Do Statins Counteract the Effect of Antidiabetic Drugs? Results of the SCEAD Study

Abstract

Purpose

Diabetes and dyslipidemia are leading causes of mortality and morbidity. According to international guidelines, statins are the cornerstone of treatment in patients with diabetes and/or dyslipidemia. However, statins and antidiabetic agents have opposite pharmacological effects, because statins, particularly atorvastatin and rosuvastatin, impair glucose homeostasis, increasing the risk of new-onset diabetes, whereas antidiabetic drugs improve glycemic homeostasis. The aim of this study was to investigate the effect of atorvastatin, rosuvastatin, and pitavastatin on glucose homeostasis in patients with type 2 diabetes mellitus (T2DM) and dyslipidemia during stable treatment with hypoglycemic drugs.

Materials and Methods

The study was conducted as a pilot, prospective, randomized, open label, parallel group with blinded-endpoints (PROBE) study. Of 180 recruited patients with T2DM and dyslipidemia, 131 were randomized to atorvastatin (n=44), rosuvastatin (n=45), and pitavastatin (n=42) and treated for 6 months.

Results

Fasting plasma glucose (FPG) marginally decreased in patients assigned to atorvastatin (-3.5 mg/dL, p=0.42) and rosuvastatin (-6.5 mg/dL, p=0.17), while it decreased much more in patients treated with pitavastatin (-19.0 mg/dL, p<0.001). Mean glycated hemoglobin A1c (HbA1c ) values remained unchanged during treatment with atorvastatin (-0.10%, p=0.53) and rosuvastatin (0.20%, p=0.40), but were significantly reduced with pitavastatin (-0.75%, p=0.01). Atorvastatin, rosuvastatin, and pitavastatin significantly lowered (p<0.001) plasma levels of total cholesterol, low-density lipoprotein-cholesterol, and triglycerides, while high-density lipoprotein-cholesterol (HDL-C) levels increased significantly (p=0.04) only in the pitavastatin group.

Conclusion

The results of the present study suggest that pitavastatin affects FPG and HbA1c less than atorvastatin and rosuvastatin in patients with T2DM and concomitant dyslipidemia. Lipid-lowering efficacies were not significantly different among the three statins, with the exception of HDL-C, which increased significantly with pitavastatin. Although the pharmacological mechanism of pitavastatin on glucose homeostasis in patients with T2DM during stable antidiabetic therapy is not known, it can be assumed that pitavastatin has less drug interaction with hypoglycemic agents or that it increases plasma levels of adiponectin.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) and dyslipidemia are often concomitantly detected in the same patient and remain major risk factors for fatal and non-fatal cardio-renal events.¹,² Statins and glucose lowering drugs are first-line treatments for patients with T2DM associated with dyslipidemia, and current guidelines³,⁴,⁵ recommend specific targets for glycemic and lipid control to delay or prevent disease progression. Indeed, lowering low-density lipoproteins-cholesterol (LDL-C) by 1–3 mmol/L and glycated hemoglobin A1c (HbA1c) by 1% is associated with 21%–50% and 26% risk reductions in major cardiovascular events, respectively.⁶,⁷,⁸,⁹ However, several studies and meta-analyses¹⁰,¹¹,¹²,¹³ have shown that statins may have a pro-diabetogenic effect that increases the incidence of new onset T2DM. While the relationship between high/moderate doses of statins and the risk of T2DM is well documented,¹⁴,¹⁵,¹⁶ the effect on glucose homeostasis in patients treated with hypoglycemic agents remains controversial.

Several studies¹⁷,¹⁸,¹⁹,²⁰,²¹ have reported varying results that could be dependent on the use of different statins, doses, duration of follow-up, and particularly, changes in antidiabetic agents during treatment. These conflicting findings have been properly highlighted in recent systematic reviews.²⁰,²²,²³,²⁴

Among statins, moderate/high daily doses of atorvastatin and rosuvastatin have been shown to be associated with dysregulation of glucose homeostasis in patients with or without T2DM.¹⁹,²³,²⁴,²⁵,²⁶,²⁷,²⁸ Meanwhile, researchers suggest that the risk of new onset diabetes and deterioration of glucose homeostasis is less evident with pitavastatin¹⁴,¹⁶,²⁴,²⁹,³⁰,³¹ and pravastatin.³²,³³,³⁴ Based on this background, the present study was designed to investigate whether and to what extent atorvastatin, rosuvastatin, and pitavastatin affect fasting plasma glucose (FPG) and HbA1c in patients with T2DM and dyslipidemia during stable treatment with hypoglycemic drugs.

