Low Carb Paleo, and LDL is Soaring – Help!

To Kindy, Zach’s parents, and the NBIA/PKAN kids: I’ve been reading papers on the disease and trying to figure out the best diet for the disease. But the biochemistry is a bit complex, more complex than I realized last week, and I want to make sure my advice is sound. So I’m delaying my NBIA/PKAN/ketogenic diet posts until next week.

My sincere apologies for the delay!

I’m a little busy this week – busy with work, busy with learning about NBIA/PKAN, and eager to spend time with my brother who is visiting from Germany – and so I thought I’d do a “You be the doctor” quiz.

Here’s the puzzle. Someone adopts a low-carb Paleo diet. Very healthy diet, right? But their LDL cholesterol level starts to rise. And rise. And rise.

Larry Eshelman emailed me last December with this problem. His LDL history:

  • 103 mg/dl (1990-2002, eating a low fat diet)
  • 115 mg/dl (2002-2007, eating a low carb diet)
  • 195 mg/dl (2007-2009, after reading Gary Taubes and adding saturated fat)
  • 254 mg/dl (Dec 2009, very low-carb Paleo for 5 weeks)
  • 295 mg/dl (Jun 2010, very low-carb Paleo for 7 months)

(SI system readers, convert to mmol/l by dividing by 38.67.)

A common problem

This is not a terribly uncommon problem in the Paleo community; it afflicts famous and brilliant bloggers as well as ordinary folks. It’s been discussed by Richard Nikoley in several posts:

Some examples of high LDL on a Paleo diet, with links – most of these provided to me by Larry (thanks Larry!):

OK, that’s enough: this is a minority phenomenon, but it’s definitely not an exceptional n=1 phenomenon.

Larry’s Progress

Larry wrote me at the beginning of December asking for advice. He implemented everything I suggested. I just heard back from him this week with new data.

His LDL decreased from 295 mg/dl to 213 mg/dl in a recent test. His HDL rose from 74 mg/dl to 92 mg/dl. His triglycerides fell from 102 to 76 mg/dl.

LDL is still high, but improving; the others are excellent and improving.

So, quiz questions:

  • Can you guess what my December advice to Larry was?
  • What causes these cases of soaring LDL on Paleo? (Of course, there are multiple possible causes of high LDL, but I think among Paleo dieters one explanation is more likely than others, and that’s what I’m looking for.)

My answers tomorrow night.

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

Leave a comment ?


  1. Resolution of non alcoholic fatty liver disease

  2. I have been puzzling over this, a couple of people on Robb Wolf’s forum have been asking the same question.

    I know I can get perfect cholesterol with a strict Zone diet, and have seen this in clients. It is moderate protein, not high. It is moderate carbs (100 – 200 grams day) It is lowish in fat and mainly monounsaturated fat. Low in saturated fat, with an optimal n-3 to n-6.

    Saturated fat increases my LDL as does omega 3.

    There is the theory of large fluffy LDL vs small dense rendering the calculated LDL incorrect. Friedewald vs Iranian formula http://homepages.slingshot.co.nz/~geoff36/LDL_mg.htm
    But I suspect this is not the full story as the HDL, TG and LDL improved with a change in diet.

    Stress? Cortisol’s effect on production of LDL, TG?
    Stress from too low carbs over time, combined with too much exercise, other stressors?
    Found this study:
    VARIATIONS UNDER STRESS IN MALE ADULTS http://www.ayubmed.edu.pk/JAMC/PAST/21-1/Mujadid.pdf
    The vitamin C connection to low carb diets as you have mentioned before – another study:
    Vitamin C supplementation lowers serum low-density lipoprotein cholesterol and triglycerides: a meta-analysis of 13 randomized controlled trials

    Other micro-nutrient deficiencies?
    Thyroid function depressed on low carb diet affecting cholesterol? The number and activity of LDL receptors decrease in hypothyroidism.

    Glyceroneogenesis? http://www.ncbi.nlm.nih.gov/pubmed/14739071

    My guess is that you may have told Larry to increase his carbs to 100 – 150 grams day, especially from safe starches. Add some vitamin C. Don’t be excessive in protein. Use coconut oil. Tackle stress. Add choline.

