SAMe for Depression: Independent Evidence, Risks & Who Funds It

Table of contents

• Evidence summary
• What SAMe is
• How it works
• What works and what does not
• Benefits by claim
• Risks and all side effects
• All interactions
• Who should avoid SAMe
• Doses studied, forms and product quality
• Follow the money: who makes it and who funded the evidence
• Related research
• Frequently asked questions
• Sources and funding notes

Evidence summary

Independent evidence and credibility scorecard

Independence tiers: 1 = regulator or government body; 2 = academic or Cochrane team with no declared commercial funding for the work; 3 = publicly or academically funded work with a material industry tie (donated product, author ties to SAMe makers), or text written by a licence holder; 4 = manufacturer-funded or manufacturer-authored. Credibility runs from A (highest) to D.

Five of the twelve sources are from the US, and the two public-funded US trials came from one research group. The evidence base is less international than it looks.

What SAMe is

S-adenosyl-L-methionine is a natural chemical found in every cell. It is also called S-adenosylmethionine, SAMe, SAM-e, SAM or AdoMet, and ademetionine in Europe. The body makes it from methionine, an amino acid in food. It was discovered in the early 1950s (NCCIH). Its ATC drug code is A16AA02, which groups it with amino acids and derivatives used in metabolism (Samyr SmPC 5.1).

Forms and salts

• Stabilised salts: SAMe is chemically unstable. The Italian label explains that oral SAMe has to be stabilised as a salt. Products use the 1,4-butanedisulfonate salt or a sulfate p-toluenesulfonate (tosylate) salt (Samyr SmPC 2, 5.1). The NIH-funded US escitalopram comparison trial used SAMe tosylate (Mischoulon 2014).
• Salt weight vs SAMe content: in Samyr, 760 mg of the butanedisulfonate salt provides 400 mg of SAMe (Samyr SmPC 2). A label that gives only the salt weight overstates the active amount.
• Forms: enteric-coated (gastro-resistant) tablets; in Europe also powder for intramuscular or intravenous injection. Many early depression trials used injections, which NCCIH warns "may not have the same effects" as oral supplements (NCCIH).

Legal status by country

How it works

SAMe is the body's main donor of "methyl groups", small chemical tags that cells add to DNA, proteins, fats and neurotransmitters. It is also the starting point for making cysteine and glutathione, the liver's main antioxidant. Folate and vitamin B12 are needed to recycle it (Samyr SmPC 5.1).

What is shown in humans:

• Oral SAMe is absorbed. Peak blood levels of 0.5–1 mg/L occur 3–5 hours after single doses of 400–1,000 mg and fall to baseline within 24 hours. Absorption is better on an empty stomach (Samyr SmPC 5.2).
• Oral or injected SAMe raises levels in cerebrospinal fluid, the fluid around the brain. Low SAMe levels in cerebrospinal fluid have been reported in people with depression (Sharma 2017).
• Absorption differs between products. A pharmacokinetic study of the MSI-195 formulation (Cameron 2020) reports its blood levels; we did not verify the exact ratios.

What is not proven: SAMe may lift mood by increasing the turnover of serotonin and noradrenaline and by changing brain cell membranes (Samyr SmPC 5.1). This rests mainly on laboratory and animal work. No human study has shown which mechanism, if any, explains a clinical benefit. An Australian trial found no genetic or blood marker that predicted who responded (Sarris 2018).

What works and what does not

Benefits by claim

Depression: SAMe on its own

The US government's 2002 AHRQ report pooled depression trials published up to 2000. We could not verify its pooled numbers from the sources we opened. It noted that most trials were small and of varying quality (AHRQ 2002).

Cochrane took a stricter approach in 2016, accepting only trials in adults with diagnosed major depression. Only 2 placebo-controlled trials of SAMe alone qualified (142 people). The result, SMD −0.54 (95% CI −1.54 to 0.46), was not statistically significant and was rated very low quality (Galizia et al. 2016).

