Some thoughts (and accounting) on the Lean Mass Hyper Responder Phenotype

Hey Folks!

I met Dave Feldman in 2016 at a Ketogains shindig in Las Vegas. 

Dave presented some of his earliest findings on what would later come to be known as the “Lean Mass Hyper Responder” (LMHR) phenotype which is described as an individual who is lean, active, eating a ketogenic diet, is ostensibly metabolically healthy, but shows remarkably elevated LDL cholesterol and the associated LDL-particles. At that time (and still to this day) I was on the board of a medical risk assessment program that routinely uses LDL-P (LDL particle count) as a data point in assessing metabolic health and in particular risk for type 2 diabetes and cardiovascular disease, so I had my own bag of familiarity and biases as they relate to lipoproteins. Dave presented some pretty mind-bending stuff, including the ability to raise and lower lipoproteins in a matter of days with various over and under feeding strategies. 

Within standard lipidology circles, this did not seem remotely possible. It’d long been thought that these values were stable over at least weeks and more likely months. So that point was and still is, quite interesting.

When I first met Dave I BELIEVE he suggested that the elevation we see in lipoproteins, specifically LDL, was a compensatory mechanism to deliver more energy. If I have that wrong, if that’s not been the supposition, then this whole exercise is just reflective of my poor hearing and or poorer memory…but I’m pretty sure that was at least the initial thought, although upon reading the current literature on the LMHR, that seems to suggest the LDL elevations are an artifact of throughput, but I’m getting a bit ahead of myself. When Dave first floated this idea to me, it made sense. I mean, why not? It was perhaps 6 months later however that a question finally percolated up to the conscious portion of my mind: Does the basic accounting line up with the proposed mechanism? Said another way: given the understood energy needs for the average LMHR, does the observable increase in lipoproteins provide a plausible supplemental energy source? Or, might this process be an artifact of energy throughput? Or are there even other possible contributors to this process?

Before I launch in I want to make it clear that I think this LMHR topic is remarkably interesting and as I’ve already said, the “simple” fact that lipoproteins and cholesterol can change so dramatically, in such a short period of time has forced a rewrite of what we know about lipidology. That’s a big deal and Dave deserves huge accolades for bringing that to light. I’m going to do my best to unpack this energy delivery story but I’d greatly appreciate other folks checking my math and kicking holes in any dubious assumptions I make here. 

Enough about you, let’s talk about me…

So, I’m going to back into all of this starting at the big picture of my energy needs (because I eat a low carb diet and have gone in and out of having LMHR type numbers…I’ll try to unpack some of that later)  and then work back from that. When I use something like the KetoGains macro calculator (and I’ve used others as confirmation, they are all pretty close) my basal energy needs (170ish lbs, modest base activity) puts me at about 2400 calories/day. Depending on if I do some Zone 2 cardio, jiu jitsu, weights etc, my daily energy needs can easily go up by 1,000 calories. Today is a good example of a modest jiujitsu day:

I know there is all kinds of gnashing of teeth about how accurate these types of platforms are, but when I’ve looked at other estimates of calorie output for a given activity (in this case Jiujitsu) it’s been pretty spot on. ALSO: When I track this stuff, and match my calorie intake as I describe below, things work quite well with regards to being adequately fuelled and having decent body composition. 

The way I eat would likely be termed a “Modified Atkins” type of diet, higher in protein than the 4/1 or 3/1 classic ketogenic diets which put a pretty tight limit on protein to ensure ketone levels are as high as possible for therapeutic reasons. I will usually get around 170-190g of protein, 20-50g of carbs some days, effectively “zero” others, the remainder fat. I use Cronometer to track all this and I just hit my protein targets, cap my carbs and then shuffle in fats to reach my calorie needs for a given day. (This is probably all more info than y’all need, but it helps me to walk through this process, so apologies if this seems pedantic). 

When I’ve worn a cgm my blood glucose tends to be quite stable, typically in the 70-90 mg/dl range) while my ketones are usually on the lower side, around .5-1.0 mmol/L. This is LIKELY due to the fact I’ve been in and around ketosis for 25 years and have likely shifted a good bit of my metabolism to burn fats directly (ketones are often higher in the early stages of ketosis, tend to decline with time, which can present all kinds of problems for folks who then do squirrely things to boost ketones…an important topic I’ve talked about previously, but we can dig into if there is interest). My triglycerides have historically been  low, in the 60-70mg/dl…while I really do not know what my free fatty acid levels are but word on the street is that for the average population this ranges from .1-.6mmol/L. Perhaps the only other things to consider at this point are the lipoproteins themselves, namely HDL, LDL and VLDL…but I honestly think we can simplify this and just look at LDL as that’s the only lipoprotein that elevates dramatically in these situations. This has quite a range as there are folks who walk around with LDL cholesterol as low as 80mg/dL, many are in the 120-150 ranger, while LMHR can be 300-500mg/dL. My LDL-choilesterol has been as low as 80 and as high as 120. My lipoproteins have been as low ass 800 and as high as 2400. What I’ve personally found is that if I keep a decent contribution from monounsaturated fat, not a ton of dairy fat, my numbers are much lower than otherwise. 

