Does saffron help with depression?

On questionnaires people fill in themselves, yes, by a few points. On the scale a clinician scores, the largest analyses found no significant change. A 2026 meta-analysis of 34 placebo-controlled trials with 1769 adults, rated with GRADE, found saffron lowered self-rated Beck depression scores by 4.39 points (95% CI 2.15 to 6.64) in the 14 trials that used that scale, with very high disagreement between trials (I² 92.3%). The same analysis found no significant effect on the clinician-rated Hamilton scale.

Unsettled. Saffron looks better than placebo in small trials and not clearly better when a clinician does the rating. The answer depends on which scale is used. A 2019 analysis of 9 small trials in mild to moderate depression found a large effect against placebo (g 0.891, 95% CI 0.369 to 1.412), and its trials were few, mostly Iranian and 6 to 8 weeks long, the kind of evidence where large effects are often inflated. A 2022 umbrella review of 7 meta-analyses found Hamilton scores did not change significantly (-2.10 points, 95% CI -5.05 to 0.86).

What the trials found

Against antidepressants

The 2019 meta-analysis found saffron non-inferior to the antidepressants it was tested against (g -0.246, 95% CI -0.495 to 0.004). A 2025 meta-analysis of 8 head-to-head trials against SSRI antidepressants found 6 fewer adverse events per 100 people on saffron (risk difference -0.06, 95% CI -0.09 to -0.04). Those trials were small and were not designed to prove saffron as good as an SSRI, and fewer side effects does not mean saffron can replace prescribed treatment. A 2025 network meta-analysis of 192 trials with 17,437 patients ranked saffron among four supplements that did better than antidepressants alone (0.69, 95% CI 0.34 to 1.04). That ranking rests on indirect comparisons and a small number of saffron trials, so it is fragile.

The largest trial on this page

In 202 adults aged 18 to 70 with depressive symptoms, 12 weeks of a commercial saffron extract lowered the DASS-21 depression score by 2.92 points more than placebo (95% CI 0.71 to 5.13, Cohen's d 0.39). A clinically significant improvement came in 72.3% on saffron and 54.3% on placebo, so more than half of the placebo group improved as well. These were people with symptoms, without a diagnosis of depression, and the extract was a commercial product, Affron.

Smaller trials, both directions

In a 12-week trial of 73 women with overweight and mild to moderate depression, Beck scores fell by 8.4 points on saffron and by 3.9 on placebo, and only 52 of the 73 finished. In 51 healthy adults with low mood below the level of a diagnosis, 6 weeks of saffron extract did not improve depression, anxiety or fatigue scores. In 50 smokers not selected for depression, 3 months of crocin, the saffron pigment, did not change depression or anxiety scores. A 2023 meta-analysis of antioxidant supplements found crocin improved depression scores by 6.04 points, with mixed populations and scales and no confidence interval in the abstract.

Saffron in food is a different thing

The trials used capsules of extract or crocin. A teaspoon of saffron spice weighs about 0.7 g. We found no trial of saffron eaten in food for mood.

Who should be careful

Not for everyone. Depression that is more than mild belongs with a doctor. The WFSBP and CANMAT taskforce advises that supplements like saffron be used mainly alongside standard medical care, especially in more severe mental illness. Do not stop a prescribed antidepressant to switch to saffron.

A review of 12 studies in pregnant women found no harm from low doses of about 250 to 750 mg at term in clinical trials. Higher doses and work in saffron fields were linked to more miscarriages, and the review gives no number for what counts as a high dose. One 64-year-old man taking the blood thinner rivaroxaban developed nosebleeds and bleeding gums after starting a saffron supplement. That is a single case and does not prove cause. Across 9 trials, saffron did not change kidney markers, and those trials were short. A review of 23 studies estimates that 20 to 30% of commercial saffron is adulterated, from 3.5% in regulated EU markets to 60% in India, sometimes with carcinogenic Sudan dyes, and those estimates come from different methods and regions.

