Statins or HMG-CoA reductase inhibitors provide several cardiovascular benefits in addition to lowering cholesterol. This could lead individuals to believe that statins may potentially aid in reducing diabetes risk. However, in numerous cardiovascular disease (CV) prevention studies, it has been consistently found that diabetes risk is increased with statin therapy. Because diabetes is not usually a direct measure in these CV disease studies, participants are often low-risk.
The following study aimed to evaluate the effect of statin therapy on diabetes patients who are considered high-risk. Population data was analyzed from a 3-year study called the Diabetes Prevention Program (DPP), and an extension of this study called the DPP Outcomes Study (DPPOS). The DPP is a randomized, controlled trial that studied the effect of lifestyle changes, metformin use, and placebo on high-risk patients with obesity or overweight. There were 3,234 participants randomized to receive 1 of the 3 interventions. Participants were included if they were older than 25 years of age, had obesity or were overweight, had high fasting blood sugar levels, and had impaired glucose tolerance. Following the DPP, participants were given the option to join the DPPOS extension study.
In both the DPP and DPPOS, use of statins and other medications was obtained through patient self-report at baseline and twice yearly at follow-up visits. Statin therapy along with hypertensive therapy was determined by the participants’ primary physicians outside of the study. Lipid panels and blood pressure were recorded once yearly. Diabetes was diagnosed using a 75 g oral glucose tolerance test once yearly, or by obtaining fasting plasma glucose levels twice yearly. Cox proportional hazard models were utilized to determine the time-dependent relation between the use of statins and risk of developing diabetes on the DPP/DPPOS population.
Results at 10 years of follow-up show that the use of statins before a diabetes diagnosis was not statistically significant among the 3 interventions. Statin use prior to diabetes diagnosis was 33% in the lifestyle intervention group, 37% in the metformin intervention group, and 35% in the placebo intervention group (P=0.36). 40% of participants were taking simvastatin, 37% were taking atorvastatin, 9% were taking lovastatin, and 8% were taking pravastatin. The use of statins increased throughout the study, becoming more prevalent after diabetes diagnosis.
It was found that risk of developing diabetes was elevated in the participants using statins in all 3 interventions. The combined hazard ratio (HR) for the 3 interventions was 1.36 (95% Cl, 1.17 to 1.59). The study also assessed statin use duration and its association to diabetes risk. Diabetes risk was increased in participants who had been using statins for a longer period of time, with higher risk in the lifestyle intervention group. The HR per visit with statin use for the lifestyle intervention was 1.06 (1.02 to 1.11), P=0.007). In the metformin intervention, the HR was 1.01 (0.96 to 1.06). And finally, the HR for the placebo intervention was 1.02 (0.97 to 1.07). Low vs. high potency statins were also evaluated in relation to diabetes risk and no difference was found with an HR of 0.96 (0.68 to 1.35).
Mechanisms behind how statins can potentially increase diabetes were also studied through insulin sensitivity and insulin secretion analysis. The Insulinogenic Index, a measure of insulin secretion, was statistically significant between statin users and non-statin users. Insulin secretion decreased in statin users and increased in non-statin users (P=0.013). There were no significant changes in fasting insulin values suggesting that statins have little to no effect on insulin sensitivity.
Overall, this study showed that diabetes risk is increased in hig- risk patients who use statins. Although this study provides evidence that statins reduce insulin secretion, mechanisms behind this finding are not clear and need to be studied further. Limitations in this study include lack of randomization of participants using statins, statin use was confirmed through patient self-report, and statin dose in relation to diabetes risk was not evaluated due to limited access to this information.