Creatinine & Determinants of GFR - Nephrology Lectures 2025

🎯 Why This Lecture Matters

Serum creatinine is the single most frequently used biomarker in clinical medicine to assess kidney function. Yet it is profoundly misunderstood.

Most changes in creatinine reflect changes in GFR.
The dominant clinical reality — and the basis for eGFR equations.

But some changes in creatinine are NOT caused by GFR changes.
Drugs, diet, muscle mass, and tubular handling can all change creatinine independent of filtration.

In AKI, the rate of rise of creatinine is disconnected from GFR.
A concept explored in our AKI module that has profound clinical implications.

🔬 Interactive Glomerular Hemodynamics

Explore how arteriolar tone, drug effects, and glomerular pressure determine GFR and albuminuria. Click the condition buttons on the right panel to interact.

📐 The Determinants of GFR

GFR is governed by the Starling forces across the glomerular capillary wall. The fundamental equation is:

GFR = K_f × P_UF

GFR = K_f × (P_GC − P_BS − π_GC)

Where P_UF = net ultrafiltration pressure

K_f — Ultrafiltration Coefficient

Product of capillary surface area and hydraulic permeability.

Surface area regulated by mesangial cells — contraction reduces K_f
Permeability determined by the 3-layer filtration barrier: fenestrated endothelium → GBM → podocyte slit diaphragms
Diseases that destroy glomeruli (e.g., GN, DKD) reduce K_f
Mesangial contraction by Angiotensin II, endothelin → ↓K_f

P_GC — Glomerular Capillary Pressure

The dominant driving force for filtration. ~60 mmHg normally.

Afferent arteriolar dilation → ↑P_GC (more inflow)
Efferent arteriolar constriction (Ang II) → ↑P_GC (backs up pressure)
Afferent constriction (NSAIDs, SGLT2i via TGF) → ↓P_GC
Efferent dilation (ACEi/ARB) → ↓P_GC
Systemic hypotension below autoregulatory range → ↓P_GC

P_BS — Bowman's Space Pressure

Opposes filtration. Normally ~15 mmHg.

Elevated in urinary obstruction (BPH, stones, tumors)
Tubular obstruction (cast nephropathy, crystal nephropathy)
Interstitial edema compressing tubules
Rarely significant unless obstruction is present

π_GC — Glomerular Oncotic Pressure

Opposes filtration. Rises along the capillary length.

At afferent end: ~25 mmHg (plasma protein concentration)
As filtrate leaves, protein concentrates in remaining plasma
At efferent end: may reach ~35 mmHg → filtration equilibrium
Hypoalbuminemia (nephrotic syndrome, liver failure) → ↓π_GC → may increase GFR transiently but does not increase net filtration long-term due to compensatory mechanisms
Filtration fraction (FF) = GFR/RPF — higher FF means more rapid rise in oncotic pressure along the capillary

🎛️ Arteriolar Tone: The Master Controller of GFR

The afferent and efferent arterioles are the principal regulators of glomerular hemodynamics. Understanding their independent effects is the key to understanding drug effects on GFR.

Change	RBF	P_GC	GFR	FF	Clinical Example
Afferent Dilation	↑↑	↑	↑	↓ or →	Prostaglandins (PGE₂/PGI₂), early DKD
Afferent Constriction	↓↓	↓	↓	→	NSAIDs, SGLT2i (via TGF), cyclosporine
Efferent Constriction	↓	↑	↑ (mild)	↑↑	Angiotensin II
Efferent Dilation	↑	↓	↓	↓	ACEi/ARB (block Ang II)
Afferent Constriction + Efferent Dilation	↓↓	↓↓↓	↓↓↓	↓	Triple Whammy: NSAID + ACEi/ARB + Diuretic

⚠️ The "Triple Whammy" — A Predictable Catastrophe

NSAID (afferent constriction) + ACEi/ARB (efferent dilation) ± Diuretic (volume depletion) → catastrophic loss of glomerular capillary pressure.

Teaching point: This is hemodynamic AKI — entirely predictable from first principles. Explore this in the interactive graphic above by selecting "NSAID + ARB (Triple Whammy)."

🛡️ Renal Autoregulation

The kidney maintains GFR across a wide range of systemic blood pressures through two mechanisms:

1. Myogenic Response

Afferent arteriolar smooth muscle contracts in response to increased intraluminal pressure (Bayliss effect).

Rapid response (seconds)
Intrinsic vascular smooth muscle property
Protects glomerulus from hypertensive injury
Impaired in diabetes → permits barotrauma → hyperfiltration injury

2. Tubuloglomerular Feedback (TGF)

The macula densa of the JGA senses NaCl delivery to the distal tubule and adjusts afferent arteriolar tone.

