Association of Inflammation Markers with Impaired Insulin Sensitivity ...

0021-972X/03/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 88(11):5321–5326 Copyright © 2003 by The Endocrine Society doi: 10.1210/jc.2003-030508

Association of Inflammation Markers with Impaired Insulin Sensitivity and Coagulative Activation in Obese Healthy Women MARIO ROMANO, MARIA TERESA GUAGNANO, GIOVANNI PACINI, SERGIO VIGNERI, ANGELA FALCO, MARINA MARINOPICCOLI, MARIA ROSARIA MANIGRASSO, STEFANIA BASILI, AND GIOVANNI DAVÌ Departments of Medicine and Aging (M.T.G., A.F., M.M., M.R.M., G.D.) and Biomedical Sciences (M.R.), University of Chieti “G. D’Annunzio,” School of Medicine, 66013 Chieti; Metabolic Unit (G.P.), Institute of Biomedical Engineering (ISIBCNR), 35127 Padova; Department of Medicine (S.V.), University of Palermo, 90128 Palermo; and University of Rome “La Sapienza” (S.B.), 00161 Rome, Italy Insulin resistance is associated with a low chronic inflammatory state. In this study we investigated the relationship between impaired insulin sensitivity and selected markers of inflammation and thrombin generation in obese healthy women. We examined 32 healthy obese women (body mass index > 28), with normal insulin sensitivity (NIS, n ⴝ 14) or impaired insulin sensitivity (n ⴝ 18), and 10 nonobese women (body mass index < 25). Impaired insulin sensitivity patients had significantly higher levels of C-reactive protein (CRP), TGF-␤1, plasminogen activator inhibitor-1 (PAI-1), activated factor VII (VIIa), and prothrombin fragment 1 ⴙ 2 (F1 ⴙ 2) compared with either control subjects or NIS patients. On the other hand, NIS patients had higher CRP, TGF-␤1, PAI-1, and

O

BESITY, INSULIN RESISTANCE syndrome, and atherosclerosis are closely linked phenomena, often connected with a chronic low-level inflammatory state (1, 2) and a prothrombotic hypofibrinolytic condition (3). Increased levels of inflammation markers, such as C-reactive protein (CRP) and TNF-␣, and of the acute-phase proteins, fibrinogen and plasminogen activator inhibitor-1 (PAI-1), have been found in the insulin resistance syndrome (4 –7). However, obesity is associated with increased procoagulant activity and decreased fibrinolytic potential in mice (8). Adipocytes, not just a fat storage, are metabolically active secretory cells that may alter the hemostatic balance by abnormally expressing proteins like PAI-1 or tissue factor (TF) (9, 10). Cultured adipocytes exposed to TNF-␣ also express the multifunctional cytokine TGF-␤ (11). TGF-␤ may be involved in the regulation of circulating PAI-1 levels by adipose tissue (11), as well as by other cellular types, including endothelial cells (12). Moreover, it potently induces TF expression by adipocytes (10). Thus, a complex interplay between inflammatory cytokines and coagulation/fibrinolysis factors may result in a procoagulant/prothrombotic state in obesity and insulin resistance. Abbreviations: BMI, Body mass index; CRP, C-reactive protein; F1 ⫹ 2, prothrombin fragment 1 ⫹ 2; factor VIIa, activated factor VIIa; FSIGT, frequently sampled iv glucose tolerance test; H, Kruskal-Wallis method; IIS, impaired insulin sensitivity; NIS, normal insulin sensitivity; PAI-1, plasminogen activator inhibitor-1; SI, insulin sensitivity index; TF, tissue factor; WHR, waist to hip ratio.

factor VIIa, but not F1 ⴙ 2, levels than controls. Significant inverse correlations were observed between the insulin sensitivity index and TGF-␤1, CRP, PAI-1, factor VIIa, and F1 ⴙ 2 levels. Moreover, significant direct correlations were noted between TGF-␤1 and CRP, PAI-1, factor VIIa, and F1 ⴙ 2 concentrations. Finally, multiple regressions revealed that TGF-␤1 and the insulin sensitivity index were independently related to F1 ⴙ 2. Our results are the first to document an in vivo relationship between insulin sensitivity and coagulative activation in obesity. The elevated TGF-␤1 levels detected in the obese population may provide a biochemical link between insulin resistance and an increased risk for cardiovascular disease. (J Clin Endocrinol Metab 88: 5321–5326, 2003)

