- Email: [email protected]
Relation of resistin levels with cardiovascular risk factors and insulin resistance in non-diabetes obese patients
diabetes research and clinical practice 84 (2009) 174–178
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice journal homepage: www.elsevier.com/locate/diabres
Relation of resistin levels with cardiovascular risk factors and insulin resistance in non-diabetes obese patients D.A. de Luis *, M. Gonzalez Sagrado, R. Conde, R. Aller, O. Izaola, J.L. Perez Castrillon, A. Duen˜as Institute of Endocrinology and Nutrition, Medicine School and Unit of Investigation, Hospital Rio Hortega, RD-056/0013 RETICEF, University of Valladolid, C/Los perales 16, Simancas 47130, Valladolid, Spain
article info
abstract
Article history:
Background: The aim of the present study was to explore the relationship of resistin levels
Received 23 September 2008
with these above mentioned factors.
Received in revised form
Subjects: A population of 213 obese was analyzed. A complete nutritional and biochemical
24 January 2009
evaluation was performed.
Accepted 26 January 2009
Results: The mean age was 45.1 + 16.7 years, the mean BMI was 35.6 + 5.7. Higher weight, fat
Published on line 24 February 2009
mass, fat free mass, waist to hip ratio, RMR, insulin and HOMA levels were observed in men than women. In all group, the analysis with a dependent variable (resistin) showed that fat
Keywords:
mass remained in the model (F = 2.48; p < 0.05), with an increase of 0.033 ng/ml (CI95%:
Adipocytokines
0.011–0.055) with each 1 kg of fat mass and a decrease of
Anthropometry
with each mmHg of diastolic blood pressure. In a second model (only women) (resistin), fat
0.29 ng/ml (CI95%:
Cardiovascular risk factors
mass remained in the model (F = 6.06; p < 0.05), with an increase of 0.037 ng/ml (CI95%: 0.032 ng/ml (CI95%:
0.53,
0.054,
0.01)
Obesity
0.015, 0.06) with each kg of fat mass and a decrease of
Resistin
with each mmHg of diastolic blood pressure. The third multivariate analysis (only men) did
0.01)
not show any relation among resistin levels and other parameters. Conclusion: Resistin levels are related with different cardiovascular risk and anthropometric parameters, without relation with insulin resistance. A sex interaction has been observed. # 2009 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Obesity and insulin resistance are associated with cardiovascular risk factors, including altered levels of inflammatory markers and adipocytokines [1]. The incidence of obesity and associated co morbidities is dramatically increasing worldwide. Obesity is characterized by a low grade systemic inflammation. Epidemiological evidence of this rising tide of obesity and associated pathologies has led, in the last years, to a dramatic increase of research on the role of adipose tissue as an active participant in controlling the body’s physiologic and pathologic processes [2].
The current view of adipose tissue is that of an active secretory organ, sending out and responding to signals that modulate appetite, insulin sensitivity, energy expenditure, inflammation and immunity. Adipocytokines are proteins produced mainly by adipose tissue [3]. These molecules have been shown to be involved in the pathogenesis of the metabolic syndrome and cardiovascular disease, for example; resistin, adiponectin, leptin, TNF alpha and interleukin 6. Resistin is a 12.5 kDa, cysteine-rich protein identified by screening for the genes that are induced during the differentiation of the adipocytes. Resistin is a dimeric protein that received its name from its apparent induction of insulin
* Corresponding author. E-mail address: [email protected] (D.A. de Luis). 0168-8227/$ – see front matter # 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2009.01.017
diabetes research and clinical practice 84 (2009) 174–178
resistance in mice. It belongs to the FIZZ (found in inflammatory zones) family. As noted above, it has been postulated that resistin mediates insulin resistance, but this role may be limited to rodents. Initial enthusiasm for this theory, which provides a direct link between adiposity and insulin resistance [4], was quickly quenched by contradictory findings in both mice and humans. It nonetheless appears safe to assert that resistin levels depend upon both nutritional state and hormonal environment. Its effects on insulin action have been extensively investigated in mice [5,6], where resistin is involved in lipid metabolism and hepatic glucose [7] and appears to be a major determinant of hepatic insulin resistance induced by high-fat diet [8]. In humans, data on the role of this adipocytokine in insulin sensitivity and obesity are controversial. Some authors indicated that increased serum resistin levels are associated with increased obesity, visceral fat [9] and type 2 diabetes [10], while other groups failed to observe such correlations [11]. Accordingly, the aim of the present study was to explore the relationship of resistin levels with cardiovascular risk factors, anthropometry, dietary intake and resting energy expenditure in obese patients without diabetes mellitus.
2.
Subjects and methods
2.1.
Subjects
A population of 213 obesity (BMI > 30) patients was analyzed in a prospective way and enrolled in a consecutive population way. These patients were studied in a Nutrition Clinic Unit after signed informed consent.
2.2.
Procedure
All patients with a 2 weeks weight-stabilization period before recruitment were enrolled. Weight, blood pressure, basal glucose, insulin resistance, C reactive protein (CRP), insulin, total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides blood and resistin levels were measured in fasting condition. Exclusion criteria included active infectious disease, history of cardiovascular disease or stroke during the previous 36 months, total cholesterol > 300 mg/dl, triglycerides > 400 mg/dl, blood pressure > 140/90 mmHg, fasting plasma glucose > 126 mg/dl, as well as the use of sulphonilurea, thiazolidinedionas, insulin, glucocorticoids, antineopplasic agents, angiotensin receptor blocker, angiotensin converting enzyme inhibitors, psychoactive medications.
2.3.
Assays
Serum total cholesterol and triglyceride concentrations were determined by enzymatic colorimetric assay (Technicon Instruments, Ltd., New York, NY, USA), while HDL-cholesterol was determined enzymatically in the supernatant after precipitation of other lipoproteins with dextran sulphatemagnesium. LDL-cholesterol was calculated using Friedewald formula. Lipoprotein (a) was determined by immunonephelometry with the aid of a Beckman array analyzer (Beckman Instruments, CA, USA). Plasma glucose levels were deter-
175
mined by using an automated glucose oxidase method (Glucose analyser 2, Beckman Instruments, Fullerton, CA). Insulin was measured by enzymatic colorimetry (Insulin, WAKO Pure-Chemical Industries, Osaka, Japan) and the homeostasis model assessment for insulin sensitivity (HOMA) was calculated using these values [12]. CRP was measured by immunoturbimetry (Roche Diagnostics GmbH, Mannheim, Germany), analytical sensivity 0.5 mg/dl. Resistin was measured by ELISA (Biovendor Laboratory, Inc., Brno, Czech Republic) with a sensitivity of 0.2 ng/ml with a normal range of 4–12 ng/ml.
