Both of the studies referenced were meta-analsyses. These are studies that pool together the individual data points from many other individual studies. The drawback is that often times the individual studies will have used different methods or definitions or some will have been of poorer quality than others. Both of these meta-analyses, however, appear to be well done and to have used sound criteria for inclusion. The benefits are that they give a sense of the totality of the evidence and that they are statistically much more powerful because they have such a dramatically larger sample size.
The Importance of the Total-to-HDL Cholesterol Ratio
The first study was a meta-analysis of 61 prospective studies published in 2007 that included almost 900,000 people followed for an average of over 13 years and over 55,000 deaths from cardiovascular diseases. It found that total cholesterol was the worst predictor of heart disease mortality risk while the total-to-HDL cholesterol ratio, which is essentially the same thing as the LDL-to-HDL cholesterol ratio since most non-HDL cholesterol is LDL cholesterol, was the best predictor.
In both men and women, a 1 mmol/L (just under 39 mg/dL) lower ratio was associated with half the risk of heart disease mortality in those aged 40-49 years, two thirds the risk in those aged 50-69 years, and five sixths the risk in those aged 70-89 years.
The first thing we can observe is that while the importance of the ratio diminishes as people grow older, it is still statistically significant in a massive meta-analysis. This shows that when studies find blood lipid levels are not risk factors in old age, they are confusing statistical significance with clinical significance — the ratio is meaningful, but the smaller magnitude just isn't detected in studies with lower statistical power.
Why does the risk decline with age? First, let's consider what the total-to-HDL cholesterol ratio even means. The fact that it is associated with heart disease risk does not mean it causes heart disease risk. But it could.
As described in my article, Cholesterol and Heart Disease: Myth or Truth?, the initiating factor in atherosclerosis is the oxidation (or glycation) of LDL particles in the blood. In the later stages, oxidized LDL also contributes to the inflammatory action of the foam cells it finds itself stuffed in, but these cells also recruit T cells that make their own inflammatory contribution independent of oxidized LDL, so its importance declines. Oxidized LDL contributes to plaque rupture, but so do other inflammatory factors as well as deficiencies of collagen production, which are probably influenced by vitamin C status. Thus, oxidized LDL is central to the initiation of the disease, but as the disease progresses the contribution of oxidized LDL is diminished.
Where does HDL fit in? Much has been made of its role in reverse cholesterol transport, but that has little to do with the oxidation of LDL or the oxidized derivatives of linoleic acid that have been shown to turn on the “foam cell” genes in the monocytes that first swallow them up. As described in the article I linked to above, when the contribution of oxidized LDL was first discovered in the late 1970s and early 1980s, researchers found that HDL or vitamin E both prevented the oxidation that would occur when LDL was exposed to the cells that line the blood vessels for a prolonged period of time. As I have pointed out on my site elsewhere, HDL is responsible for delivering vitamin E to these cells.
So that suggests a protective, causal effect of HDL, but not HDL cholesterol. And in fact, interventions to try to specifically boost HDL cholesterol have not been terribly successful. The most notorious case was torcetrapib, a drug that was designed to block the transfer of cholesterol from HDL particles to LDL particles. Not only was it toxic, but a recent trial concluded the following:
The absence of an inverse relationship between high-density lipoprotein cholesterol change and cIMT progression suggests that torcetrapib-induced high-density lipoprotein cholesterol increase does not mediate atheroprotection.
In other words, keeping the level of cholesterol in the HDL particles high does not reduce the progression of atherosclerosis. This suggests that while HDL protects against atherosclerosis by preventing the oxidation of LDL, HDL cholesterol does not protect against atherosclerosis by transporting cholesterol away from peripheral tissues and back to the liver. This is unsurprising, considering it is the oxidation of LDL and not the transport of cholesterol to peripheral tissues that contributes to atherosclerosis.
So what would the total-to-HDL cholesterol mean? The longer LDL stays in the blood, the more two things happen: it is exposed to oxidants, and as its limited supply of antioxidants run out, the polyunsaturated fatty acids in its membrane oxidize, leading to the further oxidation of its proteins and cholesterol; it is exposed to cholesterol ester transfer protein (CETP), which transfers cholesterol from HDL to LDL, thus boosting the total-to-HDL cholesterol ratio.
So the total-to-HDL cholesterol ratio should be a marker for the amount of time LDL particles spend in the blood. This, in turn, is dictated by the activity of the LDL receptor, which brings LDL into the liver and other tissues that need it. Since the liver only packages lipoproteins with a finite amount of antioxidants, it is critical that they reach cells, where antioxidant enzymes are regularly produced, quickly and efficiently. To whatever extent the total-to-HDL cholesterol ratio is high, this probably isn't happening.
How Do Dietary Fats Affect the Total-to-HDL Cholesterol Ratio?
The second study was a meta-analysis published in 2003 of sixty trials testing the effect of feeding different types of fats to humans on the total-to-HDL cholesterol ratio.
The study found that saturated fats did not change the ratio when substituted for carbohydrates. Carbohydrates, however, did raise triglyceride levels and shift LDL to the small, dense pattern associated with atherosclerosis when they were substituted for saturated fats.
Unsaturated fats, especially polyunsaturated fats, decreased the ratio. But so did specific saturated fats like stearic and lauric acids. “As a result,” the authors wrote, “lauric acid had a more favorable effect on total:HDL cholesterol than any other fatty acid, either saturated or unsaturated.”
They pointed out further that even highly saturated fats like dairy and tropical oils contain some unsaturated fat, so they will all decrease the ratio relative to carbohydrate. And since coconut oil is rich in lauric acid, it would be especially effective in reducing the ratio.
In contrast to all of these fats, trans fats raised the ratio.
What Does This Tell Us About Dietary Fat?
Before we conclude anything about what type of fat we should eat, we must remember that correlation does not prove causation. As I pointed out above, evidence suggests that the total-to-HDL cholesterol ratio is a marker for the time LDL spends in the blood rather than a causal factor itself. So we cannot conclude that PUFA oils and coconut oils are the best fats to prevent heart disease based on this meta-analysis by itself.
The latest edition of the widely respected textbook Modern Nutrition in Health and Disease states that linoleic acid (a PUFA found abundantly in vegetable oils) decreases cholesterol levels because the enzymes that store cholesterol by connecting it to fatty acids, called esterification, will selectively use linoleic acid. Thus, while liver cells get stuffed with cholesterol linoleate esters, their level of free cholesterol declines and they produce more LDL receptors on their surface, which bring LDL in from the blood.
Increased expression of the LDL receptor is good. But this meta-analysis considered blood lipids to reach a steady state by 13 days. Maybe cholesterol linoleate can accumulate in the liver for 13 days without adverse effects, but what happens over the long term when cholesterol esters progressively accumulate in that organ? Can that go on forever? Longer studies would be needed to find out.
It isn't quite clear how some saturated fats decrease the total-to-LDL cholesterol ratio. Perhaps they enhance LDL receptor function by decreasing oxidative stress, but there could be many possibilities.
As the authors of this meta-analysis pointed out, we should rely on controlled trials testing the effects of dietary fats on heart disease risk rather than extrapolating from surrogate markers. They cite several trials showing that unsaturated fats reduce heart disease risk compared to saturated fats.
These trials, however, were often poorly conducted or deceptively interpreted. Moreover, a number of trials showed the opposite, like the Rose, et al. (1965) trial that found corn oil to quadruple the risk of heart disease when substituted for butter over the course of two years. I have discussed those trials here.