Friday, February 25, 2011

The Origin of the Lipid Hypothesis -- And Proposal of a New Term

by Chris Masterjohn

In my last post (The Proper Use of the Term "Lipid Hypothesis"), I traced the origin of the term "lipid hypothesis" to Edward Ahrens in 1976, who defined it as the hypothesis that reductions in blood cholesterol levels will lead to reductions in heart disease risk.  This can be seen as a testable prediction of the underlying hypothesis that high concentrations of cholesterol in the blood cause heart disease.

Ahrens popularized the phrase and published what was apparently the first full-length paper devoted to defining it, but was not the first to use the phrase.  I included a footnote in the last post with a reference to Daniel Steinberg using it in print in 1974 with the exact same meaning.  There may be earlier references.

These authors used the term to refer to a hypothesis quite distinct from the diet-heart hypothesis, or the "diet-heart question," terms that had first been used in the early 1960s and defined more clearly in the late 1960s under the supervision of a panel chaired by Ahrens.

While these events represent the origin of the term "lipid hypothesis," they do not represent the origin of the hypothesis itself.  After all, while the lipid hypothesis does not in any way depend on the diet-heart hypothesis, the diet-heart hypothesis is directly dependent on the lipid hypothesis because it proposes that dietary fat, or dietary saturated fat, causes heart disease precisely by raising the concentration of cholesterol in the blood.  If the diet-heart hypothesis existed in the 1960s, certainly the lipid hypothesis also did.

The actual lipid hypothesis itself pre-dates the diet-heart hypothesis by decades and traces to Anitschkov's cholesterol-fed rabbit.  I have discussed this model in detail in my article, "High Cholesterol and Heart Disease — Myth or Truth?"

Anitschkov and his partner Chalatov never intended to investigate cholesterol and were following up on earlier work suggesting that dietary protein accelerated aging.  They were investigating atherosclerosis because it was seen as a sign of aging.  

They found something quite different, and Anitschkov abandoned the idea of a dietary model of atherosclerosis in favor of a metabolic model.  Anitchkov was very clear that his work suggested nothing about dietary cholesterol in humans.

In a 1932 review,* cited in my Myth or Truth article, Anitschkov wrote the following (I will use bold for emphasis but all of the italics are his):
[I]n human atherosclerosis the conditions are different. It is quite certain that such large quantities of cholesterin are not ingested with the ordinary food. In human patients we have probably to deal with a primary disturbance of the cholesterin metabolism, which may lead to atherosclerosis even if the hypercholesterinemia is less pronounced, provided only that it is of long duration and associated with other injurious factors. . . . [O]n the basis of the experimental results we are certainly justified in stating that atherosclerosis belongs to the "metabolic diseases," and that cholesterin is the "materia peccans." [the "ill-making substance"]  It is not possible to deny that some other aspects of the metabolism can be disturbed at the same time, but in this respect hardly anything is known. 
Anitschkov rejected the myopic view that cholesterol is "the cause" of atherosclerosis, as if any one molecule could be the single cause of a complex disease.  Nevertheless, he considered it essential to the disease.  A pre-requisite, but not the only factor.
It would be entirely wrong if, on the basis of these conclusions, we were to describe cholesterin or rather hypercholesterinemia as "the cause" of atherosclerosis.  But that cholesterin plays an important part in this process, as far as its etiology is concerned, has now been definitely established as a fact by these experiments. . . . In brief, atherosclerosis never develops without cholesterin. 
Anitschkov was very decisive in arguing that atherosclerosis is a multi-factorial disease:
But, to say it again, it would be wrong if, on the basis of these experimental results, we were to regard this factor as of exclusive etiologic significance.  On the contrary, all pertinent observations recorded in these experiments point to the probability that several factors contribute to the genesis of atherosclerosis  — factors both of a general and of a local nature.
 Anitschkov reviewed experimental evidence that the following factors also contribute to atherosclerosis:
  • High blood pressure
  • Local arterial injury
  • Inflammation 
  • Other dietary factors (in the rabbit, protein)
He reviewed evidence that the following factors protect against athrosclerosis:
  • Sex hormones
  • Thyroid hormone
  • Iodine
 And he considered it likely that the following factors may also play a role:
  • Infection
  • Disturbances in the nervous system
He concluded thus:
If we now survey all the experimental results in their entirety, we find that quite a number of different factors are involved in the etiology of atherosclerosis. . . . The views here set forth concerning the etiology of atherosclerosis constitute what I have called the "combination theory"of its origin.
The "lipid hypothesis," then, from the beginning, never claimed exclusivity for cholesterol, and never denied the role of inflammation, hormones, infection, and other dietary and metabolic factors.  Nevertheless, it did give primacy to elevated levels of blood cholesterol.

