ICU creatinine study should not scare creatine users

Key Takeaways

• In 952 ICU patients aged 90+, admission creatinine was not independently linked with in-hospital mortality.
• Higher admission creatinine was independently associated with 90-day and 1-year mortality in this very elderly ICU cohort.
• Removing the kidney component from the SOFA score did not change short-term predictive performance, suggesting limited short-term prognostic value of creatinine here.
• This study examined serum creatinine in critically ill nonagenarians, not creatine supplements in healthy or athletic people.
• Creatine users should not confuse creatinine, a lab marker, with creatine, the supplement.
• If you take creatine and are getting bloodwork, tell your clinician so results are interpreted in context.

What the study actually found

This BMC Geriatrics retrospective cohort study looked at 952 critically ill patients aged 90 and older admitted to ICU at a single German academic centre between 2008 and 2019. The main question was whether serum creatinine measured at ICU admission still works as a useful prognostic marker in very elderly patients, whose low muscle mass and frailty can make creatinine harder to interpret.

The headline finding was nuanced. Admission creatinine was not independently associated with in-hospital mortality after adjustment for confounders, with an adjusted hazard ratio of 1.07 and a 95% confidence interval of 0.92 to 1.26. In other words, for short-term hospital survival, creatinine on arrival did not add much by itself.

But higher admission creatinine did independently predict worse longer-term outcomes. The adjusted hazard ratio was 1.16 for 90-day mortality and 1.20 for 1-year mortality. Mortality in this population was high overall: 28.5% in hospital, 42.2% at 90 days, and 57.1% at 1 year.

The authors also tested whether the kidney component of the SOFA score was helping with short-term risk prediction. It apparently was not. The full SOFA score and a modified SOFA score with the kidney component removed both had an AUC of 0.79.

That is an important geriatric-critical-care finding. It is not evidence that taking creatine products raises mortality, harms kidneys, or causes ICU risk in the general population.

Why creatinine is not the same thing as creatine

For supplement users, the key point is simple: creatinine is a blood marker; creatine is a nutrient-like compound used as a supplement. They are related, but they are not interchangeable.

Creatinine is a breakdown product formed from creatine and phosphocreatine. Clinicians commonly use serum creatinine as part of kidney-function assessment, but it is an imperfect marker because it is influenced by muscle mass, age, hydration, illness severity, and other factors. That is exactly why this study matters in nonagenarians: people in their 90s often have lower muscle mass and altered physiology, so a given creatinine value may not mean the same thing it would in a younger adult.

This distinction is especially relevant because creatine supplementation can sometimes slightly affect lab interpretation without indicating kidney damage. As the broader literature explains, creatine use may change creatinine-related readings in some circumstances, while creatine monohydrate remains the most studied form and is generally considered safe in healthy people when used appropriately. For a deep evidence review, see the ISSN position stand and the review of common creatine questions and misconceptions.

So if a headline about “creatinine” makes creatine users nervous, the first question should be: Who was studied, and what exactly was measured? In this paper, the answer is critically ill adults aged 90+ in ICU, and the biomarker was admission serum creatinine.

How the study was designed and where caution is needed

This was a retrospective cohort study, which means the researchers looked back at existing clinical data rather than randomly assigning any treatment or exposure. That design is useful for studying real-world ICU populations, especially very old adults who are hard to enrol in trials, but it also limits how confidently we can interpret cause and effect.

Several limitations matter here:

• Single-centre data: all patients came from one institution, which may limit generalisability.
• Observational design: higher creatinine may be a marker of underlying illness burden rather than a direct driver of later mortality.
• Very specific population: these findings apply to critically ill nonagenarians, not younger adults, gym-goers, or typical creatine supplement users.
• Admission value only: the study focused on creatinine at ICU admission, not serial trends, supplement exposure, or dietary creatine intake.

The authors also built adjustment models for confounders, which strengthens the analysis, but residual confounding is always possible in retrospective ICU research.