What expert bodies say

The 2022 WFSBP and CANMAT taskforce guideline lists saffron (++) among plant products with Grade A evidence and positive directionality for unipolar depression, below St John's wort (+++). It flags poor quality and standardization of plant products as a key limitation. The EU refused to authorize a health claim that the saffron extract Affron helps maintain a healthy mood. A refusal means the evidence fell short when it was assessed, which is different from proof that it does not work.

How we searched

What would change this answer

• Large trials at several centers outside Iran, so the result does not rest on one research community.
• Trials longer than 12 weeks that measure relapse, the outcome that matters in depression.
• Clinician-rated scales as the main outcome. Self-rated scores are where saffron looks good, and clinician ratings are where it does not.
• Standardized products. The taskforce names product quality as a limit, and adulteration of the spice is common.

More on the spice itself: saffron.

Sources

1. ev-safd-01 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized controlled trials (PRISMA, Cochrane risk of bias) · n = 9 trials pooled
   Available data from randomized, controlled clinical trials support that saffron is significantly more effective than placebo (g = 0.891; 95% CI: 0.369 - 1.412, p = 0.001), and non-inferior to tested antidepressant drugs (g = - 0.246; 95% CI: - 0.495 - 0.004, p = 0.053).
   
   Who: Patients with mild to moderate depression in randomized controlled trials of oral saffron
   Effect: saffron vs placebo g = 0.891 (95% CI 0.369 to 1.412); saffron vs antidepressants g = -0.246 (95% CI -0.495 to 0.004), non-inferior
   Certainty: Few small trials, mostly from Iran, 6 to 8 weeks; large effect sizes in small trials are often inflated.
   Toth et al., Planta Medica, 2019 · checked 2026-10-07 · we read the abstract
2. ev-safd-02 · Meta-analysis or systematic review · systematic review and random-effects meta-analysis of randomized controlled trials · n = 8 trials (depression)
   With regard to safety, participants receiving saffron had fewer adverse events than the SSRI group (risk difference: -0.06; 95% CI: -0.09, -0.04; I2: 0%).
   
   Who: Adults with depression or anxiety in RCTs comparing saffron with SSRIs
   Effect: depression SMD 0.10 (95% CI -0.09 to 0.29), no difference; adverse events risk difference -0.06 (95% CI -0.09 to -0.04), I2 0%
   Certainty: Equivalence trials were small and not powered for non-inferiority; fewer side effects does not mean saffron replaces prescribed treatment.
   Shafiee et al., Nutrition Reviews, 2025 · checked 2026-10-07 · we read the abstract
3. ev-safd-03 · Meta-analysis or systematic review · umbrella review and meta-analysis of meta-analyses · n = 7 meta-analyses
   However, saffron did not change the HAMD scores (ES: -2.10; 95% CI: -5.05, 0.86, p = 0.164) and mixed scores (HAM-D/BDI/DASS) (ES: 0.02; 95% CI: -0.39, 0.43,p = 0.941).
   
   Who: Adults with depression in meta-analyses of saffron trials
   Effect: HAMD ES -2.10 (95% CI -5.05 to 0.86), not significant; mixed scales ES 0.02 (95% CI -0.39 to 0.43); BDI ES -3.87 (95% CI -5.27 to -2.46)
   Certainty: Overlapping primary trials across the pooled meta-analyses; the authors conclude saffron cannot be a single therapeutic approach.
   Musazadeh et al., Pharmacological Research, 2022 · checked 2026-10-07 · we read the abstract
4. ev-safd-04 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized placebo-controlled trials, GRADE · n = 1769
   No significant effects were observed for the HDRS, HARS, or POMS.
   