2.4.
Indirect calorimetry
Indirect calorimetry (MedGem; Health Tech., Golden, USA) was performed in a standard way (fasting conditions and 8 h of previous resting). Resting metabolic rate (kcal/day) and oxygen consumption (ml/min) were calculated [13].
2.5.
Anthropometric measurements
Body weight was measured to an accuracy of 0.1 kg and body mass index computed as body weight/(height2). Waist (narrowest diameter between xiphoid process and iliac crest) and hip (widest diameter over greater trochanters) circumferences to derive waist to hip ratio (WHR) were measured, too. Tetrapolar body electrical bioimpedance was used to determine body composition [14]. An electric current of 0.8 mA and 50 kHz was produced by a calibrated signal generator (Biodynamics Model 310e, Seattle, WA, USA) and applied to the skin using adhesive electrodes placed on right-side limbs. Resistance and reactance were used to calculate total body water, fat and fat free mass. Blood pressure was measured twice after a 10 min rest with a random zero mercury sphygmomanometer, and averaged.
2.6.
Metabolic syndrome
Presence to the metabolic syndrome was assessed according to the National Cholesterol program Adult Treatment Panel III (NCP-ATPIII) guidelines. These criteria’s are defined as waist circumference > 102 cm in men and > 88 cm in women, HDLcholesterol < 45 mg/dl in men and < 50 mg/dl in women, blood pressure > 130/85 mmHg, triglycerides > 150 mg/dl and fasting plasma glucose > 110 mg/dl.
2.7.
Statistical analysis
The results were expressed as average standard deviation. The distribution of variables was analyzed with Kolmogorov– Smirnov test. Quantitative variables with normal distribution were analyzed with a two-tailed, paired Student’s t-test and ANOVA test. Non-parametric variables were analyzed with the Friedman and Wilcoxon tests. Qualitative variables were analyzed with the chi-square test, with Yates correction as necessary, and Fisher’s test. Correlation analysis was performed with Pearson and Spearmen tests. A multiple regression model (step by step) was used to study the dependent variable (resistin). A p-value under 0.05 was considered statistically significant.
176
diabetes research and clinical practice 84 (2009) 174–178
Table 1 – Clinical and epidemiological characteristics of study population. Parameters Age (years) BMI (kg/m2) SBP (mmHg) DBP (mmHg) Glucose (mg/dl) Total chol. (mg/dl) LDL-chol. (mg/dl) HDL-chol. (mg/dl) Insulin (mUI/l) HOMA CRP (mg/dl)
All (n = 213)
Male (n = 59)
Female (n = 154)
45.1 + 16.7 35.6 + 5.7 136.3 + 16 81.8 + 12.6 99.4 + 20.6 205.8 + 43 126.1 + 40 50.1 + 16 17.2 + 12.8 4.4 + 4.1 5.99 + 9.1
43.4 15.2 35.9 4.9 138.3 + 15.6 81.7 + 10 101.7 24 200.9 38.9 127.3 33 46.9 16.3 23.8 19.6 6.4 6.8 3.12 3.1
46.2 + 16 35.5 5.9 135.6 + 17 81.9 + 13.4 98.5 + 19.7 207.9 + 45 125.6 + 42.6 55.8 + 12.4 14.8 + 8.4 3.6 + 1.4 6.65 + 10.2
p ns ns ns ns ns ns ns <0.05 <0.05 <0.05 <0.05
Chol: cholesterol; CRP: C reactive protein; SBP: systolic blood pressure; DBP: diastolic blood+ pressure; ns: no significative.
Table 2 – Anthropometric characteristics by sex. Parameters Weight (kg) Fat free mass (kg) Fat mass (kg) Waist circumference Waist to hip ratio RMR (kcal/day)
All (n =213)
Male (n = 59)
Female (n = 154)
94.6 18.3 51.7 15 39.8 13.6 107.9 + 14.6 0.91 0.08 2135 + 749
104.3 17.5 71.1 13.9 31.3 11 113.6 + 13 0.96 0.07 2464 + 785
90.8 17.3 44.5 7.1 43.1 13 105.9 14.6 0.89 0.08 2025 706
p <0.05 <0.05 <0.05 ns <0.05 <0.05
RMR: resting metabolic rate; ns: no significance; normal values on local general population: weight, 71.1 + 10.4 kg; fat free mass, 50.4 + 12 kg; fat, 18 + 9.3 kg; waist circumference, 84 + 12 cm; waist to hip ratio, 0.78 + 0.06; RMR, 2345 + 890 kcal/day.
3.
Results
3.1.
Univariate analysis
Two hundred and thirteen patients gave informed consent and were enrolled in the study. The mean age was 45.1 + 16.7 years, the mean BMI was 35.6 + 5.7. Sex baseline biochemical characteristics of patients were presented in Table 1, with higher HDL-cholesterol and CRP levels in women than men and higher insulin and HOMA levels in men than women.
Table 2 shows differences between men and women, with higher weight, fat mass, fat free mass, waist to hip ratio and RMR in men than women. Table 3 shows the correlation analysis among resistin and other parameters. In all group, a positive correlation among resistin and fat mass, glucose, triglycerides and C reactive protein was observed. In all group a negative correlation among resistin and HDL-cholesterol, systolic and diastolic blood pressures were detected. In women, a positive correlation among resistin and weight, fat mass, glucose, triglycerides and C reactive protein levels was observed. In this group, a negative correlation among resistin and systolic and diastolic
Table 3 – General correlation analysis characteristics. Characteristics BMI Weight (kg) Fat mass (kg) Fat free mass (kg) Glucose (mg/dl) Triglycerides (mg/dl) HDL-chol. (mg/dl) CRP (mg/dl) Insulin (mg/dl) HOMA SBP (mmHg) DBP (mmHg)
r All 0.01 0.12 0.148 0.03 0.107 0.106 0.104 0.217 0.02 0.01 0.285 0.285
p ns ns 0.01 ns 0.04 0.043 0.049 0.007 ns ns 0.001 0.013
r Men 0.02 0.11 – 0.02 0.205 0.09 0.099 0.362 0.03 0.10 0.098 0.112
p ns ns ns ns 0.01 ns ns 0.008 ns ns ns ns
r Women 0.01 0.15 0.142 0.04 0.121 0.12 0.100 0.202 0.09 0.11 0.175 0.356
p ns 0.002 0.036 ns 0.049 0.049 ns 0.003 ns ns 0.004 0.037
BMI: body mass index; HDL-chol: high density cholesterol; HOMA: homeostasis model as assessment; SBP: systolic blood pressure; DPB: diastolic blood pressure.
