Saturday, June 02, 2012

Cholesterol: More epidemiology

Too hilarious not to comment on!

Lipitor because your TC "number" is "bad"?.

Ah you can say, but what about psLDL? They never measured psLDL! Mmmm, sucrose...

Thanks to Karl for the heads up

Peter

28 comments:

. said...

Good to see you're still around. Miss your posts

-- Chainey

WilliamS said...

Peter, sometime I'd love to hear your take on the elite lipidologists' current view that LDL-P (lipoprotein particle number)—not any of the standard cholesterol measures—is the real risk marker and also causal for CVD.

At least the dietary strategy for lowering LDL-P is low carb instead of low fat. But then there's also the statins, niacin, etc…

As you may know, Gary Taubes' new partner in crime has been forcefully arguing this position of late:

http://eatingacademy.com/

One of those elite lipidologists argues the case here (registration required). You won't mind watching her do so...

bit.ly/K22VPz

Peter said...

Hi Chainey, there's just so much to do. And a lot of it's not nutrition so never gets near the blog. Like the natural nuclear reactors in Oklo on earth, well there appears to have been one on Mars too. It converted to a breeder reactor, generated enough plutonium to support a prompt criticality and went bang. The hole was 4km deep and 400 across. Fascinating but irrelevant! A bit like the arguments raging about the LRNT hypothesis of radiation exposure, Chernobyl, Fukushima and the Taiwan Co60 contaminate flats... Never mind, more of nutrition to come.

William, I watched most of part1. The lady believes particle number is the problem. I believe hyperglycaemia is. You have to be careful with belief structures, they tend to produce very biased outlooks!

A number of us LCers have TCs way over 8mmol/l. Even if this is made up of astronomically diametered LDL particles I doubt there is a very low apoB particle count, so I guess I need a lipidologist... Ah well, doomed again.

Peter

IcedCoffee said...

I generally like the particle number idea, but I feel that its incomplete.

High LDL-P will only be a problem if those particles are subsequently traveling through the endothelial wall and remain there. This means to me that absent endothetial damage, high particle number isn't inherently deleterious, as there wont be any plaque formation. So to twist this in a more hyperglycaemic view, its the damage caused by hyperglycaemia to the endothelial wall that prompts the elevate particle number to become damaging.

I also wonder about the effect of hyperglycaemia on particle number. If memory serves, ApoB proteins are rather susceptible to glycation, which impairs the ability of the body to clear these lipoproteins. So chronic hyperglycaemia would increase the number of circulating lipoproteins, which would circulate for longer periods of time, becoming smaller and denser. This increased time in circulation would also deplete their protective antioxidants, making them more prone to oxidation.

So I think both arguments are fundamentally right; the issues arise when you start looking at one without the other. What happens when you have increased particle count without arterial damage? Or what happens when you have hyperglycaemia, but insufficient LDL-P to promote significant plaque development?

(Worth noting this is mostly speculative, I wont pretend to be an elite lipidologist)

Sam Knox said...

@WilliamS

I watched the entire series. She seems to favor lifestyle changes (exercise, low-carb diet) over medication, and then metformin over statins. One patient she didn't treat with drugs at all, and another she eliminated several drugs and replaced them with omega-3 and niacin.

I like her.

LeonRover said...

Hey Petro,

C S Lewis titled a book "Mere Christianity" - it was meant to be ironic.

Why not re-title this post:"Mere Epidemiology" - this might regarded as post-ironic.

Ta 4 post, will you be giving us any "Tales of a Rural Vet" anytime?

Slainte.

WilliamS said...

"You have to be careful with belief structures, they tend to produce very biased outlooks!"

Indeed. Peter Attia, Gary Tabues's new partner, has promised to present the evidence supporting the LDL-P theory of CVD and its prevention. I'll pass it along if he does.

I find Gary Taubes's involvement in all this interesting. Not only his close connection with Peter—he's been interviewed with Thomas Dayspring, another lipidologist of the same ilk, with both sounding in agreement with other about the whole subject. However, I'm not aware that Gary has specifically stated his take on this. I plan to ask Peter Attia that on his blog at the right moment.

