Cerivastatin Improves Insulin Sensitivity and Insulin Secretion in Early-State Obese Type 2 Diabetes

The study included 15 patients with type 2 diabetes diagnosed by repeated fasting glucose at time of presentation according to new World Health Organization criteria (i.e., a fasting plasma glucose concentration >7.0 mmol/l) (14). All patients had received standard treatment in an established primary diabetes care program and were on diets, and none of them had received any medication known to affect glucose metabolism. The purpose and risks of the study were carefully explained before patients gave informed consent to participate. The study protocol was reviewed and approved by the Ethics Committee of the Reina Sofía University Hospital of Cordoba. Inclusion criteria were as follows: all subjects were >35 and <70 years of age, had a history of 1–3 years of moderate hyperglycemia with fasting glycemia >110 mg/dl and <140 mg/dl and HbA_1c <7% without treatment, had a BMI (calculated as weight [in kilograms] divided by height [in square meters]) >27 kg/m² (15) (this index was not modified for at least 6 months before the study), and had a waist circumference of ≥102 cm and ≥88 cm in men and women, respectively (16,17). All patients presented with serum triglyceride (TG) levels of >150 mg/dl and <350 mg/dl. Exclusion criteria were as follows: signs of diabetic retinopathy, nephropathy, neuropathy, or macrovascular complications; cigarette smoking and alcohol consumption; and use of diuretics, steroids, and β-blockers. All subjects had normal results on screening blood tests of hepatic, renal, thyroid, and adrenal function. Upon evaluation, two female patients presented with elevated levels in fasting (8:00 a.m.) plasma cortisol. After a low-dose dexamethasone suppression test, both patients presented fasting (8:00 a.m.) plasma cortisol values of <140 nmol/l. Hypertension in five patients was treated with Enalapril 10 (3) or 20 mg/day (2).

The trial was performed as a single-center, double-blind, randomized crossover, placebo-controlled study (Fig. 1). Each subject was followed for a run-in period of 4 weeks before being chosen at random for daily administration of either cerivastatin (0.4 mg/d) or a placebo with the appearance of an alternative drug. Both active tablets and placebo tablets were supplied by Bayer (Leverkusen, Germany). Each treatment period lasted 3 months. There was a washout period of 4 weeks between placebo and cerivastatin. In the last week of each treatment period, on Tuesday an oral glucose tolerance test (OGTT) + hyperglycemic clamp (HC) was performed, and on Saturday an intravenous glucose tolerance test (IVGTT) + euglycemic-hyperinsulinemic clamp (EHC) was performed.

The subjects reported to the metabolic unit at 08:00 a.m. On Tuesday, after an overnight fast of 12 h, an OGTT with 75 g of glucose dissolved in 300 ml of water was performed. Samples were obtained at 0, 15, 30, 45, 60, 90, and 120 min. An HC was started immediately afterward. Glucose was elevated and clamped to 360 mg/dl glucose (20 mmol/l), with adjustment of the rate of infusion of a 20% glucose solution at 10-min intervals with a precision syringe pump (LifeCare Pump Model 4; Abbot Laboratories, Chicago, IL). Samples were obtained at 150, 165, and 180 min (30, 45, and 60 min of infusion). In both tests, plasma insulin, C-peptide, and glucose concentrations were determined. On Saturday, after an overnight fast of 12 h, an IVGTT was performed (18). A 300-mg/kg body wt glucose load (maximum 25 g) was briefly given intravenously over 2 min as a 50% solution. Blood samples for glucose and insulin assay were collected at − 5, 0, 2, 4, 6, 8, and 10 min. For measuring the action of insulin, an EHC was performed immediately (19). A priming dose of insulin (0.1 units/kg) was followed by constant intravenous insulin infusion at a rate of 80 mU · m⁻² · min⁻¹ using a precision syringe LifeCare pump and continued throughout the 180-min study. Glucose was determined at 5-min intervals. Plasma glucose was clamped to 100 mg/dl glucose (5.6 mmol/l), with adjustment of the rate of infusion of the 20% glucose solution at 10-min intervals with a precision syringe LifeCare pump. The desired plasma glucose level was attained within 60 min. Blood samples for glucose, insulin, C-peptide, plasma lactate, and serum free fatty acid (FFA) were collected at fasting (−5) and 120, 140, 160, and 180 min. The average value for the period from 140 to 180 was used to calculate the rates of whole-body glucose uptake.

