Effect of Pioglitazone on Circulating Adipocytokine Levels and Insulin Sensitivity in Type 2 Diabetic Patients

We examined the effect of pioglitazone (PIO) on circulating adipocytokine levels to elucidate the mechanisms by which thiazolidinediones improve insulin resistance in type 2 diabetes mellitus (T2DM). Twenty-three subjects with T2DM (age 54 ± 2 yr, body mass index 29 ± 1 kg/m²) were randomly assigned to receive placebo (n = 11) or PIO, 45 mg/d (n = 12), for 4 months. Before and after treatment, subjects received a 75-g oral glucose tolerance test (OGTT); euglycemic insulin clamp (40 mU/m²·min) with 3-³H-glucose; determination of fat mass (³H₂O); and measurement of fasting glucose, free fatty acids (FFAs), leptin, adiponectin, and TNFα concentrations. After 4 months of PIO, fasting plasma glucose concentration (Δ = −2.7 mol/liter), mean plasma glucose during OGTT (Δ = −3.8 mol/liter), and hemoglobin A_1c (Δ = 1.7%) decreased (P < 0.05 vs. placebo) without change in fasting or post-OGTT plasma insulin levels. Fasting FFAs (Δ = 168 μmol/liter) and TNFα (Δ = 0.7 pg/ml) concentrations decreased (P < 0.05 vs. placebo), whereas adiponectin (Δ = 8.7 μg/ml) increased (P < 0.01 vs. placebo). Despite the increase in body fat mass (Δ = 3.4 kg) after PIO, plasma leptin concentration did not change significantly. No changes in plasma glucose, FFAs, or adipocytokine levels were observed in placebo-treated subjects. During the insulin clamp, endogenous (hepatic) glucose production decreased (Δ = −2.67 μmol/fat-free mass·min, P < 0.05 vs. placebo), whereas metabolic clearance rate of glucose (MCR) increased (Δ = 0.58 ml/fat-free mass·min, P < 0.05 vs. placebo) after PIO. In all subjects, before and after PIO, the decrease in plasma FFA concentration was correlated with the changes in both endogenous (hepatic) glucose production (r = 0.47, P < 0.05) and MCR (r = −0.41, P < 0.05), whereas the increase in plasma adiponectin concentration was correlated with the change in endogenous (hepatic) glucose production (r = −0.70, P < 0.01) and MCR (r = 0.49, P < 0.05). These results suggest that the direct effects of PIO on adipose tissue to decrease plasma FFA levels and increase plasma adiponectin contribute to the improvements in hepatic and peripheral insulin sensitivity and glucose tolerance in patients with T2DM.

THE THIAZOLIDINEDIONES (TZDS) represent a new class of insulin-sensitizing agents that have proven effective in the treatment of patients with type 2 diabetes mellitus (T2DM). TZDs initiate their action by binding to a specific nuclear receptor termed the peroxisome proliferator activated receptor (PPAR)-γ (1, 2), and their binding affinity to PPARγ is closely related to their in vivo antihyperglycemic potency (3, 4). PPARγ is found in high concentration in adipocytes (1), whereas its concentration in muscle and liver, the primary tissues responsible for the maintenance of glucose homeostasis (glucose disposal and production, respectively), is low (1). Consistent with their tissue distribution, PPARγ activation causes preadipocytes to differentiate into mature fat cells and causes the induction of key enzymes involved in lipogenesis (1, 5, 6). Although TZD therapy in T2DM patients consistently is associated with weight gain (7–11), glycemic control improves and the increase in body weight is positively related with the reduction in hemoglobin A_1c (HbA_1c) (11–13). These results suggest that the improvement in glucose homeostasis after TZD treatment may, in some way, be related to an alteration in lipid metabolism and/or fat topography.

