• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Cholesterol Myths

Update - Waiting for Deutsch

This started out as a debate between myself and Deutsch. I sent my last message to him 3 days ago (copied in message #9 above) and have as yet received no response. If he doesn't answer, does that mean I win? :)

I was trying to take things one step at a time. There were several more errors and misinterpretations in his presentation that I wanted to address. If I don't hear from him soon, I'll proceed on my own.
 
Deutsch Replies (#3)

You state that a high-fat diet is a contributing factor to atherosclerosis,
but the Harvard School of Public Health Newsletter
places the blame on only saturated fats and trans fats (Trans
being far worse than saturated)
and states that other fats are protective. And Mary Enig (Know Your Fats:
The Complete Primer for Understanding the Nutrition of Fats, Oil, and
Cholesterol) states that saturated fats have gotten their bad rap because earlier
studies did not clearly distinguish them from trans fats or take into account the
trans fat contents of studied fats.
The Framingham studies do show a correlation between cholesterol levels and
heart disease, but only up to the age of 55 and it is hard to believe that the
etiology is different for older than for younger subjects. Also, the
Framingham studies showed that those who ate the most saturated fat had the lowest
rates of CHD and overall mortality! (Castelli WP, Archives of Internal
Medicine, Jul, 1992; 152: 1371-1372).
In the MRFIT studies, which compared mortality rates and eating habits of
12,000+ men. Those who ate less saturated fat and cholesterol showed a slightly
reduced rate of heart disease, but had an overall mortality rate much higher
than the other men in the study. I'll have to go to the medical library to
see some of the MRFIT references, since PubMed gives tantalizing titles but no
abstracts. See Corr, LA and Oliver, MF The low fat/low cholesterol diet is
ineffective. Eur Heart J. 1997 Jan;18(1):18-22.
If higher blood cholesterol levels are protective against heart attacks (as
I believe the data show) and a drug is more likely to be administered to
sufferers from first heart attacks who have higher blood cholesterol levels, then
it will appear, as Uffe Ravnskov has suggested (private communication) that
this drug is of value in secondary prevention, even if it has no effect at
all.
Please forgive my slow response. I will be a slow responder for about two
weeks during which I will be working on a mycological publication of which I am
the editor.

===========
I guess you've seen this: [JAMA Article on Low-Fat Dietary Pattern and Risk of Cardiovascular Disease - Feb 2006] My citing it doesn't mean that I'd accept a reduction in "risk factors" as meaningful. I'd want to see a reduction in heart disease, or, better, in mortality.
 
The SkepDoc Answers Deutsch (#3)

(1) I am well aware that not all fats are created equal. Current recommendations are not based on an indiscriminate low fat or low cholesterol diet, but on far more sophisticated knowledge. The fact that you quote the Harvard Newsletter seems to indicate that you DO accept that diet contributes to atherosclerosis, but that the effect is not as simple or as strong as many people think. I agree.

(2) I stated the conclusions of systematic reviews, and you continue to cite individual studies. This is a very complex issue, and individual studies can be found to support both sides. Systematic reviews have a better chance of sorting out the truth.

I would like you to read an excellent article in American Family Physician, the journal of the American Academy of Family Physicians. It can be found online by searching for "American Family Physician Treatment of Cholesterol Abnormalities." The AFP reviews all the literature to come up with practical recommendations for family physicians. In addition, they "grade" the quality of evidence their recommendations are based on. They have given an "A" grade to all but one of the recommendations in this particular article. They base their recommendations on multiple systematic reviews that have come to similar conclusions, and they give references. They also point out that reducing dietary fat consumption may have only a small effect, and that risk reduction is greatest in those at greatest risk. They point out that non-drug interventions are more cost effective than statins.

(3) I think you have set up a straw man about what doctors really tell patients. Our recommendations depend on the overall risk of each patient, determined by looking at all the known risk factors. If you have a high LDL cholesterol but no other risk factors, most doctors would not be aggressive about lowering it. If you have a moderate LDL cholesterol but are diabetic, smoke, are sedentary, overweight, have a low HDL, have high blood pressure and all of your relatives had heart attacks in their 40s, we would treat more aggressively. We are not telling everyone to go on a low fat diet or take statins, and most of us are strongly in favor of giving the patient the facts and empowering him to make his own decisions. The AFP article states the effectiveness of statins in terms of NNT (Number Needed to Treat). For instance, in one study Pravastatin reduced all-cause mortality with a NNT of 34 for 6 years. This means that out of 34 people taking the drug for 6 years, only one would be prevented from dying, and the other 33 would have taken it with no benefit. These are the facts, and the patient has the right to know them. He may interpret them in different ways: "I don't want to take this pill if there is a 33 in 34 chance that I'm wasting my money" or "I want to take this pill if there is a 1 in 34 chance that it will save my life."

