A reference work, not a recommendation to use: this profile describes mechanism and evidence and does not replace medical advice. The claims come from marketing and forums and are tested here, not endorsed.

Claims checked

  • Better mood

    Mixed results

    Mood & stress · healthy · Human, randomised

    In healthy people with low mood, an extract improved mood at only one of two dose levels [9]; a more recent RCT found no effect on the primary endpoint [10]. Both studies are small and rely on self-reports.

  • Helps against depression

    Only proven in patients

    Mood & stress · Patients: mild to moderate depression · Human, randomised

    Meta-analyses of small RCTs show improvement over placebo [6][7], but in the most recent analysis only on self-rated scales and not on clinician-rated scales [7]; the methodological quality of the meta-analyses is limited [8].[6]

  • More libido

    Mixed results

    Libido & sexual function · Patients: sexual dysfunction under fluoxetine or erectile dysfunction · Human, randomised

    In two small RCTs under fluoxetine, erection and arousal improved, but sexual desire did not [11][12]. Meta-analyses on erectile dysfunction reach contradictory results [13][14]. Not studied in healthy people.[11][12][13][14]

Mechanism of action

Evidence map: Saffron

Each line stands for a published finding; line style and color show where it comes from. The filter hides weaker evidence levels. Sources: see references.

Saffron consists of the dried stigmas of Crocus sativus. A market survey found highly variable quality, especially in the volatile aroma compounds; in binding assays, saffron extract and crocetin had affinity for the phencyclidine binding site of the NMDA receptor and for the sigma-1 receptor, whereas crocins and picrocrocin did not [1]. The water-soluble crocins barely cross an intestinal cell model but are cleaved to trans-crocetin in intestinal cells; trans-crocetin crosses the intestinal barrier well and a blood-brain barrier model slowly; these data come from cell culture and mouse tissue [2]. A review of pharmacokinetics reaches the same conclusion: crocin is not available in the blood after oral administration, and crocetin would be mainly responsible for any effect [3].

The often-claimed serotonergic mechanism is weakly supported. In a stress model in rats, crocin and crocetin raised the lowered serum serotonin and lowered corticosterone; that both bind to the serotonin transporter, at the same site as fluoxetine, was shown only by a computer simulation (docking), while crocetin reduced NMDA currents in brain slices [4]. A review names serotonergic, antioxidant, anti-inflammatory and neuroendocrine effects as possible mechanisms, supported mainly by animal and laboratory studies [5].

Clinically, a meta-analysis in mild to moderate depression found a large advantage over placebo (Hedges g 0.89) and no significant inferiority to antidepressants [6]. A more recent meta-analysis of 34 RCTs with 1769 adults confirmed lower scores on the self-rated Beck Depression Inventory, but not on the clinician-rated Hamilton scale, with very high heterogeneity [7]. A meta-research study of 19 meta-analyses criticizes their methodological quality and calls for large, high-quality RCTs in several countries [8]. In healthy people with self-reported low mood, a standardized extract improved mood in an RCT with 128 participants, but only at one of two dose levels [9]; in a more recent RCT with 51 healthy people with subclinical symptoms, the primary endpoint of depression, anxiety and fatigue was unchanged [10].

On sexual function, there are two small RCTs in people with depression taking fluoxetine: in 30 evaluable men, saffron improved erectile function and satisfaction with intercourse, but not sexual desire or orgasm [11]; in 34 women, arousal, lubrication and pain improved, but not desire, satisfaction or orgasm [12]. A meta-analysis of herbal remedies for erectile dysfunction rated the three saffron studies as contradictory [13]; an Iranian meta-analysis using a fixed-effect model found positive effects on all domains of the erectile function questionnaire, but acknowledges methodological shortcomings [14]. A systematic review estimates that 20 to 30% of saffron traded worldwide is adulterated, including with safflower, marigold, turmeric or harmful dyes [15].

Synaptik serves solely to explain biochemistry and pharmacology and is not medical advice or a recommendation to take anything.

