Do statins cause an increase in A1c?

Comment by InpharmD Researcher

The use of various statins has been associated with new-onset diabetes and increased A1c. The mechanism for this effect is possibly linked to statins’ inhibitory effects on insulin sensitivity and secretion. Certain statins may increase the risk of diabetes onset (e.g., simvastatin, atorvastatin, and rosuvastatin). A greater risk for incident diabetes has also been associated with higher doses, higher potency, and longer duration of statin use. Another meta-analysis found that lower target levels of LDL (<2.59 mmol/L) for statin therapy can potentially increase risk for new-onset diabetes.

Background

There are multiple theories as to how statin therapy may increase the risk of diabetes, but a consensus has not been met. In vitro studies have found that diabetogenic statins can reduce insulin sensitivity and insulin secretion through inhibition of HMGCoAR (the main target of statin therapy) or impairing beta-cell function. Laasko et al. have noted that the majority of studies implicate simvastatin, atorvastatin, and rosuvastatin as the most diabetogenic statins in population-based studies, clinical studies, and in vitro experiments. A 2019 review concluded diabetic risk appears to be a classwide effect, with pravastatin and pitavastatin potentially having less impact on risk. Mechanistically, pitavastatin does not appear to impair adipocyte maturation at clinical doses, which may lead to improved leptin and adiponectin secretion. However, these studies and experiments are performed in vastly different scenarios, which makes it difficult to draw universal conclusions from the results. [1], [2], [3]

A 2015 cross-sectional study examined the relationship between statin use and glycemic control in 1,093 patients with type 1 diabetes (T1DM) who were under care at the Steno Diabetes Center in Denmark. These patients, with no known heart disease, underwent a comprehensive evaluation that included physical examination, questionnaires, and echocardiography while maintaining blinding to laboratory measurements. Results found that 43.5% of the participants were receiving statin therapy, with these individuals tending to be older, possessing longer diabetes duration, and exhibiting more severe kidney disease. In multivariable analysis, statin use was significantly associated with higher HbA1c levels by 0.2% (95% CI, 0.1 to 0.4), suggesting impaired glycemic control in this cohort. Despite the observed association, the investigation could not establish a causal relationship. Nevertheless, the findings underscore the need to carefully consider insulin adjustments for patients with T1DM commencing statin therapy, while maintaining the cardiovascular protective benefits of statins. [4]

A 2016 network meta-analysis was performed on 29 randomized controlled trials (RCTs) to investigate the direct and indirect evidence of statin use and risk of diabetes development. The likelihood of developing diabetes from statin use was found to be 12% (pooled odds ratio [OR] 1.12; 95% confidence interval [CI] 1.05 to 1.21; p= 0.002; 18 RCTs). Based upon the network meta-analysis, the authors ranked the risk of developing diabetes from highest to lowest: atorvastatin 80 mg, rosuvastatin (OR 1.17; 95% CI: 1.02 to 1.35), simvastatin 80 mg, simvastatin, atorvastatin, pravastatin, lovastatin, and pitavastatin (OR 0.74 95% CI: 0.31 to 1.77). From the results, the authors concluded that intensive statin treatment and lipophilic statins are more likely to increase the risk of developing diabetes. While network meta-analyses allow for such comparisons and inferences, the findings need to be verified in subsequent studies. [5]

Another meta-analysis evaluating the effects of pitavastatin on glycemia and new-onset diabetes mellitus (NODM) in non-diabetic individuals included 15 placebo or statin-controlled RCTs (~ 1,600 person-years; average age of 60 years; body mass index [BMI] 26.7 kg/m2; fasting blood glucose [FBG] <100 mg/dL) with a follow-up of ≥ 12 weeks. No significant differences associated with pitavastatin (vs. control) were observed for FBG (mean difference [MD] –0.01 mg/dL; 95% CI –0.77 to 0.74; I2= 0%), HbA1c (MD –0.03% based on 7 trials; 95% CI –0.11 to 0.05; I2= 43%) or NODM (relative risk [RR] 0.70; 95% CI 0.30 to 1.61; I2= 0%). Analysis based on the patients that completed the follow-up, excluding 7.7% and 9.8% drop-out in pitavastatin and control arms, did not show significant changes in FBG, HbA1c or NODM. Results of this meta-analysis might be limited by the short follow-up duration of included trials and patients at a higher risk of developing NODM while on statin therapy were not adequately included in this study as reflected by the baseline characteristics. In the present meta-analysis, pitavastatin did not adversely affect glucose metabolism or diabetes development compared with placebo or other statins, even though rosuvastatin was not one of the statins being evaluated. [6]

A 2011 meta-analysis (N= 32,752; 5 RCTs) examined the association between intensive- or moderate-dose statin therapy and the risk of new-onset diabetes. Patients were considered to have developed diabetes if (1) there was an adverse event report of newly diagnosed diabetes during the trial, (2) the patient was initiated on glucose-lowering medication during the trial, or (3) the patient had 2 fasting plasma glucose (FPG) values ≥ 126 mg/dL during the trial. Data were also collected for a composite cardiovascular endpoint consisting of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary artery bypass surgery, percutaneous coronary intervention, and for specific cardiovascular events and all-cause mortality. Results revealed that 2,749 patients (8.4%) developed diabetes over a mean follow-up period of 4.9 ± 1.9 years; 1,449 of these patients (52.7%) had received intensive-dose therapy, whereas 1,300 (47.3%) received moderate-dose therapy (OR 1.12; 95% CI, 1.04 to 1.22). The authors reported 2.0 additional cases of diabetes per 1,000 patient-years among intensive-dose recipients (mean 18.9 ± 5.2 cases per 1,000 patient-years with high-dose vs. 16.9 ± 5.5 cases per 1,000 patient-years with moderate-dose). The number needed to harm was 498 per year. No significant heterogeneity was reported between trials for new-onset diabetes (X2 for heterogeneity 2.59; p= 0.60; I2= 0%; 95% CI, 0% to 79%), and included trials were reported to be of high quality. [7]