In the present study, we sought to examine the relationship between obesity, insulin resistance, and inflammation and coagulation/fibrinolysis indices. Here we report that obesity is associated with higher TGF-␤, PAI-1, prothrombin fragment 1 and 2 (F1 ⫹ 2), and activated factor VII (VIIa) plasma levels and that insulin resistance exacerbates these alterations. Subjects and Methods Subjects Thirty-two nondiabetic obese women (age, 25–56 yr) were studied on outpatient basis between September 1998 and December 1999 as a follow-up investigation of cardiovascular risk factors. Subjects had to be in good general health and physical condition and had to have a body mass index (BMI) greater than 28 kg/m2 and a normal medical history. Exclusion criteria were clinical cardiovascular disease, diabetes mellitus, smoking, hypercholesterolemia, and arterial hypertension. They received a nondiabetic glucose tolerance test by the National Diabetes Data Group criteria (13). The obese population was divided into subjects with normal insulin sensitivity (NIS) or impaired insulin sensitivity (IIS), according to the insulin-modified frequently sampled iv glucose tolerance test (FSIGT), as described below. Estimated and calculated parameters from this test are summarized in Table 1. Ten healthy women (BMI ⬍ 25 kg/m2), aged 27– 49 yr, were also recruited as a control group. The study was approved by the Medical Ethics Committee of the “G. D’Annunzio” University Medical School and was conducted according to the principles of the Helsinki Declaration. All women gave written informed consent.

5321

5322

J Clin Endocrinol Metab, November 2003, 88(11):5321–5326

Romano et al. • Inflammation, Insulin Sensitivity, and Coagulation

TABLE 1. Parameters from insulin-modified FSIGT Variable

Control subjects (n ⫽ 10)

NIS (n ⫽ 14)

IIS (n ⫽ 18)

Pa

SI [104 min⫺1/(␮U/ml)] SG (min⫺1) ⌬AIRG (␮U/ml) Disp index (102 min⫺1) Insul clear (ml/min/kg)

7.02 ⫾ 0.96b 0.049 ⫾ 0.006 27.5 ⫾ 5.7 1.63 ⫾ 0.30 14.6 ⫾ 2.2

5.28 ⫾ 0.38 0.045 ⫾ 0.006 48.7 ⫾ 12.6 2.28 ⫾ 0.51 17.0 ⫾ 3.0

1.87 ⫾ 0.17 0.030 ⫾ 0.004 54.5 ⫾ 6.6 0.99 ⫾ 0.16 8.9 ⫾ 1.0

⬍0.01 NS ⬍0.01 NS ⬍0.05

SG, Glucose effectiveness; NS, not significant; ⌬AIRG, incremental acute insulin response; Disp index, disposition index; Insul clear, insulin clearance. a P: IIS vs. NIS obese subjects. b Values are mean ⫾ SD.

Measurements Anthropometric dimensions [BMI and waist to hip ratio (WHR)] were obtained according to standardized procedures (14). Height, weight, and waist and hip circumferences were measured while the subjects wore indoor clothes without shoes. Fat mass (kg) was determined using bioelectrical impedance analysis (B.I.A.101-F-Akern System SRL, Florence, Italy), as previously described (15).