Evaluating the lipid hypothesis is somewhat difficult, not only because misuse of terminology and conflation with the diet-heart hypothesis is so widespread (indeed, conflating these two hypotheses was exactly what led to the mass media's condemnation of fatty animal products in the 1980s, which will be the subject of the next post).  It is difficult because theories change over time as they attempt to accomodate new evidence.

Someone hell-bent on destroying the lipid hypothesis might accuse its proponents of creating a "moving target," but the reality is that when a "beautiful hypothesis" confronts an "ugly fact," its two legitimate choices are to change or die.

Hardly any mainstream scientists would accuse evolutionary theory of creating "moving targets" because it has undergone a great deal of revision and expansion since Darwin.

I have described in my Myth or Truth article, and in my last Wise Traditions lecture, "Heart Disease and Molecular Degeneration: The New Paradigm," what I would consider overwhelming evidence that the degeneration of lipids in the blood is essential to the initiation of atherosclerosis and plays a less prominent role in its progression.  I will provide further evidence of this several blog posts from now.

Is this a form of the lipid hypothesis?

There is nothing about the term "lipid hypothesis" that would suggest it is not.  After all, it is a hypothesis that involves lipids. 

However, it is a hypothesis of a fundamentally different character from that of Anitschkov, Ahrens, and Steinberg, even though it heavily incorporates the experimental contributions of all of them.

Anitschkov considered one of the defining characteristics of his theory of atherosclerosis to be that it considered atherosclerosis infiltrative instead of degenerative.  He focused on the amount of cholesterol, as did Ahrens and Steinberg after him.
It may therefore be regarded as certain that in these experimental animals large quantities of the ingested cholesterin are absorbed, and that the accumulation of this substance in the tissues can only be interpreted as deposits of lipoids circulating in large quantities in the humors of the body.  The establishment of this fact is of importance, because it shows that atherosclerosis is not essentially of degenerative nature, but rather of an infiltrative character.
In fact, he even called his theory the "infiltration theory."
Above all it has made it possible definitely to demonstrate that the initial stages of the process are in the nature of a lipoid infiltration or imbibition of the intima, and that the lipoids enter into the arterial wall from the lumen.  This has greatly strengthened the "infiltration theory" of atherosclerosis.
Anitschkov saw his "infiltration theory" as running in close partnership with his "combination theory."
The views here set forth concerning the etiology of atherosclerosis constitute what I have called the "combination theory" of its origin.  This theory is closely related to the "infiltration theory" of the formal genesis of this disease, because both theories are in part based on the operation of identical factors as established mainly by recent experimental research.
I think that Anitschkov's belief that atherosclerosis was infiltrative rather than degenerative (that is, that any degeneration followed infiltration of the lipid driven by its high concentration in the blood) was the most reasonable theory at the time, given the evidence available to him.

However, our understanding of the molecular mechanisms of atherosclerosis has progressed dramatically since his time, and we have definitive evidence that it is degeneration of lipids rather than their high concentration that drives atherosclerosis.

Yet proponents of the "lipid hypothesis" have nevertheless subjugated this evidence to constitute a mechanism for a theory that remains fundamentally infiltrative in character, based fundamentally on the amount of lipid in the blood.

I think, then, that the term "lipid hypothesis" is insufficient.  Anitschkov, brilliant as he was, used much better terminology.  If only it involved the term lipid or cholesterol, it would have been perfect.

