How to Drop Your Cholesterol with Dave Feldman [transcript]

Written by Christopher Kelly

March 17, 2018


Tommy:    Hello and welcome to the Nourish Balance Thrive Podcast. My name is Tommy Wood and today I am joined by Dave Feldman. Hey, Dave.

Dave:    Hello, Tommy. Thanks for having me on.

Tommy:    Thanks for joining me. Anybody who has gone on a low carb or ketogenic diet and worried about their cholesterol I feel has probably looked up or heard of some of your work. The two of us met Low Carb Breckenridge last year and you're definitely the best example, I think, of a citizen science guy who's exploring things about health and biochemistry that the general science and medical establishments have really failed to do so far.

    For those who don't know about you, why don't you tell us about yourself, a bit about your history and then specifically tell us about the time that you became interested in cholesterol.

Dave:    Sure. About almost three years ago now, in April of 2015, I started on a ketogenic diet, really enjoyed the results I was having, was really setting personal records with distance running and felt better than I had ever did in my life. My father and my sister likewise got inspired and they went on it shortly after I did.

    They had gotten their cholesterol test before I did seven months later. They had gotten theirs and I had warned them that their cholesterol might be a little bit higher. That's not uncommon with the ketogenic diet. Whereas mine seven months later went super high. My cholesterol effectively doubled, my total and my LDL effectively doubled and I found out this word that I kind of heard in passing called hyper-responder which gets applied to a lot of people who go on a ketogenic diet that in response to the diet it appears that their cholesterol is very high.

    Well, I am a software engineer, before this point in time had no interest in biochemistry or even medicine or science. I just liked writing code. The irony is I started learning about cholesterol and the lipid system that governs it and I found some familiar patterns that I see in software development that I relate to quite a bit, distributed object networks.

    With that, I then started, as you said, doing a lot of self experimentation really kind of obsessively and it's been that way for about two and a half years. What I found was that I could manipulate my cholesterol numbers substantially by changing my dietary fat and since then there's kind of been a phase two of like the last, I want to say, either nine months where I have also changed what I would say as my energy status.

    I've been doing things like swap in and out fat for carbs and so forth. A lot of this you can read about in the blog much more than I could go into here. But in general, I found that the lipid system, and in particular cholesterol, is actually very dynamic and easy to change based not even on changing medication or supplements but simply on changing dietary energy particularly in the form of fat.

Tommy:    You mentioned that you could manipulate your cholesterol levels pretty quickly based on fat intake. I think for those who aren't familiar with your work, this is going to happen in a very non-intuitive way or not in the way that people expect. You developed something called cholesterol drop protocol and they are based on just changing how much fat you take in and then seeing how your lipids change. Can you tell us how the protocol works and then what you see?

Dave:    Sure. It's based on the pattern that I observed that the three days before you take a cholesterol test. Right now, we're recording this on a Friday, for example. Let's say that I had just now taken my blood draw for cholesterol test. If on a ketogenic diet, because almost all my data right now comes from those people who are on low carb, if in the last three years, the Thursday, yesterday, and the Wednesday before that and the Tuesday before that, if in that three-day window period of time I had had a lot of fat in my diet then it actually would have inversely pushed down, my LDL-C would actually drop in response to me eating a lot of fat.

    In the reverse, if I hadn't had a lot of fat, if I just didn't eat as many calories or especially if I fasted the last three years, my cholesterol would skyrocket. It would be a lot. The protocol that a lot of people have done now, we've lost count, it's somewhere over 100 now who have actually done this, have themselves done this simple version of the protocol which is just that they've eaten enormous amounts of fat for three days just before getting their cholesterol test. While it's not a formal study, we have about an 85% success rate. We're seeing their LDL-C drop from what they would have expected to be.

Tommy:    So, when you're looking at this whole kind of network, I think you've talked about how this is basically just an energy delivery system and particularly in those who are low carb or in ketogenic diet. Can you explain how that kind of theory came together and how the various things that we can measure, say on a basic lipid test, move and how that's based on both energy requirements and delivery and production?

Dave:    Absolutely. It's cool because this will be a great opening for where we'll get to a little bit later when you and I had that great conversation about endotoxins that later led to the Twitter thing.


    I'm coming from the energy perspective of, again, I'm sort of thinking like an engineer. I'm thinking, okay, you have effectively 100 trillion customers in your body, cells that need energy delivered to them because the vast majority of your cells are stuck where they're at.

    You always start in, again, reverse engineering. Where do my cells get their energy? In the ketogenic world, we love to talk about glucose and we love to talk about ketones. The irony is, there's not actually that much talk about fatty acids. Glucose and ketones, as you know, are water soluble and, therefore, they can travel natively in the blood because the blood is water soluble. They're hydrophilic. They can swim natively. They don't need any additional carriers of any kind. They can get straight to the cells especially the endothelial cells and the bloodstream and, therefore, they can distribute relatively easy.

    Fatty acids are different. Fatty acids are a lipid. Fatty acids, if they're going to be delivered to cells, they need something that carries them that can swim, that is hydrophilic. For that, we make something called a lipoprotein. Whenever somebody hears LDL or HDL, that last L is lipoprotein. Lipoproteins are actually an ingenious way that the body says, "Hey, I'm going to put every kind of lipid, not just these fatty acids for cells to use but also fat soluble vitamins, also cholesterol. I'm going to put them all in the same lipoprotein."

    That's usually a low density lipoprotein which oftentimes are used as a class but I should just say that's a lipoprotein that's usually more on the provider side. It's either in the gut and those created in the gut are chylomicrons or created at the liver and that's usually a VLDL and they're bloated with these fatty acids that are used for energy.

    When they come into the bloodstream, that's their first job. Their first job is to distribute these fatty acids like crazy throughout the bloodstream and they do that within a fairly short period of time. That's how cells can get the energy that is not glucose or ketones. This is why I often kind of joke that, and it's not even a joke, that I'm not on a ketogenic diet. I'm actually on a fatty acid diet.

