Statins and Diabetes: Rethinking the Data
On February 28, 2012 the United States Food and Drug Administration (FDA) announced important safety label changes warning of an association between statin medications and increased glycosylated hemoglobin (HbA1c) and fasting serum glucose levels.(1) Clinicians and investigators are now questioning whether or not patients are “overdosing” on statins, thus putting them at risk for type 2 diabetes and other adverse complications.(2) Based on review of the current literature, we believe that the association between new-onset diabetes and statin therapy is real, however, the cause(s) of this association remain(s) unclear.
While the risk of developing incident diabetes is modest, it does not appear to blunt the overall cardioprotective effects of statins in high-risk primary and secondary prevention patients. In this article, we briefly review the key evidence linking statins with new-onset diabetes and highlight the relative strengths and weaknesses of existing studies. We also offer clinical insight to help health care providers weigh the potential risks and benefits of statin therapy.
Do Statins Really Cause Diabetes?
The FDA specifically cites data from Justification for the Use of Statins in Primary Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER) to support the new safety label changes. JUPITER is of particular significance since it is the largest statin primary prevention trial to date, randomizing 17,802 men and women free of baseline cardiovascular disease and diabetes to treatment with either rosuvastatin 20 mg daily or placebo.(3) The trial was stopped earlier than expected when interim analysis showed a 44% reduction in the primary endpoint in the rosuvastatin arm (HR 0.56; 95% CI 0.46-0.69). An unexpected outcome of JUPITER was a 26% (OR 1.26; 95% CI 1.04–1.51) increase in physician-reported new-onset diabetes in patients receiving rosuvastatin.(3)
While the increase in new-onset diabetes in JUPITER is surprising, there are several caveats in the trial design that limit generalization of the results. First, JUPITER subjects are described as “apparently healthy” individuals; however, a significant number of participants had metabolic syndrome at baseline (~40% in both trial arms) and 77% of patients in the rosuvastatin arm who developed diabetes had impaired fasting glucose at baseline.(3) Furthermore, the median HbA1c in both trial arms was 5.7% at entry – a level that the American Diabetes Association uses to identify “individuals with high risk for future diabetes” commonly referred to as “prediabetes.”(4) Glucose intolerance, elevated HbA1c, metabolic syndrome, and elevated hs-CRP are all risk factors for developing diabetes thus putting the JUPITER cohort at high-risk for new-onset diabetes regardless of whether or not they received statin therapy.
A major limitation of the JUPITER data that must be taken into account is the fact that there was no standardized ascertainment protocol for detecting cases of new-onset diabetes. Furthermore, new-onset diabetes events were not adjudicated by the end-point committee.(3) This lack of pre-specified screening protocols might explain the difference in rates of new-onset diabetes in the two trial arms. JUPITER data may also be skewed as a result of frequent testing and/or screening biases that are known to occur for patients on study protocols.
The existence of these biases is evident in discordant trial data. For example, there was an increase in physician reported diabetes and median HbA1c values (5.9% in rosuvastatin and 5.8% in placebo groups respectively, p=0.001) at 24 months, yet protocol-specified measurements showed no significant difference in fasting blood glucose level (98 mg/dL in both groups, p=0.12) or new-onset glycosuria (36 versus 32 patients in rosuvastatin and placebo groups respectively, p=0.64).(3) Due to these limitations in trial design, we must remember that data from JUPITER indicate an association between rosuvastatin and new-onset diabetes but do not demonstrate causality.
A subsequent 2010 meta-analysis by Sattar et al. sought to further characterize the association of new-onset diabetes with statin therapy by pooling published and unpublished data of 91,140 patients in 13 randomized controlled trials (RCTs) including JUPITER. Compared to matched-controls, patients treated with statin medications had a 9% (OR 1.09; 95% CI 1.02–1.17) increased risk of developing diabetes with little heterogeneity between trials (I2=11%).(5) This data corroborate that of smaller 2009 meta-analyses demonstrating a 13% (RR 1.13; 95% CI 1.03–1.23) increase in the risk of new-onset diabetes associated with statin therapy.(6)
Of the trials included in the Sattar et al. meta-analysis, six (HPS, JUPITER, CORONA, MEGA, AFCAPS/TexCAPS, 4S) either partially or completely relied on non-standardized physician reporting as a means of data acquisition for incident diabetes. Interestingly, when investigators analyzed only those trials utilizing pre-specified fasting glucose measurements to diagnose diabetes (n = 75,033), the risk of new-onset diabetes was attenuated and no longer achieved statistical significance (OR 1.07; 95% CI 0.97–1.17; I2=32%). Similar to JUPITER, data from Sattar et al. indicate an association between new-onset diabetes and statin therapy but do not offer proof of causality.