   Who: Adults in randomized placebo-controlled trials of saffron lasting at least 4 weeks
   Effect: no significant effect on HDRS, HARS or POMS; self-rated BDI -4.39 points (carded elsewhere)
   Certainty: Heterogeneity very high (I2 above 90%) for the self-rated outcomes; the authors call for high-quality trials.
   Nutr Neurosci, 2026 · checked 2026-10-07 · we read the abstract
5. ev-safd-05 · Meta-analysis or systematic review · systematic review and frequentist random-effects network meta-analysis · n = 17437
   Additionally, four nutraceutical monotherapies demonstrated superior efficacy compared to ADT: EPA + DHA (0.6, 0.32-0.88), SAMe (0.52, 0.18-0.87), curcumin (0.62, -0.17 to 1.40) and saffron (0.69, 0.34-1.04).
   
   Who: Adults with depressive disorder in trials of 44 nutraceuticals
   Effect: saffron monotherapy vs antidepressant SMD 0.69 (95% CI 0.34 to 1.04); most nutraceuticals tolerated like placebo
   Certainty: Indirect comparisons; saffron trials are small and few, so its rank is fragile.
   Psychol Med, 2025 · checked 2026-10-07 · we read the abstract
6. ev-safd-06 · Meta-analysis or systematic review · meta-analysis of randomized controlled trials · n = 4049
   The meta-analysis found that the positive effect of antioxidant supplementation, such as magnesium (SMD = 0.16, p = 0.03), zinc (SMD = 0.59, p = 0.01), selenium (SMD = 0.33, p = 0.009), CoQ10 (SMD = 0.97, p = 0.05), tea and coffee (SMD = 1.15, p = 0.001) and crocin (MD = 6.04, p < 0.00001), on depressive status were all significant.
   
   Who: Participants in RCTs of antioxidant supplements and depression
   Effect: crocin MD 6.04 (p < 0.00001) on depression scores
   Certainty: Crocin extract, not culinary saffron; populations and scales mixed; no confidence interval in the abstract.
   J Affect Disord, 2023 · checked 2026-10-07 · we read the abstract
7. ev-safd-07 · Randomized controlled trial(s) · 2-arm, 12-week, parallel-group, randomized, double-blind, placebo-controlled trial · n = 202
   On the primary outcome measure, compared to the placebo, saffron was associated with greater improvements in the Depression, Anxiety, and Stress scale - 21 depression score (β: -2.92 points; 95% confidence interval: -5.13, -0.71 points; Cohen's d = 0.39), with 72.3% of participants in the saffron group achieving a clinically significant change (a reduction of ≥ 7 points) compared to 54.3% of participants in the placebo group (P = 0.010).
   
   Who: Two hundred and two adults aged 18-70 with depressive symptoms
   Effect: DASS-21 depression beta -2.92 points (95% CI -5.13 to -0.71), Cohen's d 0.39; clinically significant change 72.3% vs 54.3%
   Certainty: Subclinical symptoms, not diagnosed depression; secondary outcomes did not differ; industry extract (Affron).
   J Nutr, 2025 · checked 2026-10-07 · we read the abstract
8. ev-safd-08 · Randomized controlled trial(s) · randomized, double-blind, placebo-controlled trial, 6 weeks · n = 51
   Saffron extract supplementation do not affect subclinical depressive symptoms, either as a composite score or individual symptom categories.
   
   Who: 51 adult healthy individuals with subclinical neuropsychiatric symptoms
   Effect: no significant effect on composite or individual depression, anxiety and fatigue scores
   Certainty: Small and short; a null result in non-depressed people.
   Am J Clin Nutr, 2025 · checked 2026-10-07 · we read the abstract
9. ev-safd-09 · Randomized controlled trial(s) · 12-week double-blind, placebo-controlled randomized clinical trial · n = 73
   Mean depression scores in the saffron group significantly decreased compared to placebo (mean ± SD: -8.4 score ± 5.9 vs -3.9 ± 5.5; t[50] = 2; P = .007; 95% CI: 1.3-7.7).
   