A subgroup analysis revealed the odds of developing diabetes with intensive- vs. moderate-statin therapy was similar in patients with different ages, BMI, high-density lipoprotein (HDL) cholesterol levels, and FPG levels at baseline but was higher in patients with triglyceride concentrations below the median compared to patients with above the median triglyceride levels (OR 1.27; 95% CI, 1.11 to 1.45 below mean vs. OR 1.06; 95% CI, 0.96 to 1.17 above mean; p= 0.04). Intensive statin therapy was, however, associated with fewer cardiovascular events (OR 0.84; 95% CI, 0.75 to 0.94). These benefits were reported to be consistent across all subgroups and for each component of the primary cardiovascular efficacy endpoint. The authors reported that the relative low-density lipoprotein (LDL) reduction was greater in patients who used atorvastatin 80 mg vs. those used simvastatin 80 mg, but the odds of developing diabetes was similar in both groups; however, there was a significantly lower odds of developing cardiovascular events with high-dose atorvastatin as opposed to high-dose simvastatin. The authors concluded that intensive-dose therapy was associated with an increased risk of new-onset diabetes compared with moderate-dose statin therapy. [8]

A 2016 retrospective cohort study evaluated the association between statins and NODM in Korean patients with ischemic heart disease (IHD). The study was conducted using data from the Korean Health Insurance Review and Assessment (HIRA) database. Adult patients with IHD who initiated statin therapy at any point during the study period were included; patients with previous use of statins within the last year or existing diabetes diagnosis were excluded. Of the total study population that used statins (N= 94,370), 11,851 patients were initiated on rosuvastatin and 4,075 on pitavastatin. Among rosuvastatin users, the risk of developing NODM was 8.1% (incidence rate 3.77/100 patient years, adjusted hazard ratio [HR] 2.00, 95% CI 1.85 to 2.15) compared to pitavastatin users, with a risk of 6.8% (incidence rate 3.17/100 patient years, adjusted HR 1.70, 95% CI 1.50–1.92). Patients exposed to more than one type of statin, (the ‘complex group’), were at the highest risk of NODM (adjusted HR 2.18, 95% CI 1.89 to 2.51). Subgroup analyses revealed that the adjusted NODM HR for statin users versus non-statin users was more significant among male patients (adjusted HR 1.88, 95% CI 1.79 to 2.13) and the risk of NODM was most significant among statin users under the age of 40 years (adjusted HR 5.71, 95% CI 4.00 to 8.18). The authors concluded that all statins are associated with the risk of NODM in patients with IHD and recommended periodic screening and monitoring for diabetes during prolonged statin therapy. [9]

References:

[1] Laakso M, Kuusisto J. Diabetes Secondary to Treatment with Statins. Curr Diab Rep. 2017;17(2):10. doi:10.1007/s11892-017-0837-8
[2] Carmena R, Betteridge DJ. Diabetogenic Action of Statins: Mechanisms. Curr Atheroscler Rep. 2019;21(6):23. Published 2019 Apr 30. doi:10.1007/s11883-019-0780-z
[3] Robinson JG. Statins and diabetes risk: how real is it and what are the mechanisms?. Curr Opin Lipidol. 2015;26(3):228-235. doi:10.1097/MOL.0000000000000172
[4] Ooba N, Tanaka S, Yasukawa Y, et al. Effect of high-potency statins on HbA1c in patients with or without diabetes mellitus. J Pharm Health Care Sci. 2016;2:8. Published 2016 Mar 18. doi:10.1186/s40780-016-0040-0
[5] Thakker D, Nair S, Pagada A, Jamdade V, Malik A. Statin use and the risk of developing diabetes: a network meta-analysis. Pharmacoepidemiol Drug Saf. 2016;25(10):1131-1149. doi:10.1002/pds.4020
[6] Vallejo-Vaz AJ, Kondapally Seshasai SR, Kurogi K, et al. Effect of pitavastatin on glucose, HbA1c and incident diabetes: A meta-analysis of randomized controlled clinical trials in individuals without diabetes. Atherosclerosis. 2015;241(2):409-418. doi:10.1016/j.atherosclerosis.2015.06.001
[7] 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(21): 2195-207
[8] Preiss D, Seshasai SR, Welsh P, et al. Risk of incident diabetes with intensive-dose compared with moderate-dose statin therapy: a meta-analysis. JAMA. 2011;305(24):2556-2564. doi:10.1001/jama.2011.860
[9] Lee J, Noh Y, Shin S, et al. Impact of statins on risk of new onset diabetes mellitus: a population-based cohort study using the Korean National Health Insurance claims database. Ther Clin Risk Manag. 2016;12:1533-1543. Published 2016 Oct 11. doi:10.2147/TCRM.S117150

Literature Review

A search of the published medical literature revealed 9 studies investigating the researchable question:

Do statins cause an increase in A1c?

Level of evidence

B - One high-quality study or multiple studies with limitations  Read more→



Please see Tables 1-9 for your response.


 

Effect of high-potency statins on HbA1c in patients with or without diabetes mellitus

Design

Retrospective cohort study

N= 1,136

Objective

To examine the effect of high-potency statins on HbA1c in patients with or without diabetes

Study Groups

Patients with diabetes (n= 153)

Patients without diabetes (n= 165)

Inclusion Criteria

New statin users identified via the electronic healthcare database of the general hospital in Japan, who had not been prescribed a statin within the preceding 6-month period

Exclusion Criteria

Patients with diabetes receiving low-potency statins, patients receiving fibrates or ezetimibe

Methods

This observational study analyzed data from patients in a single center in Japan with hyperlipidemia who began statin treatment after a six-month period of non-use. High-potency statins (atorvastatin, pitavastatin, and rosuvastatin) were compared to low-potency statins (pravastatin, simvastatin, and fluvastatin) on patients' HbA1c levels, using both the National Glycohemoglobin Standardization Program (NGSP) and Japan Diabetes Society (JDS) criteria for analysis.

