To clarify the significance of the serum levels of tumor necrosis factor-α (TNF-α) in the mechanism of insulin resistance, we studied 12 obese patients with noninsulin-dependent diabetes mellitus (NIDDM). We evaluated the relationship of TNF-α 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-α 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-α with the area of body fat, the GIR, and the resultant change in the TNF-α level.
Serum levels of TNF-α in the obese patients with NIDDM significantly exceeded those observed in normal subjects (P < 0.01) or in the nonobese patients with NIDDM (P < 0.01). Serum levels of TNF-α in obese NIDDM patients showed a significant positive correlation with the area of visceral fat before (r = 0.662, P < 0.03) and after (r = 0.508, P < 0.05) the treatment; similar correlation was observed in all patients with NIDDM before (r = 0.537, P < 0.02) and after (r = 0.430, P< 0.05) the treatment. Serum levels of TNF-α in obese NIDDM patients showed a significant negative correlation with GIR after the treatment (r = −0.595, P < 0.05). Serum levels of TNF-α 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-α levels may play an important role in mechanism of insulin resistance associated with obesity.
OBESITY 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-α (TNF-α) 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-α in adipose tissue would inhibit the transport of glucose in an autocrine or paracrine fashion. Until recently, blood levels of TNF-α could not be accurately measured because sensitive assays for TNF-α were not available (6).
In the present study, we measured the serum levels of TNF-α 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 tomography (CT) and with the glucose infusion rate (GIR) evaluated by the euglycemic hyperinsulinemic clamp technique.
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-α levels in 8 obese male subjects without NIDDM (age; 49.10 ± 2.29; BMI, 27.20 ± 0.20).
Clinical characteristics of the study subjects
| Normal | Nonobese NIDDM | Obese NIDDM | |
|---|---|---|---|
| n | 12 | 12 | 12 |
| Age (yr) | 54.42 ± 2.83 | 55.83 ± 1.53 | 50.00 ± 2.43 |
| Sex (M/F) | 11/1 | 11/1 | 11/1 |
| BMI (kg/m2) | 22.88 ± 0.45 | 23.71 ± 0.44 | 27.47 ± 0.33ab |
| Duration of DM (yr) | 8.50 ± 0.98 | 8.78 ± 0.57 | |
| HbAIc (%) | 9.05 ± 0.48 | 8.62 ± 0.56 | |
| Fasting glucose (mmol/L) | 4.96 ± 0.14 | 8.97 ± 0.37c | 8.92 ± 0.52b |
| Fasting insulin (pmol/L) | 28.62 ± 4.68 | 35.52 ± 5.28 | 46.14 ± 6.00b |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 45.02 ± 3.72d | |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 120.75 ± 10.13a | |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 138.00 ± 11.73d | |
| Systolic blood pressure (mm Hg) | 121.33 ± 4.33 | 127.58 ± 2.84 | 124.42 ± 3.48 |
| Diastolic blood pressure (mm Hg) | 72.50 ± 3.18 | 79.17 ± 2.55 | 80.58 ± 3.40 |
| Normal | Nonobese NIDDM | Obese NIDDM | |
|---|---|---|---|
| n | 12 | 12 | 12 |
| Age (yr) | 54.42 ± 2.83 | 55.83 ± 1.53 | 50.00 ± 2.43 |
| Sex (M/F) | 11/1 | 11/1 | 11/1 |
| BMI (kg/m2) | 22.88 ± 0.45 | 23.71 ± 0.44 | 27.47 ± 0.33ab |
| Duration of DM (yr) | 8.50 ± 0.98 | 8.78 ± 0.57 | |
| HbAIc (%) | 9.05 ± 0.48 | 8.62 ± 0.56 | |
| Fasting glucose (mmol/L) | 4.96 ± 0.14 | 8.97 ± 0.37c | 8.92 ± 0.52b |
| Fasting insulin (pmol/L) | 28.62 ± 4.68 | 35.52 ± 5.28 | 46.14 ± 6.00b |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 45.02 ± 3.72d | |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 120.75 ± 10.13a | |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 138.00 ± 11.73d | |
| Systolic blood pressure (mm Hg) | 121.33 ± 4.33 | 127.58 ± 2.84 | 124.42 ± 3.48 |
| Diastolic blood pressure (mm Hg) | 72.50 ± 3.18 | 79.17 ± 2.55 | 80.58 ± 3.40 |
Data are mean ± se.
P < 0.01; Nonobese NIDDM vs. obese NIDDM.
P < 0.05; Normal vs. obese NIDDM.
P < 0.01; Normal vs. nonobese NIDDM.
P < 0.05; Nonobese NIDDM vs. obese NIDDM.
