Diet Doctor Podcast #8 — Dave Feldman


Welcome to the DietDoctor podcast
with Dr. Bret Scher. Today I’m joined by Dave Feldman
from cholesterolcode.com. If you’ve been around low-carb circles
for a while you’ve undoubtedly heard of Dave
and all his amazing self-experimentation and what he’s done to sort of further
this concept of lipids in a low-carb lifestyle. And are they different than lipids in general
and the information we have specifically when in LDL cholesterol
and LDL particle numbers elevated, what does that mean? And that’s sort of a big cornerstone
of a lot of his research and controversy. And we have to emphasize
there’s no answer to this. We don’t know that the LDL elevation
is harmful or not harmful in this specific subset and that’s what Dave wants
to explore so much. So please don’t construe any of this
as medical advice, don’t make any decisions
based on this podcast. This is an exploration of what we know,
what we don’t know and how we can further this. Now a quick intro –
cholesterol is all over our body; every cell in our body needs cholesterol. We typically talk of LDL cholesterol
as the bad cholesterol, HDL as the good cholesterol. Those are misnomers –
cholesterol is just cholesterol. It’s only good or bad
depending where it ends up. But what’s important is when we talk
about LDL-C or LDL cholesterol, that’s the total amount of cholesterol
contained in LDL lipoproteins. When we talk about LDL-P or LDL particles
that’s the number of lipoproteins. The way you can think about it
is cars on the road. You’ve got a number of cars,
that’s the LDL-P, and then you’ve got the people in the cars
that’s the LDL-C. So you could have 100 people in one car
or you could have 100 cars with one person. They’re each going to have an LDL-C of 100 but the LDL-P is going to be either 1 or 100
in those scenarios. And that’s important because the literature
does support that LDL-P, the particle number is a better marker than LDL-C. Now frequently people are going to use
them interchangeably and for most cases that’s okay, but I wanted
you to understand the terminology because we throw that around a lot. We also talk about V-LDL,
the very low density lipoprotein, which can be degraded into IDL,
Intermediate Density Lipoprotein, and then down to LDL and that’s
an important path that we talk about. And we’ll explain more about that. As well as chylomicrons which are basically
a collection of triglycerides coming straight from your intestines and being packaged so that they can
circulate in the bloodstream. I think that’s probably all the terminology
you need to know to begin with. Hopefully we’ll explain everything else
as we go. But as everything with Dave Feldman, this is a whirlwind of discussion
with some great analogies and really focusing
on this very important topic; how we can learn more about the subset
of people and what the LDL means. So please enjoy this interview
with Dave Feldman. Dave Feldman, thank so much for joining me
on the DietDoctor podcast. Thanks for having me, Bret. Well, you have had a whirlwind
of a couple weeks I’m sure, you were recently
on the Peter Attia podcast which has caused a lot of controversy,
a lot of comments and I want to talk about that. But I think for the benefit of our listeners
we actually should rewind a little bit and talk about sort of the building blocks
of what got you to this point. And I know, between the two of us,
we love this topic and we could talk all day on this topic
so I’m going to try and summarize. And you fill in as you think appropriate
for this summary. And the basics is talking about LDL cholesterol,
LDL lipoproteins, as an energy model that in people who are low-carb
and burning fat for fuel and have potentially an increase level
of their LDL particle number, that one hypothesis could be it is driven as part of an increased usage
of the body of fat energy and LDL lipoproteins
are a byproduct of that. And as part of that you’ve shown
with your inversion pattern that you can dramatically change
your LDL particle numbers and your LDL cholesterol numbers either by fasting for three days and making
the LDL number go up dramatically, or almost binging on high levels
of saturated fat and dropping the LDL dramatically, things that really goes
against conventional wisdom in medicine. Did you say that’s an adequate summary
of the building blocks of your work? I think you did a very good job
just stating it right there. I mean I would even go to say
a little bit more simply, LDL cholesterol and LDL particles
are very much influenced by the larger energy metabolism. And much of my work especially in using
myself as my own guinea pig is to find what those levers are, what are the levers that modify
this energy metabolism both for fat-based energy
and even for glucose coming from carbs. The more that I find I can play
with these levers, the more I finally have the downstream
result of altering LDL particle count and LDL cholesterol that of course
is along with it. So let’s talk about the big picture
for a second. Why is this important?
Why do we care? Well, obviously a lot of people care about
LDL cholesterol the so-called bad cholesterol. So we right now have a very large industry
that of course sells medication in order for us to bring it down, because the assumption is
the lower the cholesterol the better. And I think you and I’ve had several
conversations outside of this, for where we talk about, you know, we try to be cautious to not say
that we know for sure whether the so-called lipid hypothesis which is the hypothesis that the more
serum cholesterol you have the more cholesterol in your blood, the more at risk you are for heart disease
in any context, that we know for sure
whether that’s right or wrong. Because as right now I would say
for the record I don’t know for sure, but this does help to expose a context
for which it may not only be not a concern but possibly even be an official
to have higher levels of LDL particles because it may be reflective
of a healthy lipid metabolism when you take other lipid biomarkers
into account along with it. And I think that’s the key,
that we’re looking for a specific subset, we’re looking at a specific subset
not the general population and that can be a definite problem
that we run into in medicine as a doctor in medicine guidelines,
in medical research, that we try and apply our findings
to the entire population. Because we have a study
with 10,000 patients, therefore it must apply to everybody. And sort of what you’re saying
and what I agree a lot with is, “Well, hold on a second! Maybe this
doesn’t apply the same for everybody.” So one of the main theories about why it would not apply for someone
who is low-carb and burning fat is because our energy utilization
is different. So talk a little bit about that. Why would we assume it would be different
in someone who is low-carb? Well, certainly a lot
of the existing literature is focused on people
who are on a mixed diet, such as having both carbohydrates and fat
or even being carb centric, where they have primarily carbohydrates
fueling them. So now we have a circumstance in which there are people who are
specifically trying to be powered by fat. Both the fat they get in their diet
and the fact that they have in their bodies. Well, this is pretty relevant because as you learn more and more
about these lipoproteins, lipid caring proteins, the more you find that actually
it’s a fairly complex system, but much of its primary purpose is to shuttle around these fatty acids
that we’re powered by. And once you learn that, then you find out
