点击显示 收起
【摘要】
Objective- The purpose of this research was to evaluate the short-term effects of pioglitazone (PIO) on high-density lipoprotein cholesterol (HDL-C) and other metabolic parameters in nondiabetic patients with metabolic syndrome (MetSyn).
Methods and Results- Sixty nondiabetic adults with low HDL-C and MetSyn were randomized to PIO or matching placebo for 12 weeks. PIO increased HDL-C by 15% and 14% at 6 and 12 weeks, respectively, compared with placebo ( P <0.001). Changes in HDL-C were correlated to changes in adiponectin ( r =0.34; P =0.01) but not to changes in insulin resistance. PIO did not affect serum triglycerides or low-density lipoprotein (LDL) cholesterol concentrations but reduced the number of small LDL particles by 18% ( P <0.001). PIO reduced median C-reactive protein levels by 31% ( P <0.001) and mean resistin levels by 10% ( P =0.02) while increasing mean serum levels of adiponectin by 111% ( P <0.001) compared with placebo. PIO did not affect weight and modestly decreased insulin resistance.
Conclusions- In nondiabetic patients with low HDL-C and MetSyn, PIO significantly raised HDL-C and favorably affected lipoprotein particle size, markers of inflammation, and adipokines without changes in triglycerides, LDL-C, or weight. These results suggest that PIO has direct effects on HDL, which may contribute to its antiatherogenic effects.
We performed a detailed evaluation of the lipid effects of pioglitazone (PIO) in nondiabetic patients with metabolic syndrome. The primary finding was that PIO raised high-density lipoprotein cholesterol by 14% compared with placebo, without significant changes triglycerides or low-density lipoprotein cholesterol. Our results suggest that PIO may be useful as an antiatherosclerotic strategy in this nondiabetic population.
【关键词】 atherosclerosis lipids lipoproteins inflammation metabolic syndrome
Introduction
Metabolic syndrome (MetSyn) is a cluster of metabolic abnormalities, which is characterized by abdominal obesity, insulin resistance (IR), dyslipidemia, elevated blood pressure, and a proinflammatory and prothrombotic milieu. 1 It is estimated that 47 million Americans (approximately one-fifth of the US population) have MetSyn in some form, and prevalence increases with age. 2 Although some have recently questioned the relative value of MetSyn, 3 others have demonstrated that MetSyn significantly increases the risk of both type 2 diabetes mellitus (T2DM) and atherosclerotic cardiovascular disease (ASCVD). 4 One of the hallmarks of MetSyn is depressed levels of high-density lipoprotein cholesterol (HDL-C), which occurs in 37% of patients with MetSyn. 2 Whereas HDL-C is not the primary target of lipid-modulating therapy, it is recognized as an important secondary target of therapy, and, thus, treatments that raise HDL-C may be important in reducing ASCVD risk. 5
Pioglitazone (PIO), a synthetic peroxisome proliferator-activated receptor ligand, is approved to treat hyperglycemia in patients with T2DM. In these patients, PIO significantly improves IR, lowers serum fasting triglycerides (TGs), and increases HDL-C, whereas it generally has a neutral effect on low-density lipoprotein cholesterol (LDL-C) concentrations and increasing body weight. 6-8 Smaller studies have also shown that PIO lowers blood pressure 9 and decreases markers of inflammation, whereas it raises adiponectin levels. 10 Thus, PIO improves several components of MetSyn; however, the mechanisms are uncertain and could be secondary to improved glycemic control. Limited data are available regarding the effects of thiazolidinediones (TZDs) in patients without diabetes. The purpose of this study was to evaluate the short-term effects of PIO compared with placebo on HDL-C, other lipoproteins, and metabolic parameters associated with ASCVD risk in nondiabetic patients with MetSyn.
