The mood lift is the line: two capsules a day of standardised extract matched fluoxetine in mild-to-moderate depression across several head-to-head trials and beat placebo in nearly every trial that's been run. PMS severity drops for three women in four. The sexual side effects that drive most SSRI quitting often get rescued by adding it. Cheap to try; eight weeks to know if you're a responder. The catches: pregnancy is out, severe depression needs a clinician, and the threads in your spice rack aren't a supplement.
Saffron's red stigmas — three per flower, weighing about a tenth of a gram all together — contain three molecules that do most of the work: crocin, the red pigment; safranal, the aroma compound that makes saffron smell like saffron; and crocetin, the small fragment of crocin that crosses into the bloodstream. They behave a lot like a polypharmacological antidepressant. In animal and binding studies they inhibit monoamine oxidase A, slow the reuptake of serotonin, noradrenaline, and dopamine the way SSRIs and SNRIs do, nudge the BDNF growth signal that sits downstream of nearly every effective mood drug, and antagonise NMDA receptors the way ketamine does — at a much smaller scale than any of those, but on most of the same levers Lopresti and Drummond 2014. The retinal effect runs on a separate track: crocetin reaches the back of the eye through the choroidal circulation and scavenges the reactive oxygen molecules that drive age-related macular degeneration. The same compound, two different jobs.
What the trials actually show
For mild-to-moderate depression, the trial pattern is unusually consistent. Two divided doses of standardised stigma extract a day, six weeks, and the Hamilton Depression scale falls by roughly a third — beating placebo in every Iranian trial that ran and matching fluoxetine 20 mg in the head-to-head Akhondzadeh et al. 2005 Noorbala et al. 2005. By 2019 the meta-analysis had pooled eleven trials and roughly five hundred patients and found the same direction: significantly better than placebo, statistically indistinguishable from standard antidepressants Marx et al. 2019. The honest caveat is that most of this evidence is Iranian, the trials are small, and the largest pooled dataset is still smaller than a single phase-three SSRI registration trial. The replications in Australian adolescents and in Western adults experiencing low mood have come in positive too Lopresti et al. 2018.
For PMS, the head trial is concrete enough to anchor on by itself.
For SSRI-induced sexual dysfunction — the side effect that drives more discontinuation than any other — two trials from the same Tehran group added 30 mg of saffron to fluoxetine in men and in women. In both, the relevant index scores improved on saffron and not on placebo: erectile function and intercourse satisfaction in men Modabbernia et al. 2012; arousal, lubrication, and pain in women Kashani et al. 2013. The trials were small and the effect didn't reach every domain — desire and orgasm were unchanged — but the rescue is real for arousal-side problems specifically.
The other claims are real but smaller. For early age-related macular degeneration, 20 mg a day for three months improved focal-ERG retinal-function scores and lifted Snellen visual acuity slightly in an Italian group's crossover trial; the gains held over a fourteen-month follow-up, and an independent Australian group replicated the visual-acuity signal Falsini et al. 2010 Piccardi et al. 2012 Broadhead et al. 2019. For appetite, a specific extract called Satiereal cut snacking by more than half and produced about 1.5 kg of fat loss over eight weeks in mildly overweight women — modest but real, and from a single industry-sponsored trial Gout et al. 2010. For sleep, four weeks of evening saffron improved the Insomnia Severity Index in a small Australian trial Lopresti et al. 2020. None of these are at the level of the depression and PMS signals. They are the secondary reasons, not the headline.
The dose, and how to actually run a trial
Across nearly every indication, the dose has converged on the same number: 28 to 30 milligrams a day of a standardised stigma extract, split into a morning and evening dose, taken with food, run for at least six to eight weeks before judging whether you're a responder. Anything shorter is too soon — the Hamilton Depression curve in the trials separates from placebo around week three or four and keeps widening through week six. The onset window is closer to an SSRI's than to aspirin's, and the same patience applies.
