under construction
Subpage of science-and-medicine
Christopher Labos on February 15, 2019
He starts out with a conclusion. He does acknowledge that there was controversy, but claims it is time to consider it closed. I see a problem in that introduction. First of all, do we know the etiology, the cause and course of arteriosclerosis? Is cholesterol a cause or an associated “risk factor”? What do we know and how do we know it? What is possible, what is probable, and how do we assess these? These are questions I have in mind as I go over the article. If some measure is only a risk factor, associated but not causative, altering the measure will not necessarily reduce the actual risk.
The photo caption:
Two bags of fresh frozen plasma. The bag on the left was obtained from a patient with hypercholesterolemia, and is cloudy with undissolved cholesterol particles.
Source. No information on cholesterol levels. Okay, but the significance? So blood with lots of cholesterol looks different than blood with little. So?
A recent article in The Guardian raised an interesting question. Is cholesterol denialism a valid form of skepticism or pseudoscience? Is there valid debate surrounding the benefit of cholesterol medication or is the evidence and the scientific consensus clearly on one side of the issue?
It is true that we argue about cholesterol far more than the other cardiovascular risk factors. It is hard today to find anyone who doubts the harmful effects of smoking, diabetes, hypertension or the lack of exercise. So why is there a cholesterol controversy but unanimity on other risk factors?
Okay, the Guardian article, our subpage.
I found value there, but only by searching for papers she referred to and related documents. The article itself was next to useless except as a great example of assuming the status quo is better than whatever is proposed to replace it. If lots of people criticize something, and if danger is asserted without evidence other than established belief, well, dangerous ideas should not be allowed to be published. I find it so ironic that advocates of evidence-based medicine, allegedly scientists, will declare criticism of what they believe “denialism,” when skepticism and criticism is essential to science even if later shown to be wrong. And who decides when later is? There a many who appear to believe that they represent the “consensus,” but they do not actually measure consensus. Signatures for the open letter were solicited on a blog, and there were
Bosely claimed “More than 170 academics signed a letter.” This shows what? The actual solicitation and signatures were not limited to academics, nor by field of study. There are currently 169 signatures, but if we include the original authors, it becomes 173. Looking at affiliations and counting those that do not show an academic affiliation, there are roughly 55, leaving 128. This was meaningless, in fact, given the population involved. Yes, it would show that 169 people agreed with the letter, but out of how many? Science is not a vote and votes are meaningless, unless conditions are set for it to truly represent a community. This was on the order of a petition requesting investigation of charges of bias.
The essence here is a conspiracy theory, that journals are publishing articles favoring low-carb diets and the like as a conspiracy to promote some crank ideas. Perhaps book authors are pimping fads to make money selling books. Boseley, however is also a book author, with her own advice. Perhaps she has a conflict of interest? Were it not for her implication that others are promoting dangerous ideas to sell books, I wouldn’t comment on it. But she is implying that, which is a huge insult to any academic, as many of the cholesterol skeptics are.
I have concluded that Boseley had an axe to grind, there are way too many signals of high bias.
Why is there a cholesterol controversy? It is very obvious why. What is controversial? He does not begin with a definition. Cholesterol is found inside arterial plaque. That is not controversial. What is controversial is whether or not cholesterol causes the plaque, and, further, how blood levels influence this process or exacerbate it, and, further back, whether or not dietary cholesterol leads to harmful blood cholesterol, or saturated fats, or all fats, depending on what point in history we go back to.
Very many of the original cholesterol hypotheses (i.e, there are more than one) have been disconfirmed by more careful study, but the attack on skepticism has remained constant, never recognizing that, at least in some ways, the skeptics were correct. For decades, Dr. Atkins “nutritional approach,” he called it (not “diet” it is actually not restrictive, but prescriptive, eat what he suggests and you may not crave the things he suggests be avoided), it was called a “fad diet,” though it was actually quite old and whether it worked or not did not depend on its age, he was called a quack, etc., etc….. But when I told the nurse at my doctor’s office that I was starting out on Atkins, she had one comment, “Oh, that works!” And then that the Atkins diet works is ascribed to many asserted causes that are not necessarily real for the diet as it is, and misinformation about Atkins abounds. It is not a high-protein diet. Atkins was correct, and eventually funded research to test his program against other common ones. Surprise! In spite of being high-fat, Atkins eaters improved cardiac risk factors. And then, of course, he was accused of influencing the outcome of the study, but studies funded by companies with billions of dollars at stake are just good science? He had chosen a skeptical professor to fund. Smart guy, rest his soul.