I'm in your boat: LDL-C so high that it seems doubtful LDL-P could be as low as these folks recommend. On the other hand, Dayspring, Dall, and Attia have all said that in their experience it is impossible to confidently predict LDL-P from a standard cholesterol panel. Dayspring has said anyone who thinks they can is delusional.

Ian said...

I agree with Chainey, your posts are great and I really like the sound of more nutrition to come!

karl said...

Here is the rub. The correlation of LDL with heart disease is not so good. Is the weak correlation due to the e4/e4 folks that seem to have a special problem? (don’t know )

If we look at oxLDL, the correlation is better - oxLDL somewhat tracks LDL (there is more of it to oxidize) - but the correlation of CAD( Coronary Artery Disease) is much better for oxLDL than LDL.

The established med folks are thinking that small LDL ( sLDL ) is more oxidizable, but they forget that both sLDL and oxLDL go up with BG. I've seen lots of experiments designed to keep the grant money flowing, but non that remove the effects of BG from the levels of oxLDL in testing interventions.

Do people with normal Postprandial BG get high values of oxLDL? We don't know - only OBGNY docs mess with glucose tolerance - the rest only test fasting BG ( which IMO only tells you about the problem after it is too late ). Postprandial BG of over 110 is probably a bad thing IMO.

So what value are the expensive NMR and ultra-centrifuge cholesterol tests? Are they more useful than reading tea leaves? What changes in medical interventions should be based on these tests? Are they really a huge waste of money?

My take is they should be testing oxLDL ( a cheap test ) - but they don't - instead they are still telling people their total chol levels - a completely useless number. There are simple dietary interventions that will lower oxLDL - and this might well be due to lower BG - the detailed research to test this just hasn’t been done.

We now know that there are over 100 forms of HDL - some apparently good - others bad - and a similar take on LDL. But both LDL and HDL are actually part of the innate immune system - which we evolved for a reason other than inducing CAD. Taking a statin will lower LDL - but there are about 10 other effects of statins that appear to reduce systemic inflammation. Other LDL lowering drugs lower LDL, but don’t appear to have much effect on CAD.

Then there is the trygly number - healthy athletes will have a number around 50mg/dL. Trygly is also correlated with CAD and we have an easy ways to lower this number - don’t eat fructose containing sugars ( not patent here so mostly ignored by drug-salesmen-trained MDs ).

In the 1960s Milton Wintz developed a synthetic diet for astronauts ( Vivonex100 ). Pure amino acids, minerals, essential nutrients and glucose (no waste is wanted in space). They tested it on prison volunteers so compliance was not an issue. They complained it had a boring taste, so they substituted sucrose for the glucose - and the trygly spiked. This was a good experiment - provided an important clue - and was promptly ignored as it did not fit the fats-R-bad mantra.

So a high t-cholesterol number is due mostly to trygly and seems to be controlled by sugar intake. YET we replaced fats in our diets in the next years with more sugar - and got more CAD thanks to farm vote buying political pressures that urged us to eat more and more carbs.

We have ended up with standard medical practice based on treating blood-test numbers rather than the disease. In the end I am guessing we will find that CAD is an auto-immune disease and has little to do with fat metabolism.

If I’m right, then it is only the inflammation modulating effects of statins that do some good. Low-carb diets - low in O-6 - might be just as helpful to lower systemic inflammation - perhaps C60 might also help. When will the cholesterol theory of CAD be pronounced dead?

Stan (Heretic) said...

Re: Oklo on Mars

Fascinating, then it must be more uranium there in the mantle than on Earth, probably. Can you post a link? Space research physics was my previous life before 1988. My last project (participation in) was high energy solar particles detector carried on Phobos II (ESA/SU/NASA). When the spacecraft was shot days after making it into Marsian orbit they never told us anything. I was disgusted. I accidentally found out what (probably) happened when a Soviet scientist (principal investigator) gave an interview to BBC a year after.

Peter said...