Recently adipose tissue has been recognized as an important source of metabolically active secretory products (adipocytokines), including free fatty acids (FFAs), leptin, TNFα, IL-6, plasminogen activator inhibitor-1, resistin, and adiponectin (14). The deleterious effect of elevated plasma FFA levels on hepatic and muscle insulin sensitivity are well documented (15–19). However, the relationship between these other adipocytokines and hepatic/muscle insulin sensitivity in T2DM subjects has been less well studied. TNFα, which is expressed in adipocytes from lean individuals, is overexpressed in large adipocytes from obese people and has been shown to cause insulin resistance in both animals (20) and humans (21) through impairment of insulin-mediated signaling pathways (22). Deficient leptin secretion by adipocytes causes profound obesity in rodents (23) and humans (24), and leptin has been shown to exert important effects on glucose metabolism in rodents (25). Although restoration of leptin levels to normal in leptin-deficient rodents ameliorates the insulin resistance independent of its effects on food intake and body weight (26), the relationship between plasma leptin levels and insulin resistance in T2DM has been poorly studied in humans. Most recently, the adipocytokine, adiponectin, and its role in the modulation of glucose and lipid metabolism has received considerable attention. Decreased plasma adiponectin levels have been demonstrated in a variety of insulin-resistant states, including obesity and T2DM, in both animals and humans (27–30). Although the mechanisms by which adiponectin enhances insulin action in muscle and liver are unclear, it has been reported in animal studies that adiponectin enhances insulin-stimulated tyrosine phosphorylation of the insulin receptor in skeletal muscle (27) and decreases hepatic expression of the glucogenic enzymes, phosphoenolpyruvate carboxykinase, and glucose-6-phosphatase (31).

In the present double-blind, placebo-controlled study, we evaluated the effect of pioglitazone therapy on circulating adipocytokine levels (FFA, leptin, TNFα, adiponectin) and their relationship to changes in body fat mass, glucose tolerance, and hepatic and peripheral tissue sensitivity to insulin, in T2DM individuals.

Subjects and Methods

Subjects

Twenty-three patients with T2DM were recruited from the outpatient clinic of the Texas Diabetes Institute (Table 1). All subjects were taking a stable dose of sulfonylurea drug for at least 3 months before the study and remained on the same dose of sulfonylurea throughout the study period. Patients who previously had received insulin, metformin, or any TZDs were excluded. Entry criteria included age from 30 to 70 yr, body mass index (BMI) less than 36 kg/m², and fasting plasma glucose concentration (FPG) between 140 and 240 mg/dl. All patients were in good general health without evidence of cardiac, hepatic, renal, or other chronic diseases as determined by history, examination, and screening blood tests. Subjects were not consuming any medicines known to affect glucose metabolism, and none were participating in an excessive physical activity program. All subjects gave signed, voluntary, informed consent before participation. The protocol was approved by the Institutional Review Board of the University of Texas Health Science Center at San Antonio.

TABLE 1.

Baseline clinical and fasting laboratory characteristics

	Pioglitazone (n = 12)	Placebo (n = 11)
Age (yr)	54 ± 3	55 ± 4
Gender (M/F)	11 /1	6 /5
Body weight (kg)	85 ± 4	82 ± 5
BMI (kg/m2)	29 ± 1	30 ± 1
Fat mass (kg)	26 ± 3	24 ± 3
HbA1c (%)	8.8 ± 0.3	8.2 ± 0.3
FPG (mmol/liter)	9.6 ± 0.7	9.0 ± 0.4
FPI (μU/ml)	17 ± 3	20 ± 4
C-peptide (ng/ml)	2.7 ± 0.3	2.6 ± 0.4
Total cholesterol (mg/dl)1	169 ± 5	172 ± 6
Triglycerides (mg/dl)2	138 ± 15	123 ± 12
HDL cholesterol (mg/dl)1	35 ± 3	37 ± 2
LDL cholesterol (mg/dl)1	107 ± 4	111 ± 5
FFA (μmol)	647 ± 39	659 ± 47
Adiponectin (μg/ml)	7.4 ± 0.8	7.6 ± 0.7
TNFα (pg/ml)	5.3 ± 0.4	4.7 ± 0.2
Leptin (ng/ml)	8.1 ± 3.1	12.7 ± 2.7

M, Male; F, female; FPI, fasting plasma insulin.

¹

To convert total HDL/LDL cholesterol from milligrams per deciliter to millimoles per liter, multiply by 0.02586.

²

To convert triglyceride from milligrams per deciliter to millimoles per liter, multiply by 0.01129.

