Szalkowski D, White-Carrington S, Berger J, Zhang B. 1995 Antidiabetic thiazolidines ... Saghizadeh M, Ong JM, Garvey WT, Henry RR, Kern PA. 1996 The ...
0021-972X/98/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1998 by The Endocrine Society
Vol. 83, No. 3 Printed in U.S.A.
Serum Levels of Tumor Necrosis Factor-a Are Increased in Obese Patients with Noninsulin-Dependent Diabetes Mellitus* AKIRA KATSUKI, YASUHIRO SUMIDA, SHUICHI MURASHIMA, KAZUYA MURATA, YOSHIHIRO TAKARADA, KUNINOBU ITO, MASASHI FUJII, KEN TSUCHIHASHI, HIROYUKI GOTO, KANAME NAKATANI, AND YUTAKA YANO Third Department of Internal Medicine (A.K., Y.S., K.I., M.F., K.T., H.G., K.N., Y.Y.) and Department of Radiology (S.M.), Mie University School of Medicine; Department of Internal Medicine (K.M., Y.T.), Yamada Red Cross Hospital, Mie, Japan icantly exceeded those observed in normal subjects (P , 0.01) or in the nonobese patients with NIDDM (P , 0.01). Serum levels of TNF-a in obese NIDDM patients showed a significant positive correlation with the area of visceral fat before (r 5 0.662, P , 0.03) and after (r 5 0.508, P , 0.05) the treatment; similar correlation was observed in all patients with NIDDM before (r 5 0.537, P , 0.02) and after (r 5 0.430, P , 0.05) the treatment. Serum levels of TNF-a in obese NIDDM patients showed a significant negative correlation with GIR after the treatment (r 5 20.595, P , 0.05). Serum levels of TNF-a were significantly reduced in the obese patients with NIDDM after the treatment (P , 0.01), while those in the nonobese NIDDM patients were unchanged. These results suggest that serum TNF-a levels may play an important role in mechanism of insulin resistance associated with obesity. (J Clin Endocrinol Metab 83: 859 – 862, 1998)
ABSTRACT To clarify the significance of the serum levels of tumor necrosis factor-a (TNF-a) in the mechanism of insulin resistance, we studied 12 obese patients with noninsulin-dependent diabetes mellitus (NIDDM). We evaluated the relationship of TNF-a levels with the visceral, subcutaneous, and total fat areas measured by computed tomography (CT), and with insulin resistance evaluated by the glucose infusion rate (GIR) observed during an euglycemic hyperinsulinemic clamp study. Controls consisted of 12 normal subjects and 12 nonobese patients with NIDDM. TNF-a levels were measured using a high sensitivity enzyme-linked immunosorbent assay. Following admission, all patients with NIDDM participated in a 4-week program of diet and exercise. After this treatment, we evaluated the relationship of the serum levels of TNF-a with the area of body fat, the GIR, and the resultant change in the TNF-a level. Serum levels of TNF-a in the obese patients with NIDDM signif-
O
phy (CT) and with the glucose infusion rate (GIR) evaluated by the euglycemic hyperinsulinemic clamp technique.
BESITY and noninsulin-dependent diabetes mellitus (NIDDM) are important health problems in industrialized countries. Insulin resistance in peripheral target tissues (e.g. fatty and muscular tissues) is usually an early event that oftentimes precedes the onset of NIDDM in subjects with obesity (1). Tumor necrosis factor-a (TNF-a) was reported to inhibit insulin action and to play a role in insulin resistance in obesity (2–5). Previous investigators speculated that the overexpression of TNF-a in adipose tissue would inhibit the transport of glucose in an autocrine or paracrine fashion. Until recently, blood levels of TNF-a could not be accurately measured because sensitive assays for TNF-a were not available (6). In the present study, we measured the serum levels of TNF-a in obese and nonobese patients with NIDDM using a high-sensitive system, and we determined their relationship with the area of body fat measured by computed tomogra-
Subjects and Methods Subjects We evaluated 12 patients with obesity (body mass index: BMI . 26.4) and NIDDM (obese group). Data obtained in 12 normal subjects (normal group) and 12 nonobese NIDDM patients (nonobese group) served as controls. The age, gender, fasting plasma glucose level, HbAIc level, and duration of diabetes mellitus in the control group were matched with those of the obese group (Table 1). Patients with NIDDM that participated in this study were physical workers with night duty in factories. They did not have any complaint and were found to have NIDDM during a group medical examination. We also evaluated the serum TNF-a levels in 8 obese male subjects without NIDDM (age; 49.10 6 2.29; BMI, 27.20 6 0.20). The BMI was estimated by dividing the body weight (in kilograms) by the square of the height (in meters). NIDDM was diagnosed according to the criteria of the World Health Organization (WHO), using the 75-g oral glucose tolerance test (OGTT) (Trelan G 75, Shimizu, Shizuoka, Japan). They did not receive any medication that could affect their insulin level or their insulin sensitivity. Before admission to the study, the patients were not engaged in any dietary regimen or exercise program. Six patients with NIDDM had simple diabetic retinopathy, and 5 had microalbuminuria. Neuropathy or macrovascular complications were not detected. No abnormality in the 75-g OGTT was found in the group of normal subjects. None of the normal subjects reported any change in body weight before the investigation. Each subject underwent physical and laboratory investigations to
Received January 14, 1997. Revision received March 28, 1997. Rerevision received June 3, 1997. Accepted November 12, 1997. Address correspondence and requests for reprints to: Akira Katsuki, M.D., Third Department of Internal Medicine, Mie University School of Medicine, 2-174 Edobashi, Tsu, Mie 514, Japan. * This work was supported in part by a grant from the Mie Prefecture Medical Association.