    Ketones are kind of secondary to the actual direct delivery of fatty acids either by these lipoproteins or actually from natural ongoing lipolysis that's occurring with my adipose tissue, my body fat. The more that I get into this the more I see that it's actually a very elegant system by which it's distributing energy to cells in a very regular way.

    A thing that I kind of try to keep bringing people back to to sort of understand this from a systems standpoint is it's better to not think of cells getting energy like they're customers in a diner. It's better to think of them as like customers at a buffet. It's the job of the body to keep this buffet constantly stocked that they can get energy on demand. Because the body can't actually monitor every single cell's energy preference in the minute. It can only do it in the larger aggregate.

    From that, it's actually very good. It's very good at constantly up and down regulating the amount of the buffet availability of energy. That's really the core of my whole research is that I keep finding that that homeostasis of making that energy available is extremely sophisticated. It's one of the best networks I've ever looked at and it wasn't written with code.

Tommy:    So, then a simple summery of this point would be that particularly eating the way that you do and maybe you'd expand that from there with future experiments, so you're doing that currently, the more fat that you're taking in the less LDL needs to be shipped out of the liver to provide energy, right? So then the more that you're taking in directly, the lower your LDL cholesterol needs to be. Is that essentially where it comes to you?

Dave:    Yes. And I'll only qualify this slightly because the biochemists always ding me on this, even though I try to simplify it for audiences, technically the thing that's being secreted from your liver is a VLDL. That's a precursor to an LDL. VLDLs are very bloated lipoproteins, bloated with triglycerides, and they're going out into the bloodstream and their job is to drop off those triglycerides to the cells all throughout the bloodstream that need them. This is why you find people who are down the road in metabolic syndrome and intense insulin resistance.


    They have a lot of VLDLs resident because they aren't getting that job done. They're not finding anywhere to drop off these triglycerides. There's no parking space, therefore, that precursor to LDLs, the VLDLs, don't get that initial job done. But it's important for your listeners and everybody to know this. VLDLs get remodeled to IDLs and then finally to LDLs.

    The VLDL and IDL stage is really like 60 to 90 minutes total. The rest of its life span, that particle's life span, if it does not get absorbed by the liver and remains, it will be two to four days later before it gets absorbed by the liver. That means 99% of its job is actually patrolling after its first and biggest job, that first part, that 1% of dropping of the energy.

    This is very, very relevant to me because when you're getting a fasted cholesterol test and they're pointing to the LDL cholesterol. We're talking about that last stage. It encompasses 99% of the whole life span of that particle. If you're powered by fat, and in particular if you're not powered as much by glucose, then it stands to reason that you are naturally in a fasted state going to be trafficking more of these VLDLs. They are the fatty acids you were powered by.

Tommy:    At this point, let me jump forward a bit to a question that I had and I sent you previously which is about the coupling and what you see on a blood test versus what is actually happening in the body. One thing when you look at the different lipids in your experiments, the LDL cholesterol tracks much more tightly with fat intake than triglycerides do.

    If triglycerides are the cargoes being carried around for energy you'd expect those to be tightly coupled, the fat intake, because that's where you're mainly doing in terms of energy delivery. However, you made a very good point which is that, just like you said now, when you're measuring the LDL, it's already done all of its job of delivering those triglycerides. It's already done that part as a VLDL and IDL and you're not measuring that.

    However, that's kind of convenient or inconvenient, depending on the way you look at, because that doesn't mean you can disprove your hypothesis with an experiment because you'll never be able to measure it doing that sort of flux through the system. Do you have any thoughts on how we could maybe confirm that that's what's going on?

Dave:    I did. This is a fairly elaborate experiment, but it's possible. Let me first kind of expand on the answer that I gave you then and maybe it will help if I kind of do it in a chatty way here which is the depletion of the triglycerides on a per particle basis would have higher variability which to me makes sense.

    Because again, getting back to what cells need the energy, what cells don't, I wouldn't be surprised if that had a higher standard deviation, if that was a noisier up and down chart than the larger aggregate of cholesterol. It's kind of hard to illustrate why except to say this. As far as I understand the secretion of these VLDLs, and LDLs if they don't ultimately start as LDLs or if they don't start as VLDs, they come actually fairly standard with the fairly fixed amount of cholesterol.

    That's kind of like cars coming from the dealership. They're going to have exactly one spare tire. Well, the variability of the size of the VLDL can be high. It can be triglyceride rich or maybe triglyceride poor. But as far as I know up to this point, I don't know of a circumstance in which the body intentionally generates a VLDL that is cholesterol rich of cholesterol poor or even an LDL that is secreted that never means to catch APOC2 or something in a more pathogen binding state and so forth.

    Basically, when it gets secreted there's going to be a fairly fixed amount of cholesterol relative to the amount of triglycerides. There is that component there that is going to stabilize the amount of cholesterol more to begin with. There's the additional component of how much is actually taken up by the cell, the actual hydrolysis of the fatty acids coming off and the difference of needs from one part of the body to the next.

    Here's a good example. Would you expect that -- and this is basically the experiment. I'd be curious if you would think that this might work. What if I could get my blood drawn from all four of my limbs at the same time? Would you expect that the triglycerides would vary more than the cholesterol for each of the four blood tests?


Tommy:    If you're doing a venous sample from after it's been through a larger body of, say, muscle tissue like it's been through your quads, say your quadriceps, then you might have lower triglycerides because more has been dropped off.

Dave:    Correct. Now, my guess is that, yes, you would see a higher variability but probably on a much tighter delta. I don't actually think it would be dramatically different between my left arm and my right leg but I'd be willing to bet that there is a bigger and more defined delta than there would be with the cholesterol. That's my guess.

Tommy:    I mean, that makes sense. The obvious question next is when are you going to start instrumenting your hepatic and portal veins to find out the cholesterol that's directly going in and out of the liver?