Statin Potency and the Risk of New-Onset Diabetes
In their online safety communication, officials at the FDA also allude to relative differences in the risk of hyperglycemia associated with high-dose statin therapy compared with moderate–dose therapy. They cite data from Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 Investigators (PROVE-IT TIMI 22)(7) which randomized 4,162 patients with acute coronary syndrome to treatment with either moderate-dose statin therapy (pravastatin 40 mg daily) or intensive-dose therapy (atorvastatin 80 mg daily).
In a post-hoc analysis, Sabatine et al. noted a larger increase in HbA1c values for patients treated with atorvastatin compared with pravastatin (0.37% versus 0.18% increase respectively). Those treated with atorvastatin also had an increased risk of developing an A1C >6% compared with pravastatin (RR 1.84; 95% CI 1.52–2.22).(8) A subsequent meta-analysis of 5 RCTs by Preiss et al. demonstrated a 12% increased risk (OR 1.12; 95% CI 1.04-1.22; I2=0%) of new-onset diabetes associated intensive-dose statin therapy as compared with moderate-dose therapy.(9) Once again, findings in both these studies are limited since diabetes was not rigorously screened for in trial participants.
Ignoring the Elephant in the Room: Metabolic Syndrome
A key 2011 study by Waters et al. analyzed the risk of new-onset diabetes in 3 trials of high-dose atrovastatin.(10) Two of the three trials (TNT and IDEAL) compared atorvastatin 80 mg with moderate-dose therapy (atorvastatin 10 mg and simvastatin 20 mg respectively), while a third trial compared atorvastatin 80 mg with placebo (SPARCL). Both TNT and IDEAL showed a trend toward increased new-onset diabetes associated with high-dose atorvastatin (for TNT, HR 1.10; 95% CI 0.94-1.29; for IDEAL, HR 1.19; 95% CI: 0.98-1.43) while SPARCL demonstrated a statistically significant increase in new-onset diabetes compared with placebo (HR 1.37; 95% CI 1.08-1.75).(10)
These data are similar to those presented above in that they reiterate the association between statin therapy and new-onset diabetes. What makes this study unique is that investigators go on to identify 4 independent predictors of new-onset diabetes including impaired fasting glucose, obesity, elevated fasting triglycerides and hypertension (all of which are defining features of the metabolic syndrome).(10) Interestingly, in patients with 0 or 1 of these risk factors, the risk of developing new-onset diabetes was relatively low and there was no statistical difference in the rate of new-onset diabetes between the high-dose atrovastatin arm and controls. Conversely, in those patients with 3 or 4 risk factors, the incidence of new-onset diabetes increased in both trial arms, but was more pronounced in those treated with high-dose therapy. These data suggest that the increase in risk of new-onset diabetes associated with statin medications may actually be driven by underlying metabolic syndrome.
Implications for Clinical Practice
Many questions have been raised as to whether or not the potential harms of new-onset diabetes outweigh the potential benefits of statin therapy. To date there does not appear to be an increase in mortality or major adverse cardiovascular events in those who develop new-onset diabetes. Data from the 3 high-dose atorvastatin RCTs indicate that major cardiovascular events occurred in 11.3% of patients with and 10.8% of patients without new-onset diabetes (HR 1.02; 95% CI: 0.77-1.35).(10) If anything, those patients who developed new-onset diabetes while taking statins enjoyed a lower rate of major cardiovascular events compared with those patients who had diabetes at baseline (10.1% and 17.5% respectively; p-value not reported).
In their meta-analysis, Sattar et al. note that in order to cause one new case of diabetes, 255 (95% CI 150-852) patients would have to be treated with statin for 4 years.(5) Conversely, the Cholesterol Treatment Trialists (CTT) meta-analysis demonstrated that reduction in LDL-C by 1mmol/L with statin therapy could prevent 5.4 major coronary events (defined as coronary heart disease deaths and non-fatal myocardial infarctions) for each predicted case of new-onset diabetes.(11) The benefits of statin therapy are even greater when taking into account reductions in strokes, revascularizations and hospitalizations.(5)
The same cardioprotective effects hold true in cases of high-dose statin therapy. Data from Preiss et al. indicate a 16% (OR 0.84; 95% CI 0.75-0.94) decrease in cardiovascular events in patients treated with intensive-dose compared with moderate-dose therapy.(9) The authors estimated that one new case of diabetes developed for 498 people treated. On the other hand, treatment of 155 patients was needed to prevent one cardiovascular event. Clearly, the scale remains tipped in favor of statins.
Similar to statins, thiazide diuretics have been associated with worsened glycemic control and new-onset diabetes.(12,13) Interestingly, long-term follow-up (mean of 6.9 years) from ALLHAT indicate that patients who developed new-onset diabetes while taking chlorthalidone had no difference in risk of cardiovascular disease mortality (HR 1.04; 95% CI 0.74–1.47), all-cause mortality (HR 1.04; 95% CI 0.82–1.30), and non-cardiovascular disease mortality (HR 1.05; 95% CI 0.77–1.42) compared to those patients who did not develop incident diabetes.(14)
These data suggest that the long-term micro- and macrovascular complications typically associated with type 2 diabetes are not necessarily evident in those who develop thiazide-associated diabetes. While data and follow-up times are limited, the same also appears to be true for incident diabetes associated with statin use.
