Comprehensive Metabolic Panel (CMP)

What Comprehensive Metabolic Panel (CMP) actually measures — the analyte, and the assay

Fourteen analytes, four unrelated chemistries, and one of the fourteen is not measured at all. Glucose comes off a hexokinase or glucose-oxidase reaction; creatinine off either a Jaffe picrate reaction or an enzymatic cascade; the transaminases off coupled NADH-consumption assays read as catalytic activity; albumin off a dye binding. eGFR is a calculation performed on the creatinine, your age and your sex.

The creatinine method changes the answer, and it changes diagnoses. Jaffe measures the color formed when creatinine meets alkaline picrate, and a long list of other things in serum form that color too — ketones, bilirubin, proteins, some cephalosporins. Enzymatic methods use creatininase and are far more specific. Both are calibrated to be traceable to isotope-dilution mass spectrometry, and calibration does not fix non-specificity. Across 41,144 parallel measurements, the Jaffe result averaged 0.07 mg/dL higher than the enzymatic one; in 19% of cases the two methods assigned a different CKD stage, and 98% of those disagreements had Jaffe staging the patient as more severe Boss 2023.

The eGFR equation changed in 2021 and the change was not cosmetic. The CKD-EPI equations were refitted without a race coefficient, validated in 4,050 participants across 12 studies. The old race-inclusive creatinine equations overestimated measured GFR in Black participants by a median 3.7 mL/min/1.73 m², and the combined creatinine-plus-cystatin-C equation without race was more accurate than either single-marker equation and showed smaller differences between groups Inker 2021. Which equation your report used is not printed on it Delanaye 2023.

The transaminase assays carry a quieter version of the same problem. The IFCC methods for ALT and AST exist in versions with and without added pyridoxal-5'-phosphate, the vitamin B6 cofactor the enzymes require. Without it, a B6-depleted sample under-reports enzyme activity that is genuinely there. Albumin has the same fork: bromocresol green binds other proteins and over-reads at low albumin, while bromocresol purple reads lower. Neither the P5P status nor the dye is on the report.

Comprehensive Metabolic Panel (CMP): what changes the blood, and what only changes the reading

What changes the analytes in your blood — grouped, because a CMP is five tests wearing one name:

1. Creatinine: muscle, first and largest. Creatinine is the spontaneous breakdown product of muscle creatine phosphate, so its generation rate is proportional to skeletal muscle mass, and the literature on it is explicit that it doubles as a body-composition marker with a highly variable individual generation rate Avila 2025. A 105 kg lifter and a 48 kg 80-year-old with identical kidneys have very different creatinines.
2. Creatine supplementation, which raises creatinine without touching the kidney. Supplemental creatine converts non-enzymatically to creatinine at a fixed daily rate, so serum creatinine rises; the narrative review of controlled trials concludes creatine is safe in healthy people and that the elevation reflects spontaneous conversion, not dysfunction Longobardi 2023.
3. A large cooked-meat meal. Heating meat converts its creatine to creatinine, which is then absorbed. This is a genuine analyte change and it is diet, not filtration.
4. A hard training session. Skeletal muscle contains ALT and a great deal of AST, and eccentric work releases both alongside creatine kinase; the transaminases on a post-training draw can be muscle-derived in their entirety.
5. Hepatic fat, for ALT. The commonest cause of a mildly raised ALT in the general population, and the reason the reference interval is contested Valenti 2021 Makker 2021.
6. Fasting state, for glucose — and note that the high-normal band is not inert: in 1,125 non-obese people with normal glucose metabolism, a fasting glucose at or above 5.3 mmol/L (about 95 mg/dL) carried a 2.1-fold risk of incident prediabetes over a median 3.7 years, and a fatty liver index at or above 16.5 carried a 2.2-fold risk Aizawa 2022.
7. Androgens, for the kidney. Anabolic-androgenic steroid use is linked with focal segmental glomerulosclerosis, while high protein intake and moderate creatine are low-risk in people whose kidneys are normal Tidmas 2022.