The 2024 meta-analysis by Limveeraprajak and colleagues used wider entry rules and found 11 trials of SAMe alone against placebo. SAMe was better, with SMD −0.58 (−0.93 to −0.23). That is a moderate effect, but the trials disagreed with each other a lot (I² 68%). The authors rated the certainty as moderate (Limveeraprajak 2024). A second 2024 meta-analysis of 14 trials was seen only through trade coverage, so its placebo result is unverified (Peng 2024).

The most recent placebo-controlled trial of oral SAMe alone was small. In 49 unmedicated adults, SAMe 800 mg a day did not beat placebo significantly, and the placebo response was high, about 53% (Sarris 2020).

A larger US trial that compared SAMe alone with both placebo and escitalopram was also negative; it is described below.

What this means: SAMe alone may help some people with milder depression, but the result depends on which trials are counted. The positive pooled result rests heavily on older trials, several using injections.

Depression: SAMe added to an antidepressant

This is how SAMe is most often promoted in the US. The evidence comes from four trials:

• Papakostas 2010 (positive): 73 people whose depression had not responded to a serotonin reuptake inhibitor antidepressant took SAMe (target 800 mg twice a day) or placebo for 6 weeks. Response was 36.1% vs 17.6% and remission 25.8% vs 11.7%, a number needed to treat of about 6 for response (Papakostas 2010). Cochrane rated the add-on evidence as low quality (Galizia 2016). A published letter questioned discrepancies in the reported response and remission rates, and the authors replied (Fleisch et al. 2010).
• Sarris 2018 (negative): 107 Australian outpatients took SAMe 800 mg a day or placebo with their antidepressant for 8 weeks. Response was 54.3% vs 50.0% and remission 43.5% vs 38.3%, with no significant difference (Sarris 2018).
• Targum 2018, the "Horizon" trial (negative): a proprietary SAMe formulation (MSI-195) added to ongoing antidepressants. The company announced that its primary endpoint was missed; a post hoc subgroup excluding patients with high BMI and unstable symptoms was reported as positive. The published paper randomised 234 people (118 MSI-195, 116 placebo); we did not verify the site count (Targum 2018; NCT01912196).
• Pooled: across 5 add-on trials, SMD −0.16 (−0.44 to 0.13), I² 57%, not significant (Limveeraprajak 2024).

Guidelines have moved with these results. In 2016 the Canadian CANMAT guideline listed SAMe among first- or second-line options for mild to moderate depression (Ravindran 2016). The 2022 WFSBP/CANMAT taskforce gave add-on SAMe only a weak recommendation and did not support SAMe alone (Sarris et al. 2022).

A 2025 network meta-analysis of nutraceutical trials (Cheng 2025) reportedly ranked SAMe highly. We could not confirm its SAMe effect sizes or funding statement, and network models estimate many comparisons indirectly, so we do not rely on this ranking.

Depression: SAMe vs prescription antidepressants

Several trials compared SAMe with imipramine, an older tricyclic antidepressant; the pooled figures could not be verified from our search (Galizia 2016). The largest SAMe–imipramine trials were two Italian studies of 1,600 mg oral or 400 mg injected SAMe; their abstract gives no funding statement (Delle Chiaie 2002).

The only trial against a modern antidepressant was funded by the US National Center for Complementary and Alternative Medicine (NCCAM, now NCCIH). It randomised 189 outpatients to SAMe 1,600–3,200 mg a day, escitalopram or placebo for 12 weeks. Response was 36% on SAMe, 34% on escitalopram and 30% on placebo; none of the differences was significant (Mischoulon 2014). The authors called it a failed trial: because even the standard drug did not beat placebo, it cannot show whether SAMe works.

How quickly does SAMe work?

A lay health website states that the antidepressant effect appears within 5–7 days. We could not find this in the Italian SmPC sections we checked, so it is unverified (a non-authoritative lay website). We found no independent trial designed to test speed of onset. The AHRQ estimate was measured at 3 weeks (AHRQ 2002).