something pretty fascinating, which is these two larger classifications
of these boats that kind of carry around these fatty acids,
our chylomicrons, and VLDLs very low-density lipoproteins. If you’re getting direct delivery
of fatty acids to your cells from one of these larger boats, you’ve got to know that the first one,
the chylomicrons, they come from food you just ate. And that the second one VLDLs
they come from your liver, primarily from fatty acids that came
from your adipose tissue, your body fat. Now why that’s important is it explains
this inversion pattern you just mentioned where if I’m going to have
a whole lot of dietary fat, I’m just going to eat a whole bunch of fat, whether my body has less reason to release
as much of these fatty acids from my own fat stores,
my own body fat and therefore less and less gets made from
the liver and packaged into these VLDLs. That matters because these VLDLs,
when they drop off their triglycerides, this form of fat-based energy, will remodel to LDLs. And the LDL particles are the ones that
contain the so-called bad cholesterol, LDL cholesterol, which is a little bit
of a misnomer as you well know. Of course. There’s really only two kinds
of cholesterol if you will, two flavors there,
esterified and un-esterified, but basically speaking that’s the cargo that ride-shares with these triglycerides
in those boats. And so then the question becomes, “Is there a good reason and a bad reason
for why you may have high LDL cholesterol?” And I myself think that
that actually is the case. I think there can be bad reasons
you have high LDL cholesterol, but typically speaking the big hint is
whether or not you have high triglycerides, meaning you’re past
your personal fat threshold, so you’ve kind of maxed out the amount of parking there is for the fat
that you can stash. And for that matter
if you have low HDL cholesterol you have you have an engagement to where it showing that you’re not utilizing
your fat as well and a lot of that comes out
in the literature as well. So the phenotype we’re trying to look for
as what they call it, is those people who have
high HDL cholesterol, low triglycerides and high LDL cholesterol. And thus far as you know
I’ve been very active in social media trying to hunt down
any and all studies I can that can show that that phenotype is at risk
particularly for cardiovascular disease. Because it stands to reason that they should
be if the existing lipid hypothesis is true. There should be no exceptions really
if all elevated LDL puts you at risk. So what you’re saying is if you’re not using
fat for fuel you store it in your body,
more spills over in your blood and your triglycerides
are therefore higher. And that’s a measurement that you can use
to say which camp are you in. One that has elevated LDL
for a bad reason or one that has it
potentially for good reason. Now when you had this interview
with Dr. Peter Attia who obviously knows a ton about lipids, one of the main issues– well, there were
a couple of different issues I guess. And he was not a big fan
of your energy model. So let me ask you first, did you think
it was disproven in this interview? First of all I have
a tremendous respect for Peter and I’m really happy they had me on. Disproven, no, I don’t feel
like it was disproven and I actually had a number of responses to a number of the points
that he brought up from before. He had brought up kind of three
in particular that he felt was disproving it. The first one was mass balance. In particular he was interested in where were those cholesterol molecules
coming from or going to or whether they were drawn
from other pools. But as I brought up a lot in the podcast
there’s also a fourth one, which is how much they’re recirculated. The liver can recirculate
as many times as it wants any particular cholesterol molecule. So if you were to look
at a cholesterol molecule in an LDL particle you don’t know whether or not
it was synthesized that day, or if it was recycled yesterday,
or if it was recycled a week ago. Or a month ago…
there is no way to know. Unless you get
into more complex tracer models which of course is something
I would be interested if they would do, but that’s how we know right now
that there isn’t as much synthesis. Also there’s been a recent paper
by Volek and Phinney that further emphasized that a lot
of these lean fat adapted athletes weren’t actually having
high levels of synthesis as he was… as Peter was postulating
at the beginning of the podcast. Yeah, let’s talk about that paper
for a little bit. So this came out recently,
Phinney and Volek study, they looked at 10 competitive
ultra endurance runners and compared
the 10 low-carb high-fat athletes to the higher carb low-fat athletes. And interesting what they found was
LDL in the low-carb athlete of 161 versus 88 in the high carb. HDL of 102 versus 64,
triglycerides of 63 versus 70 and in the low-carb athletes
the LDL were the larger particles and HDL were the larger particles. And interestingly what you’re talking about they measured these other sterol markers,
desmosterol and sitosterol, basically markers of synthesis
or absorption, and you would expect
that they would have been through the roof with an over twofold rise in the LDL,
but they weren’t. So how do you interpret that? I definitely interpreted it
as greater recirculation. Let’s use a racetrack for an example, let’s say that you were to ask what
the mass balance of an empty racetrack is. And it starts at zero. But now you go ahead
and add five racecars. Well, the mass balance has changed
to where five more have been added. So we produced five more
to add to the total. Now if those five racecars
make a circle around the track the mass balance actually hasn’t changed,
they’ve just circled the track. And until they leave can we say
that they’ve been distracked? Well, if you go ahead
and add five more racecars, you’ve increased it five more but the five that are already on the track
are still added to the total. Now there’s 10. And they can continue making loops
as many times as they need to. Well certainly for those people who are
going to have higher levels of LDL cholesterol or LDL particles
in their circulatory system there’s some amount of synthesis
to add them. But as far as constantly recycling them
they will continue doing it as many times the body seeks to do so
and rather effortlessly in comparison to actually
constantly making them and degrading them or destroying them
in some fashion, sending them back out the other side. There is no good reason for the body
to just go ahead and re-create cholesterol especially as expensive as it is
to do so biologically. So I guess one way
to think about it then would be the timeframe
in which these lipoproteins and the cholesterol concentration
increased, because if it happened in a day, maybe you need to dramatically increase
your absorption or production for that to happen… If it happened over six months or a year that can be more of a slow minimal increase
in production, but a continuous recirculation. Is that one way to think about it,
or am I sort of saying that wrong? For me, again I’m coming
from network engineering, it’s all a pool. And how much of it is a pool
that’s in circulation versus how much is in the liver
at any given time? The only thing that really matters
at the end of the day is just how much is needed and
whether or not that’s easily achieved. And as far as this circulatory system goes, I mean literally is a circulatory system,
it is like a racetrack, a fairly complex one, but that’s basically how it’s going to work. Okay so let’s take your inversion pattern
