Note from Dave: It’s no secret that I consider myself “cautiously optimistic” about high LDL on a low carb diet given what I’ve learned to this point (see my most recent presentation on it). But as always, I encourage everyone to listen to every major side of a debate, which is why I’m pleased for our site to feature this article by Dr. Nadolsky providing a detailed counterpoint. As always, let’s keep comments respectful and productive.
High Cholesterol on Low Carb
Many who go onto a low-carb high fat diet observe an increase in total and LDL cholesterol (LDL-C). In the low carb community these are often referred to as “Hyper-Responders”. This can often be coupled with increased HDL cholesterol (HDL-C) and reduced triglycerides. Dave often refers to these three together as “The Triad” (High LDL-C, High HDL-C, low triglycerides).
Which begs the question:
Would this same hyper-responder phenotype of high LDL cholesterol have increased risk of cardiovascular disease (CVD) compared to a similar healthy individual with likewise high HDL-C and low triglycerides, yet low LDL cholesterol?
In this article I’ll be focusing not just on this comparison of cholesterol, but in LDL particles (LDL-P) and apoB containing lipoproteins as well (apoB). Thus, it isn’t just about comparing high LDL-C hyper-responders to those with low LDL-C, it’s comparing LDL-P and ApoB as well. (Note: moving forward in this article, “hyper-responder” will refer to anyone observing high LDL-C/-P, high HDL-C, and low triglycerides while on a low carb diet)
Sticking to this clinical question is very important because this discussion can go south quickly with many strawman arguments.
There is no study that has looked exactly at this. It would take at least a prospective cohort following hyper-responders and a similar group without the large response in apoB/LDL-particles. A shorter term trial could look at surrogate atherosclerosis markers such as CTA or IVUS (two advanced cardiovascular risk markers).
Dave wanted an outside perspective to hopefully keep an open dialog because right now it’s a lot of bickering on social media back and forth. Let’s see how close we can get to the question at hand.
Here is a quick background on me. I am a former NCAA heavyweight wrestler turned obesity/family physician (will be taking my lipid boards this spring). In high school and college I used my passion for nutrition and exercise science to improve my athletic performance, but felt more fulfilled helping the general population rid disease. Hence the medical school and lifestyle as medicine route.
Early in medical school I became a low-carb diet proponent. In fact I had many dinners with the likes of Dr. Westman, Dr. Phinney, et. al. while attending ASBP meetings. Along with being very pro low-carb, I had mentors early on who instilled an LDL-hypothesis (and statin) skepticism. I took a lot of what my mentors said as gospel as any student usually does in the beginning of their training and schooling.
It wasn’t until I started reading the studies myself and aggregating the data where I started realizing my skepticism was likely unwarranted. Early in residency training I reached out to experts like Dr. Dayspring and went to NLA meetings where they were all able to show me with much more convincing data that my LDL-hypothesis skepticism was uninformed.
The reason I became somewhat obsessed with lipids is because I began having patients in 2011 who had skyrocketing LDL cholesterol levels after they would go to a low-carb high saturated fat diet e.g. drinking bulletproof coffee. I wanted to know more about the risks of high levels of LDL cholesterol (and particles) because these patients were reading blogs and books telling them it was okay and safe.
As a physician, we want the best for our patients. Most of us do NOT get money from pharmaceutical companies despite what people think. We want to do no harm. I am also a stickler for physiology. I believe as physicians we should know physiology and pathophysiology better than anyone. If we don’t have the deepest understanding then how can we really understand our treatments for patients? More reason for me to be obsessed with the pathophysiology of atherosclerosis.
I have ZERO ties to lipid lowering pharmaceuticals unless you count some crappy high calorie lunches brought to a clinic I worked for a few years back. I would actually be excited to find there is a paradoxical decrease in risk with a low carb diet induced increase in LDL-particles. That would mean more options to treat patients. We need as many options as possible in order to find a lifestyle they can stick to for life.
Response to Retention
Anyway, today in 2019 we have a lot of data pointing towards an increase in LDL-particles = an increase in cardiovascular disease risk. The totality of data including epidemiology, GWAS and mendelian randomization, clinical trials of statins, pcsk9 inhibitors, bile acid sequestrants, and ezetimibe, show from multiple angles that having higher LDL particles over time increases your risk of atherosclerotic disease in an independent manner. The data also show that lowering your exposure over time decreases your risk. It’s very strong data to ignore. We will touch on each of these (e.g. epidemiology etc.) after we discuss how atherogenesis occurs.
