Sunday, 16 June 2013

Book review: The Great Cholesterol Myth

Title: The Great Cholesterol Myth (2012)
Authors: Jonny Bowden and Stephen Sinatra
Publisher: Fair Winds Press

There is nothing I like more than finding out that something that has been widely believed for decades is wrong. The cholesterol myth fits the pattern of so many items in my blog: it sounds good, it superficially makes sense, and there is a biologically plausible explanation, but when put to the (scientific) test, it fails.

This book covers blood levels of cholesterol, cholesterol in the diet, and drugs to lower cholesterol, and while not everything we hear is a myth, most of it is. The authors make the following claims, and support them with good references to key articles.

Firstly, the relationship between blood cholesterol levels and heart disease is, at best, weak and certainly more complicated than often portrayed. Half of patients with heart attacks have normal cholesterol levels, and half of those with high cholesterol have heart disease. Also, the simplification of ‘good’ cholesterol and ‘bad’ cholesterol is much more complicated than most people think, as there are good types of bad cholesterol and vice versa. Also, this complex association with heart disease is weak compared to other associations such as smoking.

The most important message is that regardless of the complex correlation between blood levels of certain cholesterols and cardiac risk, there is no correlation between dietary intake of cholesterol and blood levels. Despite this, government bodies still recommend restricting dietary cholesterol. This assumed correlation is an example of conclusion jumping and oversimplification, the evidence behind it provides good examples of research bias, and the result of pushing this myth is a good example of the law of unintended consequences.

Like almost anything stored and produced in the body, the level of cholesterol is regulated by a complex homeostatic mechanism. Cholesterol is vital for many bodily functions and makes up the walls of our cells and many of our hormones. It is produced (mainly in the liver) and absorbed from cholesterol in our diet. If we eat more, our body produces less. If we eat less, the body produces more. Specifically, the authors state that dietary saturated fats are not bad for you, and the alternatives (such as hydrogenated vegetable fats and sugar) are worse.

The unintended consequences come from the use of statins (cholesterol lowering drugs). The effect they have is not necessarily mediated by any effect on cholesterol levels (they possibly work via an anti-inflammatory effect). They are only effective in a restricted population (in high risk middle-aged males who have had a heart attack) but even then, they have no effect on all cause mortality and they have significant side effects. Yet, statins are seen as the answer to the problem of heart disease because of the sheer weight of marketing and lobbying behind them and because of our belief that blood cholesterol levels are correlated with heart disease. The presence of statins and the amount of money at stake is one of the reasons the cholesterol myth is perpetuated.

Statin drugs block cholesterol production. For such an important substance, how could we think that blocking production of something that does not directly cause heart disease, and something that is so important for the function of our body, would not have unintended consequences. The side effects are not uncommon and are important, including sexual dysfunction, memory loss, muscle dysfunction and diabetes.

The other unintended consequence of demonising dietary cholesterol is that dietary fats have been replaced with dietary sugar. The authors make a case for the increased sugar in our diets being partly responsible for our poor health (via "insulin resistance" and other mechanisms).

Another repeated point is one that I have made before. Although some of the studies reported in this book show some benefits, they repeatedly show no benefit in overall (all-cause) mortality. See my earlier blog post for why this is the most important outcome.

The more likely culprits causing heart disease are discussed, including sugar, inflammation and insulin resistance.

The last part of the book covers sugars, other dietary factors and stress, and gives recommendation for heart health. The evidence offered regarding eating berries and dark chocolate, as well as laughing and crying (for example), is not convincing and I found this section of the book disappointing, given the high evidence level the authors (rightly) demanded from the cholesterol literature. The book would have been stronger if it finished a few chapters early and stuck to debunking the cholesterol myth. If you have already proven that someone is innocent, finding other suspects does not strengthen your case.

The bottom line

These two authors are not alone in their opinion that cholesterol is not a direct culprit in heart disease, that dietary cholesterol does not influence heart disease or longevity, and that statins are overprescribed and their side effects are under-reported. Their arguments are sound, but turning the cholesterol ship around will take some time.

Addit 18 Nov 2013: It might not take as long as I thought. Recent changes to treatment guidelines in the US seem to detach cholesterol from cardiovascular disease. Although still used in risk calculators, treatment targets for cholesterol levels have been removed (no need to monitor). Statins are still recommended for high risk patients, but the mechanism may not be cholesterol lowering. Other cholesterol-lowering drugs are not recommended. Overview and comment on the changes can be found here.