Insulin-modified FSIGT All studies were performed in recumbent position beginning at 0800 h, after a 10- to 12-h overnight fast. A Teflon catheter was inserted into each forearm for blood sampling and for glucose and insulin administration, respectively. Basal blood samples were collected at time ⫺10 and ⫺1 min, after which glucose (300 mg/kg body weight) was infused in a vein within 30 sec, starting at time 0. At time 20 min, rapid insulin (0.03 IU/kg, Humulin R; Eli Lilly, Indianapolis, IN) was infused for 5 min. The sampling schedule was 2, 4, 8, 19, 27, 30, 40, 50, 70, 100, and 180 min according to Steil et al. (16), with slight modifications. Plasma and serum were stored in aliquots at ⫺20 C until used for the various analyses. Blood glucose was measured by the glucose-oxidase method, and plasma insulin was measured by RIA (Coat-A-Count Insulin kit; Diagnostic Products Corp., Los Angeles, CA). Total cholesterol, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol concentrations were determined as previously described (17). TNF-␣ and TGF-␤1 were measured by ELISA (R&D Systems Europe, Abingdon, UK). CRP was measured using a highly sensitive nephelometric assay (BN-II Nephelometer; Dade Behring, Deerfield, IL). Factor VIIa was determined by STAclot VIIa-rTF (Diagnostica Stago, Asnieres-Sur-Seine, France). F1 ⫹ 2 was measured by ELISA (Enzygnost F1 ⫹ 2; Behringwerke AG, Marburg, Germany). PAI-1 antigen was determined by ELISA (Imubind plasma PAI-1 ELISA kit; American Diagnostica, Greenwich, CT). Interassay and intraassay variations of all measurements were less than 10%.

Data analysis and statistics FSIGT glucose and insulin concentrations were analyzed using the minimal model method (18), which provides an index of insulin sensitivity (SI), i.e. the effect of insulin on glucose uptake (19) by taking into account glucose disposal in the tissues and net hepatic balance. The other parameter provided by this method is the glucose effectiveness, i.e. the effect of glucose per se, independently of any change in insulin levels (20), to accelerate glucose disposal. Insulin secretion was evaluated as the incremental acute insulin response, which was calculated by averaging insulin concentration above basal from 3–10 min after glucose injection. Plasma insulin clearance was calculated as the ratio of insulin dose to dynamic area under the insulin concentration curve from 20 –180 min (21). SI multiplied by the incremental acute insulin response gives the disposition index, a measure of the combined effect of insulin secretion and sensitivity on glucose disposal (22). A cut point SI value of 3.5 10⫺4 min⫺1/(␮U/ml) was used to separate the obese women into the two groups. This value was chosen according to the following criteria. Given the individual SI values of the control group and their distribution, we accepted that less than 5% of these values were allowed to lie outside the normality range (P ⬍ 0.05). Then we used the lower 2.5% quantile of the distribution as the cut point between normal and impaired (lower)

SI. This quantile resulted as 3.27 for our control people (see Results for complete figures of SI), and we made the conservative choice of 3.5. More details on this procedure have been published elsewhere (23). The data were analyzed by nonparametric methods to avoid assumptions about the distribution of the measured variables. Comparisons between groups were made with the Kruskal-Wallis method (H) and Mann-Whitney U test. The association between different measurements was assessed by the Spearman rank correlation test. Multiple linear regression analysis was conducted to assess independent predictors of F1 ⫹ 2 plasma levels. All values are reported as mean ⫾ sd. Statistical significance was achieved when P ⬍ 0.05.

Results

Clinical and laboratory parameters relative to the controls and the two groups of obese women are given in Table 2. Notably, although fasting plasma glucose levels were different between the two obese groups, none of these patients showed indices of impaired fasting glucose or impaired glucose tolerance. The relationship between glucose tolerance, inflammatory reaction, and coagulation activation was assessed by the simultaneous determination of a number of established peripheral markers of the aforementioned conditions. As shown in Fig. 1, significantly higher levels of CRP (H ⫽ 28.8, P ⬍ 0.0001), TGF-␤1(H ⫽ 26.5, P ⬍ 0.0001), factor VIIa (H ⫽ 19.9, P ⬍ 0.0001), and F1 ⫹ 2 (H ⫽ 28.2, P ⬍ 0.0001) were observed comparing IIS patients to NIS patients or to control subjects. In particular, IIS patients had increased levels of CRP, TGF-␤1, PAI-1, and factor VIIa compared with either control subjects or NIS patients, whereas NIS individuals had higher CRP, TGF-␤1, PAI-1, and factor VIIa than healthy controls. A similar behavior was observed for PAI-1 plasma levels (IIS patients: 39.6 ⫾ 12.8 ng/ml; NIS subjects: 24.9 ⫾ 7.3 ng/ml; controls: 12.8 ⫾ 2.4 ng/ml; H ⫽ 24.6, P ⬍ 0.0001). No difference in TNF-␣ levels was observed between obese (both IIS and NIS) women and controls (results not shown). Thus, obese women presented peripheral signs of inflammation and coagulative activation that were accentuated in patients with insulin resistance. In fact, a Spearman rank correlation analysis of the variables analyzed in the study demonstrated that SI levels inversely correlated with all inflammation and coagulative activation indices (Table 3). F1 ⫹ 2 plasma levels directly correlated with factor VIIa (P ⬍ 0.0001) and PAI-1 (P ⬍ 0.0001), as well as with CRP (P ⬍ 0.0001) and TGF-␤1 (P ⬍ 0.0001) (Table 3). TGF-␤1 levels directly correlated not only with another marker of inflammation, namely CRP (P ⬍ 0.0001), but also with indices of the coagulation/fibrinolytic system, such as PAI-1 (P ⬍ 0.0001), factor VIIa (P ⬍ 0.0001), and F1 ⫹ 2. However, no significant