Stan, absolutely. They think it came from a meteorite load and it stayed in tact through there being no plate tectonics. About a quarter of a cubic kilometer is estimated to have been involved in the prompt criticality. I'm guessing there is a much bigger motherload below the crater and only to top section was water moderated in thermal mode to generate the high enrichment needed for the hard neutron criticality which exploded. It's a great read:

http://www.lpi.usra.edu/meetings/lpsc2011/pdf/1097.pdf

I picked up where a large chunk of uranium in a meteorite might have come from in Narlikar's book Violent Phenomena in the Universe, another of those books you pick up as a youngster and can't put down...

Email the Mars probe scandal off blog please!

Re lipidologists. I like Peter Attia's site but obviously have trouble with his belief structure on LDL-C. My problem is I've read so much bollocks from lipidologists in the past you just get past caring or being able to make the intellectual effort when THIS time they've gotten it right. Especially when THIS time involves chronic normoglycaemia. Dr Davis comes to mind, I'm assuming he'e still saturophobic and getting results from normoglycaemia?

Peter

Nigel Kinbrum said...

RE Hyperglycaemia and Cardiovascular Events: In Fasting Compared With Nonfasting Triglycerides and Risk of Cardiovascular Events in Women, it states (just below Table 4.):-
"The inclusion of other potential confounders, including alcohol consumption, exercise, and level of glycated hemoglobin, had no substantive impact on these results."
This suggests that fasting hyperglycaemia is not a significant contributory factor to Cardiovascular Events. Nonfasting hyperglycaemia could be a major contributory factor however, just as nonfasting hypertriglyceridaemia appears to be.

Stan (Heretic) said...

Nigel Kinbrum wrote: This suggests that fasting hyperglycaemia is not a significant contributory factor to Cardiovascular Events. Nonfasting hyperglycaemia could be a major contributory factor however, just as nonfasting hypertriglyceridaemia appears to be.

Interesting!

That may indicate indirectly that the hyper-insulinemia may be the main primary factor!

Peter said...

Stan and Nigel,

I am being pushed towards the idea that FAILURE of insulin resistance may be the problem. Glucose+insulin in Veech's isolated heart model was as good as ketones for metabolic efficiency but by differing mechanisms, ie glucose with a decent dose of insulin is safe because insulin adjusts the ETC function to maximise efficiency. Ketones appear to do this by simply lowering the inner membrane voltage so reduce proton leakage, even when NADH is high from ACoA metabolism. In those days no one knew how insulin does this, but it does. Certainly not by reducing the IMM potential. But glucose without insulin, what does that do? Bad Things methinks. We know GLUT4 can be sidestepped by hyperglycaemia acting on GLUT1 and this would fit well with post prandial hyperglycaemia (with inadequate insulin) being the problem, as Jenny Ruhl has idenified for years now. This would mean that hyperinsulinaemia is protective if it can allow "correct" control of the ETC per unit glucose metabolised. A free flow of glucose through GLUT1s without insulin would probably generate apoptotic levels of free radicals. My guess, from Jenny's site, is 140mg/dl or 7mmolish/l is where it starts... I'd bet fasting glucose is a poor surrogate for post prandial hyperglycaemia.

Oh, and fructose = glucose without the insulin...

Lipids????? You have to patch the damage up with something, let's shoot the messenger. Especially Lp(a).

Have to get on to Veech's papers!

Peter

Peter said...

Leon, probably not except my face is recovering quite well from the attempted hemimandibulectomy by a particularly unpleasant dog last week!

Peter

karl said...

Peter wrote:
Dr Davis comes to mind, I'm assuming he'e still saturophobic and getting results from normoglycaemia?

There are more than a few CAD patients that get there because they have the e4/e4 gene type - those unlucky folks may do better if they avoid sat-fats. I've seen him point out that for people with high Lp(a) that sat-fats actually appear to lower the level. (Whether this results in positive outcomes has not been tested).

RE: Oh, and fructose = glucose without the insulin...

Not sure what you mean? - fructose is a 5-carbon sugar so that when circulating appears to have a much higher chance of forming AGEs(Advanced Glycation Endproducts). Fructose also goes down a different metabolic path that spikes trygly more than glucose ..

You might want to look at the Milton Winitz papers:
http://www.ajcn.org/cgi/reprint/23/5/546.pdf

A paper I would love to have a copy of (hint hint) is here - 1964 EFFECT OF DIETARY CARBOHYDRATE ON SERUM CHOLESTEROL LEVELS.

