March 16th, 2012
Helping Patients Understand the Statins Controversy
Harlan M. Krumholz, MD, SM
After the Topol op-ed in the New York Times last week I began to get a lot of inquiries about the safety of statins. At the beginning of his piece he emphasizes that our use of statins could cause a “sharp increase in the incidence of Type 2 diabetes.” At the end, he states emphatically that “the problem of statin-induced diabetes cannot be underplayed.” Many patients reading these words were worried about their decision to take statins. I decided to write a missive to patients to help them sort through all the information about statins– and to help them in their decisions about treatment. I told them that there were five things that they needed to know:
- It is your decision.
- Statins reduce risk.
- Statins have risks.
- Your risks matter.
- Minimize your statin risk.
You may find it useful for your patients. The title is Five Things You Need to Know When Deciding on a Cholesterol Lowering Drug. Feel free to share your thoughts about it – or to use it if you think it may be helpful to you.
I had just shared this with a soccer teammate of mine, a 30 year old male with LDL 170 whose PCP was considering starting a statin. In my opinion, that is the group to whom this new data is most relevant. In these low risk patients (such as my central defender) with 10 year risk of <1% with high LDL, I am going to shy away from statins, in light of this diabetes risk.
I do not think the FDA pronouncements will have an effect on my statin decisions. Rescinding the need for monitoring liver function tests is long overdue and should alleviate some patients’ concerns. As stated in the article, the clinical significance of the diabetes issue is unclear, particularly in light of the probable ncreased cardiovascular risk likely present in those patients destined to develop diabetes on a statin. I do have an issue with the dosing suggestions. I believe that a 30 to 40 percent decrease in the LDL is a reasonable goal, likely regardless of risk. Some evidence indeed suggests that early initiation is beneficial (HPS and Mendelian randomization studies support a legacy effect). I would have preferred that the article make a distinction between moderate and high doses rather then just low and high doses. I am not convinced of any good evidence to support a change in the risk – benefit relationship according to statin dose in low to intermediate risk patients. Higher risk patients certainly tend to benefit more with higher dose statins.
The statin controversy is yet another argument for the accuracy of coronary calcium imaging as a tool to direct therapy.
Treating an elevated HS-CRP with statins in the face of “normal” cholesterol is a bad idea. The “Jupiter equivalent” in MESA demonstrated that of the “Jupiter like” subjects, 50% had a calcium score of 0 and therefore such a low risk that statin is more likely to harm than improve risk.
The assessment that Dr. Ryan makes is a good one. The problem is that what percent of heart attacks occur in “low risk” individuals. Answer, over 50% of heart attacks occur in subjects determined to be low risk by NCEP until their chests began to hurt. In men under the age of 50, that statistic is closer to 75%.
90% of the Women in MESA were considered low risk by conventional risk factors. 30% of these “low risk” women had calcified plaque in their vessels and therefore no longer “low risk”.
My point: Use an accurate technology (CAC) to measure risk and treat those who will benefit from treatment and avoid the toxicities in those who will not likely benefit from treatment.
very informative tool for patients..
There are many other reasons to be restrictive with statin treatment. In a study based on data from some 2 million primary care patients, of whom 225,922 were new statin users without heart disease, Hippisley-Cox and Copeland reported that more than 4 % suffered from severe myopathy, moderate or severe liver dysfunction, acute renal failure or cataract (doi: 10.1136/bmj.c2197). Even this numbers of side effects was underestimated, because liver damage was recorded only if the alanine transaminase concentration was more than three times the upper limit of normal, and to record muscle problems they required the creatinine kinase concentration to be four times the upper limit of normal. Add to this the risk of diabetes, which was not investigated. There is also increasing evidence from cohort and case-control studies and from several controlled, randomized trials that cholesterol lowering may result in cancer (QJM 2012:105:383-8). It has been underestimated in the meta-analyses because after the publication of HPS, where the incidence of non-melanoma skin cancer was close to statistically significant, this type of malignancy has not been recorded in any trial. Another reason to restrain statin treatment is that more than twenty studies have shown that old people with high cholesterol live the longest (references on request). One of the first such studies was indeed published by Harlan Krumholz and his group (JAMA 1994;272:1335-40).
After several years of reading Uffe Ravnskov papers and comments I have discovered another great scientist – Stephanie Seneff. She has described the link between statins and diabetes mellitus.
(…Once the muscles can no longer keep up with lactate supply, the liver and heart will be further imperilled. They’re now worse off than they were before statins, because the lactate is no longer available, and the LDL, which would have provided fats as a fuel source, is greatly reduced. So they’re stuck processing sugar as fuel, something that is now much more perilous than it used to be, because they are depleted in membrane cholesterol. Glucose entry into muscle cells, including the heart muscle, mediated by insulin, is orchestrated to occur at lipid rafts, where cholesterol is highly concentrated. Less membrane cholesterol results in fewer lipid rafts, and this leads to impaired glucose uptake. Indeed, it has been proposed that statins increase the risk to diabetes (Goldstein and Mascitelli, 2010, Hagedorn and Arora, 2010). Our data bear out this notion, with the probability of the observed distributions of diabetes references happening by chance being only 0.006.)