Rosiglitazone Improves Postprandial Triglyceride and Free Fatty Acid Metabolism in Type 2 Diabetes

Nonsmoking men and nonfertile women aged 35–70 years with documented type 2 diabetes were considered eligible. Patients on insulin treatment were excluded. All patients were treated with oral antihyperglycemic agents, which continued during the study. Exclusion criteria were current or previous treatment with TZD, HbA_1c >9%, serum creatinin >200 μmol/l, abnormal thyrotropin, aspartate aminotransferase (AST), or alanine aminotransferase (ALT) ≥2 times the upper limit of normal, congestive cardiac failure, blood pressure >160/>95 mmHg, total cholesterol >8 mmol/l and/or triglycerides >5 mmol/l, and an alcohol intake >3 units/day. The study protocol was approved by the local research ethics committee of the University Medical Center Utrecht. All participants gave written informed consent.

The study was designed as a randomized, crossover, placebo-controlled, double-blind trial. Eligible patients were randomly assigned to receive rosiglitazone 4 mg b.i.d. or placebo in addition to their current oral antihyperglycemic agents for 8 weeks. A 6-week wash-out period was included between the two treatment periods. At the end of each treatment period, a standardized 6-h oral fat-loading test was carried out. At the beginning and at the end of each 8-week treatment period, anthropometric and fasting laboratory parameters were determined. Patients were instructed to fast at least 12 h before each visit. No study medication or other medication was used on the morning of the study days.

After placing a cannula for venous blood sampling, subjects rested for 30 min before administration of the fat load. Fresh cream (a 40% weight/volume fat emulsion representing a total energy content of 3,700 kcal/l) was ingested within 5 min at a dose of 50 g fat and 3.75 g glucose/m² body surface (9,16). Participants remained supine during each test and were only allowed to drink mineral water. Peripheral blood samples were obtained in sodium EDTA (2 mg/ml), kept on ice, and centrifuged immediately for 15 min at 800g at 4°C, then plasma was stored at −80°C. Lipoproteins were subfractionated by ultracentrifugation as described previously in detail (16,17). Consecutive runs were carried out to float Svedberg flotation rate (Sf) >400 (chylomicrons), 60–400 (VLDL1), 20–60 (VLDL2), 12–20 (IDL), and 2–12 (LDL).

Total cholesterol, HDL cholesterol obtained after precipitation with phosphotungstate/MgCl₂, and triglycerides were measured in duplicate by colorimetric assay with the CHOD-PAP and GPO-PAP kits, respectively (Roche Diagnostics, Mannheim, Germany). FFAs were measured by an enzymatic colorimetric method (Wako Chemicals, Neuss, Germany). For FFA measurement, a lipase inhibitor was added to the plasma in order to block ex vivo lipolysis. Total plasma apolipoprotein B (apoB) was measured by nephelometry using apoB monoclonal antibodies. Glucose, creatinin, serum ALT, and AST were measured by standard enzymatical laboratory methods. Insulin was measured by enzyme-linked immunosorbent assay (Mercodia, Uppsala, Sweden). For estimation of insulin sensitivity, the homeostasis model assessment (HOMA) (glucose × insulin/22.5) was calculated.

All values are expressed as the mean ± SE in the text, tables, and figures. The area under the curve (AUC) for triglycerides and FFAs was calculated by the trapezoidal rule using GraphPad Prism version 4.0. Incremental integrated AUCs (dAUC) were also calculated after correction for baseline values. Differences between rosiglitazone and placebo were analyzed by paired t test. During serial measurements, time effects when compared with t = 0 were tested using repeated-measures ANOVA with Bonferroni correction for multiple comparisons. Bivariate correlations were calculated using Spearman’s correlation coefficients. Calculations were performed using SPSS/PC +11.5 (SPSS, Chicago, IL). Statistical significance was taken at the 5% level. A reduction in triglyceride dAUC of 20% was considered clinically relevant. Power analysis with β = 0.10 and α = 0.05 revealed that 15 patients had to be included to find such a reduction. Since a small drop-out rate was expected, we aimed to include 20 patients.

