Jan 20


Please consider supporting my Patreon. All funding for my research and this site come solely from individuals like you. Thank you!

  • If you know little to nothing about cholesterol->
    • And you want to learn the basics->
      • You can check out my Simple Guide to Cholesterol series. It’s full of illustrations and is written for laypeople. Enjoy!
      • Likewise, I have this video that goes over the basic markers for cholesterol while on a low carb diet. (Pictured to the right)
    • You can enter your cholesterol numbers into our popular Report tool to check them against many risk calculations at the same time.
  • If you’re wanting to know about my research->
    • You want an overview->
    • You want the most recent breakthroughs->
      • 1/2/2018: In this latest video, I demonstrate massive changes to my LDL Cholesterol over 5 stages in a matter of days. LDL 207 to 103 mg/dL in seven days with high carb, up again to 146 on mixed, down again to 113 on high fat. (Pictured to the right)
  • If you have seen your cholesterol rise considerably on a low-carb high-fat diet (like myself):
    • You may want to first visit the FAQ.
    • I would strongly encourage you to read through this blog and my own journey revealing the Inversion Pattern. Key moments were the Identical Diet experiment and the Extreme Cholesterol Drop experiment that I wrapped around the first presentation of my data for the Ketogains Seminar.

Oct 18

The Tandem Experiment Part 1: Fat

Two Experiments in One

On the plane to #NLANashville!

As soon as we started to discuss possibly attending #NLANashville Dave and I began to toy around with the idea of trying out a dual experiment, similar to the identical diet experiment, with a twist. Instead of doing the same experiment the two of us would be doing two different experiments – but both of us would be manipulating our cholesterol numbers intentionally. All while attending a lipid conference and trying to soak up all the knowledge there (because who doesn’t want to add more work on top of an already busy working weekend? Never doubt our masochistic nature).

Originally, two possibilities were floated for my experiment “track”: option 1) fasting for five days or option 2) high calorie ketogenic for five days. I ultimately decided on high calorie, as it would be more interesting to do, and I wanted to be able to concentrate on the presentations while at the conference (fasting longer term often makes me restless). As a result of this choice, I needed something I could easily prepare and eat 1) while traveling to Nashville and 2) while actually at the conference. Luckily, Chris Bair of Ketochow stepped in and offered to sponsor the experiment. This was helpful in a few ways as I would already have most of what I needed before I even left for the conference, it wouldn’t require any cooking, and I could prepare everything for the following day the night before and not have to think about it after that. With that out of the way, the experiment began!


Instead of starting out with fasting, as I did with my Feldman Protocol experiment I instead decided to start with Ketochow intake with baseline calories in the 4 days prior, with some added whey protein powder after day 1 as I had been craving lean meat after a full day of Ketochow only. Calorie intake ended up being around 1500 calories with the added whey. The fat source used was heavy whipping cream – 99 ml per shake, 3 shakes a day. The flavors used were banana and orange cream for baseline, and pumpkin spice caramel and strawberry for intervention (noted here, as some of them have slightly different nutrient info – all Ketochow used was version 2.1).


A bag full of heavy cream for the conference weekend.

Since I was aiming to drop my cholesterol over the course of the following five days, the intervention was to increase the amount of fat while keeping all else the same. Unfortunately, I forgot to bring my whey protein powder with me, so I just went without it after day 1 of intervention. I also found that trying to cram 550~ ml of cream into a bottle for the shake ended up making the texture very…. unfortunate. In order to avoid this problem, I kept the amount of powder the same, but split it up into 6 shakes per day for a goal of 6000 calories (with 4000~ calories achieved on days 1-3, and 6000~ calories achieved on days 4 and 5). For those curious this involved 297 ml of cream for baseline, and 1.6 liters of heavy cream for the highest calorie portion of the experiment. The liquid nature of the fat this time definitely made this possible!

The results

All results include data from my 12 1/2 day Feldman Protocol attempt in the shaded area for comparison – new results are highlighted. The correlations listed are for all data points between 8/20 and 9/24.