↑NaCl delivery → adenosine release → afferent constriction → ↓GFR
↓NaCl delivery → ↓adenosine → afferent dilation → ↑GFR
Also modulates renin release from JGA granular cells
SGLT2 inhibitors restore TGF: by blocking proximal Na/glucose reabsorption → ↑NaCl at macula densa → afferent constriction → ↓P_GC → nephroprotection

📊 Autoregulatory Range

GFR remains relatively constant between MAP of ~80–180 mmHg. Below this range, GFR falls linearly with pressure. NSAIDs and calcineurin inhibitors impair autoregulation, making the kidney more vulnerable to hypotension.

In CKD, the autoregulatory curve shifts rightward — patients become "dependent" on higher perfusion pressures and are more vulnerable to hypotension-induced AKI.

🔥 Hyperfiltration — When GFR Is Too High

Paradoxically, an elevated GFR is one of the earliest markers of progressive kidney disease. This is the physiology that drives diabetic kidney disease.

Mechanism in Diabetes (Early DKD)

Hyperglycemia → ↑proximal tubular glucose/Na reabsorption via SGLT2
↓NaCl delivery to macula densa → TGF interprets this as "low GFR"
TGF response: afferent dilation + Ang II-mediated efferent constriction
Result: markedly elevated P_GC (~78 mmHg vs normal 60 mmHg)
High P_GC → elevated shear stress → podocyte injury
Podocyte foot process effacement → slit diaphragm widening → albuminuria
Tubular exposure to filtered albumin → tubular injury → interstitial fibrosis

✅ Therapeutic Targets

Modern nephroprotection directly targets glomerular hemodynamics:

ACEi/ARB: Efferent dilation → ↓P_GC → ↓albuminuria 40–50%
SGLT2i: Restores TGF → afferent constriction → ↓P_GC → ↓albuminuria 30–40%
Finerenone (MRA): Anti-fibrotic + anti-inflammatory → ↓albuminuria 30%
GLP-1 RA: Additional nephroprotection via hemodynamic and metabolic pathways

The initial GFR "dip" seen with ACEi/ARBs and SGLT2i is hemodynamic (reducing P_GC) and is a sign of nephroprotection, not injury — as long as it is modest (<30%) and stabilizes.

🧬 The Three Fates of Creatinine

Creatinine is a 113-dalton molecule produced from the non-enzymatic degradation of creatine and phosphocreatine in skeletal muscle at a nearly constant rate (~20 mg/kg/day). It is handled by the kidney in three ways:

1️⃣ Glomerular Filtration (~85–90%)

Creatinine is freely filtered at the glomerulus — it is small (113 Da), not protein-bound, and not charged. This is why it serves as a GFR marker. At steady state, the amount filtered ≈ the amount excreted.

This is the dominant pathway and is the basis for all eGFR equations.

2️⃣ Tubular Secretion (~10–15%)

Creatinine is actively secreted by the proximal tubule via the organic cation transporter 2 (OCT2) on the basolateral membrane, and MATE1/MATE2-K on the apical membrane.

Because of secretion, creatinine clearance (CrCl) overestimates true GFR by ~10–15%. This becomes proportionally more significant as GFR falls.

3️⃣ Tubular Reabsorption (minimal in health)

Under normal conditions, creatinine reabsorption is negligible. However, at very low urine flow rates (severe dehydration, prerenal states), back-diffusion of creatinine from the tubular lumen into peritubular capillaries can occur.

This is why CrCl may falsely underestimate GFR in severely volume-depleted patients — more creatinine leaks back.

📊 Why This Matters Clinically

At normal GFR, tubular secretion contributes ~10–15% of total creatinine excretion, so CrCl only slightly overestimates GFR. But as GFR declines in CKD, tubular secretion becomes a proportionally larger share of total creatinine clearance — meaning CrCl may overestimate true GFR by 50% or more in advanced CKD. This is one reason eGFR equations were developed to replace CrCl.

💊 Drugs That Raise Creatinine WITHOUT Changing GFR

These medications inhibit tubular secretion of creatinine at OCT2 or MATE transporters, causing serum creatinine to rise even though GFR is unchanged. This is a pseudo-AKI that must be recognized.

