Statins are the cornerstone of dyslipidemia management, lowering lowdensity lipoprotein (LDL) cholesterol levels and substantially reducing the risk for cardiovascular disease (CVD).
Statins are the cornerstone of dyslipidemia management, lowering lowdensity lipoprotein (LDL) cholesterol levels and substantially reducing the risk for cardiovascular disease (CVD). However, many patients on statins have CVD despite reductions in LDL cholesterol.1 Furthermore, the cooccurrence of dyslipidemia, type 2 diabetes, and hypertension is common.2 A considerable body of evidence demonstrates substantial benefits of statins in terms of reduction of major cardiovascular (CV) events, stroke, and mortality in patients with or without diabetes.3 Higher doses and more intensive (i.e. more potent at LDL reduction) statin therapy are more effective than lower doses in both improving lipid levels and reducing adverse CV events.4 However, statin therapy has been demonstrated to increase the risk for incident type 2 diabetes,5 more so at higher doses.6,7 Type 2 diabetes and the metabolic syndrome are strong, independent risk factors for CVD.8–10 As a result, the US Food and Drug Administration (FDA) has added an adverse warning to statin labels. Seven statins are approved by the FDA: simvastatin, lovastatin, fluvastatin, atorvastatin, rosuvastatin, pravastatin, and pitavastatin. This article aims to examine the differential metabolic effects among the various statins, the clinical evidence outlining the effect of statins on glucose tolerance, the results of recent meta-analyses on the risk for statins and incident diabetes and, finally, discuss how clinical data influences dyslipidemia management in patients with a high risk for type 2 diabetes.
Statin Pharmacokinetics and Efficacy
The pharmacokinetics of individual statin drugs are influenced by their lipophilic or hydrophilic nature. Atorvastatin, lovastatin, simvastatin, and pitavastatin are lipophilic molecules; however, they are not all metabolized in the same way. Atorvastatin, lovastatin, and simvastatin undergo extensive metabolism via the cytochrome P450 isoenzyme CYP3A4. A study of >950,000 patient records from two US databases in 2005–6 showed that 83 % of patients with dyslipidemia used a CYP3A4-metabolized statin and that, of these, 25 to 30 % also received a CYP3A4 inhibitor,11 suggesting that there is a need for physician education regarding the impact of these inhibitors on the metabolism of lovastatin, simvastatin, and atorvastatin. The most recently introduced statin, pitavastatin, has a characteristic structure consisting of a heptenoate basic structure with a quinoline ring, moieties of fluorophenyl, and a cyclopropyl side chain (see Figure 1), which provides optimal activity and enhanced drug absorption. Unlike other lipophilic statins, pitavastatin is predominantly metabolized via glucuronidation and only minimally by CYP450, marginally by CYP2C9 and to a lesser extent CYP2C8. Pravastatin, fluvastatin, and rosuvastatin are relatively hydrophilic. Rosuvastatin is minimally metabolized by CYP2C9; fluvastatin is metabolized via CYP2C9 and pravastatin is not metabolized via cytochrome pathways.
While no direct comparisons of the efficacy of all of the various statins exist, the following commonly prescribed dosage levels are considered to be therapeutically equivalent, with the ability to reduce LDL cholesterol (LDL-C) by 20–30 %: atorvastatin 10 mg, rosuvastatin 5 mg, fluvastatin 80 mg, lovastatin 40/80 mg pravastatin, and simvastatin 20 mg.12,13 Pitavastatin 4 mg has been found to be significantly more effective than pravastatin 40 mg in terms of reduction of LDL-C (see Figure 2).14–16 However, there is a lack of randomized controlled trial data regarding CV morbidity and mortality with pitavastatin.
