Saffron and Mood: What the Clinical Trials Measured
Saffron (Crocus sativus L.) has been tested in randomized trials for mild to moderate depression since 2004. An earlier version of this article claimed the herb works as well as antidepressants with fewer side effects, and I have rewritten it because the trials cannot carry a claim that strong. Below I cover who was enrolled, what was scored, how saffron compared with fluoxetine, imipramine, and placebo, and where the evidence stops.
Key takeaways
- Tóth and colleagues pooled nine randomized trials in 2019; Yang and colleagues pooled twelve in 2018. Most enrolled 30 to 60 adults and ran six weeks.
- Against placebo, saffron lowered depression scores with a pooled effect size (Hedges' g) of 0.89.
- Head-to-head trials against imipramine and fluoxetine found no statistically significant difference. With 15 to 20 people per group, those trials were too small to prove equivalence.
- Every trial in the 2019 meta-analysis was run in Iran, and the first comparator trial used saffron donated by a commercial producer.
- Ayurvedic and Persian traditions used saffron for low mood long before any trial. I treat that history as context rather than proof.
- Saffron is not a replacement for a prescribed antidepressant. Any change to medication belongs with the prescriber.
This article is educational and is not medical advice, diagnosis, or treatment. Statements have not been evaluated by the FDA. If you are in crisis or thinking about harming yourself, call or text 988 in the United States or contact local emergency services.
What the trials measured
Evidence: moderate. Multiple randomized controlled trials with small samples and short follow-up.
Nearly every saffron trial used the Hamilton Rating Scale for Depression, a clinician-scored questionnaire covering mood, sleep, appetite, energy, and related symptoms. Some added the Beck Depression Inventory, which the participant fills out. Researchers recorded a score at enrollment, repeated it at set intervals, and compared the size of the drop between groups.
Participants were adult outpatients who met DSM-IV criteria for major depression at mild to moderate severity. The 2004 imipramine comparison required a baseline Hamilton score of at least 18. Most protocols excluded people with severe depression, active suicidal thinking, psychosis, or other psychiatric medication, so the results apply to the group that was enrolled and say nothing about severe or treatment-resistant depression.
Two outcomes matter when reading these papers. A response usually means the score dropped by half, while remission means the score fell below a set threshold, often 7 on the Hamilton scale. A trial can show a larger average drop without showing that more people reached remission, so I look for both numbers before drawing a conclusion.
Saffron against antidepressants
Evidence: moderate, from small trials. No difference detected is not the same as equivalence shown.
Shahin Akhondzadeh's group at Tehran University of Medical Sciences ran the first comparison in 2004. Thirty outpatients took either 30 mg per day of saffron stigma extract or 100 mg per day of imipramine for six weeks. Hamilton scores fell in both groups, and the difference between them did not reach statistical significance. The imipramine group reported more dry mouth and sedation, which is expected with a tricyclic drug. The saffron used in the study was donated by Novin Zaferan Co., a commercial producer, which I note because it lowers the independence of the finding.
The same group compared 30 mg per day of stigma extract with 20 mg per day of fluoxetine in 40 patients in 2005 and again found no significant difference over six weeks. A 2007 trial used saffron petal rather than stigma, 15 mg twice daily against fluoxetine, in 40 outpatients with the same result. Both drug doses sit at the low end of what prescribers use, which shapes how the comparison should be read.
These results are often summarized as saffron matching the drugs, but that reading skips a step, because with 15 to 20 people per arm a trial can only detect a large gap between treatments. A moderate advantage for the drug would go unnoticed at that size. Proving two treatments are equivalent requires a non-inferiority design with a prespecified margin and several hundred participants, and none of the saffron trials were built that way.
The 2019 meta-analysis by Barbara Tóth and colleagues at the University of Szeged pooled the comparator trials and found a small effect favoring the drugs, with a p value of 0.053. The authors described saffron as non-inferior to the SSRIs tested. I read the same number as a result sitting at the edge of significance, drawn from trials too small to settle the question.
Saffron against placebo
Evidence: moderate. Consistent direction across trials, nearly all from one country, with short follow-up.
The placebo comparisons are more informative. Akhondzadeh's group published a 40-patient trial in 2005 in which 30 mg per day of stigma extract lowered Hamilton scores more than placebo over six weeks. A 2006 trial from the same group tested petal extract against placebo with a similar outcome. Mohsen Mazidi and colleagues published a 60-person trial in 2016 using 100 mg per day for twelve weeks, the longest placebo comparison to date, and measured both depression and anxiety with Beck inventories.
Adrian Lopresti's team in Australia tested a standardized extract sold as affron in adolescents aged 12 to 16, at 14 mg twice daily for eight weeks. Self-reported depression and anxiety improved more than placebo, while parent-reported scores did not separate from placebo. It remains one of the few trials run outside Iran.