J Clin Endocrinol Metab, November 2003, 88(11):5321–5326 5323

TABLE 2. Characteristics of study subjects Variable

Age (yr) (NS) BMI (kg/m2) (H ⫽ 22.3, P ⬍ 0.0001)d Waist circumference (cm) (H ⫽ 14.8, P ⬍ 0.001) WHR (H ⫽ 4.13, P ⫽ 0.127) Fat masse (kg) (H ⫽ 10.5, P ⬍ 0.01) Free fat masse (kg) (H ⫽ 10.5, P ⬍ 0.01) SBP (mm Hg) (NS) DBP (mm Hg) (NS) Triglyceride (mg/dl) (H ⫽ 13.5, P ⬍ 0.01) (mmol/liter) HDL cholesterol (mg/dl) (NS) (mmol/liter) LDL cholesterol (mg/dl) (NS) (mmol/liter) Fibrinogen (mg/dl) (NS) (␮mol/liter) Fasting plasma glucose (mg/dl) (H ⫽ 4.9, P ⫽ 0.087) (mmol/liter) 2-h Plasma glucose (mg/dl) (NS) (mmol/liter) Fasting insulin (␮U/ml) (H ⫽ 8.7, P ⬍ 0.02) (pmol/liter)

Controls (n ⫽ 10)

P

NIS (n ⫽ 14)

P

37 ⫾ 8b

NSc

37 ⫾ 7

23.1 ⫾ 1.8

0.0001

85.4 ⫾ 7.9

IIS (n ⫽ 18)

Pa

NS

42 ⫾ 9

NS

37.4 ⫾ 6.1

NS

38.0 ⫾ 7.3

0.0001

0.0004

110.4 ⫾ 15.3

NS

107.7 ⫾ 18.8

0.002

0.84 ⫾ 0.05

0.034

0.91 ⫾ 0.07

NS

0.88 ⫾ 0.11

NS

19.2 ⫾ 5.3

0.0002

38.1 ⫾ 8.6

NS

40.5 ⫾ 9

0.0002

47.0 ⫾ 4.4

0.0007

57.9 ⫾ 8.7

NS

60.5 ⫾ 7.1

0.0003

113.6 ⫾ 5.4

NS

110.3 ⫾ 11.4

NS

115.4 ⫾ 12.3

NS

74.8 ⫾ 5.1

NS

70.7 ⫾ 6.7

NS

77.0 ⫾ 7.7

NS

46.4 ⫾ 15.6 (0.52 ⫾ 0.18) 55.3 ⫾ 10.6 (1.43 ⫾ 0.27) 113.1 ⫾ 14.7 (2.93 ⫾ 0.38) 369.1 ⫾ 73.2 (10.85 ⫾ 2.15) 89.3 ⫾ 13.1 (4.96 ⫾ 0.72) 98.3 ⫾ 20.2 (5.45 ⫾ 1.12) 11.60 ⫾ 7.45 (80.56 ⫾ 51.74)

0.0008

103.6 ⫾ 66.0 (1.17 ⫾ 0.74) 46.3 ⫾ 10.9 (1.20 ⫾ 0.28) 124.7 ⫾ 42.4 (3.23 ⫾ 1.10) 399.0 ⫾ 102.5 (11.73 ⫾ 3.01) 91.1 ⫾ 9.5 (5.06 ⫾ 0.53) 108.0 ⫾ 26.4 (5.99 ⫾ 1.46) 30.78 ⫾ 25.54 (213.77 ⫾ 177.3)