for a second. With those three days what would you expect the absorption
and production measures to be? Would you expect those
to be significantly different? I was talking about a hyper-responder
on a low-carb diet versus somebody in a typical diet. Right, who does your three-day protocol and sees their their LDL
increase dramatically with fasting? Or the other way around. Who sees they fast and the LDL goes up? Correct, if they fast
their LDL goes up. So would you expect
the production markers in the blood to be much higher
under that circumstance? Or again is still recirculating
because it’s coming from your fat stores so you would expect a change
in your production or absorption? If we were to set a hypothesis I would say it’s a little of the former
and a lot more of the latter. I think that there may be some degree
of an increase in production, but I highly doubt that it’s a 100% increase
in production to match what we then see
as the increase in total LDL particles. I think more likely we’re going to see
a greater recirculation against the pool that the liver
already had resident. Now to one degree the liver is also making
use of the cholesterol for other purposes, going on in the bile or being used
for sex hormones and so forth. All of these things are of course
a little bit tricky to track, but the larger question is
will it make more if it needs to? Certainly. When you actually look at the total quantity
and capability of what it can make versus what’s actually
in the circulatory system at any given time, it’s actually quite daunting,
it really has enormous amounts of control. But again I’m still thinking about this
from the mechanistic standpoint of is it worthwhile to constantly make
and degrade it in a rapid rate? For the most part I don’t find that the engineering of the system itself
reflects that. It seems to be very greedy
about its cholesterol. It wants to get ahold of it
and keep ahold of it as long as it can for what purposes it can use it for. And let’s be clear, just because it’s there
does not by default make it dangerous. Yet that’s sort of what the lipid hypothesis
would assume. And even Peter agreed
there are other factors; inflammation, vascular injury,
being two of the most important ones. So what about this concept though
that it’s just a diffusion gradient? That the more there is,
the more it’s going to get in the vessel wall and the more damage
is going to come some? Some would say that will counteract
your hypothesis. I think if it worked that way,
they would be right and this is effectively
what Peter was kind of advancing. That it’s kind of like fire and oxygen
that it’s a– I often like to call this
the accelerant hypothesis. As in sure maybe it takes a spark
to ignite the flame, but the more would you have,
the bigger the flame. However I have yet, and this is part of why I brought up
the challenge that I’ve had on Twitter and that I circulated on social media… I have yet to see what I would consider to be probably the template
of a healthy lipid metabolism. And how it’s reflected
in these lipid numbers, showcase the same gradient, showcase that as people’s level
of LDL-C or LDL peak at higher that there is a correspondingly higher level
of atherosclerosis. Something I was trying to bring up
in the podcast at the time and as I’m sure you’re well familiar with, I have been getting
a carotid intima media thickness test on both my left and right carotid arteries,
they are the arteries around the neck. And it’s an ultrasound which will capture
both your intima and your media together. Basically it’s capturing its thickness. Now I’ve been getting it about every six months
since I went on a low-carb diet and for the first two years outside of a few experiments I’d pretty
much averaged in LDL-C of above 200 mg/dL which is very high and an LDL-P, the particle count,
of over 2000 nmol. And that is something that without
question most doctors… most doctors,
you probably personally know, wouldn’t let me leave the office
without at least two drugs. Absolutely. Statin police would be
right around the corner for me. So during that period of time both my left
and my right carotid arteries regressed whereas there were supposed
to stay the same or for that matter progress
and get thicker with my age. They had actually regressed on both sides. And the left side was actually a bit better
and dropped by around 50 nm. In the right side around 150,
it started out a lot thicker. Interesting. And it wasn’t until I did
a recent experiment where I actually intentionally gained weight
for my research… that’s a bigger story, that I saw it reverse, hit the other direction,
it actually shot up on both sides. And what were you eating
to gain weight in that time period? I was eating a lot of bread
and just a lot of starches. I actually tried to avoid sugar
and fructose in its many forms and for the most part
I just had a lot of starches. So did your HDL and triglyceride
also change in that timeframe? They did. So you were no longer in that phenotype
that you’re describing and that’s when he saw the carotid intima
media thickness increase in your tests? Right. Now as you all know the CIMT is often used
as a proxy in many of these different studies as a risk marker for atherosclerotic burden. Now the reason I was trying
to bring this up with Peter and the reason trying
to bring this up with everyone is that it is commonly known that a risk factor
for having higher and higher levels of CIMT is LDL particle count. Yet here I am in the 90th percentile in what I would consider to be theoretically
a very well-functioning lipid metabolism that sure had high LDL-C but also had very high HDL-C
and low triglycerides. That phenotype continues to impress me
with how many other people I know turn around these incredible
cardio metabolic markers in spite of having
these very high levels of LDL. And in that regard
I do have a lot of difficulty believing that it’s as simple as fire and oxygen because it seems to be that there really is
more to the mechanism itself. Now I’d suggest to touch briefly
on the CIMT test. It has gone a bit of a bad wrap, because it doesn’t have the same
predictive value as a calcium score when compared head-to-head. But the point is that’s a one-time test. Now its greatest value I think
is exactly how you’re describing it. You can quickly measure small degrees
of progression and regression and the data you have there is pretty
powerful to say the LDL is certainly… an elevated LDL is not harming
the test for you and I think that’s pretty powerful. Now of course it’s not outcome data,
of course it takes years in larger studies and I want to get to talk to you
about that a little bit more, about how we could potentially design
a study that would answer that question. So we talked a little bit
about the mass balance and I think we can kind of move
past that one now. The other one though was this concern
about the VLDLs and how they are so low and if you are having a higher production
of VLDLs for energy that they should then be
higher in the blood. I mean wasn’t that
one of the arguments as well? Yes, so let’s actually kind of impact that
a little bit. So we’ve kind of talked about this
a little bit earlier in that VLDLs are those lipoproteins
that can come from the liver with… I would argue is their primary purpose
as to distribute energy from fat and in particular their cargo is coming
from your fat cells for the most part. So with that in mind
after they’ve dropped off these fatty acids then they remodeled to LDL. Now your triglycerides is measured
in a fasting cholesterol test. Are going to be almost entirely
