Estimating GFR when creatinine lies

Clinical stem

A 60-year-old on tirzepatide has lost 18 kg over eight months. Serum creatinine has fallen from 1.1 to 0.8 mg/dL and the creatinine-based eGFR now reads higher, which is interpreted as “improving kidney function.”

The trap

Rapid weight loss strips muscle; less muscle generates less creatinine, so the creatinine-based eGFR rises even if true GFR is unchanged — falsely reassuring. The mirror image, sarcopenia masking established CKD, is just as common.

Diagnostic reasoning

Creatinine is a muscle-derived marker, and eGFRcr assumes an average muscle mass. When muscle mass is low or falling, eGFRcr overestimates true GFR. Cystatin C depends much less on muscle (though it is affected by adiposity, inflammation, and steroids). The CKD-EPI 2021 combined creatinine–cystatin C equation better approximates measured GFR in these patients (KDIGO 2024).

Safety pearl

When weight or muscle is changing quickly, or body habitus is extreme, confirm with cystatin C or the combined estimate before making drug-dosing or CKD-staging decisions — and consider measured GFR when the stakes are high.

What not to overclaim

No estimating equation is perfect in obesity, and cystatin C has its own confounders. Use the combined estimate and clinical context, not any single number in isolation.