Answer Day: What Causes High LDL on Low-Carb Paleo?

First, thank you to everyone who commented on the quiz. I enjoyed reading your thoughts.

Is High LDL Something to Worry About?

Perhaps this ought to be the first question. Jack Kronk says “I don’t believe that high LDL is necessarily a problem” and Poisonguy writes “Treat the symptoms, Larry, not the numbers.” Poisonguy’s comment assumes that the LDL number is not a symptom of trouble. Is it?

I think so. It helps to know a little about the biology of cholesterol and of blood vessels.

When cells in culture plates are separated from their neighbors and need to move, they make a lot of cholesterol and transport it to their membranes. When cells find good neighbors and settle down, they stop producing cholesterol.

The same thing happens in the body. Any time there is a wound or injury that needs to be healed, cholesterol production gets jacked up.

When people have widespread vascular injuries, cholesterol is produced in large quantities by cells lining blood vessels. Now, to repair injuries cells have to coordinate their functions. Endothelial cells are the coordinators of vascular repair: they direct other cell types, like smooth muscle cells and fibroblasts, in the healing of vascular injuries.

To heal vascular injuries, these cells not only need more cholesterol for movement; they also need to multiply. It turns out that LDL, which carries cholesterol, also causes vascular cells to reproduce (“mitogenesis”):

The best-characterized function of LDLs is to deliver cholesterol to cells. They may, however, have functions in addition to transporting cholesterol. For example, they seem to produce a mitogenic effect on endothelial cells, smooth muscle cells, and fibroblasts, and induce growth-factor production, chemotaxis, cell proliferation, and cytotoxicity (3). Moreover, an increase of LDL plasma concentration, which is observed during the development of atherosclerosis, can activate various mitogen-activated protein kinase (MAPK) pathways …

We also show … LDL-induced fibroblast spreading … [1]

If endothelial cells are the coordinators of vascular repair, and LDL particles their messengers to fibroblasts and smooth muscle cells, then ECs should be able to generate LDL particles locally. Guess what:  ECs make a lipase whose main effect is to decrease HDL levels but can also convert VLDL and IDL particles into LDL particles and remove fat from LDL particles to make them into small, dense LDL:

Endothelial lipase (EL) has recently been identified as a new member of the triglyceride lipase gene family. EL shares a relatively high degree of homology with lipoprotein lipase and hepatic lipase …

In vitro, EL has hydrolyzed phospholipids in chylomicrons, very low density lipoprotein (VLDL), intermediate density lipoprotein and LDL. [2]

Immune cells, of course, are essential for wound healing and they should be attracted to any site of vascular injury. It turns out that immune cells have LDL receptors and these receptors may help them congregate at sites of vascular injury. [3]

I don’t want to exaggerate the state of the literature here:  this is a surprisingly poorly investigated area. But I believe these things:

1.      Cholesterol and LDL particles are part of the vascular wound repair process.

2.      Very high LDL levels are a marker of widespread vascular injury.

Now this is not the “lipid hypothesis.” Compare the two views:

  • The lipid hypothesis:  LDL cholesterol causes vascular injury.
  • My view:  LDL cholesterol is the ambulance crew that arrives at the scene of the crime to help the victims. The lipid hypothesis is the view that ambulance drivers should be arrested for homicide because they are commonly found at murder scenes.

So, to Poisonguy, on my view high LDL numbers are a symptom of vascular injury and are a cause for concern.

Big-Picture View of the Cause of High LDL

So, on a micro-level, I think vascular damage causes high LDL. But what causes vascular damage?

Here I notice a striking difference in commenters’ perspectives and mine. I tend to take a big-picture, top-down view of biology. There are three basic causes of nearly all pathologies:

1.      Toxins, usually food toxins.

2.      Malnutrition.

3.      Pathogens.

The whole organization of our book is dictated by this view. It is organized in four Steps. Step One is about re-orienting people’s views of macronutrients away from high-grain, fat-phobic, vegetable-oil-rich diets toward diets rich in animal fats. The other steps are about removing the causes of disease:

1.      Step Two is “Eat Paleo, Not Toxic” – remove food toxins.

2.      Step Three is “Be Well Nourished” – eliminate malnutrition.

3.      Step Four is “Heal and Prevent Disease” – address pathogens by enhancing immunity and, where appropriate, taking advantage of antibiotic therapies.

So when someone offers a pathology, any pathology, my first question is: Which cause is behind this, and which step do they need to focus on?

In Larry’s case, he had been eating low-carb Paleo for years. So toxins were not a problem.

Pathogens might be a problem – after all, he’s 64, and everybody collects chronic infections which tend to grow increasingly severe with age – but Larry hadn’t reported any other symptoms. More to the point, low-carb Paleo diets typically enhance immunity, yet Larry had fine LDL numbers before adopting low-carb Paleo and then his LDL got worse. So it wouldn’t be infectious in origin unless his diet had suppressed immunity through malnutrition – in which case the first step would be to address the malnutrition.

Step Three, malnutrition, was the only logical answer. The conversion to Paleo removes a lot of foods from the diet and could easily have removed the primary sources of some micronutrients.

So I was immediately convinced, just from the time-course of the pathology, that the cause was malnutrition.

Micronutrient Deficiencies are Very Common

In the book (Step Three) we explain why nearly everyone is deficient in micronutrients. The problems are most severe for minerals:  water treatment removes minerals from water, and mineral depletion of soil by industrial agriculture leads to mineral deficiencies in farmed plants and grain-fed animals.