Stature and weight were measured. The waist circumference of each subject was measured at the umbilicus, and the widest hip circumference was measured at the gluteal fold. Bioelectrical impedance was measured with a multifrequency bioelectrical impedance analyzer in fasting state with empty bladders. This method estimates lean body mass (LBM) in kilograms. Body fat mass was determined by subtracting LBM from body weight (20).

The energy production and respiratory quotient (Rq) were measured by continuous indirect calorimetry using a computerized, flow-through canopy gas analyzer system (Deltatrac; Datex, Helsinki, Finland). Energy expenditure was measured after an equilibration period of 10 min. The gas-exchange rate was recorded for 30 min in the fasting state and during the last 30 min of the euglycemic clamp, between 150 and 180 min. From knowledge of the Vo₂ and VCO₂ (l/min) and the urinary nitrogen (g/min), glucose, lipid, and protein disappearance rates (g/min) can be calculated according to calorimetric formulas (21). The protein oxidation rate was estimated from urinary nitrogen excretion (1 g of nitrogen = 6.25 g of protein) in 24-h urine collections from the same day and over the clamp studies (22). The energy expenditure data in minutes were adjusted for differences by LBM. Nonoxidative glucose metabolism was determined by subtracting the glucose disappearance (oxidation) rate, measured by indirect calorimetry, from the whole insulin-mediated glucose intake during the EHC.

Plasma glucose were measured using a Hitachi 917 analyzer by the glucose oxidase method (GOD-PAP; Boehringer, Mannheim, Germany). Plasma insulin concentrations were measured by microparticle enzyme immunoassay (Abbott Diagnostics, Matsudo-shi, Japan; lower limit 1 μU insulin/ml; total assay variation 6.1%) using an AxSYM system Analyzer (Abbott Laboratories, Dallas, TX). Plasma C-peptide content was measured by RIA (BYK Sangtec, Dietzenbach, Germany; lower limit 0.003 ng/ml). Total glycosylated hemoglobin was determined in fresh samples by high-performance liquid chromatography (Bio-Rad, Paris, France; normal range 4.5–6.4%; interassay coefficient of variation 3.92%) using the fully automated Glycosylated Hemoglobin Variant II Analyzer (Bio-Rad, Tokyo). A Hitachi P-800 analyzer (Roche Diagnostics, Tokyo, Japan) was used to measure fructosamine by an enzymatic colorimetric method (Roche, Mannheim, Germany; intra-assay coefficient of variation 4.3%). Automated methods were used for measurement of cholesterol and TGs (CHOD-PAP and GPO-PAP, respectively; Boehringer). Plasma HDL cholesterol was determined by a dextran sulfate-magnesium precipitation procedure. FFA (nonesterified fatty acids [NEFAs]) concentration was measured in triplicate in each sample using an enzymatic colorimetric assay (Roche Molecular Biochemical, Mannheim, Germany; total coefficient of variation 7.8%). An enzymatic colorimetric lactate oxidase method (Roche Diagnostics; normal range 0.5–2.2 mmol/l; interassay coefficient of variation 2.3%) was used to determine plasma lactate by a Cobas Integra 400 Analyzer. The high-sensitivity C-reactive protein (CRP) was measured using immunonephelometric assays on a BN II analyzer (Dade Behring, Newark, DE). A Quantikine HS Immunoassay kit (R & D Systems, Minneapolis, MN; intra-assay coefficient of variation 14%) was used to determine human tumor necrosis factor α (TNF-α) in serum. An enzyme-linked immunosorbent assay was used for the quantitative measurement of leptin in serum (DRG Diagnostics, Germany; intra-assay coefficient of variation 4.28%).