   Who: Seventy-three women with BMI of 25 or more and mild-to-moderate depression
   Effect: BDI-II change -8.4 vs -3.9 points (95% CI of difference 1.3 to 7.7); no effect on food craving
   Certainty: Only 52 of 73 finished; single-center trial.
   J Clin Pharm Ther, 2020 · checked 2026-10-07 · we read the abstract
10. ev-safd-10 · Randomized controlled trial(s) · randomized placebo-controlled trial, 3 months · n = 50
    The primary outcome results such as nicotine dependence, depression, and anxiety inventory did not have a significant difference among the intervention groups (P > 0.05).
    
    Who: 50 smokers
    Effect: no significant difference in depression and anxiety inventories (P > 0.05)
    Certainty: Crocin, not whole saffron; participants were not selected for depression.
    Naunyn Schmiedebergs Arch Pharmacol, 2024 · checked 2026-10-07 · we read the abstract
11. ev-safd-11 · Position of an expert body · international clinical guideline based on meta-reviews and RCTs · n = —
    Phytoceuticals with supporting Grade A evidence and positive directionality included St John's wort (+++), saffron (++), curcumin (++), and lavender (+) in the treatment of unipolar depression, while rhodiola use was not supported for use in mood disorders.
    
    Who: Adults with unipolar depression
    Effect: saffron (++); St John's wort (+++); curcumin (++); lavender (+); rhodiola not supported
    Certainty: Grade A evidence was taken as given; the taskforce flags poor quality and standardization of plant products.
    World J Biol Psychiatry, 2022 · checked 2026-10-07 · we read the abstract
12. ev-safd-12 · Position of an expert body · international clinical guideline · n = —
    Finally, the taskforce noted that such use of nutraceuticals or phytoceuticals be primarily recommended (where supportive evidence exists) adjunctively within a standard medical/health professional care model, especially in cases of more severe mental illness.
    
    Who: People with psychiatric disorders considering nutraceuticals or phytoceuticals
    Effect: adjunctive use within standard care recommended; product quality and standardization a key limitation
    Certainty: General advice for all agents in the guideline, not specific to saffron.
    World J Biol Psychiatry, 2022 · checked 2026-10-07 · we read the abstract
13. sfrn-daily-r7-01 · Meta-analysis or systematic review · meta-analysis of RCTs · n = 1769
    A total of 34 RCTs comprising 1769 participants were included. Saffron supplementation significantly reduced scores on BDI (14 trials; 817 participants; WMD = -4.39, 95% CI: -6.64 to -2.15, p < 0.001, I2 = 92.3%) and the BAI (6 trials; 339 participants; WMD = -5.06, 95% CI: -8.44 to -1.68, p = 0.003, I2 = 94.8%). No significant effects were observed for the HDRS, HARS, or POMS.
    
    Who: adults in 34 placebo-controlled RCTs lasting at least 4 weeks
    Effect: BDI WMD -4.39 (95% CI -6.64 to -2.15), 14 trials, I2 92.3%; BAI WMD -5.06 (-8.44 to -1.68); HDRS, HARS, POMS no effect; GRADE moderate
    Certainty: high heterogeneity; effect only on self-rated scales; most RCTs from Iran
    Nutr Neurosci, 2026 · checked 2026-09-29 · we read the abstract
14. sfrn-daily-r7-05 · Meta-analysis or systematic review · meta-analysis of RCTs · n = 9 RCTs
    The pooled analysis showed that saffron supplementation had no significant effect on serum urea concentrations (WMD: - 1.05 mg/dl; 95% CI - 5.1 to 3; P = 0.6, I2 = 93%, P < 0.001) and serum creatinine levels (WMD: - 0.006 mg/dl; 95% CI - 0.08 to 0.06; P = 0.8, I2 = 79%, P < 0.001) when compared to the placebo group.
    