the cargo of VLDLs, because chylomicrons get cleared
relatively quickly. So with that in mind it does seem to be
a rather basic conclusion that if you’re making
more of these VLDLs you should therefore have
more triglyceride cargo. And this was the point Peter was making and actually a point there was further
reinforced by Dr. Thomas Dayspring who also went through the transcript
and made a note of that as well. That if the energy model that I’m proposing
is correct and that people who were
on a low-carb high-fat diet were actually secreting more VLDLs, but in particular this phenotype
we’re talking about, people who are very, very lean, very fit, seemed to have very high demand
and therefore a lot more turnover, that they should therefore have
higher levels of triglycerides to match what is theoretically
the higher levels of VLDL. And this I was trying to explain
on the podcast and I’ll explain once again, this assumes that there is
the same rate of turnover of VLDLs across all of these different profiles
of people, which I definitely would argue against. I think that VLDLs obviously can drop off
their cargo at different rates of speed and I think it has a lot to do with where
you are on your personal fat threshold. So to kind of simplify it a little bit – do your fat stores have
a lot of parking available? Because if they do then they’re going
to pick up a lot of these triglycerides at a fairly rapid rate. And it’s fine because the whole point
of your fat cells is especially to provide fat-based energy to nearby tissues like the muscle tissues
in the same area. That’s why it’s a good thing to have adipose tissue
that’s nearby the muscle tissue to constantly give it resupply. So in many ways VLDLs are just repleting it, they are kind of like the wholesaler,
if you will, that’s providing. I’m stealing that one
from Siobhan Huggins by the way. So if you are a fat burner
and you are utilizing your fat stores, you’re going to have
more fat parking available, so a higher turnover of these VLDLs. So even though more being secreted
you’re still not going to measure more because you’re getting rid
of the triglycerides out of your system quickly. They’re not sticking around in your system
because you utilize it. And the VLDLs are transforming rapidly
to LDLs. So that’s why…
I can’t emphasize this enough, once you get a better sense
of the energy model, it’s hard for me as an engineer to come to the conclusion
that your LDL particle count or your LDL cholesterol count
are truly deleterious to a high degree if in fact your triglycerides are so low, it suggests that you probably needed that many of those LDL particles because if they were originating
as VLDL particles you needed that much traffic to deliver that
much triglyceride back to your adipose tissue and of course to fuel your muscle tissue
and other non-adipose tissues. That makes a lot of sense. Now another thing that was brought up that may have been a new topic
for a lot of people was this concept of a ApoC-3. So all these lipoproteins have proteins
on the outside of them that come with awful names
like ApoB-100, ApoB-48, it’s a little confusing,
ApoC-3 being one of them, but this is thought to be
sort of one of the worst ones you can have. One of the most harmful proteins
you can have, because it increases the residence time
of the VLDLs. It makes them stick around for longer where they can then be
potentially pro-atherogenic. But that only happens really in people
who are insulin resistant, diabetic, have high triglycerides. So how does that even play in
to this phenotype or does it at all? Well, this is where
I was a little surprised that early on in the podcast Peter had said
ApoC-3 typically follows insulin resistance or maybe it was the other way around, that you tend to have less ApoC-3
if you have more insulin sensitivity. But the point being as that obviously
if you can change your level of insulin resistance you’re changing what is probably going
to be the detected level of ApoC-3. And also with that same subject matter I was saying, “Well, I’m actually very
interested if an assay does get generated for which I can track ApoC-3,
I believe it’ll be dynamically determined.” Because I believe again
as I was sort of stating early in the podcast, I believe in something
that has a composite response and that a lot of these things
like the Apo’s, these little snaky bumps you’re talking
about that are on the outside, they’re kind of like metadata. They are actually…
for those people who… I’m going to speak to software people, for these people who are familiar
with the emails they’re kind of like headers. And so headers will tell an email
about where it’s going to go and it’s used with packets and so forth because they’re like address labels
on a package. And in a way what ApoC-3s do is they allow for not only binding
to some other things, but they help to kind of reduce
the likelihood of clearance. Now again if you already start with
the lipid hypothesis being effectively fact, then of course you’ll think ApoC-3 is
just a terrible thing to have on there because anything that increases
residence time is a bad thing. But we find lots of things particularly
in the immunological response of LDL particles that suggest they actually are probably
a lot of good things that can come from having more exposure
of LDL particles in the system. So for example a lot of people don’t realize
that they carry antioxidants, alpha-tocopherol for example, a lot of people don’t realize they can bind
the pathogens to help clear it. So one of the things that– it’s an anecdote,
one of the things that I can’t help but notice, and I was just talking about this
last night dinner, was that since I moved to Las Vegas
10 years ago I’ve gotten the flu three or four times,
I’ve gotten a cold almost every year and I’ve gotten bronchitis three times
in Las Vegas in the desert of all places. After I’d gone keto I have not gotten
any of those. I was excited at one point because I thought
that I might be getting sick and immediately went
and rushed to get blood work because I had not once gotten some sick
blood work and I wanted to see the difference. You are one of the few people
actually excited to get sick. That says a lot about you Dave. Absolutely but virtually anything
that was infectious disease related, not even like a persistent cough…
I just didn’t seem to get anything. And I tend to find this is reported a lot by the Facebook group that we have
for this phenotype, which is you know I typically refer to
as lean mass hyperresponders. They often report just getting sick less. Well this happens to match those people
who have the genetic disease that can result in higher LDL particles
known as familial hypercholesterolemia, which is you know is a genetic disease that can result in higher levels of LDL
particle count, therefore higher LDL-C. And they have of course seen
many different studies and it’s been suggested that this was
an advantage before the 20th century when there was more risk
of dying, mortally from infection. And so this was one of those things
where if you had it, it was actually a little more
than immunological benefit even if as it was assumed there would be
a greater risk of cardiovascular death. And in literature today
at least what literature we have, which be honest is not great, suggest that if you’re not one of those people
who has premature cardiovascular disease that you actually do get some benefits
possible for longevity which could be related to hypothesis, decrease risk of infectious disease
down the road. And it was always curious me