When not to take it
The hard line is pregnancy. At very high doses — far above any supplement, but inside the range of traditional Persian medicinal preparations — saffron stigma stimulates uterine contractions and animal models show fetal toxicity. Supplement doses are nowhere near that threshold, but no safety trial in pregnant women has been done or will ever ethically be done, and the medical tradition that knew the plant longest warned against it explicitly Hosseinzadeh and Nassiri-Asl 2014. Breastfeeding inherits the same caution. Bipolar disorder is the second caveat: any compound acting on monoamine oxidase carries a theoretical risk of tipping a mood into mania, and saffron has not been studied in bipolar populations — treat it the way you'd treat any antidepressant.
Who this is most clearly for
Two reader pictures the trial numbers describe most cleanly.
For women with significant PMS — irritability that scuttles a date night, breast tenderness that sits under every workout, the depressive dip on day 25 that you've learned to schedule around — saffron at 30 mg/day through two full cycles is the option the data most cleanly supports. Three in four felt the difference in the Agha-Hosseini trial; one in twelve felt it on placebo Agha-Hosseini et al. 2008. The wider PMS landscape has its own merits — luteal-phase SSRIs, magnesium, B6, calcium, cycle-tracked exercise — and saffron isn't the only option. It's the one that doesn't need a prescription, isn't a hormone, and travels with the cleanest side-effect profile.
For anyone on an SSRI and quietly losing their sex life to it — and that's roughly half of people on one — adding saffron is the rescue with the best trial evidence. It isn't perfect: desire and orgasm don't always come back, but arousal, lubrication, and erectile function usually do Kashani et al. 2013 Modabbernia et al. 2012. The pre-condition is that the SSRI itself is otherwise working — saffron is the add-on, not the swap.
For the reader with low-grade depression who has either tried an SSRI and disliked the trade or never quite qualified for one, saffron is a serious eight-week trial. Mild-to-moderate depression is the slice of the evidence base; severe depression isn't.
Three things saffron isn't
The threads in your spice rack are not the supplement. A paella's worth of saffron carries a fraction of the standardised extract's dose, and the bulk-spice market is heavily adulterated besides (see failure modes below). The mood and PMS trials are all run on extract; the cooking spice is for cooking.
Saffron is not a "natural SSRI" in the lay sense of weaker-but-safer. The receptor profile is actually broader than any SSRI — it touches MAO-A, every monoamine transporter, NMDA, GABA, and BDNF — and the trial evidence is specifically for mild-to-moderate depression. Severe depression, psychotic depression, and treatment-resistant depression have not been adequately studied; the right assumption is that this is the wrong tool for them.
The skin-glow, hair-growth, libido-without-an-SSRI-baseline, and limitless-energy claims that cluster around saffron in the supplement-marketing universe do not have the trial backing that depression and PMS do. They are extrapolation from the molecule's antioxidant chemistry, not findings.
Why a trial seems to fail when it does
Three reasons, in roughly descending order of frequency.
- The product is adulterated. Independent testing of bulk saffron has found 30–40% of samples padded with safflower, marigold, or beet, and even saffron classified to the ISO standard can carry up to a fifth of its weight in adulterants invisible to that standard Petrosino et al. 2018. A standardised extract with HPLC-verified active-compound content — affron, Satiereal, safr'Inside — sidesteps this. A $5 bottle of "saffron extract" with no standardisation listed is a different product, and may be a mostly-red-coloured-plant-matter product.
- The trial is too short. Most people quit at week three. So would most people on an SSRI, at the point the curve is just starting to bend. Eight weeks is the minimum honest test.
- The indication is wrong. Saffron has been tested in mild-to-moderate depression, PMS, SSRI sexual-side-effect rescue, retinal preservation, snack suppression, and sleep. For severe depression, generalised anxiety as a standalone, weight loss without an emotional-snacking pattern, or the assorted lookmaxxing claims, the evidence is thin and the result will be too.
