Field manual — single compounds, not whole-diet evidence
Do supplements actually help mood?
Food first: whole-diet evidence is consistently stronger than single-nutrient evidence — see Food & Mood before this page, not instead of it. "Natural" is not "safe": several compounds below can injure organs or interfere with prescription medication, and supplements aren't regulated like drugs — dose, purity, and contamination vary between products and batches. Tell your prescriber and your pharmacist about everything you take. Pharmacists check interactions for free; that's what they're for, and that one sentence does more good than any entry on this page.
Nothing on this page is a reason to stop, start, or change a prescribed medication. That decision belongs with the prescriber who started it — bring the page to them instead.
A few supplements have genuine trial support for mood; several widely-sold ones don't, and at least three carry safety risks bigger than most people assume.
- St John's Wort works and is dangerous to combine — both are true at once.
- Vitamin D helps only if you're deficient — a large trial found no prevention effect.
- Ashwagandha and kava have real efficacy signals and real safety flags.
- The site sells nothing here and recommends no brand.
Well-supported = backed by replicated randomized controlled trials · Promising = smaller studies or a single trial not yet replicated · anecdotal = clinical report only, no controlled studies. This page is psychoeducational, not a diagnostic or treatment tool.
Depression-side entries
Omega-3 (EPA-dominant) Promising
What it is: a fatty acid, usually from fish oil.
What the evidence shows: meta-analytic support is modest; EPA-predominant formulations outperform DHA-dominant ones.
What trials used: participants commonly received roughly 1–2 g/day of EPA — reported as what was studied, not a recommendation.
Who should not take it: anyone on anticoagulants, or before surgery, without medical supervision.
What it interacts with: blood thinners — higher intakes raise bleeding risk.
Bottom line: a reasonable adjunct with modest evidence, not a primary treatment.
Vitamin D Promising for deficiency, null for prevention
What it is: a hormone-like vitamin, synthesized from sunlight or taken orally.
What the evidence shows: deficiency correlates with low mood, but the VITAL trial (Okereke et al., JAMA, 2020) found no effect on preventing depression in a large randomized sample.
What trials used: VITAL administered 2,000 IU/day — reported as what was studied.
Who should not take it: generally low-risk, but high doses can cause toxicity over time.
What it interacts with: thiazide diuretics (raises calcium); some anticonvulsants.
Bottom line: correcting a documented deficiency is a different action from taking it hoping for a mood benefit. Test, don't guess.
Magnesium (glycinate / citrate) Promising
What it is: an essential mineral, several supplement forms with different absorption.
What the evidence shows: thinner than the internet suggests — the main trial (Tarleton, 2017) was open-label. Serum magnesium is a poor marker of body stores, so "most people are deficient" isn't established.
What trials used: glycinate and citrate forms, better tolerated than oxide — reported as what was studied.
Who should not take it: anyone with significant renal impairment — the kidneys clear it, and it can accumulate.
What it interacts with: some antibiotics and osteoporosis medications (absorption changes).
Bottom line: low risk for most people, but the mood evidence doesn't yet clear a high bar.
B12 and folate Well-supported for deficiency
What it is: B-vitamins essential for nervous-system function.
What the evidence shows: genuine deficiency causes depressive and cognitive symptoms, and is common in vegan diets, after bariatric surgery, on metformin, and with age.
What trials used: standard repletion doses for documented deficiency.
Who should not take it: no major contraindication at typical doses.
What it interacts with: metformin and acid-reducing medications lower B12 absorption over time.
Bottom line: the action here is a blood test, not a bottle.
L-methylfolate Promising
What it is: the active form of folate, marketed heavily around MTHFR genotype.
What the evidence shows: the genotype marketing outruns the evidence; some adjunctive trial support exists in treatment-resistant depression specifically.
What trials used: adjunctive dosing alongside an existing antidepressant.
Who should not take it: not indicated as a first move for most people.
What it interacts with: no major interaction beyond general folate cautions.
Bottom line: genotype testing is not indicated for this decision; talk to a prescriber if treatment-resistant depression is the actual situation.
SAMe Promising
What it is: a compound involved in neurotransmitter synthesis.
What the evidence shows: adjunctive trial support for depression.
What trials used: adjunctive dosing alongside existing treatment.