how we can identify those people who are going to be a greater CVD risk
with FH and those who don’t, because it’s not clear across the board. And that brings up this concept
of Mendelian randomization or the genetic studies
that look at genetic reasons for higher LDL associated with increased
cardiovascular death risk and genetic reasons for low LDL associated
with decreased cardiovascular risk. And people will like to point to that to say, therefore it proves
high LDL bad, low LDL good. But again it doesn’t necessarily address
this specific circumstance. So why again with this circumstance
be different and not necessarily correlate
with those Mendelian studies? Actually I am glad you brought this up because this is probably
one of the most interesting subjects I think in that from a distance it seems like
such an excellent piece of evidence for the lipid hypothesis. I mean how great would it be if we could just randomly pick
a bunch of people in the population who would just magically have
higher levels of LDL particles and therefore higher levels
of LDL cholesterol and then just see with their risk of death is
especially by cardiovascular disease. And that’s what the Mendelian
randomization trials are attempting to do. Peter Attia actually had a great illustration
of this as kind of a hypothetical in his series, The Straight Dope on Cholesterol. And he describes as a wand, like what would be cool
is if we could get say 100,000 people and wave a wand and then those people
would to have more LDL particles in them. And then have another cohort
and it’s like less than that, and another cohort that’s less than that
and then just track them lifelong. And that’s what Mendelian randomizations
are attempting to do. But early on I realized that there were
a number of issues that I was having with the SNP’s,
that’s the single nucleotide polymorphisms that they were tracking and a lot of them had to do with…
the kind of term is lipid malabsorption. Effectively if a cell, any cell in the body,
needs what it needs, it has something known as receptors, receptors that can extend that can
effectively actively pick what it wants. In fact the way these receptors bind
is almost like a key in a lock. So with that in mind it seemed
pretty obvious to me that if you are depriving a cell
of its capability to get what it wants… so if I were to say, you get the cells that
don’t have a receptor, that can either gather lipids
or lipoproteins… You are inhibited from that. Especially if they are endothelial cells,
which endothelial cells line your vessel walls. Well, then my concern is
that could cause a dysfunction. And again without question
this is somewhat theoretical but even so, why not just go ahead and exclude those? Why not focus on those SNP’s that result
in high LDL particle or high LDL cholesterol but don’t actually impact
the health of the cells specifically, particularly in regards to lipids
or lipoproteins? Now as it happens there are actually
such diseases that result in this. In fact one that I’d like to talk about
is glycogen storage disease. The reason I’d like to talk about this one
is because glycogen storage disease is… there’s many different versions of it, but particularly one version for which
you have low levels of glycogen, there can be very high levels of lipids,
in particular lipoproteins, and yet they’re dumbfounded because these people don’t have
high levels of atherosclerosis. In fact study after study they tend to find
that they have they call it an athero-protective– I forget exactly how they state it, but effectively they somehow have
a protective measure and they’ve tried to isolate with special
mineral or nutrient of some kind that they must be higher in levels of, in order to protect against these high levels
of lipids that they have at the same time. And naturally the first thing
I gravitated towards was well maybe this time around these people don’t likewise have these SNPs
associated with lipid malabsorption. And therefore why not go ahead
and find as many things as you can like glycogen storage disease, whether
it was for or against the lipid hypothesis, as long as it doesn’t impact
the health of cells and their ability to uptake lipids
or lipoproteins. Interesting, but if they have
some protective mechanism it wouldn’t necessarily be the same one
you’re talking about, because it’s not that their energy demands
are higher or actually since they aren’t burning
that glucose as well, yes, their energy demands are higher,
I take that back, right. So they can’t burn glucose,
they can’t use glucose as efficiently for fuel so maybe they are transitioning
to using the fatty acids more so their energy demand is higher
so they do sort of fit. Now I got to remember,
I remember looking at this, but do they have low triglycerides
and high HDL? No, they tend to have
higher levels of triglycerides. But this doesn’t surprise me that they are still having
lots of carbohydrates in spite of being effectively being pushed in the direction of having
a high-fat metabolism. Right, because the treatment generally
is very frequent carbohydrate meals, so you’re constantly replenishing
your glucose supplies since you can’t metabolize it as well,
you need more in your system. And that’s why I theorize that actually– I realize this may be a controversial position
but I’m just following the logic as it follows, that those people
that have glycogen storage disease may benefit substantially
from being on a low-carb high-fat diet. Right, absolutely. So at this point I want to take
a slight detour here to talk about the sort of the problem
with all this though, because one of the points
I think Peter made very well is I have a patient sitting in front of me and
I need to know what to do with that patient and there’s a lot of theory here,
there’s a lot of hypothesis and it makes a lot of good sense to me,
but we don’t have the proof. And one of his big points was I need to go
where the majority of the evidence is and where I think the highest probability is. So talking in terms of probability
and address my patient that way. So did you see a fault in that type of logic
in the way he approaches his patients? You know, I want to answer
with a little bit of an anecdote. While I was in college I had a friend,
we were really close and he was very much a pro-science guy. Whenever we would be doing tests he would talk about much he was an atheist
and how I can believe all this religious nonsense etc. but when midterms or finals came around
he would bring his lucky rabbit’s foot. And I would get surprised by this
and the first time I sort of let it pass. The second time I said,
“Why are you bringing that? I thought that you didn’t believe
in anything superstition related?” And he said, “Well I could use
all the help I can get.” And I think that there are
a lot of people understandably so, we’ve grown up with this,
our whole lives, who would say, “Look I hear you,
it’s still theoretical. All I know is I can’t help but recognize…”
and this is true, “…that the vast majority of doctors
still very much believe in this.” And so for me it’s a little bit difficult
to overcome, a lot of people will just say, “I’ll feel better
just having my LDL lower.” And for what it’s worth we could be right
in our cautious optimism and it could still actually be better
for people if they can bring their LDL down if it actually makes them less stressed. That’s a good point. It’s an important point because at the end of the day
if you don’t feel comfortable with where things are at metabolically, but you can still feel good at a different level
