The answer to that foolish question is certainly going to be “both.” But the finger pointing at the guilty party has wobbled around a good deal in the past few years, and just before Christmas, the American Heart Association came out with a set of dietary recommendations once again apparently putting at least some blame for heart disease on dietary cholesterol. This is a pretty big switch. In the past decade, the recommendations have bounced around a good bit; for example, as late as the 2018 cholesterol management guidelines from the American College of Cardiology and the American Heart Association shifted the emphasis on cholesterol management away from dietary cholesterol.
What, we may ask, is going on? Here’s how the American Heart Association science advisory puts it:
“Historically, nutrition guidelines for reducing cardiovascular disease (CVD) risk and achieving optimal plasma lipoprotein profiles have included recommendations to limit dietary cholesterol. However, contemporary guidelines for CVD risk reduction from the American Heart Association (AHA) and American College of Cardiology (ACC) and the “2015–2020 Dietary Guidelines for Americans” have not issued explicit guidance for dietary cholesterol. Because of the inconsistencies in the evidence base and the inherent difficulty in conducting and interpreting studies to isolate the independent effect of dietary cholesterol on CVD risk, controversy has ensued about whether dietary cholesterol should be a target for CVD prevention and management. (Carson et al “Dietary Cholesterol and Cardiovascular Risk” Circulation. 2019;140:00–00)
From where Doc Gumshoe sits, the AHA is being more than a bit weasely. To answer my previous question, what has been going on is that for several decades, the official positions of the major medical associations emphatically advised limiting the consumption of foods that were high in cholesterol, which specifically included red meat, eggs, and full-fat dairy products. But in the past five years or so, they have effectively backed away from those recommendations, without explaining why they did so. And then, in this past year, the AHA perhaps realized that they may have backed away a bit too far.
What do we know about cholesterol that got us to this point?
More than a century ago, it was determined firmly and clearly, and without “inconsistencies in the evidence base,” that cholesterol was the substance deposited in the walls of human arteries and was the main component in arterial plaque.
It was understood that cholesterol is a simple molecule, solid at body temperature, and not water soluble. It is present in virtually all our tissues, providing structure, and is absolutely essential for life. But, since cholesterol is not water soluble, in order to be transported in our bloodstream, it has to hitch a ride with substances that can be carried around in blood. These are lipoproteins – little particles containing both proteins and lipids, not in any fixed chemical combination, but bundles of varying sizes. Cholesterol attaches to the lipid part of these bundles, and the protein part permits them to be transported in blood.
These particles range in size and density. The low-density, loosely packed bundles, called low-density lipoprotein cholesterol, or LDL-cholesterol, are the ones nicknamed “bad” cholesterol, because they are the ones that are apt to shed the cholesterol molecules themselves, which can attach to the walls of the arteries and even penetrate those walls. But, we have to remember, the LDL-C is absolutely essential to our lives, because those particles are the ones that convey cholesterol to where it’s needed, which is pretty nearly everywhere in our bodies.
The smaller, denser bundles, called high-density lipoprotein cholesterol, or HDL-cholesterol, carries cholesterol back to the liver, where it is taken up by the bile and carried in the bile duct to the colon for excretion in the feces. Therefore, HDL-C is dubbed “good cholesterol.”
It’s worth just mentioning that only about 15% to 20% of the total cholesterol in our bodies enters our digestive systems as cholesterol. The rest we synthesize daily. Some of the foods we eat are more easily transformed into cholesterol, such as solid fats, especially partially hydrogenated fats (transfats). But no matter what we eat or don’t eat, we’ll go on synthesizing cholesterol, because we need it.
"reveal" emails? If not,
just click here...
The balance between LDL-C and HDL-C is a homeostatic mechanism. Those two types of particles have properties beyond their role as cholesterol delivery systems. LDL-C appears to be highly susceptible to oxidative damage, and may also inhibit the synthesis of nitric oxide, which is thought to be one of the body’s natural mechanisms to combat atherosclerosis. And HDL-C has the opposite effect. Its principal lipoprotein, apo A-1, plays a part in preventing particles involved in atherosclerosis from adhering to arterial walls.
So, not only do LDL-C and HDL-C play different roles in cholesterol transport, they do the same with regard to arterial damage. We could say that the nicknames “bad” and “good” cholesterol are fully justified.