This is why our “essential supplements” include a multimineral supplement plus additional quantities of five minerals – magnesium, copper, chromium, iodine, and selenium. Vitamins get a lot of attention, but minerals are where the big health gains are.

Copper Deficiency and LDL

Some micronutrient deficiencies are known to cause elevated LDL.

Readers of our book know that copper causes vascular disease; blog readers may be more familiar with an excellent post by Stephan, “Copper and Cardiovascular Disease”, discussing evidence that copper deficiency causes cardiovascular disease. As I’ve just argued that cardiovascular disease causes high LDL, it shouldn’t be a surprise that copper deficiency also causes hypercholesterolemia:

Copper and iron are essential nutrients in human physiology as their importance is linked to their role as cofactors of many redox enzymes involved in a wide range of biological processes, as well as in oxygen and electron transport. Mild dietary deficiencies of both metals … may cause long-term deleterious effects in cardiovascular system and alterations in lipid metabolism (3)….

Several studies showed a clear correlation among copper deficiency and dyslipidemia. The main alterations concern higher plasma CL and triglyceride (TG) concentrations, increased VLDL-LDL to HDL lipoproteins ratio, and the shape alteration of HDL lipoproteins.  [4]

The essentiality of copper (Cu) in humans is demonstrated by various clinical features associated with deficiency, such as anaemia, hypercholesterolaemia and bone malformations. [5]

Over the last couple of decades, dietary copper deficiency has been shown to cause a variety of metabolic changes, including hypercholesterolemia, hypertriglyceridemia, hypertension, and glucose intolerance. [6]

Copper deficiency is, I believe, the single most likely cause of elevated LDL on low-carb Paleo diets. The solution is to eat beef liver or supplement.

So, was my advice to Larry to supplement copper?  Yes, but that was not my only advice.

Other Micronutrient Deficiencies and Elevated LDL

Another common micronutrient deficiency that causes elevated LDL cholesterol is choline deficiency that is NOT accompanied by methionine deficiency. That is discussed in my post “Choline Deficiency and Plant Oil Induced Diabetes”:

Choline deficiency (CD) by itself induces metabolic syndrome (indicated by insulin resistance and elevated serum triglycerides and cholesterol) and obesity.

A combined methionine and choline deficiency (MCD) actually causes weight loss and reduces serum triglycerides and cholesterol …

I quote both these effects because it illustrates the complexity of nutrition. A deficiency of a micronutrient can present with totally different symptoms depending on the status of other micronutrients.

Julianne had a really nice comment, unfortunately caught in the spam filter for a while, with a number of links. She mentions vitamin C deficiency and, with other commenters, noted the link between hypothyroidism and elevated LDL. As one cause of hypothyroidism is iodine or selenium deficiency, this is another pathway by which mineral deficiencies can elevate LDL.

UPDATE: Mike Gruber reduced his LDL by 200 mg/dl by supplementing iodine. Clearly iodine can have big effects!

Other commenters brought up fish oil. They may be interested to know that fish oil not only balances omega-6 to modulate inflammatory pathways, it also suppresses endothelial lipase and thus moderates the LDL-raising and HDL-lowering effect of vascular damage:

On the other hand, physical exercise and fish oil (a rich source of eicosapentaenoic acid and docosahexaenoic acid) suppress the activity of EL and this, in turn, enhances the plasma concentrations of HDL cholesterol. [7]

Whether this effect is always desirable is a topic for another day.

My December Advice to Larry

So what was my December advice to Larry?

It was simple. In adopting a low-carb Paleo diet, he had implemented Steps One and Two of our book. My advice was to implement Step Three (“Be well nourished”) by taking our recommended supplements. Eating egg yolks and beef liver for copper and choline is a good idea too.

Just to cover all bases, I advised to include most of our “therapeutic supplements” as well as all the “essential supplements.”

Since December, Larry has been taking all the recommended supplements and eating 5 ounces per week of beef liver. As I noted yesterday, Larry’s LDL decreased from 295 mg/dl to 213 mg/dl, HDL rose from 74 mg/dl to 92 mg/dl, and triglycerides fell from 102 to 76 mg/dl since he started Step Three. This is all consistent with a healthier vasculature and reduced production of endothelial lipase.


Some people think there is something wrong with a diet if supplements are recommended. They believe that a well-designed diet should provide sufficient nutrition from food alone, and that if supplements are advised then the diet must be flawed.

I think this is quite mistaken. The reality is that Paleolithic man was often mildly malnourished, and modern man – due to the absence of minerals from treated water and agriculturally produced food, and the reduced diversity and higher caloric density of our foods – is severely malnourished compared to Paleolithic man.

We recommend eating a micronutrient-rich diet, including nourishing foods like egg yolks, liver, bone broth soups, seaweed, fermented vegetables, and so forth. But I think it’s only prudent to acknowledge and compensate for the widespread nutrient depletion that is so prevalent today. Even when nutrient-rich food is regularly eaten, micronutrient deficiencies are still possible.

Eating Paleo-style is not enough to guarantee perfect health. Luckily, supplementation of the key nutrients that we need for health and that are often missing from foods will often get us the rest of the way.


[1] Dobreva I et al. LDLs induce fibroblast spreading independently of the LDL receptor via activation of the p38 MAPK pathway. J Lipid Res. 2003 Dec;44(12):2382-90.

[2] Paradis ME, Lamarche B. Endothelial lipase: its role in cardiovascular disease. Can J Cardiol. 2006 Feb;22 Suppl B:31B-34B.

[3] Giulian D et al. The role of mononuclear phagocytes in wound healing after traumatic injury to adult mammalian brain. J Neurosci. 1989 Dec;9(12):4416-29.

