement in the right direction of those Alzheimer's associated biomarkers.

01:33:16.260 Now, the next step would have to be in a clinical trial, continue to monitor them.

01:33:20.120 And that was a very quick study.

01:33:22.680 You would monitor it over time, but maybe you'd throw in some cognitive function in

01:33:27.380 some of the studies to see, geez, could Obisetropib actually, because it's improving these biomarkers,

01:33:34.280 really affect what we want to do, better brain function.

01:33:37.300 And the plausibility is because they make your HDLs very big and have many copies of APOA1 on them, which can break off. So you generate some APOA1 in the plasma. But when the HDLs are big on a CETP inhibitor, the liver senses, oh, we have a deficiency of APOA1 because they're not seeing it. It's all on the HDL particles.

01:34:04.100 So the liver actually starts overproducing APOA1. So APOA1 goes up in the plasma. But once APOA1

01:34:14.540 goes up, what does it start doing? It starts binding to some of these potentially protective

01:34:18.760 proteins we've talked about. And guess what? So if you're increasing, if OB-Cetrapid is increasing

01:34:27.340 either ApoA1 or the really tiny protein-laden HDL species that can cross the blood-brain barrier,

01:34:35.720 they believe the potential would be that, hey, some protective proteins are getting into the

01:34:40.600 brain. They've looked at some anti-inflammatory, anti-oxidative aspects of those proteins.

01:34:45.940 And they believe the ApoA1 can jump on an E4, a brain ApoE HDL particle and rescue it,

01:34:54.300 and maybe turn dysfunctional brain HDL particles into functional brain particles.

01:35:00.880 So, boy, it's a wonderful story on paper right now.

01:35:03.820 But the fact that the biomarkers are moving in the right direction, I think, gives us all great hope.

01:35:08.840 And I believe the company is going to put money into investigating this with further cognitive studies and more advanced studies and perhaps even some imaging studies, PET, things like that.

01:35:21.480 Although these biomarkers really, if you have those biomarkers, some people say you don't even

01:35:26.580 need that PET scanning anymore because they reflect that easier. So that's the quick story

01:35:32.200 with obesity or PIB. Yay for its APOB ability. We're all going to certainly be using for that.

01:35:37.620 That'll be its FDA indication. But if we get more and more information like this, that it's looking

01:35:43.160 good, especially in the E4 carriers, I think the real people would look at any downside. So far,

01:35:49.920 not any, or they would have had arrested their trials, but it's not FDA approved yet. So they

01:35:55.000 have more data to collect yet and we will see, but the hope is high. Yeah. I remain very optimistic

01:36:01.460 based on the data so far. And I think the key is going to be doing the right clinical trial.

01:36:05.940 Again, I think a lot of these things, if you look too late in the pathology, you might not make

01:36:11.380 enough of a difference. So the key I think is going to be patient selection and duration. You've

01:36:16.920 got to be able to select people who are high enough risk, E4 carriers, and catch them right

01:36:23.700 at that window. You know, I always go back to a study that I think did a great job of this,

01:36:26.780 even though it was a completely unrelated study, which was the PREDIMED study. This is more than

01:36:30.920 10 years ago, which was a primary prevention trial of dietary therapy for, at a minimum MACE,

01:36:38.180 but also I believe it even looked at all-cause mortality, or maybe it was cardiac mortality.

01:36:41.980 And again, it was primary prevention, which I always thought was, I thought the study would

01:36:48.620 fail. I really did. I was like, you're not going to do a dietary primary prevention study. Come on.

01:36:53.280 And not only did the trial succeed in demonstrating the superiority of a Mediterranean diet to a low

01:36:59.540 fat diet, it was halted early. And again, I think that's just a great example of if you pick the

01:37:05.720 right population, as I thought of it as people who were just about to drive off the cliff,

01:37:10.220 but, but weren't quite there, you could get an answer to a question in a few years. And I think

01:37:14.800 that's, that's, that's the way to think about doing this. And I, and I hope they can do that.