Drug	Brand/Class	Mechanism	Expected ΔCr	Clinical Pearl
Trimethoprim	Bactrim (TMP-SMX)	Blocks OCT2 and MATE1 → inhibits Cr secretion	↑ 0.2–0.5 mg/dL	Most common clinical scenario. Dose-dependent. Reversible on discontinuation.
Cimetidine	Tagamet (H2 blocker)	Competitive inhibitor of OCT2 → blocks Cr secretion	↑ 0.2–0.4 mg/dL	Historically used therapeutically to block Cr secretion when measuring CrCl to better approximate true GFR.
Triamterene	Dyrenium (K-sparing diuretic)	Inhibits tubular Cr secretion	↑ 0.1–0.3 mg/dL	Less commonly used today. Also seen with amiloride to lesser degree.
Dolutegravir	Tivicay (INSTI for HIV)	Inhibits OCT2 → blocks Cr secretion	↑ 0.1–0.2 mg/dL	Consistent effect seen in clinical trials. Occurs within first 2–4 weeks and stabilizes. GFR unaffected.
Cobicistat	CYP3A4 booster (HIV regimens)	Inhibits MATE1 → blocks Cr secretion	↑ 0.1–0.3 mg/dL	Seen with all cobicistat-containing HIV regimens. Predictable, stable rise.
Dronedarone	Multaq (antiarrhythmic)	Inhibits tubular Cr secretion	↑ 0.1 mg/dL	Mild effect. Stabilizes within a week.
Fenofibrate	Tricor (fibrate)	Increases creatinine production AND may inhibit secretion	↑ 0.1–0.2 mg/dL	Paradoxically may also have renal-protective properties in DKD.

✅ How to Recognize Pseudo-AKI

Cr rises modestly (0.1–0.5 mg/dL) after starting the drug
Rise occurs within days to 2 weeks and then stabilizes
Patient is clinically well — no oliguria, no symptoms
Urine sediment is bland
If available, cystatin C will be unchanged (not affected by tubular secretion)
Cr returns to baseline on drug discontinuation

⚠️ The Clinical Danger

Misinterpreting a Trimethoprim-induced Cr rise as true AKI can lead to unnecessary drug discontinuation, needless imaging, inappropriate fluid boluses, or even nephrology consultation for a non-existent problem. Conversely, assuming a Cr rise is from the drug when true AKI is occurring is equally dangerous.

Key: Pattern recognition + cystatin C confirmation when uncertain.

📝 The Historical Cimetidine Trick

Before modern eGFR equations, clinicians sometimes administered cimetidine prior to a timed urine collection to block tubular creatinine secretion. This made CrCl a better approximation of true GFR (since the secreted component was eliminated). While this practice is largely obsolete, it beautifully illustrates that creatinine clearance = filtration + secretion, and that blocking secretion brings CrCl closer to GFR.

⚖️ Non-GFR Factors That Affect Serum Creatinine

Remember: serum creatinine at steady state = production / excretion. Anything that changes production or non-GFR excretion changes creatinine without changing GFR.

↑ Increased Creatinine (Without ↓GFR)

High muscle mass — bodybuilders, young muscular males
High-meat diet — particularly cooked red meat (contains creatine → creatinine)
Creatine supplements — directly increases substrate for creatinine production
Drugs inhibiting secretion — Trimethoprim, Cimetidine, etc. (see above)
Rhabdomyolysis — massive release from damaged muscle (though GFR often falls too)

↓ Decreased Creatinine (Without ↑GFR)

Low muscle mass — elderly, cachectic, amputees, paraplegia
Vegetarian diet — less dietary creatine intake
Liver disease — impaired creatine synthesis (creatine is made in the liver)
Pregnancy — increased plasma volume (dilution) + increased GFR (40–50%)
Malnutrition — sarcopenia and reduced creatinine production

⚠️ The Sarcopenia Trap

An elderly, frail patient with a serum creatinine of 1.0 mg/dL may have a true GFR of only 30–40 mL/min. Their low muscle mass generates so little creatinine that even severe renal impairment cannot raise it above the "normal range." This is why eGFR equations incorporate age and sex — but they still may overestimate GFR in cachectic patients.

Clinical pearl: Always interpret creatinine in the context of the patient's body habitus and muscle mass. A "normal" creatinine in a small, elderly patient is not reassuring.

📉 Limitations of eGFR

The estimated GFR is a regression equation derived from populations with stable kidney function. It has important limitations.

CKD-EPI 2021 Equation

Uses creatinine, age, and sex
Removed race coefficient (2021 update)
Most accurate for stable CKD
Validated for eGFR 15–90 mL/min/1.73m²

CKD-EPI Cystatin C (or Combined)

Cystatin C is not affected by muscle mass, diet, or tubular secretion
More accurate in extremes of body composition
Combined creatinine-cystatin C equation is the most accurate
Recommended by KDIGO when confirmatory testing is needed

When eGFR Fails — Known Limitations

Limitation	Explanation	Solution
Non-steady state (AKI)	Creatinine is changing — eGFR formula assumes stability. In AKI, eGFR dramatically overestimates true GFR.	Do not use eGFR in AKI. Follow Cr trends. See AKI module.
Extremes of body size	Very muscular or very cachectic patients have abnormal creatinine production.	Use cystatin C–based eGFR or 24-hour CrCl.
Drugs affecting Cr secretion	Trimethoprim, cimetidine, dolutegravir → false ↑Cr → false ↓eGFR.	Recognize pattern. Confirm with cystatin C.
High-protein diet / creatine supplements	Increased creatinine production → false ↑Cr → false ↓eGFR.	Dietary history. Cystatin C if uncertain.
Extremes of age	Not well validated in very young or very elderly (>85 years).	Clinical judgment. Combined equation.
Pregnancy	GFR increases 40–50% in pregnancy; normal values differ.	eGFR equations not validated in pregnancy. Use measured CrCl.
Amputees / Paraplegia	Markedly reduced muscle mass → low Cr production → falsely high eGFR.	Cystatin C essential. Do not rely on creatinine-based eGFR.
Hyperfiltration	High eGFR (>120) may represent early DKD, not health.	Check ACR. Evaluate for hyperfiltration etiology.

