Atorvastatin Lowers C-Reactive Protein and Improves Endothelium-Dependent Vasodilation in Type 2 Diabetes Mellitus

Endothelial dysfunction is frequently found in diabetic subjects. This study was performed to investigate whether atorvastatin therapy was able to reverse endothelial dysfunction in type 2 diabetes and, if so, whether the effect was due to its antiinflammatory action. Eighty patients (baseline low density lipoprotein, 4.37 ± 0.71 mmol/liter) were randomized to atorvastatin (10 mg daily for 3 months, followed by 20 mg daily for 3 months) or placebo in a double blind study. Endothelial function was assessed by high resolution vascular ultrasound, and high sensitivity C-reactive protein (CRP) was assessed by immunoturbidimetric assay. Diabetic patients had higher CRP (P < 0.01) than matched nondiabetic controls, and both endothelium-dependent and independent vasodilation were impaired (P < 0.01). Atorvastatin (10 and 20 mg) lowered plasma cholesterol by 32.9% and 38.0%, triglyceride by 15.4% and 23.1%, and low density lipoprotein by 43.4% and 50.1%, respectively. At 6 months, plasma CRP decreased in the atorvastatin group compared with baseline (P < 0.05). Endothelium-dependent vasodilation improved in the atorvastatin group compared with the placebo group (P < 0.05). The percent change in endothelium-dependent vasodilation at 6 months correlated with the percent change in CRP (r = −0.44; P < 0.05), but not with changes in plasma lipids. In conclusion, treatment with atorvastatin in type 2 diabetes led to a significant improvement in endothelium-dependent vasodilation, which might be partly related to its anti-inflammatory effect.

IT IS WELL known that the risk of atherosclerosis is markedly elevated in patients with diabetes, and coronary heart disease (CHD) occurs up to 2–4 times more often in patients with diabetes (1, 2). This increase is seen in almost every cultural and racial group regardless of their background incidence of CHD. In addition to having a greater risk of CHD, diabetic patients have a worse prognosis with a greater case of fatality and 1-yr mortality after myocardial infarction (3, 4). Hence, prevention of atherosclerosis is of prime importance in patients with both type 1 and type 2 diabetes mellitus.

It has been shown that endothelial dysfunction is an early event in atherogenesis and precedes the thickening of the intima and the formation of atherosclerotic plaques (5, 6). Endothelial dysfunction has been consistently demonstrated in patients with type 1 and type 2 diabetes mellitus (7–10), and there is evidence to show that endothelial dysfunction contributes to the pathogenesis of both micro- and macroangiopathy in diabetes (11). The etiology of endothelial dysfunction is complex in patients with diabetes. In addition to the classical risk factors associated with endothelial dysfunction in nondiabetic subjects, such as smoking, hypertension, and hyperlipidemia, additional factors, such as the increased formation of advanced glycation end products, insulin resistance, and activation of PKC, may also play a role in patients with diabetes (12, 13). There is also recent evidence to suggest a relationship between activation of the endothelium and chronic inflammation. Levels of C-reactive protein (CRP), a marker of systemic inflammation, are elevated in patients with diabetes (14–16), and CRP has been shown to correlate with markers of endothelial dysfunction (17). Cleland et al. (18) demonstrated a relationship between low grade chronic inflammation and basal endothelial nitric oxide (NO) synthesis, suggesting that endothelial dysfunction may be a link between inflammation and atherosclerosis.

Restoring normal endothelial function in diabetic patients may have a beneficial effect and reduce cardiovascular risk. Cholesterol-lowering therapy has been shown to reverse endothelial dysfunction in nondiabetic patients with CHD (19–21), but whether lipid-lowering therapy has a similar beneficial effect on endothelial function in patients with diabetes is still controversial. There are at present only very limited data available on the effect of statin therapy on endothelial dysfunction in patients with type 2 diabetes. Sheu et al. (22) recently reported that treatment with simvastatin has no effect on endothelial dysfunction in a small uncontrolled study of type 2 diabetic patients with hypercholesterolemia. This study was therefore performed to 1) investigate whether lowering low density lipoprotein (LDL) cholesterol with atorvastatin was able to reverse endothelial dysfunction in patients with type 2 diabetes mellitus, and 2) to determine whether this might be related to the antiinflammatory effect of atorvastatin by measuring CRP concentrations.

