Obesity, Underweight, and Accuracy of eGFR Using Cystatin C and Creatinine in a Northern European Population

Key Points Indexed eGFR cr-cys outperformed eGFR cr or eGFR cys across the body mass index (BMI) spectrum, improving GFR classification and treatment decisions. Nonindexed eGFR cr-cys led to further but smaller improvement across the BMI range that was not consistently related to BMI category. These findings inform clinical decisions on how to estimate GFR in the large segment of society living with low or high extremes of BMI. Background The presence of a low or high body mass index (BMI) in patients may influence the accuracy of eGFR. This study evaluates the performance of eGFR equations across the range of BMI. Methods This is an observational study of 4707 adults (7503 repeated observations) referred for measured GFR (mGFR) in Stockholm, Sweden. We calculated indexed eGFR (in ml/min per 1.73 m 2 ) and nonindexed eGFR (in ml/min) with validated equations that use creatinine (eGFR cr ), cystatin C (eGFR cys ), or both (eGFR cr-cys ). We assessed equation performance against indexed and nonindexed mGFR across categories of BMI with median bias, P 30 (the percentage of estimated values within 30% of mGFR), and classification of GFR categories, and modeled the implications of choice of filtration marker and indexing on clinical decisions regarding dose adjustment or eligibility for treatment. Results The mean age (SD) was 57 (16) years (39% female), and the median (interquartile range) indexed and nonindexed mGFR were 59 (39–79) and 65 (42–87) ml/min, respectively. In total, 9% of participants were underweight (BMI <20 kg/m 2 ) and 18% were obese (BMI ≥30 kg/m 2 ). For indexed and nonindexed eGFR for all equations, eGFR cr overestimated mGFR at BMI <20 and ≥30 kg/m 2 , and eGFR cys underestimated mGFR at BMI ≥30 kg/m 2 . eGFR cr-cys had the least bias, acceptable P 30 , and the highest correct classification throughout the BMI range. In theoretical modeling, using indexed eGFR cr-cys versus eGFR cr would allow more accurate clinical decisions across all BMI categories. Using nonindexed versus indexed eGFR cr-cys led to further but smaller improvement that was not consistently related to BMI category for these decisions. Conclusions In a clinic population of northern European individuals referred for GFR measurement, indexed eGFR cr-cys was more accurate than indexed eGFR cr across the BMI spectrum. Using nonindexed eGFR cr-cys further improved accuracy for some treatment decisions.