[4] Tosco A et al. Molecular bases of copper and iron deficiency-associated dyslipidemia: a microarray analysis of the rat intestinal transcriptome. Genes Nutr. 2010 Mar;5(1):1-8.

[5] Harvey LJ, McArdle HJ. Biomarkers of copper status: a brief update. Br J Nutr. 2008 Jun;99 Suppl 3:S10-3.

[6] Aliabadi H. A deleterious interaction between copper deficiency and sugar ingestion may be the missing link in heart disease. Med Hypotheses. 2008;70(6):1163-6.

[7] Das UN. Long-chain polyunsaturated fatty acids, endothelial lipase and atherosclerosis. Prostaglandins Leukot Essent Fatty Acids. 2005 Mar;72(3):173-9.

Leave a comment ?


  1. Paul, I have avoiding the blogsphere for a bit but I checked by your blog every so often because it always offers something new. This post was exceptional.

  2. By the way, do you recommend any specific brands for your recommended supplements?

  3. Hi Bill,


    We don’t have any real preferences about brands but we always appreciate it when people buy from Amazon through our site because it helps support the blog!

  4. I agree with Bill, I really enjoyed this post because it puts all of your knowledge into a real world situation. For me, that helps drive home the essential points.

    This could make a great monthly topic…solve the mystery ailment, symptoms, lab test abnormality, etc.

    Thanks for an eye opening discussion, and thanks to all of yesterdays contributors too!

  5. Excellent work, Paul.

  6. Paul,
    Another excellent blog post. I agree with all the points you have mentioned.

    Here is something you’d find interesting. I am 28 yr old make in great shape – 145 lbs, lean, strength train 3 days a week. When I started eating primal my lipid profile was great (150 TC with 75 HDL and 30 Tri). At this point my weight was stable. What I observed after that was, anytime I lost some significant weight (8-15 lbs), irrespective of how much fat I ate (15-55%) and the kind of fat I ate (SaFa or not), my LDL shot up. First time my TC went to 330 and second time it went to 350. My HDL was at 80 and my triG under 70 all times.

    With the help of Ned Kock, Chris Kresser and Chris Masterjohn I have self-diagnosed this as acute subclinical hypothryroidism. Whenever I diet down, my thyroid function seems to slow down (as a survival mechanism) thereby slowing metabolism thereby resulting in excess plasma LDL. My body temperature now (as I am dieting down) is ~ 96.5 F which confirms the acute diet related hypothyroidism.

    But after reading your post it also seems like I might be deficient in some micronutrients (since I am at a calorie deficit) which might contribute to the high LDL.

    Do you think just losing body fat (which means mobilizing stored adipose tissue) could cause a high LDL spike?

    Thanks! – Raj

  7. Well, if you are right, my numbers should be much improved next blood draw, since I’ve been supplementing according to your book since I bought it. As I mentioned, just the initiation of my supplementation with iodine (months prior to buying your book) was followed soon after by a ~200 drop in my TC (and my HDL going up by ~30%). So now I’m on *all* the supplements you espouse. We’ll see …

  8. Hi Paul, great post! my question is why didn’t larry have high LDL before going paleo? are there foods he cut out rich in copper and other minerals, but just not paleo?


  9. Hey Paul,

    This is a good blog post, though I don’t think it is quite a complete view of the role of LDL in cardiovascular disease.

    This stuff is in my blog queue, and will be up in a week or so, but I think the evidence is quite overwhelming that the degeneration of LDL particles *causes* vascular injury. No doubt that native LDL particles and cholesterol are involved in mostly positive roles in the body, and no doubt a “high LDL-cholesterol” level cannot distinguish between LDL being produced at a higher rate on the one hand, and on the other, LDL accumulating because it is not being metabolized, which, all things equal, leads to its oxidative degeneration. However, I think there is enormous evidence that degeneration of lipids in the blood is a pre-requisite for cardiovascular disease and that in humans this is primarily although not entirely associated with degeneration of LDL-associated lipids (primarily PUFAs rather than cholesterol). So I think that a comprehensive view of LDL — although it should, as you have done, emphasize the pro-life and pro-health roles of cholesterol and native LDL particles — must account for the role of LDL degeneration in promoting atherosclerosis, and this must invoke a variant of “lipid hypothesis” even if it is, as I have called them, a ‘degenerative’ rather than ‘infiltrative’ lipid hypothesis.


  10. Hi Raj,

    Try the supplements and see!

    I suppose there are mechanisms by which weight loss could lead to toxicity effects. Cholesterol poisoning happens in animal models in case of zinc/copper deficiency. Various fat soluble toxins could be released.

    Hi Mike,

    Let me know when you get your numbers!

    Hi Eric,

    Larry could answer that better than I could, but beans, cereals, seeds, and nuts are fairly rich in copper.

    Hi Chris,

    Agreed, it’s certainly not a complete view of LDL in cardiovascular disease. I brought the vascular injury part up only to explain why large numbers of native LDL particles might be generated. Of course there are many processes that oxidize LDL, create foam cells, etc.

    I agree that oxLDL causes vascular injury. Infections, lipid peroxidation, and other inflammatory processes can generate oxLDL from native LDL.

    The “lipid hypothesis” that I disagree with is that all LDL particles are equally injurious, so that a simple weighing of LDL particles per deciliter of blood is sufficient to gauge risk. It is as you say the degenerate LDL not the native LDL that is harmful.

    However I would still stick to my main thesis, which is that infections, diet (e.g. excess omega-6 or trans-fat consumption), and nutrient deficiencies are the ultimate source of degenerate LDL and other harmful compounds.