(4) The data do not show that higher blood cholesterol levels are protective against heart attacks. In individual studies it may sometimes seem that way because of confounding factors. You cited one study where elderly people with low cholesterol were more likely to die. There are several factors that can lower cholesterol levels in the elderly: chronic disease, malnutrition, cachexia, or simply poor appetite. They may indeed die sooner, but you can't draw any conclusions about cholesterol as a cause. Cholesterol might be just an innocent bystander. If a patient is chronically ill and dies of the chronic disease, their cholesterol might drop to very low levels incidental to that chronic disease, but it would not be the low cholesterol that killed them. None of this has any bearing on the fact that ALL OTHER THINGS BEING EQUAL, a higher cholesterol is associated with a higher risk of heart attack.

(5) Ravnskov's hypothesis about an illusory effect of drugs in secondary prevention is not supported by the data; the studies were properly controlled to rule out that kind of error.
 
Deutsch Answers (#4)

I guess I can't fault you for not memorizing my Amaz!ng talk, but here's a
quotation from Ravnskov which represents a summary of all papers on the subject
up to that time: “By 1998, a total of 27 studies had been published
including 34 groups (cohorts) of patients and control individuals and more than
150,000 individuals. In three of these 34 cohorts, patients with coronary disease
had eaten more animal fat than the control individuals, and in one cohort
they had eaten less. In the rest of the groups—30 in all—investigators found
no difference in animal fat consumption between those who had heart disease
and those who had not. In three cohorts the patients had eaten more
polyunsaturated vegetable oils than the control individuals, and in only one they had
eaten less.” Of course now we have this weeks JAMA paper which, based on a
study of 48,835 women over an average of 8.1 years, suggest that dietary
intervention does not affect cardiovascular disease, although it can lead to a
lowering of LDL. The fat composition of normal diets is unrelated to
cardiovascular risk, although laboratory administration of abnormal diets (large amounts
of trans fats; more cholesterol than is achievable with foods), are probably
harmful but irrelevant (I avoid trans fats but enjoy zabaglione).
As for statins, Joel Kauffman
[two websites - I can't give the urls because I haven't posted 15 messages yet]criticizes them far more effectively
than I could. Note his references to the higher all-cause death rates in the
two Lovastatin trials and the horrendous side effects.
 
The SkepDoc Answers (#4)

You did not comment on the American Family Physician article I asked you to read, you didn't respond to what I said, and you are trying to pick a fight about something different that I didn't say.

(1) I did not say that eating a high fat diet causes atherosclerosis. I said that blood cholesterol (LDL) level was an independent risk factor for heart attacks and that lowering it reduces risk.
(2) The JAMA study found that a relatively small reduction in dietary fat in a general population of women did not lower risk; it did not address a larger reduction in dietary fat in patients at risk. This is in no way incompatible with the conclusions of the systematic reviews of patients with elevated cholesterol levels as reported in the American Family Physician article.
(3) The sources you cite below are not credible sources. They are secondary sources from a biased fellow member of your International Network of Cholesterol Skeptics, and an attempt to verify the sources of his claims results in glaring discrepancies. For example, he states, "In fact, a recent meta-analysis noted side-effects in 20% of patients above the placebo rate (65% vs. 45%), and no change whatever in the all-cause death rate for atorvastatin. [8]"

This bears no resemblance to the abstract of this reference [8] which says:

"This analysis assessed the safety of atorvastatin in the 10- to 80-mg dose range using pooled data from 44 completed trials comprising 16,495 dyslipidemic patients treated with atorvastatin (n = 9,416), placebo (n = 1,789), and other statins (n = 5,290). A retrospective analysis was conducted and included treatment-associated adverse events, serious adverse events, and musculoskeletal and hepatic adverse events. Only 3% (n = 241) of atorvastatin-treated patients withdrew from studies due to treatment-associated adverse events, compared with 1% of those (n = 16) on placebo and 4% of those (n = 188) receiving other statins; the most frequently reported treatment-associated adverse events were related to the digestive system. Serious adverse events were rare and seldom led to withdrawal. Persistent elevations in hepatic transaminases to >3 times the upper limit of normal (ULN) were experienced by 0.5% (n = 47) of atorvastatin-treated patients. A persistent elevation in creatine phosphokinase (CPK) (>10 x ULN) was observed in only 1 atorvastatin-treated patient and was not associated with myopathy. The incidence of treatment-associated myalgia was low in the atorvastatin (1.9% [n = 181]), placebo (0.8% [n = 14]), and other statin (2.0% [n = 105]) groups, and was not related to the atorvastatin dose. No cases of rhabdomyolysis or myopathy were reported. Thus, the overall incidence of treatment-associated adverse events observed with atorvastatin did not increase in the 10- to 80-mg dose range, and was similar to that observed with placebo and in patients treated with other statins. Specific analysis of musculoskeletal and hepatic adverse events showed that these occurred infrequently and rarely resulted in treatment discontinuation."
 
SkepDoc,

You have *so* much more patience than I would with this kind of thing. For those who were wondering, this is *exactly* what he was doing at the meeting: never answering the questions and just going off on some tangent about some completely unrelated claim he wanted to make.

I still have the paper you gave me and actually it's been helpful in studying for an upcoming exam. So thank you! I will email you directly when I have a little more time.