Osteoarthritis

The idea came from depression trials, where some participants said their joint pain improved (NCCIH). AHRQ reported osteoarthritis findings that we could not verify here (AHRQ 2002). The Cochrane review (Rutjes 2009) pooled 4 trials (656 people) and found a small pain effect whose confidence interval includes zero (Rutjes et al. 2009).

Liver disease and cholestasis

• Alcohol-related liver disease: one Spanish placebo-controlled trial in alcoholic cirrhosis (dose and numbers not verified here). Its authors concluded that long-term SAMe may improve survival or delay transplantation, particularly in less advanced disease; the exact rates could not be verified (Mato 1999). The 2006 Cochrane review (Rambaldi & Gluud 2006) was not retrieved in full, so its trial count and conclusions are unverified here.
• Chronic liver disease generally: a 2015 meta-analysis found some liver blood tests improved, but no difference in long-term outcomes (Guo 2015).
• Cholestasis of pregnancy: The 2020 Cochrane review, with larger modern trials, found insufficient evidence that SAMe works (Walker 2020).

NCCIH's summary: whether SAMe benefits liver disease in people "has not been established" (NCCIH).

Risks and all side effects

NCCIH describes side effects as uncommon and usually minor, mainly nausea and digestive upset. Long-term safety data are limited, although one 2-year liver trial reported no serious side effects (NCCIH).

Common and less common side effects

The Italian SmPC pools trials in 2,115 people (Samyr SmPC 4.8):

• Common (1–10%): abdominal pain, diarrhoea, nausea, headache, anxiety, insomnia, itching.
• Uncommon (0.1–1%): dry mouth, indigestion, wind, vomiting, gastrointestinal bleeding, dizziness, tingling, agitation, confusion, joint pain, muscle spasms, sweating, flushing, low blood pressure, swelling, fever, urinary infection, allergic reactions, and injection-site reactions with injectable forms.
• Rare: bloating, inflammation of the oesophagus, biliary colic, suicidal thoughts or behaviour (in people being treated for depression).
• Heartburn: tablets are kept acidic for stability, which can cause heartburn and a feeling of fullness.

In the US escitalopram trial, stomach discomfort (19%) and diarrhoea (20%) were the most common SAMe complaints. Placebo patients also reported many stomach complaints.

Other safety notes

• Ammonia in cirrhosis: blood ammonia should be monitored in people with cirrhosis or high ammonia taking oral SAMe (Samyr SmPC 4.4).
• Immune suppression: SAMe enhances the growth of Pneumocystis, a fungus that infects people with weak immune systems. NCCIH calls this a theoretical concern (NCCIH).
• Driving: some people feel dizzy, so the label advises not driving until you know how it affects you (Samyr SmPC 4.7).
• Dependence: no cases of dependence have been reported (Samyr SmPC 4.8).

All interactions

Most interaction evidence is from case reports and mechanism, not formal interaction studies. The NIH-funded add-on trial gave SAMe alongside serotonin reuptake inhibitor antidepressants under trial monitoring (Papakostas 2010).

Who should avoid SAMe

• Bipolar disorder or a history of mania or hypomania: avoid except under a psychiatrist's supervision (NCCIH; Samyr SmPC).
• Inherited methionine-cycle disorders: contraindicated in homocystinuria, raised homocysteine and defects such as cystathionine beta-synthase deficiency or vitamin B12 metabolism disorders (Samyr SmPC 4.3).
• Possible B12 or folate deficiency: the label advises blood tests in people at risk (anaemia, liver disease, pregnancy, vegan diets) and correcting deficiency first (Samyr SmPC 4.4).
• Parkinson's disease on levodopa: discuss with your neurologist first (NCCIH).
• HIV or other immune suppression: theoretical infection concern (NCCIH).
• Pregnancy and breastfeeding: NCCIH says safety in pregnancy "has not been established". The Italian label reports no adverse effects in late-pregnancy cholestasis trials, but advises first-trimester use only if absolutely necessary and breastfeeding use only if benefit outweighs risk (NCCIH; Samyr SmPC 4.6).
• Children: safety and efficacy not established (Samyr SmPC 4.2).
• Cirrhosis: ammonia monitoring is advised (Samyr SmPC 4.4).
• Kidney disease: limited data; the label advises caution (Samyr SmPC 4.2).
• Surgery: we found no official guidance. The MSD Manual notes that SAMe acts as a platelet inhibitor in mechanistic studies (MSD Manual), so tell your surgical team you take it.

