Cystatin C and Creatinine Concentrations Are Uninformative Biomarkers of Sarcopenia: A Cross-Sectional NHANES Study

Objectives Differences in creatinine and cystatin C-based estimates of glomerular filtration rate (eGFRDiff = eGFRCr – eGFRCysC) may reflect differences in muscle mass. We sought to determine if eGFRDiff (1) reflects lean mass, (2) identifies sarcopenic individuals beyond estimates based on age, body mass index (BMI), and sex; and (3) demonstrates associations differently in those with and without chronic kidney disease (CKD). Design and Methods This cross-sectional study included 3,754 participants, ages 20-85 years, with creatinine and cystatin C concentration levels, and dual-energy X-ray absorptiometry scans from National Health and Nutrition Examination Survey data (1999-2006). Dual-energy X-ray absorptiometry-derived appendicular lean mass index (ALMI) estimated muscle mass. Non-race-based CKD Epidemiology Collaboration equations estimated glomerular filtration rate using eGFRCr, eGFRCysC, and both biomarkers (eGFRCysC&Cr). CKD was defined as eGFRCysC&Cr <60 mL/minute/1.73 m2. ALMI sex-specific T-scores (compared with young adult) < −2.0 defined sarcopenia. In estimating ALMI, we compared the coefficient of determination (R2) values from: 1) eGFRDiff, 2) clinical characteristics (age, BMI, and sex), and 3) clinical characteristics plus eGFRDiff. Using logistic regression, we evaluated each model's C-statistic to diagnose sarcopenia. Results eGFRDiff was negatively and weakly associated with ALMI (No CKD: R2 = 0.006, p-value 0.002; CKD: R2 = 0.001, P value .9). Clinical characteristics explained most of the variation in ALMI (No CKD: R2 = 0.851, CKD: R2 = 0.828), and provided strong discrimination of sarcopenia (No CKD C-statistic: 0.950; CKD C-statistic: 0.943). Adding eGFRDiff improved the R2 by 0.025, and the C-statistic by 0.003. Tests for interaction between eGFRDiff and CKD were not significant (all P values > .05). Conclusions Although eGFRDiff has statistically significant associations with ALMI and sarcopenia in univariate analyses, multivariate analyses demonstrate that eGFRDiff does not capture more information beyond routine clinical characteristics (age, BMI, and sex). Differences in creatinine and cystatin C-based estimates of glomerular filtration rate (eGFRDiff = eGFRCr – eGFRCysC) may reflect differences in muscle mass. We sought to determine if eGFRDiff (1) reflects lean mass, (2) identifies sarcopenic individuals beyond estimates based on age, body mass index (BMI), and sex; and (3) demonstrates associations differently in those with and without chronic kidney disease (CKD). This cross-sectional study included 3,754 participants, ages 20-85 years, with creatinine and cystatin C concentration levels, and dual-energy X-ray absorptiometry scans from National Health and Nutrition Examination Survey data (1999-2006). Dual-energy X-ray absorptiometry-derived appendicular lean mass index (ALMI) estimated muscle mass. Non-race-based CKD Epidemiology Collaboration equations estimated glomerular filtration rate using eGFRCr, eGFRCysC, and both biomarkers (eGFRCysC&Cr). CKD was defined as eGFRCysC&Cr <60 mL/minute/1.73 m2. ALMI sex-specific T-scores (compared with young adult) < −2.0 defined sarcopenia. In estimating ALMI, we compared the coefficient of determination (R2) values from: 1) eGFRDiff, 2) clinical characteristics (age, BMI, and sex), and 3) clinical characteristics plus eGFRDiff. Using logistic regression, we evaluated each model's C-statistic to diagnose sarcopenia. eGFRDiff was negatively and weakly associated with ALMI (No CKD: R2 = 0.006, p-value 0.002; CKD: R2 = 0.001, P value .9). Clinical characteristics explained most of the variation in ALMI (No CKD: R2 = 0.851, CKD: R2 = 0.828), and provided strong discrimination of sarcopenia (No CKD C-statistic: 0.950; CKD C-statistic: 0.943). Adding eGFRDiff improved the R2 by 0.025, and the C-statistic by 0.003. Tests for interaction between eGFRDiff and CKD were not significant (all P values > .05). Although eGFRDiff has statistically significant associations with ALMI and sarcopenia in univariate analyses, multivariate analyses demonstrate that eGFRDiff does not capture more information beyond routine clinical characteristics (age, BMI, and sex).