diabetes research and clinical practice 84 (2009) 174–178
Table 4 – Levels of resistin according to number of determinants of the metabolic syndrome. Number determinants
All group
Men
0 1 2 3 4
3.44 + 0.21 2.88 + 1.6 3.45 + 1.27 3.4 + 1.7 3.4 + 0.6
3.46 + 0.41 3.74 + 1.7 3.49 + 1.3 3.14 + 1.5 3.3 + 1.1
(n = 127) (n = 33) (n = 28) (n = 14) and 5 (n = 11)
Women 3.44 + 1.9 2.02 + 0. 1 3.43 + 1.25 3.47 + 1.3 3.41 + 1.6
blood pressures was detected. In men, a positive correlation among resistin and C reactive protein and glucose levels was observed.
3.2.
Multivariate analysis
After univariate analysis, we performed a multivariate analysis. In all group, the analysis adjusted by age and sex with a dependent variable (resistin) showed that fat mass remained in the model (F = 2.48; p < 0.05), with an increase of 0.033 (CI95%: 0.011–0.055) ng/ml with each 1 kg of fat mass and a decrease of 0.29 (CI95%: 0.53, 0.01) ng/ml with each mmHg of diastolic blood pressure. In a second model (only women) adjusted by age with a dependent variable (resistin), fat mass remained in the model (F = 6.06; p < 0.05), with an increase of 0.037 (CI95%: 0.015, 0.06) ng/ml with each kg of fat mass and a decrease of 0.032 (CI95%: 0.054, 0.01) ng/ml with each mmHg of diastolic blood pressure. The third multivariate analysis (only men) adjusted by age did not show any relation among resistin levels and other parameters. Patients with metabolic syndrome (n = 86) assessed according to NCP-ATPIII had the same resistin levels than nonmetabolic patients (n = 127) (3.39 + 1.25 ng/ml vs. 3.37 + 1.48 ng/ ml: ns). Patients were classified according to the presence of increasing number determinants of the metabolic syndrome; resistin levels did not increase with the increasing presence of the metabolic syndrome determinants (Table 4).
4.
Discussion
The main finding of this study is that resistin levels are related with different cardiovascular risk factors and anthropometric parameters. However, these associations are specific-gender. Initial studies have demonstrated that obesity in mices, insulin resistance is associated with increased circulating resistin levels [6]. Given the incomplete homology between human and mouse resistin and the absence in humans of one of three murine resistin isoforms, resistin in humans may have a different physiologic role than that in mice. There is controversial correlation between body weight, adiposity, cardiovascular risk factors and resistin [15], as shown by our data. The role of resistin in the metabolic syndrome is controversial, too. Some articles, reported that in humans resistin levels correlate with insulin resistance and obesity [4,6,16], while other investigations failed to observe any correlation of metabolic markers with resistin levels, and no significant difference was observed in resistin levels in
177
subjects with metabolic syndrome compare to controls, as our results [17,18]. Utzschneider et al. [19] failed to find any correlation of resistin levels with metabolic syndrome. In our study and other previous study [20], resistin levels were correlated with fat mass, HDL-cholesterol, triglycerides, C reactive protein and blood pressure, without correlation with insulin resistance. These differences may be due by a different genetic background of the participants and different average of body mass index. One explanation for the lack of correlation with insulin resistance is that many hormones affect insulin resistance, and resistin may not be a major determinant of insulin resistance. Other fact is that, some of these correlations disappeared in men and they persisted in women. In a previous study [20], triglycerides, HDL-cholesterol and systolic blood pressure remains significant in women while in men only a correlation with glucose and C reactive protein. A different role of resistin according to gender was also observed by both authors [19,20]. A direct effect of resistin in vascular endothelial has been described, it induces the release of endothelin 1 and other molecules that change vascular tone [21]. Other studies have confirmed these gender-differences [22,23]. Pischon et al. [23] demonstrated that resistin levels were significantly associated with the presence of coronary heart disease only in women. In our study, some correlations disappeared in multivariant model adjusted by age and sex, showing these variables as confounding factors in this spurious association. Using ageadjusted resistin levels as the dependent variable, fat mass and diastolic blood pressure resulted to be independent predictors of resistin in women, while no factors were independent predictors of resistin in men. We could speculate that the gender-specific effect of resistin could depend on gender-specific different adipose tissue depots that are responsible for its production in female or male. We found that resistin concentrations were not related to BMI or other index of obesity such as waist circumference of waist to hip ratio. These data suggest that circulating resistin is not related to the pattern of fat distribution, only with total fat mass measured by bioimpedance, and this relation persisted with multivariate analysis in women. In univariate analysis, we observed a relation between resistin levels and C reactive protein. Kunnari et al. [24] showed a positive correlation with C reactive protein, too. These data suggest that in humans resistin could be related to the cardiovascular inflammatory state. Accordingly, the correlation between fasting glucose and resistin levels might be explained by this inflammatory state produce by C reactive protein secondary to resistin by a direct effect without insulin resistance [25]. There might be the possibility that resistin is rather associated with inflammation markers that would appear at different stages of metabolic syndrome development but not with an established metabolic syndrome. Secondary to this hypothesis there are studies about the reducing effects of thiazolidinedione (TZD) class of insulin sensitizers on resistin levels [26] and the notion that these agents could decrease CRP values is reported [27]. In conclusion, circulating resistin concentrations are associated with different cardiovascular risk factors and anthropometric variables in non-diabetic obese patients. A sex interaction has been observed. Further studies are needed
178
diabetes research and clinical practice 84 (2009) 174–178
to analyze this unclear topic area with clinical and therapeutically implications.
