A reader pointed me in the direction of this podcast on Chris Kresser’s show where he interviews Chris Masterjohn about cholesterol, and I have to say, this is perhaps one of the most informative things I’ve read on the subject.
Chris Kresser is the author of Your Personal Paleo Code and Chris Masterjohn is the author of Cholesterol-and-health.com.
Here are my notes from Part 1.
If you want to listen to the whole podcast, you can listen here.
Here are my notes from Part 2 (there are three parts in total):
[otw_is sidebar=otw-sidebar-1]
What are ‘normal’ cholesterol levels?
What anthropological studies from various traditional cultures show:
Masai – Pastoralists; cattle herding tribe in Africa on the border of Tanzania and Kenya. They raise cattle to eat the meat, drink the milk, etc.
- Their diet varies with the seasons and the economy, because they’ll trade with other groups when their own supply of food is low or other groups are more relatively wealthy.
- At other times they may have an abundance of milk and might focus mostly on consuming milk.
- They have very low cholesterol levels, some of the lowest in the world.
- Men had cholesterol levels between 130-135 mg/dl
- Women had somewhat higher cholesterol levels of about 140 mg/dl, that would increase with age to 170 mg/dl in their 40s. The cholesterol also seemed to increase in their 3rd trimester of pregnancy, up to 205 mg/dl.
- On the whole, the Masai had a population mean cholesterol of 130-140 mg/dl and an LDL:HDL cholesterol ratio of 2.
- They have a very low rate of heart disease.
Kitavans – Horticulturalists who live in the Pacific Islands. They eat a little fish, a little coconut, a lot of tubers, and a little bit of fruit. They have very low fat intake, very high intake of saturated fat from coconut.
- Males had cholesterols around 180 mg/dl
- Females were a big higher at around 200-210 mg/dl, that rise with age to around 250 mg/dl in their 40s or 50s.
- A lot of Kitavans smoke.
- Their LDL:HDL ratio is 3.3 among the smokers and 2.5 among the non-smokers.
Puka-Pukans – Another group of in the Pacific Islands have a high intake of starches and other carbohydrates and a fairly meaningful intake of coconut.
- Men have cholesterols of 150-180 mg/dl
- Women have cholesterol of 170-190 mg/dl
- Tokelau – Yet another group of Pacific Islanders, who eat a much higher amount of coconut.
- Men have cholesterol of 180-220 mg/dl
- Women have cholesterol of 200-245 mg/dl
Masterjohn points out that if we examine these populations, what we see is a range of cholesterol. He thinks that the Masai should constitute the low range of what is possibly healthy, and the Kitavans should probably serve as the high range.
[otw_is sidebar=otw-sidebar-2]
So we could probably say that anything between 140-220 mg/dl and maybe even up to 250 mg/dl might be considered normal.
Masterjohn also points out that we DO NOT see healthy populations with cholesterols going through the roof in the high 200s or over 300.
There can many things that lower cholesterol and many things that raise cholesterol, and some of these things can be good, and some of them bad.
- If we’re lowering cholesterol because we’re clearing it into an atherosclerotic plaque, that’s clearly bad.
- If we’re lowering cholesterol because we’re converting it into sex hormones and bile acids and things that increase our fertility, virility, strength, and digestion, then its clearly good.
- If it’s low because we’re not making it, then we don’t have those substances that we need to make all those good things, so this is clearly bad.
So we can never judge the amount of cholesterol in our blood alone, but have to use it as a clue to understand what’s happening in the body.
[otw_is sidebar=otw-sidebar-3]
Advanced Cholesterol Panels
Masterjohn points out that from the data, if you want to have a pattern A large, fluffy LDL particle type, then the best thing to do is to change the sort of test you get rather than changing diet or life style, because LDL particle size measured by the different methodologies (NMR, VAP, and gel tube electrophoresis) have such widely varying results.
- Gel tube electrophoresis – 80% of people who measure their particle size using this method are classified as pattern A.
- VAP – Only 8% of people using this are classified as pattern A.
- NMR – some other number between 80 and 8%.
There’s currently a lot of disagreement between these different tests, and until we get the science to the point where we can actually agree on particle size, what test to use, and what standards to apply, we’re not at the point where we should be ordering these tests for people and telling them to change their diet or modify this or that, to alter their LDL particle size.