Tóth's meta-analysis pooled nine trials and found a Hedges' g of 0.89 against placebo, a large effect by convention. Yang and colleagues pooled twelve trials in 2018 and reached the same direction. Tóth's review noted that every included trial came from Iran and that many were led by the same investigators, which limits how far the finding generalizes.
Saffron and anxiety
Evidence: limited. Anxiety was a secondary outcome in most trials.
Anxiety has fewer dedicated trials. Mazidi's 2016 study scored it with the Beck Anxiety Inventory and reported a larger drop with saffron than placebo. Lopresti's adolescent trial found the same pattern in self-reports. In the remaining trials, anxiety was either unmeasured or recorded as a single item on the Hamilton scale. I would describe the anxiety evidence as early and nothing more.
Proposed mechanisms
Evidence: preliminary. Drawn from animal and cell studies, not from measurements in the human trials.
Saffron stigma carries crocin, crocetin, picrocrocin, and safranal. Rodent studies and cell assays show these compounds can reduce serotonin reuptake, lower markers of oxidative stress, and blunt the cortisol response to stress. Adrian Lopresti and Peter Drummond reviewed this work in 2014 and concluded that any antidepressant effect likely runs through several pathways at once rather than one.
None of the human depression trials measured serotonin, cortisol, or inflammatory markers. The mechanisms are plausible and unproven in people, and I present them as hypotheses rather than explanations.
Traditional use in Ayurveda and Persian medicine
Evidence: traditional use. Documented in classical texts; not a measure of effect size.
Saffron appears in Ayurvedic materia medica under the names kesar and kumkuma, classified as balancing to all three doshas and grouped with herbs used for the mind and the heart. It was combined with milk, ghee, or other warming herbs rather than taken alone. Persian medicine recorded saffron as a mufarrih, a substance taken to lift the spirit, and Avicenna's Canon of Medicine lists it under that heading.
These records show sustained use for low mood across centuries and two separate traditions. They do not tell me the effect size, the failure rate, or how the preparations compared with anything else. In my own practice I keep the history in view when I read the trials, and I do not use it to fill the gaps in them.
Doses used in the trials
Evidence: descriptive. These are study protocols, listed for interpretation, and not recommendations.
Each figure below is tied to the paper that used it so you can read the studies with the numbers in hand.
- 30 mg per day of stigma extract, divided through the day, in Akhondzadeh 2004, Akhondzadeh 2005, and Noorbala 2005.
- 15 mg twice daily of petal extract in Akhondzadeh Basti 2007.
- 14 mg twice daily of the affron stigma extract in Lopresti 2018.
- 100 mg per day in Mazidi 2016.
- 30 mg per day of isolated crocin added to an SSRI in Talaei 2015.
The trials used capsules of dried extract standardized to crocin and safranal content. Culinary saffron threads in a recipe are a different material at an uncontrolled dose, and the trial results do not transfer to them.
Saffron has a known toxic threshold. Doses above 5 grams have caused serious poisoning, and classical texts treat saffron as a uterine stimulant and exclude it during pregnancy. Trial doses sit far below that threshold, though the pregnancy exclusion stands at any dose.
Limits of the research
Evidence: this section assesses study design rather than saffron itself.
- Every trial pooled in the 2019 meta-analysis was conducted in Iran, and one research group produced most of the early data. Replication by independent teams in other countries is thin.
- Sample sizes ran from 30 to 80. Trials this small can miss moderate differences and can overstate effects by chance.
- Follow-up lasted six, eight, or twelve weeks. There is no data on relapse, long-term use, or what happens after stopping.
- Comparator drugs were given at low-end doses, which favors a finding of no difference.
- Preparations varied between stigma, petal, and proprietary extracts with different crocin content, so the trials are not testing one product.
- At least one trial received its saffron from a commercial supplier. Sponsorship does not invalidate a result, and it does lower my confidence in it.
- Severe depression, suicidal ideation, and people on other psychiatric medication were excluded, so the findings do not reach those groups.
If you take an antidepressant
Evidence: safety guidance. One small adjunct trial exists; interaction data is limited.
Do not stop, reduce, or replace a prescribed antidepressant on the basis of this article. Stopping abruptly can cause withdrawal symptoms and a return of depression, and the saffron trials excluded people who were changing medication.
Combining saffron with an SSRI has one small trial behind it. Ali Talaei and colleagues added 30 mg per day of crocin to fluoxetine, sertraline, or citalopram in 40 patients for four weeks and reported greater improvement than placebo. Four weeks in 40 people cannot rule out an interaction, and the animal data on serotonin reuptake gives a reason for caution. If you want to try saffron alongside a prescription, bring the trial list from this article to your prescriber and make the decision together.
Common questions
Evidence: short answers drawn from the sections above.