What month two looks like, if you respond
The trials don't promise a transformation, and the dek shouldn't either. What they project, for the readers the evidence covers, is something specific you'd been losing, given back.
The low-mood reader gets to week six and notices the morning weight has lifted — not gone, but no longer the thing they wake up under. The session at the gym they kept skipping, they're doing again. They haven't gained weight, haven't lost interest in sex, the emotional flatness they remember from a previous antidepressant trial hasn't arrived Marx et al. 2019.
The PMS reader gets to day 25 of the second cycle and realises she's been bracing for the dip and it didn't come. Or it came and lost its tooth. By the third cycle the relationship that used to fray on days 24–28 stops fraying; the partner notices before she tells him Agha-Hosseini et al. 2008.
The SSRI-with-sexual-dysfunction reader finds intercourse mostly works again by week four. Desire and orgasm aren't back to baseline; arousal and the rest of the response cycle largely are. The relationship survives the medication instead of being slowly bartered against it Kashani et al. 2013 Modabbernia et al. 2012.
None of this is a transformation. The honest verb is recover — a specific thing the depression, the cycle, or the medication had been taking from you, returned. For about half of responders, that's enough to keep taking it.
Adjacent ground
If saffron is the question, these are usually the next questions.
- SSRIs themselves — the comparator the saffron depression trials are built around. The trade-off shape worth understanding before deciding either way.
- Other PMS approaches — luteal-phase SSRIs, magnesium, B6, calcium, and the cycle-tracked exercise and sleep pieces. Saffron is one option, not the field.
- Light therapy and morning sunlight for the low-mood reader with a seasonal pattern.
- AREDS-2 for established AMD. The retinal saffron signal is interesting but doesn't displace what's already guideline-grade.
- SAMe and St. John's wort — the other botanical antidepressants with their own evidence shapes. Saffron's the cleanest of the three on side effects and trial replication, but the field overlaps.
Substance and claimed effects
Saffron is the dried red stigma of Crocus sativus, a small autumn-flowering crocus. Three flowers yield three stigmas, around 30 mg of spice; roughly 150 000 flowers are needed for a kilogram, which is why dry weight is sold at gold-comparable prices. As a supplement, saffron is taken not as the loose threads but as a standardised stigma extract — most commonly affron (Pharmactive) at 28 mg/day standardised to ≥3.5% Lepticrosalides, or Satiereal (Inoreal) at 176.5 mg/day. The bioactive constituents are the carotenoid glycoside crocin (and its aglycone crocetin), the monoterpene aldehyde safranal, and the bitter glycoside picrocrocin Lopresti and Drummond 2014.
The claimed effects clinically tested in humans, in descending order of evidence weight, are: antidepressant activity in mild-to-moderate depression at 30 mg/day; reduction in PMS symptom severity at 30 mg/day; rescue of SSRI-induced sexual dysfunction; suppression of snacking and modest weight loss at the Satiereal dose; improvement of central retinal function in early age-related macular degeneration (AMD) at 20 mg/day; and improvement of self-reported sleep quality. The entry covers each of these as a real consequence of the substance. Out of scope: cardiometabolic indices in diabetes, schizophrenia adjunct therapy, oncology adjuncts, and culinary saffron (kitchen doses of a few stigmas per dish do not reach pharmacological levels).
Evidence by addressing question
Mechanism
Saffron's antidepressant action is attributed to a combined monoaminergic profile rather than a single receptor hit. In vitro and animal work shows: monoamine reuptake inhibition at serotonin, noradrenaline, and dopamine transporters; monoamine oxidase A/B inhibition mediated by crocin (molecular modelling supports non-competitive binding); NMDA-receptor antagonism; GABA-A agonism; and upregulation of BDNF signalling in hippocampal tissue, plausibly the slow-onset mechanism shared with conventional antidepressants Lopresti and Drummond 2014. Crocin also has substantial antioxidant and anti-inflammatory activity — relevant both to neuroinflammation models of depression and to the retinal mechanism below.