859
SERUM TNF-a LEVELS AND OBESITY
860
TABLE 1. Clinical characteristics of the study subjects Normal
n Age (yr) Sex (M/F) BMI (kg/m2) Duration of DM (yr) HbAIc (%) Fasting glucose (mmol/L) Fasting insulin (pmol/L) GIR (mmol/ kg/min) Visceral fat area (cm2) Subcutaneous fat area (cm2) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg)
12 54.42 6 2.83 11/1 22.88 6 0.45
4.96 6 0.14 28.62 6 4.68
Nonobese NIDDM
Obese NIDDM
12 55.83 6 1.53 11/1 23.71 6 0.44 8.50 6 0.98
12 50.00 6 2.43 11/1 27.47 6 0.33a,b 8.78 6 0.57
9.05 6 0.48 8.97 6 0.37c
8.62 6 0.56 8.92 6 0.52b
35.52 6 5.28
46.14 6 6.00b
53.68 6 3.61
45.02 6 3.72d
81.11 6 6.59
120.75 6 10.13a
99.84 6 14.12
138.00 6 11.73d
121.33 6 4.33
127.58 6 2.84
124.42 6 3.48
72.50 6 3.18
79.17 6 2.55
80.58 6 3.40
JCE & M • 1998 Vol 83 • No 3
assay kit (DAINABOT Corp., Tokyo, Japan). Blood pressure was determined in supine position after a 5-min rest. After admission, the patients with NIDDM participated in a program of diet and exercise for about 4 weeks. The dietary treatment was as follows: 1400 –1720 kcal/day with a diet consisting of 20 energy percent (en %) protein, 25 en % fat, and 55 en % carbohydrates. As exercise therapy, the patients walked about 15,000 steps daily, as counted by a pedometer. Serum TNF-a level, body fat area, and insulin sensitivity were measured in each subject before and after the initiation of treatment. The body fat area was evaluated by a previously described method (7). The total cross-sectional area, the intra-abdominal visceral fat, and the subcutaneous fat areas were measured by abdominal computed tomography (CT) taken at the umbilical level. Any intraperitoneal region having the same density as the subcutaneous fat layer was defined as a visceral fat area. Insulin sensitivity was evaluated by the euglycemic hyperinsulinemic clamp technique using the artificial pancreas (STG-22, NIKKISO, Tokyo, Japan) (8). The clamp study was performed for 120 min, and the desired level of serum insulin was fixed to 1200 pmol/L. The mean values of insulin reached a stable level between 90 min and 120 min after starting the clamp study (obese group before treatment: 1186.45 6 43.42 pmol/L, after treatment: 1195.45 6 57.48 pmol/L, nonobese group before treatment: 1188.30 6 54.76 pmol/L, after treatment: 1111.02 6 32.65 pmol/L). The blood glucose was clamped to desired level (5.24 mmol/L), and the mean amount of glucose administered in the last 30 min was regarded as the glucose infusion rate (GIR).
Statistical methods
Data are mean 6 SE. a P , 0.01; Nonobese NIDDM vs. obese NIDDM. b P , 0.05; Normal vs. obese NIDDM. c P , 0.01; Normal vs. nonobese NIDDM. d P , 0.05; Nonobese NIDDM vs. obese NIDDM.
exclude the presence of systemic inflammatory disease. Informed consent was obtained from each subject before the study admission.
