What Is eGFR? Understanding Kidney Function

GFR categories

Full staging of chronic kidney disease pairs the GFR category with an albuminuria category, A1 to A3, from a urine albumin-to-creatinine ratio. Someone at G2 with A3 albuminuria is in a substantially different position from someone at G2 with A1, even though their eGFR is identical — which is why the two tests are read together.

Why a kidney number turns up in heart conversations

People are often surprised to find eGFR discussed at a cardiology appointment, or to see it requested alongside a cholesterol panel. There is a specific reason, and it has become more explicit in US guidance over the last three years.

The American Heart Association’s PREVENT risk model, which replaced the older Pooled Cohort Equations and is now used across US cardiovascular guidelines, takes eGFR as a direct input. Kidney function is not an afterthought in that calculation; it sits alongside blood pressure, cholesterol, smoking and diabetes as one of the terms. A lower eGFR raises the estimated risk, which can change where someone falls relative to the treatment thresholds in the blood pressure and cholesterol guidelines.

Behind that is a broader reframing. The AHA first defined cardiovascular-kidney-metabolic syndrome in a 2023 advisory, describing how excess weight, raised glucose, abnormal lipids, high blood pressure and reduced kidney filtration tend to arrive together and worsen one another. In June 2026 that became a full guideline from the AHA, ACC, American Diabetes Association and American Society of Nephrology, with stages from 0 to 4 and a recommendation that eGFR and urine albumin be assessed together rather than separately.

Trends matter more than single values

A single eGFR is a snapshot with meaningful measurement noise. Day-to-day variation of a few mL/min/1.73m² is entirely normal, and a change of that size between two results usually reflects variation rather than a real change in kidney function.

What clinicians look for instead is the trajectory across results over months and years. A stable eGFR of 55 in someone who has sat at 55 for a decade is a different clinical picture from an eGFR of 55 in someone who was 85 a year ago — even though the number is identical.

• A gradual decline with age is expected physiology, not disease.
• A sudden drop suggests an acute cause — dehydration, a new medication, an infection, or obstruction — and is investigated differently from a slow decline.
• Comparing results measured in the same laboratory is more reliable, since methods differ slightly between laboratories.

When eGFR misleads

Because eGFR is a calculation from creatinine, anything that changes creatinine independently of filtration changes the estimate without kidney function having moved at all.

Where a creatinine-based estimate is likely to be unreliable, guidelines recommend a cystatin C–based equation instead or in addition. Cystatin C is produced by all nucleated cells rather than muscle, so it sidesteps the muscle-mass problem entirely, and a combined creatinine–cystatin C equation is generally the most accurate option available.

You may also meet a second, older number. Creatinine clearance from the 1976 Cockcroft-Gault equation estimates something slightly different, is not size-normalized, and is less accurate than CKD-EPI — but a great many drug dosing instructions were written against it and still name it, so it has not gone away. The two are not interchangeable, and substituting one for the other in a dosing calculation is a real error rather than a rounding difference.