Mechanism of action only. Schematic section through an excitatory synapse, not to scale: of the NMDA receptor (a tetramer of two GluN1 and two GluN2B subunits) the two rear subunits are drawn and the front two are cut away to expose the pore. The sequence repeats in three beats: glutamate with depolarisation, glutamate alone, depolarisation alone. Particle numbers, timings, openings and the Ca²⁺ influx are outputs of this figure's model, not measurements, and are not shown as numbers. The numbers in the bottom-right readout come from laboratory and animal experiments and are not comparable with each other.
Drug
View

Glutamate  Glycine / D-serine  Mg²⁺  Ca²⁺  Na⁺  K⁺  red = drug, shown in the scene as a miniature skeletal formula · shapes as in the figure

In binding assays, saffron extract and its carotenoid crocetin showed affinity for the phencyclidine binding site of the NMDA receptor, whereas the crocins and picrocrocin did not [13]. In rat brain slices crocetin reduced the NMDA peak current; binding to GluN2B was shown only by computer docking [14]. There are no measurements in humans, so the figure shows crocetin in the pore less often than the drugs (illustrative).

Tap or focus a labelled part for a one-line explanation.

Also acting here, not drawn

  • CX-717: ampakine; NMDA effect only indirect, no action on the receptor itself
  • Huperzine A: NMDA inhibition only in a review by the developers’ group, without a binding site
  • L-theanine: only very weak binding to glutamate receptors
  • Phenylpiracetam: no displacement of ligands at NMDA receptors
  • Piracetam: only changed receptor numbers (mice)
  • Selank: only changed numbers of binding sites (mice)

What the figure shows, and how sure we are

Mechanism-of-action explainer only. Not medical advice, and no statement about benefit, risk, amount or use. Ketamine, esketamine, memantine and atomoxetine are medicines; most of the other substances have been studied at the NMDA receptor only in the lab or in animals.

What was simplified. One receptor stands for many; sizes, cleft width, membrane thickness, ion numbers and all timings are illustrative. The rear two subunits of a GluN1-GluN2B receptor are drawn; NMDA receptors with GluN2A and other subunits, and the AMPA receptors that help produce the depolarisation, are left out. The positions of Mg²⁺ in the filter and of ketamine in the vestibule follow [2] and [8]; for atomoxetine and crocetin the position in the pore has not been determined, and for agmatine the binding site at all. The numbers in [3], [4], [9], [11] to [18], [20] to [22], [24] and [26] to [32] come from the Synaptik drug profiles, where they were checked against the abstracts; [1], [5] to [8], [19], [23] and [25] were checked against their PubMed abstracts (25 Oct 2026), and the statements taken from the review [2] (M1, M3, M4 and the pore loop M2, M3 as the gate, the Mg²⁺ site at the tip of M2, Na⁺, K⁺ and Ca²⁺ as charge carriers, kinases and scaffold proteins at the C-terminal domain of GluN2B) against the relevant sections of the full text.

If medicines, alcohol or other substances are causing problems, a GP or a local addiction counselling service can help.

Assigned to the site of action Monoamine transporters DAT, NET and SERT at a dopaminergic synapse; not drawn in its figure: at SERT only a computer simulation (docking), no measured binding or inhibition.

Targets

TargetTypeActionKey dataEvidence
NMDA receptor (phencyclidine binding site)Ion channelBindingBinding by extract and crocetin, not by crocins or picrocrocin (binding assay) Laboratory[1]
NMDA receptor (NR2B-containing)Ion channelNegative modulatorNMDA peak current reduced by crocetin (rat brain slice); binding to NR2B only in docking Laboratory[4]
Sigma-1 receptorReceptorBindingBinding by extract and crocetin (binding assay) Laboratory[1]
Serotonin transporter (SLC6A4)TransporterBindingcomputer simulation (docking) only, no measured inhibition Laboratory[4]

Effect profile

Which signaling systems the compound affects according to current data. ↑ enhanced, ↓ dampened, ≈ modulated, ○ no effect; dots = evidence level.

  • Dopamineno data
  • Noradrenalineno data
  • Serotoninunclear · Serum serotonin raised in the rat stress model; binding to the serotonin transporter only in docking[4][5]
  • Acetylcholineno data
  • Glutamate · AMPAno data
  • Glutamate · NMDAmodulated · Crocetin binds to the PCP site and reduces NMDA currents in rat brain slices[1][4]
  • GABAno data
  • Histamineno data
  • Orexinno data
  • Adenosineno data
  • Opioidsno data
  • Melatoninno data
  • Neurotrophinsno data
  • Blood flowno data
  • Nitric oxideno data
  • Energy metabolismno data
  • Insulin · glucoseno data
  • Inflammation · oxidativeno data
  • Stress axisunclear · Corticosterone lowered in stressed rats; no effect on the stress-axis response in an RCT with 51 healthy people[4][10]
  • Testosterone axisno data
  • Endocannabinoidsno data
  • Autophagy · senescenceno data
  • Immune systemno data
  • Lipid metabolismno data
  • Coagulation · fibrinolysisno data
  • Methylationno data