    Just thoughts, look forward to what your answer is.

  3. Advice: watch and wait.

  4. Add more safe starches? 😛

    Was the Larry having his LDL “calculated” with a “formula” or was it a true lipid count. 295 is a little far out there for it to be all in the math, but the tris may still have bearing. For example, I just got mine back. TC: 212. HDL: 44. Tris: 75. LDL: 154.

    Now, I know 154 isn’t all that high, but it’s still a bit curious. I was instructed to follow a low-cholesterol diet (which isn’t happening). After recalculating using the Iranian method, my LDL “dropped” from 154 to 139. Curiouser and curiouser…

  5. weight loss (and VLC diet) resulting in hypothyroidism resulting in elevated cholesterol due to less pronounced LDL receptors?

    Seriously… cannot wait for the post tomorrow night!

  6. well… subclinical hypothyroidism of course.

  7. Not sure what you advised, but I advised my BF to correct his Omega 6 to 3 balance from 18:1 to 1:1, and after 3 months his LDL went from 245 to 190, trigs went down by same percentage and HDL increased by 20%. He cut way back on the canola and olive oil, and he megadosed fish oil for a month at 2000g per day and then went to 400g per day maintenance dose, plus started taking magnesium at 600mg per day.

    We set this up before some of the warnings about too much fish oil started showing up on the paleo blogs, so I’d add the “don’t try this at home” disclaimer. But we’re happy with the results.

  8. “What causes these cases of soaring LDL on Paleo?”

    —Shift of VLDL to large particle LDL.

  9. What Lacie said!
    Maybe from nuts. Some paleo-folks go nuts on the nuts. 😉

  10. I’m with Mike on the large particle LDL. The increased fat intake may raise lipid numbers but the increased LDL would be made up of preferred large particle LDL.

  11. I’m thinking something similar to what Mike mentioned. And is it a case of grouping sdLDL and the light fluffy LDL together? From my understaning you must specifically ask for seperate measurements.

    So an increase in dietary fat has has made for healthy LDL particles as they’re now loaded with fat and cholesterol…making them light and fluffy. The higher numbers is more a reflection of larger particles sizes rather than total numbers. And it seems his overall improving lipid profile would back this up.

    Did you suggest any fasting?

  12. I don’t know what I was smoking when I gave those fish oil numbers, but I meant to say 2000mg DHA/EPA per day and 400mg DHA/EPA. Other data was right the first time. BTW, he accomplished all this while still hypothyroid (TSH 6.25) so once he gets that cleared up, his trigs should improve even more.

  13. anand srivastava

    How about adding back some starches?

  14. Hypothyroidism from low carbs.

    It may also be an antioxidants deficit (vitamin C helped me lowering LDL).

  15. Paul, don’t have a link handy but there was a recent Chris Masterjohn post about this phenomena. His hypothesis seems to be that the high LDL might be the liver cleaning itself out after years of sugar abuse.

    I suppose one possible way to speed the process would be high dose fish oil, like 10g per day for a time. Stephan has had at least a couple of readers bring down their AST/ALT enzymes substantially within a very short time, from clinical fatty liver levels to normal.

    Just speculation on my part.

  16. This is perhaps the post that Richard has in mind:


    (This is also from Chris, –on paleohacks 12hours ago: “Two posts from now in the latter series, which should come out in the next week, I’ll be covering whether LDL matters.

  17. ok, so story so far as best as I can piece together (sorry for lack of attributions):

    (a) hypothyroid…(more safe starch…

    (b) fatty liver… (more fish oil (but likely better not to and just wait), …

    (c) LDL receptor… (curcumin???…

    (d)don’t worry advice: (i)pay attn to sd vs fluffy LDL, and/or oxLDL and/or TG/HDL, HDL/LDL … ratios, (ii) don’t pay attn to lipid No’s on PHD/PaNu/… perhaps unless it’s FH (then perhaps look at CAC…)

    very much looking forward to Paul’s evening post

    • Hey donat,
      You post is the first I’ve seen mentioning FH. Seems that is what I’m dealing with. You recommend looking at CAC. What does that stand for? Any other advice/suggestions regarding FH much appreciated.