Conflict of interest There are no conflicts of interest.
references
[1] G. Fantuzzi, Adipose tissue, adipokines, and inflammation, J. Allergy Clin. Immunol. 115 (2005) 911–919. [2] U. Meier, A.M. Gressner, Endocrine regulation of energy metabolism: review of pathobiochemical and clinical chemical aspects of leptin, ghrelin, adiponectin and resistin, Clin. Chem. 50 (2004) 1511–1525. [3] M. Matsuda, I. Shimomura, M. Sata, Role of adiponectin in preventing vascular stenosis. The missing link of adipovascular axis, J. Biol. Chem. 277 (2002) 37487–37491. [4] C.M. Steppan, S.T. Bailey, S. Bhat, The hormone resistin links obesity to diabetes, Nature 409 (2001) 307–312. [5] H. Satoh, M.T. Nguyen, P.D. Miles, T. Immamura, I. Usui, J.M. Olefsky, Adenovirus-mediated chronic hyper resistinemia leads to in vivo insulin resistance in normal rats, J. Clin. Invest. 114 (2004) 224–231. [6] C.M. Steppan, M.A. Lazar, The current biology of resistin, J. Intern. Med. 255 (2004) 439–447. [7] N. Sato, K. Kobayashi, T. Inoguchi, N. Sonoda, M. Immamura, N. Sekiguchi, et al., Adenovirus-mediated high expression of resistin causes dyslipemia in mice, Endocrinology 146 (2005) 273–279. [8] E.D. Muse, S. Obici, S. Bhanot, B.P. Monia, R.A. McKay, M.W. Rajala, et al., Role of resistin in diet-induced hepatic insulin resistance, J. Clin. Invest. 114 (2004) 232–239. [9] C. Pagano, O. Marin, A. Calcagno, P. Schiappelli, C. Pilon, G. Milan, et al., Increased serum resistin in adults with Prader Willi syndrome is related to obesity and not to insulin resistance, J. Clin. Endocrinol. Metab. 90 (2005) 4335–4340. [10] M.S. Burnett, J.M. Devaney, R.J. Adenika, R. Lindsay, B.V. Howard, Cross-sectional associations of resistin, coronary heart disease and insulin resistance, J. Clin. Endocrinol. Metab. 91 (2006) 64–68. [11] C.C. Zou, L. Liang, F. Hong, J.F. Fu, Z.Y. Zhao, Serum adiponectin, resistin levels and non alcoholic fatty liver disease in obese children, Endocrine J. 52 (2005) 519–524. [12] D.R. Mathews, J.P. Hosker, A.S. Rudenski, B.A. Naylor, D. Treacher, Homesotasis model assessment: insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man, Diabetologia 28 (1985) 412–414. [13] I.D. Feurer, J.L. Mullen, Bedside measurement of resting energy expenditure and respiratory quotient via indirect calorimetry, Nutr. Clin. Pract. 1 (1986) 43–49.
[14] H. Lukaski, P.E. Johson, Assessment of fat-free mass using bioelectrical impedance measurements of the human body, Am. J. Clin. Nutr. 41 (4) (1985) 810–817. [15] G. Shetty, P. Economides, E. Horton, C. Mantzmoros, A. Veves, Circulating adiponectin and resistin levels in relation to metabolic factors, inflammatory markers, and vascular reactivity in diabetic patients and subjects at risk for diabetes, Diabetes Care 27 (2004) 2450–2454. [16] K. Azuma, E. Katsukawa, S. Oguchi, M. Murata, H. Yamazaki, A. Shimada, et al., Correlation between serum resistin level and adiposity in obese individuals, Obes. Res. 11 (2003) 997–1001. [17] A. Pfutzner, M. Langenfeld, T. Kunt, M. Lobig, T. Forst, Evaluation of human resitin assays with serum from patients with type 2 diabetes and different degrees of insulin resistance, Clin. Lab. 49 (2003) 571–576. [18] N. Iqbal, P. Seshadri, L. Stern, J. Loh, S. Kundu, T. Jafar, et al., Serum resitin is not associated with obesity or insulin resistance in humans, Eur. Rev. Med. Pharmacol. Sci. 9 (2005) 161–165. [19] K.M. Utzschneider, D.B. Carr, J. Tong, T.M. Wallace, R.L. Hull, S. Zraika, et al., Resistin is not associated with insulin sensitivity or the metabolic syndrome in humans, Diabetologia 48 (2005) 2330–2333. [20] G.D. Norata, M. Ongari, K. Garlaschelli, S. Raselli, L. Grigore, L. Catapano, Plasma resistin levels correlate with determinants of the metabolic syndrome, Eur. J. Endocrinol. 156 (2007) 279–284. [21] S. Verma, S.H. Li, C.H. Wang, P.W. Fedak, R.K. Li, R.D. Weisel, Resistin promotes endothelial cell activation: further evidence of adipokine-endothelial interaction, Circulation 108 (6) (2003) 736–740. [22] C.C. Chen, T.C. Li, C.I. Li, C.S. Liu, H.J. Wang, C.C. Lin, Serum resistin level among healthy subjects: relationship to anthropometric and metabolic parameters, Metabolism 54 (2005) 471–475. [23] T. Pischon, C.M. Bamberger, J. Kratzsch, B.C. Zyriax, P. Algenstaedt, H. Boeing, et al., Association of plasma resistin levels with coronary heart disease in women, Obes. Res. 13 (2005) 1764–1771. [24] A. Kunnari, O. Ukkola, M. Paivansali, Y.A. Kesaniemi, High plasma resistin level is associated with enhanced hsCRP and leucocytes, J. Clin. Endocrine Metab. 91 (2006) 2755– 2760. [25] S. Bo, R. Gambino, A. Pagani, S. Guidi, L. Gentile, M. Cassader, et al., Relationships between human serum resitin inflammatory markers and insulin resistance, Int. J. Obes. 29 (2005) 1315–1320. [26] M. Lehrke, M.P. Reilly, S.C. Millingyon, N. Iqbal, D.J. Rader, M.A. Lazar, An inflammatory cascade leading to hyperresistinemia in humans, Plos Med. 1 (2004) e45. [27] N. Chu, A.P.S. Kong, D.D. Kim, D. Amstrond, S. Baxi, R. Deustch, et al., Differential effects of metformin and troglitaznoe on cardiovascular risk factors in patients with DM2, Diabetes Care 25 (2002) 542–548.
