The trajectory of glomerular filtration rate (GFR) in relation to glomerular hyperfiltration (GHF) has been unknown. It was evaluated retrospectively in 23,982 GHF-free health examinees who were followed for 2-10 yr (mean: 5.1 yr). GFR was estimated by the serum creatinine concentration, and GHF was defined as age- and sex-specific estimated GFR (eGFR) ≥ 95% of the Japanese general population. The temporal profile of eGFR was plotted in a GHF-centered way, which was fitted to a random coefficient linear mixed model. Of the 23,982 subjects, 797 and 23,185 subjects developed or did not develop GHF, respectively, so that they were termed as the GHF(+) and GHF(-) groups. At baseline, median eGFR was significantly elevated in the GHF(+) group compared with in the GHF(-) group: 94.1 versus 77.3 mL/min/1.73 m2 (P < 0.001). Elevation of basal eGFR lasted for a mean (SD) of 3.3 (1.9) yr in the GHF(+) group; mean eGFR then rose to the GHF range, which was 108.5 mL/min/1.73 m2. The eGFR decline after the peak was steeper in the GHF(+) group than in the GHF(-) group: -0.984 versus -0.497 mL/min/1.73 m2/yr (P < 0.001). Baseline eGFR, but no other variable, well predicted incident GHF, with an area under the receiver operating characteristic curve of 0.87 (95% confidence interval: 0.86-0.88). In conclusion, GHF occurs as a chronic, multiphasic phenomenon: initially with a sustained GFR elevation for years, followed by a GFR surge to the GHF range, which was accompanied by accelerated GFR declining.
Keywords: glomerular hyperfiltration; multiphasicity; renal failure; trajectory.