It could be that the article was in Nature 1965 Evaluation of Chemical Diets as Nutrition for Man-in-Space

Peter said...

Karl, I'm viewing the presentation of large amounts of glucose derived pyruvate to the mitochondria as causing minimal problems when accompanied by the changes to the ETC from insulin and diversion of the excess glucose to glycogen, also directed by insulin. Fructose runs to pyruvate without insulin's effects and without insulin's control and has to go through the mitochondria before being converted to lipid. Or be off loaded as lactate (both happen). The question is what happens at complex I under fructose overload. Developing insulin resistance may limit the entrance of glucose to the cell through GLUT4s, and inhibit lipolysis to limit fatty acid supply, but I don't see that it can limit fructose ingression.

Was the paper this one?

I had a look at it here and liked it.

Peter

Peter said...

Hi Karl,

Just followed the link from your email, same study. Would you mind if I put up a one line post pointing to your web page?

Ta in advance

Peter

karl said...

http://wiki.xtronics.com/index.php/History_of_Fructose_Carbs_and_weight-loss

A bit of background - Robinson used to work with Pauling and it was Robinsons research that failed to show that vit C was a panacea - so Pauling destroyed the results. ( Pauling also destroyed the results of his other work with mass-spec on urine - that could be bearing fruit today.. )

His later work has to do with biological clocks - deamination -

See -
http://scholar.google.com/scholar?hl=en&q=Robinson+AB+&btnG=&as_sdt=1%2C26&as_sdtp=

Digging into much of the dietary work is maddening - what is going on is utterly complex and everything seems to effect everything else - yet they rely on correlative evidence and assume causation! ( And then they distort the next study out of confirmation bias!) What I thought was known turns out to be opinion - I'm doubting more and more as time goes by. Pretty soon I won't know anything..

The cholesterol theory ( which Robinson has had doubts about for years according to his letter ) appears to be based on a weak correlation that may well be limited to those unfortunate e4/e4 folks. There are some workers that seem to understand that while LDL might move fats around - it may have a first duty as part of the immune system. My hunch is the oxidation of LDL is an activation step. oxLDL recruits monocytes that become macrophages - inappropriately in the case of artery walls. The statins block LOX-1 stopping this reaction - that they lower LDL may just be a side effect.

This 'reuse' of bits of biochemistry for wide ranging functions is a byproduct of evolution - an engineer would make the right tool for the job - while natural selection ends up with hammers being used for doorstops and forks get used for bookmarks and letter openers.

When a protein is used for 2 or more purposes - evolutionary mutations may cause problems for one function while aiding another - untangling the actual way things work becomes insanely complex.

Research that detects the arrow of causation seems much too rare these days.

Robinson is a good scientist for who's judgment I have great respect, yet he can't let go of creationism! I'm not sure how he does it - if I look at an eye, as an EE, I notice that the power supply cables(blood vessels) are on the wrong side of the receptors - not possibly intelligently designed - then I see the inappropriate reuse of proteins - it is clear that life was not 'engineered', but evolved. So Robinson is a puzzle to me - yet he thinks so clearly in other places and is a really good scientist..

Ken said...

Peter,
In support of your idea about "FAILURE of insulin resistance", consider "isolated fasting hyperglycemia". I (and Jenny Ruhl's brother) have this, and pp-BG does not exceed 140mg/dL (in either of us). The following study by DeFronzo et al is important, I think: http://www.ncbi.nlm.nih.gov/pubmed/18000182
They suggest mass-action (GLUT1) causes some or all of the progressively decreasing skeletal-muscle basal glucose uptake in i-IFG. If so, I suggest this is part of a systemic protective adaptation to prevent glucose toxicity in skeletal muscle. Similar things are probably going on in other tissues, as you seem to suggest. I like your idea. The functionality of the endothelium seems to be key in CVD.
BTW, my LDL-c = ~250mg/dL. I just had a coronary CT -- zero calcified plaques. 53 years old. LCHF for 2 1/2 years.

Ken said...