Clinical characteristics of the study subjects
| Normal | Nonobese NIDDM | Obese NIDDM | |
|---|---|---|---|
| n | 12 | 12 | 12 |
| Age (yr) | 54.42 ± 2.83 | 55.83 ± 1.53 | 50.00 ± 2.43 |
| Sex (M/F) | 11/1 | 11/1 | 11/1 |
| BMI (kg/m2) | 22.88 ± 0.45 | 23.71 ± 0.44 | 27.47 ± 0.33ab |
| Duration of DM (yr) | 8.50 ± 0.98 | 8.78 ± 0.57 | |
| HbAIc (%) | 9.05 ± 0.48 | 8.62 ± 0.56 | |
| Fasting glucose (mmol/L) | 4.96 ± 0.14 | 8.97 ± 0.37c | 8.92 ± 0.52b |
| Fasting insulin (pmol/L) | 28.62 ± 4.68 | 35.52 ± 5.28 | 46.14 ± 6.00b |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 45.02 ± 3.72d | |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 120.75 ± 10.13a | |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 138.00 ± 11.73d | |
| Systolic blood pressure (mm Hg) | 121.33 ± 4.33 | 127.58 ± 2.84 | 124.42 ± 3.48 |
| Diastolic blood pressure (mm Hg) | 72.50 ± 3.18 | 79.17 ± 2.55 | 80.58 ± 3.40 |
| Normal | Nonobese NIDDM | Obese NIDDM | |
|---|---|---|---|
| n | 12 | 12 | 12 |
| Age (yr) | 54.42 ± 2.83 | 55.83 ± 1.53 | 50.00 ± 2.43 |
| Sex (M/F) | 11/1 | 11/1 | 11/1 |
| BMI (kg/m2) | 22.88 ± 0.45 | 23.71 ± 0.44 | 27.47 ± 0.33ab |
| Duration of DM (yr) | 8.50 ± 0.98 | 8.78 ± 0.57 | |
| HbAIc (%) | 9.05 ± 0.48 | 8.62 ± 0.56 | |
| Fasting glucose (mmol/L) | 4.96 ± 0.14 | 8.97 ± 0.37c | 8.92 ± 0.52b |
| Fasting insulin (pmol/L) | 28.62 ± 4.68 | 35.52 ± 5.28 | 46.14 ± 6.00b |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 45.02 ± 3.72d | |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 120.75 ± 10.13a | |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 138.00 ± 11.73d | |
| Systolic blood pressure (mm Hg) | 121.33 ± 4.33 | 127.58 ± 2.84 | 124.42 ± 3.48 |
| Diastolic blood pressure (mm Hg) | 72.50 ± 3.18 | 79.17 ± 2.55 | 80.58 ± 3.40 |
Data are mean ± se.
P < 0.01; Nonobese NIDDM vs. obese NIDDM.
P < 0.05; Normal vs. obese NIDDM.
P < 0.01; Normal vs. nonobese NIDDM.
P < 0.05; Nonobese NIDDM vs. obese NIDDM.
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 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-α in serum samples was measured using a commercially available sandwich immunoassay kit (Quantikine HS Human TNF-α immunoassay kit, R&D systems, Minneapolis, MN) and following the manufacturer’s instructions. Briefly, 200 μL of standard or serum samples were added to microtiter plate wells coated with a monoclonal antibody specific for TNF-α, followed by incubation at 4 C for 16 h. The wells were then washed 4 times with a buffered surfactant solution, and thereafter, 200 μL of anti-TNF-α polyclonal antibody conjugated to alkaline phosphatase were added to each well and incubated for 3 h at room temperature. After appropriate washing, 50 μL of substrate solution (NADPH) were added to each well and incubated again for 60 min at room temperature. After this, 50 μL 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-α levels were extrapolated from a curve drawn using standard TNF-α. 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-α or other cytokines was observed. The precision of the TNF-α assay using the above described immunoassay is better than that of other methods previously reported. Previous methods showed TNF-α 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 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-α 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 ± 43.42 pmol/L, after treatment: 1195.45 ± 57.48 pmol/L, nonobese group before treatment: 1188.30 ± 54.76 pmol/L, after treatment: 1111.02 ± 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 expressed as the mean ± se. Comparison between groups was done using the Mann-Whitney U test. The statistical difference between TNF-α 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-α were not statistically different between the normal and nonobese groups. Subjects of the obese group presented significantly higher serum concentrations of TNF-α than those of the normal (P < 0.01) and nonobese (P < 0.01) groups (Fig. 1). Serum TNF-α levels (1.89 ± 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 serum levels of TNF-α in normal subjects, obese, and nonobese patients with NIDDM. Serum levels of TNF-α 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).