Methods
Subjects
Men and women between the ages of 21 and 75 years with National Cholesterol Education Panel (NCEP)-defined MetSyn were recruited from the local Philadelphia metropolitan area between January 2003 and July 2004. Subjects were required to have an HDL-C concentration of <40 mg/dL in men and <50 mg/dL in women and have a diagnosis of MetSyn as defined by NCEP (Adult Treatment Panel III) to include 2 of the remaining criteria for MetSyn: waist circumference 102 cm (males) or 88 cm (females); systolic blood pressure (SBP) 130 mm Hg, diastolic blood pressure (DBP) 85 mm Hg, or 1 antihypertensive agent; fasting serum glucose 110 mg/dL; or fasting serum TG concentrations 150 mg/dL. Subjects were excluded for any of the following: T2DM; use of statins or niacin within 6 weeks or fibrates within 12 weeks; known ASCVD, congestive heart failure, or any major active medical 100 mm Hg; 800 2.0 mg/dL, liver function tests 2 times the upper limit of normal, or hemoglobin <10 mg/dL. T2DM was defined as self-reported history of the disease. In addition, a screening fasting glucose was obtained, and subjects 126 mg/dL were excluded. Subjects with a fasting glucose 110 mg/dL were given a 2-hour oral glucose tolerance test to exclude the presence of T2DM. The protocol was approved by both the General Clinical Research Center and the Institutional Review Board at the University of Pennsylvania. The study was fully explained to all of the volunteers, who provided written, informed consent.
Study Protocol
Subjects who met entry criteria were randomized to 30 mg of PIO per day or matching placebo for 6 weeks. At 6 weeks, subjects randomized to PIO increased the dosage to 45 mg/day, whereas placebo-treated subjects continued on this regimen for an additional 6 weeks. Subjects were seen for screening, baseline, 6-week and 12-week visits after a 12-hour fast. Participants were asked to maintain their usual diet and physical activity regimen starting at the time of the screening visit until the completion of the study.
Lipoproteins and Lipoprotein Subclasses
Lipid parameters were analyzed from EDTA plasma collected after a 12-hour fast in a Centers for Disease Control and Prevention-standardized lipid laboratory. Plasma total cholesterol (TC), HDL-C, and TG were measured enzymatically on a Cobas Fara II autoanalyzer (Roche Diagnostic Systems Inc) using Sigma reagents (Sigma Chemical Co). LDL-C and very low-density lipoprotein cholesterol (VLDL-C) levels were determined after ultracentrifugation at a density of 1.006 g/mL. Apolipoprotein (Apo) B and ApoAI were measured using Wako reagents (Wako Chemicals USA Inc). Lipoprotein (a) [Lp(a)] was measured using an immunoturbidometric assay (Wako Chemicals USA Inc.). Cholesteryl ester transfer protein (CETP) mass was measured using an ELISA assay (Wako Chemicals USA Inc). NMR spectroscopic assay was performed at LipoScience, Inc, as described previously 11 and modified. 12 Pattern B LDL particles (P) were defined as LDL-P size 20.5 nm.
Insulin Sensitivity
Insulin sensitivity was estimated using both fasting homeostasis model assessments of IR (HOMA-IR) index and the quantitative insulin sensitivity check index (QUICKI). HOMA-IR was defined as [fasting glucose (mmol/L) x fasting insulin (mU/mL)]/22.5. QUICKI was defined as 1/[log fasting insulin (mU/mL) + log fasting glucose (mg/dL)].
Inflammatory Markers and Adipocytokines
High-sensitivity C-reactive protein (CRP) was measured with an ultra high-sensitivity latex turbidimetric immunoassay (Wako Chemicals USA Inc) on a Hitachi 912 autoanalyzer (Roche Diagnostics). Plasma levels of interleukin (IL) 6, soluble tumor necrosis factor receptor 2, were measured using commercially available ELISAs according to the manufacturer?s guidelines (R&D Systems). Resistin was measured by an ELISA (Linco Research, Inc). The intra-assay and interassay coefficients of variation for pooled human plasma were 8.7% and 19.9%, respectively, for IL-6, 5.3% and 12.1% for soluble tumor necrosis factor receptor 2, and 4.1% and 4.6% for resistin. Leptin and adiponectin were measured using a commercially available radioimmunoassay by Linco Research, Inc. Lipoprotein-associated phospholipase A2 was measured using an ELISA (DiaDexus).