Maintaining good levels of HDL-C in the circulation is highly important to cardiac health. Some activities, such as exercise, have a beneficial effect on HDL-C levels. >It’s the view of many cardiologists that the ratio of total cholesterol to HDL-C is more indicative of overall cardiac risk than the total cholesterol value by itself, so, while a TC value of more than 200 is thought to be higher than the optimal range, this can be offset by an HDL-C value higher than 50.
However, despite the necessity of cholesterol in our bodies, for most of the 20th century it was taken for granted that cholesterol was the cause of atherosclerosis and a principal cause of heart attacks (myocardial infarctions, or MIs).
Efforts to reduce the risk of MIs by cutting back on foods rich in cholesterol failed to lead to reliably clinically effective results. In other words, people who significantly reduced their intake of eggs, butter, and beefsteak, did not have fewer MIs – at least, statistically.
It wasn’t until the mid-1990s that a drug, simvastatin, was clearly and definitively demonstrated both to lower cholesterol in the blood and to reduce the incidence of heart attacks. Initially, this beneficial effect was demonstrated only in individuals thought already to be at an elevated risk for heart attacks.
Statins target the formation of cholesterol, termed cholesterol biosynthesis, which takes place mostly in the liver. They inhibit an enzyme labeled HMG-CoA reductase, which is active in cholesterol biosynthesis. However, statins have no effect on cholesterol that enters our bodies as cholesterol – only on the process of cholesterol biosynthesis.
In the years since the results of the trial which demonstrated the effectiveness of simvastatin, several other statins have been developed, approved, and have become widely used. One, atorvastatin, trade named Lipitor, became the world’s all-time best-selling drug, racking up $125 billion in sales in the years before it became a generic.
If heart disease rates had remained at their peak in the years immediately before the widespread use of statins, there would have been about 10 million more deaths attributable to heart disease since then. However, heart disease is still the leading cause of death, both in the US and worldwide. According to the World Health Organization, more than 30% of all deaths on Planet Earth are due to some form of heart disease.
The large decline in the number of heart disease deaths in the US and other developed parts of the world can hardly be attributed to the use of statins alone. Other factors – improved options for the control of hypertension, a reduction in the numbers of tobacco smokers – certainly played a part, but the evidence for the role of statins in that decline is very strong.
Looking for answers to some puzzling questions
While acknowledging the role of elevated cholesterol in cardiac disease, there were still a number of questions to be answered. One of these was what caused MIs in individuals with supposedly “normal” cholesterol. A common factor, unearthed by Paul Ridker, a cardiologist at Brigham and Women’s Hospital and the Harvard Medical School, was that many of these individuals had in their bloodstream high levels of a substance known as C-reactive protein (CRP), which was closely linked with inflammation. Ridker had long suspected that inflammation played at least a principal part in the pathology of serious cardiac events.
A closer look at what actually happens
Paul Ridker followed up his discovery about CRP with a study in which it was shown that treatment with statins not only lowered cholesterol levels, but also lowered levels of this inflammation marker. And in 2008, Ridker presented the results of the JUPITER trial at the New Orleans meeting of the American Heart Association. (Ridker P et al. New Engl J Med 2008;359:2195-2207)
This large trial (17,802 subjects) compared two cohorts of persons, all of whom had normal cholesterol levels. One group of 8,901 subjects received 20 mg. of rosuvastatin daily, and the other, also 8,901, got the placebo. The primary endpoint was incidence of cardiac events consisting of nonfatal myocardial infarction, nonfatal stroke, unstable angina, or death from cardiovascular causes. Subjects receiving rosuvastatin experienced 142 such events, while those on placebo experienced 251 events. Although the reduction was small in terms of absolute risk – about 1.2% ― it was considered highly significant, both statistically and in terms of implications for treatment. As a result of these results, the trial was stopped after a bit less than two years because the sponsors considered it unethical to continue a large cohort of patients on placebo when significant benefit had been demonstrated in the treatment arm.
The subjects in the JUPITER trial had baseline LDL-cholesterol levels of 108 mg/dL and CRP levels of 4.2/4.3 mg/L. Those LDL-C levels are considered reasonably good in patients with no established cardiac risk factors. However, CRP levels greater than 4.0 mg/L are now considered elevated and associated with significant risk.