Total Cholesterol

As expected, Total Cholesterol dropped from the baseline of 313 mg/dL to a new low of 171 mg/dL after 5 days of high fat feeding. This leaves me with a total drop of 142 mg/dL and a correlation of -0.967684 with the 3 day average of dietary fat before the blood draw.


Baseline LDL-C on the 4 days of Ketochow came out to 252 mg/dL and after 5 days of the high calorie protocol it dropped to 105 mg/dL for a total drop of 147 mg/dL. The correlation between 3 day average dietary fat remained high at -0.97497754. To note, this is the lowest LDL-C I have on record, regardless of diet (Standard American Diet, keto, or carnivore). It’s also worth noting that during the highest calorie days I was consuming 554g of total fat, 387g of which were saturated fat.


LDL-P did exactly as expected and dropped during the high fat/high calorie phase – from 3068 nmol/L to 929 nmol/L for a total drop of 2139 nmol/L. This is pretty much what I would expect from the drop in LDL-C but I was still pretty surprised it had dropped so low. I suppose I won’t complain! The correlation for this one was between LDL-P and the 3 day average of dietary fat plus a two day gap and resulted in -0.919329629.


HDL-C surprised me by starting out higher than is typical for me when at baseline calorie/fat intake – especially as fat was a bit lower than my norm during baseline. I have had a few cases of HDL around the upper 40s on record with a baseline diet, but it’s not the usual case for me. Even more surprising was that upon high fat feeding I got the highest level I have on record – suggesting that my baseline HDL-C may have shifted up during the protocol. I suspect this may be due to something that’s in the Ketochow, and I have my suspicions as to what, but I’ll save investigation into that for another time.

Regardless of the slightly higher baseline at 48 mg/dL HDL-C still did as expected and went higher during the high calorie ketogenic phase all the way up to 57 mg/dL. This is a total increase of 9 mg/dL which is about what I would expect. The correlation with the 3 day average of dietary fat was positive, at 0.815920699.


HDL-P, like HDL-C, went up during the protocol from 21.6 umol/L to 35.7 umol/L for a total increase of 14.1 umol/L and a positive correlation between the 3 day average of dietary fat (plus a two day gap) of 0.922353315.


Triglycerides were around normal baseline levels at 63 mg/dL which dropped further upon the high fat/high calorie feeding all the way down to 45 mg/dL – the lowest I have on record. Apparently, despite the high fat intake, I was able to use or store the fat-based energy quite efficiently! This was a total drop of 18 mg/dL and a correlation of -0.857910104.


Out of all the markers I had been looking at, I was most curious about what lipoprotein(a) would do. During my Feldman Protocol attempt in August, I was surprised to find that it seemed to be following the inversion pattern. As dietary fat intake went higher, lipoprotein(a) dropped lower. But, there were some questions as to why this could be – was it perhaps that having the high calorie phase immediately after a 7 day fast influenced it? Was it something else? Obviously, testing with Ketochow helped isolate out most other possible influences and once again lipoprotein(a) dropped like a rock upon the high calorie ketogenic phase. It also started off a bit higher than my typical 130-140 nmol/L range, but this could be due to the macro composition I was consuming due to using Ketochow + whey protein isolate (resulting in slightly higher protein via whey, and slightly lower fat than usual).

Regardless – despite starting out at 171 nmol/L lipoprotein(a) quickly dropped to an astounding 58 nmol/L after the high fat/high calorie phase of the experiment. Not only was this 7 nmol/L lower than the previous low from August, but it resulted in a 113 nmol/L drop in 5 days of high fat/high calorie feeding (a 66% decrease). Wow! Lipoprotein(a) can move!

Not only that, but the correlation between dietary fat stayed strong here, at -0.93368339.

It will definitely be worth exploring how energy status impacts lipoprotein(a) in the future, along with other possible influences on its levels day-to-day. This experiment (especially back to back with the other protocol) confirms that just like other lipid markers lipoprotein(a) appears to be slightly more dynamic than anticipated.