01:37:18.920 Yeah. Look, I'll just say, you know, Michael Davidson, your friend and John Castellan,

01:37:23.240 your friend, they are really driving all of these studies and they are well-experienced

01:37:28.400 trialists. So they will do the right studies. Tom, this has been an amazing tour of, of a topic

01:37:34.080 that is, is, is sort of new to the podcast. We haven't done sort of a deep dive into brain

01:37:39.280 cholesterol. But I think it's been such an important discussion because I think there's a

01:37:44.240 lot of confusion out there on this topic. I think that the completely different way in which the

01:37:49.320 brain goes about doing its business with respect to cholesterol from the periphery, I mean, hell,

01:37:54.920 most people don't even understand how the periphery deals with this. So why would we expect somebody

01:37:58.340 to understand the role of oligodendrostites and neurons and the different pathways and APOE versus

01:38:04.340 ZAPO-B. So again, I know that this podcast was a little technical, but I think you did a great job

01:38:10.520 of explaining it, anthropomorphizing it when appropriate. And obviously this might be the

01:38:16.480 podcast someone has to listen to or watch a couple of times and the show notes will be robust.

01:38:21.500 So I want to thank you. And as always, Tom, it's been, gosh, it's been 15 years since you

01:38:26.100 took me under your wing and helped me develop my understanding of this field of lipidology.

01:38:32.200 so I can never waste an opportunity to thank you publicly for your generosity. You're personally

01:38:40.320 with me, so thank you very much, Tom. And look, I'll wrap this up by saying,

01:38:45.200 yes, I was your lipid mentor for a while, but over the time, we've known each other a long time,

01:38:49.940 and I've got to experience your immense knowledge on things I had never even considered before. So

01:38:55.200 you've taught me just as much about so many things. I think it's a great partnership that

01:39:00.760 Thank God we bumped into each other and we've evolved into this role.

01:39:05.280 And I'm still going and have the honor of still working within your practice, not as

01:39:09.640 a prescriber, but just to keep the staff educated and, you know, I'm shipping you out.

01:39:15.200 Here's the newest, latest, and greatest stuff all the time.

01:39:17.960 So it's just been a phenomenal, wonderful way for me to continue my career.

01:39:23.740 So I love you eternally.

01:39:25.560 You know that.

01:39:26.900 And hey, sooner or later, it'll be another topic.

01:39:30.360 we're going to have to expound on again because the lipids keep changing and getting more and

01:39:34.620 more exciting. So I love doing it. Thank you, Tom. Thank you very much.

01:39:39.520 Thank you for listening to this week's episode of The Drive. Head over to

01:39:43.720 peteratiamd.com forward slash show notes if you want to dig deeper into this episode.

01:39:50.620 You can also find me on YouTube, Instagram, and Twitter, all with the handle peteratiamd.

01:39:55.780 You can also leave us a review on Apple Podcasts or whatever podcast player you use.

01:40:01.980 This podcast is for general informational purposes only and does not constitute the

01:40:06.160 practice of medicine, nursing, or other professional healthcare services, including the giving

01:40:10.480 of medical advice.

01:40:12.120 No doctor-patient relationship is formed.

01:40:14.940 The use of this information and the materials linked to this podcast is at the user's own

01:40:19.920 risk.

01:40:20.660 The content on this podcast is not intended to be a substitute for professional medical

01:40:24.540 advice, diagnosis, or treatment. Users should not disregard or delay in obtaining medical advice

01:40:30.240 from any medical condition they have, and they should seek the assistance of their healthcare

01:40:34.760 professionals for any such conditions. Finally, I take all conflicts of interest very seriously.

01:40:40.760 For all of my disclosures and the companies I invest in or advise, please visit peteratiamd.com

01:40:47.240 forward slash about where I keep an up-to-date and active list of all disclosures.

01:40:54.540 We'll be right back.

The Peter Attia Drive - June 08, 2026

#395 - Brain lipidology： understanding APOE, cholesterol homeostasis, Alzheimer's disease risk, and the effects of lipid-lowering therapies on brain health ｜ Tom Dayspring, M.D.

Episode Stats

Summary

Transcript