Background: Visfatin, also known as pre-B-cell colony-enhancing factor 1 (PBEF-1), recently was shown to be secreted from adipocytes and have insulin-mimetic properties in mice. Because renal failure per se is associated with both increased levels of circulating peptides and marked insulin resistance, even in the absence of diabetes mellitus, we hypothesized that visfatin could be a factor linking inflammation, kidney disease, and insulin resistance in this patient group.

Methods: Altogether, we studied 189 patients with chronic kidney disease (CKD), comprising 149 patients with CKD stage 5 (glomerular filtration rate [GFR] < 15 mL/min; mean, 7 +/- 2 mL/min [<0.25 mL/s; mean, 0.12 +/- 0.03 mL/s]; 61% men; mean age, 54 +/- 12 years) and 40 patients with CKD stages 3 to 4 (GFR, 15 to 60 mL/min; mean, 33 +/- 21 mL/min [0.25 to 1.00 mL/s; mean, 0.55 +/- 0.35 mL/s]; 72% men; age, 59 +/- 15 years). We compared these with 30 randomly selected population controls (mean GFR, 85 +/- 16 mL/min [1.42 +/- 0.27 mL/s]; 69% men; age, 64 +/- 11 years). Serum visfatin was measured by using commercially available enzyme-linked immunosorbent assay, and we also performed genotyping of 3 verified polymorphisms in the visfatin gene (-423A/G, -1001T/G, and -1535C/T). Body fat was estimated by using dual-energy x-ray absorptiometry.

Results: Serum visfatin levels were greater in patients with CKD stage 5 (41.3 +/- 18.0 ng/mL) than in those with CKD stages 3 to 4 (34.0 +/- 9.8 ng/mL; P < 0.01 versus CKD stage 5) or healthy controls (29.3 +/- 8.1 ng/mL; P < 0.0001). However, there were no significant differences between patients with and without diabetes, and the significant differences in circulating visfatin levels between genotypes disappeared after adjustment for differences in age, sex, GFR, and serum albumin level. In univariate analysis, visfatin level correlated with levels of GFR (rho = -0.22; P = 0.001), interleukin 6 (IL-6; rho = 0.17; P = 0.01), high-sensitivity C-reactive protein (rho = 0.14; rho < 0.05), and soluble vascular cell adhesion molecule 1 (sVCAM-1; rho = 0.39; P < 0.0001), but not total or truncal fat mass, insulin resistance, or hemoglobin A(1c) level. High plasma visfatin level predicted mortality in patients with CKD, also after adjustment for age and sex (likelihood ratio, 18.2; P < 0.0001), but not after additional correction for GFR, sVCAM-1, serum albumin, and serum IL-6 levels.

Conclusion: Circulating levels of the cytokine visfatin/PBEF-1 are influenced by renal function, but are not associated with fat mass or surrogate markers of insulin resistance in patients with CKD. Visfatin was associated independently with level of sVCAM-1, a marker of endothelial damage.