Aim: This study aimed to investigate the relationship between fibroblast growth factor 21 (FGF-21) and left ventricular systolic dysfunction and cardiac death.

Methods: The optimal cutoff of FGF-21 for identifying left ventricular systolic dysfunction at baseline was analyzed via receiver operating characteristic (ROC) curves. The identification of different serum levels of FGF-21 and their association with cardiac death was analyzed via Kaplan-Meier survival curves.

Results: A total of 253 patients were recruited for this study at baseline. Patients were excluded if they lacked echocardiography or laboratory measurement data, and there were 218 patients enrolled in the final analysis. The average age was 66.32±10.10 years. The optimal cutoff values of FGF-21 and NT-pro-BNP for identifying left ventricular systolic dysfunction at baseline were 321.5 pg/ml and 131.3 ng/l, respectively, determined separately via ROC analysis. The areas under the curves were non-significant among FGF-21, NT-pro-BNP and FGF-21+NT-pro-BNP as determined by pairwise comparisons. Both a higher serum level of FGF-21 and a higher serum level of NT-pro-BNP were independent risk factors for left ventricular systolic dysfunction at baseline (odd ratio (OR) 3.138 [1.037-9.500], P=0.043, OR 9.207 [2.036-41.643], P=0.004, separately). Further Kaplan-Meier survival analysis indicated an association between both a higher serum level of FGF-21 and a higher serum level of NT-pro-BNP with cardiac death in 5 years (RR5.000 (1.326-18.861), P=0.026; RR 9.643 (2.596-35.825), P=0.009, respectively).

Conclusion: Serum FGF-21 level was significantly correlated with left ventricular systolic dysfunction at baseline. Patients with higher serum levels of FGF-21 tended to suffer greater risks of cardiac death than patients with lower serum levels of FGF-21. The identification of FGF-21 and its relationship with left ventricular systolic function and cardiac death were non-inferior to NT-pro-BNP.

Disclosure

Y. Bao: None. Y. Shen: None. X. Zhang: None. X. Pan: None. Y. Xu: None. Q. Xiong: None. Z. Lu: None. X. Ma: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.