Renal function tests: urea, creatinine and eGFR

This topic is about reading the three numbers that most commonly say how well the kidneys are clearing waste: blood urea nitrogen (BUN, reported as urea), serum creatinine and the estimated glomerular filtration rate (eGFR). It is a small laboratory panel whose meaning comes from how the parts fit together. The kidneys filter urea and creatinine mainly by glomerular filtration, so their blood levels rise when filtration falls; because both are also pushed around by muscle mass, diet, drugs and the hydration state, the panel is interpreted in combinations rather than in single values. The reader should leave this page able to explain what each number measures, state the usual reference intervals with their units, list common reasons for high and low values, name the common patterns and what they suggest, stage chronic kidney disease (CKD) and acute kidney injury (AKI), and work through a realistic result sheet.

Throughout, units are international (SI) first, with conventional units in brackets where useful, and the illustrative sample result sheets use obviously invented numbers that are labelled as teaching cases. Every statement about a threshold, cause or rule is tied to a source that was opened. Where guidelines differ, both positions are stated. The content is educational and is not a substitute for clinical judgement or local guidelines.

Urea (usually reported as BUN) is the nitrogen-containing waste produced when protein is broken down. The liver makes it from amino acid nitrogen and the kidneys excrete it. Blood urea therefore reflects both how much protein is being broken down (diet, catabolism, tissue breakdown) and how well the kidneys clear it. Because urea is handled largely by glomerular filtration with little regulation or adaptation as function falls, it is a rough mirror of glomerular filtration rate (GFR), but it is a less accurate index of filtration than creatinine and it rises earlier in renal disease ^[3]. Urea is also reabsorbed in the proximal tubule, which is why it is more variable than creatinine with changes in hydration and perfusion ^[3].

Creatinine is a breakdown product of creatine phosphate in skeletal muscle. It is produced at a fairly steady rate that depends on muscle mass, with a small extra contribution from cooked meat, which can raise the blood level for a short time ^[2]. It is freely filtered at the glomerulus and is not reabsorbed, but a little is secreted by the tubules; some drugs block that secretion and raise serum creatinine without any real fall in GFR. Because its production is relatively stable and it is not reabsorbed, creatinine is a more accurate assessment of renal function than urea ^[2]. Its weakness is that it is generated from muscle, so it overestimates function in people with little muscle and underestimates it in people with high muscle mass ^[2].

eGFR is not a measured substance; it is an estimate of the glomerular filtration rate calculated from serum creatinine, age and sex. The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation is the one recommended for adults, used with a creatinine assay calibrated to standardised reference material and a zero-biased enzymatic method ^[5]. The estimate is reliable when the underlying assumptions (muscle mass of a typical adult of that age and sex, steady state) hold, and less reliable in pregnancy, acute kidney injury, very high or very low muscle mass, oedema, amputation and extremes of diet or supplement use ^[5]. NICE also advises that estimates become less accurate as the true GFR rises, and cautions against over-interpreting values of 60 mL/min/1.73 m2 or more ^[5]. eGFR is reported as a whole number when it is 90 mL/min/1.73 m2 or less, and as "greater than 90 mL/min/1.73 m2" when above that ^[5]. ^[1][7]

Cystatin C is a small protein produced by all nucleated cells at a fairly constant rate. It is freely filtered and almost completely catabolised by the tubules, so its blood level reflects filtration and is not influenced by muscle mass or diet in the way creatinine is. Because it is more expensive and less widely available, creatinine-based eGFR remains the first test, with cystatin C reserved for confirming kidney function when creatinine may be misleading (for example, abnormal body habitus, malnutrition, low-protein diet, malignancy, cirrhosis, catabolic states or steroid use) ^[4][6]. Estimates that combine creatinine and cystatin C are the most accurate equations without a race term and show the smallest differences between race groups ^[4].

The panel therefore measures: filtration (urea, creatinine, eGFR), muscle protein turnover (creatinine), protein breakdown and hydration (urea), and, when requested, an alternative filtration marker (cystatin C). It does not measure concentrating ability, tubular acid handling or imaging, which need different tests.