P.S.
In i-IFG (isolated impaired fasting glucose) there is early degradation of insulin secretion and normal insulin sensitivity (my own is quite supra-normal). Most are unfamiliar with i-IFG. I am trying to reverse it, with some hints of success. But it appears to be a slow process. Beta cells regenerate very slowly.
My point is this -- my body seems to know I still have a beta-cell deficit.
Mitochondrial dysfunction may likely be involved in the beta cells, and in skeletal muscles. I have read papers suggesting the same for endothelial cells as well.

Peter said...

Ken,

Very interesting. I think I'll go to the paper about insulin resistance as an antioxidant defence mechanism before Veech and the ETC, although Veech has good pointers as to why the free radicals might be generated (or not). We're getting a little closer to what insulin resistance actually is...

Peter

Hasan Hannachi said...
This comment has been removed by the author.
Hasan Hannachi said...

It may be worth examining LDL-P and its relationship to the “War on Insulin" ala Attia. Rather than being a consequence of too much insulin, large numbers of LDL particles is largely a result of insulin not being able to do its job property; i.e. insulin resistance. Through mechanisms I won’t detail here, insulin resistance results in elevated triglycerides generating large numbers of small, LDL particles and a corresponding decrease in large particles. The net result is an increase in particle number (LDL-P) without necessarily an increase in cholesterol concentration (LDL-C). (For more detail see http://clinical.diabetesjournals.org/content/26/1/8.full)

Since insulin resistance is a feature of Metabolic Syndrome, it therefore makes sense that people with MetSyn would be characterized by high LDL-P without necessarily high LDL-C and that is what is being called “discordance.” However, MetSyn is also associated with high triglycerides and low HDL and, as already noted, when those factors were controlled for, the predictive power of LDL-P for CVD was eliminated (see http://www.theheart.org/article/767865.do). In other words, when you take MetSyn out of the picture, LDL-P doesn’t seem so important and if it is picking up risk, it is through the same mechanisms as MetSyn.

That still leaves the question as to whether or not the increase in small LDL-P, essentially the reason for the increase in LDL-P, is causing CVD and a number of hypothesis have been suggested. For example, some think that small particles are more easily oxidized and others that they can more easily penetrate the arterial wall. However, when researchers stratified subjects with MetSyn by small LDL-P, they found no association with CVD suggesting that other mechanisms part of MetSyn are the culprits (inflammation, glycation, ???). (See http://circ.ahajournals.org/content/113/1/20.full.pdf for more detail.)

Bottom line is that insulin does not cause either high LDL-P nor does high LDL-P cause CVD. High LDL-P and discordance with LDL-C result from the insulin resistance that is a feature of Metabolic Syndrome and as such, discordance is a marker for MetSyn. It would seem that restoring insulin sensitivity rather than declaring a “War on Insulin” is the order of the day. If somebody suggests medication based on LDL-P in healthy people, please ask them to produce the randomly controlled trials which showed meaningful reduction in hard, clinical endpoints from such treatment (NNT- numbers need to treat). Otherwise, consider any cholesterol measurement to be a risk marker at best.

Calling for everybody to have an NMR and to be medicated for "high LDL-P" would seem to be another in the long line of attempt to enlarge the pool of statin customers, not to mention the market for NMR testing.

Hasan Hannachi said...
This comment has been removed by the author.
Hasan Hannachi said...

It may be worth examining LDL-P and its relationship to the “War on Insulin" ala Attia. Rather than being a consequence of too much insulin, large numbers of LDL particles is largely a result of insulin not being able to do its job property; i.e. insulin resistance. Through mechanisms I won’t detail here, insulin resistance results in elevated triglycerides generating large numbers of small, LDL particles and a corresponding decrease in large particles. The net result is an increase in particle number (LDL-P) without necessarily an increase in cholesterol concentration (LDL-C). (For more detail see http://clinical.diabetesjournals.org/content/26/1/8.full)

Since insulin resistance is a feature of Metabolic Syndrome, it therefore makes sense that people with MetSyn would be characterized by high LDL-P without necessarily high LDL-C and that is what is being called “discordance.” However, MetSyn is also associated with high triglycerides and low HDL and, as already noted, when those factors were controlled for, the predictive power of LDL-P for CVD was eliminated (see http://www.theheart.org/article/767865.do). In other words, when you take MetSyn out of the picture, LDL-P doesn’t seem so important and if it is picking up risk, it is through the same mechanisms as MetSyn.