Hawkeye
 
Deutsch (#5)

I forwarded to Deutsch a post from the Quackwatch Healthfraud list about the JAMA study stating (among other things), "Media reports on a recent major study on low-fat diets may mislead the public because they suggest that current advice to reduce the risk of heart disease and cancer are wrong. In fact, the study tested diet advice that is 10 to 20 years out of date."

He responded: Quotations from the JAMA paper results:
"Low-density lipoprotein cholesterol levels...were significantly reduced by
3.55 mg/dl"
"The diet had no significant effects on incidence of CHD..."
Therefore LDL level not related to incidence of CHD according to this very
large study.
What am I missing?
 
The SkepDoc Answers (#5)

What you are missing is very simple. You are mixing apples with oranges.

This study was not designed to detect a relationship between the LDL level and the incidence of CHD, and no conclusions about that relationship can be drawn from its findings. In fact, the authors assumed there WAS a relationship, and did the study to see if they could take advantage of that relationship to prevent heart attacks. The failure of their intervention only shows that that intervention failed, not that their underlying assumptions were incorrect.

It was designed to test whether giving low-fat diet advice to a large group of women (unselected for cholesterol level or risk) would make a significant difference in actual dietary habits, whether the dietary changes would significantly lower their LDL levels, and whether it would reduce their incidence of CHD events. The results showed that it made some difference in their actual diet and lowered their LDL levels by a STATISTICALLY significant (but apparently CLINICALLY insignificant) amount, and that that amount of LDL lowering did not result in a detectable decrease in CHD events - in an unselected population that included large numbers of people who were at low risk. This study simply did not have the power to show any effect. A similar study that got patients to follow a stricter diet might have enough power to show a trend towards decreased risk even in this mixed group - the study itself points out that "Stratification by attainment of lowest levels of saturated or trans fat or highest intakes of vegetables and fruits showed positive trends, with women who achieved the more optimal levels showing lower LDL-C levels and rates of CHD."

This is a very different situation from other studies that have shown a strong correlation between LDL level and risk. And other studies of patients at high risk, with high LDL levels, showing that lowering their LDL cholesterol by a clinically significant amount lowers risk. Lowering the LDL enough to see this effect in these studies required far more than the type of diet advice given to the JAMA study participants; it involved stricter dietary controls, and other interventions such as exercise, weight loss, and cholesterol-lowering medications. As the AFP article points out, diet by itself helps but is not very effective by itself.

From a study of minimal cholesterol lowering in an unselected population, you cannot draw any conclusions about the effects of a more substantial cholesterol lowering in a population at high risk.

Now, to make it perfectly clear, I do not advocate a low fat diet for everyone. I do think a diet based on the latest knowledge about "good" and "bad" fats should be part of a comprehensive risk reduction program in those at high risk.
 
From my reading, as a concerned patient, the studies show that statins do at least two things:

1) they lower cholesterol levels.
2) they lower CAD.

There is an ASSUMED correlation between those 2 effects.

But, since statins also:

3) lower inflammation.
4) prevent blood clot/platelet clumping formation

Since aspirin shows about 80% of the benefit of statins vs CAD, I'm inclined to disbelieve the cholesterol connection, in favor of statin- inflammation/clots-heart disease.

Statins have been in use for...um...20 years now? Where is the 30% drop in ACTUAL heart attack rates, that was predicted by all those studies?
 
Good point.

Statins do much more than just lower cholesterol, and yet they are no panacea (as the NNT figures make clear).

The fact remains that lowering LDL cholesterol by any means has been shown to lower risk. Since we know that from other studies with non-statin drugs and non-drug measures, my guess would be that the effect of statins may be a combination of their cholesterol-lowering and their other effects.
 
Deutsch (#6)

I have never denied that dietary lipids play a role in health. What I do deny is that cholesterol plays a significant role in cardiovascular disease. It plays no role at all as a dietary component and only a very indirect and minor role in etiology in that it is a substrate for the production of oxidized cholesterol.

I have checked Joel Kauffman's reference to the Atorvastatin review (for which I thank you). In his only reference to the review he says only "In fact, a recent meta-analysis noted side-effects in 20% of the patients above the placebo rate (65%vs. 45%), and no change whatever in the all-cause death rate for atorvastatin." Table 3 in the review gives the following figures for Placebo and Atorvastatin (all doses) respectively: Patients experiencing 1 or more adverse events 45% and 65%; Deaths (no. of patients (%)) 12 (1%) and 66 (1%).
This seems to me to be exactly concordant with his statement. What do you
find inaccurate about his statement?
 
The SkepDoc Replies (#6)

I'm not arguing that cholesterol "causes" cardiovascular disease. I can't say whether it does anything directly or is an innocent bystander that serves as a marker for something else. But it clearly IS important to cardiovascular disease, in the sense that LDL cholesterol is an independent risk factor for heart disease and lowering it reduces risk. You have denied that by claiming that a high cholesterol level actually protects against heart disease, and you have presented no evidence to support that claim.

Kauffman's statement was not technically inaccurate, but it grossly misrepresents the findings of the study.

In the first place, the way he cites the death statistics seems to imply that atorvastatin was not effective in preventing death. The study did not address effectiveness at all; effectiveness was established by the original premarketing drug trials, and the only purpose of this study was to look at safety data. One percent of patients in the trials reviewed died whether they were on atorvastatin or placebo; this supports the safety of atorvastatin but says nothing about its ability to prevent deaths as a long-term secondary preventive measure.