The data above go along with the current working model of atherosclerosis pathophysiology, which is the response-to-retention model. The gist of that model is that the apoB containing lipoproteins (mostly LDL-particles) that you have circulating your blood can cross the inner lining of your arteries (endothelium to intima) and become retained and start the atherosclerosis process. Based on the data and pathophysiology, the more apoB containing particles you have, the higher your burden and the higher the chance you continue the atherosclerosis process.
It’s absolutely important to understand the pathophysiology because it will give us clues on what could help or hinder our risk of atherosclerosis. You have probably read about the analogy of the lipoproteins (LDL-particles) being the cargo ships and the cholesterol being carried as part of the cargo.
I want to further this analogy by imagining that the lining of your artery walls (endothelium) are the harbor walls. If you have healthy endothelial function, imagine you have big bumpers that don’t allow as many cargo ships to crash and get stuck. If you have unhealthy endothelial function, you may have rocks or broken docks or piers that increase the risk of those ships crashing and getting stuck.
To go even further, imagine there are people on shore and what they do once the boat has crashed. They might represent your immune function (monocytes/macrophages). Maybe there is something about the ship that makes it more likely to get stuck (size, shape, and contents of the lipoprotein).
So you can see that it’s not just how many cargo ships (LDL-particles) you have floating around. It’s multiple other factors. They are not mutually exclusive.
A Deeper Look at Atherogenesis
To further this analogy, there are regions in the arterial tree that are more susceptible to atherosclerosis simply due to anatomical reasons (bends and curves). Maybe we think of these areas as having reefs. Regardless, these areas will develop atherosclerosis first just because of location.
Anyway this is the analogy I try to explain to patients because things like smoking, insulin resistance, hypertension and a sedentary lifestyle will disrupt both the harbor walls and the cargo ships and those are easily modifiable with lifestyle, which I like helping with the most.
I wanted to point out some of these different factors because it will come into play when we discuss the counterpoints made by those who don’t agree fully with the response-to-retention model or at least the higher ldl-particles being an independent risk factor.
We have to make sure we understand that it is the apoB containing particles that get into the endothelium so it would make sense to not conflate measurements of apoB/LDL-particles with LDL cholesterol or even Non-hdl cholesterol and sure as heck not total cholesterol. If you listened to Dr. Dayspring on Peter Attia’s podcast, you’re probably well aware of it by now.
Also understand that in the response-to-retention model, the retention of the apoB containing particle is essential to start atherosclerosis. Without that particle retention, there is no start of atherosclerosis. This is why in the lipid world, the lower the better. The low carb world contends that apoB containing particles do not matter, as long as insulin resistance (or other inflammation) is not present. The lipid world agrees that insulin resistance plays a role and even accelerates the atherosclerosis process, but with the notion that the LDL-particles are independently atherogenic. The lipid world does not think you need insulin resistance or systemic inflammation to start atherosclerosis.
A good quick recent review of the pathophysiology of atherosclerosis you can read is right here. It gets much more technical than what I mentioned.
The Low Carb Perspective
So here is the low carb logic as I see it:
- Low carb diets can SOMETIMES increase LDL cholesterol/particles/apoB although not always (see Virta’s data) especially if done in a more “mediterranean” fashion (high nut, olive oil, etc. content). Of course Dave discusses the hyper-responders (recent study here), which is really the question at hand here.
- Many low carbers believe low carb is the best diet, and that anything that occurs with low carb diets must not be bad.
- Therefore, increases in LDL cholesterol/particles are not harmful.
This is a generalization by the way. Not every low-carber feels this way.
I think that if LDL-particles ALWAYS went down, low carbers would be cheering about it (like how the plant-based crowd does now) instead of finding ways to be skeptical. From the outside, it looks like low-carbers are trying to justify the elevations in LDL-particles that occur in a subset of the population.
To clarify, I usually prescribe a lower carb diet to patients with a heavy emphasis on olive oil and nuts and vegetables and protein. I am not at all anti-low carb. I just want the best for my patients and do not want to do harm as mentioned prior.
There are multiple arguments made by staunch low-carb advocates. It would be nice to lay them out here.
Let’s start with epidemiology. For much of these epidemiological studies, they look at a group of people and follow them for years and see what happens. There is no intervention. This can give clues as to what could be increasing risks of various diseases. Here we are talking about cardiovascular disease.