In total, 22 diabetic patients were screened. Two patients were excluded after the screening visit because of abnormal thyrotropin and HbA_1c >9%. One patient withdrew informed consent during the study. General characteristics of the 19 remaining participants are listed in Table 1. All patients were using oral antihyperglycemic agents, which was unchanged during the study. Eight patients were treated for dyslipidemia with statins (four with simvastatin and four with atorvastatin). ACE inhibitors (n = 3), a β-blocking agent (n = 1), a calcium antagonist (n = 1), and a diuretic (n = 1) were used in six patients with hypertension. Rosiglitazone was well tolerated, and no patient showed significant side effects other than headache, dizziness, and gastrointestinal complaints (n = 3). Rosiglitazone significantly reduced ALT compared with placebo (36 ± 2 vs. 42 ± 4 units/l, P < 0.05). Significant decreases in hemoglobin and hematocrit were also observed after treatment with rosiglitazone (data not shown).

Rosiglitazone significantly decreased fasting plasma glucose (6.2 ± 0.3 vs. 7.2 ± 0.5, P < 0.01) and HOMA (2.06 ± 0.35 vs. 3.76 ± 0.50, P < 0.01) compared with placebo but did not change HbA_1c (6.2 ± 0.6 vs. 6.3 ± 0.7%). The effects of rosiglitazone and placebo on fasting lipoprotein profile are listed in Table 2. Rosiglitazone increased total cholesterol due to a significant increase in LDL cholesterol, leading to an increased non-HDL cholesterol compared with placebo.

Rosiglitzone did not change fasting triglycerides in plasma or in the chylomicron VLDL1 and VLDL2 fractions compared with placebo. After rosiglitazone, the postprandial triglyceride increase in plasma was lower compared with placebo (Fig. 1), which resulted in a significant lower triglyceride dAUC (Table 2). Total plasma triglyceride AUC was not different between rosiglitazone and placebo. A significant reduction in the postprandial triglycerides content of the chylomicron fraction was achieved with rosiglitazone, which resulted in a significant lower triglyceride AUC (−22%, P < 0.05) in this fraction. The postprandial triglycerides rise in VLDL1 was also lower after rosiglitazone (−27%, P < 0.05), but this did not result in a significant lower triglyceride AUC. In VLDL2, there were no significant differences in triglyceride AUC and triglyceride dAUC between rosiglitazone and placebo. Fasting FFA levels tended to be lower after treatment with rosiglitazone compared with placebo (0.35 ± 0.03 vs. 0.41 ± 0.03 mmol/l, respectively, P = 0.06). Rosiglitazone significantly decreased FFA dAUC (−18%, P < 0.05) compared with placebo, leading to significantly lower FFA AUC (−12%, P < 0.05). The reductions in FFA AUC and triglyceride dAUC upon treatment with rosiglitazone were not related to the reduction of fasting plasma glucose (r = 0.14 and r = 0.07, respectively), but they were significantly related to the reduction of ALT (r = 0.53 and r = 0.42, respectively, P < 0.05). A subgroup analysis of subjects on (n = 8) and off (n = 11) statin treatment showed similar effects of rosiglitazone on plasma triglyceride dAUC (−42 and −34%, respectively) and FFA AUC (−14 and −11%, respectively).

The main finding of the present study is that rosiglitazone improves postprandial triglyceride metabolism in patients with type 2 diabetes. Because humans are nonfasting most of the day and nonfasting triglycerides are strong predictors of atherosclerosis (2–9), this may convey increased protection from cardiovascular disease in these patients. Significant effects on triglyceride clearance were found in the chylomicron and VLDL1 fractions, suggesting a preferential action of rosiglitazone on large TRL. Finally, rosiglitazone significantly decreased postprandial FFA concentrations.

Increased postprandial lipemia is a common feature of diabetic dyslipidemia and is associated with accelerated atherosclerosis, even in fasting normolipidemic subjects (1–9). The increase of large VLDL1 particles, in particular, is associated with the generation of atherogenic remnants (1). In one study (2), postprandial triglycerides levels distinguished between case subjects with future myocardial infarction and control subjects even better than fasting plasma triglycerides levels. The data from the present study show that rosiglitazone does not change fasting triglycerides but decreases the postprandial triglyceride increase in plasma (−37%), chylomicrons (−20%), and VLDL1 (−27%). Although this did not lead to a significantly lower total triglyceride AUC in plasma, the total triglyceride AUC in the chylomicron fraction significantly decreased by 22%. Whether these effects are sufficient to produce clinical benefit is an open issue. However, it is tempting to hypothesize that the antiatherosclerotic effects observed with TZDs may involve the improvement of postprandial lipemia (18–20). Another interesting aspect of the study is the fact that the effects of rosiglitazone were found on top of statin treatment. In our opinion, this is of interest because statins are used by a majority of patients with type 2 diabetes, and they have been shown to improve postprandial triglyceride metabolism. Whether further improvement of triglyceride metabolism by rosiglitazone exerts additional vascular benefit in diabetic patients on statin treatment remains to be shown.