NS

96.2 ⫾ 51.8 (1.09 ⫾ 0.58) 55.5 ⫾ 15.2 (1.44 ⫾ 0.39) 106.8 ⫾ 26.7 (2.77 ⫾ 0.69) 398.1 ⫾ 78.3 (11.70 ⫾ 2.30) 98.3 ⫾ 9.8 (5.45 ⫾ 0.54) 119.9 ⫾ 26.2 (6.65 ⫾ 1.45) 28.05 ⫾ 23.14 (194.8 ⫾ 160.7)

0.0021

NS NS NS NS NS NS

NS NS NS 0.048 NS NS

NS NS NS NS NS 0.004

SBP, Systolic blood pressure; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein. Conversion factors: metric units–SI units: trygliceride 0.01129; HDL cholesterol 0.0259; LDL cholesterol 0.0259; fibrinogen 0.0294; glucose 0.0555; insulin 6.945. a Controls vs. IIS. b Values are mean ⫾ SD. c NS, Not significant (P ⬎ 0.05). d H, Kruskal-Wallis test. e Fat mass and free fat mass (kg) evaluated by bioelectrical impedance analyzer.

correlation could be observed between TNF-␣ and the above mentioned variables. To further investigate the relationship between signs of coagulative activation, inflammation, and SI, we performed a multiple regression analysis. As shown in Table 4, when all variables were included, the TF-dependent coagulation marker, F1 ⫹ 2, was independently correlated only with TGF-␤1 (directly) (Fig. 2) and SI (inversely). Together these results indicate that insulin resistance, TGF-␤1 increment, and coagulative activation are closely related events in obesity. Discussion

A low chronic inflammatory state, characterized by increased levels of proinflammatory molecules and acutephase proteins, can be present in obesity and insulin resistance (1, 2, 4, 5). These conditions are also associated with an elevated procoagulant activity and a reduced fibrinolytic potential that may enhance the risk for cardiovascular events (3, 6, 7). In the present report, we examined the relationship between circulating levels of inflammation mediators, namely TNF-␣, TGF-␤, and CRP, and components of the coagulation and fibrinolytic cascades in obesity and insulin resistance.

Insulin sensitivity was assessed using the minimal model approach (18, 19). This method has been extensively used to investigate rare pathologies affecting a few subjects (reviewed in Ref. 24) as well as in large epidemiological studies (1). It provides an SI, which takes into account glucose disposal in the tissues and net hepatic balance. SI is comparable with the corresponding parameter from the gold standard glucose clamp (25), and it can distinguish between subjects with NIS or IIS. We found that obesity, in general, was associated with increased circulating levels of TGF-␤1, CRP, factor VIIa, F1 ⫹ 2 (Fig. 1), and PAI-1, but not with TNF-␣. On the other hand, within obese subjects, those with insulin resistance displayed the highest levels of these parameters plus an increment in F1 ⫹ 2 levels. Moreover, simple correlation analyses showed significant inverse correlations between the SI and TGF-␤1, CRP, PAI-1, factor VIIa, and F1 ⫹ 2 and between TGF-␤1 and CRP, PAI-1, factor VIIa, and F1 ⫹ 2, whereas multiple regression analyses showed independent correlations between TGF⫺␤1, SI, and F1 ⫹ 2 (Tables 3 and 4). Together these results are consistent with previous findings of increased CRP levels and reduced fibrinolytic activity in the insulin resistance syndrome (26, 27). Moreover, they show, for the first time in obese individuals, a direct corre-

5324



FIG. 1. Plasma levels of CRP (A), TGF-␤1 (B), factor VIIa (C), and F1 ⫹ 2 (D) in nonobese women and in obese women with NIS or IIS. Dots represent individual measurements; horizontal bars represent the mean values for each group of subjects. TABLE 3. Spearman’s correlation coefficients among the various parameters analyzed in 32 obese women

F1 ⫹ 2

Factor VIIa

PAI-1

TGF-␤1

CRP

TNF-␣

SI

0.66 (P ⬍ 0.0001)

0.67 (P ⬍ 0.0001) 0.70 (P ⬍ 0.0001)