Duration

January 1, 2010 to July 31, 2014

Outcome Measures

Primary: Change in HbA1c levels

Secondary: Hazard ratio for new-onset diabetes with high-potency statins

Baseline Characteristics  

Patients with diabetes (n = 276)

Patients without diabetes (n = 659)
Mean follow-up days 262 243
Male, % 75.9 52.0
Age, years, mean ± SD 66.2 ± 11 68.1 ± 11
Hypertension, % 72.4 48.4
Myocardial infarction, % 12.6 17.6
Chronic heart failure, % 18.4 16.7
Renal disease, % 12.6 11.7
Liver disease, % 8.0 9.5
Pulmonary disease, % 13.8 14.9
CCI, mean 1.3 1.2
Results  

Before Statin HbA1c (%)

After Statin HbA1c (%) Difference p-value
With diabetes 7.18 ± 1.37 7.57 ± 1.58 0.39 ± 1.27 0.0002
Without diabetes 5.78 ± 0.38 5.92 ± 0.45 0.15 ± 0.31 <0.0001
Adverse Events

Not specifically reported in the study

Study Author Conclusions

The use of high-potency statins may increase HbA1c levels in patients with or without diabetes.

Critique

The study provides valuable insights into the effects of high-potency statins on HbA1c levels, but it is limited by its retrospective design and the small number of new-onset diabetes cases. The generalizability may be limited due to the single-center setting and the specific population studied.

 

References:

Ooba N, Tanaka S, Yasukawa Y, et al. Effect of high-potency statins on HbA1c in patients with or without diabetes mellitus. J Pharm Health Care Sci. 2016;2:8. Published 2016 Mar 18. doi:10.1186/s40780-016-0040-0

 

Statins and the Risk of Diabetes: Evidence From a Large Population-Based Cohort Study

Design

Population-based cohort study

N= 115,709

Objective

To investigate whether increasing levels of adherence with statin therapy increases the risk of developing physician-diagnosed diabetes and estimated the magnitude of the dose-response relationship

Study Groups

All patients (n= 115,709)

Inclusion Criteria

All National Health Service (NHS) beneficiaries in Lombardy, Italy; aged 40 to 80 years; at least one prescription of statins was dispensed during 2003 and 2004

Exclusion Criteria

Received one or more prescriptions of a statin within three years before the index prescription; received at least one anti-diabetic agent or were hospitalized for a primary or secondary diagnosis of diabetes within three years before the index prescription; did not reach at least one year of follow-up; received only one dispensation of statins during the first year after the date of index prescription

Methods

The period covered by a prescription was calculated from the number of tablets in the dispensed canisters, assuming a treatment schedule of one tablet per day. Adherence to therapy was assessed as the cumulative number of days during which the medication was available divided by the number of days of follow-up, a quantity referred to as the proportion of days covered (PDC). The first dispensation was defined as the index prescription.

The date of the first dispensation of an antidiabetic agent (Anatomical Therapeutic Chemical Classification System code A10) or of hospitalization with a primary diagnosis of diabetes (ICD-9 code 250) was assumed as the date of outcome onset, whichever was earlier. Treatment with high-potency statins was defined ≥10 mg rosuvastatin, ≥20 mg atorvastatin, and ≥40 mg simvastatin, whereas all other statin treatments were defined as low potency.

Duration

Statin treatment: January 1, 2003, through December 31, 2004

Follow-up: ∼ 6.4 years

Outcome Measures

New cases of diabetes, the effect of statins on diabetes risk

Baseline Characteristics

 

All patients (N= 115,709)

Age, years

62.4 ± 9.4

Male

56,258 (48.6%)

First-line therapy with statins

Atorvastatin

Fluvastatin

Pravastatin

Rosuvastatin

Simvastatin

 

37,742 (32.6%)

12,253 (10.6%)

19,474 (16.8%)

9,596 (8.3%)

36,644 (31.7%)

Antihypertensive drugs

ACE inhibitors

Angiotensin receptor blocker

Calcium channel blocker

ß-Blockers

Thiazides

Others

 

54,495 (47.1%)

34,103 (29.5%)

37,068 (32.0%)

48,246 (41.7%)

595 (0.5%)

31,995 (27.7)

Concomitant use of other drugs

High-dose corticosteroids

Oral contraceptives

Cyclosporine

Antipsychotics

 

17,909 (15.5%)

5,193 (4.5%)

426 (0.4%)

2,396 (2.1%)

History of cardiovascular disease

41,607 (36.0%)

Results

Endpoint

All patients (N= 115,709)

New cases of diabetes

Number, n

Rate (every 1,000 person-years)

 

11,154

14.9

There was a continuous and significant trend toward an increase of diabetes risk as adherence to statin therapy increased both in the unadjusted and in the adjusted risk models; 58% (95% confidence interval [CI], 51 to 66%) vs. 32% (95% CI, 26 to 39%) for high adherence vs. very-low adherence, respectively).

There was no statistical evidence that the effect of statins on diabetes risk differed according to their potency (p= 0.372).

Compared with patients with very-low adherence (PDC <25%), those with low (PDC 26 to 50%), intermediate (PDC 51 to 75%), and high (≥75%) adherence to statin therapy had hazard ratios [HRs] 1.12 (95% CI, 1.06 to 1.18), 1.22 (1.14 to 1.27), and 1.32 (1.26 to 1.39), respectively.

Adverse Events

N/A

Study Author Conclusions

In a real-world setting, the risk of new-onset diabetes rises as adherence to statin therapy increases. The benefits of statins in reducing cardiovascular events clearly overwhelm the diabetes risk.

InpharmD Researcher Critique

Several limitations included the possibility of detection bias, potential unmeasured confoundings, evaluating statin adherence based on pharmacy-dispensing information, and lack of comprehensive baseline clinical information. Additionally, since identification codes of prescription records were not available for analysis, drug-based diagnoses of diabetes cannot be validated.