Effect of Statins on Blood Glucose
On February 28, 2012, the FDA published a safety communication to alert the public and healthcare providers to changes in the prescribing information for statins.17 The adverse event information has been updated to include the potential for increased blood sugar and glycated hemoglobin (HbA1c). Clinical studies, however, have yielded inconclusive results on the association between statin usage and blood sugar. A large-scale study showed that pravastatin reduced the rate of new-onset diabetes by 30 %,18 while, conversely, multiple studies have implicated atorvastatin, rosuvastatin, and simvastatin in increasing both HbA1c and fasting plasma glucose (FPG) levels.19–22
Several mechanisms have been proposed to explain the increased risk for diabetes in patients taking statins. The increased production of plasma-derived LDL cholesterol in response to the statin-induced inhibition of cholesterol synthesis might result in direct inflammation and oxidation within the b-cell, resulting in cellular apoptosis and impaired insulin secretion.23 Other potential mechanisms include the effects of statins on hydroxymethylglutaryl-CoA reductase inhibition, isoprenoid synthesis, calcium (Ca2+) release, glucose transport, Ca2+-mediated pancreatic insulin secretion, decreasing various isoprenoids that enhance glucose uptake via glucose transporter 4 (Glut 4) in adipocytes, decrease in adiponectin, and/or insulin resistance.23,24
Preclinical and clinical data have demonstrated the effect of statins on various parameters of glycemic control. Atorvastatin attenuates adipocyte maturation and SLC2A4 expression by inhibiting isoprenoid synthesis, and impairs glucose tolerance.25 Simvastatin has been demonstrated to block L-type Ca2+channels resulting in decreased insulin secretion.26 Lovastatin was shown to downregulate insulin-responsive Glut 4 and upregulate Glut 1 in 3T3-L1 adipocytes, leading to inhibition of insulin-stimulated glucose transport.27 Atorvastatin has been demonstrated to reduce sensitivity to insulin in rats with streptozocin-induced diabetes.28 A systematic literature review concluded that statins had no significant impact on insulin sensitivity compared with placebo (standardized mean difference [SMD] –0.084; 95 % confidence interval [CI] –0.210 to 0.042; p=0.19). Pravastatin was associated with improved insulin sensitivity (SMD 0.342, 95 % CI 0.03–0.621; p=0.03), whereas simvastatin significantly impaired insulin sensitivity (SMD –0.321; 95 % CI –0.526 to –0.117; p=0.03).29
It has been suggested that lipophilic and hydrophilic statins may havedifferent effects on glycemic parameters.19 Insulin secretion has been demonstrated to decrease in MIN6 b-cell lines treated with simvastatin and atorvastatin but not with hydrophilic pravastatin.30 The Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 (PROVE-IT TIMI 22) trial found that although 80 mg atorvastatin and 40 mg pravastatin were both associated with a small increase in HbA1c (atorvastatin 0.37 % and pravastatin 0.18 %), atorvastatin significantly increased the risk for developing HbA1c >6 % compared with pravastatin (relative risk [RR] 1.84; 95 % CI 1.52–2.22; p<0.0001).31
Statins decrease levels of ubiquinone (CoQ10), especially at higher doses, which delays insulin release by decreasing levels of pancreatic b-cells.7,32 It has been postulated that, as a result of decreased levels of CoQ10, mitochondrial glycerol-3-phosphate dehydrogenase (G3PD) is underexpressed, and that correction of suboptimal CoQ10 status, by supplementation, may improve the glucose-stimulated insulin secretion of b-cells and thereby decrease postprandial glucose.33
Studies of the effect of statins on insulin levels have noted the association between changes in adipocytokine levels, including leptin and adiponectin. Adiponectin is a hormone secreted by adipocytes that reduces insulin resistance and suppresses fatty acid-induced b-cells apoptosis.34 A study found that other statins tend to decrease adiponectin levels, although this study reported no effect on insulin sensitivity.35 Other studies have reported inconsistent effects of statins on adiponectin levels.25,36–40 Unlike the other statins, pitavastatin consistently raises adiponectin levels, which in turn lowers insulin resistance and improves insulin secretion.20,41–43 This might explain the observation that, to date, pitavastatin has not precipitated new-onset diabetes or decreased glycemic control.44,45 Pitavastatin has also been shown to increase glucose uptake, improve insulin sensitivity, and the response to intraperitoneal insulin in KKAy mice.46
These data do not provide a clear explanation of the effects of statins on glycemic parameters. Evidence suggests that pravastatin and pitavastatin are the two statins least associated with worsening insulin sensitivity. However, a unifying mechanism to explain the association between statins and diabetes risk has not yet been identified.