Clinical Parameters
SBP and DBP were measured with an automatic electronic sphygmomanometer in the sitting position after resting for 5 minutes. Body weight was measured on a calibrated scale (Tronix digital scale) in light clothes with shoes off to the nearest 0.1 kg, and height was measured to the nearest 0.1 cm. True waist circumference was measured midway between the tenth rib and the iliac crest with the subjects in the standing position, recorded at the end of a gentle expiration. Bioimpedance analysis (BIA) was performed after an overnight fast in the supine position using a Quantum II Analyzer. Subjects kept 3-day diet records, which were analyzed by Food Processor (NDS-R version 4.05_33).
Outcomes and Sample-Size Calculations
The primary end point of the study was the percentage change in HDL-C at 12 weeks from baseline. This was calculated as [(week 12 HDL - baseline HDL)/baseline HDL] x 100. Secondary end points included the percentage changes in all of the other major lipoproteins at both 6 and 12 weeks. We anticipated net increases in HDL-C of 8% between the PIO and placebo groups at both 6 and 12 weeks. Accounting for a 15% estimated dropout rate, we estimated that a sample size of 30 per group would provide 80% power to detect this difference between the 2 groups, using a 2-tailed of 0.05 and an estimated within-group SD of 10%.
Statistical Analysis
The primary analysis included data from all 60 subjects meeting the study criteria who were randomized. Two subjects (1 in each group) dropped out after completing both baseline and week 6 visits, leaving 58 subjects who completed all of the study procedures. Thus, for the primary lipoprotein analyses, data for week 6 were carried forward to week 12. For all of the other analyses, data from 58 subjects were included. For continuous variables, differences between treatment groups were evaluated using unpaired Student t tests or Wilcoxon rank-sum test. For discreet variables, group differences were assessed using a 2 test. Skewed variables, such as serum TG, VLDL-C, insulin, and HOMA-IR, were log transformed before analysis. For the primary lipid analyses, 1-way ANOVA was performed with percentage change from baseline as the dependent variable. For simple univariate correlational analysis, Pearson?s product-moment correlation was used and, when necessary, Spearman correlation coefficients. All of the analyses were performed using SAS software versions 8.2 (SAS Institute Inc.). All of the P values were 2-tailed.
Results
Patient Characteristics
Demographic characteristics of 60 enrolled subjects are presented in Table 1. There were no statistically significant differences between the groups in any baseline characteristic or in the baseline lipid profile at the =0.05 level. Although all of the subjects had a low HDL-C, 65%, 52%, and 75%, respectively, met NCEP criteria for elevated TG, BP, and waist circumference. Only 1 subject met the initial NCEP criteria for impaired fasting 110 mg/L), whereas 6 met the newer definition of 100 mg/dL).
TABLE 1. Baseline Characteristics of Study Population
Effects on Weight, Body Composition, Blood Pressure, Glucose, and Insulin Sensitivity
There was no change in weight in either group. BIA analysis suggested that PIO did not affect lean muscle mass or fat mass; basal metabolic rate; or intracellular, extracellular, and total body water (data not shown).
PIO did not significantly reduce blood pressure overall but 140 mm Hg (n=14). In this small group, PIO reduced mean SBP (SD) by -17 (12) mm Hg versus -5 (5) mm Hg in placebo at 6 weeks (net difference -12 mm Hg; P =0.02), which was maintained at 12 weeks ( P =0.03).
There were no significant effects on serum glucose or free fatty acids (FFAs), whereas fasting insulin levels were modestly reduced ( Table 2 ). PIO modestly decreased IR in this population as assessed by both HOMA-IR and QUICKI.