0.76 (P ⬍ 0.0001) 0.73 (P ⬍ 0.0001) 0.67 (P ⬍ 0.0001)

0.67 (P ⬍ 0.0001) 0.60 (P ⬍ 0.001) 0.81 (P ⬍ 0.0001) 0.69 (P ⬍ 0.0001)

0.11 (NS) 0.09 (NS) 0.35 (P ⬍ 0.05) 0.09 (NS) 0.29 (NS)

⫺0.84 (P ⬍ 0.0001) ⫺0.61 (P ⬍ 0.001) ⫺0.72 (P ⬍ 0.0001) ⫺0.77 (P ⬍ 0.0001) ⫺0.75 (P ⬍ 0.0001) ⫺0.17 (NS)

VIIa PAI-1 TGF-␤1 CRP TNF-␣ NS, Not significant.

lation between TGF-␤1 and markers of the coagulation/ fibrinolytic cascade and a close relationship between insulin resistance, increased circulating TGF-␤1 levels, and coagulative activation, which is independent of other variables (Table 4). The two groups of obese volunteers examined in the present study were in fact comparable for age, menopausal status, hormone replacement therapy, overall obesity (BMI), regional fat distribution (WHR), systolic and diastolic blood pressure, smoking status, plasma lipid levels, glucose tolerance test, fasting insulin, and fibrinogen levels (Table 2). Thus, from the present results, a key involvement of TGF-␤1 in the procoagulant, hypo-fibrinolytic state associated with insulin resistance in obesity may be hypothesized. Increased circulating TGF-␤ levels have been previously documented in type 2 diabetes mellitus (28) and hypertension (29, 30) but not in insulin resistance per se. Within this context, an elevation in TGF-␤ levels may carry relevant

FIG. 2. Correlations between TGF-␤1 and F1 ⫹ 2 plasma levels in obese women with NIS (E) or IIS (F). The broken line represents the regression plot for NIS individuals, whereas the solid line corresponds to the regression plot for IIS subjects.


J Clin Endocrinol Metab, November 2003, 88(11):5321–5326 5325

TABLE 4. Multiple regression analysis between F1 ⫹ 2 and indices of inflammation, coagulative activation, and insulin sensitivity in 32 obese women

between insulin resistance and an increased risk for cardiovascular disease.

Independent variable

Regression coefficient (␤)

SE

P

Acknowledgments

VIIa PAI-1 TGF-␤1 CRP SI

0.08 ⫺0.12 0.37 0.28 ⫺0.34

0.15 0.20 0.17 0.18 0.17

0.613 0.559 0.0387 0.127 0.0498

We thank Dr. E. Porreca for helpful discussion and Dr. M. Nutini for technical assistance.

pathophysiological implications. In fact, TGF-␤ potently stimulates monocyte chemotaxis and endothelial transmigration and promotes smooth muscle cell proliferation and migration by up-regulating platelet-derived growth factor gene expression (31). These represent early events in atherosclerosis development. TGF-␤ also induces PAI-1 both in vivo and in cultured adipocytes (11), as well as in other tissues (12). Moreover, it up-regulates TF gene expression in adipocytes (10). The increase in PAI-1 and F1 ⫹ 2 plasma levels, detected in our obese women, may be a reflection of these TGF-␤ bioactions. In fact, F1 ⫹ 2, a marker of factor Xa activation, was associated with a higher plasma concentration of factor VIIa, indicating that thrombin generation in obese subjects derives from TF-dependent mechanisms. However, the origin of TGF-␤ hyperproduction in our obese women remains to be established. A relationship between accumulation of adipose tissue and TGF-␤ expression has been previously documented. In particular, TGF-␤ mRNA levels were chronically elevated in the adipose tissue from genetically obese mice (8). However, this does not explain why TGF-␤ levels were significantly higher in IIS than in NIS individuals because they had comparable indices of fat accumulation (Table 2). It is possible that these subjects had a different distribution of visceral fat. But it can be also hypothesized that other factors related to insulin resistance may contribute to TGF-␤ up-regulation. Along these lines, TNF-␣ has been proposed as a key determinant of the obesity-linked up-regulation of TGF-␤ and PAI-1 in the adipose tissue (10). In our study we were unable to find changes in the TNF-␣ plasma levels of obese women. However, this does not exclude the possibility that TNF-␣ may be overexpressed within the adipose district and that circulating levels of this cytokine do not precisely reflect a regional up-regulation. Moreover, no correlation was found between measurements of obesity and TGF-␤1 levels or insulin resistance. One possible explanation of these results is that the two groups of women selected for this study had comparable BMI and WHR. It may be possible that a relationship between obesity and SI or TGF-␤1 could be observed if the study population had a broader range of obesity. On the other hand, the apparent lack of relationship between obesity measures and TGF-␤1 or SI does not necessarily exclude a direct contribution of obesity to TGF-␤1 or F1 ⫹ 2 concentrations. In fact, our study population was limited to 32 subjects, and therefore, we cannot exclude a type II error. In conclusion, here we provide the first evidence of a close relationship between insulin sensitivity, increased TGF-␤1 levels, and coagulative activation in obesity. The present results indicate that TGF-␤1 may represent a biochemical link