Study protocol and methods Blood samples were drawn from each subject before breakfast in the early morning, after an overnight bed-rest. TNF-a in serum samples was measured using a commercially available sandwich immunoassay kit (Quantikine HS Human TNF-a immunoassay kit, R&D systems, Minneapolis, MN) and following the manufacturer’s instructions. Briefly, 200 mL of standard or serum samples were added to microtiter plate wells coated with a monoclonal antibody specific for TNF-a, followed by incubation at 4 C for 16 h. The wells were then washed 4 times with a buffered surfactant solution, and thereafter, 200 mL of anti-TNF-a polyclonal antibody conjugated to alkaline phosphatase were added to each well and incubated for 3 h at room temperature. After appropriate washing, 50 mL of substrate solution (NADPH) were added to each well and incubated again for 60 min at room temperature. After this, 50 mL of amplifier enzyme solution were added to the well, followed by incubation for 30 min at room temperature. The reaction was then stopped by the addition of 2N sulfuric acid to the wells, and absorbance was measured at 490 nm with corrections set at 650 nm using a microplate reader. The values of serum TNF-a levels were extrapolated from a curve drawn using standard TNF-a. The minimum detectable concentration by this assay was 0.01 pmol/L, and the intra- and interassay coefficients of variation of the assay were 5.6% and 7.5%, respectively. No significant cross-reactivity or interference with other factors related to TNF-a or other cytokines was observed. The precision of the TNF-a assay using the above described immunoassay is better than that of other methods previously reported. Previous methods showed TNF-a minimum detection levels of 0.24 pmol/L. The plasma glucose level was measured by an automated enzymatic method. The HbAIc (normal value: 4.3– 5.8%) was measured by high performance liquid chromatography (HPLC). Serum insulin was measured using an immunoradiometric
Data are expressed as the mean 6 se. Comparison between groups was done using the Mann-Whitney U test. The statistical difference between TNF-a levels before and after treatment was analyzed by the Wilcoxon’s rank sum test. The strength of correlation between variables was calculated using Spearman’s rank correlation. A level of P less than 0.05 was accepted as statistically significant.
Results
The serum levels of TNF-a were not statistically different between the normal and nonobese groups. Subjects of the obese group presented significantly higher serum concentrations of TNF-a than those of the normal (P , 0.01) and nonobese (P , 0.01) groups (Fig. 1). Serum TNF-a levels
FIG. 1. The serum levels of TNF-a in normal subjects, obese, and nonobese patients with NIDDM. Serum levels of TNF-a in the obese patients with NIDDM were significantly increased compared with those in the normal subjects (P , 0.01) and nonobese patients with NIDDM (P , 0.01).
KATSUKI ET AL.
(1.89 6 0.26 pmol/L) in 8 obese subjects without NIDDM were significantly elevated compared with those in the normal (P , 0.01) and nonobese (P , 0.01) groups. However they were not significantly different compared with those in the obese group. The clinical characteristics of the patients with NIDDM before and after treatment are shown in Table 2. There was no significant difference between the obese and nonobese groups in the levels of fasting plasma glucose and HbAIc after treatment. Serum TNF-a levels in the obese group significantly decreased after treatment (P , 0.01), but no significant change was observed in the nonobese group (Fig. 2). In the obese NIDDM patients, serum levels of TNF-a were positively correlated with the visceral fat area before (r 5 0.662, P , 0.03) and after (r 5 0.508, P , 0.05) the treatment. There was also a significant positive correlation between serum levels of TNF-a and visceral fat area before (r 5 0.537, P , 0.02) and after (r 5 0.430, P , 0.05) the treatment in all patients with NIDDM, including obese and nonobese subjects (Fig. 3). Serum levels of TNF-a were not significantly correlated with the subcutaneous fat area or total fat area, either before or after the treatment, in all and in obese patients with NIDDM. After the treatment, a significant negative correlation was observed between the serum TNF-a levels and the GIR in obese patients with NIDDM (r 5 20.595, P , 0.05) (Fig. 4).
861
FIG. 2. Effect of treatment on the serum levels of TNF-a in obese and nonobese patients with NIDDM. The levels of TNF-a decreased significantly (P , 0.01) only in the obese patients with NIDDM.