I propose, then, that my view — that degeneration of lipids plays an essential role in the initiation of atherosclerosis and a less prominent role in its progression — is a lipid hypothesis, but it is not THE lipid hypothesis.

I propose a new terminology to distinguish between these two related but fundamentally different hypotheses:
  • The infiltrative lipid hypothesis is the hypothesis of Anitschkov, Ahrens, Steinberg, and the medical establishment, focused on the amount of cholesterol in the blood driving its infiltration into the arterial wall, and considering this infiltration to drive any subsequent degeneration.

  • The degenerative lipid hypothesis is the hypothesis that considers the amount of cholesterol largely or perhaps entirely irrelevant, and considers the degeneration of lipids in the blood to drive their infiltration into the arterial wall, and this infiltration to actually be partly protective against worse forms of degeneration that would have occurred in its absence.
Lest we become myopic, we should fully retain Anitschkov's combination theory as a powerful and intimate partner of the degenerative lipid hypothesis.

Read more about the author, Chris Masterjohn, PhD, here.

* Anitschkow N. Experimental Arteriosclerosis in Animals. In: Cowdry EV. Arteriosclerosis: A Survey of the Problem. New York: Macmillan. 1933.


  1. I'm going to start throwing around "ILH" and "DLH" to help get the word out.

  2. Chris,

    The lipid hypothesis ultimately states that lowering the amount of lipoproteins in the blood will lower your risk of heart disease. So falsifying the hypothesis would be a matter of producing cases where lowering lipoproteins was not accompanied by lowered heart disease.

    It reads to me that your theory states that lowering the degeneration of lipoproteins in the blood will decrease heart disease. So again, falsifying it would be a matter of producing material cases where decreasing degeneration of lipoproteins does not decrease heart disease.

    I'm curious for your response to this, I wonder if I'm thinking about this correctly.

    I wonder about the adaptive nature of lipoprotein degeneration. Why do they do that? I'm swayed by the idea that oxidized LDL is sticking plaster, intended to patch damage to arterial walls. If that idea is accurate, then reducing oxLDL may not decrease heart diseases, because you now have damaged arteries going un-patched.

    Finally, I put together a list of modern studies where decreasing lipoproteins did not yield better outcomes. I find the list persuasive against the infiltrative lipid hypothesis.

    All studies found referenced from de Lorgeril's paper:

    ENHANCE: cholesterol was lowered, IMT increased. (

    CASHMERE: cholesterol was lowered, IMT increased. (

    ACHIEVE: stopped early because sponsors and investigators thought that they would not be able to show benefit. (

    SEAS: cholesterol was lowered, CHD and aortic valve complications were not reduced. ( (Oh, and the intervention group died of cancer more frequently than those in the placebo group. Minor issue. Please ignore, move along.)

    GISSI-HF: cholesterol lowered, intervention group had more deaths than placebo group, although not statistically significant. (

    CORONA: cholesterol and CRP were lowered, but the composite of cardiac death, nonfatal infarction, and nonfatal stroke were not lowered. (

    Not directly discussed in his paper, but referenced by them:

    ASPEN: cholesterol was reduced, no reduction in the composite of cardiovascular death, nonfatal infarction, nonfatal stroke, recanalization, heart bypass, resuscitated cardiac arrest, and worsening or unstable angina requiring hospitalization. (

    4D: cholesterol was lowered, no reduction in composite of cardiovascular death, nonfatal myocardial infarction, and stroke. (

    IDEAL: cholesterol was lowered, no reduction in major coronary events or cardiovascular or all-cause mortality, however risk of composite secondary endpoints was reduced. (

    ILLUMINATE: HDL way up, LDL way down. Deaths: way up! This one is really a doozy. (

  3. Hi Ed,

    I almost agree with you, in principle, about falsification, but I think this is somewhat problematic.

    Falsification of a hypothesis in a biological system must be based on consistently negative tests from a variety of angles, since there is rarely if ever an intervention that does one and only one thing. Many treatments might affect the endpoint in question as desired, but do it in such an indirect way so as to cause some other confounding effect (whether positive or negative in nature, which is why even supportive trials must be viewed very carefully and skeptically).

