Mea culpa. Almost three weeks ago I promised to do a little reading and report back on what to believe about cholesterol and heart disease. It was hubris to imagine that I would be able to untangle the thicket of conflicting claims with a short course of study. Today, my goals are far more modest, and I offer my scaled-back conclusions. Here, I offer a tentative analysis, which I hope will prompt people more knowledgable than I to refine and correct the message.
The standard medical paradigm says that eating saturated fats contributes to higher LDL cholesterol in the blood, and that higher LDL is associated with greater risk of CVD (cardiovascular disease = heart attack and stroke). The mystery which I identified last month is that there is strong evidence both for and against this hypothesis in the mainstream medical literature.
Why the disparity?
Certainly one reason there is so much difficulty obtaining agreement in the field is that the majority of scientists doing research on cholesterol are funded by drug companies that have billions riding on the results. This is not a question of “honest” science vs “corrupt” science. It is simply unreasonable to expect that a scientist working for a statin manufacturer feels free to pursue his study objectively, and to follow the data wherever it leads. It is hard to know the extent to which drug money has corrupted the science of cholesterol, but we can be assured that it is substantial.
Despite this major bias, I don’t think that the relationship between cholesterol and heart disease is entirely a mistake or a fraud. Even the harshest critics of the cholesterol establishment (Fred Kummerow, Stephanie Seneff, John Abramson, Sally Fallon) admit that there is some correlation between saturated fat in the diet and cholesterol in the bloodstream. Similarly, there is good evidence that risk of heart disease is correlated with LDL cholesterol. A new test based on the size of the LDL particles is much more predictive – see below.)
Correlation and Causation
If A is correlated with B and B is correlated with C, it does not follow that A must be correlated with C. Furthermore, even if A is correlated with C, it may be that A does not cause B, or B does not cause C, so that changing A may have no effect (or the wrong effect) on C.
To be concrete: It may be true that saturated fat in the diet is correlated with high LDL, and high LDL is correlated with high CV mortality, and still cutting saturated fat does not affect CV mortality.
Most studies linking saturated fats with LDL cholesterol have been positive [ref], and most studies linking LDL cholesterol with risk of CVD have been positive [ref]. But studies attempting to link saturated fats in the diet with a higher risk of CVD have mostly been negative [ref]. This suggests that we don’t understand the underlying mechanism behind heart disease very well, and that much of what we have done in the name of prevention may have been misdirected at intermediate targets that were not causally related to heart disease.
What is true of saturated fat reduction is even more true of statin drugs. It is certain that statin drugs improve the cholesterol profile. It is less clear that statin drugs save lives, and it is certainly not true that statin drugs are the best prevention strategy.
It is very convenient to have an intermediate measure (like LDL cholesterol) so that you can tell whether an intervention is working without having to wait for large numbers of people to die. If we look for the effect of some new treatment on LDL, we can test a small number of patients and know within a few months whether it is working; but if we are looking all the way to the bottom line (how many people taking this new treatment are dying of heart disease?) then the study must involve large numbers of subjects, followed over many years.
Almost all of what we thought we knew about saturated fats, cholesterol and CVD has been based on indirect inference. We may have to start over, using the more costly and rigorous approach based on counting CV events in large numbers of people.
In the 1960s, people were tested for total cholesterol in the blood, which was certainly the wrong indicator. So bad, in fact, that total cholesterol is inversely correlated with all-cause mortality risk. In recent decades, the standard test has been the ratio of LDL to HDL cholesterol. These are not two different kinds of cholesterol, but rather variations in the chemistry of the droplets that transport cholesterol in the bloodstream. Cholesterol is not water-soluble, so it is carried through the blood to all the places it is needed packaged in little droplets, These droplets come in low-density and high-density, large and small. “LDL” is “bad cholesterol” and it refers to the low-density droplets. “HDL” is “good cholesterol” and it refers to the high-density droplets.
Just in the last few years, there is a new blood test, which is a much better indicator of risk of heart disease and stroke than were LDL and HDL. It is a subset of the LDL droplets, the small ones, that are the basis of the new test. [Early article, 1997. large Scandinavian study, 2014.]
(I don’t want to say the small droplets are the culprit, or the cause of heart disease. It is still unclear whether relationship between the blood test and heart disease is a causal one. In other words, if we target therapies to reduce the scoe on the small particle test.)
You may want to make sure that small particle LDL is reported the next time you get a health check-up and blood test. It is sometimes reported as LDL III, or LDL-B. But the larger significance of this new, more accurate test is that it should make future research studies quicker and more efficient. Assuming that causality can be established, we will have, for the first time, a way of knowing whether a preventive therapy is working without having to wait ten years and see how many people have heart attacks.
It is probably true that statin drugs have saved millions of lives, if the alternative is defined as inaction in the face of heart risk. It is also true that statins are the most over-prescribed medications in the Western world, that their benefits are modest and come at a cost in side-effects that can be devastating in the long run. There are better ways to reduce cardiovascular risk, many of which have no side-effects, or side-effects that are wholly beneficial.
The way statins work is by impeding the synthesis of cholesterol. This is the wrong target. We need cholesterol. As I explained in my last post, cholesterol is what makes cell membranes in animals pliable. Cholesterol is also an intermediate product that the body uses to make essential hormones. Cholesterol is necessary for neurons in the brain. If we artificially suppress the body’s production of cholesterol, we are asking for trouble.
Statins also have a powerful anti-inflammatory effect, which may be the entire reason for their effectivenesss in lowering mortaliy. There are safer antiinflammatory agents available, including fish oil and aspirin, which seem to be the easiest route to lower cardiovascular risk.