Dave has posted a contest about finding a study that shows increased CVD risk when one has high HDL-C (above 50 mg/dL) , low triglycerides (below 100 mg/dL), and high LDL-C (above 130 mg/dL). This information would at least give clues into the risk of those who have elevations in LDL-C as opposed to those who don’t following a low-carb diet. It’s not perfect as we wouldn’t know the causes of the elevations in LDL-C, but at least it would be the similar phenotype (observable characteristics) to the hyper-responders.
Dave is trying to tease out if LDL-C is true independent risk factor when otherwise metabolically healthy. Why does Dave want to ensure proper HDL-C and triglycerides? Because HDL-C and triglycerides are surrogate markers of insulin sensitivity.
In the low carb community, insulin resistance is believed to be the underlying issue in cardiovascular disease. Meaning if you don’t have insulin resistance, you don’t get cardiovascular disease regardless of your apoB/LDL-particles. It’s an interesting concept we will discuss more later.
ApoB containing particles are a necessary part of atherosclerosis, but their thinking is more that you also need insulin resistance or some other endothelial dysfunction along with the particles to make it happen. Not just the particles.
This hypothesis differs from longstanding atherosclerosis and lipidology researchers. Most agree that not only can insulin resistance and other factors increase or accelerate atherosclerosis, but an increased burden of apoB/LDL particles will also increase it. Remember the response-to-retention model above.
LDL Particle Burden
Anyway, ideally instead of LDL-C, we would be looking at apoB or LDL-particle levels as it is the particles that have been shown to be more predictive than the cholesterol measures in terms of CVD risk as discussed in the pathophysiology above. A non-hdl-C would be better than the LDL-C as well. We would also have other more accurate measures of insulin sensitivity (fasting insulin, or even dynamic testing e.g. Kraft) instead of HDL-C and triglycerides, but we have to make due with what the literature has reported. In insulin sensitive individuals, LDL-C is an okay surrogate marker for apoB/LDL-particles.
So far this contest question hasn’t been answered. We have many studies that adjust and use statistical analyses showing that LDL-C is an independent risk factor, but we don’t have a pooled cohort looking at his exact question. There are a few reasons for this as I will discuss.
I thought I had found one in this recent paper that just came out from the Cooper Center Longitudinal Study:
In this paper they wanted to see the effect of LDL-C and Non-HDL-C on cardiovascular risk over time in low risk individuals. Low risk was defined as less than a 7.5% 10-year risk of atherosclerotic cardiovascular disease as calculated by the Pooled Cohort Equations. I am showing LDL-C above, but the Non-HDL-C graph looked similar. As you can see, those with the highest levels of LDL-C has the highest risk of CVD death.
It does not meet Dave’s criteria though as the table below shows.
As the LDL-C goes up, the triglycerides go up and the HDL-C goes down. Multivariable analyses were used to adjust for a few known risk factors and it was found that LDL-C and non-HDL-C were still independent risk factors. However, as shown it does not meet Dave’s criteria.
Remember the hypothesis within the low carb community is that LDL particles do not matter if one is otherwise metabolically healthy.
One problem with these types of studies is having enough subjects AND events over enough time. Because each subject would have otherwise low risk, you would need enough time to show (or not show) separation in a graph like the one above (Kaplan-Meier). Another problem would be a lot of confounding. Dave insists not adjusting the data, which means we would need a lot of subjects to have a proper study. Age is also a factor because for those who are too young we would have to wait decades to get a significant difference in events. If the subjects are too old, we may have missed the previous lifelong exposure to the risk factors.
Let’s move next to genetics as I feel this shows us the most similar to hyper-responder phenotype. It has been long-known that patients with familial hypercholesterolemia (FH) are at risk for early atherosclerotic cardiovascular disease. It’s also known that these patients will be incorrectly listed as “low risk” if put into one of the risk calculators.
FH patients are exposed to a much higher burden of LDL-particles early in life compared to non-FH patients. It’s important to note that FH isn’t due to one single genetic cause. FH is a phenotype. Lifelong elevations in LDL cholesterol (and particles) due to decreased clearance. This can be due to mutations in the LDL-receptor gene, apoB gene, PCSK9, and other rarer mutations. If you have issues with your LDL receptor, you won’t have as much uptake of the LDL-particle to get out of your blood. Same thing happens when you have mutations in your apoB. The receptor works but the LDL-particle won’t be taken up as easily. PCSK9 is involved with LDL-receptor metabolism so this would be similar to issues with the LDL receptor.