    Looking forward to your series!

    Best, Paul

  11. Thanks for this Paul. Very interesting. Does this then mean that efforts to reduce LDL, without addressing micronutrient deficiencies, are likely to be counterproductive, as they are interfering with the body’s natural vascular repair mechanisms?

    And can those of us suffering from CVD use LDL concentration as a measure of the severity of our disease? I’m (not very) low carb paleo taking large doses of niacin and looking for a way to measure the progress of my CVD

  12. VERY interesting stuff. Curious is C-reactive protein or any other markers of systemic inflammation/immune activity were tracked?

  13. Hi Dave,

    Yes, I think simply reducing LDL is counterproductive. That’s why drugs that reduce LDL as their only effect make cardiovascular disease worse. Benefits from statins seem to come from non-LDL effects, e.g. anti-inflammatory effects, raising vitamin D production, etc.

    I do think that when LDL is very high, it can serve as a crude measure of vascular damage for many people.

    Hi Robb,


    C-reactive protein tends to be an indicator of infection severity. I think Larry’s CRP level was low. I don’t know about any of the other people mentioned in yesterday’s post.

    Best, Paul

  14. Paul – Excellent post and series!

    However, from my own n=1 experience, I believe there may be other factors at play with regards to “High LDL on low-carb Paleo”. My lipid history:

    2008 – 166 Total, 103 LDL, 44 HDL, 116 Trig
    This was pre-paleo, moderate but not your typical high carb intake, lots of conventional wisdom “healthy” foods whole grains, oatmeal, etc.

    2009 – 176 Total, 78 LDL, 67 HDL, 55 Trig
    After a few months of mostly Paleo eating, moderate egg consumption, no liver

    2010 – 295 Total, 163 LDL, 97 HDL, 66 Trig
    Over 1 year strict Paleo, heavy consumption of Pork, Eggs 3-4/day, Liver, Coconut Oil

    I’m not too concerned with the last test given my excellent ratios. I also had a VAP which verified very low VLDL numbers… However, as time passes I wonder if this upward trend in LDL/Total will continue and what if anything I can/should do to lower my LDL? What are your thoughts on the increase/trend and would you recommend I tweak something in my diet to lower LDL?

    Thanks in advance Paul!

  15. Hi Paul, thanks for this post (also).

    Re CRP:

    (after about 2 months PHD/panu with no eggs but lots of liver. Typically somewhat hypothyroid, 200mcg iodine supplement significantly decreased TSH, but not enough. I’ll follow your advice to increase it gradually)

    TC 374
    HDL 123
    TG 105

    CRP 0.0 (repeat test: still 0.0)

    (two weeks earlier, it was TC 324, HDL 121, LDL184, TG 92)

    (No VAP here, can’t find a place in Europe, but I have apoB: 2.00 (.5-1.00) apoA1:2.41 (1.1-1.6) A year ago it was ApoB 1.26, ApoA1: 1.95 on low cal low fat diet)

    This does not seem unique, I’ve come across at least one similar high TC/LDL .1 CRP case on a paleo thread recently.

  16. Hi Henry,

    I also wouldn’t be too concerned with your numbers, but I do think you should try our supplement regimen and eat some “safe starches” for glucose.

    Hi donat,

    Thanks for the data.

    Like I said to Robb, CRP is an indicator of systemic infections and I think these cases of high LDL on Paleo are usually not infectious in origin, but nutritive. So I’m not surprised that Paleo dieters, especially younger ones, have very low CRP.

    That’s good news because it’s a heck of a lot easier to fix a nutrient deficiency than a chronic infection!

    Best, Paul

  17. Hi Paul,

    When I was low carb also my CRP and homocysteine were high along with LDL cholesterol. TSH was high and FT3 low (I was not iodine deficient since I was using table salt).

    CPR was 3 mg/L so high risk;

    Before the low carb high fat diet my thyroid was fine and my CRP was 1 mg/L so I’m very worried about this situation.

    Taking vitamin C lowered my LDL (from 125 to 85) without changing my HDL (still high: 88). The other values however were still out range. Then I took a multivitamin for 4 months but nothing changed.

    I’ll do lab tests again after 3 months on PHD, if they are not in the range I think I should return to an high carb (>50%).

  18. Hi kratos,

    Table salt has fairly minimal iodine. I would try to build up the iodine a little.

    Starches should help, and the vitamin C. I think that low free T3 may arise from a vitamin C deficiency. Low-carb can cause scurvy as I’ve blogged.

    Will look forward to your lab tests. I don’t think it’s necessary to go up to 50% carb. All the benefit should come in the first 600 calories.

    Best, Paul

  19. Chris Masterjohn said it much more eloquently than I could ever, so I’ll piggyback his response as it closely matches my views on the matter.

    BTW, I do take all the supplements you suggest, so I’m not arguing from that standpoint. I’m arguing the science.

    One can dramatically alter one’s total LDL quite rapidly, and easily, with a simple change in dietary macronutrient ratio, so LDL levels don’t necessarily reflect cardiovascular damage at all. It’s more likely they reflect what foods one is eating. Are Larry’s rising LDL levels more likely to be a reflection of him acquiring vascular damage or are they more likely a result of his dietary changes? I’m with the latter.

    LDL is a carrier of lipids, so its quite plausible that the more lipid one eats, the more movers one needs. So, its quite plausible that a raise in LDL is a result of diet, rather than damage. Or it could be a combo of both. That’s why there’s a VAP test now. Serial VAP test results would be much more meaningful in theorising whether Larry has sustained any cardiovascular damage.