Discussion
The present study showed that TNF-a is significantly increased in NIDDM patients with obesity. In obese patients with NIDDM, the serum levels of TNF-a were significantly correlated with visceral adiposity, and it was applicable to all (obese and nonobese) patients with NIDDM. However, it is worthy to note that the obese subjects of the present study were not typically obese subjects. The observation of increased intraperitoneal fat accumulation indicated that they have visceral fat type of obesity (9, 10). Our NIDDM patients had increased visceral fat, but without significant increase in their total body fat. Nevertheless, the results of the present
FIG. 3. Correlation between the serum level of TNF-a and the area of visceral fat in all patients with NIDDM before and after treatment. A significant positive correlation was observed before (r 5 0.537, P , 0.02) and after (r 5 0.430, P , 0.05) treatment. E, Obese NIDDM; F, nonobese NIDDM.
TABLE 2. Clinical characteristics of NIDDM patients before and after treatment
Nonobese NIDDM BMI (kg/m2) HbAIc (%) Fasting glucose (mmol/L) Fasting insulin (pmol/L) GIR (mmol/kg/min) Visceral fat area (cm2) Subcutaneous fat area (cm2) Obese NIDDM BMI (kg/m2) HbAIc (%) Fasting glucose (mmol/L) Fasting insulin (pmol/L) GIR (mmol/kg/min) Visceral fat area (cm2) Subcutaneous fat area (cm2)
Before treatment
After treatment
23.71 6 0.44 9.05 6 0.48 8.97 6 0.37 35.52 6 5.28 53.68 6 3.61 81.11 6 6.59 99.84 6 14.12
22.93 6 0.36a 7.84 6 0.40a 5.81 6 0.15a 29.28 6 3.06 56.91 6 3.33a 69.81 6 9.42b 90.07 6 10.67
27.47 6 0.33 8.78 6 0.57 8.92 6 0.52 46.14 6 6.00 45.02 6 3.72 120.75 6 10.13 138.00 6 11.73
26.15 6 0.30a 7.68 6 0.45a 5.75 6 0.29a 33.60 6 3.06b 50.07 6 4.16b 105.47 6 9.59a 122.44 6 10.21a
Data are mean 6 SE. a P , 0.01; before vs. after treatment. b P , 0.05; Before vs. after treatment.
FIG. 4. Correlation between the serum TNF-a level and the glucose infusion rate (GIR) in obese patients with NIDDM. A significant negative correlation was observed only after treatment (r 5 20.595, P , 0.05).
study in obese NIDDM subjects would probably not be observed in typical obese subjects. Namely, although in our present study the serum TNF-a levels were significantly correlated only with the visceral fat area, they might also be significantly correlated with the subcutaneous fat area and
SERUM TNF-a LEVELS AND OBESITY
862
with the total fat area in typically obese patients with NIDDM. The elevation of the serum levels of TNF-a significantly decreased after treatment in obese subjects with NIDDM but not in the nonobese group. A lesser degree of fatty tissue decrease in the nonobese group than in the obese group may be a potential explanation for this finding. The decrease in the accumulation of fatty tissues in the nonobese group after the treatment was probably insufficient to produce a significant change in the serum levels of TNF-a. These results suggest that increased total body fat may be an important factor in the regulation of serum TNF-a levels. On the other hand, previous studies have shown that hyperglycemia or the presence of diabetes mellitus may enhance the production of TNF-a from monocytes in vitro (11, 12). Our present results suggest that hyperglycemia does not affect the serum levels of TNF-a. The association between insulin resistance and TNF-a has been previously reported (13–22). In the current study, we found that the serum levels of TNF-a are inversely correlated with GIR in obese NIDDM patients after, but not before, the treatment. The lack of correlation between the serum TNF-a levels and GIR before the treatment was probably the result of the influence of obesity and glucose toxicity on GIR. Kroder et al. (23) reported that the mechanism of insulin resistance caused by TNF-a differs from that induced by hyperglycemia. Our results suggest that serum TNF-a levels may play an important role in the mechanism of insulin resistance associated with obesity. The serum levels of TNF-a found in the present study were relatively low, and circulating TNF-a may not be biologically active at such low concentration. However, it is probable that locally produced TNF-a may act synergistically with circulating TNF-a on fatty and muscular tissues. In conclusion, our results suggest that serum TNF-a levels are influenced by body fat accumulation and that they may contribute to the insulin resistance associated with obesity. Acknowledgments We thank Dr. Y. Adachi and Dr. E. C. Gabazza for their helpful discussion.
References 1. Moller DE, Flier JS. 1991 Insulin resistance-mechanisms, syndromes, and implications. N Engl J Med. 211:938 –948. 2. Hotamisligil GS, Spiegelman BM. 1994 Tumor necrosis factor-a: a key component of the obesity-diabetes link. Diabetes. 43:1271–1278. 3. Hotamisligil GS, Shargill NS, Spiegelman BM. 1993 Adipose expression of
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
17.
