November 10, 2008

Should You be Taking a Statin? What the Crestor Study Really Found.

A new study has been reported all over the media as proving that just about everyone should be taking Crestor because of the way it lowered the incidence of stroke and heart attack.

As usual, when we read the actual study, we find that it said something very different.

The entire study as published in The New England Journal of Medicine is available for free this month here:

Rosuvastatin to Prevent Vascular Events in Men and Women with Elevated C-Reactive Protein

Here is a brief summary of how the study was conducted.

The researchers found a group of men over 50 and women over 60 years old who had normal cholesterol but whose C-reactive protein was elevated. C-reactive protein believed to be a marker for inflammation, particularly inflammation in your arteries.

Take a look at this table to see the other characteristics of the participants in this study:

Baseline Characteristics of the Study Group

Several things leap out that were not made clear in the popular press reporting of this study.

First of all, they were older people. The median age was 66. Secondly, though they had normal cholesterol they had other significant risk factors for heart disease and stroke:

They were overweight. The median BMI was 28.3. That would translate into a weight of 165 lbs for the average 5' 4" woman or 192 lbs for the average 5' 9" male.

More importantly, they had higher than normal blood sugars. The median A1c was 5.7% which I have been told by an endocrinologist is high enough to suggest undiagnosed diabetes.

The interquartile range(a measure of the area in which most values concentrate) for this study group was 5.4% - 5.9%. These figures alone tell us that these people had a higher than normal risk of heart attack since several large studies have shown that the risk of heart attack rises in lock step with the rise of A1c as it goes over 4.7%. At 5.7% the risk is already significantly higher than at true normal.

Finally, These people had elevated blood pressure--which is another well known pointer to early heart disease. The median blood pressure was 134/80. Doctors now believe 120/80 is the upper bound of normal.

You need to also look at who was excluded from the study: Women taking hormone replacement therapy, people with any indicator of liver or kidney abnormality, people with high blood pressure, people with thyroid disease, people with autoimmune disease and people with a history of alcoholism or drug abuse. These people were excluded because, among other things, earlier studies have shown Crestor might be very dangerous for them.

The study participants were then put on Crestor, a newer and very expensive statin drug. Note that the study was run exclusively by people with strong financial links to the company that makes Crestor. (More about this at the bottom of this post.)

After two years the study was stopped because it found that people in this group taking Crestor had about half the risk of "end points" than people taking a placebo. These "end points" included were fatal and non fatal myocardial infarction [heart attack], fatal and nonfatal stroke, arterial revascularization [putting in a stent], hospitalization for unstable angina, or confirmed death from cardiovascular causes.

But lets look a bit closer at the statistics. When ever anyone uses the "Risk" statistic, you can be sure they are trying to amplify a very small statistic into a larger one. Here's what the actual statistics came up with:

There were a total of 393 "end points" reported among the 17,802 people involved in the study. There were a grand total of 109 more "end points" among the people taking placebo compared to those taking Crestor.

In this group of 17,802 people involved in the study, half of whom were taking Crestor, there were 99 heart attacks or 5 per 1,000. Of these 15 were fatal or 8 in 10,000. There were 97 strokes of which 9 were fatal, a rate of 5 in 10,000. These numbers are for the group as a whole.

The most frequent endpoint was getting a stent put in, which occurred in 202 cases or about 1% of the group. When considering this statistic, note that Dr. Davis the cardiologist who writes the Heart Scan Blog believes that this procedure is dramatically overused because there is a strong economic motivator involved: this kind of procedure is the primary way that cardiologists make a living.

It is also worth noting that the way these statistics were collected one person may be recorded as having several "end points" though in fact they had only one incident or hospitalization. In fact, it is worth noting that the way that these end points are reported is so confusing, with so much overlap that it is very hard to know exactly what is being reported. The total given for "primary end points" is much lower than the total of all the various end points listed.

So right away it is important to note that the actual likelihood of death or stroke in this group of middle aged people with clear cardiovascular risk factors, while regrettable, was quite small.

But no one wants to experience any of these "end points." So let's look more closely at what happened when people took 20 mg of Crestor every day for two years.