Clinical Studies of the Effect of Statins on Glucose Tolerance
Clinical studies have yielded conflicting results regarding glycemic control and the incidence of new-onset diabetes in patients taking statins, although taken as a whole, the data suggest an association between statin therapy and new-onset diabetes (see Table 1).47 Nevertheless, caution should be used in interpreting these data. The studies were not designed to detect incident diabetes and therefore were not statistically powered to evaluate this outcome; the definition of diabetes varies across trials; and some studies did not rigorously screen for new-onset diabetes. Furthermore, baseline populations differed considerably between trials, and therefore have varying risk factors for diabetes.
Atorvastatin
A significant positive relationship has been reported between the dose of atorvastatin (10, 20, 40, and 80 mg) with fasting plasma insulin and HbA1c levels, together with an inverse relationship with insulin sensitivity.21,24 The Incremental Decrease in End Points through Aggressive Lipid Lowering (IDEAL) and Anglo-Scandinavian Cardiac Outcomes Trial – Lipid Lowering Arm (ASCOT-LLA) studies have found nonsignificant increased relative risks for diabetes in patients taking atorvastatin.48–50 High-dose atorvastatin treatment is also associated with a slightly increased risk for new-onset diabetes. A post-hoc analysis of clinical trials found that in the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, incident diabetes was higher in patients randomized to atorvastatin 80 mg/day compared with placebo (RR 1.37; 95 % CI 1.08–1.75).51 In the Treating to New Targets (TNT) study, a higher incidence of new-onset diabetes was also found in atorvastatin 80 mg/day compared with 10 mg/day, but this was not statistically significant.51
Rosuvastatin
In the Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER) trial comparing 20 mg of rosuvastatin to placebo, a statistically significant increase in physician reported type 2 diabetes was noted (p=0.01).52 The Controlled Rosuvastatin Multinational Study in Heart Failure (CORONA) study reported a nonsignificant risk for developing diabetes while taking rosuvastatin (RR 1.15; 95 % CI 0.91–1.44).53
Pravastatin
The West of Scotland Coronary Prevention Study (WOSCOPS) trial identified pravastatin as a protective rather than deleterious factor in the development of diabetes (RR 0.70; 95 % CI 0.50–0.99), reducing the onset of diabetes by 30 %.18 However, these analyses were based on a nonstandardized definition of diabetes (FPG >126 mg/dl plus a 36 mg/dl increase from baseline). When a subsequent meta-analysis reanalysed these data using a more standardized definition, a smaller effect was found.5 Furthermore, the patient characteristics in the WOSCOPS study differed from other studies in that it did not enrol female participants, and the body mass index (BMI) among participants averaged 25.9 mg/ m2 whereas the mean BMI exceeded 27 kg/m2 in other trials.52,53 In the Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) trial, there was a 32 % higher incidence of diabetes with pravastatin therapy (RR 1.32; 95 % CI 1.03–1.69), particularly in those with the metabolic syndrome.54 The Long-term Intervention with Pravastatin in Ischemic Disease (LIPID) trial reported no significant effect of pravastatin on the development of diabetes (RR 0.95; 95 % CI 0.77–1.16).55
Simvastatin
The Heart Protection Study (HPS) of simvastatin 40 mg/day reported a nonsignificant trend towards an increased development of diabetes (RR 1.14; 95 % CI 0.98–1.33).56 The Study of the Effectiveness of Additional Reductions in C, Homocysteine (SEARCH) trial found a small increased incidence of new-onset diabetes with high-dose (80 mg) compared with low-dose (20 mg) simvastatin (RR high dose versus low simvastatin, 1.07; 95 % CI 0.95–1.19).57 A dose-dependent reduction in insulin sensitivity was reported in patients in treated with simvastatin (10, 20, 40, or 80 mg daily) versus placebo.20
Pitavastatin
Numerous studies have concluded that pitavastatin has a neutral and possibly beneficial effect on glycemic control. A study comparing the longterm efficacy and tolerability of pitavastatin compared with atorvastatin in patients with type 2 diabetes found no significant differences in their effects on LDL-C; however, unlike atorvastatin, pitavastatin had no significant effect on FPG.58 In a retrospective analysis that compared the effects of atorvastatin 10 mg/day, pravastatin 10 mg/day, and pitavastatin 2 mg/day on glycemic control in patients with type 2 diabetes during a 3-month treatment period, only atorvastatin significantly increased FPG and HbA1c. Neither pravastatin nor pitavastatin had any significant influence on FPG and HbA1c.44
