Geriatric assessment of estimated glomerular filtration rate: a cross-sectional study

Background Estimated glomerular filtration rate (eGFR) is routinely calculated based on the serum creatinine level. However, the validity of such calculation in the geriatric population has not been sufficiently assessed. To examine whether the discrepancies between the eGFR determined based on the serum creatinine (eGFRcr) and that based on the serum cystatin C (eGFRcys) may be influenced to a lesser degree, by factors such as aging and muscle mass. Methods We measured the cystatin C and creatinine levels in 19,764 subjects (mean 77.0 years) and the eGFRcys and eGFRcr using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), Japanese, and Berlin Invitation Study (BIS) equations were calculated. Results The mean measured eGFRcys and eGFRcr values by the CKD-EPI equation were 48.2 and 66.6 ml/min/1.73 m(2) body surface area, respectively. The correlation between the eGFRcr (x) and eGFRcys (y) was y = 0.728x (r = 0.867; p < 0.001). Analysis of the slope among all ages could be shown by the relation, eGFRcys = (0.43 + 0.33/(1 + 10(boolean AND)((82-age)* - 0.046)))*eGFRcr. The correlation between the eGFRcr and eGFRcys by the Japanese equation were also similar. However, when it was calculated by the BIS equation, no drop of the slope of the linear regression line was observed with age. Conclusions The eGFRcr was overestimated irrespective of whether the CKD-EPI or the Japanese equation was used. We could convert eGFRcr into eGFRcys by an equation using age. Estimation of eGFR including serum cystatin C was more accurate in elderly people.