The JUPITER trial cannot be said definitely to demonstrate that lowering CRP was the determining factor in reducing the numbers of signal cardiac events. Treatment with rosuvastatin not only reduced CRP from the baseline level to about 1.8 mg/L, but also lowered the LDL-C levels from a pretreatment 108 mg/dL to 55 mg/dL, so the benefit may have in part been due to the LDL-C reduction. But the reduction in that marker of inflammation was certainly an eye-opener.
In further analysis of the JUPITER trial results, Ridker came to the conclusion that elevated CRP levels signal more heart disease risk than do elevated LDL-C levels, although the highest risk is in patients in whom both of those are elevated. CRP levels below 1 mg/L are related to low risk, between 1 and 3 mg/L to medium risk, and higher than 3 mg/L to higher risk. The chart below traces the relationship between LDL-C and CRP levels and cardiac risk over an 8-year period.
Here’s another way of looking at the way cholesterol and inflammation (as shown by CRP levels) have a causal relationship with cardiac events.
According to the data behind that chart, a person could have a relatively high level of LDL cholesterol, such as greater than 160 mg/dL, and not be affected by a high level of cardiac risk. A person with elevated levels of CRP, indicating systemic inflammation, would be likely to have a higher level of cardiac risk even if his/her LDL cholesterol level was moderate, lower than 130 mg/dL. But if both the LDL cholesterol and the CRP levels were in the highest range, the cardiac risk would be apt to be very high indeed.
The LDL-C levels considered as low in these graphics are based on the guidelines promulgated by the National Cholesterol Educational Program (NCEP) Adult Treatment Panel III in 2001, which used 130 mg/dL as the cutpoint for borderline high LDL-C. More recently, many clinicians have argued for a target LDL-C reading of about 100 mg/dL in individuals with no specific cardiac risk factors, and below 70 mg/dL in persons with one or more risk factors. Thus, the data above could be criticized as minimizing the risk presented by elevated levels of LDL-C.
Although the data as presented strongly suggests that inflammation is a greater risk factor for heart disease than elevated cholesterol, it is far from presenting satisfactory vindication for the supporters of the view that inflammation is the whole story and cholesterol is a mere ruse cooked up by the pharmaceutical industry as a way to push their products.
Interaction between cholesterol and inflammation
We can accept the data showing the link between cholesterol and inflammation, on the one hand, and serious cardiac events such as heart attacks and strokes on the other hand. The statistical link is evident. But how exactly does inflammation cause these cardiac events? And why is it that the cholesterol that lodges in our arteries becomes the blood clots that cause strokes and myocardial infarcts?
At about the same time that Paul Ridker was doing his preliminary investigations , another Brigham and Women’s/Harvard cardiologist, Peter Libby, learned that cholesterol didn’t just swim around in the bloodstream. It actually worked its way into the arterial wall. This appeared to constitute a kind of insult to the arterial wall and provoked an inflammatory response, which in turn resulted in the formation of blood clots. It was these blood clots that, at least in some cases, blocked coronary arteries, causing heart attacks, and also blocked cerebral arteries, causing strokes. Peter Libby coined the term “vulnerable plaque” for plaque affected by inflammation that was prone to clot formation.
Libby demonstrated that the interior walls of blood vessels are made from smooth muscle cells which are lined with the endothelial cells that are in direct contact with the circulating blood. These endothelial cells act as sentries. If they detect the presence of anything other than blood cells in the blood stream, they summon white blood cells, which are the immune system’s front-line guardians. Naturally occurring adhesion molecules could attract the white blood cells and get them to stick to the endothelium lining the arteries. This action triggered an inflammatory response in the arteries, and the release of a cytokine called interleukin-1 (IL-1). IL-1 had been discovered in the late 1970s, and which had been found to cause fever in laboratory animals. IL-1, classified as a pyrogen (a fever-causing agent) is considered fundamental in the inflammatory process. Libby found that IL-1 could be produced in the lining of arteries, and that it actually amplified its signal at that site.