In fact, though the way the data was reported was carefully arranged to obscure this finding, it appears that while there were indeed slightly less than half as many heart attacks within the group taking Crestor, there were more fatal heart attacks in the group taking Crestor.

The way that this is reported is thus:

Nonfatal myocardial infarction: Crestor 22 Placebo 62
Any myocardial infarction: Crestor 31 Placebo 68

Subtract "Nonfatal myocardial infarctions" from "Any myocardial infarctions" and you get Fatal Myocardial infarctions, a statistic which is NOT reported in the list of "end points." But simple math gives us the information that there were 9 fatals in the Crestor group as opposed to 6 in the placebo group.

Okay, so it looks like taking Crestor cut down the incidence of heart attack by 37 cases, but did not cut down on the deaths from heart attack. In fact, the proportion of heart attacks that were fatal in the Crestor group was 29% compared to the 9% that were fatal in the placebo group. Why is this not noted by the authors of the study?

For stroke, there were 31 fewer strokes in the group taking Crestor and 3 fewer deaths. So Crestor looks a bit more helpful here.

In total there were 49 more deaths from any cause in the group taking placebo than in the group taking Crestor, though this statistic is not explained. But it is worth pointing out that the death rate for the study group as a whole (2.5%) was better than the death rate for this age group reported by census data and identical to the death rate for the age group being studied according to CDC statistics.

But here's the bad news buried in this study: Both A1c and the incidence of diagnosed diabetes rose in the group taking Crestor.

There were 54 more cases of diabetes diagnosed in the Crestor group than in the placebo group. This is brushed off as barely significant by the authors, though, in fact, it is greater than the difference in heart attack, stroke, etc.

You can examine the side effect statistics here.

So what does this study tell us about Crestor?

1. Crestor makes a small difference in the cardiovascular health of people with high CRP.

2. This study confirms the suspicion that cholesterol levels are a red herring and that the impact of statins, where they have an impact, is on lowering inflammation in the cardiovascular system.

3. Crestor is most effective against microvascular problems--most notably stroke, than heart attack death, which makes sense if the main thing they do is decrease inflammation in small arteries most likely to lead to stroke.

4. Crestor lowers the number of heart attacks, but not the number of heart attack deaths.

5. Crestor appears to raise the incidence of diabetes in those taking it. The mechanism of this is not known.

6. The study did not find an increase in muscle or liver problems, but it carfully screened out those who most likely to develop these problems. Crestor has been linked to severe muscle and liver damage. Read this FDA Advisory before you take Crestor and make sure your doctor does the same screening that this study group was given:

7. The study did not investigate or report the incidence of cognitive decline in this group. Statins have been linked to cognitive decline (i.e. memory loss and other signs of dementia).

Some information about statins and memory loss can be found here:

and here: Blood Sugar 101: Other Dangerous Drugs

8. This study did not look at the impact of a much cheaper generic statin on the same endpoints. It was run and reported by people with strong financial links to AstraZeneca, the company that makes the expensive, patented Crestor. It is very possible that the benefits found with Crestor may also be achieved by taking the inexpensive generic statin, Simvastatin (Zocor).

9. This study was run by people with very strong financial links to the manufacturers of Crestor and who have additional financial motives to promote CRP testing. The way that they obscured the heart attack death statistic should make you careful about trusting their conclusions. See the list of conflicts of interest reproduced below.


1. Get your CRP tested before you let any doctor talk you into taking a statin. Your level of CRP is probably a much better indicator of heart disease than your cholesterol levels. If your CRP is dramatically high, Crestor or a much cheaper generic statin may be helpful, but balance the possible impact the statin may have on your blood sugar control and cognitive function against the modest benefits it may present.

2. Before you take Crestor or any other statin, insist your doctor run a full Liver Panel and repeat this test every few months.

3. Stop taking Crestor or any other statin if you experience muscle pain.

4. Do not take Crestor or any statin if you are over 65 and have a family history of dementia or any symptoms of early memory loss.

5. Given the very strong link between A1c and heart disease it is possible that you might achieve similar benefits by lowering your blood sugar via a lower carbohydrate diet and exercise. Checking your CRP after six months of adopting such a regimen might help you evaluate whether or not you need a statin.