⏱️ Creatinine Kinetics & the AKI Disconnect

The rate of rise of serum creatinine is disconnected from the actual GFR in AKI. This is a foundational concept covered in our AKI module. Here we expand on the why.

The Steady-State Assumption

At steady state: Creatinine production = Creatinine excretion. Serum Cr is constant. eGFR is valid.

When GFR abruptly drops (AKI), production continues at the same rate (~20 mg/kg/day from muscle), but excretion falls. Creatinine accumulates — but it takes time to accumulate enough to raise serum levels.

Why 3–5 Days?

Think of creatinine like filling a bathtub with a constant-rate faucet (production) and a partially closed drain (reduced GFR). The tub fills gradually, not instantly.

The time constant ≈ volume of distribution / new clearance rate. At normal body water (~42L) and severely reduced clearance, it takes ~3–5 half-lives to reach a new steady state — roughly 3–5 days.

📊 Clinical Implications Reviewed

Day 0 of complete renal failure: GFR = 0, but Cr may still be 1.0 mg/dL and eGFR = 90
Day 1: GFR = 0, Cr = 2.0, eGFR ≈ 38 — still dramatically wrong
Day 3–5: Cr finally plateaus near 8–12 mg/dL and eGFR "catches up"
The maximum rate of rise of creatinine in anuric patients is approximately 1.0–1.5 mg/dL per day (depending on muscle mass)
A Cr rise of only 0.3 mg/dL in 48 hours (KDIGO Stage 1) may represent a massive GFR drop
Declining Cr after AKI does NOT mean GFR has normalized — it means creatinine is being cleared faster than it is being produced, but the system may not yet be at steady state

📊 BUN/Creatinine Ratio — Context Matters

BUN (blood urea nitrogen) and creatinine are both markers of kidney function, but they are handled differently by the kidney:

Creatinine

Freely filtered + secreted
Not significantly reabsorbed
Production relatively constant

BUN (Urea)

Freely filtered
Significantly reabsorbed (40–60%), especially in volume-depleted states under ADH influence
Production varies with protein intake, GI bleeding, steroids, catabolic states

BUN/Cr Ratio	Interpretation	Common Causes
10:1 – 20:1	Normal	Normal renal function
>20:1	Prerenal pattern (BUN rises disproportionately)	Volume depletion, CHF, GI bleed, high-protein diet, steroids, catabolic states
<10:1	Intrinsic renal damage or decreased urea production	ATN, rhabdomyolysis (Cr disproportionately elevated), liver failure (low BUN production), low-protein diet

🧪 Cystatin C — The Creatinine Complement

When creatinine cannot be trusted, cystatin C provides an independent GFR estimate.

✅ Advantages

Produced by all nucleated cells at a constant rate
Not affected by muscle mass, diet, or sex
Not significantly secreted by tubules
Freely filtered at the glomerulus
Combined Cr-CysC equation is most accurate eGFR

⚠️ Limitations

Affected by high-dose corticosteroids (↑ production)
Thyroid dysfunction: hyperthyroid ↑, hypothyroid ↓ cystatin C
Obesity and inflammation may affect levels
More expensive and less widely available than creatinine
Still has the same steady-state requirement for eGFR

🎯 When to Order Cystatin C

Suspected drug-induced creatinine elevation (trimethoprim, dolutegravir)
Extremes of body composition (amputees, bodybuilders, cachectic patients)
eGFR near critical decision thresholds (e.g., 15 mL/min for dialysis planning, 45 for referral)
Discrepancy between creatinine-based eGFR and clinical assessment
KDIGO recommends confirmatory cystatin C testing when eGFR 45–59 without other markers of kidney damage

🏅 Measured GFR — The Gold Standard

Measured GFR (mGFR) using exogenous filtration markers provides the most accurate assessment. The ideal marker is freely filtered, not reabsorbed, not secreted, not metabolized, and not protein-bound.

Inulin Clearance

Historical gold standard. Fructose polymer meeting all ideal filtration marker criteria. Requires continuous IV infusion, timed urine collections, and multiple blood samples — impractical for routine use.