MATERIALS AND METHODS

This study was designed as a pilot, prospective, randomized, open label, parallel groups, with blinded-endpoints (PROBE) study to assess the effect of atorvastatin, rosuvastatin, and pitavastatin on FBG and HbA1c in patients with T2DM associated with dyslipidemia. One hundred and eighty patients, of both genders, were recruited from individuals referred by general practitioners to an outpatient specialist clinic of Internal Medicine at Kartal Hospital in Istanbul, Turkey. Patients were considered eligible for enrollment if they met the following criteria: consensus to participate in the study, age >20 years, and confirmed diagnoses of T2DM and dyslipidemia. Major exclusion criteria were type 1 diabetes, triglycerides (TG) ≥500 mg/dL, clinical manifestations of atherosclerotic cardiovascular disease (previous myocardial infarction or coronary revascularization, stable angina, cerebrovascular accident, heart failure, peripheral artery disease), abnormal thyroid function, renal disease, hepatic dysfunction, pregnancy, and concomitant medications that would interfere with glucose homeostasis.

After collection of demographic data, medical history, and total cardiovascular risk assessment, patients underwent clinical examination and received advice on healthy dieting. For laboratory tests, blood samples were collected in fasting conditions, two times on different days, before the intake of any drugs. Plasma glucose and lipid concentrations were assessed with available automated enzymatic colorimetric methods, and HbA1c was evaluated with high performance liquid chromatography. T2DM was diagnosed in patients with FPG level ≥126 mg/dL, HbA1c ≥6.5% or on- or pre-treatment with antidiabetic drugs,³,⁴ and dyslipidemia was diagnosed as LDL-C ≥70 mg/dL.³⁵ Patients that fulfilled the inclusion and exclusion criteria were randomly assigned to receive moderate doses of atorvastatin (20 mg), rosuvastatin (10 mg), or pitavastatin (2 mg), once daily, and followed for 6 months. The doses were chosen as suggested by international guidelines⁴,³⁵ for patients without atherosclerotic cardiovascular disease. Subjects on lipid-lowering therapy were shifted, at random, to study statins, and in patients with inadequate glycemic control, glucose-lowering drugs were started or modified according to international guidelines.³,⁴ Hypoglycemic drugs, statins, and doses were maintained stable during the 6 months of follow-up. After 3 months, individuals who required statins, antidiabetics, or dosage modifications were excluded from the study. All adverse events occurring before or at the trial closure visit were recorded and analyzed.

The primary endpoints of our trial were 1) change in FPG and 2) change in HbA1c levels from baseline. Secondary endpoints included changes, from baseline, in total cholesterol (TC), LDL-C, high-density lipoprotein-cholesterol (HDL-C), and TG. The study was approved by a local independent Ethics Committee (Kartal Dr. Lütfi Kırdar City Hospital Ethical Committee NUMBER: 2019/514/150/25) and conducted in accordance with the Declaration of Helsinki.

Data analysis

Data analysis was performed in all randomized patients that completed the 6 months of treatment. Distribution of data was tested using the Kolmogorov–Smirnov test. Continuous and categorical variables are expressed as means (SD) and percentages, respectively. Plasma glucose and HbA1c data not normally distributed are expressed as medians and 25th–75th percentiles. ANOVA, Kruskal-Wallis, and chi-square tests were used to compare continuous and categorical variables between treatments. Bonferroni correction was used for post-hoc test. Changes from baseline were compared using the paired t-test or Wilcoxon signed rank test. Two-sided p values<0.05 were considered statistically significant. All statistical tests were performed with the statistical package SPSS 21.0 (IBM Corp., Armonk, NY, USA).

RESULTS

Of the180 patients recruited, 49 were excluded: 20 lost to follow-up, nine for changes in hypoglycemic drugs, and 20 for withdrawing consent during the study (Fig. 1). Thus, 131 patients completed the 6 months of treatment: 44 were assigned to atorvastatin, 45 to rosuvastatin, and 42 to pitavastatin as an add-on to hypoglycemic agents. The baseline demographic, clinical, and laboratory characteristics of patients are reported in Table 1. The three drug groups were homogeneous, particularly in terms of age, sex, treatment allocation, rate of hypertension, FPG, HbA1c, lipids, and anti-diabetic agents.