by making some changes, a lot of times just your overall peace of mind
can be just as valuable as anything else. And I think that that’s something that even
on the low-carb side of the fence we should take note of. So I will have a little bit of an issue again being somebody who likes to look
at just the hard numbers as they stand and somebody is saying,
“You don’t know for sure on this phenotype “even though you have studies
that you can show that show it’s low risk “when I feel like so many of the doctors
I know will insist that it’s dangerous,” it still makes me feel bad that that’s going to do it enough
for me personally, for Dave Feldman. But I can totally understand if other people
want to take a different route if they feel and judging the evidence for themselves
they come to a different conclusion, and not only can I respect that, but at our blog, cholesterolcode.com
and the Facebook groups I want everybody to respect
their decision to do so. Everybody’s on their own health journey
and that’s super important to emphasize. So while this is very exciting,
it’s absolutely true. You can say the evidence isn’t quite there, but by the same token
I’m going to push back a little bit to say that the evidence is there
on the other side. Well, and you can look back
even to the Framingham data and say that low HDL was more predictive
than high LDL and certainly when LDL was elevated
with an elevated HDL, there was a lower risk in that situation. The pushback would be, well those were
the smaller patient population in the group, but still they were not having
the same degree of atherosclerosis and someone sad it was LDL-C
and not ApoB, but we don’t have evidence
to say it’s different. We don’t have evidence to say with the high LDL and high HDL
there is an increased risk. That doesn’t exist either, does it? No, and look, let’s call a spade
a spade here. It’s not just about whether or not
the evidence as it stands right now. The two studies I was pointing to
substantiate the theory that I’m advancing that this can in fact be
a healthy metabolism and that they are at low risk when
they have high HDL-C and low triglycerides. I’m surprised that there’s not as much
genuine interest in the field right now to test this immediately. I mean certainly it’s one of the things
I was hoping for going on the Attia podcast, it’s certainly something in which I’m pinging
lots of pro lipid-lowering experts. I’m saying, “Help me disprove my theory,
help me do that.” Help me get ahold of MESA data,
help me get ahold of PURE data, help me get ahold
of Women’s Health Study. I mean it would be great
if I could get LDL particle count, but heck even LDL-C, that would be great. However as you well know there’s a lot
of firewalls you got to walk through to get to these larger data sets. But if the evidence has been there
this whole time, shouldn’t we take a look at it? So it seems we should have
some sort of study to look at this. And do you think a lot of the people
who are ingrained in the lipid hypothesis feel this isn’t even worthy
of their attention and that giving it attention
will only sort of fuel the fire even more? You know, I kind of turn that around if you feel as a lipid-lowering expert,
somebody who firmly believes in lipid hypothesis, that there are a number of people
that are not acting on this advice because they hear this energy model
and it feels somewhat compelling and that perhaps there is something
to be said for having the higher HDL cholesterol,
the lower triglycerides as being beneficial, then if I were one
and I believed the lipid hypothesis, I would be all that much more
of an advocate in trying to find and unearth this data
to disprove it. Because after all I’d rather disprove
that it could be impossibly beneficial, that in fact having high LDL
as bad in any context. Because that is in fact
what the lipid hypothesis is. Right, that’s a very good point. I hope we’ll have something soon because this question I think does need
to be answered clinically. Because the discussions about the specifics,
about the different theories back and forth can get very confusing and it doesn’t address the fact that this
is a different physiological set up, this is a different way of being
where we are using a different fuel. So a lot of the arguments and part evidence
doesn’t apply and that’s what can make this so confusing. Absolutely and I mean at a minimum I don’t know how many reasonable persons
can say, “Let’s find out.” Well, I’m sure we could talk on this
for a few more hours, but let’s not. Instead I want to ask you
about one of your more recent experiments and you are no stranger to experiments, you and Siobhan are the king and queen
of self-experimentation in lipid manipulation, so tell us about this latest
bread experiment that you did. Well to be fair we like to call
the tandem drop experiments the first time that my colleague
Siobhan Huggins and I– she by the way helps to run
the cholesterolcode website with me, we both did a different version
of an experiment at the same time. For her she had a keto shake
generously provided by Ketochild by the way that had dropped her LDL cholesterol throughout the entire time
she’d be drinking it, because she was invoking
the inversion pattern by having a very large caloric amount of it. I think getting somewhere close
to like 5000, 6000 cal a day. She is not very tall like I am
so that’s like a fairly large amount. It’s a lot of calories for her. Yeah, definitely a lot of calories. And sure enough her LDL-C dropped
like a rock and you can read more about that
on the blog cholesterolcode. It’s actually tomorrow, at the time that we’re recording this
I haven’t revealed this yet, but tomorrow I’ll be revealing that in fact
I was on a different diet at the same time and I guess I’m revealing this to you
for the first time. I actually went the other direction. I dropped off of keto
and went high carb low-fat. And I intentionally chose– I wanted
to demonstrate the energy model by choosing food that nobody would
ever provide in any diets or recommend. So it was white bread
and processed lean meat. Wow! That was the only two things that I ate
over and over and over again. For how long? I want to say 10 to 11 days. How did you feel? Horrible. I can’t believe you do this to yourself. I’m glad you do,
but I can’t believe you do this to yourself. Now it gets even more interesting, at some point I’ll get a chance
to post the graphs, but I have a continuous glucose monitor
match this time around. So you can see
the first continuous glucose monitor like almost a straight line
while I was on keto. In fact you couldn’t even tell when I ate
meals because I had almost no response. And as I like to joke on this experiment
it was the Alps. You could actually just see these huge rises
in my glucose and honestly the CGM I have to say that is a really powerful device
continuous glucose monitor, because it can pick up your rises and falls
in glucose especially postprandially. And it made me not want to fully engage
the experiment as I had planned just because watching this moment
after moment was so daunting. Well sure enough I saw the fastest drop
in my LDL cholesterol, but before I talk too much more about it,
I have to insert a quick warning. I knew that this would be a drop because I knew I was switching
from a fat-based metabolism over to a carb centric metabolism. And I knew that also I would be filling up
my glycogen stores which is also a part of my theory as to why it is that the body would feel
less need to mobilize as much fat for fuel, but on top of that I’m inducing
a state of hyperinsulinemia. Needless to say, Bret, as with the prior
