This topic is very close to my heart as a “preventative, non-invasive” cardiologist. I have 20 years practical experience in both primary prevention (in heathy individuals) and aggressive secondary risk modification (in those with established vascular disease) and have a database of ~ 10 000 individuals treated with both weak and potent Statin therapy, in high and low doses over many years.
The association of statin use and new onset diabetes mellitus has been known since 2010 but only hit the headline with 2 major publications in 2012 and the FDA announcement that clinicians should be aware of this association (February 2012).
The debate has centred on the following:
- Is this a real effect of statins?
- Is there “cause and effect” or is the association driven by confounding factors?
- What is the clinical relevance?
- Should we in any way change our practice of prescribing (high dose potent) statins?
Ultimately anything we do in life shares this very important paradigm of RISK versus BENEFIT and the costs involved. Driving an incredibly dangerous weapon (your motor car) at slow or high-speed carries profound risk-benefit. Investing in your future economic well-being and preventing the commonest disease process on planet earth has exactly the same implications. Investing in your cardiovascular health through prevention of atherosclerosis of your arteries requires dedicated attention to detail life-long. The one thing we do know is that attainment of “ideal cardiovascular health” is extremely difficult to achieve just with lifestyle and diet.
The prevalence of “sub clinical” atherosclerosis and cardiovascular disease is equally VERY high in “so-called” healthy 50 year individuals.
So what are the risks of diabetes with statin therapy?
Statins have been available since the 1980s but their risk of inducing diabetes did not surface for nearly 20 years. When all the data available from multiple studies was pooled in 2010 for more than 91,000 patients randomly assigned to be treated with a statin or a sugar pill (placebo), the risk of developing diabetes with any statin was one in every 255 patients treated.
The weaker statins like pravastatn introduced in the early 1990’s seemingly do not carry any clear-cut risk. It is only with the more potent statins like Zocor (simvastatin), Lipitor (atorvastatin) and Crestor (rosuvastatin), particularly at higher doses, that the risk of diabetes shows up. The cause-and-effect is unequivocal, because the multiple large trials of the more potent statins had a consistent excess of diabetes.
Overall the incidence of new onset diabetes mellitus is 9-12% over 5 years of statin therapy with high dose more potent statins offering the higher risk. This has to be balanced by the benefit in cardiovascular reduction over the same period of 5 years ~25-50% or with potent high dose Statins 20% risk reduction per 1 mmol/l reduction LDL.
Essentially a nice way of looking at risk verses benefit is to look at the number of people needed to treat (NNT) to prevent ONE MAJOR cardiovascular event (CVE) verses the number needed to treat to HARM (NNH) one patient (in this case a new onset of diabetes).
With statin therapy in primary prevention the NNT is 90 people over 2 years & 65 over 5 years. With secondary prevention the NNT is 150 over one year and only 10-40 over 5 years to save one major CVE. The NNH is 250 over a 5 year; 300 over 2 year & 500 over 1 year. In other words we will generally see one new case of diabetes in 500 treated with generally a high potency high dose statin.
With potent statins like rosuvastatin (Crestor) and high dose atorvastatin (Lipitor) the benefit in reducing MCE is very impressive due to plaque reduction and plaque stabilisation.
For every 1 mmol of LDL reduced (with statins) there is approximately 20% risk reduction independent of the risk of the person (<5% to high risk > 30%).
Of importance is the typical phenotype predicting which patients develop diabetes on statin therapy. Those individuals with “metabolic syndrome” with elevated fasting blood sugar, increase visceral adiposity, high normal blood pressure or hypertension, and elevated triglycerides are those at risk for developing diabetes on statins. If patients had four of these risk factors, they have a 25% risk of developing diabetes. If they have none of these risks factors, their risk was just 2%. That is not surprising, as we know these are just the factors that predict new onset diabetes.
The possible confounding factors relating statins to new onset diabetes are also relevant. Many people who take statins to lower atherogenic lipoproteins (LDL) rely on the “pill approach” rather than a holistic approach with correct lifestyle, exercise and nutrition and as I see every day the risks for progressive metabolic syndrome are exacerbated despite statin therapy. In my practice we focus on all 10 of the parameters that drive cardiovascular disease and consequently I see extremely few cases of diabetes on statin therapy:
- ApoB/ ApoA1 (total atherogenic/ non atherogenic lipoproteins)
- Ultrasensitive CRP < 1.0
- Non smoking of all tobacco products
- BMI < 25%
- Optimal nutrition tailor prescribed for the individual phenotype
- Moderate intensity exercise on most daily/ week
- BP ~ 120/80
- GLucose < 6.0
- Alcohol in moderation
- Stress reduction
Remember the cumulative effect of statin therapy on our arteries over many decades is profound and forms the basis of the philosophy of “optimal medical management.” This accounts for the incredible outcome in trials like STENO-2; BARI-2D and COURAGE where major cardiac event rate was impressively reduced compared to conventional management and even invasive (coronary stents) or surgical intervention (coronary bypass surgery).
So in summary:
The consistency of the statin clinical-trial findings, the temporal association and the dose-dependency of the effect all provide a strong argument for causality for statins and diabetes.
Whilst the precise mechanism for cause-and-effect is NOT known further work is needed to elucidate the mechanisms of action at hepatic, skeletal-muscle, adipocyte, beta-cell, and mitochondrial levels that could impair insulin action and glucose homeostasis. There is certainly no evidence to support statin worsening insulin resistance nor evidence to suggest increasing risk via accelerated oxidised LDL due to statin.
The bottom line is a fairly simple, straightforward clinical message. Even in a lower-risk primary-prevention population, the cardiovascular benefits of statin therapy outweigh the diabetes hazard, even among those with highest risks for diabetes (metabolic syndrome).
With the incredible benefit and low NNT in secondary prevention the risk-benefit ratio greatly favors benefit in treating patients to goal (generally LDL < 1.8 and ApoB/ ApoA1 0.3- 0.5) with high potency high dose statin therapy.
In other words the clinical RELEVANCE of new onset of diabetes mellitus with statin therapy is minimal as the benefit of major cardiac event reduction outweighs any proposed negative effect of the diabetes.