Either way, you have a single mutation leading to higher LDL-particle levels early on in life. There is heterozygous and homozygous FH. Heterozygous means one of the pair of genes is affected while homozygous means both genes are affected. Homozygous FH leads to extreme levels of LDL cholesterol and particles very early on in life and is associated with high levels of premature cardiovascular disease. Most hyper-responders look similar to the heterozygous FH patients (LDL-C around 200 to 300 or so mg/dL), particularly the “Lean Mass Hyper-responder” phenotype (LMHR) Dave discusses here at the site.
Someone with true FH is not comparable to someone with a low-carb induced elevation in LDL-C because those with FH have had elevations their whole life. On the other hand, the lipid numbers look very similar. Those with FH don’t always have elevations in triglycerides or decreases in HDL-C as their underlying issue is LDL-particle clearance and not triglyceride clearance (or dyslipidemia from insulin resistance). A study recently looked at those with the mutations that would likely lead to an FH phenotype in a cohort of patients.
A recent prospective epidemiological study looked at the FH phenotype (LDL-C greater than 190 mg/dL) compared to non FH subjects and found a significant accelerated risk of atherosclerosis. Looking at the supplementary tables, the triglycerides in those with the FH phenotype are not listed. The HDL-C on average was just lower than 50 though in those with the FH phenotype. The genotyping was not done in this one.
It’s not just unfair to make a direct comparison of FH to Hyper-responders for the reason of lifelong burden of LDL-C/P, but also because there may be other issues going on that increase atherosclerosis in these genetic situations. Those with FH may have changes in monocyte activity after retention of the LDL-particle within the artery. With my initial analogy, think about the ship crashing and the people (monocytes) on shore going and grabbing the cargo (cholesterol). This may differ depending on the genetic cause of the individual’s FH (LDL receptor vs. apoB vs. PCSK9).
Regardless, it would help if we had a genetic model that were closest to what is seen in those considered to be low carb hyper-responder.
What the heck is the mechanism behind these explosive apoB/LDL-p numbers in these individuals?
My hypothesis is this:
- A decrease in LDL receptor activity and also increased lipoproteins from excess saturated fat and paucity of polyunsaturated fat.
- A very low insulin state which would also decrease LDL receptor activity and potentially increase apoB formation.
- A non oxidative pathway for ketones is cholesterol.
- Ketogenic dieters have increased intakes of dietary cholesterol which does increase LDL cholesterol levels despite what some have said in the past.
- Low soluble fiber intake may increase dietary and bile cholesterol absorption.
I believe each of these are additive in nature and therefore could be testable.
There is no exact genetic model like this, but there are studies that look at some of these using something called mendelian randomization (MR). Some criticize MRs by pointing to how genes can be associated to many things, suggesting the outcome can be due to something else. Let’s unpack this to better understand the argument.
First, let us consider genome-wide association studies (GWAS). Prior to the wide availability of human genomic data, massive amounts of studies were published looking at the association of single nucleotide polymorphism (SNP) in a particular gene to some phenotype or clinical endpoint. They were based on “a priori” knowledge of the workings of that gene. For example, one could look at a specific protein and reasonably assume that it had some effect on a disease. Then, it would be logical to also infer that a SNP on that protein (that either enhances or hinders its function) would have an effect on the risk of disease in humans. It is relatively simple to do an association study from one SNP to an outcome in a given population. These studies were often limited by the amount of sequenced genetic data. N’s were in the range of hundreds.
However, with the adoption of full genomic data, it has become evident that this so called ‘candidate gene’ approach is wildly biased and produces spurious correlations that all too often don’t replicate in larger sample sizes. The explanation is that the genome is messy and if you have just a few people, stuff just tends to correlate with other stuff. Nowadays, most candidate gene -studies are considered false..
..unless they are replicated in GWAS studies. So what’s the difference?
One of the common problems in medicine is the reliance on ‘statistical significance’ as a marker of the reliability of the result. To keep it brief, people usually say their result is significant if P < 0.05. This can be seriously misleading. A very special kind of problem is related to multiple tests. In a sense, that’s equivalent to throwing a bunch of crap on a wall and seeing what sticks. If you do enough tests, P can be below 0.05 once in 20 attempts just due to random chance. In genetics, you can have MILLIONS of SNPs and you’re testing against a number of outcomes. Some of them are guaranteed to be “significant” if you just look for those P < 0.05 results but they’re not real – they’re just noise.