What changes only the reading — and this list decides more CMPs than most people realize:

1. Jaffe versus enzymatic creatinine, worth 0.07 mg/dL on average and a whole CKD stage in 19% of people Boss 2023.
2. Which eGFR equation, and whether it still carries a race coefficient Inker 2021 Delanaye 2023.
3. Ketones interfering with Jaffe. A fasted or ketogenic lifter is exactly the person whose acetoacetate inflates a picrate creatinine, and exactly the person who will read it as kidney damage.
4. Hemolysis, which raises potassium, AST and LD out of the red cells themselves and has no biological meaning whatsoever.
5. Fist clenching, a long tourniquet, cold storage or delayed separation, all of which raise potassium.
6. Pyridoxal-5'-phosphate in the transaminase reagent, and bromocresol green versus purple for albumin.
7. Posture and hydration, which move albumin, calcium and total protein by several percent through plasma volume alone.

Reference interval or decision threshold — which kind of number Comprehensive Metabolic Panel (CMP) is

This panel mixes both kinds of number on one page, which is why it is so easy to misread.

The eGFR bands are decision thresholds. G1 to G5 are administrative categories defined against outcomes, and the definition carries a duration criterion that the report never prints: an abnormality has to persist for at least three months before it is chronic kidney disease. A single eGFR of 58 is not stage 3 anything Delanaye 2023.

Creatinine's interval is a reference interval, and it is a reference interval for muscle. The population it came from had ordinary muscle mass, so a heavily muscled person sits above it while filtering perfectly, and a sarcopenic person sits inside it while filtering badly Avila 2025. The direction of the error is opposite in the two groups, and the second error is the dangerous one: in 145 patients with primary neuromuscular disease measured against iohexol clearance, every equation overestimated kidney function, by 22 to 60 mL/min/1.73 m², with the cystatin C-based estimates carrying the smallest bias Aldenbratt 2022.

The ALT reference interval is the most important sentence on this page. The upper limit of normal was derived from populations that included people with undiagnosed fatty liver, so it is an interval for how common liver fat is rather than a boundary of liver health. When it was re-derived in 21,296 apparently healthy blood donors aged 18 to 65 with no risk factors for liver disease, using the IFCC standardized method, the updated sex-specific upper limits came out at 42 U/L in men and 30 U/L in women — roughly 30% lower than the then-recommended values, and the lower limits detected steatosis (odds ratio 2.31, 1.40–3.80) and significant fibrosis (odds ratio 3.35, 1.19–9.42) in people with metabolic risk Valenti 2021. In 5,455 living liver donors with histologically normal livers the proposed limits were lower still, at 34 U/L for men and 22 U/L for women Choi 2024. A laboratory flag at 55 U/L is not a statement that 50 is fine.

Fasting glucose is a threshold too, and the same continuity applies below it Aizawa 2022.

How you would know your Comprehensive Metabolic Panel (CMP) was wrong — and when to redraw

Three analytes, three different clocks, and one retest interval for the panel is wrong.

The transaminases clear in days. ALT's plasma half-life is roughly 47 hours and AST's roughly 17, so a training-induced rise is largely gone in a week. Do not draw within 48 to 72 hours of hard training, and if enzymes are up, redraw after seven rested, sober days before drawing any conclusion.

Creatinine reaches a new steady state in about 7 to 10 days after a real change in filtration — which is why a creatinine drawn during an acute illness is not a baseline. And the chronic question has a hard-coded interval: three months, because that is what separates acute kidney injury from chronic kidney disease Delanaye 2023.

Conditions that must match: the same laboratory and the same creatinine method Boss 2023; the same eGFR equation Inker 2021; the same fasting duration; no creatine loading started or stopped between draws Longobardi 2023; no large meat meal the evening before; 48 to 72 hours clear of training; and a sample that was not hemolyzed.

What would have to change for the retest to mean something. Each analyte has a companion that tells you which explanation is true:

• Creatinine up? Run a cystatin C. Cystatin C is not made by muscle, so a raised creatinine with a normal cystatin C is muscle or creatine, and a raised creatinine with a raised cystatin C is the kidney Inker 2021 Aldenbratt 2022.
• ALT and AST up? Run a creatine kinase and a GGT. High CK with a normal GGT points at muscle; a high GGT with the transaminases points at the liver.
• AST above ALT? The De Ritis ratio is a real signal — in 2,728 participants with chronic kidney disease a higher AST/ALT ratio tracked all-cause mortality Li 2025 — but with a normal CK and no alcohol history it is a reason to look harder, not a diagnosis.
• Potassium up? Ask whether the sample was hemolyzed before asking anything about the patient.

