"Die wahrheit triumphiert nie, ihre gegner sterben nur aus." — Max Planck
(Truth never triumphs, its opponents just die out.)
In Part II of this series I wrote that this installment would be about the "best research" those supporting the saturated fat-cholesterol-heart disease link or "causal chain" might use (here’s Part I). In the process of doing my homework for that post I came across something I think fits in here because I essentially want you to have some idea of what you’ll be up against.
This supplemental entry began when I hit the blog of Dr. A (a reader & commenter here) in the UK: Livable Low Carb. I’ll show you a couple of the graphs she posted in a moment, but for now, how about a quote from the British Heart Foundation website on Diet?
It is now universally recognised that a diet which is high in fat, particularly saturated fat, sodium and sugar and which is low in complex carbohydrates, fruit and vegetables increases the risk of chronic diseases – particularly cardiovascular disease (CVD) and cancer. These risks are outlined in the World Health Organization 2003 report Diet, nutrition and the prevention of chronic diseases. The more recent World Health Organization Global strategy on diet, physical activity and health emphasised further the need to improve diets in individuals and populations across the world.
The dietary changes which would help to reduce rates of coronary heart disease (CHD) in the UK population were detailed in the 1994 report of the Government’s Committee on the Medical Aspects of Food and Nutrition Policy (COMA). This recommended a reduction in fat intake, particularly saturated fat intake, a reduction in sodium intake and an increase in fruit and vegetable and complex carbohydrate intake. In the 2003 report Salt and Health, the Scientific Advisory Committee on Nutrition (SACN) (which replaced COMA in 2000) repeated COMA’s guidance on salt intake in adults and introduced additional guidance on reducing salt intake in children. In 2005 the Government dietary objectives were reiterated in Choosing a Better Diet: a food and health action plan.
Research from the World Health Organization and others highlight the specific importance of low fruit and vegetable consumption as a cause of CHD. The World Health Report 2002 estimated that around 4% of all disease burden in developed countries was caused by low fruit and vegetable consumption, and that just under 30% of CHD and almost 20% of stroke in developed countries was due to fruit and vegetable consumption levels below 600g/day. The World Health Organization has yet to calculate the precise proportion of the disease burden due to high sodium intake or high saturated fat intake. [emphasis added]
I considered clipping some of that, but it would only serve to dilute what is clearly intended to be a VERY STRONG MESSAGE: reduce fat, especially saturated fat and eat more fruits & vegetables. But why? This is something I’ll get into in much more depth in the final entry, but here’s just a tidbit. When you really start digging into CVD stats it becomes clear that CVD deaths and CVD incidence are constantly conflated or at best conveniently left ambiguous. If a researcher wants to tell you that reducing saturated fat has been successful, he’ll haul out CVD death stats and indeed, there has been a tremendous decrease in death from CVD over the past decades. But they’ll rarely tell you in the same breath that the incidence of cardiovascular disease — i.e., heart attacks and other CVD markers — is pretty much as high as ever. So, is the reduction of death more likely a function of decreased saturated fat intake or more likely some other factor like improved urgent and emergency care in the face of a heart attack? As usual, a review of Occam’s Razor seems appropriate.
So, now, here’s a couple of Dr. A’s graphs for saturated fat consumption and fruit & vegetable consumption trends since 1975.
Oh, did I mention? This data comes from the very same website as the quote, above, the British Heart Foundation. You know what? I’m fairly confident that I could take any kid of normal mental function, explain the parameters including the distinction between CVD incidence and death, ask him or her about association and cause, and get the obvious conclusion: time to look elsewhere.
But kids are naturally honest about such things. For them, it’ll be another 20 years before they are "fortunate" enough to get paid for lying and manipulating, which brings us to the second case study in cognitive dissonance. Let’s begin with the names, ’cause I always name names.
These names, of course, are prominently displayed in large bold font right after the title of the study: Lipid levels in patients hospitalized with coronary artery disease: An analysis of 136,905 hospitalizations in Get With The Guidelines (PDF)
Amit Sachdeva, MD, Christopher P. Cannon, MD, Prakash C. Deedwania, MD, Kenneth A. LaBresh, MD, Sidney C. Smith, Jr, MD, David Dai, MS, Adrian Hernandez, MD, and Gregg C. Fonarow, MD.
Far less prominent, requiring a magnifying glass, are the financial disclosures.
- Christopher P. Cannon, MD; grants: Accumetrics, AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Merck, Sanofi- Aventis, Schering Plough
- Prakash C. Deedwania, MD; consultant of AstraZeneca and Pfizer
- David Dai, PhD; employee of Duke Clinical Research Institute
- Gregg C. Fonarow, MD; research from Pfizer and GlaxoSmithKline; consultant and honorarium from Abbott, AstraZeneca, GlaxoSmithKline, Merck, Pfizer, and Schering Plough; and chair of the Get With the Guidelines Steering Committee
No potential conflicts of interest there, eh? I guess that "disclosure" is the name of the game, now, and not the simple fact of the matter. Here’s a novel idea: how about simply publishing and paying attention to studies where financial disclosures are unnecessary? Yea, yea, it’s impractical and I also don’t want to unfairly paint with too broad a brush. Moreover, I’m no Luddite when it comes to the great contributions over time of doctors, researchers, and yes: drug companies. But isn’t this getting out of hand?