References:

Giovanni Corrao, Buthaina Ibrahim, Federica Nicotra, Davide Soranna, Luca Merlino, Alberico L. Catapano, Elena Tragni, Manuela Casula, Guido Grassi, Giuseppe Mancia; Statins and the Risk of Diabetes: Evidence From a Large Population-Based Cohort Study. Diabetes Care 1 August 2014; 37 (8): 2225–2232. https://doi.org/10.2337/dc13-2215

 

Effect of pitavastatin compared with atorvastatin and rosuvastatin on new-onset diabetes mellitus in patients with acute myocardial infarction

Design

Retrospective cohort study

N= 2,483

Objective

To investigate the impact of moderate-intensity pitavastatin (2-4 mg) compared with moderate-intensity atorvastatin (10-20 mg) and rosuvastatin (5-10 mg) on the development of new-onset diabetes mellitus (NODM)

Study Groups

Atorvastatin (n= 1,267)

Rosuvastatin (n= 961)

Pitavastatin (n= 255)

Inclusion Criteria

Acute myocardial infarction (AMI) patients enrolled in the Korean Acute Myocardial Infarction Registry (KAMIR), no previous history of diabetes mellitus, successful percutaneous coronary intervention (PCI) with drug eluting stent implantation, treated with intermediate-intensity atorvastatin, rosuvastatin, or pitavastatin

Exclusion Criteria

Previous history of diabetes mellitus or initial HbA1c level ≥ 6.5%, bare metal stent implantation, plain old ballon angioplasty, failed PCI, in-hospital major adverse cardiovascular events

Methods

Patients enrolled in KAMIR were retrospectively reviewed. Patients receiving intermediate-intensity atorvastatin (10-20 mg), rosuvastatin (5-10 mg), or pitavastatin (2-4 mg), according to 2014 American College of Cardiology (ACC)/American Heart Association (AHA) Release Updated Guideline, were enrolled. Differences between the 3 groups were evaluated by one-way analysis of variance.

Duration

November 2011 to May 2015

Follow-up: 3 years

Outcome Measures

Primary outcome: incidence of NODM (fasting blood glucose level ≥ 126 mg/dL or HbA1c level ≥ 6.5%)

Baseline Characteristics

 

Atorvastatin (n= 1,267)

Rosuvastatin (n= 961)

Pitavastatin (n= 255) p-value

Age, years

64.4 ± 12.5 63.2 ± 12.4 63.4 ± 12 0.0721

Male

971 (76.6%) 731 (76.1%) 184 (72.2%) 0.3097

Body mass index, kg/m2

23.7 ± 3.2 23.7 ± 3.1 24.3 ± 3.0 0.0162

Glucose, mg/dL

136.5 ± 45.4 140 ± 45.2 139.5 ± 39.6 0.1599

Results

Endpoint

Univariate analysis HR (95% CI)

p-value

Multivariate analysis HR (95% CI)

p-value

Incidence of NODM

Atorvastatin vs. pitavastatin

Rosuvastatin vs. pitavastatin

Atorvastatin vs. rosuvastatin

 

2.539 (1.124 to 5.736)

3.688 (1.634 to 8.326)

0.688 (0.471 to 1.006)

 

0.0251

0.0017

0.0539

 

2.615 (1.163 to 5.879)

3.906 (1.756 to 8.689)

0.669 (0.456 to 0.982)

 

0.0201

0.0008

0.0401

HR, hazard ratio; CI, confidence interval

Based on a Kaplan-Meier curve for the cumulative incidence of NODM up to 3 years, the pitavastatin group showed a significantly lower incidence of NODM compared to the atorvastatin and rosuvastatin groups (3% vs. 8.4% and 10.4%, respectively; p= 0.001)

Adverse Events

N/A

Study Author Conclusions

Moderate intensity pitavastatin therapy is associated with a lower incidence of NODM compared with moderate intensity atorvastatin and rosuvastatin therapy in patients with AMI. Although prospective, randomized trials with larger study population would be needed for clarifying the present study’s results, the result of our study would be a supportive evidence for favorable effect of pitavastatin on glycemic control in AMI patients.

InpharmD Researcher Critique

This study is limited by its retrospective design and its lack of data on compliance with statin therapy. Fasting blood glucose and HbA1c values were not evaluated in this study. Additionally, the 3-year follow-up period is relatively shorter than in previous studies.



References:

Choi JY, Choi CU, Hwang SY, et al. Effect of Pitavastatin Compared with Atorvastatin andRosuvastatin on New-Onset Diabetes Mellitus in PatientsWith Acute Myocardial Infarction. Am J Cardiol. 2018;122(6):922-928. doi:10.1016/j.amjcard.2018.06.017

 

Risk of incident diabetes among patients treated with statins: population-based study

Design

Retrospective, population-based, cohort study

N= 471,250

Objective

To examine the risk of new-onset diabetes among patients treated with different HMG-CoA reductase inhibitors (statins)

Study Groups

All new statin patients (N= 471,250)

Pravastatin (n= 38,470)

Atorvastatin (n= 268,254)

Fluvastatin/lovastatin (n= 11,923)

Rosuvastatin (n= 76,774)

Simvastatin (n= 75,829)

Inclusion Criteria

Aged ≥ 66 years; without diabetes; started atorvastatin, fluvastatin, lovastatin, rosuvastatin, simvastatin, or pravastatin between August 1, 1997, and March 31, 2010; not prescribed a statin in at least the preceding year

Exclusion Criteria

Established diabetes prior to the start of treatment

Methods

Patient information was collected from healthcare databases in Ontario, Canada. Pravastatin was used as an active comparator reference group. Primary outcome diabetes diagnosis was determined by patient records in the Ontario Diabetes Database. 

Duration

Intervention: treated with statin from August 1, 1997 to March 31, 2010

Follow-up: maximum of 5 years or end of study period

Outcome Measures

Primary: incidence of diabetes

Baseline Characteristics  

Pravastatin

(n= 38,470)

Atorvastatin

(n= 268,254)

Fluvastatin/lovastatin       

(n= 11,923)

Rosuvastatin

(n= 76,774)

Simvastatin

(n= 75,829)

Age, years

73.04 ± 5.62

73.97 ± 6.34

72.83 ± 5.49 73.24 ± 6.20  73.80 ± 6.06

Male

44.6%

46.1%

40.5%

46.0% 46.7%

Secondary prevention

History of cardiac disease or procedures

Previous acute coronary syndrome

Chronic coronary artery disease

Stroke/transient ischemic attacks

 

53.5%

31.8%

47.8%

13.8%

 

53.4%

31.5%

45.3%

16.1%

 

47.1%

26.1%

41.9%

12.0%

 

39.3%

19.1%

32.0%

12.3%

 

57.4%

35.4%

50.2%

16.1%

Charlson score

0

1

≥ 2

 

18.3%

10.9%

10.2%

 

18.1%

11.3%

10.7%

 

17.7%

9.6%

8.9%

 

17.8%

6.6%

6.9%

 

18.1%

11.9%

11.2%

Previous drug use

Steroids

First-generation antipsychotic

Second-generation antipsychotic

Immunosuppressants

ß-blockers

Thiazide diuretics

 