27 comments:

  1. I think the further we get, the more we realize that people just need to move more and eat less. I believe there is decent evidence for this, no?

    The anti-cholesterol crusade seems to be a medically co-opted diet 'fad' that on surface sounds plausible but doesn't seem to have panned out as you point out here.

    ReplyDelete
    Replies
    1. Thanks. Yes, the blood levels and ratios etc are confusing, but the bottom line is that cholesterol in the diet is not the culprit. More than that, replacing fat in many "lo-fat" foods is sugar and several other things that might not be so good for you.

      Many have made a leap of faith between blood levels and diet levels that, as you say, has not panned out as expected.

      Delete
  2. Actually, statins have been shown many times to lower total mortality in secondary prevention. For example, the Skandinavian Simvastatin Survival Study, published in The Lancet in 1994, showed that simvastatin therapy vs. placebo over 5.4 years lowered total mortality by a statistically significant amount (8 percent vs. 12 percent).

    The only argument is over whether statins lower total mortality in primary prevention.

    ReplyDelete
  3. Thanks,

    Authors agree that statins work in middle-aged males who have had a heart attack and are at high risk of another, but they make three points about statins:
    1. This effect is not necessarily (not all according to them) due to lowering cholesterol levels. Effect of blocking HMG CoA reductase is widespread, and just because it has tow effects, it doesn't mean that effect was mediated by effect A.
    2. Effect is only proven in this population, not for everybody with high cholesterol
    3. Adverse effects of statins are underestimated.

    ReplyDelete
  4. The authors are incorrect. Statins have been shown to work in all secondary prevention populations, not just middle-aged men. They also work in primary prevention in people with cardiovascular risk factors, both women and men. Take, for instance, the JUPITER trial. There was a clear benefit in women, in a primary prevention population. In primary prevention, women tend to have lower baseline risk than men until they reach their 60s, but that is just on average. It is quite possible for a woman to be higher risk than a man. What matters is an individual's risk factors, taken together. Gender is only one aspect and does not provide enough information by itself. Cardiovascular disease is multifactorial.

    Statins probably don't work solely through LDL-lowering, but the important thing is that they work, not how they work.

    Look, I am as skeptical as the next person. I have looked at the evidence myself. I have read many of the trials and meta-analyses. Check it out yourself.

    ReplyDelete
    Replies
    1. Thanks again. I agree that effectiveness is important, regardless of mechanism, and I have stated this previously. However in this case (JUPITER), which used people with high CRP and normal cholesterol levels, the mechanism is important because the results were used to justify lowering the 'normal' levels of cholesterol.

      You agree that we need to separate cholesterol levels from statin effectiveness, as there is too much evidence against cholesterol being the mechanism. That was not done here. Today, statins are being prescribed based on cholesterol levels.

      Regarding the overall effectiveness (1% difference in mortality over 2 years), other aspects of the JUPITER results cause concern for the longer term effectiveness of taking statins (see: http://www.nejm.org/doi/full/10.1056/NEJMe0808320).

      Other concerns about the study have been raised in the literature:
      most vehemently here: http://www.ncbi.nlm.nih.gov/pubmed/20585068
      more objectively here: http://www.courses.ahc.umn.edu/pharmacy/5822/Kaul_by%20Jove%20-%20What%20is%20a%20clinician%20to%20make%20of%20JUPITER_ARch%20Int%20Med%202010.pdf
      and in meta-analysis here: http://archinte.jamanetwork.com/article.aspx?articleid=416105

      The latter study is recommended. It concludes that statins do not significantly influence all-cause mortality in primary prevention. Some studies (like JUPITER) show a decrease, whereas others (most studies) show no difference. This heterogeneity between studies is concerning. The other finding is that any effect seen is independent of any effect on LDL levels.

      And it looks like the debate is still raging on statins:
      http://www.theheart.org/article/1092925.do
      http://ourhealthcaresucks.com/health-care/are-the-statin-wars-about-to-heat-up/

      I think that some of the confusion could be reduced by looking at the effect of statins on patient-specific outcomes (like all cause mortality) as I have referred to above, and not by including phrases like "cholesterol lowering drugs" in the argument (like many papers do) because it implies that this is the mechanism, something that has not been established. They are two separate debates.