Materials and Methods

The effect of atorvastatin was examined in a randomized, double blind, placebo-controlled study. The inclusion criteria were hemoglobin A_1c (HbA_1c) less than 10%, fasting LDL greater than 3.4 mmol/liter, triglycerides (TG) less than 4.0 mmol/liter, total cholesterol to high density lipoprotein (HDL) ratio greater than 4.0, and blood pressure 160/90 mm Hg or less. Exclusion criteria were the current use of lipid-lowering therapy, secondary hyperlipidemia, deranged liver or renal function, and a major cardiovascular event within the last 6 months. After a dietary stabilization period of 8 wk with placebo run-in, patients who satisfied the inclusion criteria were entered into the study. Eighty patients were recruited from the diabetic clinics at Queen Mary Hospital and were randomized in a double blind manner to 10 mg atorvastatin daily, increasing to 20 mg daily after the first 3 months, or to matching placebo. Fasting lipid profiles, endothelial function, and plasma high sensitivity CRP were assessed at baseline and 3 and 6 months after treatment. Carotid intima-media thickness (IMT) was measured at baseline to characterize the cardiovascular status of the patients. Eighty nondiabetic controls matched for age and sex were recruited from the community to establish a normal reference range for carotid IMT, endothelial function, and CRP. The study was approved by the ethics committee of the University of Hong Kong, and informed consent was obtained from all subjects.

Plasma total cholesterol and TG were determined enzymatically on a Hitachi 912 analyzer (Roche, Mannheim, Germany). HDL cholesterol was measured using a homogenous method with polyethylene glycol-modified enzymes and α-cyclodextrin. LDL cholesterol was calculated by the Friedewald equation. HbA_1c was measured in whole blood using ion exchange HPLC with the Variant Hemoglobin Testing System (Bio-Rad Laboratories, Inc., Hercules, CA). Plasma CRP was measured by a particle-enhanced immunoturbidimetric assay (Roche), using anti-CRP mouse monoclonal antibodies coupled to latex microparticles.

High resolution B-mode ultrasound was used to measure the IMT of the carotid artery. Ultrasonographic scanning was performed with an ATL HDI 3000 ultrasound system (Advanced Technology Laboratories, Inc., Bothell, WA). The anterior, lateral, and posterolateral projections were used to image longitudinally the right and left common carotid arteries. At each longitudinal projection, three determinations of IMT were made at 2 cm proximal to the bulb and at the site of greatest thickness. The values at each site were averaged, and the greatest value of the averaged IMT was used as the representative value for each individual. Carotid plaque was defined as the presence of a focal lesion measuring at least twice the thickness of the IMT (23). If a plaque was present in one of the projections, that value was excluded from the analysis, and IMT was averaged on the remaining values.

Endothelium-dependent and endothelium-independent vasodilation of the brachial artery was assessed noninvasively as previously described (10). Flow-mediated vasodilation caused by reactive hyperemia was related to the release of NO and was therefore an endothelium-dependent phenomenon, whereas endothelium-independent vasodilation was induced by glyceryl trinitrate (GTN), which acted on the vascular smooth muscle. Brachial artery diameter was measured from B-mode ultrasound images using a 10-MHz linear array transducer on the ATL HDI 3000 ultrasound system with continuous electrocardiogram recording. After optimal transducer positioning, the arm was kept in the same position, and the skin was marked. Diameter measurements of the right brachial artery were taken at rest after lying quietly for at least 15 min, and then during reactive hyperemia after occlusion by inflation of pneumatic tourniquet to a pressure of 300 mm Hg for 4.5 min. Twenty minutes were allowed for vessel recovery and then a further resting scan was taken. Sublingual GTN spray (400 μg) was administered, and measurement were repeated after 5 min. Measurements were taken from the anterior to the posterior m line at end diastole, incident with the R wave on the electrocardiogram. Three cardiac cycles were analyzed, and measurements were averaged. Flow-mediated and GTN-induced vasodilation was calculated as the percent change in diameter compared with baseline. Hard copy images of the brachial artery were taken to enable measurement of the same segment of the artery at subsequent visits.