Fig. 1. Flowchart. T2DM, type 2 diabetes mellitus.

Table 1. Demographic, Clinical, and Biochemical Characteristics of the Study Participants at Baseline.

Parameters		Atorvastatin (20 mg) (n=44)	Rosuvastatin (10 mg) (n=45)	Pitavastatin (2 mg) (n=42)	p value
Male		20 (45.4)	13 (28.8)	14 (33.3)	0.24
Age, yr		58.8±9.7	58.7±12.0	57.9±9.4	0.91
Hypertension		15 (34.1)	17 (37.8)	11 (26.2)	0.50
Antidiabetic drugs		35 (79.5)	32 (71.1)	33 (75.6)
	Metformin	8 (22.9)	10 (31.3)	9 (27.3)	0.88
	Metformin combinations*	23 (65.7)	20 (62.5)	21 (63.6)	0.75
	Others†	4 (11.4)	2 (6.2)	3 (9.1)	0.68
FPG, mg/dL		134.0 (122.5–149.7)	134.0 (121.0–147.0)	139.0 (112.0–185.3)	0.59
HbA1c, %		6.0 (6.5–7.8)	6.1 (6.8–8.2)	7.5 (6.4–9.4)	0.12
Total cholesterol, mg/dL		255.9±42.2	256.6±47.8	259.1±39.8	0.99
LDL-C, mg/dL		173.8±35	174.5±39.4	176.1±33	0.98
HDL-C, mg/dL		47.2±10.2	48.9±9.1	49±8.7	0.60
Triglycerides, mg/dL		177.5±64.4	172.6±68.2	179.5±66.9	0.88

FPG, fasting plasma glucose; HbA1c, Hemoglobin A1c; LDL-C, low-density lipoprotein-cholesterol; HDL-C, high-density lipoprotein-cholesterol.

Data are presented as means±SD, medians (Q1–Q3), or n (%). P-values refers to comparisons among drugs.

^*Metformin+ GLP-1 RA or DPP4 or SGLT-2 inhibitors; ^†Insulin+ metformin or DPP4 inhibitors.

At end of the study (Table 2), the median change in FPG, compared with baseline values, decreased only with pitavastatin (-19 mg/dL, p<0.001) (atorvastatin -3.5 mg/dL, p=0.42; rosuvastatin -6.5 mg/dL, p=0.17). Median changes in HbA1c (Table 2) were not significant with atorvastatin (-0.10%, p=0.53) and rosuvastatin (0.20%, p=0.40), while levels were significantly lowered by pitavastatin (-0.75%, p=0.01). At the end of follow-up, the differences among atorvastatin, rosuvastatin, and pitavastatin were statistically significant for both FPG (p=0.03) and HbA1c (p=0.01). Therefore, pitavastatin deteriorated glycemic control less than atorvastatin and rosuvastatin in patients with T2DM throughout concomitant stable treatment with hypoglycemic agents. Among the secondary endpoints of our trial (Table 3), atorvastatin, rosuvastatin, and pitavastatin significantly decreased (p<0.05) plasma levels of total TC, LDL-C, and TG, while HDL-C slightly increased with atorvastatin (0.65 mg/dL, p=0.28) and rosuvastatin (0.38 mg/dL, p=0.75) and significantly with pitavastatin (1.1 mg/dL, p=0.04), compared to baseline.

Table 2. FPG and HbA1c Changes from Baseline.

Parameters	Drugs	Change (baseline vs. end of treatment)		p value among drugs†
		Median (Q1–Q3)	p value*
FPG, mg/dL	Atorvastatin	-3.5 (-21.7–6.7)*	0.42	0.03
	Rosuvastatin	- 6.5 (-13.0–3.0)	0.17
	Pitavastatin	-19.0 (-40.0– -1.5)	<0.001
HbA1c, %	Atorvastatin	-0.10 (-0.5–0.2)*	0.53	0.01
	Rosuvastatin	0.20 (-0.8–0.5)*	0.40
	Pitavastatin	-0.75 (-1.3–0.1)	0.01

FPG, fasting plasma glucose; HbA1c, glycated hemoglobin.

^*Wilcoxon sign test; ^†Kruskal-Wallis test. For other explanations see previous table.

Table 3. Changes in Lipids from Baseline.

