Background: We aimed to investigate the associations between circulating endostatin and the different aspects of renal dysfunction, namely, estimated (cystatin C) glomerular filtration rate (GFR) and urine albumin-creatinine ratio (ACR). Methods: Two independent longitudinal community-based cohorts of elderly. ULSAM, n = 786 men; age 78 years; median GFR 74 ml/min/1.73 m2; median ACR 0.80 mg/mmol); and PIVUS, n = 815; age 75 years; 51% women; median GFR; 67 ml/min/1.73 m2; median ACR 1.39 mg/mmol. Cross-sectional associations between the endostatin levels and GFR as well as ACR, and longitudinal association between endostatin at baseline and incident CKD (defined as GFR <60 ml/min/1.73 m2) were assessed. Results: In cross-sectional regression analyses adjusting for age, gender, inflammation, and cardiovascular risk factors, serum endostatin was negatively associated with GFR (ULSAM: B-coefficient per SD increase -0.51, 95% CI (-0.57, -0.45), p < 0.001; PIVUS -0.47, 95% CI (-0.54, -0.41), p < 0.001) and positively associated with ACR (ULSAM: B-coefficient per SD increase 0.24, 95% CI (0.15, 0.32), p < 0.001; PIVUS 0.13, 95% CI (0.06-0.20), p < 0.001) in both cohorts. Moreover, in longitudinal multivariable analyses, higher endostatin levels were associated with increased risk for incident CKD defined as GFR <60 ml/min/1.73 m2 at re-investigations in both ULSAM (odds ratio per SD increase of endostatin 1.39 (95% CI 1.01-1.90) and PIVUS 1.68 (95% CI 1.36-2.07)). Conclusions: Higher circulatingendostatin is associated with lower GFR and higher albuminuria and independently predicts incident CKD in elderly subjects. Further studies are warranted to investigate the underlying mechanisms linking endostatin to kidney pathology, and to evaluate the clinical relevance of our findings. i 2014 S. Karger AG, Basel