Data that were normally distributed were expressed as the mean and sd. The distribution of TG and CRP were skewed, and data were logarithmically transformed before analyses were made. Results were then expressed as the geometric mean, and the interquartile range was also given. Comparisons between two different groups were made using independent sample t test and within-group changes of variables measured more than twice were assessed by ANOVA for repeated measurements. Pearson’s correlations were used to test the relationship between variables.

Results

The clinical characteristics of the controls and diabetic patients are shown in Table 1. Patients with diabetes had higher CRP levels (P < 0.01) than controls and increased carotid IMT (P < 0.05), and both endothelium-dependent and independent vasodilation were significantly impaired (Table 1). Endothelium-dependent vasodilation correlated inversely with HbA_1c (r = −0.51; P < 0.01), LDL (r = −0.17; P < 0.05), HDL (r = 0.18; P < 0.05), and CRP (r = −0.21; P < 0.01), whereas endothelium-independent vasodilation correlated with HbA_1c (r = −0.34; P < 0.01), LDL (r = −0.22; P < 0.01), and HDL (r = 0.27; P < 0.01), but not with CRP.

Table 1.

Clinical characteristics, plasma CRP, carotid IMT, and endothelial function of controls and diabetic patients

	Nondiabetic controls (n = 80)	Diabetes (n = 80)
M/F (no.)	40/40	40/40
Age (yr)	53.3 ± 9.3	55.0 ± 8.1
BMI (kg/m2)	25.0 ± 3.4	26.1 ± 3.41
Duration of diabetes (yr)		10.0 ± 7.0
Smokers (no.)	6	6
HbA1c (%)	5.7 ± 0.5	7.9 ± 1.12
TC (mmol/liter)	5.50 ± 0.95	6.24 ± 0.832
TG (mmol/liter)3	1.18 (0.80–1.50)	1.39 (1.0–1.78)1
LDL (mmol/liter)	3.53 ± 0.80	4.37 ± 0.712
HDL (mmol/liter)	1.30 ± 0.40	1.16 ± 0.232
CRP (mg/liter)3	0.93 (0.45–2.16)	1.80 (0.90–3.64)2
Carotid IMT (mm)	0.53 ± 0.09	0.59 ± 0.141
Endothelium-dependent vasodilation (%)	9.2 ± 4.3	5.2 ± 2.52
Endothelium-independent vasodilation (%)	16.9 ± 5.5	13.6 ± 4.92

Values are the mean ± sd.

¹

< 0.05 vs. controls.

²

P < 0.01 vs. controls.

³

Geometric mean (interquartile range).

The diabetic patients were randomized to receive either placebo (n = 41) or 10 mg atorvastatin daily for 3 months, followed by 20 mg daily for the subsequent 3 months (n = 39). None of the patients in the placebo group had a history of cardiovascular disease, whereas two patients in the atorvastatin-treated group had coronary heart disease. The number of patients with hypertension were similar in the two groups. The dosages of antihyperglycemic and antihypertensive drugs were kept unchanged throughout the study (data not shown). Carotid IMT was similar in the two groups at baseline (0.59 ± 0.14 vs. 0.59 ± 0.15 mm). Forty-one percent of the patients in the placebo group (n = 16) and 46% in the atorvastatin-treated group (n = 18) had evidence of carotid plaque.