TABLE 2. Changes in Measures of Insulin Resistance, Inflammatory Markers, and Adipocytokines *
Effects on HDL
PIO increased HDL-C by 11% compared with a 4% reduction in HDL-C in the placebo group (net difference, +15%; P <0.001) at 6 weeks. The increase in HLD-C was maintained at 12 weeks ( Table 3 ) and did not differ by gender or baseline HDL-C 40 mg/dL. Weight, level of physical activity, total dietary fat intake, or alcohol consumption, all known to affect serum HDL-C, did not change in either group over 12 weeks (data not shown). Changes in HDL-C were not correlated with changes in HOMA ( r =-0.15; P =0.25), QUICKI ( r =0.12; P =0.36), or CRP ( r =-0.18; P =0.18) but were modestly positively correlated with changes in serum adiponectin ( r =0.34; P =0.01).
TABLE 3. Raw and Percent Changes in Serum Lipoproteins and NMR Particle Size after 12 Weeks of Therapy
At 6 weeks, PIO significantly increased mean apoA-I by 6.8% ( P =0.02), apoA-II by 7.7% ( P =0.008), and reduced CETP mass by 13% ( P =0.01) compared with placebo; however, by 12 weeks, only the change in apoA-II remained significant ( Table 3 ). There was no change in overall HDL particle number in response to PIO; however, there was a modest but significant shift away from smaller HDL particles and an increase in HDL particle size ( Table 3 ).
Effects on ApoB-Containing Lipoproteins
PIO did not affect serum TG concentrations at 6 weeks (PIO -7% versus placebo +7%; P =0.18) or 12 weeks ( Table 3 ). PIO did not affect LDL-C at 6 weeks (PIO +5% versus placebo +3%; P =0.59) or 12 weeks or total LDL particle number (LDL-P) at 12 weeks ( Table 3 ). However, PIO significantly reduced small LDL-P by 18% compared with placebo ( P <0.001) and favorably affected the ratio of large to small LDL-P, whereas it also increased mean LDL particle size ( Table 3 ). PIO raised median Lp(a) at both 6 weeks (PIO +6% versus placebo -2%; P =0.09) and 12 weeks ( Table 3 ).
Effects on Markers of Inflammation and Adipocytokines
PIO significantly reduced total white blood cell count and hs-CRP levels ( Table 2 ). The change in highly sensitive CRP was not correlated with the changes in HOMA-IR or QUICKI. PIO significantly reduced plasma resistin levels, which were correlated with changes in white blood cell ( r =0.4; P =0.001), CRP ( r =0.37; P =0.003), and IL-6 ( r =0.5; P <0.001) but not with changes in HOMA-IR or QUICKI. PIO had no effect on plasma leptin levels, but significantly increased adiponectin levels, which were inversely correlated with the change in hs-CRP ( r =-0.37; P =0.005; Table 2 ).
Safety and Tolerability
Overall PIO was very well tolerated. There were no reported serious adverse events and 25 nonserious adverse events across both groups. There were no reports of peripheral edema in either group. The adherence to treatment was excellent in both groups, with a mean adherence of 96% by pill count at 12 weeks ( P =0.7). PIO mildly but significantly lowered levels of alanine aminotransferase by 3.4 U/L at 6 weeks ( P =0.03) and 3.6 U/L at 12 weeks ( P =0.06), whereas -glutamyl transpeptidase was reduced by 7.2 U/L and 9.4 U/L at 6 and 12 weeks, respectively ( P <0.01).