    Also, as the number of tests of the hypothesis increases, one must allow for a greater number of negative tests. See my post "The Great Unknown: Using Statistics to Explore the Secret Depths of Unpublished Research."

    If a given effect is the "true effect," then results from trials with similar design, sample sizes, and population characteristics will produce a distribution of results that randomly fall about the "true effect" and have it as their mean. How tightly they fall about this mean will depend on statistical precision, which incorporates both 1) interindividual response, and 2) sample size. Thus, it is very possible for a hypothesis to be true and for a number of negative tests to be done, and even for a handful of cases to show the opposite.

    So if you have an inordinate number of tests -- as you have in this case -- then I think what you need to do is group the trials together and examine the combined effects with and without separating them by treatment, trial design, size, duration, population, etc, and examine the data for publication bias and other issues. In other words, perform a meta-analysis.

    That said, I think the infiltrative lipid hypothesis is false for a number of reasons.

    I also think the idea that oxidized LDL is a repair substance is quite definitely false, both because it is essentially an absolute pre-requisite for atherosclerosis and because it is toxic to cells. I don't see how something that kills endothelial cells is going to repair them.


  4. Chris, thanks for the informative reply. I see that falsifying a biological hypothesis is not trivial. The general concept stands, but as you say, requires real effort and sufficient (copious) data.

    I can't shake the feeling that oxLDL might somehow be adaptive, else why do we have it, but I look forward to your continued writing on the topic.

  5. First, while your blog efforts are appreciated, you need to get these ideas published where the "medical establishment" can see them, not that I expect them to pay much attention (not unless your ideas would result in increased statin sales, in which case they would be widely adopted!). The game is about getting articles published and cited and giving talks to advance your theories. I hope you're in a position that you can do some of this.

    Second, your view is consistent with this paper The first sentences of the introduction state "Atherosclerosis was once thought to be a simple lipid storage disease that caused pathology by arterial obstruction. The pathology leads to arterial obstruction, but not simply by accumulation of lipid."
    The paper goes on to discuss toll-like receptors which mediate inflammation and inflammatory phenotype in endothelial cells, and other factors which upregulate and stimulate these receptors (like oxLDL among others).

    Also, the infiltrative theory also doesn't account for the role that triglycerides and HDL may play in exacerbating or relieving disease (other than calling HDL a garbage truck, which is a grossly oversimplified view of this complex molecule.

    So the infiltrative theory is passe among more enlightened researchers, though obviously not all. It's the degenerative theory they're not sure what to do with yet, other than prescribe statins for their supposed pleiotropic effect (maybe weakening arterial smooth muscle, the same way it causes striated muscle pathology?). Unfortunately, research efforts are often supported and advanced by what $ can be made, not by science and health concerns, so it may be a while before these hypotheses are thoroughly tested and understood.


  6. Chris youre on the right path but as the Dr's above said to gain traction with my collegues you need to get someone tied to this work. Youre at UCONN so you have a chance to talk to some of the vascular/CV/or cardiologist up there. You need to speak with Jeff Volek to get some inside tips of who to approach first. And I think recruiting a young fellow in one of these fields may be a real smart move. They need an angle to put on a CV and you need them to get this thing published in a journal we all read. Some of us (like me) constantly read all the off label journals but the sad truth is 99% dont. Moreover, that is why many still think TC and low LDL is the goal. I can tell you privately there are many cardiologist and CV surgeons who know the only reason they are writting for statins is because of medical standards and legal risks of not doing so and they know that statins have a very low anti inflammatory effect likely responsible for the small reductions in inflammation. I think the key step is getting all MDs to realize that VAP need to replace the standard lipid panel and we have to get them to understand that sdLDL is the particle that becomes oxLDL and is the cause of PAD, CAD, and atheroscerolsis. I also am encouraged because many cardiologist are beginning to realize that some of the statin trials done recently have shown calcium index scores are critical and they will mentally be able to make the connection once you get the science pubished infront of their eyes. The statin manufactures will do their best to keep this data out of journals because they "own" most of the editors in these journals.