Study design

The study was double blind and placebo controlled in design. During the 4 wk before randomization, the FPG was measured on four occasions at weekly intervals, and in each subject the variability was less than 5%. During this 4-wk period, subjects met with the dietitian for 45 min and were instructed to consume a weight-maintaining diet containing 50% carbohydrate, 30% fat, and 20% protein. During the week before randomization, all subjects received: 1) a 75-g oral glucose tolerance test (OGTT); 2) measurement of lean body mass and fat mass using an iv bolus of ³H₂O; and 3) euglycemic insulin clamp in combination with tritiated glucose to examine hepatic and peripheral tissue sensitivity to insulin. After completion of these studies, subjects were assigned randomly in a double-blind fashion to receive either placebo or pioglitazone, 45 mg/d, with breakfast for 4 months. Subjects returned to the Clinical Research Center of Texas Diabetes Institute at 0800 h every 2 wk for measurement of FPG, body weight, and blood pressure. On each visit adherence to a weight-maintaining diet was reinforced, and no attempt was made during the study to change these dietary instructions. Fasting plasma lipids [total cholesterol, triglyceride, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol] were measured monthly. HbA_1c was measured on the last day of the baseline and treatment periods. All studies were performed in the postabsorptive state at 0800 h after an overnight 10- to 12-h fast. Subjects omitted their dose of sulfonylurea on the morning of the day of the study. During the last week of the double-blind period, the OGTT, euglycemic insulin clamp, and body fat measurement were repeated.

OGTT

Baseline blood samples for determination of plasma glucose, FFAs, insulin, and C-peptide concentrations were drawn at −30, −15, and 0 min. At time zero each subject ingested 75 g glucose in 300 ml orange-flavored water, and plasma glucose, FFAs, insulin, and C-peptide were measured at 15-min intervals for 2 h. At time zero, a 100-μCi bolus of ³H₂O was given and plasma-tritiated water radioactivity was determined at 90, 105, and 120 min for calculation of fat-free mass (FFM) and fat mass (FM) as previously described (32).

Euglycemic insulin clamp

Insulin sensitivity was assessed with the euglycemic insulin clamp, as previously described (33). On arrival (0800 h) at the Clinical Research Center, blood for measurement of FPG was obtained. A primed (25 μCi × FPG/100)-continuous (0.25 μCi/min) infusion of [3-³H]glucose was started at time −180 min via a catheter placed into an antecubital vein and continued throughout the study. A second catheter was placed retrogradely into a vein on the dorsum of the hand, which was then placed in a heated box (60 C). Baseline arterialized venous blood samples for determination of plasma [3-³H]glucose radioactivity and plasma glucose and insulin concentrations were drawn at −30, −20, −10, −5, and 0 min. At time zero, a prime-continuous infusion of human regular insulin (Novolin, Novo Nordisk Pharmaceuticals, Princeton, NJ) was started at a rate of 40 mU/min⁻¹·m⁻² body surface area and continued for 120 min. After initiation of the insulin infusion, the plasma glucose concentration was allowed to drop until it reached 5 mmol/liter, at which level it was maintained by appropriately adjusting a variable infusion of 20% dextrose (33). Throughout the insulin clamp, blood samples for determination of plasma glucose concentration were drawn every 5 min, and blood samples for determination of plasma insulin and [3-³H]glucose radioactivity were collected every 10–15 min.

Assays

Plasma glucose was measured at bedside using the glucose oxidase method (Glucose Analyzer 2, Beckman Instruments Inc., Fullerton, CA). Plasma insulin (Diagnostic Products Corp., Los Angeles, CA) and C-peptide (Diagnostic Systems Laboratories Inc., Webster, TX) concentrations were measured by RIA. HbA_1c was measured by affinity chromatography (Biochemical Methodology, Drower 4350; Isolab, Akron, OH). Plasma FFA was measured by enzymatic calorimetric quantification (Wako Chemicals GmbH, Neuss, Germany). Plasma total cholesterol and triglycerides were measured enzymatically (Roche Molecular Biochemicals, Indianapolis, IN) on a Hitachi 704 autoanalyzer. HDL cholesterol was measured enzymatically on a Hitachi 704 autoanalyzer after precipitation of chylomicrons and very low-density lipoprotein and LDL cholesterol by phosphotungstic acid. LDL cholesterol was calculated from the Friedwald equation. Tritiated glucose specific activity was determined on deproteinized plasma samples. TNFα concentration was determined with an ELISA kit for human TNFα (Quantikine, R&D Systems, Minneapolis, MN). This is a very sensitive sandwich ELISA with a lower detection limit of 0.5 pg/ml and intraassay and interassay coefficients of variation of 5.7 and 7.5%, respectively. Fasting plasma leptin and adiponectin concentrations were determined by RIA (Linco Research Inc., St. Charles, MO).