He chose to cite the biggest numbers he could find anywhere in the study: the percentage of patients experiencing one or more adverse events (65% vs 45%) and he chose not to mention the explanation in the text that "Although the percentage of patients experiencing any adverse event in the placebo group was lower compared with the atorvastatin group, this can be explained in part by the fact that the total duration of exposure for atorvastatin-treated patients was markedly longer than for patients who received placebo, thereby increasing the reporting period."

He could have cited the percentage experiencing treatment-associated adverse events (18% vs 15%). He could have cited the percentage of serious, nonfatal adverse events (10% for atorvastatin vs 8% for placebo), or of the serious, nonfatal adverse events that were treatment associated (which actually favored atorvastatin at 6% for placebo and less than 1% for atorvastatin). He could have presented the data from figure 1 showing that for two of the four atorvastatin doses tested, the total adverse events were fewer than with placebo, and for three of the four dosages, the treatment associated adverse effects were fewer for atorvastatin than for placebo. He could have quoted the study's conclusion that "Overall, the incidence of adverse events for atorvastatin-treated patients, patients receiving other statins, and patients receiving placebo was low, and for the atorvastatin group, the incidence of these events was lower than that reported in the original new drug application." He could have mentioned the study's conclusion that "Specific treatment associated adverse events in the atorvastatin and other statins groups occurred infrequently." and that "Atorvastatin was found to be well tolerated overall."

He could have quoted from the last paragraph of the study: "The recently published Anglo-Scandinavian Cardiac Outcomes Study (ASCOT) further supports the safety of atorvastatin.21 In ASCOT, 10,305 patients with mild to moderate elevations of total cholesterol were randomized to atorvastatin 10 mg (n  5,168) or placebo (n  5,137) for a median treatment duration of 3.3 years. Only 1 nonfatal case of rhabdomyolysis was reported in an atorvastatin-treated man who had a very high alcohol intake and a recent febrile illness. The number of serious adverse events and rates of elevations in hepatic transaminases did not differ between placebo- and atorvastatin-treated patients.21"

His characterization of the study is as blatantly biased as you can get.
 
Where Is Deutsch?

A week ago, Deutsch posted a message to another thread. See:
http://www.internationalskeptics.com/forums/showthread.php?t=50839

In that message, he said I had not responded about the JAMA study, when in fact I had responded in detail the day before. See messages #27 and #28 above.

I don't understand what he hoped to gain by posting this false statement. I have not heard from him in a week now, and he has not answered my comments about the JAMA study nor my comments about Kauffman's misrepresentation of scientific data - see #32 above.
 
He's Back! Deutsch (#7) with SkepDoc Reply

(1) I told you that for about two weeks I'd be busy getting out a
mycological publication which I edit. It will go to the printer this week-end.
(2) When I said that you didn't reply to my comment on the JAMA study, I was referring to my asking you where I went wrong in concluding that, since the study showed that the treatment group had significantly lower LDL levels
than the control group but not significantly lower incidence of cardiovascular
disease, this suggested that, if there is a relationship between LDL levels
and cardiovascular disease, it is so small that a study this large couldn't
detect it. (I am expressing it a little better here.) I have no record or recollection of your answering this.
(3) When a paper was published showing that aspirin use was correlated with a reduced risk of fatal colon cancer (N. Engl. J. Med. 1991;325:1593-6) and I pointed out that this was of interest only if one were concerned about dying of a particular cause, but that overall mortality would be of greater concern (N. Engl. J. Med. 1992;326:1289), the original authors conceded the point. When Joel Kauffman pointed out the failuere of Atovastatin to affect total mortality and I pointed out that he precisely quoted a paper sponsored by a pharmaceutical company to make this point, you changed the subject and made an ad hominem attack on him. If Jim Jones were to publish a paper showing that his concoction of Kool-Aid and cyanide gave his followers 100% protection against both cancer and heart disease, would you criticize me for pointing out that this treatment suffers from the defect of raising total mortality?