If you want to get all kinds of geeked out on this I guess we could track things like remnant chylomicrons and things like that, but as you will see, this all quickly turns into a calculus problem!

What I’ve tried to do above is provide some baseline of the expected amounts of these various substrates that can contribute to total energy delivery….and again, what I’m doing here is providing a bit of a flow of consciousness as to how I’m looking at this so other folks can hopefully point out things I got wrong or did not think of. I think we need to consider not only the average amounts of these items in a standard blood draw, we also need to understand how much in total we could expect to be in our bodies in a given moment, and then we need to think about some upper limits as to production in say, one day. THEN, the money shot: how much energy could we hope to see contributed from that source in a given day? When we have all that we should be able to look at LDL contribution and judge if a 3-5 fold increase (as is typical for the LMHR) is plausible as a meaningful contributor to energy supply. 

This is already getting long and I’m inclined to just dump numbers into a chart and go from there, but again, I think showing my work, although tedious for you and me, is the only way to really know if I’m on the mark. So, let’s look at glucose. 

Glucose

At any given moment we have about 4g of glucose in circulation, with about 400g total stored as glycogen (about 100g in the liver, the remainder in the muscles, which can only be used in those muscles, it cannot find it’s way back into circulation due to an enzymatic feature of muscles vs the liver). That’s fine, but what about gluconeogenesis? We can convert some amino acids, lactate, and even the glycerol backbone of fats into glucose, how much can the liver churn out in a day? I poured over a lot of material trying to track this down and did not find much quantification on this. I DID try putting this question to Grok and received the following:

“The maximum gluconeogenic output of the liver for a day is around 240-300 grams of glucose. This is the amount of glucose that the liver can produce through gluconeogenesis, a process that converts non-carbohydrate sources like amino acids, lactate, and glycerol into glucose to maintain blood sugar levels. However, this value can vary depending on factors such as diet, physical activity, and overall health.” 

I’ve found Grok pretty helpful in narrowing search inquiries, it’s usually quite fast and typically provides references. This question took Grok about 30 seconds to answer and it provided no citations…so, this is a data point that could stand some deeper scrutiny. I tried looking at weird situations like how high glucose can go with a type 1 diabetic to look at another way of figuring out maximum hepatic glucose production ( Type 1’s can get over 500mg/dL) but I’m not sure that’s helpful as that’s just reflecting that these folks cannot clear glucose absent the administration of insulin. I also found another rabbit hole in non-hepatic glucose sources such as the kidneys…there are other tissues that can contribute to glucose production. I’d put more weight on this but there is a reality that for the LMHR population we are NOT seeing elevated glucose levels, quite the opposite in fact. When I looked at the average glucose production per day the ranges I saw were 180-300g/day, which is fairly in line with what we see in theoretical hepatic gluconeogenic production. So, for the sake of this model, I’d put maximum upper glucose contribution per day under these circumstances at about 300g/day which is 1200 calories. It’s almost certain that LMHR folks are NOT seeing this magnitude of glucose flux due to their consistently low glucose levels, but it provides a upper bracket. To me it also provides a bit of a question: if the body “just” needs more energy, and we are nowhere near maxed on hepatic production, why not just dial glucose levels up a bit? Given how similar the LMHR state is to fasting, this is likely a stretch, but (I thought) worth a question.  

Before I shift gears to triglycerides and fatty acids I think it’s worth asking a question around the maximum energy production a human can sustain in a given day (estimates put this at about 5,000 calories). I’d always thought that so long as adequate fuel was dumped in the front end there was no hard limit to this, but many years ago Joel Jameson made me aware of the Constrained Energy Hypothesis. The idea here is that as energy output increases the total caloric need for an individual does not simply add to basal needs (growth/repair + activity) but we will actually experience a ceiling of sorts as the body halts various aspects of recovery and repair. This is likely part of why we see significant deterioration in folks who do things like BUDS, ultra endurance events and the like, even when heroic efforts are made to adequately fuel. Again, I’m making all kinds of assumptions which I’d appreciate other folks to critique, but I think this is valuable to just put an upper limit of energy expenditure/delivery at about 5,000/day (and yes, I know this depends on a lot of factors, not the least of which is size of the individual!)

