Level 1: Likely reliable evidence

Statins are widely used for treating elevated cholesterol for both primary and secondary prevention of coronary artery disease. One concern about their use has been the possibility of increasing the risk of diabetes.

Last year, a randomized trial showed that after two months of treatment with atorvastatin (Lipitor), patients had increases in fasting plasma insulin, insulin resistance and hemoglobin A1c levels (J Am Coll Cardiol. 2010;55:1209-1216), and a systematic review found an increase in the risk of new onset of diabetes of 0.4% with statin use (number needed to harm =255) (Lancet. 2010;375:735-742).

The increased risk of new-onset of diabetes was confirmed in a subgroup analysis of patients from the SPARCL trial (J Am Coll Cardiol. 2011;57:1535-1545). This study found new-onset diabetes in 8.71% with atorvastatin 80 mg/day vs. 6.06% with placebo (P<0.05; NNH=38).

Now, a pooled analysis of five lipid-lowering trials using statins (atorvastatin or simvastatin) with 32,752 patients has found an increased risk of diabetes with high-dose therapy compared with moderate-dose therapy (JAMA. 2011;305: 2556-2564). Diabetes was diagnosed if it was listed as an adverse event, if there were two or more fasting blood glucose levels >125 mg/dL or if the patient was started on a glucose-lowering medication. The overall rate of new-onset diabetes was 8.4%.

Comparing intensive-dose statin vs. moderate-dose statin, the rate of diabetes was 8.8% vs. 8% (OR=1.12; 95% CI:1.04-1.22). By way of comparison, cardiovascular events occurred in 19.1% of patients taking intensive doses and in 21.7% taking moderate doses (OR=0.84; 95% CI: 0.75-0.94). Overall, intensive-dose statin use had a NNH of 498 per year for the risk of new-onset diabetes and a number needed to treat (NNT) of 155 per year for the benefit of decreased cardiovascular events.