Cystatin C: The Biomarker That May Be Better Than Creatinine
Urinary Peptides and Kidney Biomarkers
23% better reclassification
In a meta-analysis of 90,750 people, cystatin C reclassified death risk 23% more accurately than creatinine-based GFR estimation alone.
Shlipak et al., New England Journal of Medicine, 2013
Shlipak et al., New England Journal of Medicine, 2013
View as imageA blood test measures your kidney function. The number it returns shapes clinical decisions: medication dosing, surgical clearance, referral to nephrology, eligibility for kidney transplant. For decades, that test has been serum creatinine. The problem is that creatinine is a byproduct of muscle metabolism, and muscle mass varies enormously between an 80-year-old woman and a 25-year-old weightlifter. A 13.3 kDa cysteine protease inhibitor called cystatin C sidesteps this problem entirely. Produced at a constant rate by every nucleated cell in the body, cystatin C reflects glomerular filtration rate (GFR) without the confounding variables that make creatinine unreliable in large segments of the population. As part of the broader shift toward peptide-based kidney biomarkers, cystatin C has moved from research curiosity to clinical standard. The 2021 CKD-EPI equations from the National Kidney Foundation and the American Society of Nephrology now incorporate cystatin C alongside creatinine, and the combined equation is the most accurate GFR estimate available in clinical medicine.
Key Takeaways
- A meta-analysis of 90,750 individuals found cystatin C reclassified death risk 23% more accurately than creatinine and end-stage renal disease risk 10% more accurately (Shlipak et al., NEJM, 2013)
- Cystatin C production is unaffected by muscle mass, sex, diet, or race, eliminating the primary confounders that plague creatinine-based GFR estimation
- The 2021 CKD-EPI equations removed race coefficients from GFR calculation, relying on cystatin C to improve accuracy across racial and ethnic groups (Inker et al., NEJM, 2021)
- Combined creatinine-cystatin C equations reduce GFR estimation error by 25-30% compared to creatinine alone (Inker et al., NEJM, 2012)
- A diagnostic accuracy meta-analysis found cystatin C detected CKD with 85% sensitivity and 87% specificity, with a diagnostic odds ratio of 40 (Zhang et al., 2015)
- Cystatin C predicts cardiovascular events and all-cause mortality independently of GFR, suggesting it captures vascular risk beyond kidney function
What Is Cystatin C?
Cystatin C (also called cystatin 3 or CST3) is a 120-amino-acid protein belonging to the cystatin superfamily of cysteine protease inhibitors. Its primary biological role is regulating extracellular proteolysis by inhibiting cathepsins B, H, K, L, and S, lysosomal enzymes that degrade proteins in various tissues. This function makes cystatin C relevant to processes as diverse as bone remodeling, antigen presentation, and vascular wall integrity.
What makes cystatin C valuable as a kidney biomarker is its pharmacokinetic profile. It is produced at a remarkably stable rate by all nucleated cells (a "housekeeping" gene product). It is small enough (13.3 kDa) to be freely filtered at the glomerulus. After filtration, proximal tubular cells reabsorb and completely catabolize it. No cystatin C returns to the bloodstream from the tubules. This means serum cystatin C concentration is determined almost entirely by one variable: glomerular filtration rate.
Anders Grubb and colleagues first proposed cystatin C as a GFR marker in 1985, noting that its production appeared independent of the factors that confound creatinine. Over the following four decades, this observation has been validated in hundreds of studies across dozens of populations, culminating in its incorporation into the gold-standard GFR equations used worldwide.
Why Creatinine Falls Short
Creatinine is produced by the non-enzymatic conversion of creatine and phosphocreatine in skeletal muscle. This means its serum concentration reflects two things simultaneously: how fast the kidneys are filtering it out, and how much muscle is producing it. In a healthy young man with high muscle mass, serum creatinine can sit at the upper end of the normal range with perfectly healthy kidneys. In an elderly woman with sarcopenia, creatinine can remain within normal limits even as GFR declines to dangerous levels.
The specific confounders are well catalogued:
Muscle mass: Bodybuilders and athletes produce more creatinine; frail elderly patients, amputees, and people with neuromuscular diseases produce less. This creates systematic bias in eGFR that is directionally opposite in different populations.