Iohexol Plasma Clearance

Non-ionic, non-radioactive contrast agent. Single injection + plasma sampling. Excellent correlation with inulin. Preferred in many centers. Used in the landmark Stockholm cohort studies.

Nuclear Medicine (⁹⁹ᵐTc-DTPA, ⁵¹Cr-EDTA)

Radioactive markers with similar accuracy. ¹²⁵I-iothalamate used in many US research centers. Avoid timed urine collections with plasma clearance techniques.

When to consider measured GFR: Living kidney donor evaluation, when both eGFRcr and eGFRcys are unreliable (e.g., cirrhosis), research protocols, or critical clinical decisions (transplant evaluation, questionable dialysis timing).

📜 Evolution of Creatinine-Based eGFR Equations

Understanding the evolution of equations helps appreciate their limitations and appropriate application.

Equation	Year	Variables	Limitations	Status
Cockcroft-Gault	1976	Age, weight, sex, Cr	Estimates CrCl (not GFR). Not standardized to BSA. Overestimates in obesity.	Largely replaced
MDRD	1999	Age, sex, race, Cr	Developed in CKD population. Systematically underestimates GFR >60. Race coefficient.	Replaced
CKD-EPI 2009	2009	Age, sex, race, Cr	Improved accuracy at higher GFR. Still included race coefficient.	Replaced
CKD-EPI 2021	2021	Age, sex, Cr	Race-free. Now standard for clinical reporting. ~87–88% P30 accuracy.	Current Standard
CKD-EPI 2012 CysC	2012	Age, sex, Cystatin C	Race-free. More accurate in extremes of body composition. Affected by steroids, thyroid, inflammation.	Confirmatory
CKD-EPI 2021 Cr-CysC	2021	Age, sex, Cr, Cystatin C	Race-free. ~91–92% P30 accuracy. Most accurate available equation.	Most Accurate

The "Creatinine-Blind" Range: The inverse, nonlinear relationship between Cr and GFR means a serum Cr of 1.0 mg/dL may represent a GFR of 120 in a young muscular male or 50 in an elderly sarcopenic female. Patients can lose >50% of kidney function before creatinine rises above the "normal" reference range. Always report and interpret eGFR — never creatinine alone.

📊 Comprehensive: What Makes eGFR Wrong?

Beyond the basics covered above, here are systematic tables of all conditions that cause eGFRcr and eGFRcys to deviate from measured GFR.

Conditions Where eGFRcr OVERESTIMATES True GFR (Falsely Reassuring)

Condition	Mechanism	Clinical Significance
Sarcopenia / Low muscle mass	↓Cr generation	Common in elderly, chronic illness; may mask CKD
Cachexia / Wasting	Severe ↓muscle mass	Cancer, HF, COPD; significantly overestimates GFR
Amputations	Proportional loss of muscle	Cystatin C essential for accurate assessment
Paraplegia / Quadriplegia	Atrophy and denervation	May have very low Cr despite significant CKD
Advanced cirrhosis	↓Hepatic creatine synthesis + muscle wasting + ↑tubular secretion	eGFRcr very unreliable; use CysC or measured GFR
Vegetarian / Low-protein diet	↓Dietary creatine	May lower Cr by 10–15%
Advanced CKD (GFR <15–20)	↑↑Tubular secretion (up to 50% of excretion); extrarenal clearance via gut creatininase	eGFRcr may overestimate true GFR by 20–30% in Stage 5
Pregnancy	↑GFR, ↑plasma volume, ↑tubular secretion	Equations not validated; use measured CrCl

Conditions Where eGFRcr UNDERESTIMATES True GFR (Falsely Alarming)

Condition	Mechanism	Clinical Significance
High muscle mass / Bodybuilders	↑Cr generation	May appear to have CKD when GFR is normal
High dietary meat intake	Exogenous Cr from cooked meat	Acute Cr increase post-meal
Creatine supplements	Direct Cr precursor	Common in athletes
Trimethoprim, Cimetidine, Cobicistat, Dolutegravir, Dronedarone	Inhibit OCT2/MATE tubular secretion	Reversible; establish new baseline after 2–4 weeks
Fenofibrate	↑Cr production + ↓secretion	Reversible; may also be nephroprotective
Jaffe assay interference	Chromogens (bilirubin, ketones) react with picrate	Use enzymatic assay for specificity
DKA	Acetoacetate interferes with Jaffe reaction	Falsely ↑Cr during DKA; resolves with treatment
Acute rhabdomyolysis	Massive Cr release from damaged muscle	Cr may rise before AKI develops

Factors Affecting Cystatin C–Based eGFR

eGFRcys UNDERESTIMATES GFR (↑CysC)

Hyperthyroidism — ↑cellular metabolic rate (62% have elevated CysC)
Glucocorticoids — stimulate CysC gene expression (dose-dependent)
Obesity (Class II–III) — ↑production from adipose + inflammation
Systemic inflammation / High CRP — sepsis, autoimmune disease
Malignancy — ↑cell turnover (especially hematologic)
HIV (uncontrolled) — immune activation
Smoking, diabetes, heart failure — chronic low-grade inflammation

eGFRcys OVERESTIMATES GFR (↓CysC)

Hypothyroidism — ↓cellular production (eGFRcys ↑ when treated)
Some immunosuppressive therapy — variable effect
Much fewer confounders than Cr in this direction

📈 The eGFRcys / eGFRcr Ratio — Beyond GFR Estimation

When both markers are measured, their ratio provides diagnostic and prognostic information beyond either marker alone. A ratio <0.7 independently predicts mortality.

