Ok, it’s a silly question I admit. As any physician, and anyone else who thinks about it for a few seconds knows, there is no such thing as saving lives. You can only postpone death. So a more nuanced question is required: “Do statins postpone death?”. My mother sure thinks they do, since she’s been popping an atorvastatin a day for the last couple of years on her GP’s recommendation. Of course, the question gets more complicated when you consider two things:
- Different drugs have different effects in different people. A drug might have a net positive effect on a 70 year old with advanced heart disease, but a net negative effect on a healthy 30 year old.
- All drugs that have beneficial effects also have side effects. So any positive effect on longevity needs to be weighed against negative effects. If a drug makes you live longer but feel awful, you might not think it’s worth taking. And if a drug decreases your risk of dying of a heart attack, but increases your risk of dying of something else by as much, you probably won’t think it’s worth taking.
So the actual question becomes: For which groups, if any, do statins postpone death, and is that postponement big enough to outweigh any negative side effects and the bother of having to take a pill every day?
One of the earliest statin studies was the Scandinavian Simvastatin Survival Study, also known as 4S. 4S was a randomised controlled trial carried out in the late 80’s and early 90’s, and published in 1994. It was financed by Merck, the company that at the time held the patent on simvastatin. It followed 4,444 patients with atherosclerotic heart disease and high cholesterol levels for around five years. Half were given simvastatin, and half were given a placebo. Of the patients on simvastatin, 8% died over the five year period, while of the patients on placebo, 12% died. That is an absolute risk reduction of 4% . Sounds pretty good, right? Additionally, the probability of dropping out due to side effects was the same in the two groups, which suggests that side effects wasn’t an issue.
Personally, I would take a drug that decreased my five-year risk of dying by 4% and that didn’t have any side effects. Apparently, lots of people agreed, because this is that study that launched statins in to the stratosphere, and made them one of the best selling drug classes in history.
After 4S came WOSCOPS (West Of Scotland Coronary Prevention Study). 4S had shown that statins worked in people with known heart conditions, but no-one knew if they filled any function among people without a known heart condition, in other words, if they could be used for primary prevention (treating a disease before it happens). The purpose of WOSCOPS was to find out. It was funded by Bristol-Myers-Squibb and Sankyo, the companies that at the time held the patent on pravastatin. 6,595 middle-aged men without any known heart conditions but with high cholesterol levels were randomly assigned to receive pravastatin or placebo, and followed for around five years. The study was published in 1995. Of the patients on pravastatin 3.2% died, while of the patients on placebo, 4.1% died. This gave an absolute risk reduction of 0,9% , which was not only small but also not statistically significant (i.e. the result could very well have been caused by chance). Oops. On the plus side, the rate of drop-out was the same between the placebo group and the pravastatin group, so again side effects didn’t seem to be a big issue.
What conclusion to draw from this study? If you are a middle-aged man with high cholesterol levels, but who hasn’t had a heart attack, it’s probably not worth taking a statin, at least if you’re taking it in order to live longer.
After WOSCOPS came CARE (Cholesterol And Recurrent Events). This was a trial in which 4,159 patients with a prior heart attack (myocardial infarction) but normal cholesterol levels were randomized to receive either pravastatin or placebo and followed for around five years. So this study was basically doing the opposite thing to the WOSCOPS study – it was looking at whether statins were useful in people who had had a heart attack but had normal cholesterol levels. It was funded by Bristol-Myers-Squibb and the results were published in 1996. Unfortunately for Bristol-Myers-Squibb, this study was also negative, just like the previous one. 9.4% died in the placebo group, versus 8.6% in the pravastatin group. This absolute difference of 0.8% was not even close to being statistically significant, i.e. there was no difference between the groups. Doh! On the plus side though, again the medication seemed safe.
Hmm, so where does that leave us? Statins seem to be useful for people with high cholesterol levels who have had prior heart attacks, but they don’t seem to be useful for people who only have high cholesterol levels, but haven’t had heart attacks or have only had heart attacks, but have normal cholesterol levels. Odd.
