Creatine Raised a Kidney Blood Test. Two 2026 Reviews Ask What Actually Changed.

This article was prepared with AI assistance and has not yet been reviewed by a qualified medical or scientific professional. It is for general information, not medical advice.

A number goes up on a blood panel. Another number, calculated from the first, goes down. The report suggests that the kidneys are filtering less well.

Except there is a complication: the person taking the test has started using creatine. The supplement can increase production of creatinine, the chemical the laboratory measures. More creatinine in the blood can mean less is being removed by the kidneys. It can also mean more is being made. A routine estimate cannot always separate those explanations.

That is the puzzle behind two systematic reviews published in 2026. They do not establish that creatine is safe for every patient or that an abnormal kidney result should be ignored. They reveal why a seemingly straightforward laboratory signal can contain two different stories.

Two reviews, one recurring mismatch

A systematic review published July 27, 2026 brought together 26 randomized trials with 1,036 participants. Compared with placebo, creatine supplementation was associated with an average 0.14 mg/dL increase in serum creatinine (95% confidence interval 0.05 to 0.22). That result supports a genuine shift in the blood marker across the included studies, though its size varied considerably.

When the researchers pooled studies using creatinine-based methods to estimate filtration, the result looked worse: an apparent decrease of 10.75 mL/min (95% CI −17.48 to −4.02). But the studies measuring filtration using a different tool—a chromium-51 EDTA tracer, rather than a calculation driven by creatinine—did not show a statistically significant reduction. Their pooled difference was +5.89 mL/min, with a confidence interval from −0.30 to +12.08. That interval still allows some uncertainty; it does not prove that filtration improved.

The same review found no statistically significant pooled differences in urinary protein or albumin outcomes. Those tests provide context because abnormal leakage of proteins into urine can be another sign of kidney disease. But the absence of a pooled signal is not proof that kidney injury cannot occur, and the quality of evidence for some outcomes was low.

An earlier review published April 24, 2026 combined 19 randomized controlled trials and one randomized crossover study. It reported a mean serum-creatinine increase of 0.13 mg/dL (95% CI 0.07 to 0.18), with no statistically significant difference in urea or estimated glomerular filtration rate (eGFR) in its pooled analyses. The investigators specifically called for trials lasting longer than a year.

The two reviews are not two independent experiments. They likely draw on many of the same published trials, with different inclusion rules and analytical choices. Their agreement on creatinine is useful; it should not be mistaken for twice as many independent participants.

The most revealing contrast is within the larger review itself: a blood marker shifted while a non-creatinine measure of filtration did not show the same decline.

A measurement is not the thing it measures

Creatine helps muscles regenerate a rapidly usable energy compound during brief, intense effort. A portion of the body’s creatine pool converts spontaneously to creatinine. That creatinine enters the blood and is cleared substantially by the kidneys.

A clinician can measure serum creatinine directly. But a standard eGFR result is usually a calculation that uses creatinine together with characteristics such as age, rather than a direct measurement of how many milliliters of blood the kidneys filter each minute. Its accuracy depends partly on assumptions about the rate at which a body makes creatinine.

Imagine measuring how efficiently a room is emptying by counting how many people are inside. If more people start entering, the room looks more crowded even if the exit is working at the same speed. That analogy cannot diagnose anyone’s kidneys. It does show why a concentration can change through production, clearance, or both.

Supplementation can alter production; high muscle mass and changes in muscle tissue can also complicate interpretation. Conversely, a real decline in kidney filtration can elevate creatinine. Those possibilities are not mutually exclusive. It would be a mistake to dismiss every higher creatinine result in someone taking creatine as a harmless laboratory artifact.

This is precisely why the July review separated filtration estimates derived from creatinine from tracer-based measurements. The calculation and the tracer do not ask identical questions. If the input to a formula has changed for reasons outside the process the formula estimates, the output can mislead.

The alternative tests are useful, not magical

The 2024 KDIGO chronic-kidney-disease guidance describes circumstances in which creatinine-based estimates are less reliable, including situations involving unusual muscle mass. Where cystatin C is available, it recommends using creatinine and cystatin C together for relevant assessment, and emphasizes additional testing when a more accurate filtration estimate would change care.

Cystatin C is a different blood marker used to estimate filtration. It is not a universal answer to every clinical uncertainty. Its usefulness depends on the patient and context, and a measured GFR test may sometimes be needed. A question about a laboratory value is best resolved by interpreting the pattern, previous measurements, medical history and other findings—not by treating any single alternative test as infallible.

The evidence base is thinner here than the phrase “26 trials” might suggest. The July review found only two trials that actually assessed cystatin C. Tracer-measured filtration data came from a smaller subset than the overall trial count. The key comparison is informative but less extensively studied than serum creatinine itself.

What the studies cannot promise

The July review judged 11 of its 26 included trials at low risk of bias, 14 as raising some concerns, and one as high risk. Results for serum creatinine were highly heterogeneous—participants, protocols, doses and study designs differed. Four trials followed supplementation for more than six months. Most of the evidence therefore cannot settle uncommon or very long-latency harms.

The authors also acknowledged an especially important gap: their review did not include well-powered supplementation trials in people with nondialysis chronic kidney disease across its stages. Limited data in people receiving hemodialysis do not close that gap. It would be an unsupported leap to convert findings largely obtained in healthier trial populations into an assurance for someone with kidney disease, complex medical treatment, or multiple potentially kidney-affecting exposures.

Even in healthy people, controlled studies showing no significant change in certain markers are not guarantees of safety under all circumstances. They provide estimates within the tested populations, interventions and durations. The April review likewise called for longer-term trials.

These limitations also cut the other way. Case reports of abnormal creatinine are not sufficient by themselves to establish that creatine caused kidney damage. Both alarming stories and reassuring ones must confront the same central question: what was actually measured, and what other explanations remain?

How to read the next kidney panel

The practical lesson is not “ignore creatinine.” It is to avoid making the diagnosis from creatinine alone, especially when something that changes creatinine production is present.

A careful interpretation asks whether the value is different from a previous baseline; whether the calculated eGFR relies on that same changed marker; whether other blood, urine or clinical findings point in the same direction; and whether exercise, muscle mass, supplements, medicines or illness might matter. A clinician can decide whether additional testing is useful. The best next step depends on the actual clinical situation, not a generic testing or supplement protocol from an article.

There is an appealing simplicity to a number printed in bold. It feels like an answer. The two 2026 reviews suggest something more valuable: a number can be correct as a measurement and still incomplete as an explanation.

Creatinine may rise because the body is making more, because the kidneys are clearing less, or because both are occurring. The task is not to choose the most comforting story. It is to distinguish them.

Research reporting only. This article does not provide personal medical advice or a recommendation to begin, stop or change supplements or treatment.