SHUN-FA YANGHORNG-RONG CHANG2026-10-162026-10-162016-01-15Pure ID: 15041071Pure UUID: a4de61f2-a079-4af7-a69e-881b343d576810.1016/j.cca.2015.11.019https://www.scopus.com/pages/publications/84949032299https://scholars.csmu.edu.tw/handle/123456789/27938Background: We evaluated the ability of matrix metalloproteinase (MMP)-2, MMP-9, myeloperoxidase, osteopontin and stromal cell-derived factor 1 to predict mortality in hemodialysis (HD) patients. Methods: One hundred forty HD patients were enrolled and followed from December 2007 until December 2012. At the end of this 5-year period, data were compared between the patients who were alive and those who had died. Results: The patients who alive were younger (56 vs. 63 y), with lower frequency of diabetes mellitus (34.34% vs 58.53%), higher concentrations of albumin (4.13 vs. 3.91. mg/dl) and lower concentrations of MMP-2 (430.76 vs 521.59 ng/ml). Multivariate analysis showed that age (HR = 1.03, p = 0.02), diabetes mellitus (HR = 2.395, p = 0.012), albumin (HR = 0.475, p = 0.047) and MMP-2 (HR = 1.003, p = 0.005) were independent factors predicting mortality in HD patients. Receiver operating characteristic curve analysis showed that albumin (AUC = 0.628, p = 0.027) and MMP-2 (AUC = 0.643, p = 0.004) had a similar ability (p = 0.76) to predict survival of HD patients. Conclusions: Compared with albumin, serum MMP-2 is a non-inferior prognostic marker for predicting the survival of HD patients.6enMMP-2 serum concentrations predict mortality in hemodialysis patientsA 5-year cohort studyjournal article