For AMD specifically, the mechanism is retinal-protective rather than CNS: crocetin is small and lipophilic enough to reach the retina via the choroidal circulation, where it scavenges reactive oxygen species generated by the photo-oxidative stress that drives outer-retinal degeneration. Animal models show preservation of photoreceptor morphology and function under light damage and ischaemia Falsini et al. 2010.
For PMS, the mechanism is plausibly the same serotonergic action that explains the depression signal — SSRIs work for PMDD via the same pathway. For appetite, Satiereal's marketing mechanism is a mood-mediated reduction in emotional/inter-meal snacking rather than a metabolic effect Gout et al. 2010.
Evidence
Depression. The original Iranian RCT series from Akhondzadeh and colleagues established the signal: 30 mg/day for six weeks beat placebo on the Hamilton Depression Rating Scale in a n=40 trial Akhondzadeh et al. 2005, and matched fluoxetine 20 mg in a same-design head-to-head Noorbala et al. 2005. The Lopresti and Drummond 2014 systematic review collated six RCTs covering placebo and active comparators (fluoxetine, imipramine) and concluded saffron was significantly more effective than placebo and non-inferior to standard antidepressants for mild-to-moderate depression Lopresti and Drummond 2014. The Marx et al. 2019 meta-analysis pooled 11 RCTs (n≈531) and confirmed both findings — significant improvement vs placebo, no significant difference vs antidepressants — with the caveat that the bulk of evidence originated from a small number of Iranian centres Marx et al. 2019. The affron 28-mg standardised extract has been tested in Western adolescents Lopresti et al. 2018 and in healthy adults with subclinical depression, replicating the placebo-superiority signal outside Iran.
Evidence (PMS)
The headline PMS trial is Agha-Hosseini et al. 2008: a double-blind placebo-controlled RCT in n=50 women aged 20–45 with regular cycles and ≥6 months of PMS, given saffron 30 mg/day (15 mg b.i.d.) for two consecutive cycles. Seventy-six percent of the saffron arm achieved ≥50% reduction in Total Premenstrual Daily Symptoms, against 8% on placebo; Hamilton depression scores also improved Agha-Hosseini et al. 2008. Effect size is large by PMS-trial standards, although the cohort is small and single-centre. Subsequent Iranian trials (Pirdadeh Beiranvand and others) have replicated the direction with similar doses.
Evidence (vision / AMD)
Falsini et al. 2010 randomised 25 patients with early AMD to 20 mg/day saffron or placebo for 90 days in a crossover design, with focal electroretinogram (fERG) as the primary outcome. The saffron arm showed statistically significant improvements in fERG amplitude and modulation threshold, plus a small but significant increase in Snellen visual acuity Falsini et al. 2010. Piccardi et al. 2012 followed a subset for an average of 14 months and showed the retinal-function gains were sustained, not transient Piccardi et al. 2012. An independent Australian RCT replicated the visual-acuity signal in mild-moderate AMD Broadhead et al. 2019. This is a small, single-group-of-investigators evidence base but the directional consistency is high; AREDS-2 antioxidant supplementation already establishes that anti-oxidative interventions can shift the AMD trajectory, so the prior is plausible.
Evidence (appetite / weight)
Gout et al. 2010 is the foundational trial: 8 weeks of Satiereal 176.5 mg/day vs placebo in 60 mildly overweight women (BMI 25–28), no dietary restriction imposed. The Satiereal arm reduced snacking frequency by 55% vs 28% on placebo, and lost 1.65 kg vs 0.95 kg over 8 weeks Gout et al. 2010. This is a single industry-sponsored trial; the weight loss is modest and consistent with the snack-suppression mechanism rather than thermogenesis or absorption blockade.