Here is the disclosure paragraph published with this study. The names listed are those of the study's authors:

Dr. Ridker reports receiving grant support from AstraZeneca, Novartis, Merck, Abbott, Roche, and Sanofi-Aventis; consulting fees or lecture fees or both from AstraZeneca, Novartis, Merck, Merck–Schering-Plough, Sanofi-Aventis, Isis, Dade Behring, and Vascular Biogenics; and is listed as a coinventor on patents held by Brigham and Women's Hospital that relate to the use of inflammatory biomarkers in cardiovascular disease, including the use of high-sensitivity C-reactive protein in the evaluation of patients' risk of cardiovascular disease. These patents have been licensed to Dade Behring and AstraZeneca. Dr. Fonseca reports receiving research grants, lecture fees, and consulting fees from AstraZeneca, Pfizer, Schering-Plough, Sanofi-Aventis, and Merck; and Dr. Genest, lecture fees from AstraZeneca, Schering-Plough, Merck–Schering-Plough, Pfizer, Novartis, and Sanofi-Aventis and consulting fees from AstraZeneca, Merck, Merck Frosst, Schering-Plough, Pfizer, Novartis, Resverlogix, and Sanofi-Aventis. Dr. Gotto reports receiving consulting fees from Dupont, Novartis, Aegerion, Arisaph, Kowa, Merck, Merck–Schering-Plough, Pfizer, Genentech, Martek, and Reliant; serving as an expert witness; and receiving publication royalties. Dr. Kastelein reports receiving grant support from AstraZeneca, Pfizer, Roche, Novartis, Merck, Merck–Schering-Plough, Isis, Genzyme, and Sanofi-Aventis; lecture fees from AstraZeneca, GlaxoSmithKline, Pfizer, Novartis, Merck–Schering-Plough, Roche, Isis, and Boehringer Ingelheim; and consulting fees from AstraZeneca, Abbott, Pfizer, Isis, Genzyme, Roche, Novartis, Merck, Merck–Schering-Plough, and Sanofi-Aventis. Dr. Koenig reports receiving grant support from AstraZeneca, Roche, Anthera, Dade Behring and GlaxoSmithKline; lecture fees from AstraZeneca, Pfizer, Novartis, GlaxoSmithKline, DiaDexus, Roche, and Boehringer Ingelheim; and consulting fees from GlaxoSmithKline, Medlogix, Anthera, and Roche. Dr. Libby reports receiving lecture fees from Pfizer and lecture or consulting fees from AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Merck, Pfizer, Sanofi-Aventis, VIA Pharmaceuticals, Interleukin Genetics, Kowa Research Institute, Novartis, and Merck–Schering-Plough. Dr. Lorenzatti reports receiving grant support, lecture fees, and consulting fees from AstraZeneca, Takeda, and Novartis; Dr. Nordestgaard, lecture fees from AstraZeneca, Sanofi-Aventis, Pfizer, Boehringer Ingelheim, and Merck and consulting fees from AstraZeneca and BG Medicine; Dr. Shepherd, lecture fees from AstraZeneca, Pfizer, and Merck and consulting fees from AstraZeneca, Merck, Roche, GlaxoSmithKline, Pfizer, Nicox, and Oxford Biosciences; and Dr. Glynn, grant support from AstraZeneca and Bristol-Myers Squibb. No other potential conflict of interest relevant to this article was reported.


Anonymous said...

There are a lot of safer, cheaper ways of reducing CRP, e.g.:

-high-dose fish oil
-vitamin D supplementation (with appropriate 25[OH]vit D blood tests) to achieve optimal range
-reduction in body weight
-reduction in BG levels via reduction in carbohydrate intake

I started out 4 years ago with:

- HbA1c = 13
- hsCRP = 3.4

After seeing Dr. Richard Bernstein in NY and getting my diabetes on track, plus using information I learn from by joining Dr. Davis' Track Your Plaque program, I currently have:

- HbA1c = 5.1
- hsCRP = 0.7

It can be done with some effort and knowledge. And without Crestor.

Anonymous said...