18. 19.
20.
21.
22.
23.
JCE & M • 1998 Vol 83 • No 3
tumor necrosis factor-a: direct role in obesity-linked insulin resistance. Science. 259:87–91. Feinatein R, Kanety H, Papa MZ, Lunenfeld B, Karasik A. 1993 Tumor necrosis factor-a suppresses insulin-induced tyrosine phosphorylation of insulin receptor and its substrates. J Biol Chem. 268:26055–26058. Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, Simsolo RB. 1995 The expression of tumor necrosis factor in human adipose tissue, regulation by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest. 95:2111–2119. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM. 1995 Increased adipose tissue expression of tumor necrosis factor-a in human obesity and insulin resistance. J Clin Invest. 95:2409 –2415. Tokunaga K, Matsuzawa Y, Ishikawa K, Tarui S. 1983 A novel technique for the determination of body fat by computed tomography. Int J Obesity. 7:437– 445. DeFronzo RA, Tobin JD, Andres R. 1979 Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 237:E214 –223. Kissebah AH, Vydelingum N, Murray R, et al. 1982 Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab. 54:254 –260. Fujioka S, Matsuzawa Y, Tokunaga K, Tarui S. 1987 Contribution of intraabdominal fat accumulation to the impairment of glucose and lipid metabolism in human obesity. Metabolism. 36:54 –59. Morohoshi M, Fujisawa k, Uchimura I, Numano F. 1996 Glucose-dependent interleukin 6 and tumor necrosis factor production by human peripheral blood monocytes in vitro. Diabetes. 45:954 –959. Ohno Y, Aoki N, Nishimura A. 1993 In vitro production of interleukin-1, interleukin-6, and tumor necrosis factor-a in insulin-dependent diabetes mellitus. J Clin Endocrinol Metab. 77:1072–1077. Ling PR, Bistrian BR, Mendez B, Istfan NW. 1994 Effects of infusions of endotoxin, tumor necrosis factor, and interleukin-1 on glucose metabolism in the rat: relationship to endogenous glucose production and peripheral tissue glucose uptake. Metabolism. 43:279 –284. Lang CH, Dobrescu C, Bagby GJ. 1992 Tumor necrosis factor impairs insulin action on peripheral glucose disposal and hepatic glucose output. Endocrinology. 130:43–52. Hotamisligil GS, Budavari A, Murray D, Spiegelman BM. 1994 Reduced tyrosine kinase activity of the insulin receptor in obesity-diabetes. J Clin Invest. 94:1543–1549. Hotamisligil GS, Murray DL, Choy LN, Spiegelman BM. 1994 Tumor necrosis factor-a inhibits signaling from the insulin receptor. Proc Natl Acad Sci USA. 91:4854 – 4858. Hotamisligil GS, Peraldi P, Budavari A, Ellis R, White MF, Spiegelman BM. 1996 IRS-1-mediated inhibition of insulin receptor tyrosine kinase activity in TNF-a- and obesity-induced insulin resistance. Science. 271:665– 668. Begum N, Ragolia L. 1996 Effect of tumor necrosis factor-a on insulin action in cultured rat skeletal muscle cells. Endocrinology. 137:2441–2446. Kanety H, Hemi R, Papa MZ, Karasik A. 1996 Sphingomyelinase and ceramide suppress insulin-induced tyrosine phosphorylation of the insulin receptor substrate-1. J Biol Chem. 271:9895–9897. Szalkowski D, White-Carrington S, Berger J, Zhang B. 1995 Antidiabetic thiazolidines block the inhibitory effect of tumor necrosis factor-a on differentiation, insulin stimulated glucose uptake, and gene expression in 3T3–L1 cells. Endocrinology. 136:1474 –1481. Hauner H, Petruschke Th, Russ M, Rohrig K, Eckel J. 1995 Effects of tumor necrosis factor-alpha (TNF-a) on glucose transport and lipid metabolism of newly-differentiated human fat cells in cell culture. Diabetologia 38:764 –771. Saghizadeh M, Ong JM, Garvey WT, Henry RR, Kern PA. 1996 The expression of TNF a by human muscle, relationship to insulin resistance. J Clin Invest. 97:1111–1116. Kroder G, Bossenmainer B, Kellerer M, et al. 1996 Tumor necrosis factor-a and hyperglycemia-induced insulin resistance, evidence for different mechanisms and different effects on insulin signaling. J Clin Invest. 97:1471–1477.