Evidence (sexual function)
Two RCTs from the same Tehran group address SSRI-induced sexual dysfunction. Modabbernia et al. 2012 randomised men whose depression was stabilised on fluoxetine to 30 mg/day saffron or placebo for 4 weeks; the International Index of Erectile Function (IIEF) improved significantly on saffron, in erectile and intercourse-satisfaction domains Modabbernia et al. 2012. Kashani et al. 2013 ran the analogous trial in women on fluoxetine 40 mg/day and found significant improvement in arousal, lubrication and pain on the Female Sexual Function Index, though not in desire, orgasm or satisfaction Kashani et al. 2013. Both are small (n≈34–38), single-centre.
Evidence (sleep)
Lopresti et al. 2020 randomised 63 healthy adults with self-reported poor sleep to affron 28 mg/day (14 mg b.i.d.) or placebo for 28 days, with significant improvements on the Insomnia Severity Index and Restorative Sleep Questionnaire on saffron Lopresti et al. 2020. A subsequent dose-finding study showed 14 mg and 28 mg before bed were similarly effective. This is a thinner evidence layer than depression or PMS but the direction is consistent.
Protocol
Across the antidepressant, PMS, AMD-vision and sexual-function trials, the converging effective dose is 30 mg/day of standardised stigma extract, taken in two divided doses (morning and evening) with food, for a trial period of at least 6–8 weeks. The Satiereal appetite extract is a separate product at a higher mass dose (176.5 mg/day) because it is less concentrated. The affron extract (28 mg/day) is the standardised marker used in most Western RCTs; the mass is slightly below the 30 mg Iranian-trial dose but the active-compound content is matched.
Contraindications
Pregnancy is the firm contraindication: high doses of saffron stigma (≥5 g/day, far above supplement levels) stimulate uterine contractions and animal models show fetal toxicity; even at supplement doses there is no safety data, and traditional Iranian medicine warned against it in pregnancy Hosseinzadeh and Nassiri-Asl 2014. Breastfeeding has no data and is treated cautiously. Bipolar disorder is a theoretical contraindication on the monoaminergic mechanism — the same caveat that applies to SSRIs and SAMe. At supplement doses (≤100 mg/day) saffron has not been associated with serotonin syndrome in trials, but the theoretical interaction with SSRIs is real; the Kashani and Modabbernia add-on trials demonstrate the combination is tolerable in monitored settings, but combined use should be clinician-supervised. Acute toxicity threshold is ~5 g (over 150× the supplement dose); lethal threshold ~20 g.
Misconceptions
Three. (1) Culinary saffron — a few threads in a paella — does not deliver a pharmacological dose; the antidepressant effect requires the 28–30 mg/day standardised extract, not the spice cabinet. (2) Saffron is not a "natural SSRI" in the lay sense of "weaker but safer"; the mechanism overlaps but is broader (MAO-A, NMDA, GABA, BDNF), and the evidence base is for mild-to-moderate depression, not severe or treatment-resistant. (3) Marketing increasingly bundles saffron with "lookmaxxing" claims — skin glow, hair, sexual potency in the absence of any SSRI baseline — that are not supported by the trial evidence. The retinal and PMS effects are real; the rest is extrapolation.
Practicalities
Standardised extract (affron, Satiereal, or saffronSelect) costs roughly $12–20 per month at supplement-retailer prices, putting the annual cost at $150–250. Bulk saffron threads — even from reputable origin (Iran, Spain La Mancha, Kashmir) — are not a sensible substitute because dose verification is impractical at the kitchen scale and because adulteration of bulk saffron is widespread.