Intervals vary with age, sex, method and population, so always use the interval printed by the laboratory that measured the sample; the numbers below are one open catalog's adult intervals and are only a guide ^[1].

Creatinine changes very little with hydration, so a trend in one patient over time, repeated from the same laboratory, is more reliable than absolute comparison with population limits ^[2]. eGFR is unreliable in pregnancy; creatinine is interpreted against pregnancy-adjusted intervals and trend rather than against population limits ^[5].

High values point either to reduced excretion or to increased production. Low values point to reduced production or dilution. The lists below are drawn from open patient information leaflets and keep the same distinction ^[2][3].

Notes that matter in practice:

• Muscle mass and age. Creatinine is generated by muscle, so it tends to be higher in people with larger muscle mass and lower in older adults, women, amputees and people who are malnourished or bed-bound. A "normal" creatinine in a frail older woman may already mean reduced filtration. An older man whose creatinine rises only modestly may still have a meaningful fall in eGFR, because muscle mass and age set the baseline; repeat the test and calculate eGFR rather than react to one value ^[2][5].
• Diet. A meat-rich meal can raise creatinine briefly and a high-protein diet raises urea. Advising patients to avoid meat for about 12 to 24 hours before the test is a routine precaution ^[2][5].
• Hydration and renal perfusion. Dehydration raises both, urea more than creatinine, because urea is reabsorbed when the proximal tubule reabsorbs sodium and water ^[3].
• Drugs that block tubular secretion. Cimetidine, trimethoprim and other agents that inhibit tubular secretion of creatinine can raise serum creatinine without a true fall in GFR. Starvation, severe liver disease and a low-protein intake lower urea ^[3]. ^[2]
• Other non-renal causes. Gut bleeding and catabolic states raise urea; burns raise BUN; liver disease lowers it ^[3].

A practical opening question before reacting to an abnormal creatinine is therefore: how much muscle does this person have, have they eaten meat recently, are they dehydrated, and are they taking cimetidine, trimethoprim or another secreting blocker? ^[2][3]

The combinations, not single values, make the diagnosis. The following are the patterns a student should be able to name and explain ^[3][5].

• Prerenal (reduced perfusion). Both urea and creatinine rise, but urea rises out of proportion to creatinine, so the urea:creatinine ratio is high. Urine output falls and the urine is concentrated. Common causes are dehydration, blood or fluid loss, heart failure, sepsis and burns ^[3].
• Renal (intrinsic parenchymal damage). Both rise together in rough proportion, with a normalish ratio. Think glomerulonephritis, acute tubular necrosis, interstitial nephritis and nephrotoxic drug injury ^[3].
• Postrenal (obstruction). Both rise, often abruptly, with a variable ratio. Suspect prostatic enlargement, stones, tumour, retroperitoneal fibrosis, or any blockage. NICE flags hyperkalaemia, severe uraemia, acidosis or fluid overload as reasons for urgent treatment of obstruction rather than waiting ^[5].
• Raised creatinine with a relatively low urea. This suggests increased creatinine generation rather than reduced filtration: high muscle mass, bodybuilding, recent heavy meat intake, or rhabdomyolysis. Recent heavy exercise also fits here ^[2].
• Low urea with low or normal creatinine. Reduced production: low-protein intake, malnutrition, pregnancy, low muscle mass or severe liver disease ^[2][3].
• Normal creatinine, low eGFR. In an older adult with small muscle mass, filtration can be substantially reduced while creatinine stays inside the printed range. Calculate eGFR rather than trust creatinine alone ^[5].
• Discordant markers. When creatinine-based eGFR and cystatin C-based eGFR disagree, body habitus, nutrition or a drug that blocks secretion is usually the explanation, and the cystatin C value is often the truer estimate of filtration ^[4][6].
• Bicarbonate and potassium. As GFR falls, potassium rises because the kidneys excrete most of the daily potassium load; hyperkalaemia becomes a real risk, and therapy with renin-angiotensin system antagonists should be avoided if the baseline potassium is greater than 5.0 mmol/L and stopped if it reaches 6.0 mmol/L or more ^[5]. Serum bicarbonate falls and a metabolic acidosis develops in progressive CKD; management is with alkali supplements and a plant-predominant diet, as summarised in the KDIGO-based review ^[6]. Anaemia becomes more likely once eGFR falls below 30 mL/min/1.73 m2 ^[5].