The actual physiologic mechanism that caused these cardiac events became better understood because of the work of Libby and Ridker. It has been very reliably confirmed that the substance in arterial plaque was cholesterol, and that blood clots containing particles of this plaque, were the material causes of MIs, strokes, and other signal cardiovascular events. What had not been completely understood prior to the research conducted by Libby and Ridker, their colleagues and numerous other medical researchers, was what happened in between cholesterol synthesis and arterial transport of cholesterol by lipoproteins, and then, at the other end of this pathophysiology, when cholesterol was absorbed into the arteries and then released in the form of potentially lethal clots.
We could state their conclusion this way: cholesterol is the missile, and inflammation is the fuel that propels it.
Ridker and Libby, along with many others, confirmed this in a large trial conducted in 39 countries and involving 10,061 subjects, all of whom had previously sustained a myocardial infarction and had a CRP level in excess of 2 mg/L, which is well above the normal range but not at a particularly high level. The purpose of the trial, named CANTOS, was to confirm the hypothesis that reducing inflammation without affecting lipid levels would result in lowering the risk of cardiovascular disease. (Ridker PM et al, N Engl J Med 2017;377(12):1119-1131)
The active agent being used in this trial was canakinunab, a therapeutic monoclonal antibody targeting interleukin 1β, one of two forms of IL-1, the fever-causing cytokine whose release was triggered by the attachment of white blood cells to the surface of arteries. Canakinunab, whose trade name is Ilaris (Novartis), is approved for a range of uncommon autoimmune diseases including cryopyrin-associated periodic syndromes (CAPS), active systemic juvenile arthritis, and others. These diseases are inherited and frequently affect children as young as one year of age.
The CANTOS trial compared three doses of canakinumab (50 mg, 150 mg, and 300 mg, given subcutaneously every three months) with placebo. The primary efficacy endpoint was nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death.
At each of the three dosage levels, subjects in the canakinumab group sustained fewer cardiovascular endpoint events than those in the placebo group. At 3.7 years into the trial, the incidence rate in the placebo group was 4.5 cardiovascular events per 100 patient-years, while those in the 50 mg, 150 mg, and 300 mg canakinumab cohorts were 4.11, 3.86, and 3.90 per 100 patient-years. Only the differences in the 300 mg group were statistically significant for the primary and secondary end-points compared with placebo. The incidence of all-cause mortality was 15% lower in canakinumab-treated subjects than in the placebo group.
While the reductions in risk were not clinically meaningful, the CANTOS trial convincingly demonstrated that targeting the IL-1β pathway could lead to a significantly lower rate of recurrent cardiovascular events, independent of any therapy directed to cholesterol lowering.
The CANTOS research team went on to examine the effects that an anti-inflammatory strategy might have on diseases unrelated to cardiac disorders. Reports of gout and arthritis fell significantly, and lung cancer incidence fell by 77%.
The CANTOS trial did not seek to validate canakinumab as a form of therapy for the prevention of cardiovascular events. For one thing, the drug is exceedingly expensive. A single subcutaneous dose of canakinumab costs about $16,700. As a treatment for gout, for example, these would be required every two weeks. The rate of side effects with canakinumab was slightly higher than in the placebo group; neutropenia and thrombocytopenia being the most common.
What the CANTOS trial – along with other investigations of the possible role of inflammation in what are generally considered to be unrelated diseases – does is give an impetus to investigations of the ways in which inflammation may contribute to apparently unrelated diseases in virtually any part of our bodies. Evidence has gradually been accumulating suggesting possible involvement of inflammation in such conditions as Alzheimer’s Disease, cancer, asthma, gout, psoriasis, anemia, Parkinson’s disease, multiple sclerosis, diabetes, and depression. But to go from the suggestion of possible involvement to clear demonstration of a causal relation will require extensive – and expensive! – clinical trials.
… so, going back to our point of departure …
What does all of this say about dietary cholesterol, which is cholesterol that has already been formed before we bring it into our digestive system – cholesterol in meat, eggs, dairy products, and even some grains?