Received March 25, 2003. Accepted August 12, 2003. Address all correspondence and requests for reprints to: Giovanni Davı`, M.D., Department of Medicine and Aging, School of Medicine, University of Chieti “G. D’Annunzio,” Via Colle dell’Ara, 66013 Chieti, Italy. E-mail: [email protected]. This work was supported by grants from the Italian Ministry of Research and Education (Cofin 2002, and Center of Excellence on Aging). No conflict of interest exists in connection with this article.

References 1. Festa A, D’Agostino Jr R, Howard G, Mykkanen L, Tracy RP, Haffner SM 2000 Chronic subclinical inflammation as part of the insulin resistance syndrome: the Insulin Resistance Atherosclerosis Study (IRAS). Circulation 102: 42– 47 2. Kuller LH, Tracy RP 2000 The role of inflammation in cardiovascular disease. Arterioscler Thromb Vasc Biol 20:901 3. Yudkin JS 1999 Abnormalities of coagulation and fibrinolysis in insulin resistance: evidence for a common antecedent? Diabetes Care 22:25–30 4. Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW 1999 C-reactive protein in healthy subjects: associations with obesity, insulin resistance, and endothelial dysfunction. A potential role for cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol 19:972–978 5. Hotamisligil GS, Arner P, Atkinson RL, Spiegelman BM 1997 Differential regulation of the p80 tumor necrosis factor receptor in human obesity and insulin resistance. Diabetes 46:451– 455 6. Juhan-Vague I, Thompson SG, Jespersen J 1993 Involvement of the hemostatic system in the insulin resistance syndrome: a study of 1500 patients with angina pectoris: the ECAT Angina Pectoris Study Group. Arterioscler Thromb 13:1865–1873 7. Festa A, D’Agostino Jr R, Mykkanen L, Tracy RP, Zaccaro DJ, Hales CM, Haffner SM 1999 Relative contribution of insulin and its precursors to fibrinogen and PAI-1 in a large population with different states of glucose tolerance. The Insulin Resistance Atherosclerosis Study (IRAS). Arterioscler Thromb Vasc Biol 19:562–568 8. Samad F, Loskutoff DJ 1999 Hemostatic gene expression and vascular disease in obesity: insights from studies of genetically obese mice. Thromb Haemost 82:742–747 9. Morange PE, Alessi MC, Verdier M, Casanova D, Magalon G, Juhan-Vague I 1999 PAI-1 produced ex vivo by human adipose tissue is relevant to PAI-1 blood level. Arterioscler Thromb Vasc Biol 19:1361–1365 10. Samad F, Pandey M, Loskutoff DJ 1998 Tissue factor gene expression in the adipose tissues of obese mice. Proc Natl Acad Sci USA 95:7591–7596 11. Samad F, Yamamoto K, Pandey M, Loskutoff DJ 1997 Elevated expression of transforming growth factor-␤ in adipose tissue from obese mice. Mol Med 3:37– 48 12. Chen YQ, Sloan-Lancaster J, Berg DT, Richardson MA, Grinnell B, TsengCrank J 2001 Differential mechanisms of plasminogen activator inhibitor-1 gene activation by transforming growth factor-␤ and tumor necrosis factor-␣ in endothelial cells. Thromb Haemost 86:1563–1572 13. National Diabetes Data Group 1979 Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28:1039 –1057 14. World Health Organization 1998 Obesity: Preventing and managing the global epidemic. Report on a WHO consultation on obesity. WHO/NUT/ NCD/98.1. Geneva: World Health Organization, 1997 15. Segal KR, Van Loan M, Fitzgerald PI, Hodgdon JA, Van Itallie TB 1988 Lean body mass estimation by bioelectrical impedance analysis: a four-site crossvalidation study. Am J Clin Nutr 47:7–14 16. Steil GM, Volund A, Kahn SE, Bergman RN 1993 Reduced sample number for calculation of insulin sensitivity and glucose effectiveness from the minimal model. Diabetes 42:250 –256 17. Davı` G, Romano M, Mezzetti A, Procopio A, Iacobelli S, Antidormi T, Bucciarelli T, Alessandrini P, Cuccurullo F, Bittolo Bon G 1998 Increased levels of soluble P-selectin in hypercholesterolemic patients. Circulation 97: 953–957 18. Pacini G, Bergman RN 1986 MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsivity from the frequently sampled intravenous glucose tolerance test. Comput Methods Programs Biomed 23: 113–122