Marshall


(1) I'm waiting.
(2) I have re-sent the message I sent you about the JAMA study on Feb. 9th in which I told you where you went wrong. Incidentally, your current
restatement constitutes a backing off from your original claim that the JAMA
study was evidence that LDL level is not related to incidence of CHD. In
fact, your restatement incorporates part of my critique of your original
statement. Are you sure you didn't read my message, incorporate it into your
thinking, and then forget that you had read it?
(3) You are mixing apples and oranges again.
- You say overall mortality is of greater concern than mortality from a
single disease. True but irrelevant since statins reduce overall mortality
as well as cardiovascular mortality.
- On Kauffman: I did not change the subject. Actually you were the one
who changed the subject, since I had said lowering LDL cholesterol lowered
risk and had simply mentioned statins as one option to help accomplish that.
You responded that statins were harmful (I believe you used the word
horrendous) and offered Kauffman's webpage as evidence. I showed why that evidence was unacceptable.
- And that was not an ad hominem attack; it was a critique of
Kauffman's misrepresentation of data, not of Kauffman himself.
- He did not "precisely quote" the paper; he took "precise data" out
of context to support his own unjustified conclusions, which were not
compatible with those of the paper itself. I gave only one example of the
misrepresentations on the Kauffman webpage. I could have analyzed the whole thing but I didn't think it was worth my time. For what it's worth, here's one more quick example of his many errors. He makes the unqualified
statement that using statins for primary prevention increases mortality by
1%. To support this, he cites a paper that was actually designed to
determine whether patients with high cholesterol but low overall risk of
heart disease should take statins. The 1% referred to a "suggestion" from a
"regression line" that the risks of statins would outweigh the benefits by
that much for patients with a CHD event risk less than 13% over 10 years. In other words, statins are probably not indicated for patients at low overall
risk of heart disease. Kauffman gives the impression that they are not
indicated for high risk patients either, but that's the oppostie of what the
study actually showed.
- Your Kool-Aid example is pretty silly. It might make some sense as
an analogy if statins prevented heart attacks while increasing mortality,
but the data clearly show that statins reduce cardiovascular events,
cardiovascular mortality AND all-cause mortality. Don't take my word for it.
Try searching PubMed for statins and mortality. You will find a wealth of
data, individual studies and meta-analyses, showing that statins reduce
overall mortality rates. The first few I found quoted reductions of anywhere
from 15% to 39%.

Harriet
 
Deutsch Answers about the JAMA study

Good answer, but the fact remains that they showed no effect, and it is
conjecture that an effect on cardiovascular disease would be shown if they could
have lowered LDL more. If smaller studies showed this effect and this large
study did not, that is something to think about. Incidentally, what do you think
of the fact that the study seems to put trans fats and saturated fats in the
same boat? I believe that the evidence against trans fats is very convincing,
but that the evidence against saturated fats is extremely flawed. Ravnskov
demonstrates extensive flaws in the evidence against saturated fats, and Enig, who
has studied fats and diet extensively, believes that the bad reputation of
saturated fats is entirely owing to early studies having failed to distinguish
them from trans fats. There is very good evidence for benefits arising from
butyrate that makes it to the large intestine.

Marshall

Of course! The JAMA study used outdated diet advice. We are learning. First cholesterol, then saturated fats, then trans fats... we are gradually narrowing it down, and I'm sure there is much more to learn. Today's advice will be refined or superseded by next year's research.

Is there really a "myth" about cholesterol, or just people who are not up to date? Maybe the general public is remembering old information, and maybe some doctors are slow to change, but the medical providers I know keep up with the latest research and understand the true complexities of these issues. Sure there were problems with some of the earlier information, but science marches on and corrects its own errors. Trans fats have been accepted to the point that they're going to be put on food labels. Where's the problem?

Harriet
 
The Debate Continues: Today's Exchanges

[MD] = Marshall Deutsch
[HH] = Harriet Hall, MD

[MD] I'm going through the undeleted e-mails from you to be sure I'm answering
all your questions. Anent experimental atherosclerosis in animals, I checked a
number of the PubMed responses to "experimentally induced atherosclerosis" and
found none that described producing it solely by feeding any diet which
could be naturally available. Various knockout rats (deliberate unnatural gene
dificiencies) are used, and, even in rats, cholesterol at 2% of the diet does
it, but as I pointed out, a diet of pure egg yolks contains less that 1.5%. I
believe that a rat could live on a diet of hen eggs, but that it would eat a
lot of egg white too

[HH] In your article in Skeptic magazine you said, "Also troubling was the inability of researchers to demonstrate an effect on fat deposition in arteries (atherosclerosis) of experimental animals by feeding them cholesterol or fats." As stated, this was an error and you already admitted that you mis-spoke. A high fat diet HAS been used to experimentally produce atherosclerosis in animals. Whether or not this has any significance for human disease is another matter entirely. But there is certainly nothing there to make us question the LDL cholesterol/atherosclerosis connection.

[MD] I never denied that cholesterol is a factor.

[HH] You said that it was a minor factor, you predicted that proper analysis would show its contribution was "vanishingly small," and then you contradicted yourself by saying also claiming that cholesterol was protective against heart disease.

[MD] Here is a quotation from my paper in SKEPTIC which you have so unfairly criticized and which this e-mail
suggests you have not read:
"Serum cholesterol levels have been touted for a long time as risk factors
for heart disease, but they are only one among literally hundreds (really!) of
such factors, and serum cholesterol level is a much poorer indicator of the
likelihood of getting clogged coronary arteries than, for instance, are plasma
homocysteine levels, or how one does on an exercise-tolerance tests."

[HH] This is demonstrably untrue. In the first place, I doubt that you can list "hundreds" of risk factors, unless you are thinking of listing the individual DNA of everyone with a family history of heart disease. In the second place, LDL cholesterol level is such an important, clear-cut risk factor that it has been accepted as a screening test for cardiovascular risk; plasma homocysteine levels and exercise tolerance tests have not been accepted for screening the general population, although they are useful clinically.