That’s where the modern GWAS studies come in. They use N’s in the range of tens of thousands of people. If a SNP actually has a better-than-chance association to a disease, it’s going to have to survive multiple testing across the entire genome – that’s how you know it’s not just random noise. In GWAS studies, the threshold for statistical significance is often set at 5 x 10-8, or 0.00000005. That’s so low it’s virtually never achieved in, for example, clinical trials.
There are some notable GWAS studies linking genetically high LDL to higher risk of CVD and vice versa (Willer et al. and Teslovich et al). Now, remember that there are countless SNPs all over the genome so if the same specific ones consistently associated with LDL and CVD in over a hundred thousand people with genome-wide statistical significance, IT’S A BIG DEAL.
But of course, you can still argue that even a statistically strong association is just an association. Fair enough. So let’s break out the MRs.
The gold standard of proving causality in medicine is to do a randomized controlled trial. It works the best in drug trials where the exposure ( = drug) is easy to conceal. You split your subjects into two groups where one gets the drug and other gets the placebo. But did you ever wonder why this is such a powerful design? The reason is that if the people are properly randomized, THE ONLY DIFFERENCE between the groups is the drug. All other variation is evenly distributed to the groups so they don’t contribute to the outcome. If there is roughly the same amount of overweight people in both groups, the outcome is likely not affected by overweight. And so on.
MR studies take this paradigm and apply it to genetics. Here, randomization is known to take place when a sperm fertilizes an egg. Each SNP travels on a bit of DNA that’s passed on from parent to the offspring but overall, these bits are randomly distributed across the genome. This is why MR studies can be so powerful in determining causality: you can compare two groups of people with one having a particular SNP and the other without it. ALL OTHER GENOMIC VARIATION IS RANDOMLY DISTRIBUTED BETWEEN THE GROUPS so it’s unlikely to have an effect. MR studies take the GWAS hits (remember the teeny tiny P-value in them) one step further and can be used to very strongly infer causality.
How strong, you ask? Well, Big Pharma companies are starting to adopt MR studies as a part of their preclinical drug development pipeline. If they’re not seeing promise of a causal effect, they’re unlikely to keep developing those drugs.
There are multiple MR studies showing that all genomic variation being equal, those SNPs that are associated with high LDL are very likely causally associated with CVD and longevity. Anyone who disputes these results doesn’t understand the gravity of this evidence.
Here is a recent look at the genetic data using MR. You can read the study here, but basically ANY decrease in apoB containing lipoproteins from a genetic standpoint using MR was associated with less risk of heart disease. They looked at both LDL cholesterol and triglyceride changes and it all came down to apoB levels as the response to retention model predicts. See the graph below.
An important excerpt from the above, “The results of this study are also consistent with prior mendelian randomization studies demonstrating that triglyceride-rich ApoB containing remnant particles appear to be causally associated with the risk of cardiovascular disease. The results of the current study extend those findings by suggesting that triglyceride-rich remnant particles have approximately the same effect on the risk of cardiovascular disease as LDL particles. Furthermore, the results of this study are consistent with a recent mendelian randomization study that demonstrated that the causal effect of LDL particles on the risk of cardiovascular disease appears to be determined by the concentration of circulating LDL particles as measured by ApoB rather than by the mass of cholesterol carried by those particles as measured by LDL-C. The results of the current study confirm and extend those findings by suggesting that the causal effect of all ApoB containing lipoprotein particles on the risk of cardiovascular disease appears to be determined by the circulating concentration of those particles rather than by the mass of cholesterol or triglyceride that they carry.”
So going back to the hyper-responder phenotype of elevated LDL particles. We have no reason based on genetic data that the increase in apoB/LDL particles is benign. We have only reason to believe it’s harmful.
Let’s move on to trials.
Remember the studies done on LDL receptor genetic differences above? Well drugs that increase LDL receptors have been shown to lower cardiovascular events. Take a look here.
The lower you bring your LDL particles via drugs that increase LDL receptor the lower your risk of vascular events.
When pointing out that statins reduce cardiovascular events, it is claimed that it is due to the pleiotropic effects (anti-inflammatory). There may be some anti-inflammatory effect, but the LDL cholesterol lowering of statins follows the same risk reduction curve as other modalities. The main effect is from the LDL lowering.
It’s not just statins that work either. Take a look at that chart. PCSK9 inhibitors and ezetemibe and even bile acid sequestrants.
