Is estimated glomerular filtration rate (eGFR) the same as measured glomerular filtration rate (GFR), and how should I adjust metformin dosing based on eGFR in a patient on chronic metformin?

Is eGFR the Same as GFR?

No, eGFR (estimated glomerular filtration rate) is not the same as measured GFR—eGFR is a calculated estimate derived from serum biomarkers that approximates true kidney function, while measured GFR uses exogenous filtration markers to directly quantify glomerular filtration and represents the gold standard for kidney function assessment. 1

Understanding the Fundamental Difference

eGFR is an estimation calculated from serum creatinine (eGFRcr) or cystatin C (eGFRcys), or both combined (eGFRcr-cys), using validated equations that incorporate demographic variables. 2, 3 In contrast, measured GFR (mGFR) directly quantifies kidney function through plasma or urinary clearance of exogenous filtration markers such as iothalamate or iohexol. 1

The KDIGO 2024 guidelines explicitly distinguish between these terms, recommending that clinicians "use the term 'GFR' when referring to the specific kidney function of glomerular filtration" and understand that eGFR serves as a practical surrogate but carries inherent limitations. 1

Clinical Accuracy and Performance

How Close is eGFR to True GFR?

The combined creatinine-cystatin C equation (eGFRcr-cys) demonstrates the best performance among estimating equations, with approximately 8.5% of estimates differing by more than 30% from measured GFR—meaning it correctly estimates GFR within 30% accuracy in about 91.5% of cases. 4 However, this still represents meaningful clinical inaccuracy in nearly 1 in 10 patients.

Creatinine-based estimates alone (eGFRcr) show 12.8% of estimates differing by more than 30% from measured GFR, while cystatin C alone shows 14.1% inaccuracy. 4 The median bias across all equations is relatively small (3-4 mL/min/1.73 m²), but precision varies considerably. 4

When eGFR is Systematically Inaccurate

eGFRcr becomes unreliable in multiple clinical scenarios where non-GFR determinants alter serum creatinine independent of kidney function: 1, 3

• Extreme muscle mass (very low or very high) alters creatinine generation
• Malnutrition, cancer, or muscle wasting diseases reduce creatinine production
• Advanced cirrhosis with high catabolic states
• Dietary extremes including low-protein, ketogenic, vegetarian, or high-protein diets with creatine supplements
• Medications such as trimethoprim, cimetidine, or anabolic steroids that affect tubular secretion
• Obesity class III (BMI >40 kg/m²)

In these situations, the American Society of Nephrology recommends measuring cystatin C and calculating eGFRcr-cys, which provides more accurate estimates. 1

Hierarchical Approach to GFR Assessment

Step 1: Initial Assessment with eGFRcr

Use serum creatinine with a validated estimating equation (2021 race-free CKD-EPI equation) as the initial test for all patients, as it is routinely available from basic metabolic panels. 1, 2, 3 This provides adequate screening for most clinical purposes when properly calibrated to international reference standards. 3

Step 2: Confirmatory Testing with Cystatin C

Measure cystatin C and calculate eGFRcr-cys when: 1

• eGFRcr is expected to be inaccurate due to conditions listed above
• GFR affects critical clinical decision-making such as drug dosing (especially nephrotoxic agents or chemotherapy)
• CKD diagnosis or staging requires confirmation
• eGFRcr falls in borderline ranges (45-59 mL/min/1.73 m²) without other markers of kidney damage

The combined equation improves classification accuracy and correctly reclassifies 16.9% of patients with eGFRcr 45-59 mL/min/1.73 m² as having GFR ≥60 mL/min/1.73 m². 4

Step 3: Direct Measurement When Precision is Critical

Measure GFR using exogenous filtration markers when: 1, 3

• eGFRcr-cys is expected to be inaccurate (very low muscle mass with high inflammation, high catabolic states, exogenous steroid use)
• Precise GFR is essential for treatment decisions (chemotherapy dosing, kidney donor evaluation)
• Even the residual 8.5% inaccuracy of eGFRcr-cys is clinically unacceptable

If measured GFR is unavailable, consider 24-hour urine creatinine clearance, recognizing it overestimates true GFR by approximately 10-20%. 3

Specific Implications for Metformin Dosing

The Problem with Using Serum Creatinine Alone

Traditional metformin contraindications based on serum creatinine thresholds (>1.5 mg/dL in men, >1.4 mg/dL in women) systematically miss patients with advanced CKD while unnecessarily restricting treatment in others. 5 Studies show that 15.3% of patients with eGFR <60 mL/min/1.73 m² had serum creatinine below traditional cut-offs and were receiving metformin inappropriately. 5

Evidence-Based eGFR Thresholds for Metformin

Use eGFR-based thresholds rather than serum creatinine for metformin prescribing decisions: 6

• eGFR <30 mL/min/1.73 m² (CKD stage 4 or greater): Absolute contraindication to metformin
• eGFR 30-45 mL/min/1.73 m² (CKD stage 3b): Exercise caution; consider dose reduction to maximum 1000 mg daily and assess other risk factors for lactic acidosis
• eGFR 45-59 mL/min/1.73 m² (CKD stage 3a): Generally safe but monitor kidney function every 3-6 months

When to Use Cystatin C for Metformin Decisions

Measure cystatin C and use eGFRcr-cys when metformin dosing decisions are affected by conditions that make eGFRcr unreliable, particularly: 1, 3

• Elderly patients with low muscle mass where eGFRcr may overestimate true kidney function
• Patients with extreme body composition (severe obesity or cachexia)
• Those on medications affecting creatinine secretion (trimethoprim, cimetidine)

This approach prevents both inappropriate continuation (overdosing risk) and unnecessary discontinuation (underdosing/withholding benefit). 1

Critical Pitfalls to Avoid

Never rely on serum creatinine alone without calculating eGFR—approximately 60% of patients with abnormal kidney function by eGFR have normal serum creatinine values. 3

Do not assume eGFR accuracy in non-steady-state conditions such as acute kidney injury, as estimating equations are validated only for stable kidney function. 3

Recognize that "normal" eGFR ranges overlap with early CKD stages—healthy adults can have eGFR values as low as 63.5 mL/min/1.73 m², so values >60 mL/min/1.73 m² do not exclude kidney disease. 7

For extreme body sizes, adjust normalized eGFR (mL/min/1.73 m²) to absolute clearance (mL/min) by multiplying by the patient's actual body surface area divided by 1.73 before making drug dosing decisions. 2

Consider clinical context systematically—ignoring factors that alter creatinine generation (muscle mass, diet, catabolic states, medications) leads to systematic errors in GFR estimation and inappropriate clinical decisions. 1, 3