Final Thoughts

On the Locals

Although I didn’t get to do much (any) sightseeing while in Nashville, except to walk to the conference area from where I was staying, I did get to interact with some of the locals. I had several long conversations with my AirBnB hostess, Meredith, for example, who was surprisingly delighted to hear about what I did for work and what I was in town for. Within about five minutes of me arriving at the AirBnB we were discussing lipidology, and she even pulled up a recent lipid panel she’d had so I could discuss what the markers were referring to. Funnily enough, she’d also had a CAC (a test I’m fairly interested in) recently, citing that she wanted to see any signs of actual disease (I brought up Ivor Cummins after this, of course!). She’d often ask how the conference was going, and said she could discuss lipidology with me for hours as it was all so fascinating (something I would have, of course, been happy to accommodate if I’d had more free time!) and over all was one of the highlights of my time in Nashville.

On the Conference

There were many interesting presentations throughout the weekend, ranging from lipoprotein(a), to lipid disorders (I find that understanding how things look when they’re going wrong can help me figure out how they should work when they’re going right), and intermittent fasting was even mentioned during a presentation regarding dietary habits! There was also a section of vendors at the conference, including ones that offered genetic testing, and more in-depth lipid testing which I found intriguing. I’ll definitely have to do a little more research into those sometime in the future.

It was honestly a bit surreal to walk past people having a conversation and hear them discussing LDL, particle counts, and other lipid-y things in passing – it’s not often I get to experience that in person! I also got the opportunity to meet some interesting people, from clinicians, to nurses, and dietitians. I was also delighted to see that one of the complimentary goodies the conference was offering its attendees was the most recent volume of The Journal of Clinical Lipidology. I got to have a bit of a study session with Dave before the conference day started in earnest as we both read through the studies it contained. It ended up being pretty productive and (dare I say it) fun.

Over all, it was a great learning experience, and I’m glad I decided to go.

On the Experiment

Chow time at Clinical Lipid Update

I’d say, in this case, the inconvenience of having to drink 6 shakes a day was far outweighed by the data I got in the process. Lipoprotein(a) coming from a baseline diet to a high calorie/high fat phase provided some useful information that – of course – leaves me with even more questions, and possible future experiments in mind, for sure. Plus, it wasn’t too bad, as the very kind staff at the hotel where the conference was taking place offered to store my shakes in the front office fridge so they could stay refrigerated until I needed them. I’m extremely grateful, as this made the whole process much less of a hassle, and allowed it to go as smoothly as it did. The food served at the conference actually did look quite appetizing, but in this case the sacrifices of citizen science won out over the freshly carved meat they were serving.

Even with that said, I must say that my diet over the conference weekend was definitely more appetizing than what Dave was eating – as he expressed multiple times! I’m sure he’ll be mentioning that himself, however, in part 2….

To be Continued in Part 2 – Stay tuned!

Oct 17

Attending #NLANashville and the Secret Tandem Experiment

Siobhan and I flew to Nashville to attend the Clinical Lipid Update put on by the National Lipid Association. We were both very excited to be hanging with like-minded lipid geeks, even if our perspectives might vary a little. We managed to make some new friends, get interesting perspectives and learned a few new things in the lectures that have us looking into some research we might not have come across otherwise.

The Organization and Staff Were Super Friendly

My one biggest worry was that NLA staff might not receive me well given I might be considered a more controversial figure with regard to cholesterol (certainly I get that feeling in social media). But nothing could have been further from the truth. Before registering, I worked closely with Brandi Rawls, the NLA Membership Manager to get the details settled. And upon arriving, I had a number of questions for their Project Coordinator,  Chealsea Schroeder, who was all too happy to answer. All in all, it was a lot of smiles and well-wishing.

Project Coordinator Chealsea Schroeder

Meeting the Luminaries of Lipidology

We got to meet quite a few Big Names in the field.

While not all are fans of the Low Carb / Ketogenic lifestyle, one particular figure Sarah Hallberg wanted me to meet was Greg Pokrywka of the Baltimore Lipid Center. He had some great suggestions on next steps for us in our research and I was quite honored to have the chat.