Labos goes into the history of other controversies, that we allegedly forget. He covers disagreement over the harm of tobacco, blood pressure, the discovery of cholesterol, and then has
One of the earliest researchers in cholesterol was Nikolay Anichkov who in 1913 reported that rabbits fed pure cholesterol dissolved in sunflower oil developed atherosclerotic lesions, whereas the control rabbits fed just sunflower oil did not. At the time, this research had little impact and its importance was only recognized in retrospect. As Daniel Steinberg states:
If the full significance of his findings had been appreciated at the time, we might have saved more than 30 years in the long struggle to settle the cholesterol controversy and [Anichkov] might have won a Nobel Prize. Instead, his findings were largely rejected or at least not followed up. Serious research on the role of cholesterol in human atherosclerosis did not really get under way until the 1940s.
And just what is the significance? Dietary cholesterol does not cause atherosclerosis. If his findings were “rejected,” that is tragic. Research findings should be respected, and problems only arise in interpretation.
Cholesterol is found in atherosclerotic deposits. That is not controversy, but is this cause or is it effect? And how does the development and progression of such deposits relate to diet and to blood cholesterol level?.
Laboratories that tried to reproduce Anichkov’s results using dogs or rats failed to show that a cholesterol rich diet caused atherosclerosis. This likely occurred because dogs and other carnivores handle cholesterol differently from rabbits and other herbivores This led many to dismiss Anichkov’s results on the grounds that rabbits were not a good a good model for human physiology and that his research was likely irrelevant to humans.
Which still holds as an objection. Rats often are close to human response. So maybe rats are reactive to the cholesterol they were given, or the taste stressed them so much that they developed the arterial lesions that lead to initiation of the processes that build up plaque.
The criticism leveled against his research was not entirely unfounded.
On the one hand, I’d like to congratulate him for admitting the obvious, but what is rather obvious to me is that he still thinks it was at least somewhat unfounded, he still thinks this is relevant to human atherosclerosis, he has an axe to grind. Otherwise, without that, he would have skipped over this irrelevancy.
We have seen countless times how animal research does not translate into humans and to accept the “lipid hypothesis” based purely on Anichkov’s work would have been premature.
To say the least.
It should have been an invitation for others to pursue this new line of inquiry.
“Should have.” By what standard? This is obvious to me: Labos believes the lipid hypothesis. That’s okay. But it means that he is not a neutral judge, unless he could truly and consciously set aside what he believes, to study and make sure he understands what he is criticizing. He would, if interested in science, be attempting to prove himself wrong, not right. But, no, he’s convinced he is right and is only going through this exercise to prove that it’s totally silly to believe anything other than what he believes about cholesterol. He is pseudoskeptical about cholesterol skepticism. But, again, the conversation can have value.
Eventually in the 1950s John Gofman would begin his research in lipoproteins and determine that there were different types of cholesterol. Today of course we acknowledge that low-density particles like LDL are atherogenic whereas high-density particles like HDL are not. Gofman demonstrated this in the 1956 Cooperative Study of Lipoproteins and Atherosclerosis although the distinction of LDL and HDL would only come later.
Notice how fact is mixed with conclusions. Is LDL “atherogenic,” or is it merely an associated risk factor, or, third possibility, it has some effect on some more powerful, more critical cause? And notice, the early cholesterol hypothesis did not discriminate between HDL and LDL, and even deeper distinctions are moving into common practice.
I very much appreciate the link provided. The theme here is, ostensibly, “Why is there a controversy.” That link is to a review of the study. From that:
The Report provided an unprecedented majority and minority statement of the investigators. The group agreed that there was predictive value in the lipid measures. It diverged in interpretation.
Why was there controversy then? It’s fairly obvious. Social issues, and probably a drive to get “useful results” which can warp science. That page is part of HEART ATTACK PREVENTION A History of Cardiovascular Disease Epidemiology which I intend to use thoroughly. But not yet.
Despite the controversy that surrounded the Cooperative Study of Lipoproteins and Atherosclerosis, there was evidence that cholesterol (regardless how you measured it) was correlated with coronary disease. The work of Carl Muller studying patients with familial hypercholesterolemia was also largely supportive of this link. The work of Brown and Goldstein and their isolation of the LDL receptor would prove the genetic cause of this disease and win the Nobel Prize, but this work was still decades off. However, it could be argued, with some validity, that individuals with a genetic cause for their high cholesterol were not representative of the general population. Nevertheless, by the mid-1950s there was enough interest in this new potential risk factor that large-scale epidemiologic studies were launched.