These were the first three big trials of statins. All in all, not extremely encouraging.
The next study to come out was the LIPID study (Long-term Intervention with Pravastatin in Ischemic Disease). It was basically a remake of the CARE study. 9014 patients were enrolled, and they had to have had a heart attack (myocardial infarction or unstable angina), but normal cholesterol levels. The patients were randomized to either pravastatin or placebo, followed for an average of six years and the results were published in 1998. Again the study was paid for by Bristol-Myers-Squibb, and finally they were able to get some positive results for their drug. In the placebo group, 14.1% died, while 11.0% died in the pravastatin group. That is a 3.1% absolute difference in mortality, and it was statistically significant.
So we have two trials that tried to do the same thing, i.e. give pravastatin to people who had ischemic heart disease but normal cholesterol levels. One (LIPID) was able to show a difference in mortality, the other (CARE) wasn’t. The one that did show a difference enrolled more than twice as many people though, so its data are more robust. Basically, after these four trials it seems that statins are good for secondary prevention, i.e. for treating people who have known ischemic heart disease, but we don’t know yet whether they work for primary prevention, although we have one study (WOSCOPS) that suggests they don’t.
AFCAPS/TexCAPS set out to answer that question definitively. It was a trial that recruited 6,605 middle aged (ages 45 to 73) people without any prior history of heart disease and with normal cholesterol levels, and randomized them to either receive lovastatin or placebo. The patients were followed for around five years and the trial was funded by Merck. The results were published in 1998. At the end of the study, 2.9% of the patients had died in the lovastatin group, and 2.7% in the placebo group. So the results were completely negative. More people in the placebo group were alive after five years than in the statin group. Proponents of the study would argue that it wasn’t “powered” to show mortality benefit; in other words, the study would have needed to have many more participants or gone on longer to show a benefit. Which I think is reasonable, to an extent. But at the same time, if over three thousand people take a drug designed to improve life expectancy every day for five years and there’s not even the slightest hint of a positive effect on life expectancy, then it’s probably not worth taking that drug. Conclusion: If you haven’t had a heart attack, and especially if your cholesterol levels are normal, there’s no point taking a statin.
Next came the ALLHAT (anti-hypertensive and lipid lowering treatment to prevent heart attack) trial, which was published in 2002. Again, this was a primary prevention trial, but it was looking at older adults with elevated cholesterol, so a group who might more reasonably be expected to benefit from statins than the one studied in AFCAPS/TexCAPS. One thing I really like about this trial is that, unlike all the previous trials, it wasn’t funded by a pharmaceutical company, but by the National Institutes of Health (NIH). That makes the results much more trustworthy in my opinion. 10,335 people were followed for an average of five years. No difference was found in mortality between the statin group and the usual care group.
For me, that is pretty conclusive. Just being old and having high cholesterol levels is not reason enough to take a statin.
In 2002, the same year that ALLHAT was published, another important paper was also published. This was HPS, the Heart Protection Study. Like ALLHAT, this was also not financed by the pharmaceutical industry. Instead it was paid for by the British Heart Foundation and the British Medical Resarch Council. Again, that means the results are more trustworty than the results from the industry financed studies. Apart from that, it was a big study, with 20,356 participants aged 40-80, so in many ways this was the best study yet. In order to qualify for the study patients had to have known ischemic heart disease, some other type of occlusive arterial disease (for example stroke or claudication), or diabetes (a known risk factor for ischemic heart disease). They were randomized to receive either simvastatin or placebo. Of 10,269 patients who received simvastatin, 12.9% died. Of 10,267 who received the placebo, 14.7% died. This is an absolute difference of 1,8% and it was statistically significant. That is a small but real difference.
The 1,8% improvement suggests why the studies looking at statins for primary profylaxis weren’t able to achieve positive results – although statins have an effect, it is so small that there is really only a noticeable benefit if you belong to a high risk group. If the risk only decreases by 1,8 percent among people with known heart disease, it’s probably going to decrease by such a tiny amount (if at all) in people without heart disease that the effect isn’t even measureable.