Dave:    Believe me I have joked about this several times. The second phase of my research has everything to do with energy status. I've mentioned this so many times. I've really tried to find a way by which I could detect the glycogen stores in my liver. As of yet I don't know of any way of doing it other than a liver biopsy and frequent liver biopsy is absolutely not an option for me.

Tommy:    Yeah. Which I think is pretty fair.

Dave:    Even then I don't know that I would trust the existing science on how accurate that would be because I don't know that it's well and evenly distributed throughout the liver tissue. Unfortunately, I have to kind of just go off of some theory to some degree.

Tommy:    Okay, that is fair enough. One thing before I ask the next question is about how you're determining these correlations. I just wonder if because of the system you might necessarily see a nice linear correlations as things go up and down but there's still a very defined pattern there but it just would require some more complex statistics to kind of pass that out. Have you looked at some of that? Maybe it's an analysis method that's changing how tightly correlated things are rather than the system itself?

Dave:    Well, first and foremost, I'm definitely not a statistician. I found the three-day average just through purely running each one. So, once I had eight data points going back to the very beginning of my research, I just crunched it, said, "Oh, wow, look, LDL-C and LDL-P seem to correlate pretty tightly to a three-day average relative to all the other times that I did this."

    As you know, this ended up getting adjusted later. Once I did an experiment where I performed over one week a day after day blood draw against a food plan. I was inducing curves up and down and my total calories to see what the curves up and down in my LDL-C and LDL-P was. Sure enough, the LDL-C helped tight the three-day window.

    The LDL-P turned out to seemingly have a delay and that's where I found out that LDL-P, the actual particle count itself, had a two-day gap that, to use that same example that I was using before, this is again recorded on a Friday, instead of using yesterday, Thursday and the day before, Wednesday, I would look at the window that is Tuesday, Monday and Sunday going backwards. That three-day window seems to have the highest inverse correlation with LDL-P. This is probably a good segue for your next question, I'm sure.

Tommy:    Yeah, it's a perfect segue because then I'm going to ask -- So, you've mentioned that the amount of cholesterol in a given particle is going to be fairly fixed. If that was the case and the LDL cholesterol was so tightly coupled then you would expect that to be most tightly coupled to a number of particles and the particles go up and down based on delivery need. But there's a lag between the LDL-P and the LDL-C which means that the amount of cholesterol in those particles must be changing. Why do you think that is?

Dave:    Here's where I'm going to give you kind of a little bit of an advanced answer that's part of my theory on this. my issue is I had fought for a long time that when I was getting NMRs that it was a direct measurement of LDL cholesterol and I was disheartened to find out that it's actually still calculated. As you probably know, and I know your listeners know, commonly on a lipid test what they're giving you for LDL-C is technically a calculated number.

    It's a fairly reliable one as long as your HDL is not too high or too low and your triglycerides are not too high or too low and it's called the Friedewald equation. Because of this, it occurred to me a long time ago, actually, that maybe my LDL-C as calculated isn't actually correct, that actually what's happening is I'm seeing a leading indicator of where my LDL-C will end up being. This is part of what's exciting about it.

    My body is seeing, for example, higher incoming non-esterified fatty acids. Those are the fatty acids released into your bloodstream from lipolysis from your adipose tissue making its way through carrier molecules albumin, binding to albumin, making its way over to the liver.


    And then the liver ends up actually, for a period of time, let's say this gets higher, the NEFAs get higher, making more triglyceride-rich LDL, I should say VLDL particles, making more triglyceride-rich VLDL particles and then catching up over the next couple of days to meet production demands for the amount of cargo that's coming in. And just constantly doing this, constantly moving up and down production of particles to carry what appears to be the triglyceride needs of packaging from these free fatty acids coming in.

    So, that's why I'm anxious. Unfortunately, I don't have a cheap way of doing this yet, of getting direct LDL-C measurements and doing constant comparisons against the calculated because I speculate that the direct LDL-D measurements would be closer to the LDL-P equation, not the LDL-C Friedewald calculated equation. Does that make sense?

Tommy:    That makes sense. And it could still be a hybrid between those two theories where you're changing what you can change most rapidly and then altering particle production more flowy as things go up and down. So, could be both the calculation thing and what's actually happening in the system itself.

Dave:    Right. I cannot emphasize enough how amazing the human body is that it could be doing this, that there really can be that speed of response. Because this is super fast when you think about the actual number of elements involved. I mentioned this recently. You know that LDL particles are measured in nanomoles and you know how much a mole is in math. It's a ginormous number. It's a number that we can't even easily describe.

    A nanomole is basically a million trillions. Your body isn't opening and closing a small pipe to determine how much or how little the LDLs are going to be absorbed and distributed. It's actually removing these and secreting them one at a time in mass. The acquisition of them, the absorption is happening through receptors. If you can just imagine these little tiny arms in your body particularly on your liver that are just boom, boom, boom, just taking these guys away.

    And imagine this is on a scale you couldn't possibly comprehend all to meet this level of homeostasis, all to keep balancing up and down this amount of energy as needed. It's astonishing. I just have to say I have almost kind of a fanboy-ness for this network now. It's just amazing.

Tommy:    You're right. When you put it like that, it's pretty incredible how it works. That then brings me to the question of how you think this may translate to people outside of the ketogenic arena. Because I feel like it's fair to say that the majority of people aren't going to be keto anytime soon in the world. So, how much of this do you think, or maybe you're starting to figure it out based on your other experiments, how much is relevant to people eating some kind of mixed diet?

Dave:    Well, actually, that's a very big deal to me because, obviously, however much the inversion pattern applies to the non-low carber is super relevant in that if their cholesterol numbers are changing very rapidly based on what happened in the three days before I'd certainly hope that they would know that and would therefore make whatever changes they needed to, or for that matter, lack of changes that they need to be mindful of eating the same way for the three days coming up to a cholesterol test that they would do that.