Take Home Points
There is an undeniable association between new-onset diabetes and statin medications; however, there is no definitive proof that statins are actually causing diabetes. While statins may be responsible for some degree of worsened glycemic control, what is more likely is that patients who develop incident diabetes have glucose intolerance at baseline and subsequently develop diabetes either as a natural progression of their disease or perhaps as a consequence of some as-of-yet uncharacterized property of statins. Based on existing data, there is no evidence demonstrating that development of statin associated diabetes results in increased risk of major cardiovascular events.
What is clear is that high-risk primary and secondary prevention patients stand to gain significant benefit from statin therapy, and the harms of new-onset diabetes are outweighed by reductions in cardiovascular morbidity and mortality.(11,15,16) Rather than discontinue statin medications, we believe that health care providers would do better to emphasize weight reduction, healthy diet and aerobic exercise as a means of improving glycemic control and overall cardiovascular health in patients who develop new-onset diabetes.
We also recommend close follow-up and regular screening of individuals with evidence of metabolic syndrome, especially those on intensive-dose therapy. Our viewpoint corroborates that of the FDA which, in spite of the new safety label warnings, “continues to believe that the cardiovascular benefits of statins outweigh these small increased risks.”(1)
- Food and Drug Administration. FDA Drug Safety Communication: Important safety label changes to cholesterol-lowering statin drugs. [Internet]. http://www.fda.gov/Drugs/DrugSafety/ucm293101.htm. 2012 [cited 2012 Mar. 1]
- Topol EJ. OP-ED CONTRIBUTOR; The Diabetes Dilemma for Statin Users. The New York Times: The Opinion Pages. 2012; Last Accessed April 3, 2012. http://www.nytimes.com/2012/03/05/opinion/the-diabetes-dilemma-for-statin-users.html
- Ridker PM, Danielson E, Fonseca FA, et al. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med 2008; 359:2195–2207.
- American Diabetes Association. Standards of medical care in diabetes--2011. Diabetes Care 2011;34 Suppl 1:S11–61.
- Sattar N, Preiss D, Murray HM, et al. Statins and risk of incident diabetes: a collaborative meta-analysis of randomised statin trials. Lancet 2010; 375:735–742.
- Rajpathak SN, Kumbhani DJ, Crandall J, Barzilai N, Alderman M, Ridker PM. Statin therapy and risk of developing type 2 diabetes: a meta-analysis. Diabetes Care 2009; 32:1924–1929.
- Cannon CP, Braunwald E, McCabe CH, et al. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004; 350:1495–1504.
- Sabatine M, Wiviott S, Morrow D, McCabe C, Cannon C. High-dose atorvastatin associated with worse glycemic control: a PROVE-IT TIMI 22 substudy. Circulation 2004; 110(suppl III):834.
- Preiss D, Seshasai SRK, Welsh P, et al. Risk of incident diabetes with intensive-dose compared with moderate-dose statin therapy: a meta-analysis. JAMA 2011; 305:2556–2564.
- Waters DD, Ho JE, DeMicco DA, et al. Predictors of new-onset diabetes in patients treated with atorvastatin: results from 3 large randomized clinical trials. J Am Coll Cardiol 2011; 57:1535–1545.
- Baigent C, Keech A, Kearney PM, et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet 2005; 366:1267–1278.
- Gupta AK, Dahlof B, Dobson J, et al. Determinants of new-onset diabetes among 19,257 hypertensive patients randomized in the Anglo-Scandinavian Cardiac Outcomes Trial--Blood Pressure Lowering Arm and the relative influence of antihypertensive medication. Diabetes Care 2008; 31:982–988.
- Elliott WJ, Meyer PM. Incident diabetes in clinical trials of antihypertensive drugs: a network meta-analysis. Lancet 2007; 369:201–207.
- Barzilay JI, Davis BR, Pressel SL, et al. Long-Term Effects of Incident Diabetes Mellitus on Cardiovascular Outcomes in People Treated for Hypertension: The ALLHAT Diabetes Extension Study. Circ Cardiovasc Qual Outcomes 2012; 5:153–162.
- National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421.
- Minder CM, Blaha MJ, Tam LM, et al. Making the case for selective use of statins in the primary prevention setting. Arch Intern Med 2011; 171:1593–1594.
Keywords: Diabetes Mellitus, Type 2, Glucose, Hemoglobin A, Glycosylated, Hydroxymethylglutaryl-CoA Reductase Inhibitors, United States Food and Drug Administration
< Back to Listings