Calculations

Under steady state, postabsorptive conditions the rate of endogenous glucose appearance (Ra) is calculated as the [3-³H]glucose infusion rate (disintegrations per minute per minute) divided by the steady-state plasma [3-³H]glucose specific activity (disintegrations per minute per milligram). During the insulin clamp, nonsteady conditions prevail, and Ra was calculated from Steele’s equation (34). Endogenous glucose production (EGP) was calculated as: EGP = R_a minus glucose infusion rate (GIR). Total glucose disposal (TGD) equals the sum of EGP plus GIR after a very small correction for any change in the plasma glucose concentration. Total body glucose metabolic clearance rate (MCR) equals the TGD divided by the plasma glucose concentration, where TGD is expressed as milligram per kilogram FFM per minute, and plasma glucose is expressed as milligram per milliliter.

Total body water was calculated from the mean plasma ³H₂O radioactivity at time 90, 105, and 120 min after the iv bolus of ³H₂O. Plasma ³H₂O-specific activity was calculated, assuming that plasma water represents 93% of total plasma volume. Fat-free mass (FFM) was calculated by dividing total body water by 0.73 (35).

The area under the glucose, insulin, C-peptide, and FFA curves during the OGTT were determined using the trapezoidal rule. The mean plasma glucose, insulin, and FFA concentrations during the OGTT were calculated by dividing the area under the curve by the duration of the OGTT (120 min).

Statistical analysis

Statistic analyses were performed with StatView for Windows, version 5.0 (SAS Institute Inc., Cary, NC). Differences between values before and after treatment (i.e. within the placebo and within the pioglitazone groups) were analyzed using the paired Student’s t test. Comparison between the placebo and pioglitazone groups was performed using ANOVA with Bonferroni/Dunn post hoc testing when appropriate. Linear regression analysis was used to examine the relationships between the changes in hepatic and peripheral tissue insulin sensitivity and the changes in circulating concentrations of plasma FFA, TNFα, leptin, and adiponectin before and after pioglitazone treatment. All data are presented as the mean ± se. P < 0.05 was considered to be statistically significant.

Results

Clinical characteristics and metabolic measurements

Demographic characteristics of the patients are summarized in Table 1. The placebo and pioglitazone-treated groups were similar in age, race, duration of diabetes, BMI, fasting insulin and C-peptide concentrations, plasma lipids levels, FM, and FFM. FPG and HbA_1c were slightly, although not significantly, higher in the pioglitazone vs. placebo group. There were more men in the pioglitazone group. However, it is known that females respond better to TZDs than men because they have a greater percent body fat (10). Before randomization the plasma leptin concentration was slightly, although not significantly higher in the placebo group, perhaps because of the greater number of women in this group. There were no differences in plasma FFA, leptin, and adiponectin concentrations and serum TNFα concentration between the pioglitazone and placebo groups before treatment.

After 16 wk of treatment, there were significant increases in body weight, BMI, and FM and significant decreases in HbA_1c and fasting plasma glucose, C-peptide, and triglyceride concentrations in the pioglitazone group (P < 0.05–0.001 vs. baseline and vs. placebo) (Table 2). There was no significant change from baseline in any parameter in the placebo group (Table 2).

TABLE 2.