    But it's certainly super relevant if their prior cholesterol scores were highly influenced by what it was that they were eating or not eating in the days before. Here's where it's going to get a bit more complicated. The whole second phase of my research basically started in May of last year and it's where I started doing carb swapping. I effectively left the ketogenic diet in the course of these experiments by adding in more carbs than I otherwise would and in doing so I would see that past a certain threshold my cholesterol would drop substantially.

    That led me to believe that there really was something associated with an awareness on the part of the body of energy status. To kind of pull it back to a 50,000-foot view, we have effectively two tanks of energy in our body. The large long-term tank which we're all pretty familiar with which is body fat. That's energy that's not very easy to get to. That's your walk in freezer. You need to actually go get it. You need to thaw it out.


    Lipolysis is not a super fast process. And there's short-term energy and the short-term energy everybody is used to talking about is glycogen stores especially in the muscle. That's like just ready to go right now. Right at this moment in time your house caught on fire and you had to just bolt at top speed out the door. You're using glycogen from your muscles.

    But there's also glycogen storage in your liver. That too is going to get provided much faster in the whole process of actually burning fat and moving in through at the system and so forth. Well, is it possible that the body is more aware of that short-term energy tank than we thought? I would submit that I now have a lot of evidence that suggest that it is particularly in how it's treating fatty acid availability.

    Because it appears as if when I get to a certain level of glycogen store top off, maybe it's not yet 100% but maybe it's like 80%, 90%, my body seems satisfied enough to not feel it needs to mobilize fatty acids or fuel in the form of triglycerides packaged in low density lipoproteins. How would I see that? I would see that through my favorite tracer which is cholesterol.

    If I'm not seeing as high LDL-P then I can assume that there wasn't as much secretion of its precursor VLDL. Why wasn't there as much secretion? Well, because my body didn't feel it needed to have as much of that second version of a short-term energy tank that's in fat because it had enough of the other version of the short-term energy in the form of glycogen stores and, I think, in particular at the liver.

    It makes perfect sense. That's the brain that's already handling this secretion of the fatty acids in the first place. It sees that its own glycogen stores are up to certain amount that it says, "Okay, we're in fairly good shape right now to where we don't need to circulate as much fatty acid for the cell tissues." Does that make sense?

Tommy:    Yeah.

Dave:    Given that, your average person is on a carb-centric diet -- Let's go to the other end of the spectrum. Let's talk about vegans. Vegans are often showing extraordinarily low levels of LDL. Of course, they also often will have very low levels of HDL. But in the low levels of LDL from their perspective this is a marker for good health. I would say that I'm not as concerned about that so long as they have low remnant cholesterol and low triglycerides.

Tommy:    Which I'll ask about.

Dave:    Yes. Yeah, we'll do that too. Basically, I believe that this is because their body is constantly in the case of topped off glycogen scores if they're getting enough carbs, and they probably are, it's a high carb low fat diet. Their liver is satisfied that, no, we don't need to mobilize as much for fat. Where this can be problematic is if they're not making and using enough cholesterol to where it can feed into hormone supply.

    Because, of course, cholesterol is needed for all sorts of things throughout the body and one of the things that you want to care about a lot is sex hormones. Unfortunately, for example, some vegan women might lose their period or something like that. I think that that's something I would be concerned about. I would say you should at least be sure that you get enough cholesterol or at least that your body is synthesizing enough cholesterol that you meet the level of sex hormones in your cycle.

    A lot of this is sort of theoretical and kind of outside my wheelhouse. I'm just more mentioning why at this other end of the spectrum it can make sense that if you're primarily fueled by carbohydrates that -- Let me qualify. You're primarily fueled by carbohydrates and you're metabolically healthy, that your LDL would be much lower because you're just simply powered more by carbohydrates and your glycogen stores are topped off more.

Tommy:    That makes perfect sense. The metabolically healthy thing, I think, is interesting because in my mind if you're metabolically healthy the system is dynamic and intelligent enough for us to be able to fuel switch fairly frequently up and down as much as we need to. I mean, if people are all the way vegan, it's obviously consistently carb based or very high carb. Keto is consistently much more fat-based. I feel like people should be able to exist in the middle and switch between the two as long as metabolic health is good.

Dave:    I actually agree with this even though I'm featured a lot in the low carb high fat community. I think it's fine if you are low carb, I think it's fine if you're vegan, I think it's fine if you're mixed. The things I'm looking for inside of the lipid system is proper metabolic flexibility which actually surprisingly I think cholesterol test say a lot about. They can actually give lots of strong indicators as to how well you're handling your insulin, how well you're handling your existing energy load and how effective it is at getting that energy to tissues.

    So, let's go to the other end of the spectrum. I was just talking about vegans. The other end of the spectrum isn't just low carb and hyper-responders but kind of profile I like to call lean mass hyper-responders.


    Lean mass hyper-responders have the highest LDL that I see, period. In spite of that high LDL they often have a stellar array of blood markers. Their inflammation is crazy low. On lipid side, the triglycerides are super low often 30 or 40 milligrams per deciliter and their HDL can be super high, 80, 100, some I've seen in like 130, 140 HDL.

    Now, again, I'm an engineer. Mechanistically, what I see is if you have an LDL of 300 and you have triglycerides of 40, I see, okay, these LDL-C, these particles that start out as VLDLs that ultimately yielded the LDL-C we see, they clearly succeeded at dropping off their triglycerides, if we believe the secretion was much higher which I believe it was.

    If they succeeded at doing it, well then actually the triglycerides of 40 milligrams per deciliter is a huge indicator that there needed to be that many boats to traffic that much triglycerides because, by golly, it's getting used. Mechanistically, it makes perfect sense to me that somebody who, in all lean mass hyper-responders, not all of them, but I would say about 95% are lean and/athletic and usually both.

    If you are a low carber and you are working out like crazy and particularly if you are very lean, if you don't have a lot of adipose mass, then it makes perfect sense that your body would be trafficking more VLDLs to feed those fatty acids to your tissues and part of the huge clue is how low your triglycerides are in a lipid test.

