Friday, November 15, 2013

Universal statins: scam....or just honest good luck for Pharma...or what?

The American Heart Association and American College of Cardiology issued new guidelines on Tuesday for reducing risk of heart disease and stroke (the first of five explanations of these new recommendations is offered here).  If you've got a 7.5% risk of heart disease or higher, as measured by their risk calculator (downloadable here), they recommend you go on statins.  This means, according to the panel, that 70 million Americans should now be considering taking these drugs.  This is perhaps 70% more than the number who now take them, and would put at least one third of all adults in the US on this drug.  For life. 

These recommendations have caused quite a ruckus, but perhaps for the wrong reasons. Before Tuesday, people were put on these drugs to lower their LDL cholesterol levels beyond a given threshold, but, confusingly to many, that threshold which we had supposed was a well-established rock-solid risk factor, has now been eliminated!  Before Tuesday, the indication for going on statins was high LDL, but the indications have now been broadened to include other risk factors such as diabetes and obesity.  So, people now taking statins wonder if they should continue, and others wonder if they need to start.  Some doctors commenting on these changes hasten to add that the most important protection against heart disease is a healthy lifestyle -- don't smoke, exercise, lose weight -- but if these can't be accomplished, statins are recommended (see Dr Harlan Krumholz on "The Newshour" on PBS, e.g.).


But the new recommendation seems strange. First, a word about statins, drugs designed to control circulating lipids (fats), which confer heart-disease risks. They inhibit an enzyme called 'HMG-CoA reductase' which is expressed in liver cells as they produce cholesterol from raw ingredients and secret it into the blood stream.  Lower enzyme activity, lower circulating lipids.  But in fact, there is evidence that for some reason statins target inflammation in irritated arteries and veins, which may be what reduces risk of heart disease rather than any effect on cholesterol, so there is mystery even in the supposed reason for their supposed effectiveness.

Now, according to John Abramson and Rita Redberg in an editorial in the Thursday New York Times, statins aren't actually effective at preventing heart disease.
Statins are effective for people with known heart disease. But for people who have less than a 20 percent risk of getting heart disease in the next 10 years, statins not only fail to reduce the risk of death, but also fail even to reduce the risk of serious illness — as shown in a recent BMJ article co-written by one of us. That article shows that, based on the same data the new guidelines rely on, 140 people in this risk group would need to be treated with statins in order to prevent a single heart attack or stroke, without any overall reduction in death or serious illness.
If the recommendation is not based on evidence that everyone can agree is reliable then, where is it coming from? Partly, we think, it's a reflection of our belief that we're immortal, partly a general belief in the benefits of drug intervention.... and partly it's a reflection of the fact that some of the recommendation-makers have a vested interest in statins.

The first reason, our belief in immortality, is cultural, and of course very natural.  Few of us, even apparently those with strong religious belief in life hereafter, want to test out that belief.  Heart disease is the number 1 killer in the US and most of us don't want to die of it.

The second reason is more problematic. The most reliable way to lower heart disease risk is through diet, exercise and not smoking.  Indeed, lean, fit, non-smoking individuals whose only risk factor is high LDL are generally at low risk of heart disease.  Until this week, the purpose of statins was to lower LDL cholesterol, but that doesn't reliably lower risk of heart disease. So, are statins a good replacement for life-style?  The answer is No.

But what about the vested interest issues?  As Abramson and Redberg say:
The process by which these latest guidelines were developed gives rise to further skepticism. The group that wrote the recommendations was not sufficiently free of conflicts of interest; several of the experts on the panel have recent or current financial ties to drug makers. In addition, both the American Heart Association and the American College of Cardiology, while nonprofit entities, are heavily supported by drug companies. 
This kind of conflict of interest means that one must be highly suspicious.  One might argue that industry reps are the most knowledgeable about the benefits of their product.  If you want to know how to cure a toothache, ask a dentist, after all.  But how are we to judge the motive behind the recommendations?  Corporate-sponsored research is notoriously biased toward findings favoring their sponsor.  This doesn't mean the bias is intentional, but the evidence suggests that often it is.  At least, the corporate-sponsored research the corporate sponsors tell us about is that which favors their product, since they aren't in fact required to report all their results, and often don't.  (This is why Ben Goldacre, physician, writer and epidemiologist, started the AllTrials campaign to require that all clinical trials be registered, and all results be reported, positive and negative.)  So, when the science isn't convincing, and vested interests are involved in decision-making, it's not unreasonable to be suspicious.

Contingency and context
In addition, the related roles of context and contingency are fundamental and important to understand here.  Risk of heart disease is based on the context--genomic and environmental exposures affect the levels of statins and their consequences.  The risk in an individual is contingent on his/her situation at present, and that can change.  This seems so hard for the established system to understand!