This was a captivating study to read for the cognitive dissonance, the subject of this entry. Remarkable to me is that I can detect no data manipulation or even evasion of data and facts (until the conclusion, which is an evasion to take your breath away). That’s what’s so interesting. It’s all there. Everything is there to make the same childlike conclusion as we did up above: time to look elsewhere.
Here, let’s take a look.
Now that’s almost as perfect of a Bell Curve Distribution as you’re ever going to see — and I’d guess that if LDL levels actually ran to zero in people that you’d see an absolutely perfect distribution. I remind you: these are the LDL levels of 137,000 people admitted to a hospital with coronary artery disease. Let’s imagine a scenario, for example, measuring blood alcohol content in people involved in automobile accidents. Would that be a Bell Curve, or, would you naturally expect that the higher the BAC, the more accidents (an ever increasing curve)? Now there you have a definite causal relationship and the data is going to show it unambiguously.
Now let’s look at some quotes from the study (PDF).
- Half the patients hospitalized with CAD had admission LDL <100 mg/dL, and LDL <70 mg/dL was observed in 17.6% of patients.
- Less than one quarter of patients had an admission LDL >130 mg/dL.
- There were 54.6% of patients hospitalized with CAD with admission HDL levels <40 mg/dL.
- HDL ≥60 was observed in just 7.8% of patients.
- Ideal levels (LDL <70 mg/dL and HDL ≥60 mg/dL) were observed in only 1.4% of patients hospitalized with CAD.
- Among the 21.1% of patients receiving lipid-lowering medications before admission, LDL levels were modestly lower (94.3 ± 36.4 mg/dL) and HDL levels were similar to those not previously treated with lipid-lowering medications (39.6 ± 2.6 mg/dL).
- Although high serum concentrations of LDL are a major risk factor for CHD, patients may present with CAD events despite LDL levels, which are not considered elevated and fall well within guideline-recommended targets. In the present study, almost half of patients hospitalized with CAD have admission LDL <100 mg/dL, and 17.6% of patients had LDL <70 mg/dL. Even when only patients without prior history of CHD, other atherosclerotic vascular disease, or diabetes were studied, 72.1% have admission LDL <130 mg/dL and 41.5% had LDL <100 mg/dL. Thus, a substantial proportion of patients present with their first or recurrent CHD events well within the current guideline-recommended targets for LDL.
- Many of these patients presenting with CAD had HDL levels, which are associated with excess risk. High-density lipoprotein cholesterol is inversely related to the risk of CAD. Even modest increases are associated with lower risk for nonfatal MI or death from CHD. There were 54.6% of patients hospitalized with CAD with admission HDL <40 mg/dL. In addition, fewer than 10% of patients had HDL ≥60 mg/dL. Ideal levels (defined as LDL <70 mg/dL and HDL ≥60 mg/dL) were present in only 1.4% of patients hospitalized with CAD.
How about let’s put this final quote in bold.
The present study demonstrates that among patients hospitalized with CAD, the admission lipid levels are below that of the general population.
And, so, what’s the OBVIOUS conclusion from all these data points, observation and reasoning?
In a large cohort of patients hospitalized with CAD, almost half have admission LDL <100 mg/dL, whereas less than a quarter have LDL >130 mg/dL. The LDL levels <70 mg/dL are observed in only 17.6% of patients. Admission HDL levels are <40 mg/dL in 54.6% of patients hospitalized with CAD, whereas <10% of patients have admission HDL levels ≥60 mg/dL. Ideal lipid levels (LDL <70 mg/dL with HDL ≥60 mg/dL) are seen in only 1.4% of patients hospitalized with CAD. There were reductions in admission LDL and HDL levels over time. These findings provide insights into the lipid levels encountered in recent clinical practice for patients hospitalized with CAD. These findings may provide further support for recent guideline revisions with even lower LDL goals. They also may suggest a clinical need for developing effective treatments to raise antiatherogenic HDL. [emphasis added]
Shorter version: lower the cholesterol guidelines even more (and, hey, why not throw in a few billions of dollars more of statin Rx?), because heart attack victims already have average cholesterol levels less than the general population. OBVIOUSLY; more lower, please. And, while we’re at it, let’s come up with a few billion dollars market for meds to raise HDL.
Hey, maybe they could hire me to consult on that last point, since my HDL runs in the 130s and directly measured LDL in the 60s. Of course, that requires no meds, but only a diet of real foods high in natural, healthful saturated fat, plenty of animal protein, some veggies, fruits & nuts — and occasional bouts with full fat, raw dairy.
So, no chance for me to cash in, I guess. I’ll have to content myself with naming names of those who put their livelihoods above your health and are willing to engage in head-spinning cognitive dissonance to hold their lucrative positions of authority.