Who should not take it: anyone with bipolar disorder, diagnosed or not — it can precipitate mania. This is the headline risk, not a footnote.
What it interacts with: serotonin syndrome risk with SSRIs/SNRIs.
Bottom line: real support, real psychiatric risk — not a self-start decision.
5-HTP Anecdotal
What it is: a serotonin precursor, one step past tryptophan.
What the evidence shows: weak, small, and dated trial base for depression specifically.
What trials used: varies widely across small older trials.
Who should not take it: anyone on serotonergic medication.
What it interacts with: serotonin syndrome risk with any serotonergic medication, including SSRIs, SNRIs, triptans, and tramadol.
Bottom line: the risk-to-evidence ratio here is poor — probably not worth it.
St John's Wort
A Cochrane review (Linde et al.) found it comparable to standard antidepressants for mild-to-moderate depression across a large trial set. The efficacy evidence is genuinely strong — which is what makes the interactions below matter so much. Well-supported (efficacy)
It is a potent CYP3A4 inducer and reduces blood levels of many drugs:
- Hormonal contraceptives — breakthrough bleeding and contraceptive failure
- Warfarin and other anticoagulants
- HIV antiretrovirals
- Some chemotherapy agents
- Transplant immunosuppressants (e.g. ciclosporin) — organ rejection has been reported
- SSRIs / SNRIs / triptans — serotonin syndrome
- Also photosensitivity, and discontinuation effects
It's well-evidenced and sold beside the vitamins, so people take it without telling anyone. Nobody should start it without the prescriber and pharmacist knowing.
Anxiety-side entries
L-theanine Promising
What it is: an amino acid found in tea.
What the evidence shows: modest acute calming without sedation.
What trials used: single doses in the low-hundreds-of-mg range for acute effects — reported as what was studied.
Who should not take it: no major contraindication known.
What it interacts with: no major interaction documented.
Bottom line: modest evidence, low risk, few known interactions. Not a treatment.
Silexan (standardized oral lavender oil) Promising
What it is: a specific, standardized lavender oil preparation — not generic lavender capsules or aromatherapy oil.
What the evidence shows: better evidenced than most people expect, with trials against both an SSRI and a benzodiazepine comparator.
What trials used: the specific Silexan preparation, orally — the tested product matters here.
Who should not take it: no major contraindication known for the standardized preparation.
What it interacts with: no major documented interaction.
Bottom line: genuinely evidenced, but specifically for Silexan — substituting a different lavender product isn't the same thing.
Ashwagandha Promising for cortisol/stress markers, with a safety flag
What it is: a root used in Ayurvedic medicine, sold widely as an "adaptogen."
What the evidence shows: real reductions in cortisol and self-reported stress in trials. The evidence stops there being simple.
What trials used: varies by preparation and extract concentration across studies.
Who should not take it: pregnancy; thyroid disease (it can raise thyroid hormone levels); anyone on immunosuppressants (it has immunostimulant activity); anyone with existing liver disease.
What it interacts with: immunosuppressants; thyroid medication.
Bottom line: the stress-marker evidence is real, and so is the liver signal — see below. Anyone taking it should know what jaundice and dark urine look like and stop immediately if they appear.
Ashwagandha is a documented cause of herb-induced liver injury. NIH's LiverTox database describes mostly cholestatic, self-limited injury — but also rare fatal cases and cases requiring emergency liver transplantation, particularly with pre-existing liver disease; the Björnsson DILIN / Iceland case series is the standard reference. As of this writing (2026), Denmark has banned it in food supplements, and Dutch, French, German, Swedish, Belgian, and Polish authorities have raised concerns or advised against its use; the UK Food Standards Agency ran a public consultation and has not yet published a final decision. Regulatory status is actively moving — check current guidance in your country before deciding.
Kava Promising (efficacy), serious safety concern
What it is: a root from the Pacific Islands, acts on GABA receptors.
What the evidence shows: short-term anxiolytic trials are genuinely positive.
What trials used: standardized kavalactone extracts, short-term, supervised.
Who should not take it: anyone with existing liver disease, anyone drinking alcohol, anyone taking other hepatotoxic drugs.
What it interacts with: alcohol and other liver-affecting substances; some CYP-metabolized medications.