  7. And Chris I have not told you here or on FB but I as a doctor appreciate your dedication and effort to helping us all improve our former selves.

  8. So, if in fact CAD is a metabolic disease and if one has a high proportion of small LDL( via NMR or VAP) as a % of total LDD after eliminating Wheat, sugar, Omega 6, etc what would be the recommendation? Clearly meds are likely needed and in these cases the best may be the disliked statins. Of course, this requires doctors to do the proper testing to make sure they treat the right patient population and not wholesale admin of statins.

  9. the taste you love without the cholesterol:

  10. Hey Chris. Need some advice.

    My total cholesterol is 250, HDL is 85, and ldl is 165. My fasting blood glucose is always around 65-70. Is the cholesterol anything to be worried about. I know the high hdl is great, but what about the high LDL?

    J Boat

  11. Please excuse the off topic commment. I just listen to your interview with Jimmy Moore. It was excellent.

  12. thank you chris i really like your blog and i find all the blog very help full and very informative, i a really appreciated you and hope that you will keep it up
    Transcript Translation

  13. Speaking of Jimmy Moore, he recently interviewed Dr. Cordain and Cordain stated that a diet which includes high saturated fats (e.g., Paleo) will indeed create atherosclerosis. However, since such a diet is anti-inflammatory, the atherosclerosis under these conditions does not create risk since there is no thinning of the fibrous cap etc. Is this a supportable assertion?

  14. Responses to Drs. Cynthia and David, John, Steve, Anonymous, J Boat, Walter, and Spiney

    Hi Drs. Cynthia and David,

    Thanks so much for your suggestions and kind words. I will indeed try to reach out beyond the choir on these issues. I do think that the "infiltrative" lipid hypothesis has a way of explaining HDL-C, although I suppose we are in agreement that the evidence is very poor for implicating reverse cholesterol transport as the primary protective mechanism of HDL particles. Or perhaps we could say the evidence flatly contradicts this sacred cow. I do think, however, that TG can be explained by the "infiltrative" lipid hypothesis as there is some evidence that even non-oxidized VLDL particles can generate foam cells, although I will have to review this evidence more closely and think about it more. I find it interesting that often writers about inflammation begin by stating that atherosclerosis was once seen purely as lipid accumulation, when Anitschkov had so clearly and extensively documented the role of inflammation back in the days of the cholesterol-fed rabbit, though now we have clarified the molecular mechanisms in so much more detail.


    Good points and thanks so much for your kind words.


    I'm not entirely convinced that small dense LDL is a causal culprit, although I think it's quite plausible. I'm not entirely sure we know enough about how small dense LDL is formed, whether there are multiple ways of forming it, and whether these all act the same or if different pathways of forming small dense LDL have different relationships to heart disease risk. That said, there is evidence that it is formed from large VLDL, which is formed when synthesis of fat in the liver is high, which would be reversible, theoretically, by eating low-carb and/or low-fructose, or by reversing insulin resistance if insulin resistance is present. I think there is also some evidence that it could form from oxidation (chicken-and-egg issue here). In any case, if various background promoters of poor LDL metabolism are present that cannot be resolved, such as heterozygous familial hypercholesterolemia, it is possible that very careful treatment with natural thyroid hormone UNDER MEDICAL SUPERVISION might prove helpful, even if one is not hypothyroid. However, I think it would be wise to focus most on the downstream events, such as normalizing inflammation, nutrient status, and thrombogenic status.



    J Boat,

    Actually your total-to-HDL-C ratio is under 3, which is a better predictor of risk than LDL-C, and indicates low risk. (Note a "predictor" is not necessarily a cause.) However, maybe to make yourself fully comfortable, perhaps take a look at more direct measures of atherosclerosis, calcification, etc.


    Thanks! Glad you liked it.


    Yes and no. I think the idea that saturated fat promotes atherosclerosis is wrong. I also think that atherosclerosis is probably a bad thing even if it does not lead to a heart attack. However, I agree with him that the initiation of atherosclerosis is not the final mediator of a heart attack, and that inflammation plays an essential role in the progression of atherosclerosis to an actual eventual fragmentation of a lesion with ensuing thrombosis and infarction.