I had a number of questions for many who gave presentations, but I decided to forego the public Q&A to instead approach them with my query privately following the session.

I was especially interested in the blood and non-blood tests taken for patient cases. How much do they focus on Lp(a), free fatty acids, and CTs like CAC? Overall, quite a bit, actually. I also asked about fasting insulin and C-Peptide, but hadn’t found anyone who actively used it before.

Final Thoughts

Overall, Siobhan and I had a positive experience and especially enjoyed the last day which had a focus on CAC’s value and a couple things we didn’t know about like Lipoprotein X.

We also wrapped an experiment around the event: the Tandem Drop Experiment. But more on that in the next post…

Oct 15

Dave and Ivor Cummins Discuss The Recent Attia Podcast


If you haven’t already tuned in to the recent sit down between Dave and Ivor Cummins, here’s your reminder to do so! They share their thoughts during the aftermath of Dave’s recent appearance on Peter Attia’s podcast, including their thoughts on mass balance, root causes, and more.


Oct 08

Guesting on the Peter Attia Drive – Response to Peter’s Prebuttal

Peter included a prebuttal at the beginning of his podcast with me that included three central points. I’ll take them one by one.

The first:

Dave was unable to explain the mass balance, meaning how does one account for the greater amount of cholesterol in, and the greater number of, LDL particles. No one, including Dave, is disputing that the phenotype of interest has more LDL-C and more LDL-P. There are only 3 ways this can happen (these are [collectively exhaustive, but not [mutually exclusive]): make more, clear less, transfer from other pools that we can’t see (e.g., cell membranes). I think the data make the first of these by far the most likely driver, but Dave seemed unable to address this and could not explain, to me at least, what could account for this increase in LDL-P/C. So on first principles, my doubt of this model has gone up from the start of this discussion, as the person who developed the model could not actually explain the mass balance. This is one of the most fundamental requirements of any model. And to be clear, even if this fundamental condition were met, it would not be sufficient to make the case that [lean mass hyper-responders or LMHRs] are not at risk. It’s a [necessary but not sufficient] criteria that, in my mind, has failed.

Let me first start with the analogy I used in the podcast of boats/life rafts and walk it all the way through:

  1. There’s a harbor with 100 boats that are regularly leaving to deliver cargo to a far away island before returning again. Each boat has 10 life rafts onboard. Thus, we have 100 boats and 1,000 life rafts “in circulation”. There’s no boats being removed or life rafts taken off in this example for illustration purposes.
  2. There was first a period where they were built that required more mass (materials for building = mass). But once in circulation, there’s no requirement for further addition mass.
  3. After a while, the demand changes on the island and more cargo is needed to be delivered, this will require 500 boats. Construction material for these boats are required, but just enough for 400 more. And more life rafts are required, but just 4,000 more. More mass is needed in this phase to meet demand for more boats in circulation. But once in circulation, there’s no more mass needed.

So yes, there’s both more creation of the boats and recirculation of the boats. The problem comes when trying to determine how much there is of each from only seeing them in the water alone. One can’t look at the 100 boats (or their life rafts) or the 500 boats (or their life rafts) while in circulation and draw a conclusion on mass balance. Indeed, in this example, once each batch is built, there’s no further change in mass balance, only ongoing recirculation.

I don’t think Peter disputes the liver’s cholesterol pool is fungible and that lipoproteins and their cholesterol cargo are continually reabsorbed, rebuilt and/or reused. There’s still much being learned as to how much, but I doubt anyone would argue it’s very dynamic. But the recirculation is especially important to acknowledge as a relevant component of the overall metric.

Hence, this is why I eventually asked this question of Peter during the podcast when we seemed stuck on this point:

“How much of this cholesterol has already made a lap? Are you thinking of this in terms of it all getting synthesized and then reabsorbed, and then recreated it again?”

If I’m understanding his position correctly — he believes either that lipoprotein cholesterol isn’t making any “laps” (re-sent again in another lipoprotein) or that it is a fixed number of total trips. How else can we not assume this is relevant toward the total balance?