[MD] and heres a quotation about statins from the same paper:
"When the statin drugs (cholesterol-lowering agents) were introduced, it
turned out that they did lower the chance of a second heart attack, but that
they did so even in subjects with already low values of cholesterol."

[HH] So what? The point is that all cholesterol-lowering drugs have been shown to reduce risk. It is quite possible that statins also reduce risk by other mechanisms, but it seems clear that reducing LDL cholesterol level by any means, including statins, reduces risk. For that matter, some researchers think that lowering an already low cholesterol may reduce a low risk of cardiovascular events to an even lower level, and that this effect may be particularly important in people who already have atherosclerosis.

[MD] Joel Kauffman accurately quoted data in a table in the review. The abstract
ignores this important table which shows no effect of atorvastatin on total
mortality.

[HH] Did you not understand my critique? The study was of adverse effects of atorvastatin in short-term trials, and the study was only powered to show that atorvastatin did not increase mortality. The study was not designed to show whether atorvastatin could decrease mortality; no conclusions about that can be drawn from this study, and other studies have conclusively shown that atorvastatin does decrease mortality. Kauffman "accurately quoted the data" but the data didn't mean what he implied it did.

[MD] Whether the benefits outweigh the risks is a judgment call, but I note that
this abstract ends with "The
clinical benefits of preventing vascular events, myocardial infarction,
stroke, and need for revascularization outweigh the low rates of adverse
events associated with high-dose statin therapy in high- and
intermediate-risk patients."

[HH] This clearly shows that the study authors recognized that atorvastatin does more good than harm, that the 1% mortality figures in the study were not pertinent to the risk/benefit question, and that Kauffman's use of the 1% mortality figures constitutes inappropriate data mining and misinterpretation.

[MD] and that statins are being prescribed at a rate
which suggests that almost the entire patient population is at high or
intermediate risk.

[HH] This is where I agree with you, and I already said so. Statins are useful for high risk patients, but they may do more harm than good for low risk patients. They are being overprescribed, and they are being prescribed in lieu of diet, exercise, and other treatments that should be tried first. When prescribed appropriately, they save lives.

[MD] Very few studies have been done on women. Here's a meta-analysis:
Treatment of hyperlipidemia in women.

_Walsh JM_
(http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&itool=pubmed_Abstract&term="Walsh+JM"[Author]) , _Grady D_
(http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Search&itool=pubmed_Abstract&term
="Grady+D"[Author]) .

[HH] Yes, it says, "There is no evidence from primary prevention trials that cholesterol lowering affects total mortality in healthy women, although the available data are limited. Limited evidence suggests that treatment of hypercholesterolemia in women with coronary disease may decrease CHD mortality. Future research should
address the role of dietary and other nondrug treatment of hypercholesterolemia in
women at high risk for CHD." I agree. Healthy women probably do not benefit from treatment. May I remind you, women are already at lower risk because male sex is one of the risk factors for cardiovascular disease, and hypercholesterolemia is only one of many risk factors. This study was of healthy women and did not imply that lowering the cholesterol level of high risk women was not of benefit.

[MD] Also, since older women with higher cholesterol levels have lower mortality
rates than older women with low cholesterol rates (See Schupf N, Costa R,
Luchsinger J, et al. (2005). Relationship Between Plasma Lipids and All- Cause
Mortality in Nondemented Elderly. Journal of the American Geriatrics Society
53:219-226.), I don't understand why one would want to treat their hypercholesterolemia.
[HH] This study concluded, "Low cholesterol level is a robust predictor of mortality in the nondemented elderly and may be a surrogate of frailty or subclinical disease." In other words, they thought the low cholesterol levels could be due to concurrent noncardiovascular disease. It is unreasonable to draw any conclusions from this study about the advisability of treating high LDL cholesterol levels in a patient at high risk of cardiovascular disease. There is evidence that lowering a high cholesterol reduces cardiovascular risk, and there is certainly no evidence that raising a low cholesterol level reduces mortality.
============================================
[HH] I agree with you that diet is not as important as many people think, that trans-fats are more important than cholesterol, that diet is not very effective by itself in lowering risk, that recommending a low fat diet for the entire population and especially for all children is probably not justified, and that statins are over-prescribed.

But I insist that:
(1) LDL cholesterol is an independent risk factor for cardiovascular disease
(2) All other things being equal, lowering LDL levels lowers risk of cardiovascular events
(3) All other things being equal, lowering LDL levels in patients who already have heart disease reduces risk of death.
(4) Treatment should be individualized, based on overall risk rather than on any single risk factor.

Can you seriously dispute any of these claims?

You have claimed that elevated LDL protects against cardiovascular disease. There is no credible evidence that is true.
 
Deutsch Concedes 2 out of 4

[MD] = Marshall Deutsch, Cholesterol Skeptic
[HH] = Harriet Hall, MD


[HH] But I insist that:
(1) LDL cholesterol is an independent risk factor for cardiovascular
disease
(2) All other things being equal, lowering LDL levels lowers risk of
cardiovascular events
(3) All other things being equal, lowering LDL levels in patients who
already have heart disease reduces risk of death.
(4) Treatment should be individualized, based on overall risk rather than
on
any single risk factor.