Ratio	Interpretation	Possible Causes	Action
>1.0–1.15	eGFRcys higher than eGFRcr	High muscle mass, meat intake, creatine supplements, TMP, hypothyroidism	eGFRcys may be more accurate; evaluate confounders
0.85–1.0	Normal concordance	Neither marker disproportionately affected	Either eGFR likely reliable; eGFRcr-cys for best accuracy
0.70–0.84	Mild discordance (eGFRcys lower)	Mild sarcopenia, low-dose steroids, subclinical inflammation	Assess for sarcopenia; use eGFRcr-cys
<0.70	Marked discordance	Sarcopenia, steroids, hyperthyroidism, malignancy, Shrunken Pore Syndrome	Evaluate confounders; if absent → consider SPS; ↑mortality risk
<0.60	Severe discordance	Shrunken Pore Syndrome most likely if confounders excluded	High mortality risk independent of GFR; close cardiovascular monitoring

🔬 Shrunken Pore Syndrome

First described in 2015, SPS is defined by an eGFRcys/eGFRcr ratio below 0.60–0.70 in the absence of known confounders (steroids, sarcopenia). The syndrome reflects selective impairment of filtration of mid-sized molecules (10–30 kDa, including cystatin C at 13 kDa) while preserving filtration of smaller molecules (creatinine at 113 Da).

SPS is associated with hazard ratios of 3.0–7.3 for mortality depending on the population studied. The 2024 KDIGO guidelines recommend measuring both markers; the ratio adds prognostic information beyond GFR alone.

💡 Prognostic Value Beyond GFR Accuracy

The CKD Prognosis Consortium meta-analysis (JAMA 2025) showed individuals with eGFRcys at least 30% lower than eGFRcr had significantly higher risks of all-cause mortality, cardiovascular mortality, heart failure hospitalization, and kidney failure — even after adjusting for eGFRcr and albuminuria. The ratio captures cardiovascular/metabolic health information beyond filtration function.

🎯 When Markers Disagree: Which Is Closer to Truth?

Evidence from the Stockholm Iohexol Cohort (Fu et al., JASN 2023)

6,185 adults with 9,404 concurrent measurements of creatinine, cystatin C, and iohexol plasma clearance — including patients with CVD, HF, diabetes, liver disease, and cancer.

Key Finding: When eGFRcys < eGFRcr, Creatinine Has the Larger Error

In the 47% of samples where eGFRcys was >20% lower than eGFRcr (the common cardiorenal pattern):

Creatinine overestimated true GFR by +15.0 mL/min/1.73m²
Cystatin C underestimated by only −8.5 mL/min/1.73m²
Combined equation (eGFRcr-cys) bias: only +0.8 mL/min/1.73m²

Creatinine's error was nearly TWICE the magnitude of cystatin C's error.

Performance Metrics by Discordance Pattern

Metric	eGFRcys < eGFRcr (47%)	eGFRcys ≈ eGFRcr (45%)	eGFRcys > eGFRcr (8%)
Median mGFR (iohexol)	47	71	74
Median eGFRcr	62 (overestimates)	73	67
Median eGFRcys	38	71	83
Median eGFRcr-cys	49 (closest to 47)	72	75
eGFRcr Bias (eGFR − mGFR)	+15.0	+3.0	−4.5
eGFRcys Bias	−8.5	−2.0	+8.4
eGFRcr-cys Bias	+0.8	+0.5	+1.4
P30 (eGFRcr)	50%	88%	70%
P30 (eGFRcys)	73%	89%	66%
P30 (eGFRcr-cys)	84%	93%	84%
Correct CKD Stage (eGFRcr)	38%	68%	61%
Correct CKD Stage (eGFRcr-cys)	62%	74%	72%

P30 = proportion of estimates within 30% of measured GFR. Bias = median (eGFR − mGFR) in mL/min/1.73m². Data: Fu et al., JASN 2023.

Why Creatinine Overestimates in Cardiorenal Patients

In HF, diabetes with complications, cancer, and advanced age, muscle mass is frequently reduced (sarcopenia, cardiac cachexia). Less muscle → less creatinine → equations interpret low Cr as preserved GFR. The patient has both reduced muscle mass AND reduced kidney function.