Effect of treatment with pioglitazone and placebo on body weight, fat mass, plasma lipids, and parameters of glycemic control

	Pioglitazone (n = 12)	Placebo (n = 11)	P
Δ Body weight (kg)	3.6 ± 1.4	0.3 ± 0.4	0.04
Δ BMI (kg/m2)	1.3 ± 0.5	0.1 ± 0.2	0.03
Δ Fat mass (kg)	3.5 ± 1.3	0.3 ± 0.4	0.04
Δ HbA1c (%)	−1.7 ± 0.3	0.0 ± 0.2	<0.001
Δ FPG (mmol/liter)	−2.8 ± 0.6	1.4 ± 1.2	<0.01
Δ FPI (μU/ml)	−2 ± 1	−1 ± 1
Δ C-peptide (ng/ml)	−0.1 ± 0.6	−0.1 ± 0.5
Δ FFA (μmol/liter)	−168 ± 52	13 ± 48	<0.001
Δ Total cholesterol (mg/dl)1	−7 ± 6	−1 ± 5
Δ Triglycerides (mg/dl)2	−33 ± 11	1 ± 11	0.04
Δ HDL cholesterol (mg/dl)1	1 ± 1	−1 ± 1
Δ LDL cholesterol (mg/dl)1	−2 ± 5	0 ± 4

FPI, Fasting plasma insulin.

¹

To convert total HDL/LDL cholesterol from milligrams per deciliter to millimoles per liter multiply by 0.02586.

²

To convert triglyceride from milligrams per deciliter to millimoles per liter multiply by 0.01129.

During the OGTT (Fig. 1), mean plasma glucose (16.3 ± 0.7 to 12.5 ± 0.8 mmol/liter) and FFA concentrations (483 ± 30 to 347 ± 33 μmol/liter) decreased significantly (P < 0.01 vs. baseline and vs. placebo) without change in the plasma insulin concentration after pioglitazone treatment. There were no significant changes in plasma levels of glucose, insulin, and FFA after placebo treatment.

Fig. 1.

Plasma glucose, insulin, and FFA concentrations during the 75-g OGTT before and after treatment with pioglitazone (top) or placebo (bottom). P values indicate significant differences between before and after treatment.

Euglycemic insulin clamp

During the pretreatment insulin clamp, the steady-state (90–120 min) plasma glucose (5.7 ± 0.4 vs. 5.4 ± 0.3 mmol/liter) and insulin (69 ± 8 and 77 ± 8 μU/ml) concentrations were similar in the pioglitazone and placebo groups, respectively. After 16 wk of pioglitazone and placebo treatment, the steady-state plasma glucose and insulin concentrations were similar to those in the baseline insulin clamp study.

Before treatment, the basal rate of EGP was similar in the pioglitazone (15.56 ± 0.56 μmol/kg FFM per minute) and placebo (13.94 ± 0.61) groups and remained unchanged after 16 wk of treatment (pioglitazone = 15.56 ± 0.56 μmol/kg FFM per minute and placebo = 14.72 ± 0.61). During the pretreatment insulin clamp suppression of EGP was similar in the pioglitazone and placebo groups (6.11 ± 0.56 vs. 5.78 ± 0.56 μmol/kg FFM per minute, respectively) After 4 months of pioglitazone treatment, there was a further decrement in EGP (Δ = −2.72 ± 1.11 μmol/kg FFM per minute, P < 0.05 vs. baseline and vs. placebo) during the insulin clamp (Fig. 2). In contrast, the suppression of EGP during the insulin clamp was less complete (Δ = +2.28 ± 1.11 μmol/kg FFM per minute) after 4 months of placebo treatment. The total body glucose MCR (3.5 ± 0.5 vs. 3.4 ± 0.4 ml/kg FFM per minute, respectively) was similar in the pioglitazone and placebo groups before treatment. After 4 months of treatment, the MCR of glucose during the insulin clamp increased in the pioglitazone group (Δ = +0.58 ± 0.41 ml/kg FFM per minute, P < 0.05 vs. baseline and P < 0.05 vs. placebo), whereas it deceased modestly (Δ = −0.47 ± 0.19 ml/kg FFM per minute) in the placebo group (Fig. 2).

Fig. 2.

Changes (Δ) in total body glucose MCR and EGP during the euglycemic insulin clamp after treatment with placebo (PLAC) and pioglitazone (PIO).