Data on the effects of TZDs on postprandial lipemia are scarce. In one study (21), there was a lack of effect of pioglitazone on postprandial triglyceride levels. However, in that study, the postprandial triglyceride level was measured only 2 h after a conventional breakfast. Our study was performed in a metabolic ward setting using an oral fat load, and triglyceride levels were measured at 2-h intervals up to 6 h postprandially. The present study is the first to show improved postprandial triglyceride metabolism by rosiglitazone. Whether rosiglitazone also improves postprandial lipemia during insulin action, such as after a mixed meal, remains to be investigated in future studies.

In most intervention studies with lipid-lowering drugs, the reduction in postprandial lipemia is more or less similar to the reduction in fasting plasma triglycerides, which are the main determinants of postprandial lipemia (22). To the best of our knowledge, this is the first study to describe improved postprandial triglyceride metabolism without lowering fasting triglyceride levels. These results suggest other mechanisms for improved postprandial lipemia than reduced competition for the common lipolytic pathway. First, TZDs are able to upregulate adipocyte LPL production through activation of PPAR-γ (14,15), which could have contributed to the improved incremental triglyceride response after rosiglitazone. Unfortunately, LPL mass and activity were not measured in the present study. Second, it has been demonstrated that rosiglitazone directly stimulates the expression and function of lipoprotein receptor–related protein in vitro (23). Third, rosiglitazone significantly decreased postprandial FFA concentrations, probably by increasing adipocyte FFA trapping, thereby decreasing the source of hepatic VLDL production (12,24–26). Finally, improvement of glycemic control may translate into improvement of postprandial lipemia, especially in patients with poor glycemic control. For example, in patients with poor glycemic control, it has been shown that metformin (27) and glipizide (28) improve postprandial lipemia. In contrast, in diabetic patients with good glycemic control, nateglinide and glibenclamide attenuated hyperglycemia, but they did not improve postprandial lipemia (29). Patients in our study had good glycemic control (HbA_1c 6.2%). Rosiglitazone decreased fasting plasma glucose, but not HbA_1c, and improved postprandial triglyceride metabolism. Improvement of postprandial lipemia upon treatment with rosiglitazone was not related to the decrease in fasting plasma glucose. Therefore, we propose that it is unlikely that the beneficial effects of rosiglitazone on postprandial lipemia are due solely to improved glycemic control, although it may have contributed.

It has been shown that in the postprandial phase, when chylomicrons and VLDL compete for clearance by LPL, the former are hydrolyzed preferentially (30). This could partly explain the greater beneficial effects of rosiglitazone on postprandial triglyceride clearance in large TRL (chylomicrons and VLDL1) compared with small TRL (VLDL2). Alternatively or additionally, it has been suggested that hepatic VLDL1 and VLDL2 production are independently regulated (31,32). Acute insulin administration suppresses the rates of VLDL1 apoB production but has no effect on VLDL2 apoB production (31). Hence, increased insulin sensitivity by rosiglitazone may have resulted in the suppressed postprandial production of VLDL1 compared with VLDL2, as suggested by our results.

Insulin resistance is commonly associated with biochemical evidence of nonalcoholic steatohepatitis, such as increases in ALT (33). Treatment with rosiglitazone improved insulin sensitivity and significantly reduced ALT. Interestingly, the reduction of ALT upon treatment with rosiglitazone was related to the decreases in FFA AUC and triglyceride dAUC, suggesting reduced liver fat content due to preferential adipocyte FFA storage (33,34).

In conclusion, rosiglitazone improves the metabolism of large TRL and decreases postprandial FFA concentrations in type 2 diabetes. This may have clinical implications, as these effects may contribute to cardiovascular risk reduction.

This study was financially supported by a grant from GlaxoSmithKline.

We dedicate this study to the memory of our mentor Prof. Dr. D.W. Erkelens.