Doses studied, forms and product quality

The figures below are doses used in trials or stated on an official label. They are not a recommendation.

Across depression trials, daily doses ranged from 200 to 3,200 mg and trials lasted 2 to 12 weeks (Peng 2024). Dose may matter. As summarised by the MSD Manual, the WFSBP/CANMAT taskforce said lower-dose SAMe alone may not help, while higher doses may help as an add-on; the Australian trials used 800 mg a day (MSD Manual summary).

How it is taken

The Italian label says tablets should be swallowed whole, not chewed, and taken away from meals. They should be removed from the blister only just before use. Tablets that are not white to yellowish, for example because the foil is damaged, should not be used (Samyr SmPC 4.2).

Stability and product quality

• "Nearly all" SAMe supplements are enteric-coated to protect the SAMe from breaking down (ConsumerLab 2011). One older negative trial used a less stable oral form whose tablets degraded through exposure to air (Sharma 2017).
• ConsumerLab reported that about half of SAMe products tested in 2000 had too little SAMe. In 2003, 1 of 8 contained 30% of its claim (NutraIngredients 2003). In 2011, 3 of 10 failed: one had no SAMe, one had 75% of its claim, and a pet product's coating failed. Price per 400 mg ranged from 53 cents to over $3 (ConsumerLab 2011).
• Absorption varies between formulations (Cameron 2020, a study of one novel formulation). Trial results with one product may not apply to another.

Follow the money: who makes it and who funded the evidence

Who makes and sells SAMe

• Gnosis by Lesaffre (raw material, Italy/France): Gnosis has plants at Pisticci (Italy) and Sant'Antonino (Switzerland). The French yeast group Lesaffre took a majority stake in 2015 and full ownership in 2018 (NutraIngredients 2018). In paid content, Gnosis calls itself the first company to commercialise SAMe and the "global leader" in SAMe manufacturing; its brand is Adonat (Gnosis sponsored content). We could not verify market-share figures.
• Pharmavite / Nature Made (US supplement brand): based in West Hills, California (Pharmavite), and listed as a group company of Otsuka Holdings, Japan (Otsuka).
• Prescription ademetionine: in Italy, Samyr, Donamet and Transmetil are held by Viatris Italia S.r.l. (AIFA); the 2023 SmPC still named Mylan Italia (Samyr SmPC 7). Elsewhere, Abbott markets Heptral, Transmetil and Samyr (Abbott).
• MSI Methylation Sciences (Canada): developed the proprietary MSI-195 formulation and funded its trials (Targum 2018; Cameron 2020).

Who funded the trials and reviews

• US public money, company product: the positive 2010 add-on trial was funded by an NIMH career grant (K23 MH-069629) (PubMed grant record). Pharmavite supplied the SAMe and placebo and issued a press release on the result (Pharmavite release). The negative 2014 trial was funded by NCCAM grant R01AT001638. We could not verify who supplied the SAMe tosylate or placebo in that trial (Mischoulon 2014).
• Australian trials: we could not confirm the funders of the two Australian trials from the papers we opened (see gaps note).
• Company money: the Horizon trial of MSI-195 was funded by MSI Methylation Sciences. The company announced that its primary endpoint was missed and reported a post hoc subgroup as favourable (Targum 2018). We did not verify whether results are posted on its registry entry (NCT01912196).
• Reviews: Cochrane reviews registered before October 2020 fell under a 2014 policy saying they "cannot be funded or conducted by commercial sponsors" (Cochrane policy). We could not confirm the funders of the 2024 meta-analyses from their full texts. The 2017 APA work-group review lists no funding; we could not confirm company links for its authors (Sharma 2017).
• How companies use independent work: in 2024 Gnosis publicised both 2024 meta-analyses to promote SAMe for mood (NutraIngredients 2024). One of them found no significant benefit over placebo.