Tommy:    That does make perfect sense to me. This is the pattern we see very frequently in the people that we work with. They have very low triglycerides. They often have high HDL which we'll talk about too. My personal bias on this is that because of the system you described, when somebody is low carb, and like you said, exercising like crazy, low triglycerides to me particularly once they're down to like 30 or 40 tells me that person isn't taking in enough food.

    We see a lot of athletes under-eating and they often feel good early on, maybe even for a year or two, ketone levels are probably super high, they feel really energetic but eventually the wheel starts to come off and we see problems with thyroid function, we see problems with sexual hormone function, everything along that sort of HPA-G and T axis. Do you have any thoughts on that or are you seeing lots of people whose triglycerides are 30 and they're actually in great shape all the time for long into the future?

Dave:    To be perfectly honest, I don't know that I had anybody come forward with this yet and this isn't to say that this isn't happening. I mean, in a sense, really Tommy, so much has happened so quickly in my life and for in as much as I'm treated as kind of a lipid expert there are a lot of things for which I feel like there's still a lot that we're learning.

    Unfortunately, there are a lot of ways -- I tend to be the proverbial skeptic. Any article I've read that has kind of made claims as to what cholesterol and the lipid system are doing and what's actually good or bad sign, I have to first start with the, okay, did they take into account the research that I've had up to this point and, of course, almost all of them haven't.

    And this isn't from an egotistical standpoint. It's that there's such a massive confounder in the short-term variability that that needs to be accounted for first before we can make huge assessments. This goes both ways. This is to say that I can't say with certainty that lean mass hyper-responders don't have a higher risk factor in the having the higher LDL, and I'm careful to say that too, that I don't know for sure.

    But I'll say this. Without first accounting for that short-term variability, that's the first part of my research, and without high frequency testing, we can't feel very confident about tests that were taken, say, every six months or a year for anybody. This is why I kind of like the protocol in that it starts establishing at least how much you have variability, metabolic flexibility and doing something that's an experiment in the short term.

    So, where I'm going with this is if you've seen people will have low triglycerides that have been athletic, have been keto up to this point for a long period of time but they then start developing other issues, it's very possible that this is the case. But I have had a number of lean mass hyper-responders who have reached out to me that had been keto for half a decade, for example, who I've asked these very things. I specifically say, are there any other blood markers you have that are out of range?

    Thyroid is one of the ones I zero in on because, of course, hypothyroidism is one of the first presumable reasons for why you would have higher cholesterol because it's very common outcome of it.


    To date, the only thing that I find that comes up with thyroid in particular is with low carbers just in general, not even with lean mass hyper-responders, there's often say a little bit of a lower T3. That's just a little outside of range at the bottom. Now, otherwise, the only other thing I get that's occasionally out of range for lean mass hyper-responders is unsurprisingly higher T. It's a little outside of range, and higher creatine which can occasionally change with the creatine/BUN ratio and so forth.

Tommy:    Okay. You mean, creatinine.

Dave:    Yeah. So, outside of that, I haven't had a lot of word on that front. But to be fair, I haven't really pursued much beyond just that statement that I said that's more on the screening side. Were there any other markers that you had there out of range? And usually they come back with no. But to be fair, to be even further skeptical of even my own screening, it's very possible that they over focused on cholesterol and anything that was outside of range but only by a little bit they kind of consider to be a non-plus.

    It's my desire to ultimately try to assemble a fairly good network of lean mass hyper-responders for lots of different tests because they kind of represent one end of the spectrum of my research. I'll be mindful of this because I like to pursue this and I might actually get offline from you. What makers you see that kind of spin out, and try to pursue this in particular with this population.

Tommy:    I think it's a great idea. Like you mentioned, I mean, the most important, if you sort of forget the details, the most important part of the research you've done is just to show how dynamic the system is. Testing your cholesterol every six months or after you've been on a statin for however long or whatever it is that's the standard medical practice, it's just completely crazy because you could be absolutely purely influenced by what you did the three days beforehand.

    So, complete change in that paradigm is a huge part of this. The other part that you mentioned was HDL, HDL going up and up and up. I listened to you and Professor Ken Sikaris on the Low-Carb Conversations podcast and Ken mentioned that HDL is good and then LDL is good until it's bad. However, I'd argue potentially that HDL is also good until it's bad. We have both epidemiological and mechanistic data to suggest that.

    And so in people on a population level and generally people eating a mixed or western diet, once you get above an HDL of 70 or 80 all cause mortality tends to creep up. Some people, so by this conversation we had on Twitter, George Henderson mentioned that if the main reason that people's HDL goes up on a population level is because of alcohol and they're drinking too much alcohol whereas if you're HDL is going up because you're on a ketogenic diet or increased saturated intake, obviously, the two outcomes are going to be very different.

    So, when we're looking at it on the population level, it's probably because we're actually looking at those alcoholics and they're being killed by the alcohol and liver failure and all the associated problems. But in keto, that HDL could go, like you said, 100, 130 or higher and it's not something that you need to worry about. Do you have any thoughts on that? I have no evidence particularly either way. I'm just asking.

Dave:    No, I definitely do. In fact, I think that you may have quoted a study that I was interested. There was a study that took it from insurance claims that I think is like a million and a half people, is that right?

Tommy:    Yeah.

Dave:    They found that there wasn't a lot of correlation with LDL and all cause mortality but there was a higher correlation with all cause mortality with higher HDL. That definitely was something that I latched on earlier on my research because I was very curious about that. I think that certainly when you find out more about alcoholic liver disease, you certainly find that it's a very common factor that HDL can increase with, and I'm curious about the pathways behind it.

    But again, I keep swinging things back to let's first find out what these variabilities are with low carb in particular and why they would increase. And in the case of LDL we kind of spend a good part of this program with me explaining energy distribution. Why would HDL go up so substantially?