  18. 1. Add safe starch
    2. Check a TSH

    Not sure I buy the fatty liver tale…

  19. The lipid hypothesis raises it’s ugly head again! LOL

    Treat the symptoms, Larry, not the numbers.

  20. Larry Eshelman

    In answer to Becky’s question, the high LDL numbers (post reading Taubes) are all measured LDL-C, not calculated LDL-C.


  21. I’m the guy with the 585 TC. It went down (to 378 8 months or so ago, time to check again) when I started supplementing with iodine. My TSH has also been trending up the last few years, even before Paleo. So hypothyroidism is my primary suspect. I also kind of like Masterjohn’s NAFLD theory.

    Looking forward to tomorrow’s post. 🙂

  22. I would say it is genetics combined with diet composition. It is my understanding that many people will increase their cholesterol on a high saturated fat diet on account of the fact that the liver is no longer afraid that all your cholesterol is going to oxidize as soon as it is released. Those cholesterol numbers on the SAD would not be good but on a high fat diet, they don’t worry me. I believe the most salient numbers related to heart disease are generally triglycerides and HbA1c.

    I bet your advice involved coconuts and fatty fish.

  23. Larry Eshelman

    Several commenters have suggested that my high LDL-C reflects large fluffy LDL. When I saw my LDL-C rise, this was what I assumed. However, I had an MNR on Dec 16, 2009 with the following results (where P refers to particle number and the numbers in brackets are the recommended values):

    LDL-P 2674 [<1000]
    sm-LDL-P 720 [=4.8]
    Large VLDL-P 1.2 [20.5 pattern A]
    Lp(a) 6 [<20]

    So Lp(a) is fine, and the pattern is type A which is good. But the LDL particle numbers are not reassuring.

    My doctor prescribed statins, which I refused to take. I agree with Poisonguy that treating the numbers is a bad idea. On the other hand, my LDL (C and P) is much higher than that found in any healthy, traditional culture — this is unchartered territory, which is a cause for concern.


  24. Larry Eshelman

    Oops, it didn’t like my less than or equal signs and messed up the numbers. Here are the MNR results again:

    LDL-P 2674 [less than 1000]
    sm-LDL-P 720 [less than 527]
    Large-HDL-P 7.2 [greater than 4.8]
    Large VLDL-P 1.2 [less than 2.7]
    LDL part size 21.9 [greater than 20.5 pattern A]
    Lp(a) 6 [less than 20]

    • Hi Larry
      Your numbers look a bit like mine. Have you been doing anything in the last year to get the LDL-P down?

  25. I would say that I don’t believe that high LDL is necessarily a problem until I see a clealy run study of large group of people that show a correlation of high LDL and CVD when those people adhere to a “paleo” or “lacto-paleo” style diet, or a close variation of whatever paleo means… (specifically a pure foods diet rich in sat and mono fats and very low in poly, adequate high quality proteins, low sugar/fructose, virtually no grains).

    Let me know when you see something like that. I just refuse to believe that all these people are at risk for heart disease, yet most other factors of their health are ok. Is the body being completely fooled, meaning that it’s very happy and content with diet from a gazillion other angles, but is confused about LDL? I think not. I think our understanding of LDL is not complete at this current time.

    I’ll take a high, healthy LDL over a low, unhealthy LDL any day.

  26. Here are my two best guesses:

    1. Sleep more, in total darkness.
    2. Add some safe starch.
    3. Fast every once in a while.

    For those saying that LDL is not a problem, look for studies showing the total cholesterol of hunter-gatherers. They don’t pass 180 mg/dl, not even the Inuits. I am not saying it’s not a problem but why does a similar diet cause such different results?