The launching of those studies was appropriate, given the evidence available. We do need to remember that correlation is not causation. Muller (article linked above) does not clearly relate to the issue under discussion. Of course there is “some validity.” I notice, again, how Labos is organizing his post. I have seen this from fanatics many times: they will assert a series of weak facts that they consider connected, and then they will assert that the preponderance of the evidence (which appears to be the number of facts claimed) demonstrates that their belief is therefore true. It is not the collection of evidence that is the problem, exactly, but the conclusions drawn from it. But then Labos does go closer to the heart of it.
The Seven Countries Study has certainly been one of the most notorious studies of the period and its originator, Ancel Keys, has become a popular target for attack. The main thrust of the attack is that he cherry picked the data in order to obfuscate the truth that saturated fats are unrelated to heart disease. The reality is slightly more nuanced and a detailed review of the Seven Countries Study highlighting its strengths and limitations can be found here for anyone who is interested. Suffice it to say, the main argument that can be leveled against the attempt to deny the role of cholesterol in heart disease is to point out that other studies have shown similar results.
Now, Labos appears to mindread Keys, which I would not do. But perhaps he is merely reflecting the claims. This is obvious: the Seven Countries were selected from a much larger possible set, and I’ve seen results plotted including the larger set. The alleged strong correlation disappears. Did Keys do this deliberately? Maybe. Maybe he had strong political motives, maybe something else. However, the link is to a remarkable document, a detailed defense of Keys that takes into account the critiques.
“Other studies have shown similar results” requires an assessment of “similarity,” which can easily be biased. Further, Labos is slipping from correlation (which Keys claimed and which may exist), subtly into causation, i.e., “role.” There is an abundance of evidence, almost too much. But what would a neutral review (if that is possible, I’m not sure) conclude? And, more to my interest (and Taubes as well, by the way) what research could be designed to definitively answer open questions?
If the opinion is spread that the question is closed, it has already been answered with overwhelming evidence, there will be two outcomes: one is some level of suppression of research and discussion, and the other is a hardening of positions. Nobody likes to be told that they are wrong, everyone knows they are wrong, and they should just shut up and, and what? Die? They are called “die-hards.” People who are willing to question authority, the popular wisdom are precious, if they do not go too far and attempt to oppress others. There is a danger in challenging the status quo. There are few who will welcome difficult questions. They condemned Socrates to death for asking inconvenient questions.
Semmelweiss, on puerperal fever, was right, and was rejected for two reasons: his study showed that physicians were causing the death of patients, many of them, and he also became highly caustic. The personal defects of critics, if we care about science and human welfare, must be set aside to examine claims. This cuts in all directions. I will be reviewing the document on Keys’ Seven Countries study and checking the information there against what is written about Keys.
Studies like the Ni-Hon-San study and the Honolulu Heart Study examined the rate of heart disease in Japanese men living in Japan, Hawaii and San Francisco. They found that compared to the men living in Japan, Japanese men who had migrated to Hawaii had higher cholesterol levels and higher rates of heart disease. Japanese men who migrated to San Francisco had still higher rates. The not-unreasonable conclusion was that the increase in heart disease was environmentally mediated and that as these Japanese men adopted the diet and lifestyle of their adopted country, their cardiovascular risk rose accordingly.
I will need to look at those, but Taubes, for example, attributes the rise in heart disease to the common modern diet, and what is stated here does not show that fat was the causal factor, nor does it show that cholesterol is causal, which is the substantial factual issue. If cholesterol is not causal, but merely associated, then treatments to reduce cholesterol are unlikely to work, except possibly through some associated effect. One of the predictions of the cholesterol hypothesis is that reducing cholesterol will reduce atherogenesis, and a strong effect would be expected, not a weak one. What is the reality?
Finally, we cannot forget the impact of the Framingham Heart Study. Begun in 1948 and still ongoing, this project has provided many insights into the causes of heart diseases. It established that risk factors like cholesterol, hypertension, smoking, lack of exercise, and obesity all affected the risk of cardiovascular diseases. In fact, it coined the term “risk factor”.
Suffice it to say, whatever criticisms one wants to level against the Seven Countries Study, there was plenty of other data suggesting a link between cholesterol and heart diseases. Not unsurprisingly, researchers eventually resolved to try and do something about it.
Looking forward to seeing what Labos writes about this.
(to be continued!)