Circulating adipocytokine concentrations

After pioglitazone treatment fasting plasma FFA and TNFα levels decreased significantly, whereas the fasting plasma adiponectin concentration increased from 7.4 ± 1.0 to 16.2 ± 2.1 μg/ml (P < 0.05–0.001 vs. baseline and vs. placebo) (Fig. 3). Fasting plasma leptin did not change significantly in either the pioglitazone or placebo groups (Fig. 3).

Fig. 3.

Change (Δ) in circulating FFA, TNFα, leptin, and adiponectin concentrations after treatment with placebo (PLAC) and pioglitazone (PIO).

Regression analysis

The changes in body weight (r = −0.65, P < 0.001), BMI (r = −0.67), P < 0.001, and FM (r = −0.63, P < 0.001) all correlate strongly with the change in HbA_1c. When the data before and after 16 wk of treatment with pioglitazone or placebo are analyzed collectively in all 23 subjects, the change in fasting plasma FFA concentration was correlated with the changes in both EGP (r = 0.47, P < 0.05) and MCR (r = −0.41, P < 0.05) during the insulin clamp (Fig. 4). The change in fasting plasma adiponectin concentration was correlated strongly with the changes in both EGP (r = −0.70, P < 0.001) and MCR (r = 0.49, P < 0.05) during the insulin clamp (Fig. 4). The change in fasting plasma adiponectin concentration also was correlated with the change in fasting plasma FFA concentration (r = −0.57, P < 0.01) and mean plasma FFA concentration during the OGTT (r = −0.64, P < 0.01). The changes in fasting serum TNFα and plasma leptin concentrations were not correlated with the changes in either EGP or MCR during the insulin clamp. The changes in body fat after treatment did not correlate with the changes in any circulating adipocytokines.

Fig. 4.

Relationship between the changes (Δ) in total body glucose MCR and EGP during the insulin clamp vs. the changes (Δ) in plasma FFA (left) and plasma adiponectin (right) concentrations after treatment with placebo and pioglitazone.

Discussion

In the present double-blind, randomized, placebo-controlled study, we examined the effect of 16 wk of pioglitazone (45 mg/d) treatment on fasting adipocytokine (FFA, TNFα, leptin, and adiponectin) levels in T2DM subjects and related changes in these circulating adipocytokines to changes in body composition, glycemic control, and peripheral and hepatic insulin sensitivity. Consistent with previous findings from our laboratory and others (12, 13, 36–41), we demonstrated that pioglitazone lowered the HbA_1c and fasting plasma glucose concentration and improved glucose tolerance in T2DM patients without significant change in the plasma insulin concentration, indicating enhanced tissue sensitivity to insulin. The improvement in insulin sensitivity was confirmed with the euglycemic insulin clamp/tritiated glucose technique, which demonstrated that both peripheral tissues (muscle) and liver were responsible for the amelioration of insulin resistance (Fig. 2).

Recent studies demonstrated that the adipocyte is a metabolic factory capable of producing a number of adipocytokines (14). Of these adipocytokines, fatty acids have been the most extensively studied (15–19, 42). However, the adipocyte also produces a variety of other adipocytokines that play important roles in regulating glucose and lipid metabolism (adiponectin, resistin, TNFα, leptin), the coagulation cascade (plasminogen activator inhibitor-1), and inflammation (IL-6, resistin, TNFα) (14). In the present study, we focused on the effect of TZDs on those adipocytokines that have been shown to play an important role in the regulation of glucose metabolism in animals and humans with T2DM.

Many investigators have demonstrated that TZD treatment reduces the fasting plasma FFA concentration by approximately 20–30% and that this decrease is paralleled by a 20–30% reduction in fasting plasma glucose concentration and HbA_1c in T2DM patients (8, 10, 13, 36–38, 43). In agreement with these previous clinical studies, pioglitazone reduced both the fasting and post-OGTT plasma FFA by 25–35%,and the reduction in plasma FFA concentration was correlated with the decreases in HbA_1c, fasting plasma glucose concentration, and mean plasma glucose during the OGTT as well as with the increases in hepatic and peripheral (muscle) insulin sensitivity. Because elevated plasma FFAs are known to induce both hepatic and peripheral (muscle) tissue insulin resistance (15–19, 44, 45), the present results suggest a major insulin sensitizing action of the TZDs in media via their FFA lowering effect.

