Introduction
Acute kidney injury (AKI) is common among endurance runners. Current practice to diagnose AKI when baseline serum values are unavailable uses an age-based estimation of glomerular filtration rate (GFR) and the Modification of Diet in Renal Disease (MDRD) equation for back-calculation of baseline creatinine (Cr). However, the accuracy of this practice in a cohort of healthy runners is unknown.
Objective
To evaluate the accuracy of the age-based GFR and MDRD equation in ultramarathoners.
Methods
A multisite prospective analysis from participants in the 155-mile (250-km) Racing the Planet multistage ultramarathons in the Sahara, Gobi, and Atacama Deserts. Baseline creatinine (Cr) measurements were taken prior to race start and compared with estimated Cr using age-based GFR and MDRD.
Results
Forty-eight study participants (27% women, age 39.3 ± 10.3 years) were analyzed with all racers combined as a single cohort. Age-based estimated baseline GFR was 95.9 ± 5.8, and the corresponding Cr was 0.88 ± 0.1 by MDRD. Measured Cr was 0.99 (± 0.2) at race start, with a corresponding baseline GFR calculated via the MDRD equation of 86.1 (± 14.6) and a candidate healthy population equation of 107.1 ± 9.3. Age-based GFR with back-calculation of Cr by MDRD underestimates measured Cr by 9% (P < .001), while back-calculation of Cr by a candidate healthy population overestimates measured Cr by 27% (P < .001). Correlation between estimated baseline GFR and MDRD-based GFR was low (r = .1). Correlation between estimated baseline GFR and the healthy population equation GFR was higher (r = .43).
Conclusions
Analysis of the accepted age-based GFR and MDRD equations resulted in an underestimation of baseline Cr, leading to an overestimation of subsequent AKI. Additional studies are needed to create a more accurate estimation for baseline renal function, and this method should consider using the healthy population equation for a more accurate estimation of disease.