Routes of administration and bioavailability

How much of a compound reaches the blood or target tissue depends on the route into the body and on the dosage form. All routes of administration explained

  • Oral (swallowed) · Stigma powder or standardized extract

    Crocins practically not, crocetin well (cell and tissue models)

    Crocins are cleaved to trans-crocetin in intestinal cells; trans-crocetin is absorbed passively and crosses a blood-brain barrier model slowly [2]. Crocin itself is not detectable in the blood after oral administration, and a large portion is excreted in the stool [3].

    Laboratory

Human studies

  • Meta-analysis of 34 RCTs on mood and anxiety

    systematic review and meta-analysis of RCTs, GRADE · Adults (mixed) · n = 1769 · at least 4 weeks

    Measured: BDI, BAI, HDRS, HARS, POMS

    Result: BDI −4.39 (95% CI −6.64 to −2.15; I² 92.3%), BAI −5.06; HDRS, HARS and POMS not significant.[7]

    Human, randomised

  • Meta-analysis in mild to moderate depression

    meta-analysis of RCTs (9 pooled) · Patients: mild to moderate depression · varied

    Measured: Severity of depression

    Result: Versus placebo g 0.891 (95% CI 0.369–1.412); versus antidepressants g −0.246 (p = 0.053).[6]

    Human, randomised

  • Extract in low mood

    RCT, double-blind, placebo-controlled, 3 arms · healthy (self-reported low mood) · n = 128 · 4 weeks

    Measured: POMS, anxiety, stress, sleep

    Result: Higher dose level: total POMS score better (p < 0.001, d = −1.10); lower dose level without effect.[9]

    Human, randomised

  • Extract in subclinical symptoms

    RCT, double-blind, placebo-controlled, registered (NCT05690126) · healthy (subclinical depressive symptoms) · n = 51 · 6 weeks

    Measured: Combined z-score of depression, anxiety and fatigue

    Result: Primary endpoint and individual symptoms unchanged; only one quality-of-life scale better; no effect on inflammatory markers and stress axis.[10]

    Human, randomised

  • Sexual function in men under fluoxetine

    RCT, double-blind, placebo-controlled · Patients: depression with fluoxetine-induced sexual dysfunction (men) · n = 36 · 4 weeks

    Measured: IIEF

    Result: Erectile function and satisfaction with intercourse better; desire, orgasm and overall satisfaction not; 30 participants analyzed.[11]

    Human, randomised

  • Sexual function in women under fluoxetine

    RCT, double-blind, placebo-controlled · Patients: depression with fluoxetine-induced sexual dysfunction (women) · n = 38 · 4 weeks

    Measured: FSFI

    Result: Total score, arousal, lubrication and pain better; desire (p = 0.196), satisfaction and orgasm not; 34 analyzed.[12]

    Human, randomised

Risks and warnings

  • Adulteration

    According to a systematic review, 20 to 30% of saffron traded worldwide is adulterated, with large regional differences (EU markets 3.5%, India 60%); substances used include safflower, marigold, turmeric and dyes suspected of causing cancer such as Sudan dyes and auramine O [15]. Even genuine goods vary greatly in quality [1].

    Laboratory

  • Side effects

    In meta-analyses, most commonly nausea, dry mouth, loss of appetite and headache; no serious side effects were reported [8].

    Human, randomised

  • Limited study quality

    Most meta-analyses on saffron have methodological shortcomings (missing protocols, registration, funding disclosures); confirmation by large RCTs from several countries is lacking [8].

    Human, randomised

Interactions

  • Fluoxetin: In two RCTs given as an add-on to fluoxetine, side effects were about as frequent as under placebo [11][12].

Fact check: common claims

  • partly supported

    “Saffron works as well as an antidepressant”

    A meta-analysis found no significant inferiority to antidepressants, close to the significance threshold (p = 0.053) [6]; the studies are small, the meta-analyses methodologically weak [8], and clinician-rated scales most recently showed no effect [7].

  • not supported

    “Saffron is a natural serotonin reuptake inhibitor”

    Binding to the serotonin transporter was shown only in the computer model; in rats, serum serotonin rose in the stress model [4]. Measured transporter inhibition was not found.

