BackgroundThe role of fibroblast growth factor 21 (FGF21) in predicting the long-term prognosis of patients with cardiovascular disease (CVD) remains unknown.MethodsA comprehensive search in PubMed, Embase, and the Cochrane Library was performed to identify studies reporting the association between FGF21 and prognosis among patients with CVD. A meta-analysis was performed, with patients stratified by coronary artery disease (CAD) or heart failure (HF). The endpoint of CAD or HF was major adverse cardiovascular events defined by each study and a composite of death or HF readmission, respectively. The I2 method and linear regression test of funnel plot asymmetry were used to test heterogeneity (I2 > 50% indicates substantial heterogeneity) and publication bias (asymmetry P < 0.05, indicating publication bias).ResultsA total of 807 records were retrieved, and nine studies were finally included. Higher FGF21 levels were significantly associated with the risk of major adverse cardiovascular events in patients with CAD (multivariate hazard ratio [HR]: 1.77, 95% confidence interval [CI]: 1.40–2.23, P < 0.05, I2 = 0%, fixed-effect model). Increased FGF21 levels were also associated with the risk of all-cause death among patients with CAD (multivariate HR: 2.67, 95% CI: 1.25–5.72, P < 0.05, I2 = 64%, random-effect model). No association was found between FGF21 and the endpoint among patients with HF (HR: 1.57, 95% CI: 0.99–2.48, P > 0.05, random-effect model), but a large heterogeneity (I2 = 95%) and potential publication bias (Asymmetry P < 0.05) existed in the analysis.ConclusionIncreased FGF21 levels were independently associated with poor prognosis of CAD, whereas the role of FGF21 in predicting clinical outcomes of HF requires further investigation.