Preventive Medicine Column

Dr. David L. Katz

All too often, drugs that perform some valuable function- fighting cancer, alleviating pain, lowering blood pressure- come at a high price in side effects.  Every now and then, though, a class of drug comes along that does what it’s supposed to do- and then confers other benefits not originally planned.

I have two favorite drug categories that exemplify this.  The first is angiotensin converting enzyme inhibitors (ACE inhibitors).  These drugs, developed to lower blood pressure, also protect the kidney in unique ways, and may actually help prevent diabetes in high-risk individuals.  This class of drug can, on occasion, cause a dry cough- and there are of course other potential side effects as well.  But by and large, ACE inhibitors do the good they were designed to do, and other good besides.

The other stand-out in this area is statin drugs.  Originally called HMG-coA reductase inhibitors, these drugs block a key step in the synthesis of cholesterol in the liver.  By turning down cholesterol production, statins lower levels of harmful cholesterol in the blood- and reduce the risk of heart disease.  There is no doubt about it; in patients with high levels of harmful cholesterol, statins save lives.

But it seems they do it in ways other than just lowering cholesterol.  Statin drugs also reduce inflammation- not what they were designed to do, but they do it just the same.

As with ACE inhibitors, statins can cause side effects.  Muscle pain is most common, often treatable with the nutrient, co-enzyme Q10.  A more serious kind of muscle inflammation –rhabdomyolysis- can occur as well, but is quite rare.  Statins can also cause liver inflammation, but usually don’t.  More often than not, these drugs help prevent heart disease and are very well tolerated.

Recently, a study called JUPITER looked into the value of statin drugs in patients who don’t even have high cholesterol.    Responding to the revelation that statins reduce inflammation as well as cholesterol, the JUPITER trial tested the hypothesis that statins can reduce the risk of heart attack and death in people who don’t even have high cholesterol- but who do have high levels of inflammation, as indicated by a chemical in the blood called CRP.

The JUPITER trial showed, in fact, that the particular statin drug called Crestor reduced the risk of cardiovascular disease by roughly 50% in people with normal cholesterol and high CRP.

Responding to this dramatic and somewhat surprising benefit of Crestor, the FDA recently approved the drug for just such patients- those with normal cholesterol numbers, but elevated CRP.  AstraZeneca, the drug company with the Crestor patent, supports the FDA decision with the expected enthusiasm, and is beginning a marketing campaign in response.

So, the question for you is: should you take Crestor if you have normal lipid levels, but a high CRP?  There is a case both for, and against.  And the reason is that risk, like everything else, is relative.  In fact, it is absolutely relative.

Clinical studies often report relative risk reduction, as the JUPITER trial did.  A relative risk reduction of 50% means that the events of interest- heart attacks, or deaths- occur half as often on the drug as they did off the drug.  But, the significance of half as often as ‘they did off the drug’ obviously depends on how often events occur off the drug.

For example, if a given group of patients has 1 heart attack per million people per year, a drug that cuts that risk down to 1 heart attack per 2 million people per year, would provide a 50% relative risk reduction.  But in ‘absolute’ terms, it only reduces the heart attack risk by 1 in a million per year.

This is the trouble with relative risk- it ignores the absolute risk reduction.  Relative risk is often used expressly because it exaggerates the magnitude of benefit when risks are low to begin with.

In the JUPITER trial, Crestor reduced heart attack risk from 0.37 percent per year, to 0.17 percent per year.  A particularly good way of understanding this is called ‘the number needed to treat,’ and is, literally, the number of patients that would need to be treated for a year before one would benefit.  Mathematically (sorry!), the NNT is 1 divided by the absolute risk reduction, 1/ARR.  The ARR in the JUPITER trial was 0.2 percent, so the NNT is 1/0.002, or 500.  So, 500 patients like those in the trial need to take Crestor for a year before one benefits.  Or 50 patients would need to take it for ten years.

Before you sign up to be one of them, consider that eating a diet rich in fruits and vegetables and other  wholesome foods close to nature can lower CRP and inflammation.  So can a supplement of omega-3 oils.  And so can moderate exercise.

For those who find diet change too hard, and fish oil ineffective, Crestor and other statins are a reasonable consideration.  But given the small absolute risk reduction, I would favor their use only among those who have heart disease in the family- especially heart disease despite normal lipid levels.

For most others, the small absolute risk reduction seen in JUPITER is apt to mean that the relative benefits of Crestor and similar drugs may be greater for the pharmaceutical company, than for the patient.

The right decision for you cannot be made by the FDA.  You and your doctor will have to consider both the meaning of relative risk, and the fate of your relatives, to determine whether expanded indications for Crestor indicate that you should take it.


Dr. David L. Katz;