  • not supported

    “Saffron boosts libido”

    In both RCTs under fluoxetine, sexual desire did not improve, but erection and arousal did [11][12]; one meta-analysis with methodological shortcomings reports positive effects on desire in erectile dysfunction [14], another rates the data as contradictory [13].

  • partly supported

    “Saffron lifts mood even in healthy people”

    Positive RCT in self-reported low mood, but only at one dose level [9]; a more recent RCT without effect on the primary endpoint [10].

  • refuted

    “Expensive saffron on the market is pure saffron”

    Systematic review: 20 to 30% adulterated worldwide, up to 60% regionally [15].

Pharmacokinetics

Absorption
Crocins are cleaved to crocetin in the gut, and crocetin is absorbed (cell and tissue models) [2][3].
Half-life
Not verified.
Metabolism
Cleavage of crocins mainly by intestinal cells, hardly by the gut flora; gut bacteria degrade the backbone further [2].
Entry into the brain
Crocetin crosses blood-brain barrier models slowly (in vitro) [2][3]; not measured in humans.

Legal status

Germany
Not verified.
EU
Not verified.
USA
Not verified.
Other countries
Traded worldwide as a spice and one of the most expensive foods, and accordingly often adulterated [15]; legal status of extracts not checked.

Open questions

  • Do large, preregistered RCTs outside individual research centers confirm the effect in depression, including on clinician-rated scales?
  • Which constituent is responsible, and what crocetin levels does the human brain reach?
  • Does saffron act on sexual desire, or only on arousal and erection?

References

  1. Lechtenberg M, Schepmann D, Niehues M, Hellenbrand N, Wünsch B, Hensel A. Quality and functionality of saffron: quality control, species assortment and affinity of extract and isolated saffron compounds to NMDA and sigma1 (sigma-1) receptors. Planta Med. 2008;74(7):764–72. doi:10.1055/s-2008-1074535 · PubMed
  2. Lautenschläger M, Sendker J, Hüwel S, Galla HJ, Brandt S, Düfer M, et al. Intestinal formation of trans-crocetin from saffron extract (Crocus sativus L.) and in vitro permeation through intestinal and blood brain barrier. Phytomedicine. 2015;22(1):36–44. doi:10.1016/j.phymed.2014.10.009 · PubMed
  3. Hosseini A, Razavi BM, Hosseinzadeh H. Pharmacokinetic properties of saffron and its active components. Eur J Drug Metab Pharmacokinet. 2018;43(4):383–90. doi:10.1007/s13318-017-0449-3 · PubMed
  4. Mohammadi S, Naseri M, Faridi N, Zareie P, Zare L, Mirnajafi-Zadeh J, et al. Saffron carotenoids reversed the UCMS-induced depression and anxiety in rats: behavioral and biochemical parameters, and hippocampal BDNF/ERK/CREB and NR2B signaling markers. Phytomedicine. 2023;119:154989. doi:10.1016/j.phymed.2023.154989 · PubMed
  5. 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):517–27. doi:10.1002/hup.2434 · PubMed
  6. Tóth B, Hegyi P, Lantos T, Szakács Z, Kerémi B, Varga G, 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 · PubMed
  7. Mahmoudi R, Mohammadi-Sartang M, Servatyari K, Rafieipour N. Effect of saffron on depression, anxiety and mood disorder: a GRADE assessed systematic review and meta-analysis of 34 randomized controlled trials. Nutr Neurosci. 2026;29(7):816–37. doi:10.1080/1028415X.2025.2602153 · PubMed
  8. Lu C, Ke L, Li J, Zhao H, Lu T, Mentis AFA, et al. Saffron (Crocus sativus L.) and health outcomes: a meta-research review of meta-analyses and an evidence mapping study. Phytomedicine. 2021;91:153699. doi:10.1016/j.phymed.2021.153699 · PubMed
  9. Kell G, Rao A, Beccaria G, Clayton P, Inarejos-García AM, Prodanov M. affron® a novel saffron extract (Crocus sativus L.) improves mood in healthy adults over 4 weeks in a double-blind, parallel, randomized, placebo-controlled clinical trial. Complement Ther Med. 2017;33:58–64. doi:10.1016/j.ctim.2017.06.001 · PubMed
  10. Amadieu C, Leyrolle Q, Farneti M, Anesi A, Bruchet E, Montet J, et al. Effect of saffron extract supplementation on mood in healthy adults with subclinical symptoms of depression: a randomized, double-blind placebo-controlled study. Am J Clin Nutr. 2025;122(6):1625–35. doi:10.1016/j.ajcnut.2025.09.050 · PubMed
  11. Modabbernia A, Sohrabi H, Nasehi AA, Raisi F, Saroukhani S, Jamshidi A, et al. Effect of saffron on fluoxetine-induced sexual impairment in men: randomized double-blind placebo-controlled trial. Psychopharmacology (Berl). 2012;223(4):381–8. doi:10.1007/s00213-012-2729-6 · PubMed
  12. Kashani L, Raisi F, Saroukhani S, Sohrabi H, Modabbernia A, Nasehi AA, et al. Saffron for treatment of fluoxetine-induced sexual dysfunction in women: randomized double-blind placebo-controlled study. Hum Psychopharmacol. 2013;28(1):54–60. doi:10.1002/hup.2282 · PubMed
  13. Borrelli F, Colalto C, Delfino DV, Iriti M, Izzo AA. Herbal dietary supplements for erectile dysfunction: a systematic review and meta-analysis. Drugs. 2018;78(6):643–73. doi:10.1007/s40265-018-0897-3 · PubMed
  14. Maleki-Saghooni N, Mirzaeii K, Hosseinzadeh H, Sadeghi R, Irani M. A systematic review and meta-analysis of clinical trials on saffron (Crocus sativus) effectiveness and safety on erectile dysfunction and semen parameters. Avicenna J Phytomed. 2018;8(3):198–209. PubMed
  15. Mehdizadeh M, Omidi A, Morya S, Abideen Z. Combating saffron fraud: a systematic review of adulteration practices, detection technologies, recommendations and challenges. Crit Rev Food Sci Nutr. 2026;66(7):1387–403. doi:10.1080/10408398.2025.2544767 · PubMed
How this profile was checked

