Complex Etiologies of the Discordance Between Cystatin C- and Creatinine-Based Estimated GFR and Its Adverse Associations: Findings From the CRIC Study.

Abstract

Rationale & objective: The difference between GFR estimated by cystatin C and creatinine (eGFRdiff, defined as eGFRCys - eGFRCr) has been repeatedly associated with adverse outcomes, often ascribed to low muscle mass. However, it is unclear to what extent putative determinants of eGFRdiff, such as low muscle mass, explain associations between eGFRdiff and the outcomes of death and heart failure. Determinants of eGFRdiff have not been investigated to assess impacts on eGFRCys, eGFRCr and ultimately eGFRdiff in a dataset with measured GFR.

Study design: Prospective cohort study for outcomes of eGFRdiff and cross-sectional study for determinants of eGFRdiff.

Setting & participants: 1,290 adult participants in the Chronic Renal Insufficiency Cohort (CRIC) Study with directly measured iothalamate GFR, 24-hour urine creatinine collections, and plasma measurements of larger molecules, such as beta-2-microglobulin.

Exposure(s): Baseline eGFRdiff (eGFRCys - eGFRCr) for prospective analysis; putative eGFRdiff determinants for cross-sectional analyses.

Outcome(s): Time to all-cause death, heart failure hospitalization for prospective analyses; eGFR for cross-sectional analyses.

Analytical approach: Cox proportional hazards regression for prospective analysis and linear regression for cross-sectional analyses.

Results: Among adults with CKD, more positive eGFRdiff was associated with decreased risk of death (aHR 0.80, 95% CI 0.74-0.88) and heart failure hospitalization (aHR 0.83, 95% CI 0.73-94). Neither association was substantially changed by adjustment for putative determinants of eGFRdiff. eGFRdiff was weakly correlated with markers of muscle mass, middle molecule clearance, and other putative determinants of eGFRdiff (e.g., obesity, inflammation). Only 36% of the variance in eGFRdiff was explained by these factors.

Limitations: Imprecision in measurements, such as 24-hour urine collections.

Conclusion: The associations of eGFRdiff with adverse outcomes were unchanged by adjustment for markers of putative determinants, such as muscle mass. Examined putative determinants of eGFRdiff accounted for a minority of the variance in eGFRdiff.