Background: Animal models implicate FGF-23 (fibroblast growth factor-23) as a direct contributor to adverse cardiorenal interactions such as sodium avidity, diuretic resistance, and neurohormonal activation, but this has not been conclusively demonstrated in humans. Therefore, we aimed to evaluate whether FGF-23 is associated with parameters of cardiorenal dysfunction in humans with heart failure, independent of confounding factors.
Methods: One hundred ninety-nine outpatients with heart failure undergoing diuretic treatment at the Yale Transitional Care Center were enrolled and underwent blood collection, and urine sampling before and after diuretics.
Results: FGF-23 was associated with several metrics of disease severity such as higher home loop diuretic dose and NT-proBNP (N-terminal pro-B-type natriuretic peptide), and lower estimated glomerular filtration rate, serum chloride, and serum albumin. Multivariable analysis demonstrated no statistically significant association between FGF-23 and sodium avidity measured by fractional excretion of sodium, or proximal or distal tubular sodium reabsorption, either before diuretic administration or at peak diuresis (P≥0.11 for all). Likewise, FGF-23 was not independently associated with parameters of diuretic resistance (diuretic excretion, cumulative urine and sodium output, and loop diuretic efficiency [P≥0.33 for all]) or neurohormonal activation (plasma or urine renin [P≥0.36 for all]). Moreover, the upper boundary of the 95% CI of all the partial correlations were ≤0.30, supporting the lack of meaningful correlations. FGF-23 was not associated with mortality in multivariable analysis (P=0.44).
Conclusions: FGF-23 was not meaningfully associated with any cardiorenal parameter in patients with heart failure. While our methods cannot rule out a small effect, FGF-23 is unlikely to be a primary driver of cardiorenal interactions.
Keywords: chloride; diuretic; heart failure; models, animal; outpatient.