experiment I did with the weight gain, I want to tell any of your listeners
who is listening right now, this is actually not something
I’d recommend for that matter. It’s something I discourage anybody
from doing. As I say with many of my experiments,
I do them so you don’t have to. This is absolutely one of those. Thank you. I don’t want anyone to go, “This would be
a great way to lower cholesterol.” This is not a great way to lower cholesterol
in my opinion. But that said it was one
that I felt confident would work, which is why I actually filmed a video
that I posted to YouTube privately and I’ll be actually making public
by the time I do the public presentation of this, where this is exactly what I predicted. That not only would it be the fastest drop
that I’d ever had with my LDL cholesterol, but then I would actually
break my own record. And sure enough I did,
I brought it from around 300 mg/dL and in seven days
I brought it down to 82. Oh my goodness. 382! That’s right. So if someone who is firmly rooted
in the LDL hypothesis saw this, they would say, “Fantastic!
You have improved your health dramatically.” Yet you felt terrible,
your blood sugar was spiking like crazy, your insulin was through the roof…
Is that really improving your health? Well, we’re leaving out
two other markers. What are those two other markers
I like to look at in the lipid panel? I don’t know, I can think of a lot though…
CRP is that one of them? HDL cholesterol and triglycerides. My HDL dropped to around 48,
I had to double check. From? From the 60s. But my triglycerides
were a lot more interesting. They bounced up above 200. Really?
Yes. So from like the 60s up to 200!
Fascinating. It almost was a flip-flop, it’s almost as if my LDL-C switched
with my triglycerides. Because my triglycerides
were nice and low when I started and my LDL cholesterol was very high. It then got all the way down to 82 and my triglycerides have gone much higher. And I’ve seen many different YouTube videos
and read many articles that say once your LDL-C is nice and low particularly if you can get down there
around like 70s or below that you’re basically bulletproof. That you don’t have to worry
about cardiovascular disease at all. And that even if your triglycerides are high
and your HDL is low, it doesn’t matter. Again no oxygen for the fire. Right. But as I’m sure you and I both know Tim Russert had actually
a very low level of LDL cholesterol unfortunately when he passed. He had very high levels of triglycerides
and his HDL was not in a good shape either. So for me again I want to look
at the whole picture, I want to look at the whole
lipid metabolism and I want to see in particular those people
who have high LDL-C when matched with high HDL cholesterol
and low triglycerides. Obviously I would rather not be on the profile that I was on
at the end of that experiment. Right. Just as a quick aside,
did you measure LP(a) with this experiment? One of the issues was because– in order for me to be able
to track this day after day I actually wasn’t able to get a lot
of blood test due to the traveling, because this was over the time
in which I was in LA at the National Lipid
Association Conference. So it may have been that I got it
in the two lab test that I got throughout, but actually most of these were tracked
on a cardio check, which is a device from PTS Diagnostics. It’s not as good as a lab draw, but I then did a further lab draw
at the very end of the experiment as well. Now one of the more fascinating things
is even though I brought it that low, it started to rebound toward the end. My LDL-C actually started
going up a little bit as my triglycerides started
going down a little bit. And this actually makes somewhat sense
because of course this will be impacted, LDL-C gets impacted
by the Friedewald equation. But the larger question is would I continue
if I wanted to keep playing with it to see if I could get my LDL-C
lower and lower? Probably could’ve, but I didn’t want to. By that time I was very unhappy
with the experiment and very happy to be leaving it and
getting back to my ketogenic diet. You must have a very patient wife
to deal with your mood swings as you go through all these
dietary changes. This is a tough episode to wrap up because
we’ve talked about a lot of different things but what you think
the next step needs to be? Well, obviously I’m a bit biased, but I do feel the discovery of this phenotype
with lean mass hyperresponders may hold the key to really unlocking whether there really is a strong validity
behind the lipid hypothesis. And I think it may actually be
one of the most important things for the low-carb community to determine. Again I’m biased
but I feel pretty strongly about that. If there are a number of people who we see
who get this phenotype and they feel just amazing– I mean this Facebook group,
I can’t emphasize this enough, the most common statement
we get incoming is people saying, “I don’t know what to do
because I feel so great. “All my markers look fantastic
but my doctor is freaking out. My LDL-C is so high…
what I do about this?” And we don’t really have
a short easy answer, but I can say anecdotally
and with some reservation that a lot of these cardio metabolic markers
coming back and looking good is part of what fuels my optimism. But what we really need is a study,
a long-term follow-up study on those people
who are lean mass hyperresponders to actually see whether or not
this is a risk factor. And while I’m cautiously optimistic I’ve got to hold
my strongest reservations out until I can get to that point. And so I think, an important point
to lead this with is we are not advocating, neither you nor myself are advocating
to ignore LDL, we’re not saying that this
is proven to be safe. We’re saying this is a fascinating topic
worthy of a lot more exploration. And if you are in this phenotype
you can learn a lot online, but the ultimate decision has to be made
by you and your health care provider about what to do about this. Because it still is very individualized, even if you are in this phenotype
that we are talking about with a high HDL, low triglycerides, that still has to factor in
your other risk factors, your family history,
what the rest of your health is like and if you’ve had any other tests like calcium scores
or carotid intima media thickness test. So still very individualized
so don’t take this as any medical advice. This is just exploring a fascinating field which you have done a tremendous job
in spearheading and bringing to the limelight and I’m very thankful for you doing that. So where can people learn
more about you? Obviously the blog cholesterolcode.com. We try to answer
as many questions as we can for people who bring their labs there
or have questions themselves. There is also a Facebook group for this
phenotype lean mass hyperresponders. You can actually do a search
for lean mass hyperresponders on Facebook. And of course we’re very active on Twitter.
I’m @Daveketo. And you can find Siobhan Huggins
also on Twitter as well. And overall we’re just very happy with how many people have helped
to carry us through this journey, as we also have patrons where people
have been very helpful with the hard costs of our blood tests,
so I got to give a shout out to them as well. It wouldn’t be possible without our patrons
being able to give us support, to each of our patrons. Dave Feldman, it has been a pleasure and I look forward to hearing more
about more experiments that you’re doing and helping further this field
so thank you so much for joining me. Thank you.