5326


19. Bergman RN 1989 Toward physiological understanding of glucose tolerance. Minimal model approach. Diabetes 38:1512–1527 20. Best JD, Kahn SE, Ader M, Watanabe RM, Ni TC, Bergman RN 1996 Role of glucose effectiveness in the determination of glucose tolerance. Diabetes Care 19:1018 –1030 21. Gibaldi M, Perrier D 1982 Pharmacokinetics. 2nd ed. New York: Dekker 22. Kahn SE, Prigeon RL, McCulloch DK, Boyko EJ, Bergman RN, Schwartz MW, Neifing JL, Ward WK, Beard JC, Palmer JP, Porte Jr D 1993 Quantification of the relationship between insulin sensitivity and ␤-cell function in human subjects. Evidence for a hyperbolic function. Diabetes 42:1663–1672 23. Kautzky-Willer A, Krssak M, Winzer C, Pacini G, Tura A, Farhan S, Wagner O, Brabant G, Horn R, Stingl H, Schneider B, Waldhausl W, Roden M 2003 Increased intramyocellular lipid concentration identifies impaired glucose metabolism in women with previous gestational diabetes. Diabetes 52:244 –251 24. Bergman RN 1997 The minimal model approach and determinants of glucose tolerance. Baton Rouge, LA: LSU Press 25. Bergman RN, Prager R, Volund A, Olefsky JM 1987 Equivalence of the insulin sensitivity index in man derived by the minimal model method and the euglycemic glucose clamp. J Clin Invest 79:790 – 800 26. Hak AE, Stehouwer CD, Bots ML, Polderman KH, Schalkwijk CG, West-


27.

28.

29.

30.

31.

endorp IC, Hofman A, Witteman JC 1999 Associations of C-reactive protein with measures of obesity, insulin resistance, and subclinical atherosclerosis in healthy, middle-aged women. Arterioscler Thromb Vasc Biol 19:1986 –1991 Abbasi F, McLaughlin T, Lamendola C, Lipinska I, Tofler G, Reaven GM 1999 Comparison of plasminogen activator inhibitor-1 concentration in insulin-resistant versus insulin-sensitive healthy women. Arterioscler Thromb Vasc Biol 19:2818 –2821 Pfeiffer A, Middelberg-Bisping K, Drewes C, Schatz H 1996 Elevated plasma levels of transforming growth factor-␤1 in NIDDM. Diabetes Care 19:1113– 1117 Li B, Khanna A, Sharma V, Singh T, Suthanthiran M, August P 1999 TGF-␤1 DNA polymorphisms, protein levels, and blood pressure. Hypertension 33: 271–275 Porreca E, Di Febbo C, Mincione G, Reale M, Baccante G, Guglielmi MD, Cuccurullo F, Colletta G 1997 Increased transforming growth factor-␤ production and gene expression by peripheral blood monocytes of hypertensive patients. Hypertension 30:134 –139 Blobe GC, Schiemann WP, Lodish HF 2000 Role of transforming growth factor-␤ in human disease. N Engl J Med 342:1350 –1358