  27. Did you advise him to take iodine? Could be hypothyroidism that causes such large increases in LDL in many of us.

  28. Larry Eshelman

    This is a quote from Stephan Guyenet, in response to a commenter:

    “I’m a little bit worried when I see people with sky-high LDL as well. The associations between LDL and heart attacks were nearly all performed on people eating a standard Western diet. So I don’t know how much it applies to someone with an atypical diet. Still, I do recognize the point you brought up that HGs [Hunter/Gatherers] had generally low TC and LDL. That makes me think it’s our natural state. Yet most HGs eat a lot of animal foods and some eat plenty of animal fat.” [Comment section of Stephan’s post “Animal Models of Atherosclerosis: Diet-Induced Atherosclerosis”, posted Sept 7, 2009, comment Sept 8]

  29. Another vote for Chris Masterjohn’s idea:
    1) Before paleo is started, some level of fatty liver exists (e.g., due to a fructose rich, choline poor diet).
    2) Paleo eating is begun, including a lot of choline rich foods: egg yolk, liver, meat etc.
    3) The choline allows the liver to finally dump the accumulated fat into the blood.
    4) End result: high lipids.

    In addition, when it is a low carb diet, insulin is likely to be low. This could lead to less mobilization of lipoprotein lipase (LPL) from adipocytes [1]. Low LPL would cause poor lipid clearing, and so they just keep circulating.


    [1] Knutson V.P. The release of lipoprotein lipase from 3T3-L1 adipocytes is regulated by microvessel endothelial cells in an insulin-dependent manner. Endocrinology. 2000 Feb;141(2):693-701. http://pmid.us/10650951

  30. Forgot the other quiz question.

    My guess about your December advice to Larry (based on the insulin-LPL connection):
    Stop very low-carb and add safe starches to the diet. (Or: drink milk. Should help get the insulin up as well.)


  31. My LDL as well as my HDL have also increased since starting paleo about 5 years ago. Switching from cooked to a RAW paleo diet two years ago decreased my LDL and increased my HDL even more. At the age of 24, my LDL levels are 137 mg/dl and my HDL is at 162 mg/dl. I eat a very high RAW cholesterol diet which I believe helps to raise my HDL levels. The key being RAW unoxidized-undenatured high cholesterol foods. Typical day would look be:

    Breakfast: 1lb. RAW WC-Salmon + 12 Oysters
    Lunch: 1lb. RAW Lamb Liver or 1lb. Lamb Chops + 1 Coconut
    Dinner: 1 lb. RAW WC-Salmon or 10 eggs + Plantain or Butternut Squash

    If hungry, I may snack on bell-peppers, green parsley-cilantro juices, veggies or raw eggs.

  32. Larry’s numbers (8 years after low fat diet LDL still rising) don’t seem to provide much confirmation for the fatty liver hypothesis at least for his case. Do we have other longer LDL number histories on paleo (as opposed to one-spot or before-and-during-the-diet data)?

  33. Personal experience from over 3 years ago (I don’t remember the exact numbers) – low carbohydrate diet caused a huge increase in measured LDL – >10 mmol/l. Thyroid was normal (borderline hyper actually). Then got retested in about 9 months and LDL was ~3 mmol/l. Take it for whatever it’s worth.

  34. Paleo newbie here.

    Ooh, ooh–can I take a guess? I’m having the same thing happen, and my doctor so badly wants to put me on Lipitor, but I’m fighting her off with a baseball bat. I’ve also been taking milk thistle and dandelion root for years to keep my organs clean, and eating chia seeds, yet this is still happening.

    My guess is polyunsaturated fats from too many peanuts (or other sources).

    I recently watched a cooking show on PBS about how type of fat affects baking (brownies), and it turns out that on the poly-to-saturated scale, the closer you get to poly fat, the chewier the brownies became. The closer to saturated fat you got, the drier and fluffier the brownies became. The proper texture for brownies (it turns out) requires a 60/40 mix of poly to sat fats. Good thing brownies aren’t Paleo!

    Maybe this goes on inside our bodies too?

  35. Hi Wenchypoo – My answer is here: http://perfecthealthdiet.com/?p=2547

    You’re right to stay off Lipitor — nutrition is the better solution! But I’m not so sure about milk thistle, dandelion root, and chia seeds.

    If polyunsaturated fats make us dry and fluffy, that might explain why I have more muscle without them!

  36. Poly fats make the brownies dense and chewy, and the sat fats make them fluffy. I wonder if it also makes cholesterol do the same things…poly makes more VLDL, while sat makes more LDL?

    I’m not a scientist, but a homemaker. The brownie answer guy on PBS was a food scientist–maybe we need more food scientists and less MD’s?