What Comprehensive Metabolic Panel (CMP) cannot tell you

eGFR is an estimate of a measurement nobody performed. The reference standard is measured clearance of an exogenous filtration marker such as iohexol, and against it the estimating equations can be wrong by tens of mL/min in specific groups Aldenbratt 2022 Delanaye 2023.

A single eGFR cannot distinguish acute from chronic. The difference is a three-month duration criterion, which no single blood draw contains Delanaye 2023.

A normal creatinine does not mean a normal kidney. Creatinine sits flat across a wide span of falling filtration before it rises — the creatinine-blind range — and in a muscular person the flat part is even wider, because a high generation rate masks a falling clearance Avila 2025 Tidmas 2022.

A normal ALT does not exclude liver disease, and the numbers on this are stark. In 59 people with type 2 diabetes and normal transaminases, elastography found hepatic steatosis in 81%, and 12% had advanced fibrosis, stage F3 or F4, against none in the control group Makker 2021. The enzyme reports leakage; fibrosis is scarring, and scarring is silent on this panel.

The panel does not include the kidney test that matters most. Albuminuria is a urine test, and a normal eGFR with significant albuminuria is still chronic kidney disease. A CMP alone cannot rule out kidney disease.

The electrolytes are homeostatically defended, which means they stay normal until the system that defends them is failing. A normal sodium and potassium is weak evidence of anything.

The wrong inference readers actually draw is one of two, and they are opposites. The muscular lifter reads a creatinine of 1.35 mg/dL as kidney damage and panics, when the confirmatory test costs little and answers it Longobardi 2023 Aldenbratt 2022. And the person with an ALT of 38 U/L reads the laboratory's 55 U/L flag as permission, when the healthy-population limit is closer to 30 and the difference between those two numbers is where fatty liver lives Valenti 2021 Choi 2024.

Sources read for these sections

• Boss K, et al. Effect of serum creatinine difference between the Jaffe and the enzymatic method on kidney disease detection and staging. Clinical Kidney Journal 2023 · PMID 37915891
• Inker LA, et al. New Creatinine- and Cystatin C-Based Equations to Estimate GFR without Race. New England Journal of Medicine 2021 · PMID 34554658
• Avila M, et al. The Metabolism of Creatinine and Its Usefulness to Evaluate Kidney Function and Body Composition in Clinical Practice. Biomolecules 2025 · PMID 39858438
• Longobardi I, et al. Is It Time for a Requiem for Creatine Supplementation-Induced Kidney Failure? A Narrative Review. Nutrients 2023 · PMID 36986197
• Tidmas V, et al. Nutritional and Non-Nutritional Strategies in Bodybuilding: Impact on Kidney Function. International Journal of Environmental Research and Public Health 2022 · PMID 35409969
• Delanaye P, et al. Diagnostic standard: assessing glomerular filtration rate. Nephrology Dialysis Transplantation 2023 · PMID 37950562
• Valenti L, et al. Definition of Healthy Ranges for Alanine Aminotransferase Levels: A 2021 Update. Hepatology Communications 2021 · PMID 34520121
• Choi J, et al. Updated Reference Intervals for Alanine Aminotransferase in a Metabolically and Histologically Normal Population. Clinical Gastroenterology and Hepatology 2024 · PMID 38750867
• Makker J, et al. Prevalence of advanced liver fibrosis and steatosis in type-2 diabetics with normal transaminases: A prospective cohort study. World Journal of Gastroenterology 2021 · PMID 33642826
• Aizawa T, et al. Hepatic Steatosis and High-Normal Fasting Glucose as Risk Factors for Incident Prediabetes. Journal of the Endocrine Society 2022 · PMID 35958436
• Li Q, et al. Association between De-Ritis ratio (AST/ALT) and mortality in patients with chronic kidney disease: a retrospective cohort study. Scientific Reports 2025 · PMID 40113904
• Aldenbratt A, et al. Estimation of kidney function in patients with primary neuromuscular diseases: is serum cystatin C a better marker of kidney function than creatinine?. Journal of Nephrology 2022 · PMID 34351595