Fats (triglycerides and free fatty acids)

Over the years of thinking about tackling this topic I’ve known this piece was going to have a lot of moving parts. We have intramuscular triglycerides, body fat (stored as triglycerides) and these triglycerides can be deesterified and released either into circulation (in the case of body fat)  to be used anywhere in the body (including going to the liver and being converted to ketones) or shuttled into the local mitochondria of working muscles. Even lean individuals can store a non-trivial amount of energy in the form of body fat. I’m 170 lbs, about 10% BF which is 17 lbs of fat…which is 59,500 calories! Now, there is a reality that our bodies utilize dietary lipids preferentially to stored supplies. This kinda makes sense in that if we are flush with dietary calories the body will be a bit less stingy than if we are digging into vital reserves. 

Ok, so what does all this mean in the context of unraveling this LMHR question? Even though I can have nearly 60K stored calories as body fat, there is a limit in how much of that I can access in a day. There is also an upper limit in how much I can digestively consume per day (before disaster pants) and also that upper limit for energy expenditure via the Constrained Energy Hypothesis. I can burn about 5K calories per day (really the upper bound before things start falling apart) and realistically I can take in about 5K per day before all manner of digestive evils set it (again, I’m making a bunch of assumptions here, if I have things wrong, let’s unpack them). But how much fat can one actually use in a given day to fuel activity? I looked through “a lot” of material and did not find a great answer to this (other than a directional idea from keto adapted athletes which I’ll touch on in a moment). So, I asked Grok again and got this: 

“The upper limit of fat utilization for humans per day can vary depending on factors such as age, sex, weight, and activity level. However, a general guideline is that the body can use approximately 31.4 calories of fat per pound of body fat per day (14.2 calories per kg of body fat per day). This means that if an individual has 20 pounds of body fat, their body can use approximately 628 calories of fat per day (20 x 31.4).” 

No reference is provided for this, so I’m not sure where Grok is getting this information. I do think it’s safe to say this is derived from a standard population that is not fat adapted. Many of you are likely familiar with this study which looked at the metabolic characteristics of keto adapted ultra endurance athletes which suggested a whopping 1.5g/min fat utilization. This rewrote our understanding of this topic as previous studies suggested an upper limit of 1.0g/mon for maximal fat utilization. What’s interesting here is the individual performed a session at 64% VO2 max for 180 min, which derived the 1.5g/min fat utilization. That’s about 270g of fat and over 1,000 calories, so Grok is clearly not savvy to keto oriented material!

So, I think it’s safe to say that our upper daily fat utilization is quite a bit higher than standard physiology circles would appreciate but this paper from 2001 (Fat utilization during exercise: adaptation to a fat-rich diet increases utilization of plasma fatty acids and very low density lipoprotein-triacylglycerol in humans) illuminated some pretty interesting things. The first is in their introduction they cited previous studies which suggested little if any intramuscular fat utilization during exercise (which they subsequently provided findings suggesting the opposite) but in comparing a normal high carb group compared to a higher fat group (60% fat, 20% carbs, so not keto adapted) They found that not only do the higher fat fuelled folks burn more fat at the intensities studied, a non-trivial amount of that energy is delivered in the form of VLDL-triglycerides. 

Let’s take a step back for a moment and consider a cartoon level look at fat digestion and distribution around the body, and what that means for lipoproteins:

Source: https://openoregon.pressbooks.pub/nutritionscience/chapter/5e-lipid-transport-storage-util/

Dietary fat is broken down passing through the intestinal epithelial cells and packaged into chylomicrons as triglycerides. These particles have two primary fates: delivering fatty acids to peripheral tissues such as adipocytes and muscle, chylomicron remnants which make their way to the liver and these can be converted to VLDL’s (very low density lipoproteins), ILDL’s  (intermediate density lipoproteins) and ultimately can contribute to the production of LDL (low density lipoproteins). From that same site this is a great comparison of these various players:

Ok, I THINK we are in a position to address this LMHR question of if LDL is elevated in response to the body distributing more energy in the form of lipids? 

Yes. 

But LDL elevation is (IMO and accounting) an artifact, not a contributor. Here’s the thing, the best I can find on this suggests the human liver can make about 1g of LDL particles PER DAY. Again, that’s a Grok number, no direct citations, but I found similar bounds in some lipidology texts. The reality is as important as lipoproteins are, they have never existed in amounts or production levels to be primary energy contributors…unless I seriously missed something in all of this. If someone has better numbers that paint a different story, lay them out in the comments please. 