    Milk thistle and dandelion root extract are supposed to keep your insides clean from fatty liver and the like. I’m eating chia seeds for the fiber and protein boost, while eating everything else raw except the meat and eggs, and so’s hubby. His recent doctor visit shows only a slight elevation from last year’s cholesterol check, and he goes in again in June. I don’t go in until May, and the baseball bat’s going with me just in case.

    It’s just so hard to get over our doctors using “the standard of norm” from fifty years ago, and when you try to bring them up to speed, they take a baseball bat to YOU! I suppose the government, insurance companies, and drug makers are in cahoots somehow to keep us ALIVE and not HEALTHY. I mean how long has Weston A. Price been around, yet nobody “official” has bothered to look into his work?

    Although it seems that slowly, some medical researchers are starting to come around, and Price and other seemingly futuristic nutrition researchers (Cordain) are being discovered to be right. Maybe by the time I die, medical heads will have been pulled out about such things as cholesterol levels.

    I’m not waiting that long. When the Obamacare program becomes official in 2014, I’m dropping off the insurance radar completely, and using a nutritionist as my primary care doctor (for cash).

    Speaking of doctors, does anyone know of a a Price or Paleo-friendly doc in VA? I’m really getting tired of having to bring these doctors up to speed! It’s like all they’ll ever know, they learned way back in med school…or whatever crap the NIH or FDA spews out.

  37. LDL Goals for Type 2 - pingback on March 7, 2011 at 6:02 pm
  38. Increase cardio in exercise will imrove HDL level. Change protien/ fat to be more seafood and less red meat. Increase good fats to coconut oil, fatty fish, and nuts. Limit animimal fat. Cut out dairy completely.

  39. FACTS:
    Male, white, 52
    approx. 130lbs fat-free mass
    Bodyfat: approx. 19%

    GOAL: <10% bf using a LC PaIeo diet combined with Intermittent Fasting which means I must lose at least 17lbs of bf, both subcutaneous and visceral.

    Resistance train 4xs/week
    Cardio interval train 2-3xs/week

    I was dx'd via ultrasound with non-alcoholic fatty liver disease (NAFLD) two years ago. It is indeterminate how long I've had it. Liver enzymes continue to remain within normal range and I notice no symptoms

    I am hypothyroid and have been taking 75mcg of sustained-release T3. My latest thyroid numbers are posted in the link below [p2, 3, 10].

    TSH has improvement over last labs. Both T4 and T3 show depressed, however, I was fasting for nearly 20 hours on the first day. rT3 is less this time, which is a good thing.

    Have been on LC Paleo diet for last three months and doing between a Lean Gains [ http://www.leangains.com/2010/04/leangains-guide.html ] 16/8 IF protocol to as much as 20/4 IF protocol. I do mini-carb refeeds (about 50-60g) on my workout days only. If I increase carbs more than this, my post-prandial BG will exceed 125.

    What's freaking me out are my latest lipid tests. I had the top three done in a matter of two consecutive days [p2, 5, 11-14].

    LATEST LABS: http://s57.photobucket.com/albums/g207/CS2006/LABCORP%2006-19-12%20and%2006-20-12%20LABS/

    If we’re clearing lipids from the liver, then this is a good thing, but HOW CAN I DETERMINE THAT IT'S THIS AND NOT FROM THE DIET ITSELF?

    In other words, is the increase due to CREATION or CLEARANCE (resolution of NAFLD)?

    I have not changed any macros in my diet, maybe slightly more lean grass-fed animal protein, but saturated fat intake has remained constant throughout. And most of the saturated fat is drained because I steam all my meats, so how can it be CREATION?

    The only other fats I eat with frequency are O3s (2-4g), coconut, flax, macadamia and olive oil.

    What evidence supports this unconventional theory?

    I understand that one of the key problems with fatty liver disease is that the lipids get stuck in the liver and they’re not being released into the bloodstream.

    How could this be the case when taking liver support supplements like milk thistle, dessicated liver, choline, lecithin, etc.? Why wouldn't such intervention spur on the purge also?

    Why wouldn't free fatty acids (FFAs) from stored subcutaneous fat be released into the bloodstream as well?