Heck, while I think some of these trials should be redone, dietary interventions lower the risk too. Saturated fat, specifically palmitic acid, increases LDL particles at least partially from LDL receptor downregulation relatively to polyunsaturated fat as mentioned before. Recent cochrane analyses of both reducing saturated fat (and replacing with polyunsaturated fat) and increasing polyunsaturated fat also show improved risk from cardiovascular events. From a mechanistic standpoint as well, polyunsaturated fat may decrease aggregation of the LDL particles once inside the endothelium, which may decrease atherosclerosis.
Many skeptics will point out that the statistics are skewed because relative risk is used as opposed to absolute risk. For example if your absolute risk went from 1% of having a heart attack to 0.7% risk, you could say you lowered your risk by 30%. That’s a relative risk difference though. This is a silly argument though for a few reasons.
- Regardless of effect size, there is clearly an effect. So you can’t say the LDL levels don’t matter.
- These trials are relatively short compared to the long process of atherosclerosis. When you extrapolate over years and over many people in the population, the effect would increase.
I’m not advocating that everyone needs to take drugs. I am just answering the question at hand, which is do increased LDL particles increase risk of cardiovascular disease. Everything I have shown points to yes. Does this answer the question of whether increased LDL particles in the hyper-responder phenotype increase risk of cardiovascular disease? Not exactly, but it should give us some big clues.
Common Questions and Statements
Let’s go over other common skeptical questions/statements.
“We need high cholesterol for our cells”
Each cell makes cholesterol. We don’t need much in our serum. In fact you can have very low levels of serum cholesterol and be perfectly healthy (see genetic data above). There is some epidemiological data that lower serum levels correlates to poorer health. This can be explained by something called “reverse causation”. Basically poor health can lead to lower cholesterol. The genetic data above show there isn’t a risk when poor health isn’t the culprit.
“But the increased HDL cholesterol from low carb high fat is protective!”
Probably not. Epidemiological studies showed an inverse relationship, but it doesn’t pan out in genetic studies or drugs that increase HDL cholesterol. It’s likely just a surrogate marker for metabolic health as described in the beginning. In fact very high levels of HDL cholesterol could be harmful as shown recently. It likely comes down to the function of these HDL particles, not the cholesterol that’s carried.
“All-cause mortality isn’t different so who cares”
Well first the question at hand is whether the increased LDL particles increases cardiovascular disease. As I showed they do. But what is often said is that the risk of dying isn’t different. Valid question, but I would be very cautious because even if you don’t die from a heart attack, it’s not something you’ll want to go through and live with. Quality of life likely decreases and cost increases. Not only that, but the genetic data show there likely is a longevity component. It just takes a lot of time and more people in the studies. That’s why the relatively short statin studies may not show an all-cause mortality difference. It comes down to stats and power of the studies.
It’s not always and in fact not USUALLY the case that apoB containing lipoproteins go up. In fact, Virta’s recent study show their patient’s apoB is unchanged while every other risk factor goes down. This is due to the therapeutic effect of their program on the patient’s insulin resistance and diabetes. Those patient’s apoB levels were driven by insulin resistance (read here for a primer on why insulin resistance increases apoB levels).
Problem is not exactly the same but LDLr differences – high burden of LDL-P. Otherwise metabolically healthy still get atherosclerosis.
Energy Model and Risk
The energy hypothesis with hyper-responders isn’t mutually exclusive with the LDL hypothesis of atherosclerosis.
- The idea that insulin resistance is the cause for vascular disease is also not mutually exclusive from increases in LDL particles increasing ASCVD risk. Insulin resistance and inflammatory states increase risk as well likely from increased retention and modification (immune response as Dave discusses) of retained LDL particles.
- Atherosclerosis takes years. Area under the curve is the most important aspect as shown in those with FH/apoB/LDLr mutations versus non carriers. Analogy of retirement
- So for the hyper-responders who have astronomical increases in LDL cholesterol, does it put them at risk? I would say so given our current evidence although it would take many years to manifest.
- What about those with just slight increases in LDL cholesterol/apoB with a lower carb diet. On a population level it might increase with all else.
If after all of this you still want to continue with your high LDL particle/cholesterol levels because you feel better then more power to you. I felt I needed to put this out there so you have informed risk on the matter.
In that case you do want to continue this lifestyle, it is crucial to study this further. I am working on resources to help further the science behind this. It may turn out that there is something protective. We won’t know until we study it more.