  15. Hi Chris, please consider writing this into a screenplay (along the lines of Thank You For Smoking [2005]). It's big, important, but way too complicated for a blog post.

  16. "Falsification of a hypothesis in a biological system must be based on consistently negative tests from a variety of angles, since there is rarely if ever an intervention that does one and only one thing. Many treatments might affect the endpoint in question as desired, but do it in such an indirect way so as to cause some other confounding effect (whether positive or negative in nature, which is why even supportive trials must be viewed very carefully and skeptically).
    no homo just bromo, but wow, why can't more ppl think this rationally (although i love my noneducated friends)? Often times in debunking others, we fail to hold our supportive stuff to a high degree.

    sometimes i think i'm wasting my life just learning/reading about things i'm interested in, cooking/eating, and relaxing (quit office work after half a year), but I can't imagine going back to deal with such politics. but dr cynthia makes such a great point reminding us that we hold significant power in change. what do ppl here think we can all do without taking too much stress to get the right info or type of thinking out besides just suggesting everyone we know and our doctors to start looking into this stuff? I wouldn't want to go to school again or deal with writing solid papers (researching/reading yes but writing no) I have been thinking about compiling some of this info in a presentation and presenting it to interested institutions (churches, other non gov affiliated places maybe?) Does anyone know a good post on how the best way to explain these fallacies to friends and loved ones so that they don't think you're crazy. Chris, maybe i can couchsurf your uconn dorm and be your assistant (50+% serious)

  17. Hi Chris,
    I'm doing a project about how the lipid hypothesis has negatively affected American health. Is there a direct source that is the exact lipid hypothesis? If so, can you please direct me to it?
    Thank you!

  18. "In my last post (The Proper Use of the Term "Lipid Hypothesis"), I traced the origin of the term "lipid hypothesis" to Edward Ahrens in 1976, who defined it as the hypothesis that reductions in blood cholesterol levels will lead to reductions in heart disease risk. This can be seen as a testable prediction of the underlying hypothesis that high concentrations of cholesterol in the blood cause heart disease."

    No, the underlying hypothesis is not that high concentrations of cholesterol in the blood cause heart disease... even limiting the lipid hypothesis to heart disease. A "causative factor" only ends up meaningful from looking at real, significant, statistical relationships in populations. To say that high concentrations of cholesterol in the blood cause heart disease, comes as to make a claim about heart disease in individuals, not populations. The lipid hypothesis basically concerns populations...

    "What the hypothesis does propose is that hypercholesterolemia is what might be called a *determining factor* i. e. it is sufficiently dominant that correcting it will significantly reduce the burden of disease and its clinical consequences even if hypercholesterolemia is the sole variable manipulated." Daniel Steinberg _The Cholesterol Wars_ p. 1.

    The phrase "burden of disease" indicates something about populations, since a sole individual having a disease does not pose a "burden" for the medical community in general.

    "The degenerative lipid hypothesis is the hypothesis that considers the amount of cholesterol largely or perhaps entirely irrelevant, and considers the degeneration of lipids in the blood to drive their infiltration into the arterial wall, and this infiltration to actually be partly protective against worse forms of degeneration that would have occurred in its absence."

    For this to meaningfully replace the lipid hypothesis, we'd have to have a rigorous definition of what a vague term like "degenerative", and then strong evidence that such degenerations of lipids as a better predictor of atherosclerosis in populations than serum cholesterol levels. Such would have an impact for medical practitioners only if such degeneration can get measured and determined more easily than serum cholesterol levels, and the benefit of such a measurement of "degeneration" would work out as significant.

    Falsification doesn't matter much in terms of theories which have already proven useful.

    "A discrepancy in Mercury's orbit pointed out flaws in Newton's theory. By the end of the 19th century, it was known that its orbit showed slight perturbations that could not be accounted for entirely under Newton's theory, but all searches for another perturbing body (such as a planet orbiting the Sun even closer than Mercury) had been fruitless. The issue was resolved in 1915 by Albert Einstein's new theory of general relativity, which accounted for the small discrepancy in Mercury's orbit.