Parameters	Drugs	Change (baseline vs. end of treatment)		p value among drugs†
		Mean (SD)	p value*
Total cholesterol, mg/dL	Atorvastatin	-74.2 (31.8)	<0.001	0.38
	Rosuvastatin	-70.9 (31.9)	<0.001
	Pitavastatin	-71 (49.5)	<0.001
LDL-C, mg/dL	Atorvastatin	-65.7 (21.5)	<0.001	0.58
	Rosuvastatin	-64.6 (28.2)	<0.001
	Pitavastatin	-62.6 (30.5)	<0.001
HDL-C, mg/dL	Atorvastatin	0.65 (4.1)	0.28	0.86
	Rosuvastatin	0.38 (8.0)	0.75
	Pitavastatin	1.1 (3.1)	0.04
Triglycerides, mg/dL	Atorvastatin	-32.8 (71.4)	0.001	0.92
	Rosuvastatin	-27.7 (64.4)	0.001
	Pitavastatin	-33.1 (73.8)	0.001

LDL-C, low-density lipoprotein-cholesterol; HDL-C, high-density lipoprotein-cholesterol.

^*Paired Student’s t test; ^†ANOVA test. For explanations see previous tables.

No clinical adverse events, particularly muscle symptoms or abnormal liver function, were observed with the treatment (data not shown).

DISCUSSION

Our study shows that moderate doses of atorvastatin, rosuvastatin, and pitavastatin exert different effects on glucose homeostasis in patients with T2DM and dyslipidemia under stable treatment with hypoglycemic drugs. Herein, significantly (p<0.01) greater decreases in FPG and of HbA1c were obtained with pitavastatin, compared to atorvastatin and rosuvastatin. It is noteworthy that, unlike other studies and meta-analyses,¹⁷,²³,²⁴,²⁷,²⁸,³¹,³³,³⁴,³⁵,³⁶,³⁷,³⁸ that have reported a dysregulation of glycemic control with low-moderate doses of atorvastatin and rosuvastatin in patients with T2DM, we did not observe this effect. On the contrary, our results show that glycemic homeostasis was not significantly impaired by atorvastatin and rosuvastatin, but, surprisingly, also not improved as expected during concomitant glucose lowering therapy. This finding is in agreement with the results of other studies that have shown modest changes in HbA1c with low to moderate doses of atorvastatin and/or rosuvastatin in patients with diabetes in stable therapy with antidiabetic agents.¹⁹,²⁰,²⁶,³⁹,⁴⁰,⁴¹ Notably, different from other trials,¹⁹,²⁹,³⁰,³¹,⁴² we found no evidence that pitavastatin has a neutral effect on glucose homeostasis. Our results indicate a positive effect on glycemic control, as shown by significant reductions in FPG and HbA1c. This suggests that pitavastatin, compared to atorvastatin and rosuvastatin, could have counteracted the effect of glucose-lowering medications less. The pharmacological mechanism of pitavastatin on glucose homeostasis in patients with T2DM during stable antidiabetic therapy is not known. It could be assumed that pitavastatin has fewer drug interactions with hypoglycemic agents or that it increases plasma levels of adiponectin.⁴³,⁴⁴ Atorvastatin, rosuvastatin, and pitavastatin exhibited similar effects on lipids, except for a significant increase in HDL-C with pitavastatin. This finding is consistent with the results of other studies.⁴⁵,⁴⁶,⁴⁷,⁴⁸

Our study has some potential limitations and strengths. Limitations include 1) that we did not calculate sample size, being a pilot trial; 2) that the study was not performed in a double-blind manner; 3) the absence of oral glucose tolerance test; and 4) small study groups. The strengths include 1) that the study was conducted as a simultaneous head-to-head comparison; 2) that antidiabetic therapy was maintained stable throughout the study; and 3) that FPG and HbA1c plasma levels were assessed in blinded conditions.

In conclusion, the results of our study suggest that not all statins are the same. Although the lipid lowering efficacies of atorvastatin, rosuvastatin, and pitavastatin are not statistically different, pitavastatin appears to less negatively affect the action of hypoglycemic drugs in patients with T2DM. This observation has important clinical relevance and may help in choosing the most appropriate statin in patients with dyslipidemia and diabetes. Moreover, our data emphasize the importance of closely monitoring glycemic control during statin therapy, being ready to change therapeutic strategy.

It is worth mentioning that the results of the present pilot study have to be considered as preliminary and that additional double-blind trials are needed to confirm our data.