Documented integrity and regulatory events

• Germany (year not confirmed): a Cologne administrative court case about 200 mg SAMe capsules marketed as a food supplement found the product may not be classed as one; an appeal was possible (Apotheke Adhoc).
• India, 2023: FSSAI ordered SAMe supplements off shelves and e-commerce platforms (NutraIngredients 2023).
• EU: SAMe disulfate tosylate was classed as a possible novel food (CIRS 2026).
• Product failures: repeated ConsumerLab findings of products below their label claim, including one with no SAMe (ConsumerLab 2011).
• We did not find an FDA warning letter or FTC action specific to SAMe in our searches.

What this means: SAMe's case rests on decades of small European trials, then a handful of better modern trials. The modern trials were mostly paid for by governments, but the product often came from companies that sell SAMe. The publicly funded results lean negative. The favourable reviews and marketing lean on older or pooled data.

Related research

For wider, evidence-graded context, see these Pure City Research guides:

• Stress, anxiety and depression supplements: evidence and safety
• Major depressive disorder
• Treatment-resistant depression
• Bipolar depression
• St John's wort: depression evidence and interaction risks
• Saffron for depression
• St John's wort vs saffron vs SSRIs
• Omega-3: which form works
• Escitalopram
• Alcohol-related liver disease
• The COA transparency audit: supplement lab results

Frequently asked questions

Sources and funding notes

What we could not verify:

• Not verified in this fact-check: the AHRQ 2002 pooled figures; Mato 1999 sample size and rates and the 2015 Cochrane withdrawal; the Limveeraprajak funding statement; Peng 2024 results and funders; Cheng 2025 SAMe effect size; Guo 2015 counts; the CANMAT 2016 and 2023 SAMe placements and the 2022 WFSBP/CANMAT recommendation strength; the 5–7 day onset claim; the Donamet and Transmetil indications; the AIFA shortage notice; the 2018 Cologne ruling date; the funders of Sarris 2018 and 2020 and of Cameron 2020; the Mischoulon 2014 supplier of SAMe and placebo; the Galizia 2016 and Rutjes 2009 author conflict statements; the Sharma 2017 company ties; the Papakostas 2010 letter; the European Commission novel-food status; and the Targum 2018 site count.
• We could not open the "Sources of support" sections of the Cochrane depression (2016) and osteoarthritis (2009) reviews; we report their declared conflicts from PubMed.
• We could not read the full text of Papakostas 2010 or the 2010 letter and reply about its response rates. Funding is from PubMed's grant record and Pharmavite's press release.
• A search summary says a Sarris-group SAMe trial reported an NHMRC grant and co-sponsorship with FIT-BioCeuticals, a supplement company. We could not confirm which trial this refers to or check the full text.
• We could not find funding statements for the older Italian imipramine trials or confirm who sponsored the 2002 journal supplement they appeared in.
• The AHRQ summary mentions "funding agencies" without naming them, and the full report on NCBI Bookshelf blocked automated access.
• We did not open the European Commission's own novel-food entry; the EU status is as reported by a regulatory consultancy.
• We did not check funding of the 2022 WFSBP/CANMAT taskforce, whose lead author also led the Australian SAMe trials.
• We found no citable market-size figures, no complete list of SAMe raw-material makers, and no MHRA classification decision for the UK.