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice journal homepage: www.elsevier.com/locate/diabres
Relation of resistin levels with cardiovascular risk factors and insulin resistance in non-diabetes obese patients D.A. de Luis *, M. Gonzalez Sagrado, R. Conde, R. Aller, O. Izaola, J.L. Perez Castrillon, A. Duen˜as Institute of Endocrinology and Nutrition, Medicine School and Unit of Investigation, Hospital Rio Hortega, RD-056/0013 RETICEF, University of Valladolid, C/Los perales 16, Simancas 47130, Valladolid, Spain
article info
abstract
Article history:
Background: The aim of the present study was to explore the relationship of resistin levels
Received 23 September 2008
with these above mentioned factors.
Received in revised form
Subjects: A population of 213 obese was analyzed. A complete nutritional and biochemical
24 January 2009
evaluation was performed.
Accepted 26 January 2009
Results: The mean age was 45.1 + 16.7 years, the mean BMI was 35.6 + 5.7. Higher weight, fat
Published on line 24 February 2009
mass, fat free mass, waist to hip ratio, RMR, insulin and HOMA levels were observed in men than women. In all group, the analysis with a dependent variable (resistin) showed that fat
Keywords:
mass remained in the model (F = 2.48; p < 0.05), with an increase of 0.033 ng/ml (CI95%:
Adipocytokines
0.011–0.055) with each 1 kg of fat mass and a decrease of
Anthropometry
with each mmHg of diastolic blood pressure. In a second model (only women) (resistin), fat
0.29 ng/ml (CI95%:
Cardiovascular risk factors
mass remained in the model (F = 6.06; p < 0.05), with an increase of 0.037 ng/ml (CI95%: 0.032 ng/ml (CI95%:
0.53,
0.054,
0.01)
Obesity
0.015, 0.06) with each kg of fat mass and a decrease of
Resistin
with each mmHg of diastolic blood pressure. The third multivariate analysis (only men) did
0.01)
not show any relation among resistin levels and other parameters. Conclusion: Resistin levels are related with different cardiovascular risk and anthropometric parameters, without relation with insulin resistance. A sex interaction has been observed. # 2009 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Obesity and insulin resistance are associated with cardiovascular risk factors, including altered levels of inflammatory markers and adipocytokines [1]. The incidence of obesity and associated co morbidities is dramatically increasing worldwide. Obesity is characterized by a low grade systemic inflammation. Epidemiological evidence of this rising tide of obesity and associated pathologies has led, in the last years, to a dramatic increase of research on the role of adipose tissue as an active participant in controlling the body’s physiologic and pathologic processes [2].
The current view of adipose tissue is that of an active secretory organ, sending out and responding to signals that modulate appetite, insulin sensitivity, energy expenditure, inflammation and immunity. Adipocytokines are proteins produced mainly by adipose tissue [3]. These molecules have been shown to be involved in the pathogenesis of the metabolic syndrome and cardiovascular disease, for example; resistin, adiponectin, leptin, TNF alpha and interleukin 6. Resistin is a 12.5 kDa, cysteine-rich protein identified by screening for the genes that are induced during the differentiation of the adipocytes. Resistin is a dimeric protein that received its name from its apparent induction of insulin
* Corresponding author. E-mail address: [email protected] (D.A. de Luis). 0168-8227/$ – see front matter # 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2009.01.017
diabetes research and clinical practice 84 (2009) 174–178
resistance in mice. It belongs to the FIZZ (found in inflammatory zones) family. As noted above, it has been postulated that resistin mediates insulin resistance, but this role may be limited to rodents. Initial enthusiasm for this theory, which provides a direct link between adiposity and insulin resistance [4], was quickly quenched by contradictory findings in both mice and humans. It nonetheless appears safe to assert that resistin levels depend upon both nutritional state and hormonal environment. Its effects on insulin action have been extensively investigated in mice [5,6], where resistin is involved in lipid metabolism and hepatic glucose [7] and appears to be a major determinant of hepatic insulin resistance induced by high-fat diet [8]. In humans, data on the role of this adipocytokine in insulin sensitivity and obesity are controversial. Some authors indicated that increased serum resistin levels are associated with increased obesity, visceral fat [9] and type 2 diabetes [10], while other groups failed to observe such correlations [11]. Accordingly, the aim of the present study was to explore the relationship of resistin levels with cardiovascular risk factors, anthropometry, dietary intake and resting energy expenditure in obese patients without diabetes mellitus.
2.
Subjects and methods
2.1.
Subjects
A population of 213 obesity (BMI > 30) patients was analyzed in a prospective way and enrolled in a consecutive population way. These patients were studied in a Nutrition Clinic Unit after signed informed consent.
2.2.
Procedure
All patients with a 2 weeks weight-stabilization period before recruitment were enrolled. Weight, blood pressure, basal glucose, insulin resistance, C reactive protein (CRP), insulin, total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides blood and resistin levels were measured in fasting condition. Exclusion criteria included active infectious disease, history of cardiovascular disease or stroke during the previous 36 months, total cholesterol > 300 mg/dl, triglycerides > 400 mg/dl, blood pressure > 140/90 mmHg, fasting plasma glucose > 126 mg/dl, as well as the use of sulphonilurea, thiazolidinedionas, insulin, glucocorticoids, antineopplasic agents, angiotensin receptor blocker, angiotensin converting enzyme inhibitors, psychoactive medications.
2.3.
Assays
Serum total cholesterol and triglyceride concentrations were determined by enzymatic colorimetric assay (Technicon Instruments, Ltd., New York, NY, USA), while HDL-cholesterol was determined enzymatically in the supernatant after precipitation of other lipoproteins with dextran sulphatemagnesium. LDL-cholesterol was calculated using Friedewald formula. Lipoprotein (a) was determined by immunonephelometry with the aid of a Beckman array analyzer (Beckman Instruments, CA, USA). Plasma glucose levels were deter-
175
mined by using an automated glucose oxidase method (Glucose analyser 2, Beckman Instruments, Fullerton, CA). Insulin was measured by enzymatic colorimetry (Insulin, WAKO Pure-Chemical Industries, Osaka, Japan) and the homeostasis model assessment for insulin sensitivity (HOMA) was calculated using these values [12]. CRP was measured by immunoturbimetry (Roche Diagnostics GmbH, Mannheim, Germany), analytical sensivity 0.5 mg/dl. Resistin was measured by ELISA (Biovendor Laboratory, Inc., Brno, Czech Republic) with a sensitivity of 0.2 ng/ml with a normal range of 4–12 ng/ml.