    As pointed out by Chris, for cardiovascular health one has to look at not only native LDL, but at oxLDL and even oxLDL-IC, etc. And one has to also look at what agents could be oxidising these lipoproteins and its constituents. That’s what I mean by “treat the symptoms, not the numbers.” If Larry is avoiding the things that are likely involved in the oxidation of LDL (i.e. PUFAs, fructose, etc), then he shouldn’t be fussing over LDL numbers.

    So, Paul, if you agree that native LDL is not a problem, then overall LDL levels are not necessarily an indicator of cardiovascular health. They could merely be a reflection of one’s diet. Nothing more, nothing less. One needs more info to make the claim of vascular damage, IMO. You may differ.

    If I were Larry, I wouldn’t make any rash moves based on native LDL levels. That’s all I was saying.

    Lastly, in my experience, VLC people tend to be some of the bigger consumers of liver and eggs, so I’m not sure how copper and their high LDL levels fit together.

  20. Paul,

    Looks like we’re mostly agreed!


  21. Larry Eshelman

    As Paul relates, I started eating a low carb diet in 2002 and for five years my lipid numbers were “good”. During that time I still feared saturated fat and so ate very little beef. My main sources of protein were fish, yogurt, and nuts (mainly almonds and walnuts). My other source of fat was olive oil.

    Then in 2007 I read Taubes and started reading Paleo oriented blogs. I overcame my fear of saturated fat and started eating more beef. I also became concerned about PUFA and started to use less olive oil and more (grassfed) butter and coconut oil. The shift was somewhat gradual, but coincidental with this shift was an increase in my LDL.

    This inspired me to become even stricter about avoiding toxins (i.e., fructose, wheat, omega6). For example, I eliminated olive oil and nuts and even beer and switched to grass fed beef. But to my surprise my LDL rose even more. (At this point I was very low carb.)

    During the last two years I have had my LDL measured directly eight times. And I also bought a home lipid testing machine (CardioCheck) which measures my total cholesterol, HDL and triglycerides. So I’m basing my findings on a lot of numbers not just a few.

    I have tried a number of self experiments with my diet, varying something and then testing my lipids after 4 weeks or so. For example, I’ve tried eliminating eggs for 10 weeks (no change) and eliminating dairy for 4 weeks (no change).

    Two changes, however, have made a big change. First, after eating a lot of coconut oil for four weeks, my HDL jumped from the 60s to the 80s. Then I eliminated coconut oil for four weeks and my HDL dropped back to the 60s. Then I reintroduced coconut oil and it went back up to the 80s and on my last measurement it peaked at 92.

    The other change that made a big difference was following Paul’s supplement recommendations. During my pre-Taubes low carb period I was taking a daily multivitamin. After I started reading Paleo blogs I stopped taking a multivitamin, but started taking D3 and K2 and later magnesium. I was already taking these three supplements when I started taking the remaining supplements that Paul recommended. Before taking Paul’s recommended supplements my last measured LDL was 295 and my calculated LDL was hovering around 310 (based on measurements using my CardioCheck machine). Four weeks later my calculated LDL had dropped to 221 (and has stayed in the low 200s on subsequent tests). I’ve also had my LDL directly measured by a lab twice since then and it came back as 217 and 213.

    Paul has suggested that it is the copper from the liver which has improved my LDL numbers. Another possibility is the vitamin C I’m now taking. Even before Paul made his recommendations, I was concerned that I wasn’t getting enough vitamin C.

    I should add that I didn’t only follow Paul’s advice on supplements, but also followed his advice in other areas — e.g., several months prior to taking his recommened supplements I had increased my consumption of starches.

    Finally, several commenters have brought up CRP. My hsCRP has consistently been low throughout this period (< 0.16).


  22. Hi Poisonguy,

    I am not sure what you mean by “LDL levels … could merely be a reflection of one’s diet.” If you mean of micronutrient composition of the diet, then we agree. If you mean macronutrient composition, I disagree. Not at the levels we are talking about.

    Dietary lipids are carried by chylomicrons, not LDL. In the absence of micronutrient deficiencies that limit processing, there shouldn’t be a big impact of diet on LDL.

    We know from the choline data that nutrient deficiencies in the liver can cause big swings in LDL. I think the copper data shows that similar swings can occur due to nutrient status of the vasculature.

    VAP results should be fairly dynamic with nutrient status and I would expect Larry’s VAP data to be much improved along with his nutrition. But we won’t know unless he gets a test.

    Let me spell this out a little more thoroughly:

    Atherosclerosis and vascular injury are two different things. Atherosclerosis is much more severe.

    I am saying that high LDL on a Paleo diet signifies vascular injury but not, generally, atherosclerosis.

    OxLDL, foam cells, plaques, infected lesions, calcification, are all characteristics of atherosclerosis.

    I have no reason to believe Larry has any of those things. However, I suspect he has had a copper-deficiency induced vascular injury.

    Native LDL is a problem because it is a symptom of that vascular injury, which should be repaired.

    I don’t think supplementing is a rash move.

    Best, Paul

  23. Since vascular injury is likely critical in initiating the pathogenesis of atherosclerosis, I’m not so sure you can separate the two so neatly. Those dietary lipids that get dumped off by chylomicrons get repossessed, they don’t disappear, so they can still affect LDL levels.

    There are plenty of studies demonstrating that small changes in macronutrient composition can have effects on LDL levels (Shai et al 2008 comes to mind, but I can find plenty more). I thought that was common knowledge.

    I still don’t see how supplementing, and the subsequent lowering of LDL is a sign of improving vascular health. It could just be a sign of supplementing changing LDL levels, just like changes in macronutrients do.

