eGFR (Creatinine)

Your kidneys filter your blood all day, and creatinine-based eGFR is the routine way labs estimate how well they are doing it. It appears on most metabolic panels. By the time kidney trouble causes symptoms, filtration may have been falling for years.

eGFR stands for estimated glomerular filtration rate. The creatinine version is a calculation, not a direct measurement. It uses the creatinine in your blood, your age, and your sex to estimate filtration. That shortcut is useful. It is also where most of the confusion comes from.

What the Number Actually Reflects

Creatinine is a waste product your muscles release at a fairly steady rate. Healthy kidneys clear it, so when creatinine rises, the estimated filtration rate falls. But creatinine depends on more than filtration. Muscle mass, diet, age, sex, and some medicines all shape how much of it is in your blood, so the same creatinine can mean different true kidney function in two people.

This is why the number can miss early loss. Serum creatinine often needs a large fall in filtration before it rises enough to stand out. As kidney function worsens, the kidneys also clear more creatinine through tubules. Tubules are small kidney channels. This side route can make creatinine-based estimates read higher than measured filtration in more advanced disease.

Death and Kidney Failure Risk

Lower filtration tracks a higher chance of dying, and the gradient is steep. In a pooled analysis of general-population cohorts, compared with a study reference eGFR near 95, the risk of death from any cause was about 18% higher at an eGFR of 60, about 57% higher at 45, and roughly three times higher at 15. The link held after accounting for urine protein and standard heart-disease risk factors.

A 2023 meta-analysis took this further. Across 27.5 million people in 114 cohorts, lower creatinine-based eGFR was tied to higher rates of ten outcomes, including kidney failure, heart attack, stroke, heart failure, hospitalization, and death. In cohorts of people already diagnosed with kidney disease, the gradient steepened once eGFR fell below 45, with each further drop carrying substantially higher risk of both death and kidney failure.

The Speed of the Drop Matters

A single value tells you where you are; the direction of travel tells you more. In a consortium of over 1.7 million people, a 30% fall in eGFR over two years came with more than five times the risk of kidney failure, and a 57% fall with about thirty-two times the risk. This held across ages, whether or not people had diabetes, and at different starting levels. It is the strongest argument for watching your own trend rather than fixating on one reading.

Heart Disease and Stroke

Kidney disease and artery disease often travel together. Lower eGFR is linked with more coronary disease, heart failure, atrial fibrillation, and peripheral artery disease. For stroke, a UK Biobank study of 429,566 people followed for about ten years found that total and ischemic stroke risk rose as eGFR fell below 75. The study did not find the same adjusted link for bleeding strokes. The creatinine-based signal for stroke was weaker than the cystatin C version, which is less tied to muscle.

Where the Pattern Breaks: Cancer

It would be easy to assume every bad outcome rises as this number falls, but that is not what the data show. In 431,263 UK Biobank participants followed for a median of 11.3 years, creatinine-based eGFR was not associated with cancer incidence or cancer death after full adjustment. This is not a general "everything gets worse" marker. Its power is specific to kidney and cardiovascular outcomes, and reading it as a global health score overstates what it can do.

When a High Number Is Not Reassuring

Most people assume higher filtration is always better. With creatinine-based eGFR, that assumption can mislead you in two directions. A very high number can reflect low muscle mass, which means less creatinine and an inflated estimate, rather than unusually strong kidneys. In coronary-disease cohorts, both very low and very high creatinine-based eGFR tracked higher death rates, and a UK Biobank analysis found elevated creatinine-based eGFR linked to more cardiovascular disease, not less.

The way to hold these facts together is to stop treating this as a simple good-number, bad-number test. It is a creatinine readout turned into a filtration estimate. The middle of the range often means healthy kidneys with typical muscle. The extremes can mean kidney trouble, or they can mean the creatinine input is distorted. A surprising number is a reason to look closer, not to celebrate or panic.

The Muscle Mass Problem

The biggest source of error runs through muscle. Because creatinine comes from muscle, a person with less of it produces less creatinine, and the formula reads that as better kidney function than they actually have. Older adults, frail people, and those losing muscle are the classic case. Studies show creatinine-based estimates can overstate true filtration and miss kidney disease in these groups.

The reverse happens too. A muscular athlete, or someone taking creatine, can carry more creatinine and post a lower eGFR without kidney damage. In hospital data, women and older patients with low muscle were more likely to have kidney injury underdiagnosed when only creatinine was used. If your body composition is unusual in either direction, this number deserves a second marker before you trust it.

When Cystatin C Sharpens the Picture

Cystatin C is another blood marker used to estimate kidney filtration. It is much less tied to muscle than creatinine, though thyroid disease, inflammation, steroid use, and some other factors can still affect it. When creatinine and cystatin C disagree, the combined estimate is usually most accurate. The disagreement also carries risk information: people whose cystatin C estimate is worse than their creatinine estimate tend to have worse outcomes. In the study that developed the combined equation, adding cystatin C reclassified some people with a creatinine eGFR of 45 to 59 upward, to a measured filtration above 60; in everyday practice it can also move people into a worse category, so the shift is not always toward reassurance.

The 2021 CKD-EPI equation removed race from the calculation. Race is a social category, not a kidney measurement. But removing it did not erase error. In one large evaluation, the 2021 creatinine equation underestimated measured filtration in Black patients by about 6 units. A creatinine-only estimate is a starting point, not a verdict.

Why One Reading Is Not Enough

A single eGFR carries real noise. Pooled data put the reference change value between about 12.5% and 16.5%, so a small swing can come from normal biological variation and lab differences alone, with no real change in your kidneys. Chase one number and you will react to static. Watch a series and you can see whether you are stable, drifting down, or dropping fast, which is the finding that predicts trouble.

Filtration and protein leak are different kinds of kidney injury. eGFR can look okay while urine albumin is already abnormal, and urine albumin can be low while filtration is falling. That is why the most useful read is usually eGFR plus urine albumin, with cystatin C added when creatinine looks suspect.

When Results Can Be Misleading

Several things move this number without moving true kidney filtration in the same direction. The ones below are the most likely to fool you.

• Muscle mass: low muscle from age, frailty, or illness makes the number read higher than your kidneys deserve; heavy muscle can push it lower with no kidney problem.
• Recent meat, hard exercise, or creatine: a heavy cooked-meat meal, intense training, or creatine supplements can raise creatinine and make eGFR look lower for a short time.
• Acute illness and dehydration: a temporary rise in creatinine during illness or after a dry spell can drop the number for days, then recover on its own.
• Creatinine-handling medications: trimethoprim, cimetidine, cobicistat, dolutegravir, fenofibrate, and some targeted cancer drugs can raise creatinine without a matching fall in measured filtration. In one palbociclib case, creatinine eGFR fell to 12 while cystatin C eGFR stayed around 33 to 35.
• Testosterone-based gender-affirming therapy: in people assigned female at birth, it can raise creatinine and lower the estimate without a clear rise in chronic kidney disease risk in one population study.