Statin therapy for healthy people with high cholesterol?

SW VandeCarr

Where does the article say anything about survival? The end point was intima-media thickness of the coronary arteries; the measurement of which has been challenged by the "cholesterol skeptics". It's in the Ravnsknov Oxford Journals paper I posted earlier. If the second drug (ezetimibe) were effective, it would have supported (not proved) the hypothesis that cholesterol levels alone are causal.

I don't consider myself a "cholesterol skeptic". However, I do think they make some good points. Atherosclerosis and its relationship to various presentations of cardiovascular disease (CVD) is complex. It's not a simple matter of blood levels of LDL-C or even the (total cholesterol)/ (HDL-C) ratio. Some of the papers I linked to clearly make this point. C Reactive Protein (CRP) appears to be an independent risk factor and there may be other independent risk factors.

Your use of the word "causal" is IMO inappropriate. Causality is established by experiments. Our ability to do medical experiments involving human beings is obviously limited by ethical considerations. In a true experiment you can more effectively isolate the variables of interest and manipulate them. The most rigorous medical experiment allowed for these kinds of questions is the randomized double blinded clinical trial. It reduces but does not eliminate confounding and the only variable you can manipulate is the treatment. The results are determined on statistical grounds based on normal theory. I'm not criticizing the methodology. This is how I once earned my living, but I do recognize the limitations.

Causality is usually described in terms of "sufficient cause", "necessary cause", and "necessary and sufficient cause". These terms might be applicable to infectious diseases where the infecting organism or agent is the necessary cause (but usually not a sufficient cause) for the relevant disease. I would again challenge you to find a paper in a reputable refereed journal that uses these terms in reference to the present subject. So far you have not supported your arguments regarding your use of the term "causal" as required in these forums.

EDIT: I also find your characterization that regulatory agencies (specifically the US FDA) operate according to their opinions to be unfounded. Major decisions like product approval and labeling are a very public affair involving experts from the academic community and the pharmaceutical companies. Even if the regulators have opinions, there are sufficient checks on any arbitrary and capricious actions on their part. I know the process very well. I would suggest you don't.

Do you consider the articles I linked in refereed journals to be at odds with mainstream medicine? Science is not (or should not be) dogmatic. As I said, I don't subscribe to the complete skeptical thesis. Cholesterol is a factor in CVD and statins are effective, but possibly not only because they lower LDL-C, but also apparently because they lower CRP (Jupiter Trial) and may also have anti-thrombotic and anti-infammatory effects as suggested in linked articles. This, however, remains to be studied.

Apparently you've translated my "healthy dose of skepticism" into some kind of irrational extreme position. I don't know if that's deliberate or not.


The rules say you must support your claims in the science forums by referencing valid sources.

What?

bohm2

Not sure if this study was mentioned but there's also the ARBITER 6-HALTS trial. In that trial ezetimibe produced greater reductions in LDL cholesterol (so-called “bad” cholesterol) but resulted in no overall improvement in carotid intima-media thickness. In fact, individual results suggested greater thickening with greater LDL reductions. The use of etezimibe was also accompanied by a higher number of heart attacks and deaths:
Settings Extended-Release Niacin or Ezetimibe and Carotid Intima-Media Thickness
http://www.ccmconsultants.com/assets..._Thickness.pdf


The ARBITER 6-HALTS Trial (Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 6–HDL and LDL Treatment Strategies in Atherosclerosis)http://content.onlinejacc.org/articl...icleid=1142914

SW VandeCarr

Since you posted 6 in a row without waiting for a response , I'll respond at my leisure to the extent I feel it justifies a response. This post indicates a poor example of scientific reasoning. All the factors you named (and many more), for the sake of this discussion, are considered to be "risk factors" for high serum cholesterol, and in particular LDL-C. Some relationships were discovered in randomized trials were it's presumed that confounding is adequately controlled. However, most of these relationships were discovered in observational studies where potential confounding factors must be identified and controlled. The best of them use very sophisticated multivariate regression models, but for many reasons which I won't list here, they can never justify the confidence one might place in a large well managed randomized trial. Nevertheless, there's little controversy regarding the amount of certain kinds of fats in a diet contribute to high serum cholesterol levels, particularly LDL-D and triglycerides. Other factors such as genetics, and exercise level are considered risk factors for elevated cholesterol (including the total cholesterol-HDL-C ratio.)

To keep things simple, I'll concentrate on just one value, LDL-C. Everything that I talked about above are considered explanatory variables contributing to this one value. This value is considered to be a risk factor for atherosclerosis. Now, we have to make the connection between atherosclerosis and the specific CVD outcomes (heart attack, stroke, etc). Here, we now know that high sensitivity C-reactive protein (hsCRP) is a major independent risk factor for MI (heart attack) through inflammation and clot formation. Your characterization that, say eating red meat causes heart attacks is a misleading and unscientific oversimplification. It may true, but "may be" is not the way you support the claim of strict causality. I will agree that a diet very high in red meat increases your risk for a heart attack if your LDL-C is elevated. I don't agree that we can say this will be the cause of your next heart attack.

Causality is an empirical concept. It's not defined in mathematics or logic. Philosophers have debated about what it really means. I prefer the simple but powerful concepts of "sufficient cause" and "necessary cause". Here the concept of LD(x)50 and LD(x)100 (lethal dose of x for a percent of a population)is useful. At LD(x)100, every individual in the population is dead. In a controlled experiment, we can say that x at D is a sufficient cause of the deaths (ie D(x)=LD(x)100) if all the subjects are dead and all or almost all the controls are alive. Presumably, the subjects in such an experiment are not human beings.

I believe saying x causes y is meaningless unless you state it in terms of a necessary and/or sufficient cause.

