Thread by @DaveKeto

1/ Thank you for the invitation, @chadinabhan.

I've now listened to the episode and was a bit surprised at a few points, but also less so with others.

Cholesterol is a very emotionally charged topic, and understandably so...

https://twitter.com/chadinabhan/status/1407419727962574855

2/ For me, this journey started 6 years ago with alarmingly high total and LDL cholesterol following my going on a ketogenic diet. I became obsessed with trying to understand why and begin reading everything I could on lipidology...

3/ I found through a series of experiments there was quite a bit of change I could induce based on dietary patterns. As I developed and executed this "citizen science" research, I turned it around back to the community to hopefully help us in advancing this important topic.

4/ Ex. -- probably the most talked about of these was my White Bread experiment (LDL 296->83 in 7 days) where I intentionally had a diet no one recommends to demonstrate it was more likely carbs v fat, not plant v meat, or nutritious v non-nutritious, etc. cholesterolcode.com/the-tandem-dro…

5/ This "metabolic influence of lipids" shouldn't be controversial, of course, there's plenty of literature looking it how metabolism has impacted lipids for decades. Not just with diet, but also with fasting as well.

(ie See my presentation )

6/ If you want a bit of a deeper dive on this #LipidEnergyModel we're working on, you can check out my presentation to Stanford here:

However, while the model may or may not explain why, does it matter if high LDL-C is of high risk regardless?

7/ It absolutely could be. When discussing risk where LDL-C may be higher due to VLDL-TG turnover (per the model), I often say I'm "cautiously optimistic". But those two words are meant to convey a leaning, not certainty.

8/ But a key part of this model in looking to the success or failure of VLDL-TG turnover is in whether it likewise results in higher HDL-C and lower triglycerides.

(I know that got a tiny bit technical, but it's relevant with regard to downstream influence on this lipid profile)

9/ The combination of these three together I call the "low carb lipid triad": High LDL-C, High HDL-C, and low triglycerides.

And this is an *extremely* common outcome, particularly for many seemingly metabolically healthy on keto.

https://twitter.com/DaveKeto/status/1317651445194260480?s=20

10/ Conversely, both low HDL-C and high TG are already criteria for Metabolic Syndrome (MetSyn) and Atherogenic Dyslipidemia (AD), each of which are substantially associated with cardiovascular disease (CVD) risk.

Put another way, the LC triad as almost the opposite of AD.

11/ For me, the obvious next step is to look to existing data we have with this triad and see how strongly it associates with CVD.

Simply stated, do we see an association of high LDL with high CVD where there's absence of MetSyn/AD?

But this is where a lot of pushback starts...

12/ In spite of this case for looking to metabolic health and disease and how it impacts lipid profiles, many feel there's more then enough evidence where looking to LDL-C/ApoB in isolation, and that greater specificity is unnecessary. (I respectfully disagree.)

13/ Which is why this divide has emerged. So many in the low carb community would say, "Yes, of course metabolism has impacts on all these lipid levels, it just happened to me (or my brother, my mom, etc)." And thus, this interest toward metabolic causality makes a lot of sense..

14/ Conversely, many others who disagree are very adamant that these markers should be looked at piecemeal instead of together.

And when doing so, you can see an association with LDL and CVD -- and given this less specific approach, I'd agree.

15/ For example, this is a great study that @ethanjweiss tweeted recently which I really love given its very recent comprehensive review on adipocyte hypertrophy research.

Note these scatterplots looking to lipid levels, particularly with visceral fat.

16/ HDL goes down, TG and LDL go up. I'm sure most who research lipids would agree the changes in HDL and TG are more likely a consequence, and not a cause of this metabolically compromised state. (Which I agree with)

But why wouldn't we consider this of LDL as well?

17/ Many including Dr Dayspring point to how low HDL and high TG is typically a proxy for higher ApoB. And yes, it usually is (but not always).

But the fact this is usually case should be the very reason we have much more curiosity in examining its counterpart.

18/ To make a claim that A causes B, we need to likewise know that B does not cause A, or to what degree it does we can account for.

But we also need to confirm there isn't a C, D, E, F, etc that is causing both. (Which is where Bradford Hill comes in very useful)

18/ Again, this can be looked at in existing datasets.

We have so much available now with Framingham, Copenhagen, UK Biobank, etc -- and many of these are longitudinal.

We can not only look to the triad in these data, we could observe its outcome association over time.

19/ Unfortunately, the nuance I explained above is often lost, which is a big part of the problem.

Getting others to consider why metabolism influences lipid profiles and how this is relevant to atherosclerosis is actually much more challenging than I originally assumed...

20/ More commonly, my position is oversimplified to "LDL has nothing to do with heart disease" or some such. And then responses or arguments are returned in kind.

(In fact, I wouldn't be surprised if this thread is no different and likewise gets single lipid study responses)

21/ Fortunately, we're working really hard on a clinical trial with those have the most pronounced version of this triad, Lean Mass Hyper-responders. cholesterolcode.com/lmhr/ It's been taking a while to get into place, but we're currently in IRB and hope to be launching soon

22/ I hypothesize we'll see #LMHRs demonstrating a low progression of atherosclerosis at a population level in spite of having LDL in the top 3% (and consistent with FH levels). But I truly don't know and hope to help us get this data ASAP.

23/ If they do have low progression of atherosclerosis and it is ultimately agreed in time that #LMHRs are low risk (at a population level), then it could go a long way at advancing this distinction of metabolic context with regard to resulting lipid profiles and CVD.

Addendum— in rereading, I realized I didn’t actually link the study page itself.

Obligatory plug👇
citizensciencefoundation.org/campaigns/lean…