Rethinking "Heart Healthy"

"Heart healthy" has become synonymous with "low fat".  The number one killer of Americans is variously called atherosclerosis, heart disease, cardiovascular disease, or coronary artery disease, but these refer to the same condition or its symptoms and complications: the accumulation of plaque in the arteries, causing a loss of elasticity (hardening) and eventually blockage.

Conventional wisdom holds that LDL cholesterol, by its very presence, causes a buildup in certain parts of the arterial walls, leading to this condition, and that the best way to prevent and treat it is with a low-fat diet, or cholesterol-lowering drugs.  If this is correct, then a low-carb, high-fat diet is dangerous indeed.  My aim here is to discredit that hypothesis and present an alternative.


The hypothesis itself was sound and worth pursuing.  It arose from the Framingham Study (thoroughly debunked in this book), the original canonical work on the matter, and the discovery that statin drugs, which lower LDL cholesterol levels, also tend to protect against heart disease.  The assumption is that these drugs achieve the latter by means of the former, which if correct means that heart disease is primarily caused by high levels of LDL cholesterol.  It turns out, however, that cholesterol levels are a poor predictor of heart disease, so we must question this assumption.

Early in the development of the hypothesis, it was believed that total cholesterol levels, including HDL, were the culprit.  As the theory evolved, LDL was implicated, and ultimately the ratio of HDL to LDL, with HDL earning a reputation as "good" cholesterol, and LDL as "bad".  LDL is, after all, the type of cholesterol responsible for arterial plaque deposits, and HDL is how the body packages cholesterol for destruction in the liver.  Thus, the current state of the art as I understand it is that LDL levels correspond to cholesterol on its way to wreak havok, such as in the arteries, and HDL is the reverse of the process.  This is seen not just as an indicator, but as an almost-root cause of heart disease, with the real root of the problem being dietary fat that raises LDL levels.

When you get your cholesterol checked at the doctor's office, the results are a reasonably complete "lipid profile", including HDL, LDL, VLDL and triglycerides, all of which have been shown in various studies as independent risk factors for heart disease.  The conclusion, again, is that one or more of these numbers, or a combination, is the cause of heart disease, and that any treatment regimen should target those numbers.  One thing they don't tell you is that the standard test doesn't even count LDL, what many doctors (including mine, quite vociferously) consider to be the most important number.  It's estimated based on total cholesterol and HDL, with the assumption (sometimes justified) that HDL molecules are about 5 times the molecular weight of LDL molecules.  This leads to the current technique for estimating heart disease risk from cholesterol: if your LDL to HDL ratio is less than 5:1, you must have more cholesterol leaving problem sites than being delivered to it, and your risk decreases.  If it's greater than 5:1, you have a problem.

All well and good, but as I mentioned (and linked) the evidence does not quite bear this out.  One reason for this may be the misleading numbers that result from the estimation of LDL.  LDL refers to a relative density of lipids in each molecule, and there's a large degree of variance within the classification of LDL.  If you are on a low-carb, high-fat diet, your lipid profile will likely show a "favorable" HDL figure, a high (unfavorable) LDL figure, very low triglycerides and very low VLDL.  This pattern has been shown to be a favorable lipid profile, as it suggests a lower concentration of what appear to be the harmful, higher-density subcategory of LDL molecules.  This in turn suggests a predominance of larger, less-dense, "fluffy" LDL molecules, which is unexpected and will not correlate with the 5:1 ratio assumed for the difference in molecular weight between LDL and HDL.  Thus your doctor will be unnecessarily alarmed, especially if he didn't graduate from medical school in the last 5-10 years.

We now know that heart disease is strongly correlated to insulin resistance, Syndrome X or metabolic syndrome, and type 2 diabetes, which I'm trying to show are all manifestations of the same root problem.  We're advised to consume a low-fat diet to reduce our risk of heart disease, but no one seems to have any success with that.  We seem to be perpetuating a dogma that can't be penetrated by science.

There is another hypothesis available, implicating chronic inflammation as the cause of heart disease. A number of recent studies have established the link from inflammation to one or more of these interrelated conditions.  Other studies suggest the link from insulin and/or insulin resistance to this inflammation, and even more directly to heart disease itself.  Insulin promotes the proliferation of smooth muscle cells in the arterial walls, and these cells produce inflammatory agents such as C-reactive protein.  From the article:

[I]t might be... that the insulin resistance itself [rather than the inslulin], by production of proinflammatory cytokines, induces atherogenesis and that the hyperinsulinaemia could be the body's compensatory attempts to suppress the inflammation and overcome insulin resistance... A vicious circle can also develop: decreased insulin sensitivity in organs such as skeletal muscle, adipose tissue, and liver leads to increased inflammatory activity, which in turn leads to further reductions of insulin sensitivity. Moreover, glucose has proinflammatory effects: it increases synthesis of reactive oxygen species and accentuates several inflammatory markers in vitro.

I'm going to be talking a lot about that last bit, how glucose itself promotes the production of "reactive oxygen species", and how all the movie stars keep talking about the health benefits of antioxidants, and how it all relates to the low-carb diet, in the near future.  Stay tuned!