Statins have side effects.  Risk of statin-related disease is estimated in the context of current culture.  If that culture changes, then the risks will change and by all that we know, they'll change dramatically.  If the major contexts change--better diets and so on, things we know a lot about--then the overall risk of heart disease will change.  Models can only do so well at estimating the interaction among the various factors (as the above quote suggests).  If people live healthier lives and if physicians actually pay attention to risk calculations, many may be able to go off their statins, or not start on them, as a result.

But how many will actually risk going off?  Will they keep taking, or their physicians not dare to recommend stopping, for various subjective, inertial, or even emotional reasons?  Will drug companies recommend cessation if, say, body weight goes below some value?  Will they fully advertise the fact that if other factors are favorable, to stop buying their product?  What does history--including their history--suggest to you?

And 10 years from now, how accurate will the predictions have been on which lifelong medications are now being recommended?  Will results be contingent, for example, on the current recommendations themselves?   For example, if you're on statins, will you be more likely to take that second helping of fries, feeling protected by the drug?  The Times Op-Ed puts it this way:
Perhaps more dangerous, statins provide false reassurances that may discourage patients from taking the steps that actually reduce cardiovascular disease. According to the World Health Organization, 80 percent of cardiovascular disease is caused by smoking, lack of exercise, an unhealthy diet, and other lifestyle factors. Statins give the illusion of protection to many people, who would be much better served, for example, by simply walking an extra 10 minutes per day.
These are not secret or new issues by any means, but they tend to be overlooked or minimized by a system that tries to be 'objective' based on current data.  The Op-Ed is written by respectable authors, but they are also known skeptics of the over-medicating problem, with its built-in conflicts of interest as we noted above.  So is their skepticism itself a disqualifying issue--does it mean they bias their views in a similar way to having Pharma-supported people on the panel that made the new recommendations?

Clearly, the issues are complex as so many issues related to late-onset disease are.  After all, you don't get a heart attack even at a young age like 40, unless you live to be 40.

Like second-hand smoke?  The real beneficiaries
Smokers get all sorts of diseases, because of the direct effects of the ugly weed.  But those who live in the same house also get diseases, indirectly, courtesy of their smoking cohabitant.  We have just the opposite story here.  The vendors of statins will get filthy rich as a direct result of their recommendations, whether or not they actually prevent heart disease.  And if they do, some other people--maybe the same people--will get even richer as an indirect result of the same recommendations!

If we don't die of heart disease or its associated diseases, we may live longer but that means more of us will get the slower, nastier, very expensive lingering ailments of old age.  The surgeons, retirement homes, cancer and dementia drug-makers will rake it in big-time!

We've written a few times about the subtle, surreptitious problem of competing causes, and this is another manifestation of the problem.  It's largely unavoidable that if you survive the quick-hitting earlier causes of death, you'll last and linger in service to the slower causes.  They're even more expensive.

We would not credit (nor blame) the statin-promoters for the diabolical scheming that would be involved in salivating over the indirect benefits of statin use.  That takes more perception and a longer view than most people, even scientists, usually have.  It is clear that most drug companies, not to mention the scientific research community itself, as we often write, are in for the quick kill, so to speak.


Hollis said...

Thanks for the info, discussion and links! I'm one of those with high LDL but low ranks for every other risk factor.

Unknown said...


one thing that confuses me here (and in othe guidelines like this) is that question(s) and goal(s) is/are not clearly and explicitely stated:

Public Health goal is to reduce prevalence/incidence of disease in a population at large and it is well documented that even a small decrease in an average (LDL) cholesterol level reduces prevalence/incidence/mortality of CVD in the proposed target population. Hence, it makes sence to introduce interventions that are designated to lower average (LDL) cholesterol.

However, I am not covinced that any single- or poly-pill approach will work here. A solution need to come from political acts. Best obesity prevention would be to promote healthy eating. Now we allow fast-food culture, which has created big food industry and business and then we ask big pharma to minimize health hazards attributable unhealthy behaviour. This serve as an example for an industrialism paradigm; we develop a technology to porvide fast and cheap food for large group pf people and then employ another (pharmacological) technology to solve the helth problems created by the first technology. This certaily serves one way to create growth economy...

A prevailing paradigm in clincial medicine today is to use population studies to estimate an individual's pobability of a disease (# of diseased cases/total # of individuals in the proposed population). I do not know, what justifies this paradigm? We can compute a probability to win in a lottery because (by definition) each number has same probability to be picked. Same logic does not, however, fit in a disease prediction because disease does not "pick" individuals with same/equal probailiity; the probability depends on an individual's biological properties as well as on the background population where s/he happens to live. Hence, the prolem is not necessarily the complex, context dependent pathobiology of CVD or pharmacodynamics of statins but math and general biology (=there are variation among individuals). I have a hard time to understand, why top medical scientist appreciate these "facts" .Maybe we have bad memes. I do not believe that they have bad will or that they have sold themeselwes to big pharma.

Ken Weiss said...

Absolutely! We have put the 'lottery' point somewhat differently, that today's methods assume replicability but the key fact of life is that people are not replicates. It is these issues -- what you call bad memes, and that we have referred to as the need for a new conceptualization -- that are key.