    I guess I’m just not convinced that all these people’s low carb diets have damaged their vasculature. It could be, but there are more plausible theories (Occam’s razor, possibly), one which I thought was common knowledge–macronutrient ratios. I guess time will tell and we won’t have to depend on theorizing.

  24. Hi Poisonguy,

    If you agree that the micronutrients are lowering the LDL, then we’re pretty much in agreement.

    Macronutrients do affect LDL levels, but the changes are usually much smaller than what Larry experienced. Changing macronutrient ratios doesn’t satisfy Occam’s razor because it doesn’t explain the phenomenon.

    I’m not insisting that the mechanism is via improved vascular health. Just suggesting that. That’s obviously a topic for biomedical research. Mechanisms aren’t yet understood.

    I’m not saying and don’t believe that low-carb dieting damages vasculature, but rather that nutrient deficiencies do.

    Yes, time will tell.

    Best, Paul

  25. Hi Paul,
    is it not possible that certain people are hyper responders to the extra saturated fat in their diet? How many of those from the list know about their apoe type? Dr. Davis has a few posts about apoe4 too.

    I was told not to worry, that I’ll be fine as long as I avoid high pufas and excess sugar, but Stephan and Chris M.’s posts are enough to make me keep my eyes peeled for new answers.

  26. Hi simona,

    Yes, it’s possible that some people are hyper-responders to saturated fat diets, but then we would have to ask why. ApoE4 is one factor. But in general, I think a big LDL response is unnatural and suggestive of nutrient deficiencies. LDL levels are regulated and if the body is healthy, they shouldn’t vary too far from the normal range.

    Avoiding high PUFAs and excess sugar is very important. But it’s not the be-all and end-all of healthy eating.

    However, not worrying is good advice in general. Chronic worry doesn’t have a good record of making health better!

    Best, Paul

  27. Fascinating! Perfect kind of material that needs to be published in reviews in medical journals. Perhaps an opinion piece or review in the Ancestral Health Journal would be appropriate once it is established.

  28. Hi again, Paul,

    I never dismiss anything you say out of hand. You have great theories, so I like challenge you when I’m not in complete agreement. It helps me understand and it might help others too.

    It would be interesting for your theory, I think, to know what the LDL levels are for patients with established atherosclerosis for comparison to these low carb values–theorectically, they should be much higher in atherosclerotics than in those with minor vascular damage.

    Also, if the liver is churning out more LDL for vascular repair, shouldn’t that be reflected in the level of HDL (i.e. if the vasculature is slurping up cholesterol, or whatnot, to fix the damage, shouldn’t the need for HDL decrease and the feedback to the liver being to stop sending them around, which if you apply the law of supply and demand, you’d get a drop in HDL in cases of vascular damage)?

    I guess I’m having difficulty accepting that the values exhibited by the folks you quoted are even high to begin with (since I believe most of them eat a healthy diet–so what is doing the damage?). Who says they aren’t in the normal range? The American Heart Association?

    I’ve ruffled enough feathers for one day. Hope I’m somewhat contributing positively nonetheless. Keep up the good work.

  29. Hi Aaron,

    Would love to publish in the Ancestral Health Journal. When will it start?

    Hi Poisonguy,

    No feathers ruffled here. I’m always glad to get your comments, and welcome disagreement. Keeps me honest. Chris M. has pointed out some issues with things I’ve said from time to time, for which I’m grateful.

    I think it’s likely, for reasons presented in the book, that many low-carb Paleo dieters are mildly malnourished. So when I see people with symptoms of malnourishment, that are responsive to supplementation, I’m not inclined to believe their symptoms are in the healthy range.

    Best, Paul

  30. I understand your point, that it is a maladaptive response nonetheless, that’s what the consultant chemical pathologist said to me, in Nov 2009. He saw that there were a few cases in some previous low-carb studies, but was unable to think of a causal mechanism. It’s strange that copper deficiency could mediate such huge changes in LDL particle number.

  31. Hi Paul,
    Adding to your conversation with Poisonguy – I can wrap my head around the fact that some of us have nutrient deficiencies that lead to vascular damage.

    But how come all of us had lower levels of LDL before eating (relatively) low carb paleo? People who read and comment here are among the 0.01% of the population is that is even remotely informed about real nutrition. If, after all this research and careful eating, our LDL’s are so high due to nutritive deficiencies, shouldn’t there be people out there eating the SAD (or other super high carb diets)with a much greater degree of vascular damage (due to nutritive deficiencies AND systemic inflammation) and hence much higher LDL values?

    I guess I’m finding it hard to believe that all these people’s LDLs (including mine) were within the normal range without any specific supplementation/liver consumption before starting to eat low carb paleo… and now after starting to eat MUCH healthier we are in need of additional supplements/liver?


  32. I think there are excellent points being made by Paul, Poisonguy and Chris Masterjohn.

    Note the word ‘necessarily’ in my statement that Paul quoted in the article. That’s because I believe that there is still much to learn with regard to LDL’s full role in cardiovascular health. Could high LDL be found to be definitively problematic? Sure. But I will continue to leave room for an explanation something closer to what Poisonguy articulates. As for macronutrient affecting LDL, I strongly believe it does. I saw a dynamic increase in my LDL very quickly when I switched to a low-carb Paleo diet. My trigs plummeted too.