Bottom line: hepatotoxicity including liver failure has driven regulatory restriction in multiple countries since the early 2000s (it remains restricted in much of Europe; legal status varies by US state and by country). Short-term and supervised only, never with alcohol.
Valerian, passionflower, CBD Anecdotal
What it is: three separate botanicals, commonly grouped together in sleep and calm products.
What the evidence shows: weak and inconsistent across trials for anxiety specifically. Evidence does not currently support meaningful anxiety benefit for any of the three at typical consumer doses.
What trials used: varies widely, inconsistent dosing across small trials.
Who should not take it: CBD can affect liver enzyme metabolism of other drugs; valerian can add to sedation with other sedatives.
What it interacts with: CBD interacts with several liver-metabolized medications; check with a pharmacist.
Bottom line: popular, low-harm for most people, but the mood and anxiety evidence doesn't support the marketing.
High-dose B6 Anecdotal
What it is: a B-vitamin included in most B-complex and many "stress formula" products.
What the evidence shows: no strong mood-specific benefit at high doses.
What trials used: not applicable — this entry exists as a safety flag.
Who should not take it: anyone taking multiple B-complex or "stress formula" products at once, since exposure stacks without anyone noticing.
What it interacts with: no drug interaction; the risk is dose accumulation across products.
Bottom line: high-dose B6 causes peripheral neuropathy, sometimes irreversible.
When this is not the lever
If you're managing an active eating disorder, this page's territory — restriction, rules about intake, "correcting" your body — is not the right one; go to Food & feelings — start here. If a supplement is being considered instead of assessment or treatment for a serious depressive episode, that's a conversation for a prescriber, not a purchase. If it's urgent, Support & crisis lines.
This site sells nothing and recommends no brand, anywhere.
Questions people ask
Which supplement has the best evidence for depression?
St John's Wort has the strongest efficacy evidence on this page — a Cochrane review found it comparable to standard antidepressants for mild-to-moderate depression. It's also the one with the most serious interaction risk, since it reduces blood levels of many prescription drugs. The two facts belong together, not apart.
Is it safe to take St John's Wort with my antidepressant?
No — combining it with SSRIs, SNRIs, or triptans carries a serotonin syndrome risk, and it also reduces the effectiveness of hormonal contraceptives, blood thinners, and several other prescription drugs. Tell your prescriber and pharmacist before starting it, every time.
Is ashwagandha safe?
The stress-marker evidence is real, and so is a documented liver-injury signal, including rare cases requiring transplantation. Several countries have restricted or are reviewing it as of this writing. It's contraindicated in pregnancy, thyroid disease, and alongside immunosuppressants — check current regulatory status before deciding.
Do magnesium supplements help with mood or anxiety?
The evidence is thinner than the internet suggests — the main trial was open-label, and a blood test is a poor way to estimate whole-body magnesium stores. It's not dangerous for most people, but it's not a proven mood treatment either, and it's contraindicated in significant kidney impairment.
Should I take 5-HTP or SAMe for depression?
Both carry a real serotonin syndrome risk with serotonergic medication, and SAMe can trigger mania in undiagnosed bipolar disorder. SAMe has some adjunctive trial support; 5-HTP's risk-to-evidence ratio is poor. Neither is a first move, and both need a prescriber's knowledge before starting.
Related
Sources
- Linde, K., et al. Cochrane review of St John's Wort for major depression.
- Okereke, O.I., et al. (2020). VITAL trial — vitamin D and depression prevention. JAMA, 324(5), 471–480.
- Tarleton, E.K., et al. (2017). Magnesium supplementation for depression (open-label trial). PLOS ONE.
- Björnsson, E., et al. — DILIN / Iceland case series on herb-induced liver injury; NIH LiverTox database entry for ashwagandha (Withania somnifera).
- Cochrane and case-report literature on kava hepatotoxicity and regulatory restriction.
- Kasper, S., et al. — Silexan randomized trials against SSRI and benzodiazepine comparators.
- National Alliance for Eating Disorders — helpline and referral information.
Regulatory status for ashwagandha and kava is current as of mid-2026 and is actively moving — verify the rules where you live before buying or using either.
Clinically reviewed by: not yet completed for this edition.