Low eGFR: what the stages mean, and why one result is not chronic kidney disease

Glomerular filtration rate is the volume of fluid the kidneys filter per minute, standardized to body surface area, so the units are milliliters per minute per 1.73 square meters. Measuring it directly is impractical in routine care, so laboratories estimate it from serum creatinine using an equation. The equation in current use in the United States is the 2021 CKD-EPI creatinine equation, which removed the race coefficient that earlier versions applied.

That word estimated matters. Creatinine is a muscle breakdown product, so anything that changes muscle mass or creatinine handling shifts the estimate without the kidneys changing at all. A muscular person may have a creatinine that makes their kidneys look worse than they are; a frail older adult with little muscle may have one that flatters them. This is the main reason KDIGO's 2024 recommendation reads as it does: "In adults at risk for CKD, we recommend using creatinine-based estimated glomerular filtration rate (eGFRcr). If cystatin C is available, the GFR category should be estimated from the combination of creatinine and cystatin C." Cystatin C is not muscle-dependent, so the combined estimate is less easily fooled.

The footnote to that table is the part almost never quoted, and it is the one that matters most to a worried reader: "In the absence of evidence of kidney damage, neither G1 nor G2 fulfils the criteria for CKD." An eGFR of 75 with a normal urine test and no structural abnormality is not chronic kidney disease. It is category G2, which is a description of filtration, not a diagnosis.

KDIGO classifies chronic kidney disease by cause, GFR category and albuminuria category, abbreviated CGA. Albuminuria is measured as a urine albumin-to-creatinine ratio, and it is a stronger predictor of outcomes than the eGFR alone at many levels of function. The guideline notes that this nuance "is often missed by healthcare providers and students".

Two people can share an eGFR of 55 and face very different futures depending on whether their ACR is 8 or 400. KDIGO's practice point is to test people at risk using both urine albumin measurement and an assessment of GFR. If you have been given a kidney stage on the strength of a blood test alone, asking whether a urine ACR was done is a reasonable and specific question.

KDIGO answers this directly. Its Table 1 gives two routes to chronic kidney disease, either one present for at least three months: a marker of kidney damage, such as albuminuria, or a decreased GFR, defined as below 60. They are alternatives. An eGFR below 60 that persists for three months meets the definition on its own, with a normal urine test.

What a normal urine result changes is the risk, not the label. KDIGO grades prognosis on both dimensions at once, and the A1 column, normal to mildly increased albumin, is the lowest-risk cell in rows G1 to G3b. In G4 and G5 every cell is very high.

So an eGFR of 52 with an ACR below 30, confirmed over three months, is CKD category G3a A1: a real diagnosis, at moderately increased risk. The same eGFR with an ACR of 100 is G3a A2, one risk step higher. Below 45, a normal ACR no longer keeps you out of the high-risk cells.

Check which test the "no protein" came from. A standard urinalysis reports protein from a dipstick, which KDIGO calls a reagent strip, and KDIGO ranks a manually read total-protein strip last among the initial tests it accepts. Its first choice is a urine albumin-to-creatinine ratio, ideally on a first morning sample, and the A1 to A3 categories are defined by that ratio. A negative strip is not an ACR, so if your report shows only a urinalysis, your albuminuria category has not been measured.

The definition is time-bound. KDIGO requires the abnormality to be present for a minimum of three months, and its practice points spell out what that means operationally: following incidental detection of an elevated ACR, hematuria or a low eGFR, repeat the tests to confirm the presence of CKD. And then, directly: "Do not assume chronicity based upon a single abnormal level for eGFR and ACR, as the finding could be the result of a recent acute kidney injury (AKI) event or acute kidney disease (AKD)."

This is not a technicality. Dehydration, a vomiting illness, a course of non-steroidal anti-inflammatories, contrast imaging, a new blood pressure medicine or a urinary infection can all drop an eGFR temporarily. Those are acute events with a different meaning and often a different treatment. Chronicity can also be established retrospectively, by reviewing past eGFR and albuminuria results, imaging showing reduced kidney size, or the medical history.

The guideline balances this with a practice point in the other direction: consider starting treatment at first presentation of a decreased GFR or elevated ACR if CKD is deemed likely on other clinical grounds. Waiting three months to confirm is not the same as waiting three months to act.

Average filtration falls across adult life, which raises an obvious objection to a fixed threshold: is an eGFR of 58 at 80 the same finding as an eGFR of 58 at 35? KDIGO addresses the objection rather than dodging it. It acknowledges an average age-associated GFR decline in longitudinal and cross-sectional studies, with substantial variation between individuals, "such that not all individuals will have a significant GFR decline with age". It also notes that the threshold of 60 is well below the average in young adults, and that a significant reduction in a younger person is usually accompanied by other markers that would capture the diagnosis anyway.

The guideline keeps the fixed threshold, and the reason is empirical: even in older populations, the risk of adverse outcomes rises with higher CKD stages. Our separate guide on eGFR by age sets out the average values by decade and the state of the age-adaptation debate.

Repeat it, and get the urine test. A confirmed value at least three months apart, with an ACR, is the minimum for a real classification.

Ask what changed. New medicines, an illness, dehydration, a recent scan with contrast. This explains a large share of one-off drops.

Look at the trend, not the label. A stable G3a for a decade is a different clinical object from a G3a that was G2 a year ago. Rate of change is what drives management.

Ask about cystatin C. Particularly if you are very muscular, very frail, an amputee, or if the number does not fit the rest of the picture.

Check the modifiable drivers. Blood pressure and glucose control do most of the work in slowing progression, and both are covered in our guide to the metabolic numbers to know after 40.

Our eGFR reference page carries the KDIGO categories with their source and effective date, and the lab result lookup will place a specific value against them.

There is no single ICD-10 code for a low eGFR, and which code appears on your paperwork tells you what has and has not been concluded. The codes below are from the FY 2027 ICD-10-CM, which applies to encounters from October 1, 2026 to September 30, 2027.

The official outpatient coding guidelines explain when R94.4 is used instead of a CKD code. Codes for signs and abnormal findings are acceptable when a diagnosis has not been established, and a diagnosis recorded as "probable", "suspected" or "rule out" is not coded at all; the abnormal test result is coded instead. That lines up with KDIGO's rule that one low result is not yet a diagnosis. An N18 code means a clinician has documented chronic kidney disease, and it is fair to ask which results the stage rests on and whether they were at least three months apart.

Arrange review for any eGFR below 60 that has not been repeated, for any result below 45, and urgently for a rapid fall, reduced urine output, swelling, or confusion. Bring your previous results so the trend is visible. Two questions get you most of what a nephrologist would want to know: has my urine albumin-to-creatinine ratio been measured, and has this value been confirmed at least three months apart.