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-α levels in the obese group significantly decreased after treatment (P < 0.01), but no significant change was observed in the nonobese group (Fig. 2).
Clinical characteristics of NIDDM patients before and after treatment
| Before treatment | After treatment | |
|---|---|---|
| Nonobese NIDDM | ||
| BMI (kg/m2) | 23.71 ± 0.44 | 22.93 ± 0.36a |
| HbAIc (%) | 9.05 ± 0.48 | 7.84 ± 0.40a |
| Fasting glucose (mmol/L) | 8.97 ± 0.37 | 5.81 ± 0.15a |
| Fasting insulin (pmol/L) | 35.52 ± 5.28 | 29.28 ± 3.06 |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 56.91 ± 3.33a |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 69.81 ± 9.42b |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 90.07 ± 10.67 |
| Obese NIDDM | ||
| BMI (kg/m2) | 27.47 ± 0.33 | 26.15 ± 0.30a |
| HbAIc (%) | 8.78 ± 0.57 | 7.68 ± 0.45a |
| Fasting glucose (mmol/L) | 8.92 ± 0.52 | 5.75 ± 0.29a |
| Fasting insulin (pmol/L) | 46.14 ± 6.00 | 33.60 ± 3.06b |
| GIR (μmol/kg/min) | 45.02 ± 3.72 | 50.07 ± 4.16b |
| Visceral fat area (cm2) | 120.75 ± 10.13 | 105.47 ± 9.59a |
| Subcutaneous fat area (cm2) | 138.00 ± 11.73 | 122.44 ± 10.21a |
| Before treatment | After treatment | |
|---|---|---|
| Nonobese NIDDM | ||
| BMI (kg/m2) | 23.71 ± 0.44 | 22.93 ± 0.36a |
| HbAIc (%) | 9.05 ± 0.48 | 7.84 ± 0.40a |
| Fasting glucose (mmol/L) | 8.97 ± 0.37 | 5.81 ± 0.15a |
| Fasting insulin (pmol/L) | 35.52 ± 5.28 | 29.28 ± 3.06 |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 56.91 ± 3.33a |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 69.81 ± 9.42b |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 90.07 ± 10.67 |
| Obese NIDDM | ||
| BMI (kg/m2) | 27.47 ± 0.33 | 26.15 ± 0.30a |
| HbAIc (%) | 8.78 ± 0.57 | 7.68 ± 0.45a |
| Fasting glucose (mmol/L) | 8.92 ± 0.52 | 5.75 ± 0.29a |
| Fasting insulin (pmol/L) | 46.14 ± 6.00 | 33.60 ± 3.06b |
| GIR (μmol/kg/min) | 45.02 ± 3.72 | 50.07 ± 4.16b |
| Visceral fat area (cm2) | 120.75 ± 10.13 | 105.47 ± 9.59a |
| Subcutaneous fat area (cm2) | 138.00 ± 11.73 | 122.44 ± 10.21a |
Data are mean ± se.
P < 0.01; before vs. after treatment.
P < 0.05; Before vs. after treatment.
Clinical characteristics of NIDDM patients before and after treatment
| Before treatment | After treatment | |
|---|---|---|
| Nonobese NIDDM | ||
| BMI (kg/m2) | 23.71 ± 0.44 | 22.93 ± 0.36a |
| HbAIc (%) | 9.05 ± 0.48 | 7.84 ± 0.40a |
| Fasting glucose (mmol/L) | 8.97 ± 0.37 | 5.81 ± 0.15a |
| Fasting insulin (pmol/L) | 35.52 ± 5.28 | 29.28 ± 3.06 |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 56.91 ± 3.33a |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 69.81 ± 9.42b |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 90.07 ± 10.67 |
| Obese NIDDM | ||
| BMI (kg/m2) | 27.47 ± 0.33 | 26.15 ± 0.30a |
| HbAIc (%) | 8.78 ± 0.57 | 7.68 ± 0.45a |
| Fasting glucose (mmol/L) | 8.92 ± 0.52 | 5.75 ± 0.29a |
| Fasting insulin (pmol/L) | 46.14 ± 6.00 | 33.60 ± 3.06b |
| GIR (μmol/kg/min) | 45.02 ± 3.72 | 50.07 ± 4.16b |
| Visceral fat area (cm2) | 120.75 ± 10.13 | 105.47 ± 9.59a |
| Subcutaneous fat area (cm2) | 138.00 ± 11.73 | 122.44 ± 10.21a |
| Before treatment | After treatment | |
|---|---|---|
| Nonobese NIDDM | ||
| BMI (kg/m2) | 23.71 ± 0.44 | 22.93 ± 0.36a |
| HbAIc (%) | 9.05 ± 0.48 | 7.84 ± 0.40a |
| Fasting glucose (mmol/L) | 8.97 ± 0.37 | 5.81 ± 0.15a |
| Fasting insulin (pmol/L) | 35.52 ± 5.28 | 29.28 ± 3.06 |
| GIR (μmol/kg/min) | 53.68 ± 3.61 | 56.91 ± 3.33a |
| Visceral fat area (cm2) | 81.11 ± 6.59 | 69.81 ± 9.42b |
| Subcutaneous fat area (cm2) | 99.84 ± 14.12 | 90.07 ± 10.67 |
| Obese NIDDM | ||
| BMI (kg/m2) | 27.47 ± 0.33 | 26.15 ± 0.30a |
| HbAIc (%) | 8.78 ± 0.57 | 7.68 ± 0.45a |
| Fasting glucose (mmol/L) | 8.92 ± 0.52 | 5.75 ± 0.29a |
| Fasting insulin (pmol/L) | 46.14 ± 6.00 | 33.60 ± 3.06b |