      Delete
  5. Hi, the fact that I mentioned the JUPITER trial was not meant to imply that I think everyone who meets the entry criteria of that trial should be on a statin. We were discussing whether or not statins work in women, etc. My position is that they work in women who have cardiovascular risk factors whose risk is at a certain level. I don't think it matters much for statin therapy what the person's baseline LDL level is, so long as their baseline risk, on the basis of all of their risk factors, is high enough.

    To say that a drug works and its effects are statistically significant is not at all the same thing as saying its effects are clinically significant. Moreover, people vary in what they consider a worthwhile benefit. Primary prevention with statins is an issue that should be decided by shared decision making, or by a fully informed patient on their own. I don't think it is any of my business or your business or anyone else's business whether someone takes a statin for primary prevention.

    The JUPITER trial was stopped early for benefit, which could mean that the size of the effect was overestimated. However, I still feel very safe in saying that the trial showed a benefit.

    The meta-analysis of total mortality in primary prevention that was published in Archives showed that total mortality was 9% lower in people who received a statin as compared to placebo. The difference was not statistically significant. To me that is at a minimum a strong indication that statins do not increase noncardiovascular or total mortality and probably lower total mortality slightly in primary prevention. Moreover, it is very well established that statins lower the risk of heart attacks, strokes, and other patient important outcomes in primary prevention. I think avoiding a stroke or heart attack is unquestionably a worthwhile benefit.

    So, once again, I think these authors are not correct in what they say about statins.

    Again, it is an individual decision whether to take a statin, but let's make decisions based on a fair assessment of the facts as they are.

    The fact that an intervention has only a small benefit can make it very difficult to decide if that benefit is worthwhile. There are many areas of medicine that are like that. Primary prevention with statins is one of them for most people.

    ReplyDelete
    Replies
    1. Thanks Marilyn, I agree with all your points. I particularly agree that the decision should be made by the patient after being provided with an accurate representation of the likely risks and benefits.

      The only thing I would point out is that I use absolute risk difference when explaining benefits and risks to patients. The 9% difference you refer to is relative, so that for somebody at moderate risk of death (say 10%) within x years, statins would decrease that expected rate to about 9%, an absolute decrease of 1%.

      Delete
  6. "there is too much evidence against cholesterol being the mechanism".

    What is your proposed explanation for different forms of familial hypercholesterolemia, where the majority of known mutated genes impact LDL-C metabolism and the phenotype is associated with significantly increased risk for CAD?

    ReplyDelete
    Replies
    1. Thanks, and I am sorry that I do not have an explanation but I can make a few points.
      1. The book does not cover hypercholesterolaemia (perhaps deliberately because it does not support their hypothesis?)
      2. The relationship between specific types of cholesterol (LDL-C) etc. and cardiac risk is there, but it is not a strong as many other associations.
      3. Association is not causation, and jumping to that conclusion has caused many mistakes in the past that have taken a long time to correct (like cholesterol in the diet).
      4. The association (that I accept is there) is more complex than relating cardiac risk to the overall amount of cholesterol in your blood. LDL, HDL etc. refer to how cholesterol is carried in the blood, something that is under the control of many hormones and feedback mechanisms.
      5. Familial hypercholesterolaemia does raise issues, as the effect on statins appears to be stronger in this group, but the mechanism of statins in this (genetic and therefore complex) condition is likely to be more complicated than just lipid lowering. For example, one article I saw in BMJ (http://www.bmj.com/content/337/bmj.a2423) noted that lower doses of statins were associated with a greater effect in this population.

      Delete
    2. Thanks for taking the time to reply to my comment. I agree that association is not causation, but I'm not sure what in your mind would qualify as causation. I am well aware of the complexity of cholesterol physiology, having worked on HDL, LDL and hepatic metabolism both in academia as well in industry. As opposed to HDL, the interventional evidence with LDL is incredibly strong, not only with statins (where there is almost a linaer correlation between on-treatment LDL levels and risk reduction) but also with LDL reduction achieved with non-drug interventions such as ileal bypass surgery. The association of LDL-C/apoB with CAD risk is quite strong but, as you point out, not the strongest risk factor which is actually age. No matter if you're a non-smoker with BP of 120/80 and an LDL-C of 90, going from age 40 to age 75 will dramatically increase your risk.

      Finally, I should point out that the BMJ article you cite does not support your stated thesis that statin effect is more than just lipid lowering. What they do show is that moderate dose statins dramatically decrease risk in high-risk patients (FH patients) in a primary prevention setting. There is no comparison of high vs low dose statins and I did not see a correlation with LDL-C levels.