2.4.
Indirect calorimetry
Indirect calorimetry (MedGem; Health Tech., Golden, USA) was performed in a standard way (fasting conditions and 8 h of previous resting). Resting metabolic rate (kcal/day) and oxygen consumption (ml/min) were calculated [13].
2.5.
Anthropometric measurements
Body weight was measured to an accuracy of 0.1 kg and body mass index computed as body weight/(height2). Waist (narrowest diameter between xiphoid process and iliac crest) and hip (widest diameter over greater trochanters) circumferences to derive waist to hip ratio (WHR) were measured, too. Tetrapolar body electrical bioimpedance was used to determine body composition [14]. An electric current of 0.8 mA and 50 kHz was produced by a calibrated signal generator (Biodynamics Model 310e, Seattle, WA, USA) and applied to the skin using adhesive electrodes placed on right-side limbs. Resistance and reactance were used to calculate total body water, fat and fat free mass. Blood pressure was measured twice after a 10 min rest with a random zero mercury sphygmomanometer, and averaged.
2.6.
Metabolic syndrome
Presence to the metabolic syndrome was assessed according to the National Cholesterol program Adult Treatment Panel III (NCP-ATPIII) guidelines. These criteria’s are defined as waist circumference > 102 cm in men and > 88 cm in women, HDLcholesterol < 45 mg/dl in men and < 50 mg/dl in women, blood pressure > 130/85 mmHg, triglycerides > 150 mg/dl and fasting plasma glucose > 110 mg/dl.
2.7.
Statistical analysis
The results were expressed as average standard deviation. The distribution of variables was analyzed with Kolmogorov– Smirnov test. Quantitative variables with normal distribution were analyzed with a two-tailed, paired Student’s t-test and ANOVA test. Non-parametric variables were analyzed with the Friedman and Wilcoxon tests. Qualitative variables were analyzed with the chi-square test, with Yates correction as necessary, and Fisher’s test. Correlation analysis was performed with Pearson and Spearmen tests. A multiple regression model (step by step) was used to study the dependent variable (resistin). A p-value under 0.05 was considered statistically significant.
176
diabetes research and clinical practice 84 (2009) 174–178
Table 1 – Clinical and epidemiological characteristics of study population. Parameters Age (years) BMI (kg/m2) SBP (mmHg) DBP (mmHg) Glucose (mg/dl) Total chol. (mg/dl) LDL-chol. (mg/dl) HDL-chol. (mg/dl) Insulin (mUI/l) HOMA CRP (mg/dl)
All (n = 213)
Male (n = 59)
Female (n = 154)
45.1 + 16.7 35.6 + 5.7 136.3 + 16 81.8 + 12.6 99.4 + 20.6 205.8 + 43 126.1 + 40 50.1 + 16 17.2 + 12.8 4.4 + 4.1 5.99 + 9.1
43.4 15.2 35.9 4.9 138.3 + 15.6 81.7 + 10 101.7 24 200.9 38.9 127.3 33 46.9 16.3 23.8 19.6 6.4 6.8 3.12 3.1
46.2 + 16 35.5 5.9 135.6 + 17 81.9 + 13.4 98.5 + 19.7 207.9 + 45 125.6 + 42.6 55.8 + 12.4 14.8 + 8.4 3.6 + 1.4 6.65 + 10.2
p ns ns ns ns ns ns ns <0.05 <0.05 <0.05 <0.05
Chol: cholesterol; CRP: C reactive protein; SBP: systolic blood pressure; DBP: diastolic blood+ pressure; ns: no significative.
Table 2 – Anthropometric characteristics by sex. Parameters Weight (kg) Fat free mass (kg) Fat mass (kg) Waist circumference Waist to hip ratio RMR (kcal/day)
All (n =213)
Male (n = 59)
Female (n = 154)
94.6 18.3 51.7 15 39.8 13.6 107.9 + 14.6 0.91 0.08 2135 + 749
104.3 17.5 71.1 13.9 31.3 11 113.6 + 13 0.96 0.07 2464 + 785
90.8 17.3 44.5 7.1 43.1 13 105.9 14.6 0.89 0.08 2025 706
p <0.05 <0.05 <0.05 ns <0.05 <0.05
RMR: resting metabolic rate; ns: no significance; normal values on local general population: weight, 71.1 + 10.4 kg; fat free mass, 50.4 + 12 kg; fat, 18 + 9.3 kg; waist circumference, 84 + 12 cm; waist to hip ratio, 0.78 + 0.06; RMR, 2345 + 890 kcal/day.
3.
Results
3.1.
Univariate analysis
Two hundred and thirteen patients gave informed consent and were enrolled in the study. The mean age was 45.1 + 16.7 years, the mean BMI was 35.6 + 5.7. Sex baseline biochemical characteristics of patients were presented in Table 1, with higher HDL-cholesterol and CRP levels in women than men and higher insulin and HOMA levels in men than women.
Table 2 shows differences between men and women, with higher weight, fat mass, fat free mass, waist to hip ratio and RMR in men than women. Table 3 shows the correlation analysis among resistin and other parameters. In all group, a positive correlation among resistin and fat mass, glucose, triglycerides and C reactive protein was observed. In all group a negative correlation among resistin and HDL-cholesterol, systolic and diastolic blood pressures were detected. In women, a positive correlation among resistin and weight, fat mass, glucose, triglycerides and C reactive protein levels was observed. In this group, a negative correlation among resistin and systolic and diastolic
Table 3 – General correlation analysis characteristics. Characteristics BMI Weight (kg) Fat mass (kg) Fat free mass (kg) Glucose (mg/dl) Triglycerides (mg/dl) HDL-chol. (mg/dl) CRP (mg/dl) Insulin (mg/dl) HOMA SBP (mmHg) DBP (mmHg)
r All 0.01 0.12 0.148 0.03 0.107 0.106 0.104 0.217 0.02 0.01 0.285 0.285
p ns ns 0.01 ns 0.04 0.043 0.049 0.007 ns ns 0.001 0.013
r Men 0.02 0.11 – 0.02 0.205 0.09 0.099 0.362 0.03 0.10 0.098 0.112
p ns ns ns ns 0.01 ns ns 0.008 ns ns ns ns
r Women 0.01 0.15 0.142 0.04 0.121 0.12 0.100 0.202 0.09 0.11 0.175 0.356
p ns 0.002 0.036 ns 0.049 0.049 ns 0.003 ns ns 0.004 0.037
BMI: body mass index; HDL-chol: high density cholesterol; HOMA: homeostasis model as assessment; SBP: systolic blood pressure; DPB: diastolic blood pressure.