Last reviewed: 2026-10-23. All sources with a DOI or PMID were resolved automatically via Crossref or doi.org and PubMed.

Translated from the German profile (version of 2026-10-23); translation checked on 2026-10-26.

Checked

  • Checked against the abstracts: quality variation, binding of extract and crocetin to NMDA (PCP site) and sigma-1, not of crocins/picrocrocin [1]; crocin barely crosses Caco-2, cleavage to trans-crocetin in intestinal cells, slow BBB passage in the model [2]; crocin not in the blood after oral intake [3]; SERT and NR2B binding only bioinformatic, serum serotonin ↑ and corticosterone ↓ in rats, NMDA current ↓ in patch-clamp [4]; mechanisms [5]; g 0.891 and −0.246 (p = 0.053), 9 pooled studies [6]; 34 RCTs, 1769 participants, BDI −4.39, I² 92.3%, BAI −5.06, HDRS/HARS/POMS not significant, GRADE moderate [7]; 19 meta-analyses, quality shortcomings, side effects [8]; n = 128, only the higher dose level effective, d = −1.10 [9]; n = 51, primary endpoint without effect, no effect on the HPA axis [10]; 36 and 30 men, respectively, desire p = 0.517 [11]; 38 and 34 women, respectively, desire p = 0.196 [12]; 3 saffron RCTs (n = 397) with mixed results [13]; 6 studies, 3 in the meta-analysis, desire MD 0.78, methodological shortcomings [14]; 20–30%, EU 3.5%, India 60%, adulterants [15].
  • All DOIs checked via Crossref and all PMIDs via PubMed (checking script); [14] has no DOI, only a PMID.

Not verified

  • Legal status of saffron extracts (dietary supplement, novel food) not checked.
  • Whether study [9] was funded or co-authored by the manufacturer of the branded extract used was not checked.
  • Country of origin of the individual studies included in the meta-analyses not checked individually; the call for studies “in several countries” comes from [8].
  • The mean values of the quality-of-life scale in the abstract of [10] appear to be swapped (placebo higher than saffron despite the claimed improvement); therefore not adopted.
  • Human pharmacokinetics of crocetin not checked against a primary study.

No dosages, no sources of supply. This information does not replace medical advice.

Spotted an error or missing something?

Request a compound, suggest an improvement or report an error. Every message is read.

Type of message

The website stores nothing. The button opens your email app with a pre-filled message.