    Well, something that doesn't have a lot of research behind it that I'd been very curious about, very interested in is the potential pathway of HDL being a helping hand in the distribution of triglycerides as well. Now, this is not fairly often discussed. But the apolipoprotein -- I'm going to get a bit geeky here but I kind of need to.

    There's a kind of apolipoprotein that's the sneaky, bumpy protein that attaches to the outside of the whole of a lipoprotein that signals or particularly binds to receptors specifically for hydrolyzing the triglycerides inside so that they break up and become fatty acids and come into the cell's membranes. That apolipoprotein is APOC2.


    Surprise, surprise, HDL particles is the one that supplies the APOC2 to the VLDL or, I should say, to the LDL that ultimately becomes the VLDL in the process of gaining these apolipoproteins from the HDL particle in a circumstance which has a lot of triglycerides in circulation from bloated VLDLs and so forth. So, VLDLs and LDLs can interact with C-reactive protein, but oftentimes they're talked about in terms of HDLs interacting with C-reactive protein.

    It's kind of like a bridge, if you will, between these lipoproteins so that things can be exchanged. Both apolipoproteins can be exchanged but also, interestingly, triglycerides and cholesterol. They found that, sure enough, HDL, which seems surprising at first, would drop off its cholesterol into the LDL particle and the LDL particle would off load some of its triglycerides into the HDL particle.

    It turned out that this was part of the larger reverse cholesterol transport. Basically, HDL went and got a bunch of cholesterol that's found in your arterial wall that's why it's called the good cholesterol. HDL is considered the good cholesterol because its particles are capturing this cholesterol bringing it back to the liver to drop it off.

    It turns out that, actually, a lot of times they don't just bring it back to the liver. They take it and put it into an LDL particle. The LDL particle in a big group of a bunch of these is taking it back to the liver for them, so it's kind of part of the larger chain. Well, what happens to the triglycerides that are then on loaded to the HDL particle?

    There's really not a lot of focus on that past that point because everyone's so focused on LDL. But there's a lot of speculation that I've read and some amount of study not enough to where I'm going to hang my hat on it that the HDL then distributes some of the triglycerides as fatty acids to the same cells that would be binding to an apolipoprotein C that was on the LDL except now they're pulling it off of an HDL particle.

    Okay. That in mind, it seems very reasonable that those who are lean mass hyper-responders representing one end of the spectrum are showcasing this because both their LDL particles and their HDL particles, more specifically their HDL cholesterol, is getting higher, seemingly in a thus dependent way.

    I often see that these two numbers track together, that as LDL goes high -- and you've probably seen the same thing -- HDL likewise goes high in somebody who appears to be metabolically healthy and flexible. That needs to be accounted for because that's a different hat that's being worn by HDL and we don't know for sure if the other reason high HDL-C that turns out to be deleterious is for a different hat that's being worn, if that makes sense.

Tommy:    Yeah. That makes perfect sense. If you're up regulating part of the system you want the whole system to go with it rather than just individual aspects. So, you kind of expect those to track together. I think that makes sense. I mean, you said multiple times very honestly that what that means in terms of the long-term scenario, we just don't know yet but we're still just -- We think that we know so much about this but in reality we don't really and so now we're just starting to figure it out.

Dave:    Right. There really needs to be extra emphasis placed on this one key factor because lipoproteins wear lots of hats and you're familiar with this because you've done much research on it and how they're involved with clearing of pathogens, for example, and their impact with lipopolysaccharides and so forth.

    I like to use this analogy. They're like cars. If I told you, "Hey, last week, man, you wouldn't believe it, I used my car a whole lot." That doesn't tell you a lot of information. I could have been using my car a lot because I was taking my sister to the hospital and you'd say, "Oh, I'm sorry to hear that." Or I could have been using my car a whole lot because I was going to movies in which case you'd go like, "It sounds like you're having a good time."

    If lipoproteins only had a single job, if they were only carrying a single lipid then it would be a lot easier for us to say, "Ah, this is indicative of a good thing or a bad thing," or even the particle itself is a good thing or a bad thing because there would be nice clean correlations that we could see.

    Instead, we have to take into account the fact that, no, they not only deliver energy, as I said in that first 1%, but that patrol period of time, the 99%, that could be either because they are being recruited more to fight an illness especially like an infection. That can also be so that they can supply the means by which for an injury to be healed because they can use more of those for raw material through endocytosis.


    There's a lot of things that move the numbers up and down that are good things and bad things altogether. The larger question we have to keep coming back to is are LDL particles and HDL particles at certain quantities themselves a problem or are they an indicator of a problem due to their up and down regulation to meet it?

    The truth is a lot of it is unknown. A lot of it is unknown. But I do feel confident that I'm helping to identify a channel by which it makes perfect mechanistic sense for these numbers to be higher both for LDL and for HDL and the means by which to deliver energy and that that has a natural outcome of there being longer residence time in the bloodstream.


    I am pretty confident that longer residence time in the bloodstream isn't by itself an independent cause of atherosclerosis. I think there can be arguments made that it's possible it is a component of a larger series of other things but as far as it being truly an independent causative agent I feel like that's not even a discussion to have with me because the correlation would be strikingly strong and there are other things like remnant cholesterol that I find to be just absolutely crushing it in head to head matches.

Tommy:    So, we'll come back to that in a second but I just briefly want to touch on the endotoxin point which I think you've actually -- If you piece together one of the things you said, I think you've covered it pretty well, which is that after your VLDL has dropped off its triglycerides and has become and LDL pathway, it has this circulation time where it does other things.

    One of those things is to bind to endotoxins that come across the gut wall and then they cleared at least partly once the LDL is then taken into the LDL receptor. So, one of the best ways we know to get endotoxin across the gut wall other than having some kind of inflammatory bowel disease or something like that is to take in particularly refined fats.

    So, part of my question always was during your drop protocol, if you're shoveling large amounts of fat into the system, is part of the drop in LDL associated with those endotoxins being cleared? In my mind, based on what you said, I think that could be the case but it doesn't necessarily need to be a bad thing because I also think that the innate immune system is going to be better coupled and function better in somebody whose metabolic health particularly is improving on a ketogenic diet.