diabetes research and clinical practice 84 (2009) 174–178
Table 4 – Levels of resistin according to number of determinants of the metabolic syndrome. Number determinants
All group
Men
0 1 2 3 4
3.44 + 0.21 2.88 + 1.6 3.45 + 1.27 3.4 + 1.7 3.4 + 0.6
3.46 + 0.41 3.74 + 1.7 3.49 + 1.3 3.14 + 1.5 3.3 + 1.1
(n = 127) (n = 33) (n = 28) (n = 14) and 5 (n = 11)
Women 3.44 + 1.9 2.02 + 0. 1 3.43 + 1.25 3.47 + 1.3 3.41 + 1.6
blood pressures was detected. In men, a positive correlation among resistin and C reactive protein and glucose levels was observed.
3.2.
Multivariate analysis
After univariate analysis, we performed a multivariate analysis. In all group, the analysis adjusted by age and sex with a dependent variable (resistin) showed that fat mass remained in the model (F = 2.48; p < 0.05), with an increase of 0.033 (CI95%: 0.011–0.055) ng/ml with each 1 kg of fat mass and a decrease of 0.29 (CI95%: 0.53, 0.01) ng/ml with each mmHg of diastolic blood pressure. In a second model (only women) adjusted by age with a dependent variable (resistin), fat mass remained in the model (F = 6.06; p < 0.05), with an increase of 0.037 (CI95%: 0.015, 0.06) ng/ml with each kg of fat mass and a decrease of 0.032 (CI95%: 0.054, 0.01) ng/ml with each mmHg of diastolic blood pressure. The third multivariate analysis (only men) adjusted by age did not show any relation among resistin levels and other parameters. Patients with metabolic syndrome (n = 86) assessed according to NCP-ATPIII had the same resistin levels than nonmetabolic patients (n = 127) (3.39 + 1.25 ng/ml vs. 3.37 + 1.48 ng/ ml: ns). Patients were classified according to the presence of increasing number determinants of the metabolic syndrome; resistin levels did not increase with the increasing presence of the metabolic syndrome determinants (Table 4).
4.
Discussion
The main finding of this study is that resistin levels are related with different cardiovascular risk factors and anthropometric parameters. However, these associations are specific-gender. Initial studies have demonstrated that obesity in mices, insulin resistance is associated with increased circulating resistin levels [6]. Given the incomplete homology between human and mouse resistin and the absence in humans of one of three murine resistin isoforms, resistin in humans may have a different physiologic role than that in mice. There is controversial correlation between body weight, adiposity, cardiovascular risk factors and resistin [15], as shown by our data. The role of resistin in the metabolic syndrome is controversial, too. Some articles, reported that in humans resistin levels correlate with insulin resistance and obesity [4,6,16], while other investigations failed to observe any correlation of metabolic markers with resistin levels, and no significant difference was observed in resistin levels in
177
subjects with metabolic syndrome compare to controls, as our results [17,18]. Utzschneider et al. [19] failed to find any correlation of resistin levels with metabolic syndrome. In our study and other previous study [20], resistin levels were correlated with fat mass, HDL-cholesterol, triglycerides, C reactive protein and blood pressure, without correlation with insulin resistance. These differences may be due by a different genetic background of the participants and different average of body mass index. One explanation for the lack of correlation with insulin resistance is that many hormones affect insulin resistance, and resistin may not be a major determinant of insulin resistance. Other fact is that, some of these correlations disappeared in men and they persisted in women. In a previous study [20], triglycerides, HDL-cholesterol and systolic blood pressure remains significant in women while in men only a correlation with glucose and C reactive protein. A different role of resistin according to gender was also observed by both authors [19,20]. A direct effect of resistin in vascular endothelial has been described, it induces the release of endothelin 1 and other molecules that change vascular tone [21]. Other studies have confirmed these gender-differences [22,23]. Pischon et al. [23] demonstrated that resistin levels were significantly associated with the presence of coronary heart disease only in women. In our study, some correlations disappeared in multivariant model adjusted by age and sex, showing these variables as confounding factors in this spurious association. Using ageadjusted resistin levels as the dependent variable, fat mass and diastolic blood pressure resulted to be independent predictors of resistin in women, while no factors were independent predictors of resistin in men. We could speculate that the gender-specific effect of resistin could depend on gender-specific different adipose tissue depots that are responsible for its production in female or male. We found that resistin concentrations were not related to BMI or other index of obesity such as waist circumference of waist to hip ratio. These data suggest that circulating resistin is not related to the pattern of fat distribution, only with total fat mass measured by bioimpedance, and this relation persisted with multivariate analysis in women. In univariate analysis, we observed a relation between resistin levels and C reactive protein. Kunnari et al. [24] showed a positive correlation with C reactive protein, too. These data suggest that in humans resistin could be related to the cardiovascular inflammatory state. Accordingly, the correlation between fasting glucose and resistin levels might be explained by this inflammatory state produce by C reactive protein secondary to resistin by a direct effect without insulin resistance [25]. There might be the possibility that resistin is rather associated with inflammation markers that would appear at different stages of metabolic syndrome development but not with an established metabolic syndrome. Secondary to this hypothesis there are studies about the reducing effects of thiazolidinedione (TZD) class of insulin sensitizers on resistin levels [26] and the notion that these agents could decrease CRP values is reported [27]. In conclusion, circulating resistin concentrations are associated with different cardiovascular risk factors and anthropometric variables in non-diabetic obese patients. A sex interaction has been observed. Further studies are needed
178
diabetes research and clinical practice 84 (2009) 174–178
to analyze this unclear topic area with clinical and therapeutically implications.