The acute insulin response (first phase) to the glucose pulse was considered to be the mean of the postglucose insulin concentrations from 4, 6, 8, and 10 min (IVGTT). The insulin index was defined as the ratio between peak and basal plasma insulin values. The ratio Δ I_30-0/ΔG_30-0, increment of plasma insulin (pmol/l) to increment in glucose (mmol/l), was also used as a measure of early insulin secretion during OGTT. Maximal insulin and C-peptide secretion (second phase) was assessed as the average plasma insulin and C-peptide concentrations at 150, 165, and 180 min of the HC. The incremental areas under the curves for insulin and glucose during 2-h OGTT were calculated as the deviations from the basal value using a trapezoidal method. Insulin sensitivity was measured as an insulin sensitivity index, calculated by dividing the average glucose infusion rate (nanomoles per LBM per minute) by the average plasma insulin concentration (picomoles per liter) at 140, 160, and 180 min of the EHC. Kinetics of peripheral insulin metabolism were evaluated by measurement of metabolic clearance rate (MCR) of exogenously administered insulin during the EHC. MCR (ml · m⁻² · min⁻¹) = IIR (μUI · m⁻² · min⁻¹)/[IC_s − IC_e (μU/ml)], where IIR = insulin infusion rate during the EHC; IC_s = steady state of mean insulin concentration during the EHC (140, 160, and 180 min); IC_e = endogenous steady-state insulin concentration assessed from the mean of two fasting insulin concentration per (steady-state C-peptide levels during the EHC/mean of two fasting C-peptide levels) (23). Plasma C-peptide/plasma insulin ratio was considered to be the index of hepatic insulin extraction in the fasting state and during the HC (mean values at 150, 165, and 180 min).

To compare the two study periods and to assess the effect of the sequence in which the patients received the placebo and cerivastatin treatment, we used repeated-measures ANOVA, followed by paired t test analyses (24). For variables with a skewed distribution, logarithmic transformation was performed. Results are presented as the mean ± SE. Covariance analysis with change in body weight as a covariant was performed to separate the effects of treatment from the effects of changes in body weight.

The mean (±SD) age of patients was 57 ± 7 years (range 38–68). The mean period observed for moderate fasting hyperglycemia was 25 ± 12 months (Table 1). At baseline, the patients presented with a mean BMI of 32.3 ± 1.1 kg/m² and a central body fat distribution. LBM and calculated body fat mass were similar during placebo (50.4 ± 3; 31.8 ± 1.7) as compared with cerivastatin treatment (51.1 ± 2.6; 32.1 ± 1.8), respectively. The weight, BMI, LBM, and waist-to-hip ratio did not change throughout the study. Mean arterial blood pressure remained unchanged during the study. The mean HbA_1c level was 6.4 ± 0.13% at randomization with no significant differences observed between placebo and cerivastatin treatment (6.7 ± 0.15% and 6.6 ± 0.2%; P = 0.8 [r = 0.63, P = 0.012]; Table 2). Plasma fructosamine concentrations measured in a single assay were also similar (271 ± 15 vs. 270 ± 14 μmol/l; P = 0.9 [r = 0.88, P = 0.000]). The mean serum cholesterol concentration at admission was 5.64 ± 0.2 mmol/l. Compared with the baseline and placebo period, cerivastatin treatment resulted in a decrease in fasting plasma total cholesterol concentration by 25 and 19.5% (P < 0.001), respectively. After cerivastatin treatment, cytokine TNF-α was significantly lowered as compared with placebo period (2.68 ± 0.3 vs. 2.16 ± 0.4 pg/ml; P < 0.05 [r = 0.747, P = 0.001]). High-sensitivity CRP also decreased with cerivastatin treatment (3.57 ± 0.6 vs. 2.23 ± 0.4 mg/dl; P < 0.05 [r = 0.45, P = 0.08]). Leptin levels did not change with cerivastatin treatment and were positively correlated with body fat mass (r = 0.537, P = 0.062; r = 0.619, P = 0.032, placebo and treatment period, respectively).