NICE recommends the CKD-EPI creatinine equation for adults, with IDMS-traceable, specific (preferably enzymatic) creatinine assays ^[5]. The race-free equations published in 2021 are now preferred where available, because race is a social rather than a biological construct and adding it created systematic differences in estimated GFR between groups ^[4]. The equations are power models in which creatinine is divided by a sex-specific constant (kappa; 0.7 for females and 0.9 for males), capped or floored at 1, then raised to fitted exponents, multiplied by an age factor raised to age and a sex factor, with no race factor in the 2021 versions ^[4]. Laboratories implement validated versions, so clinicians rarely calculate by hand, but it helps to know what is inside:

• CKD-EPI creatinine (2009, with race term): the version used historically, with a race factor for Black patients; it systematically overestimated GFR in Black participants ^[4].
• CKD-EPI creatinine (2021, race-free): the same structural model without any race factor; accurate enough for many situations, but it can introduce small biases in both race groups and differential bias between them ^[4].
• CKD-EPI cystatin C: a filtration marker model with no race term; nearly unbiased in Black and non-Black groups, but less accurate than the combined equations ^[4].
• CKD-EPI creatinine-cystatin C (2021, race-free): the combined model without race; this is the most accurate equation without a race term and has the smallest differences between race groups ^[4].

Limits worth stating plainly ^[4][5]:

• eGFR equations assume steady state; they are not valid in acute kidney injury.
• They are assumed to be less accurate once the true GFR is above 60 mL/min/1.73 m2, and NICE advises caution in that range.
• They fail or mislead in extremes of muscle mass, pregnancy, oedema, amputation, very high protein or meat intake, and supplement use.
• Cystatin C is preferable when creatinine generation is abnormal (muscle wasting, high muscle mass, diet), but it costs more and is not universally available.
• For key thresholds such as drug dosing and organ donation, a measured reference GFR using inulin, iothalamate or iohexol is more accurate than any estimate.

AKI is an abrupt reduction in kidney function, reducing urine output and/or raising creatinine. KDIGO defines it as any of the following: an increase in serum creatinine of 0.3 mg/dL (26.5 micromol/L) or more within 48 hours; an increase in serum creatinine to at least 1.5 times baseline, known or presumed to have occurred within the previous 7 days; or urine volume below 0.5 mL/kg/h for 6 hours ^[6]. Staging then uses the rise relative to baseline and urine output ^[6].

KDIGO defines CKD as a structural or functional abnormality of the kidney lasting at least 3 months, most practically an eGFR under 60 mL/min/1.73 m2 or an albumin to creatinine ratio (ACR) of at least 3 mg/mmol (about 30 mg/g), plus other markers of kidney damage ^[6][7]. Staging then combines a GFR category (G1 to G5) with an albuminuria category (A1 to A3), and both components multiply the risk of progression, cardiovascular events and death ^[5][7].

NICE (NG203) uses the same CKD-EPI creatinine equation and the same G and A categories, and it also sets out practical points: eGFR is reported as a whole number up to 90 or as "greater than 90"; a confirmed ACR of 3 mg/mmol or more is clinically important proteinuria; recheck an ACR of 3 to 70 mg/mmol on a repeat early-morning sample; and proteinuria is monitored with ACR rather than protein:creatinine ratio in adults ^[5]. NICE also notes that hyperkalaemia, severe uraemia, acidosis and fluid overload make obstruction an emergency ^[5]. ^[7]

All values are invented for teaching and are not real patient data. Reference intervals are the illustrative ones used earlier.

Sheet 1 - within normal limits (male adult, age 34)

1. All analytes sit inside the printed intervals and the eGFR is above 90 mL/min/1.73 m2 ^[1][5].
2. There is no proteinuria by ACR and no reduced eGFR, so CKD is not present by KDIGO criteria (decreased GFR under 60 or ACR 3 mg/mmol or more, confirmed) ^[6][7].
3. No action is needed; repeat the panel only as clinically indicated.