The Reframing

When eGFRcys is substantially lower than eGFRcr in a cardiorenal patient, do not assume cystatin C is wrong. mGFR-validated data shows creatinine is typically further from truth. Use eGFRcr-cys. The patient likely has more kidney disease than creatinine suggests.

Clinical Implication: Relying on eGFRcr alone in patients with HF, diabetes, cancer, or sarcopenia may substantially overestimate kidney function — delaying CKD recognition, resulting in inappropriate drug dosing, or missing eligibility for renal-protective therapies. Always obtain cystatin C and calculate eGFRcr-cys in these populations.

📉 Creatinine Sensitivity at Low GFR: The Advanced CKD Paradox

In patients with GFR below 20 mL/min/1.73m², several factors cause creatinine-based equations to systematically overestimate true GFR:

↑Tubular secretion — accounts for up to 50% of urinary creatinine excretion in advanced CKD (vs. 10–15% at normal GFR)
Extrarenal clearance — gut bacterial creatininase becomes clinically significant
Sarcopenia and cachexia — common in advanced CKD, reducing creatinine generation

Clinical impact: In patients with eGFR <20, creatinine-based equations may overestimate true GFR by 20–30%. When accurate GFR is needed for dialysis initiation timing or transplant evaluation, use eGFRcr-cys or measured GFR.

👥 Special Populations

Elderly Patients

Age-related sarcopenia reduces Cr generation → eGFRcr overestimates. Combined with age-related GFR decline, this masks significant CKD. eGFRcr-cys is preferred when accuracy matters.

Cirrhosis

↓Hepatic creatine synthesis + muscle wasting + ↑tubular secretion → eGFRcr markedly overestimates. CysC performs better but may be affected by inflammation. Measured GFR with iohexol recommended for liver transplant evaluation.

Kidney Transplant Recipients

Altered Cr metabolism from immunosuppressants + prior chronic illness muscle loss. CysC may be affected by glucocorticoids early post-transplant. Combined equation preferred; interpret trends over time.

Acute Kidney Injury

Neither Cr nor CysC equations are validated in AKI. Both lag behind true GFR changes. Interpret absolute values and trends — not calculated eGFR. See our AKI module.

💰 Cystatin C: Cost, Availability & Coverage

Laboratory Costs

CysC reagent: ~$5–10/test
Creatinine reagent: ~$0.50/test
Cost differential decreasing with ↑test volumes

Medicare Coverage (CPT 82610)

Reimbursement: ~$18.52/test
Requires documentation of medical necessity
Coverage criteria: uncertain eGFRcr range, known Cr confounders, or management impact

Tip for coverage: Document in the order why creatinine-based eGFR may be inaccurate (sarcopenia, extremes of muscle mass, cirrhosis, medication effects) and how accurate GFR will affect management (drug dosing, CKD staging, transplant evaluation, eligibility for renal-protective therapies).

🧭 Clinical Approach to GFR Assessment

Start with eGFRcr (CKD-EPI 2021) — appropriate for routine monitoring in stable patients without known confounders.
Add cystatin C (calculate eGFRcr-cys) when: eGFRcr is near decision thresholds (60, 45, 30, 15), Cr reliability is in question (sarcopenia, cirrhosis, drug effects), clinical findings are discordant with eGFRcr, or accurate GFR is needed for drug dosing.
Calculate the eGFRcys/eGFRcr ratio when both markers are available. Ratio <0.7 warrants investigation for sarcopenia, steroid use, or Shrunken Pore Syndrome, and conveys independent prognostic information.
Consider measured GFR (iohexol/iothalamate clearance) for living kidney donor evaluation, when both eGFR markers are unreliable, or when the highest accuracy is required.
Always interpret in clinical context. 80–90% of eGFR values fall within 30% of measured GFR — meaning 10–20% of patients will have significant discrepancies. No equation is infallible.

📋 High-Yield Summary: Most vs. Some

MOST Changes in Cr = GFR Changes

AKI (prerenal, intrinsic, postrenal)
CKD progression
Hemodynamic changes (ACEi/ARB, NSAIDs, SGLT2i)
Volume depletion / overload
Sepsis, cardiorenal syndrome

→ These are true GFR changes and warrant clinical action.

SOME Changes in Cr ≠ GFR Changes

Drugs blocking Cr secretion (TMP, cimetidine, dolutegravir, triamterene)
Changes in muscle mass / diet
Creatine supplements
Rhabdomyolysis (massive Cr release from muscle)
Lab interference (e.g., high bilirubin, ketoacidosis with Jaffe assay)

→ These are "pseudo" changes. Use cystatin C, clinical context, and drug history to differentiate.