Plasma lipid levels are shown in Table 2. Plasma cholesterol, TG, and LDL decreased in the atorvastatin-treated group compared with the placebo-treated group, whereas no significant changes were seen in HDL. With 10 mg atorvastatin daily, plasma cholesterol decreased by 32.9%, TG by 15.4%, and LDL by 43.4% from baseline; with 20 mg atorvastatin daily, the reductions were 38.0%, 23.1%, and 50.1%, respectively. Plasma CRP significantly decreased in the atorvastatin-treated group at 6 months compared with the baseline value (P < 0.05), whereas no significant changes were observed in the placebo group after treatment (Table 2). Although the difference in CRP between the atorvastatin-treated group and the placebo group at 6 months did not reach statistical significance (P = 0.057), the percent change in CRP from baseline was significantly larger in the atorvastatin-treated group than in the placebo group (−17.4% vs. 10.6%, respectively; P = 0.02). The percent change in CRP at 6 months did not correlate with percent changes in LDL (r = −0.01; P = NS) or TG (r = −0.05; P = NS) in the atorvastatin-treated group.

Table 2.

Effects of atorvastatin on plasma lipids, CRP, and endothelial function

	Baseline		3 months		6 months
	Placebo (n = 41)	Atorvastatin (n = 39)	Placebo	Atorvastatin (10 mg)	Placebo	Atorvastatin (20 mg)
TC (mmol/liter)	6.12 ± 0.64	6.35 ± 0.98	6.04 ± 0.86	4.26 ± 0.7323	5.95 ± 0.75	3.99 ± 0.6324
TG (mmol/liter)1	1.41	1.38	1.63	1.13	1.51	1.11
	(1.0–1.90)	(1.0–1.80)	(1.15–2.3)	(0.80–1.50)3	(1.0–2.25)	(0.80–1.25)24
LDL (mmol/liter)	4.29 ± 0.52	4.45 ± 0.87	4.11 ± 0.77	2.52 ± 0.5523	4.08 ± 0.66	2.28 ± 0.4924
HDL (mmol/liter)	1.12 ± 0.23	1.19 ± 0.23	1.12 ± 0.24	1.18 ± 0.22	1.11 ± 0.26	1.14 ± 0.22
HbA1c (%)	8.0 ± 1.1	7.8 ± 1.2	8.0 ± 1.1	8.1 ± 1.3	8.0 ± 1.2	8.2 ± 1.4
CRP (mg/liter)1	1.64	1.98	1.74	1.62	1.75	1.23
	(0.77–3.02)	(0.98–4.89)	(1.03–3.03)	(0.86–3.43)	(0.80–4.15)	(0.90–2.81)5
Endothelium-dependent vasodilation (%)	5.0 ± 2.4	5.3 ± 2.6	4.7 ± 2.3	5.6 ± 2.1	5.0 ± 2.1	6.5 ± 2.856
Endothelium-independent vasodilation (%)	12.8 ± 4.4	14.4 ± 5.3	14.0 ± 5.2	14.4 ± 5.1	14.4 ± 4.1	14.7 ± 4.8

Values are the mean ± sd.

¹

, Geometric mean (interquartile range).

²

P < 0.01, within-group comparison vs. baseline values.

³

P < 0.01 vs. placebo at 3 months.

⁴

P < 0.01 vs. placebo at 6 months.

⁵

P < 0.05, within-group comparison vs. baseline values.

⁶

P < 0.05 vs. placebo at 6 months.

The results of endothelium-dependent vasodilation and independent vasodilation are also shown in Table 2. There was an improvement in endothelium-dependent vasodilation in the atorvastatin-treated group at 6 months compared with that in the placebo-treated group (P < 0.05). No significant changes were observed in endothelium-independent vasodilation. In the atorvastatin-treated group, the percent change from baseline in endothelium-dependent vasodilation at 6 months did not correlate to the percent change in LDL (Fig. 1) or TG (r = 0.1; P = NS), but correlated significantly with the percent change in CRP (Fig. 2A). In the placebo group, no significant correlation was found between the percent change in endothelium-dependent vasodilation at 6 months and the percent change in CRP (Fig. 2B).

Figure 1.

Relationship between the percent change in LDL at 6 months and the percent change in endothelium-dependent vasodilation at 6 months in the atorvastatin group.

Figure 2.

Relationship between the percent change in CRP at 6 months and the percent change in endothelium-dependent vasodilation at 6 months in the atorvastatin group (A) and the placebo group (B).