The mean fasting blood glucose concentration at admission was 7.3 ± 0.3 mmol/l and remained unchanged during the study (placebo 6.8 ± 0.3 to cerivastatin 6.9 ± 0.3; P = 0.9 [r = 0.66, P = 0.007]; Table 2). Mean fasting plasma insulin levels presented a slight but nonsignificant increase after cerivastatin treatment (69 ± 6.8 to 76 ± 5.6 pmol/l; P = 0.31 [r = 0.3, P = 0.3]; Table 3). Neither difference was observed in mean fasting C-peptide levels (placebo 1.29 ± 0.34 to cerivastatin 1.14 ± 0.17 pg/ml; P = 0.5 [r = 0.77, P = 0.001]; Table 3). Fasting hepatic insulin extraction (pmol C-peptide/pmol insulin) was found to increase after cerivastatin treatment as compared with placebo (14.87 ± 1.5 vs. 9.34 ± 1.2; P = 0.02).

After OGTT, three patients presented with impaired glucose tolerance (2-h glucose >7.7 and <11.1 mmol/l) during the placebo period and two patients presented with impaired glucose tolerance during cerivastatin treatment (Fig. 2, top left). The rest of the patients were considered diabetic because of the oral glucose load. The incremental glucose area under the curve was not changed (placebo 21.25 ± 1.4 mmol · l⁻¹ · h⁻¹, cerivastatin 21.47 ± 1.4 mmol/l; P = 0.8 [r = 0.55, P = 0.03]).

The incremental insulin area under the curve after oral glucose load showed no significant difference (placebo 512 ± 108 pmol · l^{− 1} · h⁻¹, cerivastatin 622 ± 124 pmol/l; P = 0.3 [r = 0.59, P = 0.02]; Fig. 2, bottom left). The insulinogenic index (Δ I_30-0/ΔG_30-0) increased significantly with cerivastatin treatment (placebo 32.1 ± 7.0 pmol/mmol, cerivastatin 46.5 ± 9.9 pmol/mmol; P < 0.05 [r = 0.76, P = 0.001]; Table 3).

Mean values for glucose infusion rates during the HC studies increased slightly in patients on placebo compared with patients on cerivastatin treatment (952 ± 51 vs. 992 ± 59 μmol · LBM⁻¹ · min⁻¹; P = 0.059 [r = 0.8, P = 0.001]; Fig. 2, right). The incremental insulin area under the curve increased slightly with placebo (25%) as compared with cerivastatin treatment (1,193.9 ± 251 vs. 1,492 ± 252; P = 0.2 [r = 0.53, P = 0.04]). Insulin sensitivity index increased after cerivastatin treatment as compared with placebo (209.6 ± 38 vs. 288 ± 44 nmol · LBM^{− 1} · min⁻1 · pmol⁻¹ · l⁻¹; P = 0.036 [r = 0.67, P = 0.006]). The incremental C-peptide area under the curve during the HC was slightly but not significantly lower after cerivastatin (8,908 ± 1,488 vs. 7,460 ± 1,656 pmol/h; P = 0.4 [r = 0.45, P = 0.1]). Hepatic insulin extraction during the HC was not changed (placebo 9.38 ± 1.2 vs. cerivastatin 10.21 ± 1.4; P = 0.6).

After hyperglycemic stimulus, the average plasma glucose concentrations were similar for the two periods studied (placebo: 16.7 ± 0.6 vs. cerivastatin: 16.6 ± 0.6 mmol/l, P = 0.9; [r = 0.8, P = 0.000]; Fig. 3, left). Acute insulin response was substantially higher after treatment with cerivastatin (144 ± 25 to 210 ± 39 pmol/l; P = 0.003 [r = 0.9, P = 0.000]). The insulin index was increased significantly with cerivastatin treatment by 36% (P = 0.039 [r = 0.56, P = 0.03]).