🎓 Key Learning Points

GFR is determined by K_f × (P_GC − P_BS − π_GC). Afferent and efferent arteriolar tone are the principal regulators of P_GC.
Creatinine is handled three ways: filtered (~85–90%), secreted (~10–15% via OCT2/MATE), and minimally reabsorbed.
Because of secretion, CrCl overestimates GFR — this error grows as GFR falls (up to 50% of Cr excretion via secretion in Stage 5 CKD).
Multiple drugs (trimethoprim, cimetidine, triamterene, dolutegravir, cobicistat, dronedarone) raise creatinine by blocking tubular secretion, not by reducing GFR.
Muscle mass, diet, and supplements alter creatinine production independent of GFR. A "normal" creatinine in a sarcopenic patient may mask severe CKD.
eGFR assumes steady state — it is invalid in AKI, where the rate of rise of Cr is disconnected from the actual GFR.
The CKD-EPI 2021 race-free equation is the current standard. The combined eGFRcr-cys equation is the most accurate (~91–92% P30).
Cystatin C provides an independent GFR estimate. When eGFRcys < eGFRcr in cardiorenal patients, creatinine is typically further from truth than cystatin C (Stockholm cohort).
The eGFRcys/eGFRcr ratio <0.7 independently predicts mortality, even after adjusting for eGFR and albuminuria.
Hyperfiltration (elevated P_GC) drives early DKD — ACEi/ARBs and SGLT2 inhibitors are nephroprotective because they reduce P_GC.
The "Triple Whammy" (NSAID + ACEi/ARB + diuretic) is a predictable hemodynamic catastrophe.
Most changes in creatinine reflect GFR changes — but the clinician must recognize the exceptions.

📚 References

Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate GFR. Ann Intern Med. 2009;150:604-612.
Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function — measured and estimated GFR. NEJM. 2006;354:2473-83.
Inker LA, Eneanya ND, Coresh J, et al. New creatinine- and cystatin C-based equations to estimate GFR without race. NEJM. 2021;385:1737-49.
Delgado C, Baweja M, Crews DC, et al. A unifying approach for GFR estimation: NKF-ASN Task Force recommendations. JASN. 2021;32:2994-3015.
Levey AS, Coresh J, Tighiouart H, Greene T, Inker LA. Measured and estimated GFR: current status and future directions. Nat Rev Nephrol. 2020;16:51-64.
Soveri I, Berg UB, Bjork J, et al. Measuring GFR: a systematic review. Am J Kidney Dis. 2014;64:411-424.
Shlipak MG, Matsushita K, Arnlov J, et al. Cystatin C versus creatinine in determining risk based on kidney function. NEJM. 2013;369:932-943.
Inker LA, Schmid CH, Tighiouart H, et al. Estimating GFR from serum creatinine and cystatin C. NEJM. 2012;367:20-29.
KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of CKD. Kidney Int. 2024;105(4S):S117-S314.
Grubb A, Lindstrom V, Jonsson M, et al. Reduction in glomerular pore size — "Shrunken pore syndrome." Scand J Clin Lab Invest. 2015;75:333-340.
Purde MT, Nock S, Risch L, et al. Ratio of cystatin C and creatinine-based eGFR predicts mortality. Clin Chem Lab Med. 2022;60:675-684.
Shemesh O, Golbetz H, Kriss JP, Myers BD. Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int. 1985;28:830-838.
Delanaye P, Cavalier E, Pottel H. Serum creatinine: not so simple! Nephron. 2017;136:302-308.
Lepist EI, et al. Impact of dolutegravir on eGFR in healthy volunteers. Clin Pharmacol Ther. 2020;107:189-199.
Fu EL, Coresh J, Engstrom G, et al. Accuracy of GFR estimating equations in patients with discordances between Cr- and CysC-based estimations. JASN. 2023;34:1574-1586.
Grams ME, Estrella MM, Coresh J, et al. Discordance in Cr- and CysC-based eGFR and clinical outcomes: a meta-analysis. JAMA. 2025.
Heerspink HJL, et al. Dapagliflozin in patients with CKD (DAPA-CKD). NEJM. 2020;383:1436-46.
The EMPA-KIDNEY Collaborative Group. Empagliflozin in patients with CKD. NEJM. 2023;388:117-27.
Breyer MD, Harris RC. Cyclooxygenase 2 and the kidney. Curr Opin Nephrol Hypertens. 2001;10:89-98.
Brenner BM, Rector FC. Brenner and Rector's The Kidney. 11th ed. Elsevier; 2020.
Francoz C, Glotz D, Moreau R, Durand F. Evaluation of renal function in patients with cirrhosis. J Hepatol. 2010;52:605-613.
Ferreira JP, Girerd N, Pellicori P, et al. The difference between CysC- and Cr-based eGFR in HFrEF: PARADIGM-HF. Kidney Med. 2023;5:100629.

🔬 Creatinine & Determinants of GFR