Creatine is studied for depression because brain-scan studies report altered energy metabolism in some people with the condition, and creatine is part of the system the brain uses to buffer energy. That is a reason to run trials, not a result. The trials that have been run are mostly small, mostly add-ons to an existing treatment, and mostly in women. Here is what they found.
The first-screen answer
| Study | Who took part | Creatine dose | Length | Result against placebo |
|---|---|---|---|---|
| Eckert et al., 2025 — meta-analysis | 11 trials, 1,093 participants | As given in each trial | Varied | SMD −0.34 (95% CI −0.68 to −0.00) ≈ 2.2 Hamilton points; below the 3.0-point minimal important difference; very low certainty |
| Lyoo et al., 2012 | 52 women with major depression, all on escitalopram | 5 g a day | 8 weeks | Greater Hamilton improvement from week 2, held at weeks 4 and 8 |
| Sherpa et al., 2025 | 100 adults with depression, all receiving CBT | Not stated in the abstract | 8 weeks | PHQ-9 fell in both arms, 5.12 points more on creatine |
| Toniolo et al., 2018 | 35 people with bipolar depression, on their usual medication | 6 g a day | 6 weeks | Main score no different (p = 0.560); remission 52.9% vs 11.1%; two switches to hypomania or mania on creatine |
Every trial in the table added creatine to something else: an antidepressant, therapy or existing medication. None tested creatine on its own as a treatment for depression.
The pooled estimate: real, small and uncertain
The 2025 meta-analysis from Porto Alegre is the most complete summary. It searched four databases to February 2025 for trials comparing creatine with placebo, with or without diagnosed depression, and pooled eleven of them. The standardised effect, −0.34, translates to about 2.2 points on the 17-item Hamilton scale. The minimal important difference the authors used is 3.0 points, so the average effect sits below what is usually taken as noticeable to a patient. The confidence interval runs almost to zero.
Three more findings shape how to read it. Heterogeneity was substantial (I² = 71.3%), so the trials disagree with each other. Effects looked larger in people with clinical depression, but the authors' bias checks pointed to substantial bias favouring creatine. Remission was more frequent on creatine across three trials (odds ratio 3.60), while treatment response across two trials was not (0.72). The authors' conclusion is that the true effect "may be trivial or null" and that the evidence is very uncertain.
The trials one at a time
Escitalopram, 2012. The Seoul and Utah trial is the one most often cited. Fifty-two women with major depressive disorder were randomised to escitalopram plus creatine 5 g a day or escitalopram plus placebo for eight weeks. Hamilton scores improved more on creatine as early as week 2, and the gap held at weeks 4 and 8. Dropout was higher on creatine (8 of 25, 32.0%) than placebo (5 of 27, 18.5%), a difference the authors report as not significant; total adverse events were 36 on creatine and 45 on placebo. The trial enrolled only women, and its size limits how far its result carries.
CBT, 2025. The pilot in Uttarakhand gave 100 adults (half women, average age about 30) eight weeks of cognitive-behavioural therapy plus creatine or placebo. PHQ-9 scores fell in both groups, which is expected with therapy; they fell by 5.12 points more with creatine. Discontinuation and adverse events were comparable. The trial was published as a feasibility and exploratory study, and when a critique asked whether creatine was useful with CBT, the authors' reply was titled "Preliminary evidence should be treated as such".
Bipolar depression, 2018. The São Paulo trial randomised 35 people in a bipolar depressive episode, already on medication, to creatine 6 g a day or placebo for six weeks. The main outcome, change on the Montgomery-Åsberg scale, did not differ (p = 0.560). Remission was more frequent on creatine: 52.9% against 11.1% in everyone randomised. Two patients on creatine switched to hypomania or mania early in the trial.
The bipolar warning
The 2018 switches were not the first. A 2007 open study of ten people with treatment-resistant depression reported that both participants with bipolar disorder developed hypomania or mania on creatine; its anxiety scores sit on the anxiety evidence page. Four people across two studies cannot give a rate, and the 2018 abstract does not report how many on placebo switched, so the comparison that would matter is not available from it. The warning recurs in the later reviews found in this page's search. Other conditions where a clinician's view comes first are set out on creatine and medical conditions.