    Who: adults in 9 RCTs
    Effect: serum urea WMD -1.05 mg/dL (95% CI -5.1 to 3), creatinine WMD -0.006 mg/dL (-0.08 to 0.06), both not significant
    Certainty: short RCTs; safety with long-term intake not assessed
    Int Urol Nephrol, 2022 · checked 2026-09-29 · we read the abstract
15. sfrn-daily-r7-07 · Position of an expert body · regulatory decision · n = —
    POL-HC-08742 Affron®, aqueous saffron extract, with the sum of crocins and safranal concentration > 3,5 % and dextrin as inert carrier Affron® contributes to maintain a healthy mood by reducing the negative traits of depressive and anxiety feelings -/- Non-authorised
    
    Who: EU health claim application, Affron aqueous saffron extract
    Effect: status Non-authorized
    Certainty: refusal = the evidence did not reach the bar at the time of assessment, not 'has no effect'; the reason is in the EFSA opinion
    EU Register on nutrition and health claims, claim POL-HC-08742, snapshot 2026-09-21 · checked 2026-09-29 · we read the dataset
16. sfrn-daily-r7-08 · Laboratory or animal data · case series · n = 1
    We herein present the case of a 64-year old male with the complaint of acute onset epistaxis and bleeding gums following the concomitant use of rivaroxaban and saffron supplement.
    
    Who: one 64-year-old man taking rivaroxaban
    Effect: acute epistaxis and gum bleeding
    Certainty: K-143: one case, before decision D; causality not established
    Cardiovasc Hematol Disord Drug Targets, 2022 · checked 2026-09-29 · we read the abstract
17. sfrn-daily-r7-10 · Observational data · systematic review of analytical studies · n = 23 studies
    Findings reveal that 20-30% of commercial saffron is adulterated globally, though with significant regional disparities (e.g., 3.5% in regulated EU markets vs. 60% in India), driven by economic incentives and regulatory gaps. Common fraudulent practices include substitution with safflower, marigold, or turmeric; dilution with extraneous plant materials; and the addition of hazardous synthetic dyes such as Sudan compounds and auramine-O.
    
    Who: commercial saffron samples, 23 studies 2015-2025
    Effect: 20-30% adulterated globally; 3.5% EU vs 60% India; substitutes safflower, marigold, turmeric; Sudan dyes, auramine-O
    Certainty: frequency estimates from different methods and regions
    Crit Rev Food Sci Nutr, 2026 · checked 2026-09-29 · we read the abstract
18. sfrn-daily-r7-12 · Observational data · integrative review · n = 12 studies
    Within clinical trials, low dosage of saffron (approximately 250 to 750 milligrams at term) showed no incident of negative maternal or neonatal outcome, which significantly translates into its enhanced tolerance in controlled clinical settings, particularly for low-risk, term pregnancies.26,32 These findings were notably contradicted across studies, which reported that saffron's consumption in higher dosages or being occupationally exposed in saffron fields increases critical pregnancy-associated risks. Studies focused on women farmers in saffron fields noted elevated risk for miscarriage, particularly due to longer dermal or inhalational exposure.
    
    Who: pregnant women, 12 studies from inception to June 2025
    Effect: low dose at term without adverse maternal or neonatal outcomes; higher dose or occupational exposure with elevated miscarriage risk
    Certainty: review of mixed designs; the "high dose" threshold is not quantitative; the first trimester is the riskiest
    Journal of Evidence-Based Integrative Medicine, 2026 · checked 2026-09-29 · we read the fulltext
19. sfrn-daily-r7-14 · Position of an expert body · reference dataset · n = —
    Spices, saffron (FDC 170934, SR Legacy, Spices and Herbs): portion 1 tsp = 0.7 g; per 100 g iron 11.1 mg; magnesium 264.0 mg; potassium 1724.0 mg; vitamin C 80.8 mg; energy 310 kcal
    
    Who: Spices, saffron, per 3.5 oz (100 g) and per 1 tsp (0.7 g)
    Effect: per 3.5 oz (100 g): iron 11.1 mg, magnesium 264 mg, potassium 1724 mg, vitamin C 80.8 mg; 1 tsp = 0.7 g, about 0.08 mg iron
    Certainty: cooking amount per dish is usually less than a teaspoon; no spread
    USDA FoodData Central, SR Legacy, FDC 170934 (NDB 2037), 2019 · checked 2026-09-29 · we read the dataset