Clinical conclusion: this is a normal renal panel with preserved filtration; no follow-up testing is required beyond usual care.

Sheet 2 - chronic kidney disease with hyperkalaemia (male adult, age 61)

1. Both urea and creatinine are raised in proportion, which fits renal parenchymal disease rather than prerenal or postrenal patterns ^[3].
2. eGFR 26 mL/min/1.73 m2 places this in GFR category G4 (15-29 mL/min/1.73 m2), and the urine ACR of 180 mg/g places it in albuminuria category A2 (30-300 mg/g) ^[5][7].
3. Potassium 5.4 mmol/L is above the interval and bicarbonate 19 mmol/L is low, so hyperkalaemia and a metabolic acidosis accompany the reduced filtration ^[1][6].
4. Anaemia becomes more likely once eGFR falls below 30 mL/min/1.73 m2, so check haemoglobin and iron status ^[5].
5. NICE recommends confirming a new eGFR under 60 mL/min/1.73 m2 by repeating the test within 2 weeks in an adult not previously tested, and allowing for plus or minus 5 percent biological and analytical variability when interpreting changes ^[5].

Clinical conclusion: this is G4 CKD (eGFR 26 mL/min/1.73 m2) with moderately increased albuminuria (A2), hyperkalaemia and metabolic acidosis; the next step is to confirm the eGFR, arrange a nephrology review, check haemoglobin and bone profile, and review drugs (avoiding renin-angiotensin blockade while potassium is above 5.0 mmol/L).

Sheet 3 - acute kidney injury in a previously normal man (male adult, age 55)

1. Creatinine has risen 3.3-fold from baseline, which meets the KDIGO AKI criterion of an increase to at least 1.5 times baseline within 7 days, and reaches Stage 3 as a 3-fold rise (with creatinine under 4.0 mg/dL / 353.6 micromol/L) ^[6].
2. The rise in both urea and creatinine with marked hyperkalaemia (6.2 mmol/L) and acidosis (bicarbonate 16 mmol/L) is an emergency pattern ^[5][6].
3. eGFR equations are not valid in acute states, so the panel marks it not reportable ^[5].
4. Look for prerenal volume depletion, nephrotoxic drugs (notably NSAIDs, aminoglycosides, contrast), obstruction and sepsis, and arrange urinalysis, renal ultrasound if indicated and an urgent clinical review ^[6].

Clinical conclusion: this is KDIGO Stage 3 acute kidney injury with hyperkalaemia and acidosis; the next step is urgent medical assessment, correction of life-threatening hyperkalaemia, fluid and drug review, and investigation of the cause.

• Treating creatinine as a measure of kidney disease in isolation. A small or malnourished older adult can have a "normal" creatinine with a markedly reduced GFR. Always calculate eGFR and consider cystatin C when the body habitus is unusual ^[2][5].
• Forgetting meat and drugs before interpreting a result. Recent meat raises creatinine and high-protein intake raises urea for a day or two; cimetidine, trimethoprim and other secretory blockers can raise creatinine without any real renal injury ^[2][3].
• Using eGFR during acute illness. All creatinine-based estimates assume a steady state and are not valid in acute kidney injury, pregnancy or extremes of muscle mass. A highly accurate measure (inulin, iothalamate or iohexol) is needed for chemotherapy monitoring or living-donor evaluation ^[5].
• Ignoring the printed interval on the report. Methods differ between laboratories; only cholesterol, creatinine and HbA1c are internationally harmonised, and creatinine still shows method-to-method variation. Always compare like with like, from the same laboratory ^[1].
• Overlooking potassium and bicarbonate. A creatinine that is only mildly raised can sit beside a dangerous potassium of 6.0 mmol/L or more. Renin-angiotensin system antagonists should be avoided with a baseline potassium above 5.0 mmol/L and stopped if potassium reaches 6.0 mmol/L or more ^[5].
• Stopping after one abnormal result. NICE wants a repeat of a new eGFR under 60 mL/min/1.73 m2 within 2 weeks, accounting for plus or minus 5 percent variability, before deciding ^[5].