What the registry holds
A ClinicalTrials.gov search on 10 November 2026, with creatine as the intervention and depression as the condition, returned 26 records. Read one by one:
| What the record tests | Records |
|---|---|
| A creatine-containing intervention in depression | 20 |
| — completed | 10 |
| — withdrawn before enrolling | 5 |
| — terminated early | 3 |
| — recruiting | 2 |
| Something else the search matched (psychedelic, probiotic, ATP and nutritional-formula studies) | 6 |
The twenty include add-on trials with antidepressants, studies in adolescent girls with treatment-resistant depression, trials in women using methamphetamine, an ECT augmentation pilot and two 5-HTP-plus-creatine studies, one of them still recruiting (NCT05895747). A 2016 dose-ranging study from that adolescent programme gave 2, 4 or 10 g a day for eight weeks and found frontal-lobe phosphocreatine rose by 4.6%, 4.1% and 9.1%, against a 0.7% fall on placebo, with no group difference in adverse events or serum creatinine (Kondo et al., 2016). Two of the withdrawn records tested a combined cytidine-and-creatine drug for bipolar depression. Eight of twenty withdrawn or terminated is a high share; the records do not always say why.
Diet and depression: an association
The 2020 NHANES analysis linked dietary creatine with depression questionnaires in 22,692 US adults. Depression prevalence was 10.23 per 100 in the lowest quartile of dietary creatine and 5.98 in the highest, with an adjusted odds ratio of 0.68. The association was strongest in women, adults aged 20 to 39 and people not on antidepressants. Dietary creatine comes mostly from meat and fish, so it travels with diet, income and activity. The study adjusted for several of these but cannot show that creatine itself lowers risk. How much creatine common foods carry is on foods high in creatine.
What is not established
- Whether creatine works on its own. Every trial here added it to another treatment.
- Whether the effect is clinically meaningful. The pooled average sits below the usual threshold, and the bias checks point the other way.
- Effects in men. The most cited trial enrolled only women, and several registered trials are in women or adolescent girls.
- Interaction safety with antidepressants. No formal interaction study exists.
- The mania risk in bipolar disorder. Reported in four people across two studies; no rate can be given.
Who should ask a clinician first
Anyone being treated for depression, anyone with bipolar disorder, and anyone taking an antidepressant, mood stabiliser or other psychiatric medicine. Depression treatment decisions belong with a clinician, and these trials were run with clinicians monitoring every participant. The broader cognition claims sit on creatine and the brain claims, sleep on creatine and sleep, and drug questions on how creatine fits with prescriptions. This page reports what the studies measured; it does not say whether anyone should take creatine for depression.
Sources and dates
- Eckert I, Lima J, Dariva AA. Creatine supplementation for treating symptoms of depression: a systematic review and meta-analysis. Br J Nutr. 2025;134(11):947-959. PubMed 41189312. Opened 2026-11-10.
- Lyoo IK, Yoon S, Kim TS, et al. A randomized, double-blind placebo-controlled trial of oral creatine monohydrate augmentation for enhanced response to a selective serotonin reuptake inhibitor in women with major depressive disorder. Am J Psychiatry. 2012;169(9):937-945. PubMed 22864465. Opened 2026-11-10.
- Sherpa NN, De Giorgi R, Ostinelli EG, et al. Efficacy and safety profile of oral creatine monohydrate in add-on to cognitive-behavioural therapy in depression: an 8-week pilot, double-blind, randomised, placebo-controlled feasibility and exploratory trial in an under-resourced area. Eur Neuropsychopharmacol. 2025;90:28-35. PubMed 39488067. Opened 2026-11-10. Authors' reply: PubMed 39879837.
- Toniolo RA, Silva M, Fernandes FBF, et al. A randomized, double-blind, placebo-controlled, proof-of-concept trial of creatine monohydrate as adjunctive treatment for bipolar depression. J Neural Transm. 2018;125(2):247-257. PubMed 29177955. Opened 2026-11-10.
- Roitman S, et al. Creatine monohydrate in resistant depression: a preliminary study. Bipolar Disord. 2007;9(7):754-758. PubMed 17988366 (as cited on this site's anxiety page).
- Kondo DG, et al. Creatine target engagement with brain bioenergetics: a dose-ranging phosphorus-31 magnetic resonance spectroscopy study of adolescent females with SSRI-resistant depression. Amino Acids. 2016. PubMed 26907087. Opened 2026-11-10.
- Bakian AV, Huber RS, Scholl L, Renshaw PF, Kondo D. Dietary creatine intake and depression risk among U.S. adults. Transl Psychiatry. 2020;10(1):52. PubMed 32066709. Opened 2026-11-10.
- ClinicalTrials.gov API v2,
query.intr=creatine,query.cond=depression, read 2026-11-10: 26 records, each classified for this page by its title and status. - PubMed title search "creatine AND (depression OR depressive)", 2026-11-10: 49 records, used to find the trials above. Abstracts read on PubMed; figures as printed there.