The whole insulin-mediated glucose intake assessed during the EHC was significantly higher (6.5%) in patients after cerivastatin treatment (P < 0.05 [r = 0.87, P = 0.000]; Table 4). The steady-state insulin levels averaged during the EHC were higher in patients on placebo (817 ± 35 vs. 762 ± 20 pmol/l; P < 0.05 [r = 0.86, P = 0.000]). The mean insulin sensitivity index values during the EHC, a rate that corrects whole glucose uptake for average steady-state plasma insulin levels, was substantially higher (13.75%) after cerivastatin treatment (58.37 ± 3.7 vs. 66.39 ± 3.9 nmol · LBM⁻¹ · min⁻¹ · pmol⁻¹ · l^{− 1}; P < 0.01 [r = 0.77, P = 0.001]). The MCR of exogenously infused insulin during the EHC increased by 6% after treatment with cerivastatin (P < 0.05 [r = 0.83, P = 0.000]; Table 3).

In the fasting state, the energy expenditure and Rq were similar in both periods studied (0.65 ± 0.09 vs. 0.67 ± 0.07 kcal · m⁻² · min⁻¹ [P = 0.8] and 0.83 ± 0.01 vs. 0.83 ± 0.01 kcal · m⁻² · min⁻¹ [P = 0.9], placebo and cerivastatin, respectively; Table 4). During the steady state of the EHC, an increase in energy expenditure and Rq was observed with cerivastatin (1.69 ± 0.16 vs. 1.96 ± 0.15 kcal · m⁻² · min⁻¹, P < 0.01 [r = 0.87, P = 0.000]; and 0.91 ± 0.01 vs. 0.94 ± 0.01, P < 0.05, respectively; Table 4). The rate of glucose oxidation in the fasting state was similar. During EHC, glucose oxidation rates increased by 10.5% after treatment with cerivastatin, which represents 50% of whole-body glucose uptake as compared with placebo (16.1 ± 1.37 vs. 14.28 ± 1.48 μmol · LBM⁻¹ · min⁻¹; P < 0.05; Table 4).

Figure 4 shows profiles of TG, serum NEFA, and plasma lactate levels during the HC and EHC. The fasting serum TG concentration after 0.4 mg cerivastatin treatment was reduced by 14% (1.25 ± 0.06 vs. 1.46 ± 0.06 mmol/l; P < 0.05 [r = 0.51, P = 0.06]). Average serum TG levels during the HC also decreased by 14% with cerivastatin treatment (1.16 ± 0.1 vs. 1.36 ± 0.1 mmol/l; P < 0.01 [r = 0.8, P = 0.000]). Average serum TG concentrations during the EHC showed a greater decrease (27%) after cerivastatin therapy (1.17 ± 0.1 vs. 1.61 ± 0.1 mmol/l; P < 0.00 [r = 0.67, P = 0.006]. Serum FFA concentrations and lipid oxidation in the fasting state were similar in both periods (0.41 ± 0.03 vs. 0.39 ± 0.04 mmol/l, P = 0.2 [r = 0.62, P = 0.014]; and 0.76 ± 0.07 vs. 0.71 ± 0.06 mg · LBM⁻¹ · min⁻¹, P = 0.5 [r = 0.51, P = 0.052], placebo and cerivastatin, respectively; Table 4). Average FFA concentrations and lipid oxidation during the EHC were significantly higher in patients on placebo as compared with patients on cerivastatin (0.090 ± 0.016 vs. 0.042 ± 0.007 mmol/l, P < 0.01 [r = 0.50, P = 0.058]; and 0.48 ± 0.06 vs. 0.33 ± 0.05 mg · LBM⁻¹ · min^{− 1}, P < 0.01 [r = 0.61, P = 0.015], respectively; Fig. 3). Mean plasma lactate levels in the fasting state were similar for both treatment periods (Table 4). During the EHC, an increase was observed in average serum lactate levels in patients on cerivastatin treatment (1.24 ± 0.02 vs. 1.42 ± 0.03 mmol/l; P < 0.05 [r = 0.56, P = 0.054]; Fig. 3).