Before the start of pioglitazone, the plasma adiponectin concentration (7.4 ± 0.8 μg/ml) in T2DM was decreased, compared with values in healthy, age/gender/obesity-matched nondiabetic subjects studied in our laboratory (8.9 ± 0.5 μg/ml). After pioglitazone treatment the circulating adiponectin concentration increased in every diabetic subject by a mean of 2.2-fold. The increment in plasma adiponectin concentration was strongly correlated with the increments in hepatic and peripheral (muscle) tissue insulin sensitivity after pioglitazone treatment. We made similar observations in an open-label pioglitazone trial in a completely separate group of T2DM subjects (46). To the best of our knowledge, these two investigations represent the first demonstration that the improvement in hepatic insulin sensitivity after TZD therapy is closely associated with the increase in plasma adiponectin concentration. Our results are consistent with three previous reports that demonstrated that TZD treatment increases plasma adiponectin levels in concert with an improvement in peripheral tissue insulin sensitivity (47–49).

Improved glycemic control also has been shown to be associated with an increase in plasma adiponectin concentration in TZD-treated T2DM patients (50–52). The precise mechanism(s) by which adiponectin exerts its actions on glucose metabolism and insulin sensitivity remain unclear. Coombs et al. (31) demonstrated that ip administration of adiponectin to mice inhibited the expression of hepatic gluconeogenic enzymes and decreased EGP. Yamauchi et al. (27) reported that adiponectin administration to rodents increased insulin-stimulated tyrosine phosphorylation of the insulin receptor in skeletal muscle in association with increased whole-body insulin sensitivity. The present results are consistent with these results in animals (27, 31). In the present study, we observed a strong correlation between the increase in plasma adiponectin concentration and the decrease in both fasting and post-OGTT plasma FFA levels. This is especially noteworthy because FFA infusion in humans has been shown to inhibit the insulin signal transduction system at the level of insulin receptor substrate (IRS)-1/phosphatidylinositol 3 kinase (PI-3 kinase) (53), and we previously have shown that rosiglitazone treatment in T2DM patients enhances IRS-1 tyrosine phosphorylation, the association of IRS-1 with PI-3 kinase, and PI-3 kinase activity (43). Moreover, these improvements in insulin signaling after rosiglitazone therapy were closely correlated with the reduction in plasma FFA concentration. Because the increase in plasma adiponectin concentration in the present study was strongly related with the reduction in plasma FFA levels, one could postulate that the primary effect of adiponectin is to inhibit lipolysis in adipocytes and/or increase fatty acid oxidation in muscle (27), leading to a decrease in intracellular fatty acid metabolites, i.e. fatty acyl coenzyme A, diacylglycerol, ceramides, and enhanced insulin signal transduction (54–56). Alternatively, adiponectin may have a direct effect on the insulin signaling cascade (27). We believe that it is unlikely that the decrease in plasma FFA concentration is responsible for the increase in plasma adiponectin after pioglitazone because reduction in circulating FFA levels with acipimox has no effect on the plasma adiponectin concentration (Bajaj, M. and R. A. DeFronzo, unpublished results).

In the present study, pioglitazone treatment was associated with a weight gain of 3.6 kg, and this was entirely accounted for by an increase in FM of 3.5 kg. Edema was looked for carefully on every visit and was not observed in any subject. Despite the weight gain, oral glucose tolerance, HbA_1c, and hepatic and peripheral tissue (muscle) insulin sensitivity improved significantly. In fact, the greater the increase in body weight, the greater the decrease in HbA_1c. Improved glycemic control, despite weight gain, has been reported with other TZDs (7, 9, 10, 13, 39, 41), including troglitazone, pioglitazone, and rosiglitazone. Several studies demonstrated that TZD-induced weight gain is associated with an increase in sc adipose tissue and a concomitant decrease in visceral fat content (9, 12, 57). This redistribution of fat is explained by the remodeling of fat tissue, associated with differentiation of preadipocytes into small fat cells in sc fat depots (58) and apoptosis of differentiated large hypertrophic adipocytes in visceral and sc fat depots (59, 60). Although similar quantitative measurements in human T2DM patients have yet to be reported after TZD therapy, such a scenario (52–54) would be consistent with the present observations. Because the newly formed fat cells possess all of the enzymes required for lipogenesis and manifest increased insulin sensitivity with respect to glucose transport and inhibition of lipolysis (58, 59–61), this remodeling of adipose tissue also could explain the decline in plasma FFA levels, resulting secondarily in enhanced hepatic and muscle sensitivity to insulin.

