Effect of Steroids on Cystatin C and Estimated Glomerular Filtration Rate in Hospitalized Patients

Introduction: Serum creatinine (SCr) is the mainstay of kidney function assessment in the critically ill, but the serum concentration is affected by several non-kidney determinants not well accounted for in glomerular filtration rate estimation equations (eGFR). Cystatin C (cysC), an alternative kidney biomarker used for GFR estimation, may also be affected by non-kidney factors, including use of corticosteroids. Study purpose was to evaluate the impact of steroid exposure on cysC in acutely ill adults.


Methods: This was a single-center, retrospective, observational cohort study at a large health system from 2011-2023 of acutely ill adults with at least two concurrent SCr and cysC pairs (i.e., tests drawn within 24-hours of each other) and exposure to systemic corticosteroids during hospitalization. Eligible patients had a SCr and cysC pair within 24 hours before steroid initiation and another pair within 72-hours after steroid initiation. Patients with acute kidney injury, need for kidney replacement therapy, kidney transplantation, pregnancy, incarceration, or refusal to authorize their records to be used for medical research were excluded. The primary outcome was change in cysC before and after steroid initiation. Descriptive statistics were used to compare the pre-steroid/post-steroid within group change in cysC to SCr.


Results: 229 patients were included: mean ± SD age of 67 ± 16 years, 128 (55.9%) female. SCr and cysC were drawn median (IQR) 6.3 (3.2, 11.2) hours pre-steroid and 19.8 (13.5, 38.0) hours post-steroid. Steroid dose was 50 (25, 53) mg prednisone equivalents per day. Median cysC was 1.62 (1.20, 2.18) mg/L pre-steroid and 1.68 (1.20, 2.33) mg/L post-steroid, p=0.21. Median SCr was 1.10 (0.82, 1.40) mg/dL pre-steroid and 1.13 (0.80, 1.60) mg/dL post-steroid, p=0.089. Degree of difference in change from pre to post steroid was not clinically significant when cysC was compared to SCr (cysC change median 0.02 (IQR -0.17, 0.22) mg/L vs SCr change median 0.03 (IQR -0.11, 0.19) mg/dL; P = 0.035).


Conclusions: Initiation of steroids in acutely ill patients did not significantly alter cysC concentrations in this large observational study. Findings suggest that routine doses of corticosteroids used in hospitalized patients are unlikely to significantly alter eGFR interpretation when based on cysC.