For readers considering the best creatine option for training or healthy ageing, this is therefore the wrong paper to use as a verdict on supplementation safety. It tells us something clinically useful about prognosis in a very elderly ICU population. It does not test whether creatine supplementation changes mortality, kidney outcomes, or ICU admission risk.

That distinction is not a technicality; it is the central point of accurate interpretation.

What this means in practice for people taking creatine

For most readers, the practical takeaway is reassuring: this study should not change standard evidence-based creatine use. If you are a healthy adult using creatine monohydrate for strength, power, lean mass, or training capacity, the mainstream evidence base remains centred on forms and doses already well established in sports nutrition.

Typical approaches include:

• Maintenance dosing: about 3 to 5 g/day of creatine monohydrate.
• Optional loading: about 20 g/day split into 4 doses for 5 to 7 days, followed by maintenance.

If you want a personalised starting point, our creatine dosage calculator can help, and our creatine guides cover the basics of timing, hydration, and consistency.

Where this study does matter is around bloodwork and communication. If you take creatine and have kidney-related labs checked, tell your clinician. A serum creatinine result always needs context, including age, body size, muscle mass, medications, medical history, hydration, and acute illness. In frail older adults especially, creatinine can be an imperfect proxy for kidney function.

For older adults with chronic kidney disease, complex medical conditions, or recent hospitalisation, supplementation decisions should be discussed with a physician or pharmacist who knows the full clinical picture. But nothing in this paper suggests healthy supplement users should panic, stop taking creatine, or assume elevated creatinine automatically means harm from creatine.

How this fits with the broader creatine evidence

The broader evidence base on creatine supplementation says something very different from what a quick read of this ICU paper might imply. Reviews and position stands have consistently concluded that creatine monohydrate is the most studied form and is generally safe and effective for improving high-intensity exercise performance and supporting gains in strength and lean mass when used appropriately.

That does not mean creatine is risk-free for every person in every context. It means safety questions have to be answered with studies that actually examine creatine supplementation, not papers about creatinine as a prognostic biomarker in critically ill populations.

This new study adds something valuable to medicine: in adults aged 90+ in ICU, admission creatinine may be less useful for short-term risk stratification than clinicians often assume, while still carrying information about 90-day and 1-year prognosis. That supports more age-adapted interpretation of kidney biomarkers in very old patients.

For supplement consumers comparing creatine brand reviews or browsing our creatine product catalog, the bigger message is to separate hospital biomarker research from supplement efficacy and safety research. Those are related domains, but they answer different questions.

In plain language: this paper is about how doctors interpret a lab value in very elderly ICU patients, not about whether creatine supplementation is harmful.

Bottom line

The study’s main news is straightforward: in critically ill adults aged 90 and older, admission creatinine did not independently predict in-hospital death, but it did predict worse 90-day and 1-year survival. It also suggests that the creatinine-based kidney component of the SOFA score adds little to short-term prediction in this very old ICU population.

For creatine users, the bottom line is just as straightforward: do not confuse creatinine with creatine. This paper does not show that creatine supplements damage kidneys, increase mortality, or become unsafe because a hospital study used creatinine as a prognostic marker.

The practical response is not panic; it is precision. Use creatine monohydrate in evidence-based doses, buy from reputable products rather than chasing exotic formulas, and make sure your healthcare team knows if you supplement before interpreting lab work. If you are choosing a product, start with reputable best creatine rankings and transparent formulations.

That is the fairest reading of the evidence: an important ICU geriatrics finding, but not a warning against creatine supplementation for the general public.

What this creatinine study found and what it did not

• 952 ICU patients studied — All were aged 90 or older
• 28.5% In-hospital mortality — Admission creatinine was not independently associated with this outcome
• 42.2% / 57.1% 90-day and 1-year mortality — Higher admission creatinine was independently associated with both
• 3-5 g/day Typical creatine maintenance dose — Mainstream evidence-based guidance for creatine monohydrate

Frequently Asked Questions

Sources & Further Reading

• Kreider et al. (2017), ISSN Position Stand: Safety and Efficacy of Creatine — JISSN
• Antonio et al. (2021), Common questions and misconceptions about creatine supplementation — JISSN