18.3%

0.7%

0.6%

0.1%

26.8%

14.4%

 

18.0%

0.5%

1.4%

0%

26.3%

19.1%

 

16.2%

0.8%

0.3%

0.1%

23.7%

13.4%

 

18.3%

0.4%

1.6%

0%

22.3%

20.5%

 

18.1%

0.5%

1.0%

0%

28.2%

16.7%

Results

Endpoint

Pravastatin

(n= 38,470)

Atorvastatin

(n= 268,254)

Fluvastatin

(n= 5,636)

Lovastatin

(n= 6,287)

Rosuvastatin

(n= 76,774)

Simvastatin

(n= 75,829)

Diagnosis of diabetes

Number of outcomes

Adjusted* hazard ratio (HR) (95% confidence interval [CI])

The number needed to treat to harm

 

1,443

Reference

 

15,261

1.22 (1.15 to 1.29)

172

 

167

0.95 (0.81 to 1.11)

 

211

0.99 (0.86 to 1.14)

 

3,732

1.18 (1.10 to 1.26)

210

 

3,727

1.10 (1.04 to 1.17)

363

 

Low dose

Moderate dose

High dose

Diagnosis of diabetes by dose grouping

Adjusted HR (95% CI)

Reference

1.22 (1.19 to 1.27)

1.30 (1.20 to 1.40)

 

Pravastatin Moderate potency (simvastatin) High potency      (atorvastatin, rosuvastatin) Low potency         (fluvastatin, lovastatin)

Diagnosis of diabetes by statin potency

Number of outcomes

Adjusted HR (95% CI)

 

1,443

Reference

 

3,727

1.11 (1.04 to 1.18)

 

18,993

1.22 (1.15 to 1.29)

 

378

0.97 (0.87 to 1.09)

*Adjusted for age, sex, year of cohort entry, recent acute coronary syndrome, chronic coronary artery disease, Charlson score, previous use of diuretic (thiazide), nitroglycerin, angiotensin receptor blocker, β blocker, hormones, and analogs.

The risk of incident diabetes was similar whether statins were being used for primary or secondary prevention.

Adverse Events

N/A

Study Author Conclusions

Compared with pravastatin, treatment with higher potency statins, especially atorvastatin and simvastatin, might be associated with an increased risk of new-onset diabetes.

InpharmD Researcher Critique

Some risk factors for diabetes could not be taken into account in this study, including patient weight, ethnicity, and family history. Additionally, no data was available regarding blood lipids, hemoglobin A1C concentration, or triglyceride concentrations, thus, no potential association between these factors and risk of incident diabetes was able to be analyzed.



References:

Carter AA, Gomes T, Camacho X, Juurlink DN, Shah BR, Mamdani MM. Risk of incident diabetes among patients treated with statins: population based study. BMJ. 2013;346:f2610. Published 2013 May 23. doi:10.1136/bmj.f2610

 

Statins and Risk of Treated Incident Diabetes in A Primary Care Population

Design

Retrospective, cohort study

N= 239,628

Objective

To examine the effect of different types of statins and their duration and dose-response relationship on the new onset treated diabetes using a large national pharmacy claims database

Study Groups

Atorvastatin (n= 120,307)

Pravastatin (n= 41,899)

Rosuvastatin (n= 19,888)

Simvastatin (n= 11,458)

Fluvastatin (n= 3,125)

Inclusion Criteria

Received monotherapy of the different types of statin; under the General Medical Services (GMS) Scheme (over-represented by females, socio-economically deprived patients, and the elderly)

Exclusion Criteria

Not specified.

Methods

Duration of exposure to statins was examined according to the number of months for which the patients were treated with statins. The total cumulative dosage of statin dispensed over the study period was determined using the defined daily doses (DDD) according to WHO methodology (average recommended daily dose for a drug used for its main indication in adults; equivalent to 10 mg atorvastatin, 20 mg pravastatin, 10 mg rosuvastatin, 15 mg simvastatin, and 40 mg fluvastatin).

All patients initiated on statins were followed up until their first prescription of any antidiabetic medications or until 31 January 2009 in those not receiving antidiabetic medication. Patients who did not receive statins during the study period were considered as a control group and were followed up until they received antidiabetic agents or until 31 January 2009.

Duration

Statin treatment period: January 1, 2002, to December 31, 2007

Outcome Measures

Risk of new-onset treated diabetes

Baseline Characteristics

 

All patients in univariate analysis (N= 197,138)

 

Number with diabetes

Hazard ratio (95% confidence interval [CI])

p-value

Age groups, years

25-44

45-64

≥ 65

 

7,672 (1.5%)

12,472 (3.7%)

18,662 (4.7%)

 

-

0.96 (0.94 to 1.00)

1.14 (1.11 to 1.16)

 

-

NS

p< 0.0001

Female

19,324 (2.6%) 1.02 (0.99 to 1.04) NS

Diabetogenic agents

Antipsychotics

Corticosteroids

 

6 ,721 (2.5%)

10,179 (2.8%)

 

1.24 (1.21 to 1.28) 

0.77 (0.75 to 0.80)

 

p< 0.0001

p< 0.0001

Cardiovascular comorbidities

Hypertension

Ischemic heart disease

Obesity

 

21,391 (4.0%)

2,799 (4.8%)

1,552 (4.3%)

 

1.19 (1.17 to 1.22)

1.14 (1.10 to 1.19)

1.31 (1.23 to 1.39)

 

p< 0.0001

p< 0.0001

p< 0.0001

Other lipid-lowering agents

Ezetimibe

Omega-3

Fibrates

Nicotinic acid

 

309 (3.6%)

192 (2.7%)

153 (7.1%)

179 (8.5%)

 

1.67 (1.47 to 1.88)

1.70 (1.44 to 2.00)

1.24 (1.08 to 1.44)

1.35 (1.01 to 1.80)

 

p< 0.0001

p< 0.0001

p= 0.003

p= 0.04

NS: not significant

Results

Statins

Number with diabetes

Adjusted hazard ratio (95% CI)

p-value

All statins

Atorvastatin

Pravastatin

Rosuvastatin

Simvastatin

Fluvastatin

11,591 (5.9%) 

5,608 (4.7%)

2,729 (6.5%)

967 (4.9%)

726 (6.3%)

220 (7.0%)

1.20 (1.17 to 1.23)

1.25 (1.21 to 1.28) 

1.02 (0.98 to 1.06)

1.42 (1.33 to 1.52)

1.14 (1.06, 1.23)

1.04 (0.91 to 1.18)

p< 0.0001

p< 0.0001

NS

p< 0.0001

p= 0.0005

NS 

Adverse Events

N/A

Study Author Conclusions

A significant association was found between new-onset treated diabetes in those treated with statin therapy. This was associated with increased duration and dose of treatment suggesting a possible biologically plausible effect. Further prospective studies are required to examine this association in order to prove a possible causal relationship.

