What is normal GFR?
This is an article for those of you who want to know more about normal human GFR (glomerular filtration rate). It is a very controversial area [“Oh no, not another one” CKDEx Ed].
Normal eGFR number is 90-120 ml/min (in adults; according to the CKD/GFR classification – the higher the GFR, the better the kidney function, the better.
What is GFR used for?
The GFR is used to:
- Diagnose CKD
- Monitor – changes in kidney function over time
- Decide – whether you need dialysis or a kidney transplant.
But does is this normal range true?
But, the big but. This number range is so neat, it is clearly made up (wrong) – as are all the CKD stages, as determined by ‘overly neat ranges’ of GFR.
For example CKD3 is said to be 30-59 ml/min (split into two CKD3B, which is ‘worse’ than CKD3A). This is also made up folks [“shush, don’t tell the patients!” Ed].
To make things even more confusing, we (kidney doctors) invent another arbitrary sub-classification (on the right of the often reproduced diagram above) based on arbitrary amounts of albumin (a protein) in the urine. The idea is the more albumin the urine, the ‘worse’ (apparently). This puts patients in a table like above. All made up, again, folks.
So what is normal GFR? It’s complicated.
The truth? Mm ..
- We (kidney doctors) actually don’t know what is ‘normal’ human GFR. We do know (or think we know!) ..
- Average human GFR, adjusted for body surface area, is about 120 ml/min/1.73 m2. No one knows the exact number. And this is an average, from a range of 100-130 mls/min – not all humans are the same. Also there will be normal humans with a normal GFR of 90 or 140 mls/min.
- The GFR is adjusted for body surface area since bigger kidneys can filter more fluid, and kidneys get bigger as body surface area gets bigger
- One small but careful study of normal fit Swedish 27 year olds (roughly 50:50 men vs women) came out with an average of 117 ml/min (range 106-129) (Sterner, 2007). In this study, a technique called inulin was used to measure GFR. This is a more accurate (but laborious) method than the ones we use for patients now, which are mathematical estimates based on the blood creatinine level. Hence you sometimes see an ‘e’ before your GFR, written as ‘eGFR’. This stands for ‘estimated GFR’
- We also don’t know ..
– Whether GFR reflects % kidney function. It does not .. as the kidneys have seven functions and GFR/clearance (of toxins) is just one of them. So GFR should not be converted to (and expressed) as a % kidney function
– What GFR indicates what severity of CKD
– What level of albumin in urine is ‘OK’.
But we have to start somewhere, with some basic ‘rules’, that are vaguely right and meaningful for most patients (and us!) most of the time – the ‘80% rule’.
And, before we are too self-critical, no other branch of medicine (e.g. heart, lung, liver, joints, brain, skin) has come up with any reliable number to quantify the function of ‘their’ organ. What does 71.2% brain or 67.3% heart function mean? Exactly.
Effect of Age?
Oh no, don’t ask [“well they have, answer the question please docs” Ed]. GFR may (may) decline with age, even in people without kidney disease. See chart below for average estimated GFR used by some doctors, based on age.
| Age (years) | Average eGFR |
| 20–29 | 116 |
| 30–39 | 107 |
| 40–49 | 99 |
| 50–59 | 93 |
| 60–69 | 85 |
| 70+ | 75 |
Is this true? Sort of. But the scientific evidence for such tables is not strong. For example, such tables are made from measuring GFR – using a more accurate method than normal – in different age groups of different people. No one has tracked GFR through the lives of the same people.
Anyway, say it’s true, how do we know it reflects a decline in kidney function over a lifetime? Alternatively, it could reflect loss of muscle mass (and usage) through life; as GFR is derived from blood creatinine level that is based on muscle mass/usage, not just kidney function – the bigger the muscles, and the more they are used, the higher is the creatinine.
In other words, kidney function/GFR could be quite stable through life (unless you develop true CKD). And the ‘age-related’ fall of GFR just reflects muscle ageing.
Effect of Age and Gender?
Right. Now we are getting there [“so where exactly is ‘there’ team?” Ed].
The graphs below are taken from a 2007 study, when GFR was measured in many 100s of patients, who were thought be well and normal (without CKD). They show, like the table above, an age related decline in GFR (that may not be a disease, i.e. the computer says they have CKD but they do not).
Most patients will have GFRs within the middle three lines. But they also show slightly lower GFRs in women at all age groups, especially when younger (see 18-24 age range). Like the effect of age, this ‘difference’ could also relate to muscle bulk; as, at that age, males will be significantly more muscular than their female contemporaries.
Effect of GFR equation used
And here is another issue. Unfortunately there are several competing equations, as shown in a study of 799 potential living kidney donors who were well, youngish and presumed to have 2 normal kidneys (Giron-Luque, 2022). Three GFR equations (Cockcroft–Gault, MDRD, and CKD-EPI) were compared to a ‘gold standard method’ (there is no perfect one by the way) called ‘creatinine clearance (Cr Cl)’. This is a more accurate but laborious way of measuring GFR, like the inulin method described above.
As you can see from this scattergram below (of individual patients), comparing one method (MDRD) to Cr Cl, there is a wide spread, around the 120 ml/min line. But when the other GFR equations are used (see the paper) different scattergrams occur. We do know all three equations tend to underestimate GFR by upto 20%. In other words, the number you hospital tells you may actually be higher, suggesting your kidneys are worse than they actually are.
In the Sterner Swedish study above, estimated GFR was on average 23 ml/min (yes, a lot!) lower than the ‘actual’ GFR (on Cr Cl).
So. What is normal human GFR?
- We don’t know for certain, partly as
- The world uses different GFR equations, based on different chemical tests for blood creatinine level
- Most importantly, there isn’t one.
Why? ‘What is normal GFR?’ is like saying ‘what is normal height?’. It is a biological variable with a normal range. Its measurement is not easy and all methods have problems. So normal GFR is (probably) about 120 ml/min +/- 20 ml/min. But this is affected by age, gender, race (which we have not mentioned, that has an effect too) and how you measure it.
Can I trust the GFR number I am told?
Yes, largely .. if it is measured in the same laboratory that your GP/hospital uses. But rather than focusing on the actual number, it is best to focus on the change in number over time. That has more meaning for you.
Summary
We have described what is a normal GFR, and shown that age/gender and other factors make a difference (probably). But it’s complicated.
Nonetheless the whole ‘CKD/GFR/albumin in urine thing’ is a useful language – and has some truth/meaning – for most patients (and their doctors and nurses) most of the time. But we still need doctors to interpret and explain it all for their patients.
[“and themselves, as they don’t all agree!” CKDEx Ed].
Other resources
CKDEx has linked articles on what should your GFR be at 70 years old? and the meaning and limitations of GFR.
And this is a good review of a range of studies that have tried to answer the question posed at the start of this article: Delanaye, 2012
Last Reviewed on 25 June 2024