ABSTRACT: Background Rosiglitazone not only improves insulin-sensitivity, but also exerts anti-inflammatory effects. We have now examined in type 2 diabetic patients if these effects are reflected by changes in mRNA expression in peripheral blood mononuclear cells (PBMCs) to see if these cells can be used to study these anti-inflammatory effects at the molecular level in vivo. Method Eleven obese type 2 diabetic patients received rosiglitazone (2x4 mg/d) for 8 weeks. Fasting blood samples were obtained before and after treatment. Ten obese control subjects served as reference group. The expression of NFkappaB-related genes and PPARgamma target genes in PBMCs, plasma TNFalpha, IL6, MCP1 and hsCRP concentrations were measured. In addition, blood samples were obtained after a hyperinsulinemic-euglycemic clamp. Results Rosiglitazone reduced plasma MCP1 and hsCRP concentrations in diabetic patients (-9.5 +/- 5.3 pg/mL, p=0.043 and -1.1 +/- 0.3 mg/L p=0.003), respectively). For hsCRP, the concentration became comparable with the non-diabetic reference group. However, of the 84 NFkappaB-related genes that were measured in PBMCs from type 2 diabetic subjects, only RELA, SLC20A1, INFgamma and IL1R1 changed significantly (p<0.05). In addition, PPARgamma and its target genes (CD36 and LPL) did not change. During the clamp, insulin reduced plasma MCP1 concentration in the diabetic and reference groups (-9.1 +/- 1.8%, p=0.001 and -11.1 +/- 4.1%, p=0.023, respectively) and increased IL6 concentration in the reference group only (23.5 +/- 9.0%, p=0.028). Conclusions In type 2 diabetic patients, the anti-inflammatory effect of rosiglitazone is not reflected by changes in NFkappaB and PPARgamma target genes in PBMCs in vivo. Furthermore, our results do not support that high insulin concentrations contribute to the pro-inflammatory profile in type 2 diabetic patients.