Overall, treatment with 0.4 mg cerivastatin was well tolerated. The following blood variables were analyzed: hemoglobin, erythrocytes, white blood cells, platelets, serum urea, creatinine, calcium, phosphorus, magnesium, sodium, chloride, potassium, creatine phosphokinase, lactate dehydrogenase, alanine transaminase, aspartate transaminase, alkaline phosphatase, γ-glutaryl transferase, bilirubin, and β₂-microglobulin. The most common adverse effects were mild and transient headache (one patient) and gastrointestinal (two patients) side effects, which did not require the discontinuation of therapy in any of the patients. No adverse events in the present study were indicative of muscular, renal, or hepatocellular toxicity.

The present data demonstrated that treatment with 0.4 mg cerivastatin increases insulin-mediated glucose uptake and improves first-phase insulin secretion in mild hyperglycemic patients in early state obese type 2 diabetes without pharmacological intervention. Type 2 diabetes involves a combination of peripheral tissue insulin resistance, excessive hepatic glucose production, and impaired insulin secretion (25). Various mechanisms of action of statins may involve glucose metabolism and insulin resistance. First, at mild to high doses, statins decrease plasma TGs, reducing the production of VLDL. This effect was observed in our study. Plasma TG levels averaged during EHC decreased by 27%, whereas serum FFA decreased by 50% (from 90 to 40 μmol/l/l) with cerivastatin treatment. During the fasting state and the HC, serum TGs were lowered by 14%. Finally, whole insulin-mediated glucose uptake during euglycemic clamp was seen to increase significantly (6.5%). In addition, in our study, the weight and mean HbA_1c level was not changed throughout the study, whereas fasting plasma total cholesterol decreased by 19.5%. In contrast, in type 2 diabetes, an increase in glucose disposal rates by 13% with metformin therapy and by 54% with troglitazone therapy has been shown (26). However, weight loss or decreases in hyperglycemia induced with these treatments could partially have an effect on insulin sensitivity (27). In humans, during infusion of a TG emulsion, inhibitory effects have also been associated with whole glucose uptake and glucose storage after 3–4 h (28). The classic mechanism of FFA-induced peripheral insulin resistance was proposed by Randle and colleagues (29,30) through the inhibition of pyruvate dehydrogenase, phosphofructokinase, and hexokinase, which affect glucose transport (GLUT4) activity. Recent data suggest an early direct effect of intracellular FFA content on insulin stimulation GLUT4 transporter activity (31, 32). Contrasting results in a few experimental studies regarding the effect of statins in glucose metabolism of patients with diabetes have been published (33–35). The nonhomogeneous selection of patients together with the low doses of statins and short times of EHC used could have limited some of these results. In addition, cerivastatin could have effects beyond the pharmacological class effect.

Second, after 3 months of statin treatment, a significantly higher rate of glucose oxidation was observed and represents ∼50% of the glucose uptake observed during the EHC. This was accompanied by an increase in the Rq and a decrease in FFA oxidation. In addition, glucose oxidation during the steady state in EHC was negatively correlated with average FFA concentrations during the placebo period (r = −0.452, P = 0.08) and during cerivastatin treatment (r = −0.475, P = 0.059). It has recently been demonstrated that changes throughout the physiological range in plasma FFA concentrations (from ∼50 to ∼800 μmol/l) exert an inhibitory effect on insulin-stimulated glucose oxidation and are evident in the first hours (10). It is of note that part of the remaining ∼ 50% of glycolytic nonoxidative glucose metabolism was due to an increase in glycolytic lactate observed during the euglycemic clamp with cerivastatin treatment. It has been previously shown that statins therapy may affect mitochondrial respiration and may be associated with a high blood lactate/pyruvate ratio (36). Whether 0.4 mg cerivastatin is related to mitochondrial dysfunction is unknown; the biochemical causes and precise functional significance of these defects remain to be seen.

