Table 1.

Clinical evidence on the association of circulating FGF21 levels with HF

Reference	Population	Key findings	Limitations
HFpEF
Chou et al. [27]	n = 238 •95 HFpEF patients •143 controls Patients followed up for 1 year to assess prognostic value of FGF21	•Elevated FGF21 levels were significantly associated with HFpEF •Higher baseline FGF21 levels predicted a higher mortality risk at 1 year	•Cross-sectional study design of FGF21 levels between HFpEF and controls which precludes causality •Small sample size •Control and HFpEF were poorly matched with traditional CVD risk factors more prevalent in the HFpEF group. In particular, HFpEF patients were significantly older •No data on pre-existing liver disease among patients •Single ethnic cohort (Chinese)
Ianos et al. [22]	n = 69 •40 HFpEF patients •29 controls Various CVDs were excluded	•Elevated FGF21 levels were associated with HFpEF	•Type 2 diabetic cohort •No data on pre-existing liver disease Single ethnic cohort (Caucasian)
HFrEF
Sommakia et al. [109]	n = 60 •40 HFrEF patients •20 controls Patients with non-alcoholic fatty liver disease were excluded	•FGF21 levels were elevated in patients with HFrEF compared to controls •Tissue sample analysis revealed the presence of FGF21 in diseased cardiomyocytes, with elevated FGF21 levels likely originating from the liver	•Small sample size •Cross-sectional study design •Presence of comorbidities among HFrEF patients may confound the findings •HFrEF patients were end stage; thus, it is unclear whether findings can be generalised to patients with early-stage HF
Fan L, et al. [108]	n = 199 •128 subjects with HFrEF •71 controls Performed follow-up with a mean follow-up time of 13.36 months to assess 1-year mortality and HF readmission events In contrast to previous studies, patients with liver diseases were excluded	•Serum FGF21 was elevated in HFrEF •HFrEF patients with higher FGF21 levels had higher risk of 1-year mortality and heart failure readmission	•Small sample size •No multivariable regression analysis. Data was not adjusted for confounding factors •Single ethnic cohort (Chinese) •All HFrEF patients were NYHA functional class III or IV indicating a highly skewed population
Holm et al. [107]	n = 57 •19 patients with HFrEF and cardiac cachexia •19 patients with HFrEF and no cardiac cachexia •19 patients with IHD	•FGF21 levels were elevated in HFrEF and cardiac cachexia, as compared to the other two groups •Higher FGF21 levels were associated with higher interleukin 6 and lower muscle mass, but not cardiac function	•Cross-sectional study design •Small sample size •No data on pre-existing liver disease •Single ethnic cohort (Caucasian)
Gu et al. [20]	n = 321 •241 dilated cardiomyopathy patients •80 controls Patients followed up for a mean of 16 months. All-cause mortality and readmission were considered the end points Patients with liver disease were excluded	•Higher Serum FGF21 levels were independently associated with prognosis and severity of dilated cardiomyopathy •FGF21 levels were positively correlated with NYHA HF classification •FGF21 levels were inversely correlated with left ventricular ejection fraction in this study	•Cross-sectional study design assessing difference in FGF21 levels between groups •Patients with diabetes were not excluded •Single ethnic cohort (Chinese)