Medications Influencing Serum Cystatin C Independent of Measured GFR

KEY POINTS: Use of glucocorticoids, opioids, and loop diuretics raised cystatin C independent of GFR. Except for glucocorticoids, cystatin C increases were small and likely clinically nonmeaningful. Consequently, the underestimation of eGFR cys was larger in medication users; combined eGFR cr-cys was most accurate. BACKGROUND: Serum cystatin C is influenced by factors beyond GFR, and several medications have been hypothesized to alter its generation. Whether reported associations reflect effects on cystatin C independent of measured GFR (mGFR) or confounding by illness remains unclear. METHODS: We studied 5595 adults who underwent outpatient cystatin C (and creatinine) testing with mGFR determination using iohexol clearance in Stockholm health care. Associations between eight candidate medication classes and cystatin C were assessed using both interindividual and intraindividual analyses, sequentially adjusting for mGFR, clinical variables, and ongoing medications. We further evaluated the effect of these medications on cystatin C-based eGFR (eGFR cys ) bias compared with serum creatinine-based (eGFR cr ) and combined (eGFR cr-cys ) equations. RESULTS: After adjustment for mGFR and confounders, the use of systemic glucocorticoids, opioids, and loop diuretics was associated with higher cystatin C. The magnitude of increase was potentially clinically meaningful for glucocorticoid use, particularly for greater estimated daily dose categories, whereas the smaller increases observed with opioid and loop diuretic use likely reflect clinically nonmeaningful effects. No associations were observed for use of allopurinol, immunosuppressants, thyroid hormone analogues, or antipsychotics or for paracetamol after adjustment for concomitant medications. eGFR cys underestimated mGFR more in users than in nonusers, while eGFR cr overestimated mGFR similarly in users and nonusers. eGFR cr-cys provided the least biased GFR estimates. CONCLUSIONS: This study suggests that systemic glucocorticoids, opioids, and loop diuretics may increase cystatin C independent of mGFR. The use of eGFR cr-cys was consistently more accurate than either eGFR cr or eGFR cys alone in patients using these treatments.