First-phase insulin response increased substantially after cerivastatin treatment. Moreover, fasting insulin levels were slightly higher and hepatic insulin extraction increased after cerivastatin treatment in postabsorptive basal conditions. Direct effects on β-cell function of statins are little known. It has been shown that simvastatin inhibits glucose-induced insulin secretion in rat islet β-cells, whereas pravastatin had no effects (37). The lipid-lowering effects observed after 0.4 mg cerivastatin treatment, mainly on VLDL production with lower average plasma TG levels, could have an indirect effect in β-cell function increasing basal insulin secretor rates and first-phase insulin secretion (38). The continuous effect of 0.4 mg cerivastatin on VLDL production with lower averaged plasma TG levels could be related to higher basal insulin secretor rates and increased first-phase insulin secretion. Accordingly, in Zucker diabetic fatty rats, overaccumulation of triacylglycerol, the loss of function, and the loss of β-cells can be prevented by treatment with a thiazolidinedione if begun in the prediabetic stage (39). However, it is well established that the regulation of hepatic glucose production by insulin is a major determinant of blood glucose concentrations. In insulin-resistant states, chronic hyperinsulinemia suppresses the mRNA for insulin receptor substrate-2, a component of the insulin-signaling pathway in the liver, contributing to hepatic insulin resistance and inadequate gluconeogenesis (40). In our study, an approach to hepatic glucose uptake can be observed from OGTT results. Because the Δ I_30-0/ΔG_30-0 was increased and the incremental glucose area under the curve remained unchanged after cerivastatin treatment, it is possible that defects in liver insulin action on glucose metabolism remain unaltered after treatment with statins.

Last, during statin treatment, a significant reduction in plasma TNF-α and CRP concentrations was found. Recent data suggest that inflammation is involved in the early state of insulin resistance syndrome and predicted atherosclerotic disease (41, 42). TNF-α is produced by adipocytes with a decrease of TNF-α mRNA expression observed after weight loss (43). Moreover, TNF-α impairs insulin action by effects on the insulin receptor and its substrate (44). In addition, TNF-α induces interleukin-6 (IL-6) production (45), and CRP synthesis in the liver is stimulated by IL-6 (46). In our study, CRP levels were associated with body fat mass in both periods studied. However, as patient weight was not modified in the study, it is possible that reductions in CRP and proinflammatory cytokine levels with cerivastatin are a direct effect of the statins (47). In summary, our results show that 0.4 mg cerivastatin therapy increases whole insulin-mediated glucose uptake during the euglycemic clamp and improves β-cell function measured as first-phase insulin secretion in early-state obese type 2 diabetes without previous treatment and with clinical features of metabolic syndrome. Increased insulin action was associated with lowered plasma TG and FFA concentrations and an increase in the Rq. Moreover, there is evidence that cerivastatin treatment improved the proinflammatory state of these patients.

This study was funded by a grant from Bayer AG.

We thank the clinical investigators Alfredo Ortiz, PhD, Javier Fonseca, PhD, and Remedios Vigara, PhD, for providing the patients for this study. We are grateful to the pharmacologist Manuel Cárdenas for providing cerivastatin and placebo tablets on a randomized named-patient basis. We thank Cristobal Aguilera, PhD, chief of laboratory of Hospital Reina Sofia, and Trinidad López, PhD, for hormones and biochemical measurements. Our thanks to the nurses Rosa Berni, José Prados, Rafaela Garcia, José A. Muñoz, and Antonia Ruiz for excellent preparation of samples and patient care during tests.