The main reason this became so popular were due to early retrospective studies that looked at people with heart disease and saw that people who had the larger particle sizes seemed to have a lower prevalence of heart disease.
This observation generated hypotheses about the role of different LDL particle sizes in heart disease.
There were a number of studies that tested these hypotheses, and most of them did show that particle size, whether measured by gradient gel electrophoresis, NMR, or in one case by the apoB to LDL ratio, had some success in predicting cardiovascular disease.
BUT, in most of these studies, when they were adjusted for traditional risk factors, the predictive power was mostly lost.
[otw_is sidebar=otw-sidebar-4]
Quebec Cardiovascular Study – They studied particles using gradient gel electrophoresis and found that size was an independent predictor of heart disease risk.
Then, a number of followup studies failed to confirm this.
- Stanford Five City Project – Showed that the particle size was useful, but once they adjusted for the total:HDL cholesterol ratio, it wasn’t useful anymore.
- The Physicians Health Study – Showed that particle size was useful, but once they adjusted for triglycerides and total cholesterol, it wasn’t useful anymore.
- Norfolk Study – Measured with NMR, and found the particle size was useful, but then found that particle number was much more useful. However, when they adjusted particle number for HDL and triglycerides, once again its usefulness was questionable.
- AMORIS study – looked at particle size estimated by the Apo B to LDL ratio which once again seemed useful, but when they adjusted it to triglycerides or to Apo B, it lost its usefulness.
- Womens Health Study – Looked at particle size with NMR, which lost its usefulness after adjusting for total:HDL ratio.
Kresser succinctly states, “We can neither prove that particle size is not an issue nor prove that it is an issue with the current test because all of that rests on methodology that hasn’t been standardized or agreed upon.”
According to Masterjohn, what we are currently left with after all this are 4 hypotheses:
- Small dense LDL is more likely to oxidize and thereby contribute to plaque development
- Small dense LDL is more likely to pass through the endothelium where it is oxidized behind the blood vessel wall, contributing to plaque development
- Small dense LDL is more likely to get stuck behind the vessel wall where it can be oxidized and contribute to plaque development (I don’t really see what the difference is between 2 and 3).
- Small dense LDL serves as a marker indicating poor LDL receptor activity, and the oxidation occurs due to the amount of time LDL remains in the blood stream.
At this point we have more studies addressing the topics of total:HDL and triglycerides and better agreement on what tests to use to measure these things. There is also good data suggesting that at the population level, there is predictive value to these.
JAMA Article in 2009 – Meta analysis which showed that fasting triglycerides, HDL, and non HDL cholesterol were all useful predictors, but if they adjusted triglycerides for the other predictors, triglycerides lost their significance.
But, if they adjusted HDL and non-HDL cholesterol for triglycerides, nonHDL and HDL cholesterol persisted in their predictive power.
Lancet 2007 – Meta analysis including 61 studies with 900,000 subjects, found that if you looked at total cholesterol, HDL, or LDL, or any of the ratios, it was the total:HDL ratio that really stood out as the best predictor.
[otw_is sidebar=otw-sidebar-5]
So what can we say about this?
We can only say that on a population level, it can be used to predict risk, but we can’t extend this to an individual level. On an individual level, it should never be used without further investigation to see why the ratio is out of whack.
Maybe it’s out of whack because of thyroid issues, insulin resistance, fatty liver, etc. These need to be explored.
Why is my cholesterol so high after adopting a nutrient dense, whether we’re talking about a Paleo or a Weston A Price style diet that’s higher in saturated fat?
Masterjohn first addresses the normal variation of cholesterol tests. There is a standard deviation of about 17 mg/dl. So without any changes in diet or lifestyle it can go up about 2 standard deviations, so up about 35 mg/dl or down about 35 mg/dl for no reason at all.
Kresser points out that this is a huge range and could mean the difference between a statin or no statin.