Statin side-effects include:
- Increased blood sugar, leading to diabetes
- Reduced CoQ10, leading to muscle weakness and oxidative damage
- Reduced levels of other hormones, including sex hormones that are synthesized from cholesterol
- Muscle pains
- Low energy
- Cognitive impairment
- Parkinson’s Disease?
The first item is the most troubling. Statin drugs increase blood sugar, and increased blood sugar signals the body to age more rapidly. In the long run, this has got to be bad for heart disease risk, and also for all other causes of mortality. Trading heart disease for diabetes is not a good bargain. Blood sugar (particularly fructose) also combines with cholesterol to create the chemical species that is most damaging. Stephanie Seneffexplains:
Worse than that, once LDL particles have finally delivered their contents, they become “small dense LDL particles,” remnants that would ordinarily be returned to the liver to be broken down and recycled. But the attached sugars interfere with this process as well, so the task of breaking them down is assumed instead by macrophages in the artery wall and elsewhere in the body, through a unique scavenger operation. The macrophages are especially skilled to extract cholesterol from damaged LDL particles and insert it into HDL particles. Small dense LDL particles become trapped in the artery wall so that the macrophages can salvage and recycle their contents, and this is the basic source of atherosclerosis. HDL particles are the so-called “good cholesterol,” and the amount of cholesterol in HDL particles is the lipid metric with the strongest correlation with heart disease, where less cholesterol is associated with increased risk. So the macrophages in the plaque are actually performing a very useful role in increasing the amount of HDL cholesterol and reducing the amount of small dense LDL.
Perhaps the clearest message to come out of my reading is that there are better methods of lowering CV risk than statin drugs, and that people taking statin drugs should be looking for ways to replace them with measures that have beneficial side-effects. I divide these into two categories:
(1) those whose mechanism is like statins, and may lead to some of the same problems
- Red yeast rice is a natural product with much the same effect profile as statin drugs, but with milder side-effects. It may provide an alternative for people who experience muscle pain from statins , though it still interferes with CoQ10 and hormone production.
- Niacin=Vitamin B3 is frequently prescribed in high doses by heart doctors.
- Garlic works surprisingly well for a simple, common food, suppressing cholesterol production in the liver in a way comparable to statins.
(2) those that have a fundamentally different mechanism of action, and are likely to be safer
- Hibiscus flower helps prevent the oxidation of cholesterol in the blood to its toxic form. In tests on rabbits (more prone to heart disease than mice), hibiscus lowers the incidence of CVD.
- Berberine (a Chinese herb) signal activation of a gene that breaks down LDL (“bad”) cholesterol
- Resveratrol acts at a higher level, promoting the SIR-T genes that have many downstream effects: increasing pliability of hardened arteries, reducing blood clots that lead to heart attack, and also lowering LDL cholesterol
- Tree mushrooms used in Chinese soup (Auricularia) are quite effective in lowering LDL, and I have not found information about the mechanism. [ref, ref, ref]
- Fish oil (either from fish in the diet or from capsules) seems to reduce heart risk more consistently and more effectively than anything else, with side-effects that are almost all positive. It works on HDL:LDL ratio by a mechanism that is not well-understood, and it is also anti-inflammatory, which lowers risk of cancer and Alzheimer’s as fringe benefits.
Congenital high cholesterol
There are a number of genetic defects that can cause high LDL cholesterol. Some can be tested, but most cannot. The defects are relatively rare (1/500) but have a substantial effect on life expectancy [ref].
There is a particular mutation on chromosome 19 that is responsible for high levels of LDL cholesterol, and experts on both sides of the divide agree that it is associated with a significant decrease in life expectancy and should be treated. Traditional treatment is based on extra-high doses of statins. Alternative treatments are available that are less well-documented but show indications they may be more effective than statin treatment, with fewer long-term side-effects. I recommend this article from Life Extension Magazine.
Separating saturated fats from other factors in a meat-based diet
Vegetarians enjoy lower mortality rates. How much of this is due to lower saturated fats, how much to high fiber, how much to micronutrients from fruits and vegetables? And how much is an artifact from the association between vegetarianism and better self-care in other areas? In this same study , vegetarians who ate fish had 10% lower mortality (over 6 years) than vegetarians who ate cheese and eggs. Perhaps this is an indication that saturated fats are bad for you, and fish oil lowers inflammation as well as blood cholesterol. This study suggests that the difference may be more about benefits of fish than of hazzards of eating eggs and dairy.
A diet high in vegetables is our first-line defense.
(This is the last paradox, I promise!) I began by noting that the correlation between saturated fat in the diet and cardiovascular risk is weak or non-existant [ref]. But we know that saturated fat intake is highly-correlated with animal foods in the diet, and animal products lead to higher overall mortality. This suggests that it is something else about the plant-based diet, besides low sat-fat, that is giving the benefit.
Coming back to my daughter’s question
My daughter is 28 years old and a vegetarian. Last month she asked me, “I have high blood cholesterol. Should I be avoiding coconut oil and cocoa fat?”
From three weeks of research, my answer has to be
(E) It cannot be determined from the above information.
Cardiovascular risk in the context of other sound advice
Lots of exercise, weight control, low-carb diet, fish oil, daily aspirin all contribute to lowering risk of heart disase, and in the aggregate are far more powerful than statin drugs, with other health benefits as a free fringe benefit.
This post originally appreared on Josh’s blog here: http://joshmitteldorf.scienceblog.com/2014/06/04/cholesterol-part-ii/