Now, that might be the most anti-climatic finding in the whole interweb, as all I’ve done is backed into the same conclusions suggested in The Lipid Energy Model: Reimagining Lipoprotein Function in the Context of Carbohydrate-Restricted Diets but I THINK I looked at this in a very different way than the study authors did. And, again, I’ve been operating from the perspective that at some point it was suggested that LDL elevations are to provide additional energy. If that was never the case, this is a clear sign of my cognitive decline and I should begin my ascent to public office!

All that said, I believe that’s one layer to this story, I do believe there are others including aspects of the innate immune response and subclinical hypothyroid which may play into some of what we see in the LMHR. 

Lipoproteins and sepsis

Lipoproteins have a host of functions, not the least of which is acting as part of the innate immune system and helping us to deal with gram negative bacterial infections and the associated lipopolysaccharide (LPS) which can be so incredibly deadly. If you are not familiar, sepsis is the condition of a blood borne bacterial infection and some of the material which makes up the bacterial cell membrane (lipopolysaccharide, LPS) causes all manner of mayhem for higher organisms. Lipoproteins play an interesting role in neutralizing LPS by binding it and transporting it to the liver:

Source:https://www.researchgate.net/publication/6703271_Murch_O_Collin_M_Hinds_CJ_Thiemermann_C_Lipoproteins_in_inflammation_and_sepsis_I_Basic_science

It’s worth noting that although that graphic shows HDL, most of the lipoproteins play some role in this process. There is even data that suggests LPS is cleared in part by direct binding to the LDL-receptor in the liver. What I think may be occurring in at least some people is increased fat intake, particularly saturated fat, increases LPS translocation from the gut to the circulation. This may then increase the production of LDL as a protective mechanism as it is well understood that, all things being equal, higher lipoprotein/cholesterol levels are protective against sepsis and low cholesterol levels are a risk factor for death from sepsis. I’m not sure if we have much on the research side of this but we have a deep well of anecdote of folks shifting out saturated fat for mono’s and seeing significant decreases in LDL. 

Subclinical Hypothyroid

Another possible mechanism which may be at play in the LMHR is subclinical hypothyroid. This may seem a stretch as these individuals are quite lean and active, but it’s well understood that low carb diets dramatically reduce thyroid levels. There has been a lot of debate if this reduction in thyroid is benign or not. I suspect it generally is benign, but there may be some metabolic tradeoffs at play here. This whole situation reminds me of the cholesterol elevation seen in anorexia. It looks eerily similar to the LMHR, and explanations include increased lipid reserves being mobilized to supply energy. But it’s also clear these folks have absolutely cratered thyroid levels (and other hormones too…which I think is a good indicator for low carb folks that if androgens are amiss, we need to potentially rejigger some things.) What if some percent of the LMHR are in a limbo state of thyroid? Not low enough to cause classic symptoms, but low enough to really impact cholesterol and lipoproteins. As an example, Dr. Bernstein has long recommended that folks following his low carb high protein diet for diabetes (both type 1 and 2) should add thyroid to mitigate elevated LDL. 

Let them Eat Oreos!

A related contribution of late is a case report in which an individual added 100g of carbohydrate in the form of Oreo Cookies (a tasty option, if I do say so) to see what the effect of of this level of carb intake would do with regards to lipoprotein and cholesterol levels. I’ll let you check that out for yourself, but the interesting finding was that Oreo cookies reduced LDL-C a whopping 71% while a fairly robust dose of statin only produced a 32% reduction. 

What to make of this? First, we have to remember some of the biochemistry of ketosis and that if we really “want” ketones we need to limit both carbs and protein to a degree as carbs and some amino acids can provide carbons which end up as oxaloacetate, which will be used with acetoacetate, effectively “bumping us out of ketosis.”  Ketosis requires energy flow through HMGCO-A and this alone is sufficient in some folks to produce elevated lipoproteins:

It’s worth noting this does not happen with everyone. Peter Attia noted he has great looking lipoproteins on keto. Others see immediate elevations with just an uptick of saturated fat (which can be due to multiple mechanisms, as we’ve seen). In the case of this Oreo Cookie experiment they administered exogenous ketones to technically stay in a state of ketosis, which might be an interesting technique for adding carbs while still getting some benefits of ketosis, but the carbs provided in that example were clearly enough to halt ketogenesis as that’s why they had to use exogenous ketones. It’s a very different process with regards to lipoproteins and how we reach ketosis. Unfortunately, this experiment did not look at thyroid, so we do not know if that normalized, but most likely, it did. So I can see two mechanisms that could have contributed to the reduction of lipoproteins: sufficient oxaloacetate to offline ketosis, elevated thyroid which will foster the reduction of lipoproteeins. 

Ok, that’s what I have on this topic. I have no doubt I’ve missed a lot of things here, looking forward to what folks have to contribute to this,