      Delete
    3. Thanks for your comment. I think you have a knowledge advantage over me in this debate. I am happy for you to challenge my comments because my knowledge on this matter has been heavily influenced by the book.

      I think that many of the comments are pointing out that the mortality benefit from statins is highest in people with very high levels of cholesterol (LDL, LDL-C or whatever) and that it appears to be mediated through cholesterol lowering. Then how do statins work in people with normal levels of cholesterol? And are these patients more likely to get harms from cholesterol levels that are too low?

      These are not rhetorical questions, I just can't fit them into the cholesterol hypothesis.

      But am I right in thinking that we agree on one thing: it is not cholesterol itself that is the culprit, it is the form in which it is carried through the blood (LDL vs HDL etc.)?

      Delete
  7. Don't statins also have an effect on pro-inflammatory cytokines? As you are both suggesting, perhaps the cholesterol mechanism is not the culprit here.

    As Marilyn suggests, perhaps the statins help reduce the chronic inflammation in at risk persons?

    ReplyDelete
    Replies
    1. Yes, statins have effects (that we know of) on plaque stability, thrombosis, inflammation, endothelial function and lipid levels. HMG-CoA reductase is such a basic and important enzyme that the effects of its inhibition are likely to be widespread and manifested in many ways. We measure cholesterol and can see the changes, so that was the first guess at the mechanism, and that led to the assumption that dietary cholesterol is bad for you. The history and impact of that assumption are fascinating and are discussed at length in the book.

      Delete
  8. Before pretending to be skeptical, as Marylin Mann said, you have to explore many many books talking about the studies, and who made them, with which sponsor.

    99,9% of the statins studies ARE MADE for finding a positive answer, and used as a marketing tool.
    This is no more "science".
    This is just a construction made by a laboratory to find a positive response in the right population, with the right criteria, controled by the employes of the pharmaceutical company.

    JUPITER is a perfect exemple of a FRAUD.

    4S was never confirmed after, never never, never. 4S is an UFO, a miracle...
    Never confirmed when the study is not totally controled by the laboratory and with the FDA not checking anything in the process, just the results.

    On important book that i recommand to you is : BAD PHARMA
    And the author, Ben GOLDACRE, is an author who is skeptical about ALL the sciences, not ony the medical one.
    You have to read his book.

    ReplyDelete
    Replies
    1. Thanks. You imply bias in the literature, which I agree with, but you imply that this bias is conscious - a deliberate distortion of the truth. At some level this may be correct, but I see the bias in scientific reports to be more subtle, and largely subconscious, and I think this is worse. Let me explain.

      Fraudsters are out there but they are in the minority. It is not sensible for an academic or a doctor, with a long term career to deliberately produce a fraudulent study for personal gain. It is a short-lived gain with a large potential downside. Most of the bias in science creeps in more insidiously through things like researcher enthusiasm, confirmation bias, sloppy methods (a lack of understanding of scientific principles) and a lack of awareness of potential biases.

      I have reviewed Bad Pharma here: http://doctorskeptic.blogspot.com.au/2012/12/book-review-bad-pharma.html#more

      Delete
  9. The statins market is worth over $30 billion per annum, a highly lucrative business. The drug companies that manufacture and sell statins do not want the truth to be known. It would be nice to know how much money these drug companies 'donate' to the FDA and so stop the boat from rocking.

    ReplyDelete
    Replies
    1. To quote from the FDA budget overview: "The FY 2013 total for user fees
      is $1,969,057,000". These user fees are largely drug company fees.

      Delete
  10. Forgive me for not speaking in scientific language as I am a lay person. I keep seeing the statin drugs mentioned here, but what about a non-statin drug such as Lopid? My doctor keeps wanting me to take it, but I am wary of cholesterol drugs! I am a 49 year old woman with bad cholesterol (all the 3 kinds) according to my doctor. I'm sorry to say that i can't recall the actual numbers, but I think the highest number was in the 300's. I know there are people with numbers way higher than mine, because I read the Lopid users reviews on the RXLIST.com site. I am so unsure of what to do! To take or not to take, that is the question!

    ReplyDelete
    Replies
    1. Thanks. Lopid is another cholesterol modifying drug. I am not familiar with the evidence for Lopid and can't really give you any personal medical advice, except to be skeptical, ask the important questions about your treatment and demand to see the most scientifically valid evidence for your treatment and any benefit over alternative treatments.
      Ask your doctor for the evidence for Lopid lowering mortality or cardiovascular risk in someone like you (your cholesterol levels, cardiac history etc.)