    I know that the new diet also changed my micronutrient intake, and although I will leave room for the possibility, I am not yet convinced that mineral deficiencies are the primary cause of the raised LDL across so many Paleo eaters. I eat a diet rich in organic eggs, take FCLO, eat lots of pastured dairy, lots of salads with a wide array of veggies, eat sprouted nuts, wild sardines, choice fruits regularly, use all the right oils to cook with, consume zero veg oils. What I am saying is that I do not have a super narrow diet. I eat a very colorful variety of items all of the highest quality and I’d be quite surprised if I am suddenly short on micronutrients that my SAD diet was somehow providing.

    That being said, Paul brings some other excellent points into the picture here and it’s very nice to see the Paleo community really digging into this. Getting to the bottom of the ‘LDL’ issue will be a huge win for the Paleo community because many people wonder about this and so I’m very appreciative of Paul’s efforts here. This type of dialogue and conversation that the Jaminet’s welcome shows that they are an open book with respect to the opinions of other experts in these areas, even if they directly challenge their own findings and/or suggestions.

    Thanks again,
    Jack Kronk

  33. the quiz and answer posts were really informative and i learned a lot! thank you 🙂 best, claire

  34. “Copper supplementation of 2-3 mg/day for four to six weeks did not result in clinically significant changes in cholesterol levels (12, 23)”.
    from (I suppose it’s not clear what was the starting point, if they were correcting a possible deficiency or not, “due to a lack of a reliable biomarker of copper nutritional status”)

  35. So are we saying that Larry now has ideal numbers?

    What kind of guidance would you give for target numbers?

  36. Hi Paul,
    Excellent discussion here. Wouldn’t you suspect that most people who previously followed a SAD or even low-fat diet would suffer from some degree of atherosclerosis by their 30s or 40s? There were studies done of US soldiers killed in the Korean war where over 75% had some atherosclerosis: Subsequent analysis of trauma victims showed similar levels of atherosclerosis:

  37. Hi Paul,

    Another highly informative post, very nice.
    Every day I learn how little I actually know 🙂

    You wrote: “Readers of our book know that copper causes vascular disease.”.
    Unless I have a limited edition of the book, I’m guessing the word “deficiency” is missing here 😉

    If I were to put my current understanding of LDL into equations, I would get something like this:

    LDL level = (Liver output) + (Local production due to endothelial lipase) – (Clearance)

    The liver output seems difficult to modify by diet. Chris Masterjohn mentioned that linoleic acid can cause esterification of cholesterol in the liver, basically trapping it, causing a drop in plasma LDL [1]. But that doesn’t seem like a desirable way to reduce LDL.
    Modifying the local production due to endothelial lipase seems only possible indirectly, by tackling the underlying reason for increased local LDL generation. So adding copper etc. as described in the post.
    The clearance appears to be related to thyroid (and probably other factors as well). So another indirect dietary path. Coconut oil may do its LDL lowering “magic” via this path.

    If atherosclerosis as mainly initiated by oxidation of LDL, that equation could be as follows:

    LDL safety = (Antioxidant level in LDL particle) – ((Oxidation rate) x (Time in circulation))

    Increasing the (fat soluble) antioxidant level in LDL particles seems feasible. Stephan mentioned a study in which red palm oil (high in CoQ10 and vitamin E, both packed with LDL [2]) lowered oxLDL by an impressive 69% [3].
    Reducing the oxidation rate of LDL can be done by lowering the consumption of oxidation prone polyunsaturated fatty acids. As these are incorporated into the LDL membrane, they make the LDL vulnerable to oxidation [2]. Lowering the oxidation rate in this way means the LDL doesn’t run out of antioxidants as quickly.
    Reducing the time in circulation seems mainly influenced by the clearance of LDL. So this goes back to thyroid health as mentioned above.

    These equations are likely way too simplistic to cover all nuances. But for now I find them very useful to get a slight grasp on the whole LDL matter.


    [1] Masterjohn C. The Total-to-HDL Cholesterol Ratio — What Does It Mean?
    [2] Masterjohn C. High Cholesterol And Heart Disease – Myth or Truth?
    [3] Guyenet S. Tropical Plant Fats: Palm Oil.

  38. Hi Poisonguy,

    You wrote: “There are plenty of studies demonstrating that small changes in macronutrient composition can have effects on LDL levels (Shai et al 2008 comes to mind, but I can find plenty more).”

    Are you referring to the weight loss study based on 3 diets [1]? If so, my brief scanning of that paper paints a different picture.
    First of all, it is about 3 different diets, which not only differ in macronutrients, but also micronutrients. So it seems difficult to reliably pinpoint lipid changes exclusively to macronutrients.

    Furthermore, the paper states the following on LDL levels:
    “LDL cholesterol levels (Figure 3C) did not change significantly within groups, and there were no significant differences between the groups in the amount of change.”

    So, if my guess about the Shai et al. 2008 paper is correct (which is a big if), I don’t understand why this study came to your mind regarding LDL levels being influenced by macronutrient composition. The paper clearly states this is not the case.
    The story for HDL levels is entirely different. These appear to be significantly influenced by diet, and possibly macronutrient composition.

    Would you be willing to provide references to the other studies you had in mind? Thanks.


    [1] Shai I., et al. Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. N Engl J Med. 2008 Jul 17;359(3):229-41.

  39. Before we can be sure about recommending ‘target numbers’, we MUST first understand exactly what the numbers mean, and we must first be able to correctly assess how the numbers correlate with vascular health in all its aspects. How can we know, at this point, what numbers are a healthy range?

    We cannot rely solely on data that was collected regarding LDL numbers correlating with vascular health in people who do not eat the right foods. It’s not fair to the complex physiology of the human body.