Discussion
We show that in nondiabetic patients with low HDL-C and MetSyn, PIO significantly increased HDL-C by 15%, net of placebo, as early as 6 weeks at 30 mg, which was sustained with 45 mg for an additional 6 weeks. This robust HDL-C raising effect was not correlated with changes in IR as estimated by the validated indices, suggesting a direct effect. The mechanisms of the increase in HDL-C are unknown. Although PIO has also been shown to increase secretion of ApoA-I in vitro, 13 a recent human detailed lipoprotein kinetic study in 8 dyslipidemic patients with T2DM failed to find an increase in apoA-I production or a reduction in apoA-I catabolism. 14 The modest reduction in CETP mass with PIO treatment may contribute but is unlikely to be the sole cause. Another possible mechanism may be through upregulation of ABCA1 seen in vitro, 15 suggesting the possibility that TZDs may promote cellular cholesterol efflux and reverse cholesterol transport. PIO may also exert its effects on HDL indirectly via adiponectin, a protein that is robustly increased by TZDs, and has been associated in univariate analyses with HDL and is an independent predictor of ASCVD risk in multivariate analyses. 16 Finally, PIO may have some in vitro peroxisome proliferator-activated receptor activity. 13 Although PIO did not increase HDL particle number, it did cause a shift away from smaller HDL particles, which have recently been independently associated with a higher risk of recurrent cardiovascular events in the Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial (VA-HIT) population. 17
The increase in HDL-C that we report is higher than the increase reported in a recently published Prospective Pioglitazone Clinical Trial in Macrovascular Events trial 8 and identical to that reported in a recent, large head-to-head comparison of PIO versus rosiglitazone (ROSI; GLAI study) 18 in patients with T2DM. In the multicenter 28-week GLAI study of 735 patients, 45 mg of PIO raised HDL-C by 14.9% compared with a 7.8% increase by 8 mg of ROSI ( P <0.001). In nondiabetic populations, data suggest that PIO does not significantly affect HDL-C; 19,20 however, 1 study of PIO in nondiabetic patients with hypertension reported a significant HDL-C increase of 8%. 9 This general lack of reported effect could be explained by small samples sizes and a lack of selection of patients with low HDL-C. Our HDL-C findings are comparable to the increases in HDL-C seen in response to fibrates but higher than produced with statin drugs. 5 Whereas the increase in HDL-C with PIO is smaller than that seen with moderate-dose niacin, PIO is better tolerated and may have an advantage in terms of its effects on IR. However, unlike these 3 drugs, which have robust outcomes data, there is only a single trial suggesting that a TZD improves clinical cardiovascular outcomes. 8
The effect of PIO on TG has been modest. In the GLAI and Prospective Pioglitazone Clinical Trial in Macrovascular Events studies, PIO reduced mean fasting TG by 12% in patients with T2DM. 8,18 The mechanism was elucidated in a recent kinetic study showing that PIO increased the fractional clearance rate of VLDL-TG from the circulation most likely by increased lipoprotein lipase-mediated lipolysis and a reduction in apoC-III production. 14 In our nondiabetic population, we did not find an effect of PIO on either TG or apoC-III levels. It may also be that in patients with T2DM, the reduction in TG is directly linked to improvements in glycemic control and reduced FFA levels and that, in our relatively insulin-sensitive nondiabetic population, the very modest observed changes in FFA and insulin sensitivity were not sufficient to significantly lower TG.
We did not find that PIO altered LDL-C or apoB concentrations. This is in line with results from reviews 7 and meta-analyses 6 of PIO in patients with T2DM. Although the change in overall LDL particle number was not significant, we found that PIO significantly reduced the number of small LDL particles, consistent with previous results in hypertensive nondiabetic subjects. 20 Whereas smaller LDLs are believed to be more atherogenic, there has been inconsistent association of LDL size and coronary heart disease in multivariate analyses. 21 PIO produced a small but statistically significant increase in Lp(a) by 12 weeks. The clinical significance of this small increase is unknown but has been reported previously with troglitazone in diabetic patients. 22
In humans, TZDs have been shown to reduce various inflammatory markers in T2DM, 23 and, in some studies, this antiinflammatory effect does not seem to correlate to their insulin-sensitizing effects. 10,24 In a study of nondiabetic patients with MetSyn, ROSI 4 mg/day was shown to reduce CRP by 30%, 25 similar to our findings, and of a similar magnitude to the CRP reductions seen with higher doses of statins. 26 We also show that PIO increased serum levels of adiponectin, an adipocytokine believed to have cardioprotective properties, in both lean and obese adults. 27 Finally, we show that PIO significantly reduced resistin levels, a macrophage-derived inflammatory adipokine independently linked to atherosclerosis, 28 which is consistent with a previous report in patients with T2DM. 29 Also novel is the fact that we did not find a change in serum lipoprotein-associated phospholipase A2, a proposed novel inflammatory marker.