Ryan_m_b

The discussion here is far from civil, may I remind everyone that insults and aggressive behaviour will not be tolerated.

Secondly the discussion is descending into pedantry. When discussing articles please reference and quote specific sections when making your case.

ApplePion

They actually were adequately controlled. All of these insights you are treating us to are well-understood by the real epidemiologists who work at the American Heart Association, the Surgeon General's Office and at the National Institutes of Health. You are saying things that are obvious, and irrelevant in context.

ApplePion

I have no idea what that means.

SW VandeCarr

What do mean by "real" epidemiogists?

SW VandeCarr

If it's not a necessary or sufficient cause, what is it? Some people use the term "partial cause" but what exactly is that? It's usually a factor that is correlated to the outcome which you may believe is causally related. But how do you prove it? You're in physics I believe. Do you believe correlation equals causation? By the way, I worked at NIH for a time.

ApplePion

You are the one who is big on authorities.

The sort of problems you are discussing are problems that any serious person, and even many casual people know. While they seem like complex concepts to you, to epidemiologists they are trivial. No one is making the errors you assume are being made.

Why is it that you are always demanding quotes from authorities, while you also assume the authorities cannot get simple stuff right? Do you really think the supposed issues you raised are things only you are aware of? Everyone knows about confounding varuables, and real epidemiologists are not going to just ingnore such. When they say that someything increases the risk of something, it means the risk is increased taking into account confounding variables.

ApplePion

It often does, but not always. What I do believe is that everytime I have seen someone object that causation does not imply correlation, that person was arguing against something where the correlation was clealy causal. (Ironically, that is an interesting correlation.)

We know that the relationship between LDL cholesterol and heart disease is causal because it occurs through all subgroups of people with elevated LDL, and because almost any way of lowering itlowers risk. This is unlike the correlation between coffee and lung cancer--the elevation is found only in one subgroup--smokers--and so we know it is not really coffee that is the cause. In the coffee case, the coffee is not harmful, but rather is correlated with the thing that is really harmful. But it does not work that way with cholesterol--the correlation with heart disease is unrestricted.

SW VandeCarr

Why did you cut off the quote? The next sentence refers to observational studies where you hope you've controlled for the major confounding factors in the analysis. These comprise the majority of the studies on specific risk factors for elevated cholesterol. I also go on the say that in the aggregate I believe the overall results. Unfortunately, you can't randomize patients to study the effect of diet, and life style differences

ApplePion

There are lots of things you cannot randomize. You cannot randomly construct group of people and have them use dirty hypodermic needles, and randomly another group of people and have them not use needles. But even though no randomized trials have been done it still clear that the correlation between getting AIDS and using dirty needles is causal.

Users of illegally injected hypodermic drugs differ from the general population in many ways--higher smoking rates, worse diet, etc., but despite such confounding variables, we still really know that dirty needles is causally related to contracting AIDS.

SW VandeCarr

Well, not necessarily. Here in PF we need to post links to authoritative sources. I don't mind this at all as long as it's something that really needs support and has not already been cited in the same thread. But I can look at a paper critically and evaluate it.

I think the central issue in this thread is the recent evidence that cholesterol is not the whole story in CVD (CRP, inflammation, thrombus explanation). This is in the first few paragraphs of the 2002 "cholesterol skeptic" paper that I posted and to which you seem to object to so much. I don't accept their full thesis, but they seem to have gotten that one right. They also claim that the method of measuring arterial wall media-intima thickness is inaccurate. I don't have an opinion on that.

SW VandeCarr

Of course. You do the best you can, but you may not get the most convincing results.

SW VandeCarr

You can use the word any way you wish, but don't try publish in a medical journal that you've proved your causal hypothesis by means of a correlation. In the case of LDL-C, it is found in plaques, but there's still a lot of research regarding the stability of plaques. Stable plaques can remain asymptomatic for a long time. When they become unstable, there is a high risk of intimal rupture and inflammation leading to a clot. It's not known what causes this destabilization so you can't say cholesterol is involved at this stage for certain.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2659534/

Evo

Please quote what you are referring to so that we don't have to guess.

Thanks.

SW VandeCarr

This is the third paper listed at the bottom of post 15. I mistook this one for the 2002 paper by the "cholesterol skeptic", published in the Oxford Journals. This is a 2003 paper which expresses similar points of view regarding the role of cholesterol but uses different arguments and may not be associated with the "cholesterol skeptics".

http://jcem.endojournals.org/content/88/6/2445.full Here's the quote:

"Atherosclerosis is a complex multifactorial disease. Lipids play an important, but not an exclusive, role in its development and progression. In some persons, lipids will be a major factor, and in some, lipids as we currently understand them will play a minor role. Outstanding advances have been made in understanding the biology of the vessel wall and of atherosclerosis, but there is still a long way to go. Our concepts of atherosclerosis and coronary heart disease (CHD) have dramatically changed, and we now know that coronary atherosclerosis is a diffuse multifocal inflammatory vasculopathy (1, 2, 3, 4). Rupture of nonocclusive lesions (<50% of the lumen) are often the most dangerous, abruptly causing sudden death or the acute coronary syndrome (ACS; Ref. 5). The important fact about these nonocclusive or culprit lesions is that there are many of them in the coronary circulation that are as dangerous as the one that causes the ACS. In fact, it is these lesions, and not necessarily the one causing the ACS, that are responsible for recurrent CHD events after myocardial infarction or the development of unstable angina (4). It is for this reason that antiatherosclerotic therapy (including lipid-lowering therapy) should be aggressive in patients with CHD or at high CHD risk. The most dangerous lesions are nonocclusive, asymptomatic, and not necessarily detected by stress testing, with or without imaging, or by coronary arteriography because the lesions may be accompanied by compensatory dilation with little or no encroachment on the lumen (6)."