| GIR (μmol/kg/min) | 45.02 ± 3.72 | 50.07 ± 4.16b |
| Visceral fat area (cm2) | 120.75 ± 10.13 | 105.47 ± 9.59a |
| Subcutaneous fat area (cm2) | 138.00 ± 11.73 | 122.44 ± 10.21a |
Data are mean ± se.
P < 0.01; before vs. after treatment.
P < 0.05; Before vs. after treatment.
Effect of treatment on the serum levels of TNF-α in obese and nonobese patients with NIDDM. The levels of TNF-α decreased significantly (P < 0.01) only in the obese patients with NIDDM.
In the obese NIDDM patients, serum levels of TNF-α were positively correlated with the visceral fat area before (r = 0.662, P < 0.03) and after (r = 0.508, P< 0.05) the treatment. There was also a significant positive correlation between serum levels of TNF-α and visceral fat area before (r = 0.537, P < 0.02) and after (r = 0.430, P < 0.05) the treatment in all patients with NIDDM, including obese and nonobese subjects (Fig. 3). Serum levels of TNF-α 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.
Correlation between the serum level of TNF-α and the area of visceral fat in all patients with NIDDM before and after treatment. A significant positive correlation was observed before (r = 0.537, P < 0.02) and after (r = 0.430, P < 0.05) treatment. ○, Obese NIDDM; •, nonobese NIDDM.
After the treatment, a significant negative correlation was observed between the serum TNF-α levels and the GIR in obese patients with NIDDM (r = −0.595, P < 0.05) (Fig. 4).
Correlation between the serum TNF-α level and the glucose infusion rate (GIR) in obese patients with NIDDM. A significant negative correlation was observed only after treatment (r = −0.595, P < 0.05).
Discussion
The present study showed that TNF-α is significantly increased in NIDDM patients with obesity. In obese patients with NIDDM, the serum levels of TNF-α 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 study in obese NIDDM subjects would probably not be observed in typical obese subjects. Namely, although in our present study the serum TNF-α levels were significantly correlated only with the visceral fat area, they might also be significantly correlated with the subcutaneous fat area and with the total fat area in typically obese patients with NIDDM.
The elevation of the serum levels of TNF-α 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-α. These results suggest that increased total body fat may be an important factor in the regulation of serum TNF-α levels.
On the other hand, previous studies have shown that hyperglycemia or the presence of diabetes mellitus may enhance the production of TNF-α from monocytes in vitro (11, 12). Our present results suggest that hyperglycemia does not affect the serum levels of TNF-α.
The association between insulin resistance and TNF-α has been previously reported (13–22). In the current study, we found that the serum levels of TNF-α are inversely correlated with GIR in obese NIDDM patients after, but not before, the treatment. The lack of correlation between the serum TNF-α 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-α differs from that induced by hyperglycemia. Our results suggest that serum TNF-α levels may play an important role in the mechanism of insulin resistance associated with obesity. The serum levels of TNF-α found in the present study were relatively low, and circulating TNF-α may not be biologically active at such low concentration. However, it is probable that locally produced TNF-α may act synergistically with circulating TNF-α on fatty and muscular tissues.
In conclusion, our results suggest that serum TNF-α levels are influenced by body fat accumulation and that they may contribute to the insulin resistance associated with obesity.
Acknowledgements
We thank Dr. Y. Adachi and Dr. E. C. Gabazza for their helpful discussion.
This work was supported in part by a grant from the Mie Prefecture Medical Association.