Diet: Cooked meat contains creatinine and its precursors. A steak dinner the night before a blood draw can raise serum creatinine by 10-30%. Vegetarians tend to run lower creatinine values.
Medications: Trimethoprim and cimetidine inhibit tubular secretion of creatinine, raising serum levels without any change in actual GFR. This can trigger false alarms in patients taking these common medications.
Tubular secretion: Approximately 10-15% of urinary creatinine arrives via tubular secretion rather than glomerular filtration. At low GFR, this fraction increases to 30-50%, meaning creatinine overestimates residual kidney function precisely when accuracy matters most.
Weight change: Patients losing weight on GLP-1 receptor agonists like tirzepatide lose muscle mass alongside fat. Creatinine production drops, eGFR appears to worsen (or stagnate), and clinicians may misinterpret this as kidney injury when true filtration has not declined.
These are not theoretical edge cases. They affect millions of clinical decisions annually. The elderly represent the fastest-growing segment of CKD patients, and they are precisely the population where creatinine performs worst.
Cystatin C vs. Creatinine: The Evidence
The defining study for the cystatin C debate was published in the New England Journal of Medicine in 2013. Shlipak et al. conducted a meta-analysis of 11 general-population cohorts (90,750 participants) and 5 CKD cohorts (2,960 participants), comparing standardized cystatin C and creatinine measurements against clinical outcomes.
The results were decisive. When participants were classified by cystatin C-based eGFR rather than creatinine-based eGFR, the net reclassification improvement was 0.23 (23%) for all-cause mortality and 0.10 (10%) for end-stage renal disease. In practical terms: a meaningful fraction of patients who appeared low-risk by creatinine were actually high-risk by cystatin C, and vice versa. Cystatin C identified the true risk more accurately.
An earlier meta-analysis by Dharnidharka et al. (2002, Clinical Journal of the American Society of Nephrology) compared 46 studies head-to-head and found that cystatin C had a correlation coefficient of 0.816 with measured GFR, compared to 0.742 for creatinine. The difference was most pronounced in the so-called "creatinine-blind range" (GFR 60-90 mL/min/1.73 m2), where creatinine frequently fails to detect early kidney dysfunction.
Zhang et al. (2015) synthesized diagnostic accuracy data across 31 studies and reported cystatin C sensitivity of 85% (95% CI: 81-89%) and specificity of 87% (95% CI: 84-90%) for CKD detection, with a diagnostic odds ratio of 40. These numbers position cystatin C as a strong standalone diagnostic marker, comparable to many established clinical tests.
The pattern across studies is consistent: cystatin C matches or outperforms creatinine in nearly every population studied, with the largest advantages in elderly patients, women, and people with extreme body compositions. The combined evidence from over 100,000 participants leaves little doubt that cystatin C is at least equivalent to creatinine for GFR estimation and superior in multiple clinical contexts.
The CKD-EPI Equations: Combining Both Markers
The practical application of cystatin C in clinical medicine centers on the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equations. These equations convert raw biomarker concentrations into estimated GFR values.
Inker et al. (2012, New England Journal of Medicine) developed and validated three equations using data from 13 studies (5,352 participants with measured GFR):
- Creatinine alone (eGFRcr): the most widely used clinical equation
- Cystatin C alone (eGFRcys): performs comparably to creatinine in most populations
- Combined creatinine-cystatin C (eGFRcr-cys): the most accurate available equation
The combined equation reduced the percentage of GFR estimates that deviated more than 30% from measured GFR by 25-30% compared to the creatinine-only equation. This improvement is large enough to change clinical staging (and therefore treatment decisions) in a substantial number of patients.
Why does combining two imperfect markers outperform either alone? Because their errors are uncorrelated. Creatinine overestimates GFR in patients with low muscle mass and underestimates it in muscular patients. Cystatin C is unaffected by muscle mass but can be influenced by inflammation, obesity, and thyroid dysfunction. By averaging two markers with different error profiles, the combined equation cancels out a portion of each marker's bias.
The clinical implication: when cystatin C is available and the combined equation is used, the confidence interval around a GFR estimate narrows considerably. This matters when clinical decisions hinge on specific GFR thresholds, such as 60 mL/min/1.73 m2 (the threshold for CKD stage 3) or 15 mL/min/1.73 m2 (the threshold for dialysis eligibility). Peptidomic approaches like the CKD273 urinary peptide classifier address a different question (early detection of fibrosis and progression risk), while cystatin C addresses accuracy of current GFR measurement.
