Objective: The regulatory processes that modulate adiponectin production and the mechanisms involved in nuclear factor kB (NF-kB) transcriptional activity in human adipocytes are not yet fully known. The aim of our study was to evaluate the interrelationships between body fat, fat distribution, systemic inflammation, insulin resistance, leptin and the serum and subcutaneous adipose tissue gene expression levels of tumor necrosis factor-alpha (TNF-a), adiponectin and the inhibitor kappa B-alpha (IkB-a), in subjects with a wide range of body mass index (BMI). We also wanted to determine which of these variables was most closely related to adiponectin gene expression and adipocyte NF-kB transcriptional power. Methods: A total of 27 women aged between 50 and 80 years, with BMI ranging from 22.1 to 53.3 kg/m 2 , were studied. In all subjects BMI, waist circumference, body composition by dual X-ray absorptometry, triglycerides, cholesterol, high-density lipoprotein cholesterol (HDL-Ch), glucose, insulin, homeostasis model assessment of insulin resistance (HOMA), high-sensitive C-reactive protein (hs-CRP), serum adiponectin, leptin and TNF-a were evaluated. Subcutaneous adipose tissue biopsies were taken from the abdomen of all subjects and the mRNA levels of adiponectin, TNF-a and IkB-a were determined. Results: BMI and waist circumference were associated positively with leptin, HOMA, and hs-CRP, and negatively with HDL-Ch; waist was also associated with adiponectin and IkB-a mRNA. HOMA was negatively associated with serum adiponectin and adiponectin mRNA. Hs-CRP was negatively associated with IkB-a mRNA, and was positively associated with HOMA.Step-down multiple regression analysis was performed to determine the joint effects of BMI, waist circumference, triglycerides, HDL-Ch, HOMA, hs-CRP, leptin, serum and TNF-a mRNA on adiponectin gene expression: waist circumference and leptin were both included in the best fitting regression equation for predicting adiponectin gene expression (R 2 ¼ 0.403, P ¼ 0.006). Stepwise multiple regression analysis was performed, considering IkB-a mRNA as a dependent variable and BMI, waist, HDL-Ch, HOMA, hs-CRP and adiponectin mRNA as independent variables. Adiponectin mRNA was the only variable to enter the regression (R 2 ¼ 0.406, Po0.001). Conclusion: Our results suggest that abdominal adiposity and leptin are independent predictors of adiponectin gene expression and that in human adipocytes, adiponectin gene expression is strongly related to IkB-a mRNA.