This analysis makes sense to me:

Perhaps reducing LDL is not the principle efficacy for Statins, especially in light of the recent failed Vytorin study.


Anonymous said...

I was so happy to see your post on this topic. The hype is intense! I can only imagine how many healthy people are now convinced they should be on Crestor.

My favorite was at the very bottom:
"No other potential conflict of interest relevant to this article was reported."


Anne said...

Thanks for your thorough analysis of this study. It is not new that statins can recuce CRP. A PubMed search will bring up references about statins and CRP as far back as 1998.

Ten years ago my hs-CRP was 13. It is now 3. I cannot take statins as they cause severe muscle pain. I am not sure how I lowered this number other than supplements and an anti-inflammatory diet. Hopefully it will continue to fall.

It is worrisome that doctors who will be prescribing Crestor to people who were excluded from the study.

The study does not say what was in the placebo. Are all placebos truly inert?

Anonymous said...

I was handed a piece of paper

haven't finished reading through it yet

Simvastatin dropped my LDL without affecting my trigs or HDL which were seriously bad, I'm still taking it due to a lack of effects other than doing exactly what it says on the tin, however I am considering dropping it for a month prior to my next bloods as I want to see my unmedicated lipids.

The reason? Low carbing and nailing my BG doubled my HDL and brought my lipids down to ten percent of what they were. My latest (controversial!) experiment revealed that I now convert saturated fats into HDL in the absence of dangerous levels of carbs and hence trigs.

I would love to see what effect my regime has had on CRP but here I cannot get it tested "unless I have an inflammatory condition" (sigh)

Bottom line is that statins work for me, but on the wrong thing. Diet a la Rich, above, probably works on the right thing, but I'll never know.

Alan said...

Thanks once again Jenny. I was going to do my own analysis. You not only saved me the trouble, you did it better than I ever could have.

Great job.

Cheers, Alan

Jenny said...


It is always worth having a look on your own, as you may see something that escaped me, or which has particular relevance to you which might not have caught my interest.

Anonymous said...


I'm a 32 year (since diagnosis) juvenile onset diabetic. I write for and have been doing research on this topic. So I want to thank you. You did it all for me.

I am already anti-statin and had read the "raises blood sugar" part (among other things). I was already annoyed at the tv interview I saw with the B&W researcher. I go to an endo at B&W who is pro-statin and I got really angry when he said that no other side effects matter (other than the muscle side effects). Excuse me? They matter to us!!!

So I just thought I'd throw in a few comments of my own:

1. I already take two other meds that causes insulin resistance and have no choice in the matter. This, I have a choice int he matter

2. I am sick of the ADA shoving statins down our throats. There is no proof they reduce plaque build up. They only reduce LDL. And they do screw with our livers.

3. My pharmacist told me folic-acid just as easily reduces c-reactive protein.

4. However, I have a 40 plus year type 1 board buddy who is taking a cheap generic statin who has lowered his A1C from 5.2 to 4.4 (as well as his LDL). My A1C is 5.1 consistently and my LDL is 108. HDL and triglycerides in 70's. I don't see c-reactive protein on my recent lab list. Will have to get it done.

Thanks for your hard work and come read my articles some time.

Doris Dickson

Lily said...

Thanks! That was interesting to read. The comment that Simvastatin lowered Trinkwasser's LDL was good news since it costs 1/10th of what Crestor costs using my RxDrugCard.

Anonymous said...

Why has the cancer in this study gone undiscussed?
Crestor 35 deaths
placebo 58 deaths
This is the most interesting of the stats in this study.

Jenny said...


My guess is that in a technical sense the cancer statistic is not statistically significant. It's a very small difference in a very large sample and may be within the normal range of error.

The tiny differences in many of these statistics only became significant (again, in the technical meaning of the word) when all added together.

Unknown said...

According to a study in the University of Michigan, announced Statins may reduce the risk of dementia and loss of memory. But some people advice not to use statins in preventing dementia due to the side effects of drug.
Cholesterol-lowering statin drugs have had a rough time of it lately. Thank you for sharing this article, this is very informative.

My url:
I also invite people to use my links and read more.