InpharmD Researcher Critique

Although the study was limited by retrospective nature and lack of clinical diagnosis in the database, potential confoundings (oral corticosteroids, thiazide diuretics, β-adrenoceptor blockers, and antipsychotics) were all clearly reported and adjusted in the results.



References:

Zaharan NL, Williams D, Bennett K. Statins and risk of treated incident diabetes in a primary care population. Br J Clin Pharmacol. 2013;75(4):1118-1124. doi: 10.1111/j.1365-2125.2012.04403.x

 

Effect of statins on fasting glucose in non-diabetic individuals: nationwide population-based health examination in Korea

Design

Retrospective study

N= 379,865

Objective

To analyze the changes in fasting glucose by use of statins based on data from repeatedly performed nationwide health examinations in Korea

Study Groups

Patients exposed to any statin (n= 96,182)

Atorvastatin (n= 61,157)

Rosuvastatin (n= 11,720)

Pitavastatin (n= 8,010)

Pravastatin (n= 8,047)

Simvastatin (n= 46,773)

Lovastatin (n= 9,291)

Fluvastatin (n= 3,786)

Inclusion Criteria

Age 40-79 years old, non-diabetic patients with fasting glucose level data from at least 2 examinations during the study period

Exclusion Criteria

Fasting glucose level ≥ 7 mmol/L at the baseline health examination, diagnosis of diabetes Mellitus (DM) based on a questionnaire in health examination, health insurance claim data, or anti-diabetic prescription medication

Methods

Patients were randomly selected from a Korean database that compiles data from a free health examination offered to National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) members ≥ 40 years, every 2 years. Subjects were followed-up until death, loss of eligibility for NHIS due to emigration, or Dec 31, 2013, the end of the study period. The baseline value of fasting glucose was defined as the value at the first health examination. Fasting glucose was measured at each health examination. 

Duration

Study period: 2002 through 2013

Outcome Measures

Changes in fasting glucose, as affected by proportion of days covered (PDC), and average number of daily doses per day (anDDD)

Baseline Characteristics

 

All subjects (N= 379,865)

Age, years

51.91 ± 9.19  

Male

199,786 (52.6%)  

Body mass index, kg/m2

23.79 ± 2.87  

Current smoker

33,840 (8.9%)  

Alcohol consumption, frequency per week

< 1

1-2

3-4

≥ 5

 

 

215,773 (56.8%)

123,573 (32.5%)

25,446 (6.7%)

15,073 (4.0%)

Exercise, days per week, days

< 1

1-4

≥ 5

 

215,736 (56.8%)

129,356 (34.1%)

34,773 (9.2%)

Results

Endpoint

Atorvastatin (n= 61,157)

Rosuvastatin (n= 11,720)

Pitavastatin (n= 8,010)

Pravastatin (n= 8,047) Simvastatin (n= 46,773) Lovastatin (n= 9,291)

Fluvastatin (n= 3,786)

Cumulative duration of years exposure

78,627 14,678 7,691 8,099 46,768 6,045 3,175

Coefficient for change in fasting glucose for PDC

95% CI 

p-value

 

0.108

0.086-0.130

< 0.001 

 

0.069

0.021-0.117

0.005 

 

0.080

0.008-0.152

0.030 

 

0.026

–0.025-0.077

0.318 

 

0.118

0.093-0.143

< 0.001 

 

0.044

–0.006-0.094

0.088

 

 0.021

–0.071-0.114

0.654

Coefficient for change in fasting glucose for anDDD

95% CI 

p-value

 

 

0.154

0.119-0.188

< 0.001

 

0.061

0.013-0.108

0.012 

 

0.088

0.014-0.161

0.020

 

0.049

–0.017-0.116

0.148

 

0.162

0.126–0.197

< 0.001 

 

0.108

–0.001-0.217

0.052

 

0.040

–0.034-0.114

0.294 

Adverse Events

N/A

Study Author Conclusions

There was a significant increase of fasting glucose in non-diabetic individuals in proportion to adherent and intensive use of statins. Among the statin subtypes, use of atorvastatin, rosuvastatin, pitavastatin, and simvastatin was associated with a significant increase in fasting glucose. Pravastatin, lovastatin, and fluvastatin also had a trend toward increased fasting glucose but were statistically non-significant. These findings suggested medication class effects of all types of statins predisposing hyperglycemia although there was some difference in the degree according to the type.

InpharmD Researcher Critique

Aside from the risk of selection bias due to the retrospective design, it may be difficult to determine the degree of patient compliance and how the potential lack of compliance may have affected the final results. Additionally, this study was performed in Korea which limits the generalizability of the study to the population in the United States.



References:

Kim J, Lee HS, Lee KY. Effect of statins on fasting glucose in non-diabetic individuals: nationwide population-based health examination in Korea. Cardiovasc Diabetol. 2018;17(1):155. Published 2018 Dec 5. doi:10.1186/s12933-018-0799-4

 

Statins and risk for new-onset diabetes mellitus

Design

Retrospective cohort study

N= 41,325

Objective

To investigate the risk for new-onset diabetes mellitus (NODM) with statin treatment in real-world clinical settings

Study Groups

Statin exposed (n= 8,265)

Pitavastatin (n= 403)

Rosuvastatin (n= 1,429)

Inclusion Criteria

Statin-exposed patients, 18 years of age or older, took statin continuously for more than 90 days

Exclusion Criteria

Psychiatric disorder or received organ transplant(s), DM, abnormal glucose levels (≥ 200 mg/dL), abnormal fasting glucose levels (≥ 126 mg/dL), abnormal hemoglobin A1c (≥ 6.5%), International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) diagnosis codes related to diabetes (E10-E14), received a prescription for diabetes medication(s) including insulin before the start of observation

Methods

Statin-exposed patients consisted of patients exposed to statins for more than 90 consecutive days. Patients were divided into 6 subgroups according to the type of statin: atorvastatin, fluvastatin, pitavastatin, pravastatin, and simvastatin. Observation for statin exposure began at day 91 from the first exposure to evaluate the long-term effects of statins on NODM; observation ended when NODM occurred. A within-class comparison was performed to compare the risk for NODM in patients exposed to 1 of the 6 statins with patients exposed to other statins.