Removing Race from Kidney Function Assessment
One of the most consequential developments in nephrology over the past decade has been the removal of race from GFR estimation. The original CKD-EPI creatinine equation included a race coefficient that assigned higher eGFR values to Black patients for any given creatinine level. This coefficient was based on population-level differences in average muscle mass, but its clinical effect was to systematically overestimate kidney function in Black patients, potentially delaying referral, transplant evaluation, and treatment.
Inker et al. (2021, New England Journal of Medicine) developed new CKD-EPI equations without race coefficients. The key finding: removing race from the creatinine-only equation reduced accuracy in some subgroups, but the combined creatinine-cystatin C equation without race performed as well as or better than the old race-adjusted creatinine equation. Cystatin C made race-free GFR estimation possible without sacrificing accuracy.
This is now the recommended approach. The NKF and ASN joint task force endorsed removing race from eGFR equations in 2021, with cystatin C measurement recommended to confirm creatinine-based estimates, particularly when clinical decisions depend on the GFR value. This applies to drug dosing, transplant eligibility, specialist referral, and CKD staging.
The equity implications are substantial. Historically, Black patients with the same true GFR as white patients received higher eGFR values, which could delay transplant listing and access to medications restricted by eGFR thresholds. The race-free combined equation eliminates this structural bias.
Clinical Scenarios Where Cystatin C Excels
While the combined equation is always more accurate, certain clinical situations make cystatin C measurement particularly valuable:
Older adults with sarcopenia: Muscle loss accelerates after age 70. Creatinine-based eGFR systematically overestimates kidney function in this population. A 2024 study in the Journal of the American Geriatrics Society (Yuan et al.) found that the difference between cystatin C-based and creatinine-based eGFR correlated directly with muscle mass and frailty status.
Weight loss on GLP-1 agonists: Patients losing 15-20% body weight on semaglutide or tirzepatide lose both fat and muscle. Creatinine drops, and eGFRcr may not reflect actual GFR changes. Cystatin C provides a check that is independent of body composition shifts.
Vegetarians and patients with eating disorders: Low protein intake reduces creatinine production. eGFRcr may overestimate kidney function by 10-20% in strict vegetarians.
Drug dosing decisions: Approximately 40% of medications are renally cleared, and their doses depend on GFR. For narrow therapeutic index drugs (vancomycin, aminoglycosides, methotrexate, lithium), a 20-30% error in GFR estimation can cause toxicity or treatment failure. The combined equation with cystatin C provides the most accurate GFR for dosing calculations.
Confirmatory testing: When creatinine-based eGFR falls near a clinically important threshold (60 for CKD staging, 45 for SGLT2 inhibitor initiation, 20 for transplant referral), adding cystatin C can confirm or refute the creatinine-based estimate.
Kidney donor evaluation: Living kidney donors undergo extensive evaluation of kidney function. Because the consequences of overestimating a donor's GFR are severe (leaving the donor with inadequate kidney reserve), the combined equation is recommended for donor assessment.
The Limitations That Temper Enthusiasm
Cystatin C is not a perfect marker. Several non-GFR factors affect its serum concentration:
Obesity: Adipose tissue produces cystatin C. BMI above 30 is associated with higher cystatin C levels independent of GFR, causing cystatin C-based eGFR to underestimate true kidney function in obese patients. This is ironic given that creatinine overestimates GFR in obesity (due to higher muscle mass), meaning both markers are biased, just in opposite directions.
Thyroid dysfunction: Hyperthyroidism increases cystatin C production; hypothyroidism decreases it. Patients with uncontrolled thyroid disease may have misleading cystatin C values.
Systemic inflammation: C-reactive protein and other inflammatory markers correlate with cystatin C levels. Patients with active inflammatory conditions (rheumatoid arthritis, inflammatory bowel disease, acute infections) may have elevated cystatin C that does not reflect reduced GFR.
Corticosteroid use: High-dose glucocorticoids increase cystatin C production. Patients on dexamethasone or prednisone may have artificially elevated levels.
Cost and availability: A cystatin C test costs approximately $15-50, compared to $5-10 for creatinine. While this difference is small relative to the clinical value, it has slowed universal adoption. Not all laboratories offer cystatin C testing, and insurance coverage remains inconsistent in some regions.