In other words, this suggests the lipid pool within the lipid system is not ultimately fungible from a lipoprotein transport perspective, and that reuse/recirculation is actually limited in a predicable way.

If so, this is certainly a point we’d disagree on.


<Special note: If Peter would like to add clarification or rebuttal to the above response, I’ll include it here>

The second:

Dave argues that VLDL production is driving the LDL concentration, but the fact remains that in insulin-sensitive people (which presumably the LMHRs are), the opposite is true: there are fewer, not more, TG being exported from the liver and there is less, not more, apoC-III on the VLDL, thereby reducing, not increasing, their residence time. In other words, LMHR would have less VLDL to LDL conversion than, say, someone with T2D. So again, I can see no evidence whatsoever that his energy model, which can’t be explained on mass balance, and can’t be explained on what is known about the physiology of VLDL and LDL, is plausible.

As per the podcast, I was very up front on this being theoretical. Indeed, I qualify this very early:

Dave Feldman

“With that in mind, here’s what I would speculate. This is purely hypothetical, but I would speculate if you were to grab a whole bunch of people who are … We’ll hopefully get into this model that I’m talking about, that I call lean mass hyper-responders. People who are at the far end of the spectrum, they are athletic, they are thin, and they are very, very low carb, and therefore see very high levels of LDL-C and LDL-P, but they also have very high levels of HDL-C and low levels of triglycerides.

I suspect that they would show a very high rate, proportionally, of VLDL secretion, that they actually are trafficking a lot more, for their energy, triglycerides in VLDL particles, and therefore have succeeding LDL particles as to the explanation as to why their LDL-C and LDL-P would be higher.”

What I’m more interested in is how one can be confident in the opposite. How do we know there are fewer TG “being exported” from the liver in healthy hyper-responders, for example? If we can be, than he’s right, that would put a dent in my theory.

But don’t confuse a blood test for TG as an indicator of TG secretion by the liver. Again, you have to realize the other side of the ledger with uptake by tissues. This is why the three lowest TG blood tests I’ve ever had were following half and full marathons. We wouldn’t confuse glucose entering the bloodstream with glucose found in a blood test for the very same reason — uptake matters.

So without question, I’m extremely interested in any kinetic studies that would involve capturing TG secretion from the liver (aboard lipoproteins, of course) in payload comparisons between healthy hyper-responders vs healthy controls. And I definitely speculate the former would have higher overall TG secretion, yet likely higher TG use as well. But the “export” as Peter puts it, is certainly the key.

<Special note: If Peter would like to add clarification or rebuttal to the above response, I’ll include it here>

The third:

Even if you ignore the points above—which you can’t—I am more unconvinced than ever at the notion that we should exclude the roughly 2,000 genetic mutations known to produce a phenotype of high LDL-C, high HDL-C, and low TG. We have 2,000 natural experiments. Surely at least some of these cases (e.g., PCSK9 gain of function) are excellent proxies for the key features of LMHR. Yet to ignore them for imaginary reasons (e.g., having gain of function PCSK9 is somehow toxic to endothelial cells because it impairs their ability to take up cholesterol despite there being no evidence that endothelial cells require PCSK9 to uptake LDL in a receptor-mediated fashion) is to say, in my opinion, one does not want to know the answer to this question.

Before answering this, I’m reaching out to Peter to get clarification on his first sentence. I believe he honestly misspoke on identifying the triad that I’m interested in (high LDL-C, high HDL-C, and low TG) and instead simply meant LDL-C alone. To my knowledge, there’s no study that identifies the phenotype of that triad I’m interested in. (But if you know of one, please contact me!)

Oct 08

Guesting on the Peter Attia Drive (5 of 5) – Comments and Featured Thoughts

This final posting in the series is devoted entirely to discussion in the comments on the podcast. Share your opinions in the comments down below on anything and everything about this podcast from your perspective.

I’ll be featuring a number of these comments in this space as well. The likelihood of your comment being featured  doesn’t need to align to any particular position, but I do ask that it be respectful and without ad hominem toward anyone in or outside this podcast. (Attack ideas, not people)

Older posts «