Conflict of interest There are no conflicts of interest.
references
[1] G. Fantuzzi, Adipose tissue, adipokines, and inflammation, J. Allergy Clin. Immunol. 115 (2005) 911–919. [2] U. Meier, A.M. Gressner, Endocrine regulation of energy metabolism: review of pathobiochemical and clinical chemical aspects of leptin, ghrelin, adiponectin and resistin, Clin. Chem. 50 (2004) 1511–1525. [3] M. Matsuda, I. Shimomura, M. Sata, Role of adiponectin in preventing vascular stenosis. The missing link of adipovascular axis, J. Biol. Chem. 277 (2002) 37487–37491. [4] C.M. Steppan, S.T. Bailey, S. Bhat, The hormone resistin links obesity to diabetes, Nature 409 (2001) 307–312. [5] H. Satoh, M.T. Nguyen, P.D. Miles, T. Immamura, I. Usui, J.M. Olefsky, Adenovirus-mediated chronic hyper resistinemia leads to in vivo insulin resistance in normal rats, J. Clin. Invest. 114 (2004) 224–231. [6] C.M. Steppan, M.A. Lazar, The current biology of resistin, J. Intern. Med. 255 (2004) 439–447. [7] N. Sato, K. Kobayashi, T. Inoguchi, N. Sonoda, M. Immamura, N. Sekiguchi, et al., Adenovirus-mediated high expression of resistin causes dyslipemia in mice, Endocrinology 146 (2005) 273–279. [8] E.D. Muse, S. Obici, S. Bhanot, B.P. Monia, R.A. McKay, M.W. Rajala, et al., Role of resistin in diet-induced hepatic insulin resistance, J. Clin. Invest. 114 (2004) 232–239. [9] C. Pagano, O. Marin, A. Calcagno, P. Schiappelli, C. Pilon, G. Milan, et al., Increased serum resistin in adults with Prader Willi syndrome is related to obesity and not to insulin resistance, J. Clin. Endocrinol. Metab. 90 (2005) 4335–4340. [10] M.S. Burnett, J.M. Devaney, R.J. Adenika, R. Lindsay, B.V. Howard, Cross-sectional associations of resistin, coronary heart disease and insulin resistance, J. Clin. Endocrinol. Metab. 91 (2006) 64–68. [11] C.C. Zou, L. Liang, F. Hong, J.F. Fu, Z.Y. Zhao, Serum adiponectin, resistin levels and non alcoholic fatty liver disease in obese children, Endocrine J. 52 (2005) 519–524. [12] D.R. Mathews, J.P. Hosker, A.S. Rudenski, B.A. Naylor, D. Treacher, Homesotasis model assessment: insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man, Diabetologia 28 (1985) 412–414. [13] I.D. Feurer, J.L. Mullen, Bedside measurement of resting energy expenditure and respiratory quotient via indirect calorimetry, Nutr. Clin. Pract. 1 (1986) 43–49.
[14] H. Lukaski, P.E. Johson, Assessment of fat-free mass using bioelectrical impedance measurements of the human body, Am. J. Clin. Nutr. 41 (4) (1985) 810–817. [15] G. Shetty, P. Economides, E. Horton, C. Mantzmoros, A. Veves, Circulating adiponectin and resistin levels in relation to metabolic factors, inflammatory markers, and vascular reactivity in diabetic patients and subjects at risk for diabetes, Diabetes Care 27 (2004) 2450–2454. [16] K. Azuma, E. Katsukawa, S. Oguchi, M. Murata, H. Yamazaki, A. Shimada, et al., Correlation between serum resistin level and adiposity in obese individuals, Obes. Res. 11 (2003) 997–1001. [17] A. Pfutzner, M. Langenfeld, T. Kunt, M. Lobig, T. Forst, Evaluation of human resitin assays with serum from patients with type 2 diabetes and different degrees of insulin resistance, Clin. Lab. 49 (2003) 571–576. [18] N. Iqbal, P. Seshadri, L. Stern, J. Loh, S. Kundu, T. Jafar, et al., Serum resitin is not associated with obesity or insulin resistance in humans, Eur. Rev. Med. Pharmacol. Sci. 9 (2005) 161–165. [19] K.M. Utzschneider, D.B. Carr, J. Tong, T.M. Wallace, R.L. Hull, S. Zraika, et al., Resistin is not associated with insulin sensitivity or the metabolic syndrome in humans, Diabetologia 48 (2005) 2330–2333. [20] G.D. Norata, M. Ongari, K. Garlaschelli, S. Raselli, L. Grigore, L. Catapano, Plasma resistin levels correlate with determinants of the metabolic syndrome, Eur. J. Endocrinol. 156 (2007) 279–284. [21] S. Verma, S.H. Li, C.H. Wang, P.W. Fedak, R.K. Li, R.D. Weisel, Resistin promotes endothelial cell activation: further evidence of adipokine-endothelial interaction, Circulation 108 (6) (2003) 736–740. [22] C.C. Chen, T.C. Li, C.I. Li, C.S. Liu, H.J. Wang, C.C. Lin, Serum resistin level among healthy subjects: relationship to anthropometric and metabolic parameters, Metabolism 54 (2005) 471–475. [23] T. Pischon, C.M. Bamberger, J. Kratzsch, B.C. Zyriax, P. Algenstaedt, H. Boeing, et al., Association of plasma resistin levels with coronary heart disease in women, Obes. Res. 13 (2005) 1764–1771. [24] A. Kunnari, O. Ukkola, M. Paivansali, Y.A. Kesaniemi, High plasma resistin level is associated with enhanced hsCRP and leucocytes, J. Clin. Endocrine Metab. 91 (2006) 2755– 2760. [25] S. Bo, R. Gambino, A. Pagani, S. Guidi, L. Gentile, M. Cassader, et al., Relationships between human serum resitin inflammatory markers and insulin resistance, Int. J. Obes. 29 (2005) 1315–1320. [26] M. Lehrke, M.P. Reilly, S.C. Millingyon, N. Iqbal, D.J. Rader, M.A. Lazar, An inflammatory cascade leading to hyperresistinemia in humans, Plos Med. 1 (2004) e45. [27] N. Chu, A.P.S. Kong, D.D. Kim, D. Amstrond, S. Baxi, R. Deustch, et al., Differential effects of metformin and troglitaznoe on cardiovascular risk factors in patients with DM2, Diabetes Care 25 (2002) 542–548.