Here’s a common scenario: A patient comes back to the office a year into taking semaglutide. She has lost about 25 lb on the drug, and her creatinine-based estimated glomerular filtration rate (eGFR) has climbed from 54 mL/min/1.73 m² to 61 mL/min/1.73 m².

On paper, she is out of stage III chronic kidney disease (CKD). But is that because her kidneys improved, or because she lost muscle and produced less creatinine, which pushed the estimate up with no change in filtration?

That second possibility is a common worry as the use of GLP-1 receptor agonists spreads through primary care. Two randomized trial analyses published this year tried to answer the question, and neither found the muscle-driven inflation clinicians feared.

In a prespecified analysis of the SMART trial, 101 adults with CKD and overweight or obesity but not type 2 diabetes received 2.4 mg of semaglutide weekly or placebo for 24 weeks. Participants on semaglutide lost about 9 kg more than those on placebo, with no significant difference in lean mass. Body composition changes did not correlate with changes in creatinine-based eGFR, cystatin C-based eGFR, or measured GFR. The authors concluded that weight reductions of about 10% with semaglutide do not influence GFR estimates.

A post hoc analysis of a Danish trial found a different result. Among 48 patients with type 2 diabetes and albuminuria — all on empagliflozin — the use of 1 mg per week of semaglutide for 26 weeks produced a small rise in plasma creatinine without a significant change in the cystatin C level or measured GFR. Weight loss was minimal and did not differ significantly between the groups. The combined creatinine-cystatin C equation tracked the measured GFR more closely than either marker alone, the researchers reported.

That creatinine rise runs opposite to the muscle-loss worry. Muscle loss, if anything, should make eGFR look better than true kidney function, said Nishtha Modi, MD, board-certified internist in New York City who focuses on obesity and wellness. “So if the creatinine is actually rising, I wouldn’t simply dismiss that as an artifact of weight loss,” Modi said.

Calogero DiMaggio, MD, primary care physician in New York City, said he would take the opening patient’s higher eGFR at face value because GLP-1 agonists “have renal-protective benefits,” he said. He would hesitate, however, before restarting a previously held drug without more data.

When to Order Cystatin C

The Kidney Disease: Improving Global Outcomes (KDIGO) 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease recommended the combined creatinine-cystatin C equation when the creatinine estimate is less accurate and GFR affects clinical decision-making. Its stepwise approach starts with creatinine, adds cystatin C, and moves to measured GFR when more accuracy is needed.

Before ordering cystatin C, Modi said she typically repeats the serum creatinine and checks for common causes of a change, namely angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, SGLT2 inhibitors, diuretics, nonsteroidal anti-inflammatory drugs, and volume contraction from nausea, vomiting, or reduced oral intake associated with use of GLP-1s.

The next question is whether a prescribing threshold is at stake. Metformin is the classic primary care case. Initiation is not recommended at an eGFR of 30-45 mL/min/1.73 m², and the drug is contraindicated at an eGFR < 30 mL/min/1.73 m². For current users whose eGFR is < 45 mL/min/1.73 m², the label calls for weighing the benefit against the risk. When a patient has lost substantial weight and the eGFR is nearing a threshold that will change management, “I want to be confident that I’m acting on a reliable estimate,” Modi said.

Anticoagulants are the exception: Direct oral anticoagulant labels don’t dose by eGFR. In atrial fibrillation, the apixaban dose is reduced to 2.5 mg twice daily only when two of three criteria are met: age 80 years or older, weight ≤ 60 kg, or serum creatinine ≥ 1.5 mg/dL. Weight loss alone won’t trigger it. Rivaroxaban, dabigatran, and edoxaban are dosed by creatinine clearance, usually calculated using the weight-based Cockcroft-Gault equation, so weight loss can lower the result with no change in filtration.

When those checks still leave a consequential decision resting on a creatinine estimate that may be uncertain, the guideline calls for measuring cystatin C and using the combined creatinine-cystatin C equation.

Discordance between the markers is common. In a 2025 meta-analysis of 821,327 outpatients, 11% had a cystatin C-based eGFR at least 30% lower than their creatinine estimate. A gap that size can change the clinical picture. For the patient in the opening scenario, a cystatin C-based eGFR 30% lower than her creatinine estimate of 61 mL/min/1.73 m² would be about 43 mL/min/1.73 m²; that puts her back in stage III CKD.

Cystatin C has its own blind spots. KDIGO listed higher adiposity, smoking, thyroid dysfunction, glucocorticoid excess, and chronic inflammation as factors other than kidney function that likely skew cystatin C levels. In some health systems, Modi said, “it may be a send-out test,” with a turnaround “considerably longer than a serum creatinine.” KDIGO noted that its reagents also cost more.

What the Trials Don’t Answer

Both trials were small, short, and tightly controlled. The weight loss seen in SMART — about 10% — fell short of what higher-dose tirzepatide or bariatric surgery produces.

The Danish cohort was older, mostly male, and entirely on an SGLT2 inhibitor, with no body composition data. Neither trial was designed specifically for patients with frailty or sarcopenia, for whom creatinine-based eGFR is the least reliable, according to the KDIGO recommendations.

Even so, the Danish investigators wrote that clinicians should “consider utilizing” the combined equation “with routine cystatin C measurements” for older adults with type 2 diabetes starting semaglutide.

DiMaggio said he already orders it “pretty regularly” for patients with stage III CKD and beyond, especially older patients. “It’s helpful to track two numbers in these patients to accurately stage their CKD and trend the overall kidney function,” he said. The test has, at times, changed how quickly he refers patients to nephrology.

Modi was more selective. “The question isn’t whether cystatin C can provide a better estimate. We know that it can,” she said. “The question is whether it’s likely to change what I do for that particular patient.”

Modi and DiMaggio reported no relevant financial conflicts of interest.

Yonathan Daniel, MD, is a third-year internal medicine resident at NYU Grossman School of Medicine in New York City and a graduate of Brown University in Providence, Rhode Island. He enjoys medical journalism and has previously contributed to the ABC News Medical Unit.