Discussion

We have demonstrated in the present study that 6 months of treatment using atorvastatin resulted in an improvement in endothelium-dependent vasodilation and a lowering of plasma CRP concentrations in patients with type 2 diabetes mellitus. The lack of effect of simvastatin on endothelial function in patients with type 2 diabetes previously reported by Sheu et al. (22) might be due to the small sample size and the relatively low dosage of simvastatin used. An improvement in endothelial function has recently been demonstrated in a group of young type 1 diabetic patients with normal cholesterol levels by Mullen et al. (24) using 40 mg atorvastatin daily. Hence, similar to nondiabetic subjects with hypercholesterolemia and/or CHD, endothelial dysfunction can be ameliorated in diabetic patients with statin therapy. However, it is important to note that in both our study and the study by Mullen et al. (24), statin therapy improved, but did not completely normalize, endothelial function. Compared with our nondiabetic controls, endothelium-dependent vasodilation was reduced by almost 50% in patients with type 2 diabetes, and treatment with atorvastatin was only able to improve endothelium-dependent vasodilation by approximately 20%. Mullen et al. (24) also found that the improvement in endothelial function was relatively small in their type 1 diabetic patients compared with the responses seen in nondiabetic hypercholesterolemic subjects they had studied previously. Taken together, these findings suggest that there are other additional factors that may play an important role in causing endothelial dysfunction in diabetes mellitus.

Several mechanisms by which statins improve endothelial dysfunction have been investigated, and it has been shown that statins have pleiotropic properties that complement their cholesterol-lowering effects (25). One of these pleiotropic properties is the antiinflammatory effect of statins. Atorvastatin has a direct antiinflammatory effect on the vessel wall in animal models (26), and statin therapy has been shown to lower CRP levels in patients with hypercholesterolemia and combined hyperlipidemia (27, 28). We have shown, for the first time, that atorvastatin also reduces CRP levels in patients with type 2 diabetes, and the magnitude of reduction in CRP correlated with the degree of improvement in endothelium-dependent vasodilation. This would support the hypothesis that the improvement in endothelium-dependent vasodilation in our diabetic patients might be partly mediated by the antiflammatory effect of atorvastatin. We cannot exclude the possibility that other mechanisms might also be involved. There is both in vitro and in vivo evidence to show that the effect of statins on endothelial dysfunction is related not only to the lowering of LDL, but also to a direct effect on NO production. Statins have a direct effect on endothelial NO synthase (eNOS) expression and cause up-regulation of eNOS with increased bioavailability of NO in vivo (29–31). Statins can also reverse the inhibitory effect of oxidized LDL on eNOS (29). In addition to the effect on NO, statins have been shown to reduce the synthesis of endothelin-1, a potent vasoconstrictor, by endothelial cells (32). The metabolites of atorvastatin have potent antioxidant activities on LDL in vitro, protect HDL against oxidation, and have a paraoxonase-sparing effect (33). All these additional pleiotropic properties of statins are independent of their cholesterol-lowering effect, and this may explain why we did not find a significant correlation between the magnitude of LDL lowering and the degree of improvement in endothelium-dependent vasodilation.

In conclusion, treatment with atorvastatin leads to a significant reduction in plasma LDL cholesterol and TG, a lowering of plasma CRP, and an improvement in endothelial dysfunction in patients with type 2 diabetes mellitus. Whether this will translate into a reduction in cardiovascular risk in patients with type 2 diabetes mellitus may become clear when the results of the Collaborative Atorvastatin Diabetes Study become available (34).

Acknowledgements

This work was supported by the Hong Kong Research Grants Council (HKU 483/96 m), the Department of Medicine Research Committee, University of Hong Kong, and Pfizer, Inc.

Abbreviations:

 
• CRP,

  C-Reactive protein;

 
• eNOS,

  endothelial nitric oxide synthase;

 
• GTN,

  glyceryl trinitrate;

 
• HbA_1c,

  hemoglobin A_1c;

 
• HDL,

  high density lipoprotein;

 
• IMT,

  intima-media thickness;

 
• LDL,

  low density lipoprotein;

 
• NO,

  nitric oxide;

 
• TG,

  triglycerides.