Failure modes
Three common reasons a saffron trial appears not to work. (1) Adulterated product. Market testing of bulk saffron has found 30–40% of samples contaminated with safflower, marigold, beet, or other red plant matter; even ISO 3632-classified saffron can be padded up to ~10–20% by weight with adulterants invisible to that standard Petrosino et al. 2018. Standardised extracts with HPLC-verified active-compound content (affron, Satiereal) bypass this risk, but cheap "saffron extract" without standardisation is a different product. (2) Trial too short. Trial outcomes converge at 6–8 weeks; many users abandon at 2 weeks for "not working", which would also be premature for an SSRI. (3) Wrong indication. Saffron's evidence is concentrated in mild-to-moderate depression with felt anhedonia, PMS, retinal preservation, and SSRI side-effect rescue. For severe depression, ADHD, generalised anxiety as a standalone, or weight loss without snacking pattern, the evidence is thin and saffron is the wrong tool.
History
Saffron cultivation is documented in Crete by 1500 BCE and Persia by ~1000 BCE; it appears in the Ebers Papyrus as a remedy, in the Greek medical corpus, and prominently in Avicenna's Canon of Medicine as an antidepressant and cardiotonic Hosseinzadeh and Nassiri-Asl 2014. The modern clinical-trial programme began with the Tehran University of Medical Sciences group around 2001–2005, which is why the literature is concentrated in Iran, the world's largest saffron producer (~85% of world supply).
The credibility range
Optimist case
Saffron is one of the few botanicals with multiple RCTs across multiple indications, replicated across countries, with effect sizes for depression on a par with first-line SSRIs and a side-effect profile that strikingly excludes the sexual dysfunction, weight gain, and emotional blunting that drive SSRI discontinuation. The mechanism is broader than any single-target antidepressant (MAO inhibition, monoamine reuptake, NMDA, GABA, BDNF, antioxidant), which fits the polypharmacological reality that lots of effective mood agents share. For PMS, AMD, and SSRI-induced sexual dysfunction, the placebo-vs-saffron deltas are large enough that the underlying mechanism is almost certainly real. The 30 mg dose is well below the toxicity threshold (~5 g) by more than two orders of magnitude. This is a serious, real intervention with an unusually wide therapeutic window.
Skeptic case
The majority of the depression evidence is from a small number of Iranian centres, with trial sizes mostly 30–60 patients per arm, and several with risk-of-bias concerns (allocation concealment, blinding integrity given saffron's distinctive smell, primary-endpoint reporting). The largest meta-analyses pool fewer total participants than a single phase-III SSRI registration trial. The PMS trial is single-centre and never independently replicated at scale. The AMD evidence is largely from one Italian group with small n. The Satiereal weight-loss trial is industry-sponsored and underpowered for body-composition outcomes. Publication bias is plausible: a botanical with the saffron literature's pattern (small positive trials, single-region cluster) is exactly what funnel-plot asymmetry tends to flag. And mechanism plausibility is not the same as effect — the same MAO-inhibition logic was invoked for St. John's wort, which now has equivocal real-world performance.
Author's call
Saffron at 28–30 mg/day of a standardised extract is a real intervention for mild-to-moderate depression, PMS symptom severity, and SSRI-induced sexual side effects. The depression and PMS effects are at the level where they deserve an honest mention as a non-prescription option, especially for readers who have tried or rejected SSRIs. The AMD signal is promising but should not be presented as practice-changing without the larger replications. The appetite, sleep, and general-cognitive claims are real-but-modest and should be framed as secondary, not headline. The right framing is "a genuinely effective botanical with a narrow but real evidence base — not a panacea, not an SSRI substitute for severe depression, but a reasonable monthly $15 trial for the specific indications where it has been tested." Evidence rating: 3 (multiple replicated RCTs in the moderate-quality band; one strong meta-analysis; not at the guideline-recommendation tier yet). Controversy rating: 2 (the field is broadly accepting; methodological critiques are about evidence quality, not direction).