Duration

January 1996 to August 2013

Outcome Measures

Incidence of NODM

Baseline Characteristics

 

Statin exposed (n= 8,265)

 

Age, years

54.3 ± 12.3  

Male

4,148 (50.2%)  

Observation period, days

603.7 ± 644  

Glucose level, mg/dL

108.9 ± 22.5  

Results

Endpoint

Pitavastatin (n= 403)

Rosuvastatin (n= 1,429)

Incidence of NODM, per 1,000 patient-years

1.321 4.770

Within-class comparison of NODM, hazard ratio (95% confidence interval); p-value

0.368 (0.051 to 2.636); 0.319 1.008 (0.579 to 1.756); 0.977

Adverse Events

N/A

Study Author Conclusions

Among various statins, atorvastatin might be more prone to increase the risk for DM than their matched non-exposed counterparts. We need more careful consideration regarding the balance between the risks and benefits when determining statin treatment.

InpharmD Researcher Critique

The intensity of the statin was not considered in this study. Additionally, the comparative risk for NODM between various statins was analyzed as a secondary endpoint; therefore, it was underpowered, and further studies are needed.



References:

Yoon D, Sheen SS, Lee S, Choi YJ, Park RW, Lim HS. Statins and risk for new-onset diabetes mellitus: A real-world cohort study using a clinical research database. Medicine (Baltimore). 2016;95(46):e5429. doi:10.1097/MD.0000000000005429

 

Effect of statins on fasting glucose in non-diabetic individuals: nationwide population-based health examination in Korea

Design

Retrospective database analysis

N= 379,865

Objective

To investigate the longitudinal changes in fasting glucose level of non-diabetic individuals by use of statins

Study Groups

All (N= 379,865)

Inclusion Criteria

Non-diabetic individuals who had data on their fasting glucose level from ≥ 2 health screening examinations during the study period

Exclusion Criteria

Subjects who had a fasting glucose level ≥ 7 mmol/L at the baseline health examination; answered “yes” to the diagnosis of diabetes mellitus (DM) on the questionnaire in health examination; had a diagnosis of DM (ICD-10 code ‘E10–E15’) from the health insurance claim data; had received a prescription for anti-diabetic medication during the study period

Methods

Relevant data were collected from the National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) in Korea. Using the prescription records of statins in the database, the proportion of days covered (PDC) and the average number of defined daily doses per day (anDDD) by statins were calculated. Multivariate linear mixed models were constructed to evaluate the effects of statins on the changes in fasting glucose (Δglu). 

All included subjects were followed-up until death, loss of eligibility for NHIS due to emigration, or Dec 31, 2013, which is the end of the study period. 

Duration

Between 2002 and 2013

Outcome Measures

PDC, anDDD

Baseline Characteristics

 

All (N= 379,865)

 

 

Total number of health examinations

2,117,598    

Number of health examination per subject, median [IQR]

5 [4–6]    

Time period between baseline to last health examination per subject, years

8.96 ± 2.12    

Age, years

51.91 ± 9.19    

Male

199,786 (52.6%)    

Body mass index (BMI), kg/m2

23.79 ± 2.87    

Fasting glucose, mmol/L

5.02 ± 0.68    

Systolic blood pressure, mmHg

125.38 ± 17.55    

Current smoker

33,840 (8.9%)    

Alcohol consumption, frequency per week 

< 1

1-2

3-4

≥ 5

 

215,773 (56.8%)

123,573 (32.5%)

25,446 (6.7%)

15,073 (4.0%)

   

Exercise, days per week 

< 1

1-4

≥ 5

 

215,736 (56.8%)

129,356 (34.1%)

34,773 (9.2%)

   

Household income

Q1, low

Q2, middle

Q3, high

 

109,219 (28.8%)

139,051 (36.6%)

131,595 (34.6%)

   

Number of patients who were exposed to statins and other lipid-lowering agents

Medication Number (%) of patients who had exposure to medication Cumulative duration (years) of exposure to medication  

Any statin

Atorvastatin

Rosuvastatin

Pitavastatin

Pravastatin

Simvastatin

Lovastatin

Fluvastatin

96,182 (25.3%)

61,157 (16.1%)

11,720 (3.1%)

8,010 (2.1%)

8,047 (2.1%)

46,773 (12.3%)

9,291 (2.4%)

3,786 (1.0%)

165,083

78,627

14,678

7,691

8,099

46,768

6,045

3,175

 
Fibratea 14,926 (3.9%) 11,617  
Ezetimibe 5,847 (1.5%) 5,830  
aBezafibrate, ciprofibrate, etofibrate, fenofibrate, and gemfibrozil

Results

Endpoint

Coefficient (β)

Standard error

p-Value

PDC by any statina

0.093 0.007  < 0.001

PDC by fibrateb

0.022  0.025  0.387 

PDC by ezetimibe

0.046  0.045  0.314 

anDDD by any statinc

0.119  0.009  < 0.001 

Effects of each statin type on the change in fasting glucose

Statin

β for PDC 95% confidence interval (CI) p-Value

Atorvastatin

Rosuvastatin

Pitavastatin

Pravastatin

Simvastatin

Lovastatin

Fluvastatin

0.108

0.069

0.080

0.026

0.118

0.044

0.021 

0.086;0.130

0.021;0.117

0.008;0.152

-0.025;0.077

0.093;0.143

-0.006;0.094

-0.071;0.114

< 0.001

0.005 

0.030

0.318

< 0.001

0.088

0.654

  β for anDDD    

Atorvastatin

Rosuvastatin

Pitavastatin

Pravastatin

Simvastatin

Lovastatin

Fluvastatin

0.154

0.061

0.088

0.049

0.162

0.108

0.040

0.119;0.188

0.013;0.108

0.014;0.161

-0.017;0.116

0.126;0.197

-0.001;0.217

-0.034;0.114

< 0.001

0.012

0.020

0.148

< 0.001

0.052

0.294 

aValue is the estimated change in fasting glucose when whole time period is covered by statins (PDC= 1); bBezafibrate, ciprofibrate, etofibrate, fenofibrate, and gemfibrozil; cWhen ‘anDDD by any statin’ were included in the multivariate linear mixed model instead of ‘PDC by any statin’