A 1,8% decrease in absolut risk means that 55 (100/1.8) people need to be treated for five years in order for one person to have their death delayed – remember there is no such thing as saving a life. How big that delay is, we don’t really know, since the studies haven’t been designed to answer that question, but the small effect size in the studies suggests that it might be from a few months to at most a few years if you happen to be that lucky one in 55 who gets to postpone death.
One additional interesting result from the study that is worth noting is that the size of the benefit was the same regardless of the patients’ starting cholesterol – people with low cholesterol benefited as much from the statins as people with high cholesterol. This result was of course in line with all the previous studies. Conclusion: high cholesterol levels are not a sign that you as a patient will benefit from taking a statin.
We’re pretty much ready to wrap this up, but I’m going to mention one other study, SPARCL (Stroke Prevention By Aggressive Reductions in Cholesterol Levels). This study was published in 2006, and the objective was to see if statins are useful at preventing new strokes in people who have recently had a stroke or TIA (transitory ischemic attack – a mini-stroke that resolves on its own). The study was financed by Pfizer. It randomized 4,731 patients who had recently had a stroke or TIA to either receive atorvastatin or placebo. Patients were followed for an average of five years. In the placebo group, 13.1% had a new stroke, compared with 11.2 in the atorvastatin group. That is a 1.9% absolute difference – i.e. if you treat 53 patients for five years, you prevent one stroke. When it comes to the more important metric of death, there was no difference – 9.1% died in the atorvastatin group versus 8.9% in the placebo group.
Hmm, how to interpret that data? So you decrease the risk of stroke, but not the risk of death. Of course, most strokes are not fatal, but can still be devastating if they result in significant brain damage. Which is why I’m including this study, because someone who has had a stroke might think it’s worth taking a statin to very slightly decrease their risk of a new stroke, even knowing that it won’t affect their overall risk of dying.
There have been more studies, but they are invariably either of low quality or just copies of the ones that have already been done with minor varitions, and have similar results, so I won’t go through them here.
So, what conclusions can be drawn from all this? Here are my two takes:
Firstly, if you have some type of confirmed ischemic arterial disease, whether it be myocardial infarction, stroke, angina, or claudication, or you have type 2 diabetes, and you don’t mind taking a pill every day for the rest of your life to achieve a very minor lengthening of your life expectancy, then you should take statins. If you don’t have confirmed ischemic arterial disease and you don’t have type 2 diabetes, then it is almost certainly a waste of your money and time to be taking a statin.
Secondly, there doesn’t appear to be any correlation between cholesterol levels at the start of treatment and the benefit gained. People with low cholesterol levels gain just as much (or little) from treatment with statins as people with high cholesterol levels. The studies listed which specified that people had to have high cholesterol levels in order to be eligible to participate were no more likely to find a benefit, and did not find a bigger benefit, than those that took all comers. What this means is that it is pointless to get your cholesterol levels tested – the number you get back doesn’t tell you anything about your probility of benefiting from taking a statin.
You might also be interested in my article about whether aggressive blood pressure treatment is good or bad for your health, or my article about whether it’s sensible to take fever lowering drugs when you’re sick.
I like your way write. I would like to read more these. From Lappeenranta Finland. Matti
As a taker of statins for the last 25 years with no heart problem I find your analysis very soberring. I alway suspected the recommended cholestrol level. Given the studies how can one claim that 220 require medication and 210 do not
From the 1950 to the 1990 years normal Cholesterol (Chol) levels were under 250mg/dL. Then affluent countries began fighting obesity from overeating. The economic losses were thought to decrease without obese people. One way was to scare people and, officially, ‘desirable’ Chol levels went down to 230, then to 210 until 190mg/Dl.
As Chol is synthesized in the liver, 50 years ago a very low Chol level was typical on liver cirrhosis patients.
During and after WW2, with famine, the rate of myocardial infarction and cerebral thrombosis decreased in Europe, only to increase after the availability of food went to the usual levels.