PIO did not affect weight or other measures of body composition. Reviews and meta-analyses report that TZDs, as a class, increase body weight by 2 to 3 kg in patients with T2DM, 6,7 but this weight gain seems linked to the reduction in HbA1C. 7 The limited data in nondiabetic patients are mixed, ranging from no weight gain 9,20 to a 4-kg increase. 30,31 Detailed anthropometric studies suggest that the increase in weight in response to TZDs is secondary to increased in total body fat, 30 which we did not find using BIA. It may be that our treatment period was too short to specifically address the longer-term weight implications of TZD in nondiabetic subjects. We also found that PIO reduced SBP in a small subset 140 mm Hg, consistent with studies in hypertensive, nondiabetic patients 9,32
There are several limitations to address. First, the treatment was of relatively short duration and, thus, our results cannot be confidently extrapolated beyond 12 weeks. However, our findings are similar to those seen in the head-to-head study, which was twice as long. 18 Second, the design does not permit the differentiation between treatment duration and dose titration effects. Third, we estimated IR using static indices, such as HOMA-IR, and not the better-validated dynamic measures, such as those obtained from frequently sampled intravenous glucose tolerance tests. Fourth, by not systematically performing an oral glucose tolerance test on all of the patients at screening, our population may have included a few patients with unrecognized T2DM. Finally, because we restricted enrollment to patients with MetSyn and low HDL-C, our results may not be generalizable to the more heterogeneous MetSyn population.
In conclusion, this is the largest study of PIO in nondiabetic patients to date. We found that PIO was safe, well tolerated, and improved several of the metabolic derangements seen in patients with NCEP-defined MetSyn. In this population, PIO significantly increased HDL-C, reduced the number of small LDL particles, reduced hs-CRP and resistin, and increased adiponectin. Overall, our results suggest a direct effect of PIO on lipids, inflammation, and adipokines, rather than effects secondary to improvements in IR. We await the completion of the Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication (DREAM) trial, which will definitively answer the question of whether TZDs can also reduce the risk of coronary heart disease events in nondiabetic subjects. 33
Acknowledgments
We thank Carl Shaw for carrying out the protocol, Anna Lillethun, Linda Morrell, and Kimberly McMahon for technical support, and the General Clinical Research Center nurses for patient care. We also thank Drs. Richard Dunbar, Peter Nonack, and Mark Weiner for their help with patient recruitment.
This work was supported (in part) by the following grants: investigator initiated grant from Takeda Pharmaceuticals of North America (TPNA) as well as K-23 AT-00058 (P.O.S.) and M01-RR00040 (General Clinical Research Center), both from the National Institutes of Health. D.J.R. is a recipient of a Burroughs Wellcome Fund Clinical Scientist Award in Translational Research and a Doris Duke Distinguished Clinical Scientist Award. TPNA also provided active drug and matching placebo for the trial.
【参考文献】
Grundy SM, Hansen B, Smith SC Jr, Cleeman JI, Kahn RA. Clinical management of metabolic syndrome: report of the Am Heart Association/National Heart, Lung, and Blood Institute/Am Diabetes Association conference on scientific issues related to management. Circulation. 2004; 109: 551-556.
Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA. 2002; 287: 356-359.
Kahn R, Buse J, Ferrannini E, Stern M. The Metabolic Syndrome: Time for a Critical Appraisal: Joint statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2005; 28: 2289-2304.
Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr, Spertus JA, Costa F. Diagnosis and management of the metabolic syndrome. An American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005; 112: 2735-2752.
Szapary PO, Rader DJ. The triglyceride-high-density lipoprotein axis: an important target of therapy? Am Heart J. 2004; 148: 211-221.
Chiquette E, Ramirez G, Defronzo R. A meta-analysis comparing the effect of thiazolidinediones on cardiovascular risk factors. Arch Intern Med. 2004; 164: 2097-2104.
Yki-Jarvinen H. Thiazolidinediones. N Engl J Med. 2004; 351: 1106-1118.
Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E, Massi-Benedetti M, Moules IK, Skene AM, Tan MH, Lefebvre PJ, Murray GD. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet. 2005; 366: 1279-1289.
Fullert S, Schneider F, Haak E, Rau H, Badenhoop K, Lubben G, Usadel KH, Konrad T. Effects of pioglitazone in nondiabetic patients with arterial hypertension: a double-blind, placebo-controlled study. J Clin Endocrinol Metab. 2002; 87: 5503-5506.
Satoh N, Ogawa Y, Usui T, Tagami T, Kono S, Uesugi H, Sugiyama H, Sugawara A, Yamada K, Shimatsu A, Kuzuya H, Nakao K. antiatherogenic effect of pioglitazone in type 2 diabetic patients irrespective of the responsiveness to its antidiabetic effect. Diabetes Care. 2003; 26: 2493-2499.
Otvos JD. Measurement of lipoprotein subclass profiles by nuclear magnetic resonance spectroscopy. Clin Lab. 2002; 48: 171-180.
Festa A, Williams K, Hanley AJG, Otvos JD, Goff DC, Wagenknecht LE, Haffner SM. Nuclear magnetic resonance lipoprotein abnormalities in prediabetic subjects in the insulin resistance atherosclerosis study. Circulation. 2005; 111: 3465-3472.
Sakamoto J, Kimura H, Moriyama S, Odaka H, Momose Y, Sugiyama Y, Sawada H. Activation of human peroxisome proliferator-activated receptor (PPAR) subtypes by pioglitazone. Biochem Biophys Res Commun. 2000; 278: 704-711.
Nagashima K, Lopez C, Donovan D, Ngai C, Fontanez N, Bensadoun A, Fruchart-Najib J, Holleran S, Cohn JS, Ramakrishnan R, Ginsberg HN. Effects of the PPARgamma agonist pioglitazone on lipoprotein metabolism in patients with type 2 diabetes mellitus. J Clin Invest. 2005; 115: 1323-1332.
Cabrero A, Cubero M, Llaverias G, Jove M, Planavila A, Alegret M, Sanchez R, Laguna JC, Carrera MV. Differential effects of peroxisome proliferator-activated receptor activators on the mRNA levels of genes involved in lipid metabolism in primary human monocyte-derived macrophages. Metabolism. 2003; 52: 652-657.
Pischon T, Girman CJ, Hotamisligil GS, Rifai N, Hu FB, Rimm EB. Plasma adiponectin levels and risk of myocardial infarction in men. JAMA. 2004; 291: 1730-1737.
Asztalos BF, Collins D, Cupples LA, Demissie S, Horvath KV, Bloomfield HE, Robins SJ, Schaefer EJ. Value of high-density lipoprotein (HDL) subpopulations in predicting recurrent cardiovascular events in the veterans affairs HDL intervention trial. Arterioscler Thromb Vasc Biol. 2005; 25: 2185-2191.
Goldberg RB, Kendall DM, Deeg MA, Buse JB, Zagar AJ, Pinaire JA, Tan MH, Khan MA, Perez AT, Jacober SJ, for the GLAI Study Investigators. A Comparison of lipid and glycemic effects of pioglitazone and rosiglitazone in patients with type 2 diabetes and dyslipidemia. Diabetes Care. 2005; 28: 1547-1554.