      Delete
    2. Thank you for your reply.
      after further reading, I found that lopid, like the statin drugs, increases a person's chance of getting cancer! I think I am going to take niacin as suggested on some other sites that warn about cholesterol drugs. It's quite interesting.
      Thankfully, my doctor, although brainwashed into believing in these bad drugs, doesn't "insist" I take them. My aunt's doctor's office "fired" her b/c she was unwilling to take those drugs!
      It's getting harder and harder for us ordinary folks everyday!

      Delete
  11. I was given statins for four years at increasing dozes as it kept increasing. This happened because I became sicker and could not move to the point that I could not even stand straight. I kept asking my doctor why and got no answer. I insisted on going to an specialist for checking for diabetes and within 10 minutes he told me I had bad side effects from the statin. All my family has high levels even those young and extremely fit. I put on 20 kg from taking the pills due to decreased ability to move. I found out through book written by a guy who went up in space in USA that it was all rubbish and many countries which have high levels do not have more heart attacks eg France and Italy. I told this to a world recognised Heart Specialist at the Gold Coast Hospital who did not seem to know this. I am so angry because I now am heavier and am predisposed to getting bad muscle reaction to other drugs taken for other reason. I seem to have become sensitive to this problem. As recently as a few months ago at a new doctor I was being told to go on Statins. I hope someone in the USA takes Class Action against this con job. It must be the biggest in the world.

    ReplyDelete
  12. Don't even get me started on cholesterol. When I came to Australia in 2004 and started setting up insurance policies it turned out my cholesterol was high. Instead of taking the knee-jerk reaction advice of medical colleagues I started reading. I am not going to re-hash the history of the whole lipid history. One comment: the thing that irritates me the most is that elevated cholesterol results caused immediate knee-jerk advice from colleagues: take statins. No attempt to work up the causes, such as hypothyroidism, stress, depression etc. No, just take the tablet. When I asked if this will make me live longer the answer was always "of course", but then they could never keep meaningful conversation regarding evidence from studies.

    Elevated cholesterol is one of several risk factors for heart disease. Moreover, after the age of about 65 the correlation reverses: old age home residents with higher cholesterol live survive longer. Statins have never been shown to work as the primary prevention means, and in the recent review by Michael Lorgeril the data shows that they are not so hot for the secondary prevention either.

    Yet virtually every one of my friends is on a statin.

    ReplyDelete
  13. Just re-read your blog as TV program Catalyst (Australia ABC) is causing a stir. This thursday 20pm they air the second of a 2 part series on ABC with plenty of air time for I think both of the authors of the book you reviewed here.

    For the 1st part of the series, and the warning to Australian doctors that their patients might come knocking:

    http://www.australiandoctor.com.au/news/latest-news/warning-over-statin-backlash

    As often with these programs, one could be disappointed with the balance. I would be interested in a debate between the 2 groups, discussing the evidence.
    That would be boring television of course, so what they do is let the skeptics have their air time, then refute the heart foundations views, but not the other way around.

    ReplyDelete
  14. There has been intriguing developments in the Cholesterol myth saga with the expose published by Catalyst this week. See it here: http://healingwithintent.com/self-healing/heart-disease-risk-with-high-cholesterol/

    ReplyDelete
  15. Dr. Skeptic, I am a 53 year old man who discovered in 2011 that my coronary calcium score was 1180. An 80% blockage in my Right Coronary Artery was given a stent in 2012. I just read the The Great Cholesterol Myth and am wondering what to do. I have been on Crestor which brought my cholesterol down from 230 to 140. Should I still take it? I did pass my Stress Echo with flying colors last month. I would be grateful for your opinion. Thanks!

    ReplyDelete
    Replies
    1. I really can't give you any personal advice. I question the effectiveness of a single stent for an RCA blockage, and I question the importance of your total cholesterol level. Having said that, it appears that statins can be effective in patients like you, regardless of any effect they have on cholesterol, so I would not tell you to stop taking the drug.
      The book reviewed is a bit of a wake up call for us who believe that cholesterol is the cause of so many problems, particularly dietary cholesterol intake. The book also questions how effective statins are, and if they are, how they achieve their benefits (is it something unrelated to cholesterol?). For you personally, you need to discuss it with your doctor(s) and ask the right questions, for example: what is the evidence that taking this drug will reduce my chance of death or heart attack in the next 5 - 10 years?

      Delete