    So what are we to do in the mean time? We take care of ourselves as best we can with respect to eating a diet rich in all the right elements and macro whos its and micro whats its and have faith that the body knows exactly what to do with that. Trying to correct a function of cholesterol in the body in which the initial reason for worry in the first place stems directly out of a flawed hypothesis that has since been completely obliterated could prove to be an ‘off track’ approach when the dust settles.

    It’s like we’re on a dirt path in the back mountains with this research and the path will lead us to a fork of main roads, but we must first get to the fork so we can see what’s up ahead in order to determine which road to take. Let’s not hop on the wrong main road before we can see what it leads to.

    We’re not there yet folks.

  40. I hear you, Jack, but are we to congratulate Larry or lament his current lipid situation?

  41. How does the question of “real” vs “computed” LDL fit into all of this? Isn’t the important question the actual amount of small particle LDL?

  42. I don’t know Bob. I don’t see any conflicting ideas surrounding an increase of HDL and decrease of Trigs, so probably we should congratulate him.

    I was only responding to the ‘target numbers’, specifically regarding LDL.

  43. I’m thinking along selfish lines, Jack. A few months ago, I converted to this way of eating from a vegan way and my LDL doubled while by HDL was cut in half and my triglycerides were cut to one-third of what they were.

  44. Hi, I too am “afflicted” with quite high TC and LDL as measured by a standard lipid panel:

    TC: 391
    TGs: 54
    HDL: 55
    LDL: 325 (Friedewald)
    TSH: .86

    3 years ago it was TC 168, TGs 65, HDL 48, and LDL 107

    My paleo diet (a little less than a year) heretofore consisted of a lot of fatty cuts of grass-fed lamb, perhaps ungodly amounts of (usually clarified) pasture butter, at least two pastured eggs per day, half a pound of spinach per day, and enough potato to replete my glycogen stores. I also ate L. digitata (the most concentrated natural source of iodine I could find) every day. After reading the PHD, I tried to eat liver but it thoroughly disgusted me, but having read all of this, I’m going to give it a go or at least take a Cu supplement. I was attempting to eat a lot of mushrooms for Cu, but it, assuming all of this is correct, was insufficient. Supplements were 3g of vitamin C a day and 5000IU of d3. Still waiting for lab results for d3 and hba1c. I asked for and was told that I was getting a VAP test as well, so I’m not entirely sure why it is that a standard panel was run, but here we are. I’m certain that my doctor will now listen to me with this result.

    Before paleo I had 8 years of a vegan diet followed by a couple years of a whole grain yuppie diet.

    My personal theory, before coming upon this post was that I had, especially from the vegan diet, caused enough damage that when I finally supplied my body with the correct raw materials, it massively increased cholesterol synthesis and transport in order to repair damage. It seems that a diet high in PUFAs/n6 alone would lead to cell membranes that are improperly constructed and thus there would be a high demand for the proper materials as those cells degrade and are replaced.

    Why then does this phenomenon happen with paleo but not SAD? I am of the opinion that the average person on the SAD would be deficient in both micronutrients and/or have various types of damage that needs to be repaired but also be deficient in saturated fat. When the latter is supplied, a spike in LDL-C is seen not because there is a sudden demand for it, but because the demand is finally met with an adequate supply of dietary cholesterol and/or the materials used to manufacture it endogenously.

    As for whether this number should be troubling, I would point to the surveys of average TC among several HG groups, with none that I have seen ever crossing the magic 200 mg/dl threshold. This tells me that this aberration should at least be taken seriously.

    As I don’t appear to be afflicted with a hypothyroid condition or a deficiency in either vitamin C or choline, I can only assume that Cu-deficiency is the primary driver for this, assuming that this is all correct. It’s certainly worth a try as the toxicity at the suggested dose appears to be quite low. I generally have a strong hesitance to supplement unnecessarily, but it’s reasonable to assume that nearly all HGs would be consuming liver quite regularly and thus would not encounter this. I’ll report back after a few months of this supplementation with new results.

  45. Why not simply test micronutrient levels and confirm if/what supplements are required for the individual? This would avoid the issues of oversupplementing certain nutrients and missing or undertreating the exact ones that the individual is deficient in.

    Something like this?

    It covers all of your “essentials” and many of your “suggested” nutrients. If the person has health insurance, the cost is about what one or two months of supplements run for most people…and the “anti-supplement” crowd would only need to take the minimum number of supps.

  46. Travis, one suggestion re the liver if you’ve got the $$. I get US Wellness’ braunschweiger (40% liver) and eat that rolled up in sliced roast beef with some liberal use of spices.

  47. Stas,

    I think what you really mean is… supplementing with Copper to a point that brings you into a state of Copper toxicity is potentially dangerous to your health.

    …because ‘supplementing with copper’ is not definitely dangerous advice for everyone.

  48. Hi Stas –

    Copper and zinc do need to be in balance.

    Her book seems to be written for vegetarians. As she says, plant foods tend to be higher in copper, lower in zinc; animal foods high in zinc, low in copper.

    Copper deficiency is much more common than zinc deficiency in the general population, so it must be far more common among low-carb Paleo dieters.

    Moreover high copper does not cause zinc deficiency, but high zinc can cause copper deficiency.

    We recommend a multivitamin containing zinc, as well as food sources.

    The evidence for copper toxicity at doses up to 4 mg/day is nonexistent as long as zinc is adequate.

    So while her concerns may be relevant to vegetarians eating copper-rich foods like soybeans, I don’t think it’s a significant concern for Perfect Health Dieters or low-carb Paleo dieters generally.

    Best, Paul

Leave a Comment

NOTE - You can use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>