14 Comments

  1. Focus the camera on the person, not on the mic. A shallow depth of field looks cool, but it is not cool when the subject is not in focus.

  2. Great interview, amazing research! Hopefully in the near future we will be able to make some concrete affirmations on the lipid hypothesis. Well done!

  3. It makes a lot of sense!

  4. 38:45 Great story about the lucky rabbit's foot. "I can use all the help I can get."

  5. Great interview

  6. Absolutely fantastic job all the way around! One good thing that came out of the Attia podcast with Dave is that it has sharpened him, this is very evident and very good. Bret does a superb job with interview style, tone and anything else that you can measure about someone who does a good job with interviews/podcasts. Just really really good job to you both!

  7. I have never heard so many big words that I don't understand in a single sitting….and I have a pretty decent vocabulary. lol

  8. great and interesting  interview. thx.

  9. Very interesting. I would like to see the benefit of high LDL for APOe4. I have two copies and my LDL runs high. I have been IM fasting for one year. Periodically I fast for 2 to 4 days.

  10. Dave Feldman is one of my favorite researches. He knows way more than a lot of the cardiologists!

  11. I always like the way Feldman speaks. I am a big believer in his lipid view, at least that medicine may have things wrong, and find it great that he calmly accepts he may be wrong or right.

  12. Excellent information 🙂👍.

  13. So aside from the small group of lean Keto/LC'rs with a hyperresponding phenotype… what does the energy model suggest about a high TC, High HDL but low TG for the rest of the population?

    I eat a normal diet, and my TC is 300, HDL 90, with a TG of 60… the lipid hypothesis of heart disease says the TC is too high!
    https://www.youtube.com/watch?v=9U8L0Vx702o&t=44m34s

  14. I love these dialogues…so calm, collected, and concise.

Leave a Reply

Your email address will not be published. Required fields are marked *