Guerrero-Romero F, Rodriguez-Moran M. Pioglitazone increases serum magnesium levels in glucose-intolerant subjects. A randomized, controlled trial. Exp Clin Endocrinol Diabetes. 2003; 111: 91-96.
Winkler K, Konrad T, Fullert S, Friedrich I, Destani R, Baumstark MW, Krebs K, Wieland H, Marz W. Pioglitazone reduces atherogenic dense LDL particles in nondiabetic patients with arterial hypertension: a double-blind, placebo-controlled study. Diabetes Care. 2003; 26: 2588-2594.
Cromwell WC, Otvos JD. Low-density lipoprotein particle number and risk for cardiovascular disease. Curr Atheroscler Rep. 2004; 6: 381-387.
Tack CJ, Smits P, Demacker PN, Stalenhoef AF. Effect of troglitazone on lipoprotein(a) levels in obese subjects. Diabetes Care. 1999; 22: 1752-1753.
Haffner SM, Greenberg AS, Weston WM, Chen H, Williams K, Freed MI. Effect of rosiglitazone treatment on nontraditional markers of cardiovascular disease in patients with type 2 diabetes mellitus. Circulation. 2002; 106: 679-684.
Pfutzner A, Marx N, Lubben G, Langenfeld M, Walcher D, Konrad T, Forst T. Improvement of cardiovascular risk markers by pioglitazone is independent from glycemic control: results from the pioneer study. J Am Coll Cardiol. 2005; 45: 1925-1931.
Wang TD, Chen WJ, Lin JW, Chen MF, Lee YT. Effects of rosiglitazone on endothelial function, C-reactive protein, and components of the metabolic syndrome in nondiabetic patients with the metabolic syndrome. Am J Cardiol. 2004; 93: 362-365.
Nissen SE. High-dose statins in acute coronary syndromes: not just lipid levels. JAMA. 2004; 292: 1365-1367.
Yu JG, Javorschi S, Hevener AL, Kruszynska YT, Norman RA, Sinha M, Olefsky JM. The effect of thiazolidinediones on plasma adiponectin levels in normal, obese, and type 2 diabetic subjects. Diabetes. 2002; 51: 2968-2974.
Reilly MP, Lehrke M, Wolfe ML, Rohatgi A, Lazar MA, Rader DJ. Resistin is an inflammatory marker of atherosclerosis in humans. Circulation. 2005; 111: 932-939.
Jung HS, Youn BS, Cho YM, Yu KY, Park HJ, Shin CS, Kim SY, Lee HK, Park KS. The effects of rosiglitazone and metformin on the plasma concentrations of resistin in patients with type 2 diabetes mellitus. Metabolism. 2005; 54: 314-320.
Shadid S, Jensen MD. Effect of pioglitazone on biochemical indices of non-alcoholic fatty liver disease in upper body obesity. Clin Gastroenterol Hepatol. 2003; 1: 384-387.
The Diabetes Prevention Program Research Group. Prevention of type 2 diabetes with troglitazone in the diabetes prevention program. Diabetes. 2005; 54: 1150-1156.
Raji A, Seely EW, Bekins SA, Williams growthhormone (GH), Simonson DC. Rosiglitazone improves insulin sensitivity and lowers blood pressure in hypertensive patients. Diabetes Care. 2003; 26: 172-178.
Gerstein HC, Yusuf S, Holman R, Bosch J, Pogue J. Rationale, design and recruitment characteristics of a large, simple international trial of diabetes prevention: the DREAM trial. Diabetologia. 2004; 47: 1519-1527.
作者单位：Division of General Internal Medicine (P.O.S., L.T.B.), Institute for Translational Medicine and Therapeutics (P.O.S., L.T.B., M.L.W., D.J.R.), Division of Cardiovascular Medicine (F.S., M.P.R.), and Center for Clinical Epidemiology and Biostatistics (P.O.S., J.C., M.P.R.), University of Pennsylvani