The pharmaceutical companies went on to make their profit. They amplified all the scares. Tons of statins are sold each month but not in countries where famine raged (aren’t needed).
In my view, there are two groups of people:
* Those who have variable levels of Chol on each occasion; a hard work for education on meals. They have evidence of bad arteries throughout. They die from organ infarction or from very serious arterial disease.
* Those who always have very high Chol levels, every time, in the range of 180-210mg/dL (more or less 10mg). These were healthy, died after their eighties (shortly after a CVA), and were very intelligent; a hard work to kill their fears and to ‘be quiet, first do not wrong’.
Thanks for your attention.
Statins have significant side effects, as one would expect from a drug that interferes with a fundamental physiologic process, namely the body’s ability to make cholesterol, a molecule absolutely essential for life.
I am not a doctor and just guy-guessing.. But what if the high level of cholesterol is not the problem-indicator but the triglycerides are. What if rather than taking a statin the patient in a risk category were to change their diet in a way that the body consumes more of the blood fats rather than storing it (“sugar” and food stuff that convert to “sugar”) . Maybe a high cholesterol is good, in the right circumstances.
Great read. What is then the mechanism through which statins work? The cholesterol hypothesis seams completely dead, what else is in play?
Hi Håkan!
That is the million dollar question. It is known that statins have anti-inflammatory effects, so as I understand it that is the main hypothesis.
Perhaps not a milion dollar question since the effect is marginal anyway 😉
But what about anti-clotting, what is your take on that? Most things circulating in the blood seems to be involved in clotting one way or the other.
That is also a possibility, and could be tied to the anti-inflammatory effect. Inflammation generally increases blood clotting.
I wonder if you would comment about the observation of sometimes devastating effects on muscle, especially for those on stains who do large amounts of physical exercise, such as distance running?
Hi!
I haven’t looked in to the literature on that in detail yet, but I do plan to write an article on it in the near future.
Well, interesting. But
I had a by pass 10 years ago. My cholesterol level was not high, but LDL/HDL was high. I am on Crestor since,
For me the result from the Jupiter study not included in your overview was convincing:
#AstraZeneca announced it has decided to stop the CRESTOR JUPITER clinical study early based on a recommendation from an Independent Data Monitoring Board and the JUPITER Steering Committee, which met on March 29, 2008. The study will be stopped early because there is unequivocal evidence of a reduction in cardiovascular morbidity and mortality amongst patients who received CRESTOR when compared to placebo.#
It was unethical to go on with placebo!
Is there more studies like this that you did not want to quote?
It seems like some people are more sensitive then others to the level of cholesterol. I wonder how this was included in the studies you quoted. -same % reduction,or fixed target value or LDL/HDL level
Hi,
I did not include the JUPITER study because in my view it doesn’t add anything new on top of the earlier studies, and because it was stopped early, which makes it’s results highly questionable from a scientific standpoint and thereby of low value. The problem is due to regression to the mean. When researchers see that they have a good result, they stop early, which they wouldn’t have done if the study had had a weaker result or a negative result. This results in cherry picking of the time point at which the trial is ended, which generally makes the results seem much better than they really are. JUPITER was supposed to run 5 years, but was stopped at 1.9 years. That is bad science.
JUPITER stands ”Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin trial”. The very name is giving away the bad conduct in this study. The had decided in advance that the outcome would be a Justification of their pill.
So you prefer to take the risk that those people should die. in the name of science. But of course your bias is that this is not working.
I am happy that most GP dont agree with you and gives their patients best possible treatment. Are all those findings of reduced clogging in blood vessels also bad science?
All test of new medicine have the target to prove and justify that they are working. At least, this is my experience after 20 years of investment in biotech.
JUPITER enrolled only people with normal cholesterol but elevated CRP, obviously they had inflammation in their bodies. It is hard to understand how any benefit could be attributed to the cholesterol lowering in this particular study.
addition
What about findings in the famous FRAMINGHAM HEART STUDY. I dont think this is bad science. Why was those findings and other epidemic studies not included?