I can assert with considerable confidence that what the word “cholesterol” means to a huge number of people – perhaps even to some of you in the Gumshoe community – is the bad stuff in meat, eggs, etc. And therefore, since in the past few years, if we’ve paid much attention to the proclamations of many medical and health organizations, we may have come around to the position that cholesterol isn’t the problem, because those organizations have eased back on their dietary recommendations. They have instead doubled down on more generalized “healthy lifestyle” recommendations – the Mediterranean diet and DASH (Dietary Approaches to Stop Hypertension)-style diets. Obesity – rightly! – is the bugaboo. The 2013 AHA/ACC Guideline on Lifestyle Modification to Reduce Cardiovascular Risk” did not include a recommendation for dietary cholesterol and concluded that “There is insufficient evidence to determine whether lowering dietary cholesterol reduces LDL-C (low-density lipoprotein cholesterol.” The Dietary Guidelines Advisory Committee (part of NIH) stated in its 2015 recommendations that they would not bring forward their previous recommendation to limit dietary cholesterol “because available evidence shows no appreciable relationship between consumption of dietary cholesterol and serum cholesterol.”
So what accounts for the big switch, with the AHA’s science advisory making these statements:
“Given the relatively high content of cholesterol in egg yolks, it remains advisable to limit intake to current levels. Healthy individuals can include up to a whole egg or equivalent daily. A 3-oz serving of shrimp is equivalent to about a whole egg. Shrimp and other shellfish can be incorporated into a heart-healthy dietary pattern when paired with other lean or plant-based protein sources. Caveats exist for the following subgroups:
- Vegetarians (lacto-ovo) who do not consume meat-based cholesterol-containing foods may include more dairy and eggs in their diets within the context of moderation discussed herein.
- Patients with dyslipidemia, particularly those with diabetes mellitus or at risk for heart failure, should be cautious in consuming foods rich in cholesterol.
- For older normocholesterolemic patients, given the nutritional benefits and convenience of eggs, consumption of up to 2 eggs per day is acceptable within the context of a heart-healthy dietary pattern.”
That strikes Doc Gumshoe as a considerable restriction on dietary cholesterol. A three-ounce serving of shrimp or one whole egg for healthy individuals? For older persons (why call them patients if their cholesterol is normal?) two eggs a day is okay as long as they stick to a heart-healthy diet.
I am prepared to venture a guess. My conjecture is that the medical community has been observing, with a mixture of satisfaction and disapproval, that a very large number of people are indeed taking statins, and that by and large, the statin regime controls their cholesterol levels fairly well. But – here comes the disapproval part – once having brought their total and LDL cholesterol down to acceptable levels, these individuals aren’t paying a whole lot of attention to their diets. Yes, it has been impossible to establish a link between dietary cholesterol in isolation and cardiac problems. But the link between total cholesterol – or specifically the low-density to high-density lipoprotein cholesterol ratio – and heart disease has been definitively established. And there’s no doubt that dietary cholesterol contributes to the total cholesterol burden – perhaps not more than 15% or 20%, but those modest fractions could raise the cholesterol level past the safety point, at least in some individuals.
It may be that a return to somewhat stricter recommended limits on dietary cholesterol are a tribute to the success of statins.
* * * * * * *
Speaking now from a personal perspective, as your faithful Doc Gumshoe, what strikes me about this business of guidelines, whether from the American College of Cardiology, the American Heart Association, the Dietary Guidelines Advisory Committee, or any other august body, is that the very act of formulating guidelines is fraught with uncertainty. As I read the guidelines, I get the definite impression that they are skating on thin ice, and they know it. Their guidelines are full of caveats, as they need to be. The AHA acknowledges, albeit in microscopic print, that their so-called “data” about what people actually eat comes from either food-frequency questionnaires, or 24-hour dietary recalls. In other words, not controlled, and not even observational, but relying on the memory and reliability in their subjects. And in arriving at their dietary guidelines, they categorize the entire population into four bunches – healthy individuals, vegetarians, people with dyslipidemia, and older people. I can’t imagine how many categories it would take to cobble together recommendations that would be right for the entire population, but those four aren’t nearly enough.
My guess is that most health-care professionals will pay only a tiny bit of attention to those AHA recommendations.
Thanks for all the comments – keep them coming! Best to all, Michael Jorrin (aka Doc Gumshoe)
Ed. Note: Michael Jorrin is not a doctor, though I dubbed him “Doc Gumshoe” many years ago. He is a longtime medical writer who shares his thoughts with our readers a couple times a month. His articles generally do not focus on investment ideas, but he has agreed to our trading restrictions. You can see all of his past commentaries here.