Can you seriously dispute any of these claims?

[MD] A nice narrowing down of the disagreements. No reasonable person
could
disagree with (4), and making all other things equal in order to test (2)
would be
a formidable task, since there are so many other interrelated risk factors.
However, I can seriously dispute the claims.

[HH] I don't think you can. You have conceded (4)and you have already
conceded (1) by saying in a previous e-mail that you never denied
cholesterol was a factor. That leaves (2) and (3). The bulk of the evidence
from the medical literature supports (2) and (3). The articles that you cite
do not constitute acceptable evidence that points (2) and (3) are wrong.

Articles can be found to support any point of view. Good medical science
does not cherry-pick articles to support a hypothesis; it examines ALL the
evidence, rates the quality of that evidence, and reaches a consensus about
recommendations for treatment. That is what American Family Physician has
done in the article I asked you to read:
http://www.aafp.org/afp/20050315/1137.html. They reviewed the literature and
came up with the table showing the "strength of recommendations" and they
found that high quality evidence supported these claims:
For primary prevention, diet "may lead only to a small reduction in
cardiovascular events" and lowering cholesterol with statins "decreases
cardiovascular events in patients with elevated cholesterol levels, although
a decrease in cardiovascular and all-cause mortality has not been
demonstrated." For secondary prevention, statins "decrease cardiovascular
events, cardiovascular mortality and overall mortality."

I think these statements accurately reflect the weight of all the evidence
and the consensus of large numbers of conscientious scientists who have read
all the papers Ravnskov cites, who have seriously evaluated the evidence to
the best of their ability and who have no axe to grind. The consensus is
constantly evolving as new evidence becomes available. You belong to a
"Cholesterol Skeptics" organization; the very idea is ludicrous. Scientists
are already skeptics. Science does not progress by activist organizations.
Science is a self-correcting system that progresses by publication and peer
review. We did not need a "Gastric Acid Skeptics" organization for medicine
to recognize the role of bacteria in causing ulcers. The truth always
becomes apparent with time.

Some of what the Cholesterol Skeptics say is true and is information that
doctors already take into consideration; some of it is a misleading
distortion of what the literature actually says. None of it is news to
science.
 
What Do These Studies Really Show?

[MD] = Marshall Deutsch, Cholesterol Skeptic
[HH] = Harriet Hall, MD

[MD] Hecht and Harman (Am. J. Cardiology. August 1, 2003:92(3):334-6)
used
electron beam tomography to study the changes in calcified plaque
progression
in 182 individuals over 1.2 years of treatment and found that (as in the
recent study of middle-aged women) despite lower cholesterol levels
(achieved by
drugs, not diet this time), there was no difference in the development of
atherosclerotic plaque. They concluded that “with respect to
LDL-cholesterol
lowering, lower is better’’ is not supported by changes in calcified
plaque
progression.” Since they studied what they referred to as "more versus less
aggressive cholesterol-lowering therapy" I would appreciate your not
responding
that this reflected that they produced a statistically significant but not
a
clinically significant LDL lowering. If a clinically significant lowering
can
never be achieved, that shows that there is none.

[HH] Your citation does not support your hypothesis and your comments are
incoherent. "If a clinically significant lowering can never be achieved,
that shows that there is none."??? This study does not show that lowering
cholesterol does not decrease the number of cardiovascular events. It did
not measure cardiovascular events as an endpoint. It was designed to see if
measurable decreases in plaque progression could be detected, and it was
done by researchers who actually support the use of statins. It didn't even
have any controls with patients who were not on statins; the authors already
knew from previous studies that the rate of progression would have been
greater in patients whose cholesterol was not lowered at all. In a similar,
later article, the same authors reported, "These findings indicate that
subclinical atherosclerosis may respond to drug therapy in a fashion
parallel to that of clinical disease demonstrated in the Heart Protection
Study." The American Journal of Cardiology
Volume 93, Issue 1 , 1 January 2004, Pages 101-103.

[MD] Studies with statins show the lack of a relationship between
cholesterol
level and atherosclerosis. If high cholesterol were the cause of heart
disease, then the greatest preventive effects from statins would be seen in
these
trials among those with the highest cholesterol levels and in patients
whose
cholesterol levels were lowered the most. An exhaustive review by Ravnskov
(Quarterly J. Med. 2003;96:927-934) shows that this has never been the case

[HH] Exhaustive reviews by multiple unbiased scientists show that it IS the
case, and new studies continue to come out showing that the more you lower
cholesterol, the more you reduce the rate of heart attacks and deaths. For
instance, see
http://www.ncbi.nlm.nih.gov/entrez/...uids=16356805&query_hl=16&itool=pubmed_docsum
This large study of secondary prevention, published in Dec. 2005, showed
that patients whose cholesterol was reduced more aggressively had 16% fewer
heart attacks and deaths than those whose cholesterol was reduced less
aggressively.
.
[MD] An (admittedly short-term) study of the effects of atorvastatin on
plaque composition ( _J Intern Med._ (javascript:AL_get(this, 'jour', 'J
Intern
Med.');) 2006 Mar;259(3):267-75) showed no effect on lipid content of
plaques,
which, together with the fact that the effects of statins on
atherosclerosis
seem to be independent of their effects on LDL, suggests to me that the
effects of statins on atherosclerosis are mediated through other of their
pleiotropic effects.