In the extension study of the LIVALO Effectiveness and Safety (LIVES) postmarketing surveillance study, patients (n=6,580) with hypercholesterolemia were treated with pitavastatin 1, 2, or 4 mg/day over 5 years. Pitavastatin was demonstrated to significantly decrease HbA1c in patients with diabetes (see Figure 3).59 In a recent study, patients aged ≥65 years with primary hyperlipidemia or mixed dyslipidemia were given pitavastatin (1 mg, 2 mg, and 4 mg) or pravastatin (10, 20, or 40 mg). There were no significant changes over 12 weeks in FPG within each treatment group.60 Pitavastatin and pravastatin also had no effect on FPG over 6 months in patients with the metabolic syndrome who had multiple risk factors for diabetes.61 In the HIV-infected Patients and Treatment with PItavastatin versus Pravastatin for Dyslipidemia (INTREPID) phase IV, multicenter, 12- week, randomized, double-blind study (n=252), neither pitavastatin 4 mg nor pravastatin 40 mg had any significant effect on FPG or HbA1c levels. However, pitavastatin 4 mg was superior to pravastatin 40 mg in terms of LDL-C reduction (see Figure 2).14 Similar results were demonstrated prospectively in the Pitavastatin Compared with Pravastatin In Lowering LDL-C in the US (PREVAIL US) study comparing pitavastatin 4 mg with pravastatin 40 mg in patients with primary hyperlipidemia or mixed dyslipidemia, neither statin had clinically relevant effects on glucose metabolism.15
Post-hoc analyses of phase III noninferiority trials comparing pitavastatin with other statins have evaluated the effect of statins on FPG. In a posthoc analysis of a 12-week randomized multicenter, double-blind trial of pitavastatin 4 mg versus atorvastatin 20 mg, no significant changes in FPG were noted with pitavastatin while atorvastatin significantly increased FPG in patients with type 2 diabetes and combined dyslipidemia.62 In a post-hoc analysis of a 12-week, randomized, multicenter, doubleblind trial of pitavastatin 4 mg versus simvastatin 40 mg, no significant changes in FPG were noted with pitavastatin. No significant changes in FPG from baseline were observed with pitavastatin at 12 weeks (mean change –0.03 mg/dl; p=0.963) or at 56 weeks (2.2 mg/dl; p=0.060). With simvastatin, FPG was unchanged at 12 weeks (–1.40 mg/dl; p=0.209), but was significantly increased (7.5 mg/dl; p=0.0001) at 56 weeks. Patients had primary hyperlipidemia or mixed lipidemia and ≥2 risk factors for coronary heart disease.63
Much of the data in support of the beneficial effect of pitavastatin on glycemic parameters have been obtained from relatively small, retrospective, and/or single-center studies. In order to provide more conclusive evidence, the Japan Prevention Trial of Diabetes by Pitavastatin in Patients with Impaired Glucose Tolerance (J-PREDICT) study is a prospective randomized, open-label controlled trial evaluating the effect of pitavastatin on the development of diabetes. Patients (n=1,269) with impaired glucose tolerance (IGT) were randomized to either pitavastatin (1–2 mg/day) plus lifestyle modification or the control group (lifestyle modification only). Preliminary data from this study showed that pitavastatin in combination with lifestyle modification was associated with a lower incidence of diabetes than was lifestyle modification alone in Japanese patients with IGT (see Figure 4).64
Meta-analyses of the Effect of Statins on Glucose Tolerance
These meta-analyses involved trials with differences in treatment regimen, patient characteristics, and duration of follow-up. Such differences must be considered when interpreting results. A meta-analysis of five clinical trials, including 39,791 patients, did not identify a significantly increased risk for diabetes in the group receiving statins. However, when the WOSCOPS trial data were removed from the analysis the RR of diabetes increased to 1.14 (95 % CI 1.02–1.28).65 A similar effect was noted in a meta-analysis of six clinical trials including 57,593 patients, which found a small increase in the risk for diabetes (RR 1.13; 95 % CI 1.03–1.23) if the WOSCOPS study was removed from the analysis. The RR was lower (1.06) and not statistically significant when WOSCOPS was included.47 A larger meta-analysis of 13 major statin trials including 91,140 participants identified a 9 % increased risk for the development of diabetes in those taking statins (RR 1.09; 95 % CI 1.02–1.17).5 Physician-reported diagnosis as opposed to standardized diagnostic criteria may overstate this effect. When FPG levels alone were used as the definition for diabetes, the risk was no longer statistically significant (RR 1.07; 95 % CI 0.97–1.17). This study did not find a clear difference between lipophilic statins (RR 1.10) and hydrophilic statins (RR 1.08). Nonsignificant trends were associated with atorvastatin (RR 1.14) and simvastatin (RR 1.11). For pravastatin the RR was 1.03; for lovastatin it was 0.98, which may suggest that the more potent statins have a stronger effect on diabetes risk.