    Could this explanation be the mechanism behind the clearance of FFAs: During fasting or starvation, free-fatty-acids are released during lipolysis into the liver and muscles to be burned as energy, this is called fat-oxidation. During the fed-state and especially while eating a starch-based-diet, fat-oxidation is inhibited and replaced with carbohydrate-oxidation, insulin is what mediates this shift. When carbohydrate-oxidation is taking place, fatty-acids are shuttled back and "locked away" in adipose-tissue… where they belong.

    In addition to LC Paleo/IF, I also began taking 1g of choline nearly a month before the labs + 3mg methylfolate/day to help with a genetic methylation defect.

    Could the above combination have created a mega-purge?

    Could the answer be that the best predictor of fatty liver is obesity and insulin resistance?

    Another thing:
    I have been on warfarin for 3 months and must remain on it for another 3 months. This was the only thing I could find on warfarin and its effects on blood lipids:

    Coumadin binds to bile acids in the intestine leading to increased excretion of bile acid in the feces leading to increased oxidation of cholesterol to bile acids resulting in increased numbers of low density lipoprotein receptors with increased hepatic uptake of LDL and lower serum cholesterol levels
    source: http://www.flash-med.com/Side_Effects_Coumadin.asp

    Anomalies to purge theory:

    Why the decreased HDL when I was making nice progress before?

    Lastly, I had been fasting for 18-20 hrs prior to my blood being drawn. My fasting insulin was only 5.2. Then why an elevation in HbA1c (5.8) and FBG (95)? Should've been in the low 80s, especially when fasting for LONGER periods AND on LC Paleo.

    This even when on a broad range of BG-lowering natural agents including corosilic acid, chromium, cinammon, etc.

    So, why is my insulin sensitivity is taking a nosedive during this so-called healing crisis as well [see Insulin Resistance Score – p12, 14]?

    Can someone please interpret my lipid profiles, especially the NMR LipoProfile and tell me what is going on?

    • Your insulin sensitivity is taking a nosedive because a paleo diet that is low carb and high protein/high animal fat promotes insulin resistance. More specifically, it promotes excessive intra-myocellular lipid storage, which does 2 things: 1. inhibits insulin from effectively binding with the cell membrane to facilitate glucose getting into the cell, 2. Turns OFF the gene that codes for mitochondria – so the # of mitochondria in your muscle cells decrease so you end up not burning fat as well.

  40. CALCIUM D-GLUCARATE | hypoallergenic diet 2011 - pingback on October 17, 2012 at 10:12 am
  41. I put different variations of “ldl paleo” into google and this was one website among many that detailed how many have had high to very high LDL after doing paleo.

    Speaking as a former paleo promoter/dieter myself – I find it comical how the OP and many commenters find all these creative ways to explain away high LDL and its real significance in relation to atherogenesis. Whether its lack of choline and the liver is dumping, the large “fluffy” ldl, oxidized ldl, blah blah blah.

    Guys, the paleo diet as defined by sisson, cordain, taubes, this blog owner, or masterjohn’s variation are all ATHEROGENIC (amongst other unhealthful things). Do not take their word at face value! Instead of relying on secondary sources of information like these bloggers, how about reading the primary sources? (The actual science?). The preponderance of evidence (>10,000 studies) shows that a whole foods plant based diet with <10-12oz of meat per week (and not red meat) reverses atherosclerosis, there is no other way!

    I know where you are coming from, I was in the same boat. I loved paleo. But you can only explain away the facts for so long…as long as you read from websites like masterjohn, sisson, and this guy. If we are to be science based – we must be willing to put down our confirmation bias and blind spots, and actually acknowledge truth. Read the other side. Ornish, Esselstyn, Fuhrman, Barnard, Greger, McDougall. Read about huge former paleo blogger Don Matesz – who was paleo for 14 yrs but finally realized he was hurting himself and made it public.

    Best of luck to you all. At first it seems like there is no consensus on the lipid-hypothesis, and the ancel keys was completely wrong. That there is so much controversy over this subject and no one agrees! But if you actually get into the primary sources, you'll see that keys has been wrongfully demonized and he was right all along.

    check out http://www.primitivenutrition.com

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>

This site uses Akismet to reduce spam. Learn how your comment data is processed.