Background: Elevated fibroblast growth factor 21 (FGF21) levels have been suggested, from cross-sectional studies, as an indicator of subclinical diabetic nephropathy. We investigated whether serum FGF21 was predictive of the development of diabetic nephropathy.
Method: Baseline serum FGF21 levels were measured in 1136 Chinese type 2 diabetic subjects recruited from the Hong Kong West Diabetes Registry. The role of serum FGF21 in predicting decline in estimated glomerular filtration rate (eGFR) over a median follow-up of 4 years was analyzed using Cox regression analysis.
Results: At baseline, serum FGF21 levels increased progressively with eGFR category (P for trend <.001). Among 1071 subjects with baseline eGFR ≥ 30 mL/min/1.73 m(2), serum FGF21 levels were significantly higher in those with eGFR decline during follow-up (n = 171) than those without decline (n = 900) (P < .001). In multivariable Cox regression analysis, baseline serum FGF21 was independently associated with eGFR decline (hazard ratio, 1.21; 95% confidence interval [CI], 1.01-1.43; P = .036), even after adjustment for baseline eGFR. In a subgroup of 559 subjects with baseline eGFR ≥ 60 mL/min/1.73 m(2) and normoalbuminuria, serum FGF21 level remained an independent predictor of eGFR decline (hazard ratio, 1.36; 95% CI, 1.06-1.76; P = .016). Integrated discrimination improvement (IDI) suggested that the inclusion of baseline serum FGF21 significantly improved the prediction of eGFR decline (IDI, 1%; 95% CI, 0.1-3.0; P = .013) in this subgroup, but not in the initial cohort involving all subjects.
Conclusions: Elevated serum FGF21 levels may be a useful biomarker for predicting kidney disease progression, especially in the early stages of diabetic nephropathy.