A recent meta-analysis analyzed 17 randomized controlled trials reporting the incidence of new-onset diabetes during statin treatment between 1994 and October 2012, including a total of 113,394 patients. The study concluded that risk is both statin and dose dependent. Results demonstrated that pravastatin 40 mg/day was associated with the lowest risk for new-onset diabetes compared with placebo (RR 1.07, 95 % CI 0.86–1.30). Rosuvastatin 20 mg/day was associated with a 25 % increased risk for diabetes compared with placebo (RR 1.25; 95 % CI 0.82–1.90). Atorvastatin 80 mg/day had an intermediate risk for diabetes compared with placebo (RR 1.15; 95 % CI 0.90–1.50).66
Effect of Dose Regimen and Risk Factors at Baseline
Intensive dose statin therapy has been associated with reduced CV risk compared with low or moderate doses, but some studies have suggested that high-dose regimes confer a greater risk for diabetes. Compared with lowerdose statin therapy, atorvastatin 80 mg/day did not increase the incidence of new-onset diabetes in patients with 0 to 1 diabetes risk factors but a 24 % increased incidence was found among patients with 2 to 4 diabetes risk factors. The number of CV events was significantly reduced with atorvastatin 80 mg in both risk groups.67 This finding raises the possibility that statin therapy may increase the risk for diabetes only in those at higher baseline risk for developing the disease. In the JUPITER trial, 77 % of the patients in the rosuvastatin group who developed diabetes had impaired fasting glucose at baseline.52 Furthermore, more than 40 % of the patients in JUPITER met the criteria for the metabolic syndrome. The PROSPER trial, which also reported a higher incidence of new-onset diabetes with statin use, recruited an elderly patient population, who were therefore more likely to have a higher risk for diabetes compared with the normal population.5 However, further studies are required to confirm this hypothesis.
In 2011, a meta-analysis of 32,752 patients from five large-scale clinical trials examined the impact of the intensity of statin therapy on diabetes risk. There was a 12 % higher risk for developing new-onset diabetes in those treated with intensive dose statin therapy versus moderate dose statin therapy (RR 1.12; 95 % CI 1.04–1.22).6 However, a review of 5-year data from the Enhanced Feedback for Effective Cardiac Treatment (EFFECT) study found no association between higher diabetes risk and more intensive statin therapy. The risk for diabetes was in fact lower in the high-dose groups.68
Retrospective and Observational Studies
Observational studies have also failed to provide conclusive evidence regarding the association between statin use and diabetes incidence. A Women’s Health Initiative observational study recruited 161,808 women aged 50–79 years from 1993 to 1998. Results showed found that statin use at baseline was associated with an almost 50 % increased risk for diabetes (RR 1.48; 95 % CI 1.38–1.59).69 However, the study was limited by its observational nature, and the data set contained unequal representation of statins that may have influenced the outcomes. In addition, not all the currently available statins were in use during the study period, and women who took statins may have changed statin type prior to new onset.
In a Canadian population-based retrospective cohort study of more than 1.5 million people, compared with pravastatin (the reference drug in all analyses), there was an increased risk for incident diabetes with atorvastatin (RR 1.22, 95 % CI 1.15–1.29), rosuvastatin (RR 1.18; 95 % CI 1.10–1.26), and simvastatin (RR 1.10, 95 % CI 1.04–1.17). There was no significantly increased risk among people who received fluvastatin (RR 0.95, 95 % CI 0.81–1.11) or lovastatin (RR 0.99, 95 % CI 0.86–1.14).70 The risk for incident diabetes did not differ between patients with or without CVD. This study also found an association between development ofdiabetes and statin potency and dose.