    Although Newton's theory has been superseded, most modern non-relativistic gravitational calculations are still made using Newton's theory because it is a much simpler theory to work with than general relativity, and gives sufficiently accurate results for most applications involving sufficiently small masses, speeds and energies."

    In other words, Newtonian mechanics did get falsified. It didn't get thrown out. It got superseded by general relativity *after* it had gotten falsified. It still didn't get thrown out even with general relativity, because general relativity doesn't provide enough of a benefit for most things physicists want to do. Physicists know that Newtonian mechanics isn't the whole story, and they know it has gotten falsified. They oftentimes don't care, however, because Newtonian mechanics ends up as too useful for so many situations.

    1. Hi Spoonwood,

      I think you are making important distinctions but exaggerating them to the point of denying the clear connections between these ideas.

      The "lipid hypothesis" as stated by Ahrens and Steinberg relates to population-level clinical treatment, but it is the clinical implication of the hypothesis that high blood cholesterol causes atherosclerosis on an individual level, which was Anitchkov's inference from his animal model. If you read Steinberg's book, Cholesterol Wars, you will see that in his mind these ideas are connected in unbroken continuity.

      I disagree that a theory of causality is justified by its practical usefulness, but that seems like a semantic point I unfortunately don't have time to pursue with you. Thanks for writing.


    2. This comment has been removed by the author.

    3. I have about a chapter left in Steinberg's book. I simply do not see that he thinks those ideas connected in unbroken continuity. In other words, I do not think that he believes high blood cholesterol *always* causes atherosclerosis on an individual level. The section on inflammation appears to make this clear. On p. 113 he writes:

      "Weighing the Relative Improtance of Inflammation and Hyperlipidemia

      Whether inflammation alone, i. e. *in the absence of some elevation of blood cholesterol*, can ever initiate atherosclerosis is not clear. Various forms of arteritis can of course be generated, but the lesions do not closely resemble those of human atherosclerosis. There is no evidence for "inflammatory atherosclerosis," that is, atherosclerosis in the abscence of hypercholesterolemia. Nevertheless, the rate of progression of lesions once established, at any given level of blood cholesterol, can be strongly influenced by the multiple growth factors and cytokines produced within the lesion (or reaching it from elsewhere in the body) and by the immune system. Hopefully, with better understanding of the complex events triggered by the cells in the lesion and by the immune system it may be possible to identify one or more vital step-Achilles heel steps-at which intervention will significantly slow the growth of lesion and/or stabilize the plaque *independently* [emphasis added] of the blood cholesterol level.

      Hypercholesterolemia, on the other hand, *can* [emphasis added] of itself be a sufficient cause of atherosclerosis, as in patients with familial hypercholesterolemia or in the LDL receptor-deficient rabbit, suggesting that inflammatory changes are downstream consequences of hypercholesterolemia."

      Since Steinberg says "can" there instead of "is" or something similar, as well as the end of the previous paragraph, at bare minimum, it seems that Steinberg doesn't think the relevant ideas connected in unbroken continuity. Note that Steinberg also doesn't give any indication as to how often he thinks that such downstream consequences of hypercholesterolima occur. Perhaps extremely often, but not necessarily always.

    4. Hi Spoonwood,

      I think we miscommunicated. I did not mean "unbroken continuity" between high cholesterol and heart disease such that if cholesterol is high, heart disease necessarily ensues regardless of other factors. I meant unbroken historical and theoretical continuity between Anitchkov's work and the "lipid hypothesis." If you ask Steinberg what the first foundational evidence for the lipid hypothesis was, I am quite sure he would say it was the cholesterol-fed rabbit.

      In any case, I think it is quite clear that Steinberg believes cholesterol-lowering drugs reduce the proportion of people in a treatment population suffering from heart disease precisely because high cholesterol causally contributes to heart disease on an individual level. To tear these notions apart from one another is to render them incoherent: populations do not get heart attacks; individuals do.