Is saffron as effective as antidepressants?
The trials that compared saffron with imipramine and fluoxetine found no statistically significant difference, but they enrolled 30 to 40 people and ran six weeks, which is too small to prove equivalence. The pooled comparison in the 2019 meta-analysis leaned slightly toward the drugs at p = 0.053.
How long did the saffron depression trials last?
Most ran six weeks. The adolescent affron trial ran eight weeks, and the Mazidi trial ran twelve. No trial has followed participants beyond twelve weeks.
Which part of the saffron plant was used?
Most trials used dried stigma extract. Two trials from Akhondzadeh's group used petal extract, and the affron trials used a proprietary stigma extract standardized to its crocin content.
Does saffron help with anxiety?
Two trials measured anxiety directly and reported larger reductions than placebo. Anxiety was a secondary outcome elsewhere, so the evidence is early.
Can I take saffron with an SSRI?
One 40-person, four-week trial added crocin to SSRIs without reported problems, and that is not enough to establish safety. Ask your prescriber before combining them.
- Akhondzadeh S, Fallah-Pour H, Afkham K, Jamshidi AH, Khalighi-Cigaroudi F. Comparison of Crocus sativus L. and imipramine in the treatment of mild to moderate depression: a pilot double-blind randomized trial. BMC Complement Altern Med. 2004;4:12. doi:10.1186/1472-6882-4-12
- Noorbala AA, Akhondzadeh S, Tahmacebi-Pour N, Jamshidi AH. Hydro-alcoholic extract of Crocus sativus L. versus fluoxetine in the treatment of mild to moderate depression: a double-blind, randomized pilot trial. J Ethnopharmacol. 2005;97(2):281-284. doi:10.1016/j.jep.2004.11.004
- Akhondzadeh S, Tahmacebi-Pour N, Noorbala AA, et al. Crocus sativus L. in the treatment of mild to moderate depression: a double-blind, randomized and placebo-controlled trial. Phytother Res. 2005;19(2):148-151. doi:10.1002/ptr.1647
- Moshiri E, Basti AA, Noorbala AA, Jamshidi AH, Hesameddin Abbasi S, Akhondzadeh S. Crocus sativus L. (petal) in the treatment of mild-to-moderate depression: a double-blind, randomized and placebo-controlled trial. Phytomedicine. 2006;13(9-10):607-611. doi:10.1016/j.phymed.2006.08.006
- Akhondzadeh Basti A, Moshiri E, Noorbala AA, Jamshidi AH, Abbasi SH, Akhondzadeh S. Comparison of petal of Crocus sativus L. and fluoxetine in the treatment of depressed outpatients: a pilot double-blind randomized trial. Prog Neuropsychopharmacol Biol Psychiatry. 2007;31(2):439-442. doi:10.1016/j.pnpbp.2006.11.010
- Talaei A, Hassanpour Moghadam M, Sajadi Tabassi SA, Mohajeri SA. Crocin, the main active saffron constituent, as an adjunctive treatment in major depressive disorder: a randomized, double-blind, placebo-controlled, pilot clinical trial. J Affect Disord. 2015;174:51-56. doi:10.1016/j.jad.2014.11.035
- Mazidi M, Shemshian M, Mousavi SH, et al. A double-blind, randomized and placebo-controlled trial of saffron (Crocus sativus L.) in the treatment of anxiety and depression. J Complement Integr Med. 2016;13(2):195-199. doi:10.1515/jcim-2015-0043
- Lopresti AL, Drummond PD, Inarejos-García AM, Prodanov M. affron, a standardised extract from saffron (Crocus sativus L.) for the treatment of youth anxiety and depressive symptoms: a randomised, double-blind, placebo-controlled study. J Affect Disord. 2018;232:349-357. doi:10.1016/j.jad.2018.02.070
- Lopresti AL, Drummond PD. Saffron (Crocus sativus) for depression: a systematic review of clinical studies and examination of underlying antidepressant mechanisms of action. Hum Psychopharmacol. 2014;29(6):517-527. doi:10.1002/hup.2434
- Tóth B, Hegyi P, Lantos T, et al. The efficacy of saffron in the treatment of mild to moderate depression: a meta-analysis. Planta Med. 2019;85(1):24-31. doi:10.1055/a-0660-9565
- Yang X, Chen X, Fu Y, et al. Safety and efficacy of saffron (Crocus sativus L.) for treating mild to moderate depression: a systematic review and meta-analysis. J Nerv Ment Dis. 2018;206(4):254-262. doi:10.1097/NMD.0000000000001118
- Shafiee M, Arekhi S, Omranzadeh A, Sahebkar A. Saffron in the treatment of depression, anxiety and other mental disorders: current evidence and potential mechanisms of action. J Affect Disord. 2018;227:330-337. doi:10.1016/j.jad.2017.11.020