That still leaves the question as to whether or not the increase in small LDL-P, essentially the reason for the increase in LDL-P, is causing CVD and a number of hypothesis have been suggested. For example, some think that small particles are more easily oxidized and others that they can more easily penetrate the arterial wall. However, when researchers stratified subjects with MetSyn by small LDL-P, they found no association with CVD suggesting that other mechanisms part of MetSyn are the culprits (inflammation, glycation, ???). (See http://circ.ahajournals.org/content/113/1/20.full.pdf for more detail.)

Bottom line is that insulin does not cause either high LDL-P nor does high LDL-P cause CVD. High LDL-P and discordance with LDL-C result from the insulin resistance that is a feature of Metabolic Syndrome and as such, discordance is a marker for MetSyn. It would seem that restoring insulin sensitivity rather than declaring a “War on Insulin” is the order of the day. If somebody suggests medication based on LDL-P in healthy people, please ask them to produce the randomly controlled trials which showed meaningful reduction in hard, clinical endpoints from such treatment (NNT- numbers need to treat). Otherwise, consider any cholesterol measurement to be a risk marker at best.

Calling for everybody to have an NMR and to be medicated for "high LDL-P" would seem to be another in the long line of attempt to enlarge the pool of statin customers, not to mention the market for NMR testing.

Hasan Hannachi said...

It may be worth examining LDL-P and its relationship to the “War on Insulin" ala Attia. Rather than being a consequence of too much insulin, large numbers of LDL particles is largely a result of insulin not being able to do its job property; i.e. insulin resistance. Through mechanisms I won’t detail here, insulin resistance results in elevated triglycerides generating large numbers of small, LDL particles and a corresponding decrease in large particles. The net result is an increase in particle number (LDL-P) without necessarily an increase in cholesterol concentration (LDL-C). (For more detail see http://clinical.diabetesjournals.org/content/26/1/8.full)

Since insulin resistance is a feature of Metabolic Syndrome, it therefore makes sense that people with MetSyn would be characterized by high LDL-P without necessarily high LDL-C and that is what is being called “discordance.” However, MetSyn is also associated with high triglycerides and low HDL and, as already noted, when those factors were controlled for, the predictive power of LDL-P for CVD was eliminated (see http://www.theheart.org/article/767865.do). In other words, when you take MetSyn out of the picture, LDL-P doesn’t seem so important and if it is picking up risk, it is through the same mechanisms as MetSyn.

That still leaves the question as to whether or not the increase in small LDL-P, essentially the reason for the increase in LDL-P, is causing CVD and a number of hypothesis have been suggested. For example, some think that small particles are more easily oxidized and others that they can more easily penetrate the arterial wall. However, when researchers stratified subjects with MetSyn by small LDL-P, they found no association with CVD suggesting that other mechanisms part of MetSyn are the culprits (inflammation, glycation, ???). (See http://circ.ahajournals.org/content/113/1/20.full.pdf for more detail.)

Bottom line is that insulin does not cause either high LDL-P nor does high LDL-P cause CVD. High LDL-P and discordance with LDL-C result from the insulin resistance that is a feature of Metabolic Syndrome and as such, discordance is a marker for MetSyn. It would seem that restoring insulin sensitivity rather than declaring a “War on Insulin” is the order of the day. If somebody suggests medication based on LDL-P in healthy people, please ask them to produce the randomly controlled trials which showed meaningful reduction in hard, clinical endpoints from such treatment (NNT- numbers need to treat). Otherwise, consider any cholesterol measurement to be a risk marker at best.

Calling for everybody to have an NMR and to be medicated for "high LDL-P" would seem to be another in the long line of attempt to enlarge the pool of statin customers, not to mention the market for NMR testing.

Peter said...

Hi Hasan,

You are verging on the old adage, "the best cholesterol number is one which hasn't been measured..." which I personally have a great deal of time for! HbA1c is another matter.

Peter