    That also doesn't prevent the first bit from happening, which is the triglycerides being dropped off because then afterwards LDL can just do its job as it wants to. I think I've kind of answered my own question but do you have extra bits to add into that? One issue is that the theory is the same in somebody who's septic, they tend to see a massive drop particularly in LDL because the LDL is being used up again as part of fighting the infection.

    I think that's probably a bad thing particularly because people whose cholesterol drops the most end up doing the worst where they've studied that. Again, this is the hypothesis that I can't really disprove with basic lab testing because the way it is to measure endotoxin and all that kind of stuff is, I mean, with normal lab test is just really terrible. It's something that I can't really check as much as I'd like to. Any sort of thoughts on all that would be very much appreciated.

Dave:    Going back all the way to our first conversation, it's certainly something that while I'm very skeptical of it, I think it's very worth proving or disproving. Certainly if what's happening is by ingesting a lot of fats I am basically increasing my endotoxin load, I certainly want to know if that's one of the major if not the major component that's increasing or decreasing the total amount of LDL that we see in the resulting blood test.

    Even if it was not the majority reason, even if it was 10% or 20% of the reason, I'm certainly curious and would love to isolate that. The only thing I'd been able to think about which we would have to, you and I may have to just figure out a more complex experiment because the plus is that I have such a wealth of lipid data as me as an individual that I really can find out something with a lot of confidence relative to somebody who's just only had a lipid test every six months or something. Because I've done these extremely sensitive experiments to isolate where the inversion pattern sits and what foods cause what and so forth.

    What would be interesting is if we could change the composition of the fats that we know would likely have a higher or lower endotoxic load with the qualification that I wouldn’t be shooting for the most endotoxic fats. I'd be trying to figure out something that would be reasonable that I could do within a limited quantity of time against a likewise calorie per calorie cleaner fat if you will.

    In which case, we should be able to see if it follows more of the energy side versus the -- I think the problem, you and I are both going to see though is we're still kind of talking something that's probably more refined versus not, right?

Tommy:    Yeah.


Dave:    Unfortunately, we'd already have a natural confounder in that the refinement level could itself be a confounder.

Tommy:    Yeah, absolutely. One thing you could do but it would take a longer period of time is I've been looking a lot of endotoxin antibodies. You can do them. Cyrex do an endotoxin antibody panel, but it's obviously expensive. One of the confounders there again is in athletes, which is a population that I've read all about and athletes after intense exercise they get an increase in endotoxin in the gut.

    They also upregulate endotoxin antibody production to protect against that. It's a perfectly normal thing. But immediately after exercise, you'd actually see a drop in endotoxin antibodies as they bind to that endotoxin and then are cleared. So, there is some way if we could get Cyrex to sponsor us, you could maybe look at dynamics of endotoxin antibodies to see whether you're getting increased or decreased endotoxin because we know that that can be fairly dynamic. But again, maybe that's not going to be particularly cheap experiment.

Dave:    If somebody else is paying the bill I'm happy -- If there's anything that anybody reads 1% of 1% of my research has come to understand is I do quite a lot for science. I'd be a happy guinea pig just to get even more data for us to make use of.

Tommy:    Well, I definitely think about that. I'm really answering that question but I know that we're kind of running out of time and just a couple of things I quickly wanted to touch upon was the sort of -- You mentioned the LDL particle over time causing atherosclerosis. Even super smart guys like Peter Attia, he's detailed out that process of atherosclerosis and as far as I can tell, his theory is basically that atherosclerosis is a function of LDL particle number times by time.

    Knowing what I know and listening to all things that you figured out, I do not have it in me to think that it's that simple, and there has to be some broken aspect of the system or something else that's causing that issue which then the particles could contribute to but they cannot be causative on their own right.

    So, I'd very much like to hear you expand on that for a couple of minutes and then also talk about remnant cholesterol. That's something else that's a risk factor that you've been really interested in and you found a really nice paper that showed that as remnant cholesterol increases mortality increases. Actually in that same population, as LDL increased mortality decreased, which I thought was very interesting. Maybe you can just talk about all that and risk factors and ways people can maybe start to look at this if they're worried about their risk factors based on some basic lipid tests.

Dave:    First of all, I have enormous respect for Peter Attia and, of course, he's the protégé of Tomas Dayspring. Effectively, Peter Attia's opinion is basically the same as Tomas Dayspring. They're pretty much of the same line, that LDL-P is all that matters effectively. Again, I try to be as open minded as I can. I have an extraordinarily difficult time believing this to be the case.

    We've effectively gone through quite a bit right here. There's just too many ways in which LDL particles could be up or down regulated for good and bad reasons. While I think it's possible that it can be an indicator for something that's a problem I think that that's in the circumstance in which it's at a higher level for bad reasons.

    That makes for good segue to come into remnant cholesterol. So, we already kind of discussed remnant cholesterol when I was talking about how they start out as VLDLs and ultimately become LDLs from the liver. VLDLs and IDLs are themselves remnant cholesterol particles. Now, this is a natural part of the system to generate these to have them drop off their energy and then ultimately turn the LDL particles.

    Here's an important distinction. VLDLs and IDLs being at higher quantities inside the bloodstream at the point of getting a blood test, obviously, showcase that they're not succeeding dropping off their energy. Therefore, it seems very indicative of what is very commonly an insulin resistance state and some larger level of metabolic backup that often relates back to metabolic syndrome.

    Therefore, I'm not at all surprised that you would see a higher correlation with a higher number of these particles because they're not succeeding in the first part of their job with a deleterious state and especially with atherosclerosis. That's why this whole journey started when hearing these things like that Peter Attia and Tom Dayspring said is I was like, okay, could I be wrong about my systems theory and that low triglycerides with high LDL-C is still a bad thing because high LDL-C is a bad thing?