Recommendations for Dyslipidemia Treatment in Patients with a Risk for Diabetes
New-onset diabetes on statin therapy has not yet been associated with an increase in adverse CV events. In an analysis of trials of atorvastatin therapy, major CV events occurred in 11.3 % of patients with newonset diabetes, 10.8 % without new-onset diabetes (hazard ratio [HR] 1.02; 95 % CI 0.77–1.35), and 17.5 % of patients who had diabetes at baseline.67 However, the follow-ups of 5 years or less in these trials may be insufficient to evaluate the impact of new-onset diabetes on futureCV events. In a recent clinical trial, the use of statins was associated with a 39 % reduction in the primary end point (myocardial infarction, stroke, admission to hospital for unstable angina, arterial revascularization, or CV death) in patients with ≥1 risk factors for diabetes, but in patients with no risk factors for diabetes, statins were associated with a 52 % reduction in the primary endpoint.71
Despite the risk for future adverse CV events, clinical data to date suggest that the benefit of statin usage far outweighs the risk for diabetes. In an analysis of the JUPITER data in patients with diabetes risk factors who did not have CVD at baseline, for every 54 new cases of diabetes at follow-up, 134 CV events or deaths were prevented. For patients with no diabetes risk factors at baseline, 86 CV events or deaths were prevented with no new cases of diabetes.71 A meta-analysis estimated that treating 255 patients with standard-dose statin therapy for 4 years would avoid 9 vascular events while leading to one case of new-onset diabetes (9:1 benefit versus risk ratio).5 A meta-analysis of intensive statin therapy found that for every case of diabetes caused, 3.2 CV events were prevented per year with intensive-dose statin therapy.6 A literature review of reports of new-onset diabetes associated with statin use among women concluded that benefits outweighed risks though emphasized the paucity of long-term studies focusing on the risks and benefits of statins in women.72 A recent study concluded that the risk:benefit ratio favoring statins is extremely strong, even in those who have an increase in blood glucose levels on statin therapy.73 It is therefore recommended that statins should still be first-line therapy for the majority of patients with dyslipidemia and risk for CVD. However, it is advisable that patients treated with statins undergo regular blood glucose monitoring.
Summary and Concluding Remarks
The incidence of new-onset diabetes varies substantially among clinical trials investigating the efficacy and safety of statins, with only JUPITER and PROSPER finding statistically significant increases in the risk for type 2 diabetes. However, as a whole, clinical data suggest that statin therapy is associated with an increased risk for diabetes of about 9 %. It appears that patients with a predisposition for diabetes may be at greatest risk for developing diabetes while on statin therapy. Some studies comparing highwith lower-dose statin regimes suggest that there may be a higher risk for incident diabetes at higher doses. Furthermore, more potent statins, such as atorvastatin and simvastatin, have been associated with a higher rate of diabetes compared with lower potency statins.
There is a need for continued efforts to understand the relationships between statin therapy and diabetes, both in future clinical trials and pre-clinical investigations. Many questions remain, such as: how does diabetes associated with statin treatment progress? How should it be treated? What mechanisms underlie the observed differences between statins on glycemic parameters? How should these findings affect statin usage? It is also important to examine how the increase in absolute risk for new-onset diabetes reported in meta-analyses compares with the benefits of statin treatment in terms of CV risk reduction.
Statins are now used with the understanding that a slightly increased risk for diabetes is outweighed by the CV benefits. However, the differentialmetabolic effects of the various statins should be taken into account when deciding treatment plans for patients with a high risk for developing diabetes. Based on meta-analyses, pravastatin is considered the statin with the least risk for incident diabetes; however, pitavastatin consistently shows neutral to beneficial effects on glycemic parameters similar to that of pravastatin, and has shown superior lipid-lowering ability compared with pravastatin. In comparative studies, pitavastatin has demonstrated a favorable effect on glycemic parameters compared with other statins. Further data are needed to confirm these findings and assess their relevance to clinical outcomes.