Adverse Events

N/A

Study Author Conclusions

More adherent and intensive use of statins were significantly associated with an increase in fasting glucose in non-diabetic individuals. In a subgroup analysis of individual statins, the use of atorvastatin, rosuvastatin, pitavastatin, and simvastatin had a significant association with an increase in fasting glucose. Pravastatin, lovastatin, and fluvastatin had a non-significant trend toward increased fasting glucose. Our findings suggest the medication class effect of statins inducing hyperglycemia.

InpharmD Researcher Critique

Despite a large sample inclusion from a population-based nationwide health screening program for a 10-year period, all subjects are representative of Koreans which greatly limits the generalizability, considering the gaps in lifestyles and diet. Thus, the response to statins might be varied in other genetic populations. Also, the actual intake of statins in subjects might be different from the prescription records.



References:

Kim J, Lee HS, Lee KY. Effect of statins on fasting glucose in non-diabetic individuals: nationwide population-based health examination in Korea. Cardiovasc Diabetol. 2018;17(1):155. Published 2018 Dec 5. doi:10.1186/s12933-018-0799-4

 

Effects of Pitavastatin, Atorvastatin, and Rosuvastatin on the Risk of New-Onset Diabetes Mellitus: A Single-Center Cohort Study

Design

Retrospective cohort study

N= 8,337

Objective

To investigate the effects of commonly used moderate-intensity statins (pitavastatin, atorvastatin, and rosuvastatin) on the risk of new onset diabetes by utilizing a single-institute real-world database

Study Groups

2 mg/day pitavastatin (n= 1,312)

10 mg/day atorvastatin (n= 3,034)

10 mg/day rosuvastatin (n= 3,991)

Inclusion Criteria

Non-diabetic adults, initiated on 10 mg/day atorvastatin, 10 mg/day rosuvastatin, or 2 mg/day pitavastatin for the first time

Exclusion Criteria

Patients without consecutive prescriptions (identical prescription for 2 months or more), existing diabetes or two consecutive fasting glucose levels of 126 mg/dL or higher or two consecutive HbA1c levels of 6.5% or higher before statin therapy

Methods

Patient data was compiled from electronic medical records from the Tri-Service General Hospital, a tertiary hospital in Taiwan. 

Duration

January 2006 to July 2018

Follow-up: 4 years

Outcome Measures

Development of new onset diabetes

Baseline Characteristics

 

Pitavastatin (n= 1,312)

Atorvastatin (n= 3,034)

Rosuvastatin (n= 3,991)

 p-Value

Follow up, days

468.31 ± 446.18 558.30 ± 666.85 648.30 ± 796.06 < 0.001

New-onset diabetes

167 (12.7%) 555 (18.3%) 864 (21.6%) < 0.001

Male

728 (55.5%) 1546 (51.0%) 2236 (56.0%) < 0.001

Age, years

60.42 ± 12.44 61.53 ± 13.54 58.21 ± 13.45 < 0.001

BMI, kg/m2

25.71 ± 4.18 24.75 ± 4.08 25.11 ± 4.38 0.017

Fasting glucose, mg/dL

99.94 ± 15.74 99.27 ± 17.31 99.57 ± 17.98 0.581

HbA1c, %

6.05 ± 0.98 6.08 ± 1.17 6.09 ± 1.23 0.872

LDL, mg/dL

131.52 ± 28.51 122.94 ± 37.57 141.30 ± 45.86 < 0.001

Hypertension

796 (60.7%) 1,416 (46.7%) 1,864 (46.7%) < 0.001

There were also statistically significant differences in baseline for co-morbidities of coronary artery disease, congestive obstructive pulmonary disease, chronic kidney disease, cancer, and ischemic stroke.

BMI: body mass index, LDL: low-density lipoprotein

Results

Endpoint

Pitavastatin (n= 1,312)

Atorvastatin (n= 3,034)

Rosuvastatin (n= 3,991)

 

New-onset diabetes

Crude HR (95% CI)

p-Value

Adjusted HR (95% CI)

p-Value

 

1.00

--

1.00

--

 

1.21 (1.02–1.44)

0.032*

1.04 (0.87–1.25)

0.677*

 

1.24 (1.05–1.47)

0.011*

1.13 (0.94–1.35)

0.196*

 

*Compared to pitavastatin

The HRs in the three statin groups were not statistically significant after adjusting for baseline characteristics and comorbidities (p = 0.23). Hazard ratio was adjusted by years, gender, age, coronary artery disease, hypertension, chronic obstructive pulmonary disease, chronic kidney disease, cancer, ischemic stroke, hemorrhagic stroke, heart failure.

HR: hazard ratio, CI: confidence interval

Adverse Events

N/A

Study Author Conclusions

Among patients treated with moderate-intensity statins, a nonsignificant trend of lower risk of new-onset diabetes was associated with pitavastatin compared with atorvastatin and rosuvastatin based on a real-world clinical database. On the basis of the results, clinicians can select an appropriate statin according to patients’ comorbidities and baseline.

InpharmD Researcher Critique

Due to retrospective design, there were significant confounders between groups. The authors adjusted for gender, age, and baseline comorbidities to reduce the confounding factors in the statin groups but these adjusted hazard ratios were not statistically significant. Additionally, the study was conducted at a single institute in Taiwan; the results should be applied to other populations with caution. 



References:

Liu WT, Lin C, Tsai MC, et al. Effects of Pitavastatin, Atorvastatin, and Rosuvastatin on the Risk of New-Onset Diabetes Mellitus: A Single-Center Cohort Study. Biomedicines. 2020;8(11):499. Published 2020 Nov 13. doi:10.3390/biomedicines8110499