    I went on a journey, and I'm sure you yourself because you've gone through a lot of this research, are familiar, I can't find one paper that shows people that have very high LDL-C but very low triglycerides have high atherosclerosis. I haven't found one. Not one yet. I have found drug studies that manipulate your LDL-C and your triglycerides and your HDL and then they will come to conclusion from that.


    I'm going to get my soup box for a second. I can't stand when they do things like make a drug trial that inhibits cholesteryl ester transfer proteins and that results in lower LDL and results in higher HDL and then they had deleterious outcomes and then they say, "Oh, this shows that higher HDL isn't actually that cardio protective." I want to tear my hair out. I would be like, look, give me a study that doesn't involve drugs or genes that shows a population with higher HDL is in a bad way that also accounts for things like alcohol and then I can trust your numbers.

    Anyway, sorry, I digress. The remnant cholesterol. I don't believe that lipoproteins themselves are the problem. I believe they're a reflection of the broken system. I believe that the fact that they're not remodeling, the fact that they're not dropping off their triglycerides as designed, it's showcasing why they're a part of the system that has broken down in some fundamental way.

    What's frustrating to me is, again, given everything that I told you throughout this podcast, I read these same papers. There are a lot of them on remnant cholesterol. These same papers will then say clearly this is a very atherogenic particle. No. Why do you come to that conclusion? Why can't you? Look, I keep wondering, are these same people who are seeing this as an atherogenic particle seeing ambulances and saying ambulances are clearly a problem with society because I keep seeing them coupled with traffic accidents? No, they're coupled with traffic accidents because they're there to help and they're reflective of a problem.

Tommy:    I do have a slight theory surrounding this which is that in the general population that we're examining, so say the US adult population, the majority have pre-diabetes or diabetes. So, the majority are insulin resistant. In that setting, those particles are potentially going to be problematic. But because everybody has that problem the two immediately track together. But if you took away the insulin resistance problem then it's going to be much less of an issue. But because there's just so few people where that's the case, they're not seeing it in the data.

Dave:    Right. It's one of the issues that I have with statin studies, is a lot of the statin studies don't compare a statin treated population with a non-medicated population. They compare a statin population with another population that's on another kind of cholesterol lowering therapy. You already have muddied waters with the standard population that's being examined. This is a commonality that we're seeing over and over again with all of this data that we get or that's supposed to try to tell us this narrative on lipids.

    It's a very serious problem, Tommy. I find that the vast majority of studies people send me, I find pretty quickly what it is that make the data not clean enough for me and then they feel like my standards are too high. But my standards are high for my own data. I'm very upfront about the things that I can feel like I can make a conclusion on and things that I can't and there's more in the latter category than in the former category.

    To me, that's just good science. Look, if you really want to draw conclusions, if you really want to have a high confidence level and one thing being deleterious or not, then you need to have data that effectively isolated it and showed its true causality. Honestly, there aren't just many things that you can do that with in a functioning metabolism because of how complex it is.

    That's okay. Be upfront about that. Let people understand that there's some things that you can feel confident about, some things you can't. Getting back to remnant cholesterol, yeah, if I had to hang my hat on one marker to say this is a pretty clearly watchable marker for what's good or bad, if I were arm twisted, I'd go towards remnant cholesterol. But the truth is, I don't really like any marker especially a marker by itself as a reflection of good or bad health. I just don't like that.

Tommy:    So, remnant cholesterol is just taking total cholesterol and taking away HDL and LDL, right?

Dave:    Yes. A lot of people can be forgiven for believing that total cholesterol is just HDL cholesterol and LDL cholesterol added together because we heard over and over again that there's just two kinds which is HDL and LDL. That's not actually true. If you subtract LDL cholesterol and HDL cholesterol from your total cholesterol there should be a remaining number. That remaining number is remnant cholesterol.

    In American units, you usually wanted to be 15 milligrams per deciliter or lower. If that number is 15 or lower, you're probably in pretty good shape. A little bit higher, it's a little more medium. There's actually on my site on, there's a report tool where people can feed in their numbers and it will calculate their remnant cholesterol and then also calculate their AIP, atherogenic index of plasma, which is a whole other podcast.

Tommy:    Yes. We did mention it briefly on the podcast I did on cholesterol.


    If people are using, if people can, your tool, freely available on the website, people should definitely use. If they're using our blood chemistry calculator we also calculate remnant cholesterol plasma atherogenic index, all that stuff.

Dave:    That's great.

Tommy:    People can use those things and start to figure that stuff out. This has been brilliant, Dave. Thank you so much for taking the time to answer all of my questions and maybe we'll talk offline about some experiments we can do. Can you quickly tell people where they can find you, where they can find out more information about your work, where they could maybe hear you speak and all I'll let you go?

Dave:    Sure. Well, actually, I get a lot of engagements for speaking. I've got the Low Carb Cruise that's coming up. I've got Keto Con, Keto Fest and a chance I may be going to one of the Australian conferences this year. You can find me, of course, on and I'm pretty active on Twitter. It's just @DaveKeto.

    Overall, if there's anything that I really want to leave with any audience whether they're low carbers or not is that the lipid system is extremely dynamic. My research is definitely indicating this quite a bit. Even if you are not pro low carb or not, even if you're pro statin or not, you should at least know just how much these numbers can change from day to day based on a short-term window of what you've been eating in the days before.

    Hopefully, more of this will come to light in general for people before they start making judgments on lifelong medical therapy based on a single cholesterol test. I always want to be sure everybody gets that.

Tommy:    Yeah. That's the perfect way to summarize. I'll just thank you again for your time. You've been very generous and I really think that people are taking a lot out of this and would have learned a lot. Thanks again.

Dave:    Thanks for having me, Tom.

Tommy:    Cheers.

[1:01:44]    End of Audio

blog comments powered by Disqus

Register for instant access to your FREE 15-page book, What We Eat

© 2018 nourishbalancethrive