Stakeholder and incentive map
- Commercial pushers. Pharmactive (affron), Inoreal (Satiereal), Activ'Inside (safr'Inside) — extract manufacturers with patents and a clear interest in the clinical literature growing. Most Western RCTs since 2018 are at least partly industry-sponsored. The supplement retailers (Life Extension, Thorne, Now Foods) market saffron at "natural mood support" with claims that often exceed the evidence on appetite and beauty.
- Cultural / academic pushers. Iranian academic medicine has institutional incentive — saffron is a national export and a heritage substance — which is why the trial cluster is concentrated there. This is not a fatal bias but the reader (and the meta-analyst) should account for it.
- Counter-incentives. Pharmaceutical antidepressant manufacturers compete with any over-the-counter botanical that reaches the mood market. SSRI prescribers are appropriately cautious about uncontrolled supplementation in patients on serotonergic drugs. Insurance does not cover saffron.
- Skeptic community. Examine, Cochrane-style reviewers, and academic psychiatrists who took the same critical stance on St. John's wort generally rate saffron more favourably than St. John's wort — the depression evidence is structurally similar but methodologically a bit tighter.
Population variability
- Sex. The PMS, sexual-dysfunction-in-women, and Satiereal appetite trials are female-only by design. The men's sexual-dysfunction and most depression trials are mixed. There is no signal that depression response differs by sex.
- Age. Depression evidence is in adults 18–55. Adolescents (12–16) tested in Lopresti 2018 also responded. AMD evidence is in older adults (~55–85). Saffron has not been studied in children under 12 outside isolated case reports.
- Baseline severity. Effect sizes are largest in mild-to-moderate depression and PMS. Severe MDD, psychotic depression, and treatment-resistant depression have not been adequately studied; the assumption should be saffron is not the right tool there.
- Genetics / metabolism. Crocin metabolism to crocetin in the gut is microbiome-dependent, which may explain inter-individual variability but is not well characterised in humans. CYP450 interaction profile is light at the 30 mg dose.
Knowledge gaps
- Large multi-centre Western RCT against an SSRI in formally diagnosed MDD — has not been done. Would settle the "non-inferior to fluoxetine" claim in a regulatory-grade fashion.
- Long-term safety at 30 mg/day for years. Most trials are 6–12 weeks; the AMD longitudinal extension to 14 months is the longest published.
- Saffron in pregnancy and lactation — no safety data and unlikely to be ethically generated; will remain a precautionary contraindication.
- AMD replication in a large independent trial. The Italian and Australian work is suggestive; we need an AREDS-2-style trial to make this guideline-relevant.
- Mechanism in vivo in humans — most receptor-affinity work is animal or in vitro. PET / pharmacokinetic studies in humans at 30 mg are sparse.
- Effect on emotional blunting and anhedonia specifically, separable from depression generally. Early signals (e.g., the anhedonia-specific affron trial) are promising; replication needed.
Coverage vs the brief. The input description named four consequences — mood, appetite, vision, PMS. All four are covered in the meta and body. Mood and PMS carry the article's centre of gravity because the trial evidence is strongest there; vision (AMD) and appetite (snacking on the Satiereal dose) are framed as real-but-secondary, not headline. SSRI-induced sexual dysfunction rescue is added to the trial coverage even though the brief didn't name it, because the Kashani and Modabbernia RCTs are the cleanest concrete payoff this substance has and the omission would have been misleading.
Scoping calls.
- Culinary saffron — paella, biryani, risotto — is mentioned only to disambiguate from the supplement. The pharmacological dose isn't reached at kitchen scale, and the food-use story is a different entry if it warrants one at all.
- Cardiometabolic outcomes in diabetes and prediabetes (a 2024 meta-analysis exists) are out of scope. They are an emerging signal, not a load-bearing one for the substance's identity here.
- Saffron in schizophrenia and as a chemotherapy adjunct — both have small trial literatures — are clinician-supervised contexts not appropriate for a supplement-tier entry.
- The lookmaxxing-adjacent marketing claims (skin, hair, libido without an SSRI baseline) are addressed as misconceptions rather than separate dimensions, because scoring them non-zero would have meant fabricating evidence the literature doesn't carry.