Standardization: International standardization of cystatin C assays (traceable to the ERM-DA471/IFCC reference material) has improved consistency, but interlaboratory variation remains higher than for creatinine. Clinicians should be aware that not all cystatin C assays are equivalent.
These limitations do not negate cystatin C's advantages. They mean that cystatin C, like creatinine, works best when interpreted in clinical context rather than in isolation. The combined equation mitigates both markers' weaknesses by averaging their independent error sources.
Beyond Kidneys: Cardiovascular and Mortality Risk
One of the more compelling findings in cystatin C research is its association with outcomes beyond kidney function. Shlipak et al. (2005, New England Journal of Medicine) reported that in elderly adults without known kidney disease, elevated cystatin C predicted all-cause mortality, cardiovascular mortality, and heart failure independently of traditional risk factors and independently of creatinine-based eGFR.
This raised a question: is cystatin C capturing something about vascular health that creatinine misses? Cystatin C's biological role as a cathepsin inhibitor is relevant here. Cathepsins B, K, L, and S are active in atherosclerotic plaque remodeling, vascular smooth muscle cell migration, and extracellular matrix degradation in arterial walls. Altered cystatin C levels may reflect changes in this proteolytic balance that have direct cardiovascular consequences, not just passive glomerular filtration changes. This connection between protease inhibition and vascular disease parallels how BNP and NT-proBNP function as both biomarkers and active participants in cardiovascular physiology.
A 2025 JAMA meta-analysis found that discordance between cystatin C-based and creatinine-based eGFR (where cystatin C shows worse kidney function than creatinine) is itself a risk factor for adverse outcomes. Patients with eGFRcys significantly lower than eGFRcr had higher rates of cardiovascular events and death even after adjusting for the average eGFR. This "discordance phenotype" may identify patients with systemic vascular dysfunction who are missed by creatinine-only screening.
Where Cystatin C Fits in the Biomarker Landscape
Cystatin C addresses a specific question: how accurately can we measure GFR right now? It does not detect early kidney damage before GFR declines (that is the domain of NGAL, KIM-1, and other tubular injury markers). It does not predict fibrosis progression (that is CKD273's strength). It does not replace C-peptide for assessing insulin secretion in diabetic kidney disease patients, though both are peptide biomarkers used in the same patient populations.
What cystatin C does is fix the foundational measurement. Every clinical decision in nephrology begins with an eGFR value. If that value is systematically biased by muscle mass, race, or diet, every downstream decision inherits that error. Cystatin C, and specifically the combined creatinine-cystatin C equation, reduces that foundational error more than any other single intervention in kidney function assessment. The urinary peptide revolution explored in our pillar article on urinary peptides and kidney health builds on top of accurate GFR measurement, not as a replacement for it.
Mass spectrometry-based peptidomic approaches are advancing rapidly.[1] Good et al. mapped over 5,000 naturally occurring urinary peptides, establishing the foundation for classifiers like CKD273 that detect kidney disease at its earliest stages.[2] The CKD273 classifier, validated with 85% sensitivity and 100% specificity, represents where kidney diagnostics are heading: panels of peptide biomarkers that capture disease biology rather than a single filtration proxy.[3]
The future kidney health assessment will likely combine cystatin C-based eGFR (for accurate current filtration), tubular injury markers like NGAL and KIM-1 (for acute damage detection), CKD273 (for fibrosis risk), and genomic risk scoring into a multidimensional evaluation that makes today's creatinine-only approach look primitive by comparison. For patients with chronic kidney disease exploring peptide therapeutics, accurate GFR measurement through cystatin C is the starting point for appropriate treatment selection and monitoring.
The Bottom Line
Cystatin C is a 13.3 kDa cysteine protease inhibitor that provides a more accurate measure of kidney filtration than creatinine alone. A 90,750-person meta-analysis demonstrated that it reclassifies mortality risk 23% more accurately, and the combined creatinine-cystatin C CKD-EPI equation is now the gold standard for GFR estimation. Cystatin C is not affected by muscle mass, diet, or race, making it essential for equitable kidney function assessment. Its limitations (sensitivity to inflammation, obesity, and thyroid dysfunction) are real but do not outweigh its advantages, and the combined equation mitigates both markers' weaknesses by averaging independent error sources.