Masterjohn’s general rule if you’ve only measured your cholesterol twice:
- Total cholesterol: Must see an increase or decrease greater than 35 mg/dl before you can be 95% confident that the change is real
- HDL: Must go up or down 9-10 mg/dl before you can be 95% confident that the change is real
- LDL: Must go up or down 30 mg/dl before you can be 95% confident that the change is real
- Total:HDL ratio: Must go up or down by 0.8
- LDL:HDL ratio: Must go up or down by 0.8
- Triglycerides: Must go up or down by 40 mg/dl
Masterjohn’s next step is to ask if there has been any weight change. His working hypothesis given the absence of clear research in this area, is that if you are obese or overweight and you normalize your weight and normalize all the other metabolic parameters, in the long run, blood lipids mostly improve.
In the mean time, it’s important to always get measurements at the same time of the day. Triglycerides are sensitive to fasting, so they will be lowest in the AM after an overnight fast.
[otw_is sidebar=otw-sidebar-6]
Things also change however in the setting of a prolonged fast. Because there isn’t any food in the system, blood triglycerides will get cleared over the course of a day or so of fasting.
The body then starts to burn its own adipose tissue releasing free fatty acids into the blood. The free fatty acids have 3 important fates:
- They go to the liver where they are converted into ketones to provide energy for the rest of the body
- Free fatty acids can interfere with thyroid hormone signaling, which itself can affect blood lipids.
- There is good preliminary evidence that unsaturated free fatty acids are especially good at inhibiting thyroid function (PUFAs and oleic acid)
- Concentrations of free fatty acids can get high enough in the nucleus of the cell to inhibit signaling there, which will not be reflected by free T3 level in the blood, making this hard to detect (what about reverse T3?)
- If the liver’s capacity to burn fatty acids for energy is overwhelmed, the liver can convert them back into triglycerides and send them back out in to the blood.
- People with a high capacity to burn fatty acids for energy might see their triglycerides go down as their losing weight
- People with a lower capacity to burn fatty acids can see their triglycerides go up as they lose weight as the liver releases more into the blood.
So someone who has very good micronutrient status, has good thyroid hormone status, and has all the things that we would expect to promote burning energy at the liver, will burn energy really efficiently and their blood lipids might improve the whole time.
Someone else who’s losing weight might not have enough choline to help the liver process fatty acids, so that fats get stuck in the liver and cause fatty liver. Choline is found in liver and egg yolks.
Masterjohn summarizes his thoughts:
- First make sure the change is large enough to be considered a true change
- If you’ve been losing weight, don’t even check blood lipids until a few months after weight has stabilized
- Make sure you test at the same time of day, fasting for the same amount of time, and under similar conditions as the prior test (especially for triglycerides).
Part 3 coming up soon…
Nice summary – from what I remember from that podcast, you’re pretty much spot on!
Sorry to nitpick, but if you were talking about p-values when saying “Total cholesterol: Must see an increase or decrease greater than 35 mg/dl before you can be 95% confident that the change is real” then that’s not a correct way to state it. “95% confidence” actually refers to the there being only “a 5% chance of seeing a difference at least as big as we have done, by chance alone”. So, it’s a statement about the level of confidence that such a result arises by chance alone – not that one can be 95% sure that the result is true. It seems to be a distinction without a difference but it really isn’t, it’s massively important when interpreting data. Reference: “An investigation of the false discovery rate and the misinterpretation of p-values”
Unfortunately I don’t know enough about statistics to comment on this… but what you’re saying seems to make sense!
Have you seen this?
http://www.sciencedaily.com/releases/2008/07/080718075357.htm
Regular Walking Protects The Masai — Who Eat High Fat Diet — From Cardiovascular Disease
http://www.researchgate.net/profile/Per_Waendell/publication/51399410_Daily_energy_expenditure_and_cardiovascular_risk_in_Masai_rural_and_urban_Bantu_Tanzanians/links/00b7d528f4bca72c20000000.pdf
Daily energy expenditure and cardiovascular risk in Masai, rural and urban Bantu Tanzanians
Hadn’t read that yet! Thanks for sharing.
It’s interesting that they expend so much more energy… but then I again, I guess it wouldn’t be too hard to expend 2500 kcal more than us Americans who are driving all over the place and working in an environment where we’re sitting on our asses all day.
Not sure I agree with the initial supposition that “There is strong evidence that the high consumption of animal fats increases the risk of developing cardiovascular disease”