[HH] Stick to the main issue. I am not claiming statins affect the lipid
content of plaques; I am not claiming statins don't have benefits other than
lowering cholesterol. The details of how they work may not be entirely
clear, but it is entirely clear that they do are effective drugs for primary
and secondary prevention. I am simply agreeing with the consensus of medical
scientists that statins lower LDL cholesterol levels and lowering LDL
cholesterol levels decreases risk of cardiovascular events.

[MD] Here's another reference showing no effect of statin treatment on
total mortality
of women: _JAMA._ (javascript:AL_get(this, 'jour', 'JAMA.');) 2004 May
12;291(18):2243-52..

[HH] Stick to the main issue. The fact that lowering lipids lowers risk in
general does not mean it does so in every particular case or in every
subgroup. We know women are generally at lower risk and we know the benefit
of cholesterol lowering is greater in men. That is why we must look at
overall risk and prescribe individually. This JAMA study concluded, "For
women without cardiovascular disease, lipid lowering does not affect total
or CHD mortality. Lipid lowering may reduce CHD events, but current evidence
is insufficient to determine this conclusively. For women with known
cardiovascular disease, treatment of hyperlipidemia is effective in reducing
CHD events, CHD mortality, nonfatal myocardial infarction, and
revascularization, but it does not affect total mortality."

If you think this proves the failure of statins to be of benefit to women,
you are clearly wrong. Reducing events is a benefit even if total mortality
is unchanged. This study addressed women as a class and could not conclude
anything about the benefits to individual women at high risk. The results
are consistent with the hypothesis that statins may have the greatest
benefit in women at highest risk, and that the risks may outweigh the
benefit in women at lower risk.

You are being inconsistent. You agreed with me that treatment should be
individualized based on overall risk, and yet you are implying statins
should be categorically avoided in women.

Just as you were inconsistent when you admitted LDL cholesterol was a risk
factor for heart disease and then claimed it was protective against heart
disease. You can't have your cake and eat it too.
 
Practicing Medicine Without A License?

[MD] = Marshall Deutsch, Cholesterol Skeptic
[HH] = Harriet Hall, MD

[MD] Present at my Amaz!ng talk were two women who had been prescribed
statins: Linda, Randi's right-hand woman and Susan Dennett, Daniel's wife.
I had
hoped to convince them as to the unwisdom of their prescriptions, and would
have welcomed any help I could have received from others present familiar
with
the literature showing the failure of statins to be of benefit to women.

[HH] As for others present familiar with the literature, we were there, but
we had no wish to help you in your misguided quest. I am familiar with the
literature showing the "failure" of statins in women, as well as the
literature showing the benefit of statins, and in my judgment (which is
consistent with the consensus of medical experts) there are good reasons to
prescribe statins for individual women at high overall risk.

As for convincing Linda and Susan, I think the idea is totally
inappropriate. I am licensed to practice medicine, but it would be unethical
for me to advise a patient without knowing any more than the fact that they
were taking a certain prescription medication. You are not licensed to
practice medicine, and you know nothing about the risk status of those
patients or the reasons their doctors chose to prescribe statins. Saying
their prescriptions are "unwise" amounts to practicing medicine without a
license. It is illegal, inappropriate, unethical, and unfair. Wanting them
to stop taking their statins puts you in the same category as anti-vaccine
activists who talk people out of immunizations on the basis of their biased
selection of evidence. Their advice can kill people, and yours might too.
The most you can ethically do is point out information that their doctor may
not have seen and ask them to discuss it with him.
 
Deutsch Sends for Reprint

Deutsch is sending for a reprint of an article I cited two messages ago. I answered him:

What's the point? I only mentioned this study as one of a great many that support the consensus that greater lowering of cholesterol leads to greater benefit. If you read the entire paper, there are two possibilities: either you will find flaws in it (in which case I could cite lots more papers) or you will agree that its conclusions are valid. Even if you agree that it is valid, no one study in isolation means much; it is the accumulation of many confirming studies that convinces.

The point is that a great many excellent scientists have looked at all the published data INCLUDING all the data that Ravnskov relies on and have come up with considered conclusions; and they reconsider those conclusions constantly as new evidence emerges. Ravnskov has looked at the published data and come up with a different conclusion. It is clear that he is reading the literature to prove a point rather than to objectively balance the pros and cons to find out the truth. Since science is a collaborative and self-correcting enterprise with a proven success record, I can trust the combined brainpower of the world's scientists far more than I can trust any one individual's judgment.

I think you have been conned. You have been reading secondary sources like Ravnskov and Kauffman that are partly right and partly wrong, that misinterpret and misrepresent some of the evidence to prove a point. I suggest you try reading analyses by unbiased scientists who have no point to prove but are sincerely trying to find out the truth. Like the American Family Physician article (which I keep citing and which you have not yet commented on) and the studies that its recommendations are based on.
 
You must log in or register to reply here.

Back
Top Bottom