Obviously you can find arguments outside your study.
I think studies that not considering influences of LDL/HDL quota is not relevant as high HDL take away the risk from high cholesterol as such.
When there is data from randomized controlled trials, that data trumps observational studies, because observational studies can only show correlation, not causation, and the scope for confounding effects is huge.
Dr. Rushworth, now I start with quotations:
It is dangerous to be sincere unless you are also stupid — George Bernard Shaw.
Genius may have its limitations, but stupidity is not thus handicapped — Elbert Hubbard.
Some people are hurt with truth. The industry are not. What follows is for readers not involved in Medicine.
Cholesterol is essential to animal life — plants do not use it. For cellular walls and transmitters (hormones). Cholesterol is a major component of brain cells. Almost 30% of it is in the brain.
I always feel that to give a diet or drug designed to gave results in some 20-40 years is non-sense [there are exceptons] when given to people more than 50 year-old. We do not treat (yet) people with Məṯūšélah DNA.
By the way, when young (I’m 77) I didn’t see dementia patients. They appeared after 1980.
A life of good work, joy and peace!
With regards to dementia and meds like statins. I’m in my late 60s and been in the health business for about 45 yrears and soemtimnes notice trends. One is dementia.
I wrote this article:
https://www.garymoller.com/post/is-this-the-real-cause-of-the-dementia-epidemic
garymoller (Dr), I liked what you wrote.
It covers what is intentionally forgotten: “First, never do harm”.
Thanks
Thank you! Love your work. Keep it coming.
The UK BBC have just reported https://www.bbc.co.uk/news/health-54951648 on a study that suggests some of the reported statin problems are nocebo effects, which perhaps means we may need triple blind trials!
I haven’t read the study, but seems to be small (60 people total) so I don’t think any big conclusions can be drawn. But the idea of having a third arm that knows they are not getting treatment is interesting, to get more clarity on what the size of the placebo effect and nocebo effect is, and can make sense in some situations.
I was also thinking that the ’empty’ bottle treatment was sort of obvious so could create certain effects, especially if all participants were briefed equally about the trial (i.e. they’d know what was going on and react accordingly).
9 years ago I had a stent inserted to clear a heart artery blockage and was advised to take blood thinners and statins for the rest of my life. Some years later, after a pharmacist asked if I was also taking CoQ10, I started to research papers on statins. I was also concerned about muscle cramps I was experiencing after exercise and an occasional lightheadedness. At this time I also discovered that electro-therapy and overnight earthing to restore body potential to the natural slightly negative potential of the earth’s surface has beneficial effects on blood circulation, causing increased flow and agglutination. As a natural skeptic, and health and safety professional, I checked this out and found bio-physics reasons supported by my own voltage tests and blood tests. No more statins, no cramps and healthier than ever. At 77, still running, cycling, mountain climbing and free of medications.
Correction: increased blood flow and avoiding agglutination. as similarly charged blood cells do not attract to one another electrostatically.
I dispute your assertion that side effects do not seem to be a problem with statins. On the contrary statins are notorious for their severe and often permanent affects on muscles and joints. Of the twelve or so of my friends who were prescribed statins all have abandoned them due to the severity of side effects. Some still have lingering effects years after abandoning statins. In my own case I was prescribed a 10mg dose of a statin to attempt to lower my LDL which was ~ 90 at the time. I have a long history of stent restenosis – 26 installed, 25 restenosed 100% – hence the statin. While taking this dose I began to experience muscle and joint pain at a low level but did not connect it to the statin. When the 10mg dose had no effect whatsoever on my LDL level the statin dose was increased to 20mg. Within a week I had so much pain in my hip joints and right leg that I could barely walk from kitchen to bedroom. The second night I just stopped taking the statin and within a few days was experiencing less pain than I had since starting the statin regime.
Don’t even try to tell me these poisons don’t have side effects. Incidentally after I abandoned the statin my LDL went down to 80….