Rating difficulties.
- Evidence at 3: Marx 2019's meta-analysis would justify a 4 on its own, but the regional cluster (mostly Iranian centres), small per-trial sample sizes (mostly n=30–60/arm), and absence of any guideline-level recommendation hold it at 3. A large Western multi-centre replication would move it to 4.
- Longevity at 0: AMD slowing is disease-prevention, which technically belongs on this axis, but the evidence base is one Italian group plus one Australian replication at small n on a single condition, and there's no mortality signal. Conservative 0 chosen over a marginal 1.
- Sleep at 2: One small Lopresti trial plus a dose-finding replication from the same group. Could be argued at 1; left at 2 because the ISI/RSQ deltas were real and the affron extract is the same one used in the depression trials, so the cross-indication consistency raises confidence.
- Applicability at 3: Counted on the decision audience (anyone considering a non-prescription mood option, anyone on an SSRI with sexual side effects, anyone with significant PMS) rather than only current saffron users. This matches the avoidance / decision audience rule in the meta spec.
Future-link candidates. SSRIs as their own entry (the trial comparator); SAMe and St. John's wort as the other botanical antidepressants; AREDS-2 for established AMD; luteal-phase SSRIs, magnesium, B6, calcium, and cycle-tracked exercise as the wider PMS landscape; morning sunlight / light therapy for the seasonal-mood reader. None of these need to exist for this entry to stand, but the out-of-scope closing section assumes they will exist later.
Separate-entry candidates. SAMe, St. John's wort, and possibly a dedicated entry on emotional-snacking patterns (where Satiereal's mechanism actually sits). The cardiometabolic literature on saffron in diabetes may also be substantial enough to warrant a brief crossref entry once the evidence body grows.
Dream narrative. Written below the 40-score threshold (overall computed at ~20) because the entry's honest hook genuinely is a relief-leaning do, not a debunking. The dek and tagline lift the marketing-words ban only modestly — the "fluoxetine-without-the-trade" register is grounded in the head-to-head trial result, not extrapolated.
Saffron
Two capsules a day with food. The only ongoing effort is remembering to take it; the 6–8-week response window means a real trial requires sustained adherence rather than a one-off.
Standardised extracts (affron, Satiereal) run $12–20/month, ~$150–250/year. Falls in the minor band — not negligible but well below most other supplement subscriptions.
Multiple RCTs at 30 mg/day for 6 weeks show saffron outperforms placebo on the Hamilton Depression Rating Scale and is non-inferior to fluoxetine 20 mg and imipramine in mild-to-moderate depression (Akhondzadeh 2005, Noorbala 2005, Lopresti 2014 review, Marx 2019 meta-analysis of 11 RCTs). Clear stabilisation of inner life in the responder; not transformative (severe MDD untested).
Multiple small-to-moderate RCTs and at least one solid meta-analysis (Marx et al. 2019, 11 RCTs) for depression and PMS; mechanistic plausibility (MAO-A inhibition, monoamine reuptake, BDNF). Held back from 4 by regional clustering of trials in Iran, small per-trial sample sizes (mostly n=30–60/arm), and no guideline-level recommendation.
Within weeks, the documented felt effects are large reductions in PMS symptom severity at 30 mg/day (Agha-Hosseini et al. 2008: 76% achieved ≥50% symptom reduction vs 8% on placebo) and rescue of SSRI-induced sexual dysfunction in both sexes (Kashani 2013, Modabbernia 2012). Real day-to-day functional improvement for the indication subgroups, not a general wellness lift.
Lopresti et al. 2020 (n=63, 28 days) showed affron 28 mg/day improved Insomnia Severity Index and Restorative Sleep Questionnaire scores vs placebo; a follow-on dose-finding RCT replicated. Real but single-research-group evidence base; modest effect.