In human and rodent models of obesity (29, 62), circulating adiponectin levels are decreased. Therefore, one would not expect further weight gain, induced by pioglitazone, to result in an increase in plasma adiponectin levels, as was observed in the present study. Moreover, if the weight gain was associated with enlarged fat cells, as occurs with overeating, one might even expect to observe a further reduction in plasma adiponectin concentration. We postulate that the differentiation of preadipocytes into small fat cells in sc fat depots, in association with enhanced expression of adiponectin mRNA (52), is responsible for the increase in plasma adiponectin after pioglitazone treatment.

Pioglitazone treatment resulted in a significant decrease in serum TNFα concentration. However, we failed to observe any correlation between the decrease in circulating TNFα concentration and the improvements in glucose tolerance or hepatic and peripheral tissue insulin sensitivity. However, one cannot exclude a local paracrine effect of reduced tissue TNFα levels in muscle or liver. Although the plasma leptin concentration correlated strongly with FM before treatment (r = 0.53, P < 0.01), plasma leptin levels did not change after pioglitazone treatment, despite the significant increase in FM. Because circulating leptin levels normally increase with weight gain, the failure to observe any change in plasma leptin concentration, despite a 3.5 kg increase in FM, also suggests that the pioglitazone-induced increase in FM does not simply represent the enlargement of adipocytes in the body. The lack of change in circulating leptin levels after pioglitazone treatment also suggests that this adipocytokine does not contribute to the improvements in hepatic and peripheral tissue (muscle) sensitivity to insulin.

In summary, the present results demonstrate that 16 wk of pioglitazone treatment decreases the fasting and post-OGTT plasma FFA levels, reduces the serum TNFα concentration, and increases the plasma adiponectin concentration despite weight gain in T2DM patients. Consistent with previously published results, pioglitazone also decreased the fasting plasma glucose concentration, improved oral glucose tolerance, reduced HbA_1c, and enhanced hepatic and peripheral tissue sensitivity to insulin. The reduction in plasma FFA levels and increase in plasma adiponectin concentration after pioglitazone treatment both were strongly associated with the improvements in peripheral tissue (muscle) and hepatic insulin sensitivity and glycemic control. These observations enhance our understanding of the mechanism of action the TZDs in T2DM patients.

Acknowledgements

We thank our nurses, Socorro Mejorado and Magda Ortiz, for their assistance in performing the insulin clamp and OGTTs, and the care of the patients throughout the study. Elva Chapa and Lorrie Albarado provided expert secretarial assistance in preparing the manuscript.

This work was supported in part by a grant from Takeda Pharmaceuticals North America.

Abbreviations:

 
• BMI,

  Body mass index;

 
• EGP,

  endogenous glucose production;

 
• FFA,

  free fatty acid;

 
• FFM,

  fat-free mass;

 
• FM,

  fat mass;

 
• FPG,

  fasting plasma glucose concentration;

 
• HbA_1c,

  hemoglobin A_1c;

 
• HDL,

  high-density lipoprotein;

 
• IRS,

  insulin receptor substrate;

 
• LDL,

  low-density lipoprotein;

 
• MCR,

  metabolic clearance rate;

 
• OGTT,

  oral glucose tolerance test;

 
• PI-3 kinase,

  phosphatidylinositol 3 kinase;

 
• PPAR,

  peroxisome proliferator activated receptor;

 
• Ra,

  rate of endogenous glucose appearance;

 
• T2DM,

  type 2 diabetes mellitus;

 
• TGD,

  total glucose disposal;

 
• TZD,

  thiazolidinedione.