1. NCCIH. S-Adenosyl-L-Methionine (SAMe): In Depth (last updated January 2017) — US government fact sheet; NIH, funded by Congress.
2. NCCIH funding appropriations history — US government budget record.
3. AHRQ. SAMe for Treatment of Depression, Osteoarthritis, and Liver Disease. Evidence Report No. 64, summary (2002; archived); PubMed record — US government; Southern California EPC, contract 290-97-0001.
4. Galizia et al. SAMe for depression in adults. Cochrane 2016; plain-language summary — Cochrane review; author conflicts not verified.
5. Limveeraprajak N et al. Efficacy and acceptability of SAMe for depressed patients. Prog Neuropsychopharmacol Biol Psychiatry 2024 — Thailand; funding not verified in our search; competing interests listed as "None declared".
6. Peng TR et al. SAMe as an adjuvant therapy for depression. Gen Hosp Psychiatry 2024 — Taiwan; Buddhist Tzu Chi Medical Foundation; no conflicts.
7. Cheng YC et al. Nutraceuticals for depressive disorder: network meta-analysis. Psychol Med 2025 — Taiwan; Ministry of Science and Technology; no conflicts.
8. Papakostas GI et al. SAMe augmentation of SRIs. Am J Psychiatry 2010 — US; NIMH K23 MH-069629; SAMe supplied by Pharmavite.
9. Pharmavite press release, 31 August 2010 (reproduced by ScienceBlog) — company communication.
10. Fleisch SB et al. Discrepancy in response and remission rates. Am J Psychiatry 2010 (letter) — academic correspondence.
11. Mischoulon D et al. SAMe vs escitalopram vs placebo. J Clin Psychiatry 2014; full text; NCT00101452 — US; NCCAM R01AT001638; SAMe from Pharmavite.
12. Sarris J et al. Adjunctive SAMe in non-remittent MDD. Eur Neuropsychopharmacol 2018 — Australia; funder not confirmed in our search.
13. Sarris J et al. SAMe monotherapy for depression. Psychopharmacology 2020; ANZCTR record — Australia.
14. Targum SD et al. Augmentation study of MSI-195. J Psychiatr Res 2018; NCT01912196 — funded by MSI Methylation Sciences (Canada).
15. Cameron BR et al. Pharmacokinetics of MSI-195. BMC Pharmacol Toxicol 2020 — funding and author roles not verified in our search.
16. Sharma A et al. SAMe for neuropsychiatric disorders. J Clin Psychiatry 2017; full text — APA Council on Research work group; funding listed as none; company ties not confirmed.
17. Delle Chiaie R et al. SAMe vs imipramine in 2 multicentre studies. Am J Clin Nutr 2002 — Italy; funding not stated in abstract.
18. Ravindran AV et al. CANMAT 2016, complementary and alternative treatments — Canada; authors with pharma ties declared.
19. Lam RW et al. CANMAT 2023 update on major depressive disorder — Canada; funding not checked.
20. Sarris J et al. WFSBP/CANMAT nutraceutical guidelines. World J Biol Psychiatry 2022 — international taskforce; funding not checked.
21. NICE NG222: Depression in adults — UK government-funded guideline; no SAMe recommendation.
22. Rutjes AW et al. SAMe for osteoarthritis of the knee or hip. Cochrane 2009 — Switzerland; conflicts "None".
23. Rambaldi A, Gluud C. SAMe for alcoholic liver diseases. Cochrane 2006; 2015 withdrawal — Cochrane Hepato-Biliary Group.
24. Mato JM et al. SAMe in alcoholic liver cirrhosis. J Hepatol 1999 — Spain; funding not stated in abstract.
25. Guo T et al. SAMe for chronic liver disease. PLoS One 2015 — China; no competing interests declared.
26. Walker KF et al. Pharmacological interventions for cholestasis of pregnancy. Cochrane 2020 — UK/Australia; authors involved in trials of other drugs.
27. Samyr Summary of Product Characteristics (AIFA, made available 5 October 2023) — licence-holder text approved by the Italian regulator.
28. Donamet Summary of Product Characteristics (AIFA, made available 28 May 2026) — licence-holder text (Viatris Italia) approved by the Italian regulator.
29. AIFA database: Samyr, Transmetil, ademetionine — Italian regulator data (accessed 6 October 2026).
30. German Medicines Prescription Ordinance (AMVV), Annex 1 — German federal law.
31. Apotheke Adhoc: court rules ademetionine is a medicine (2018) — German pharmacy trade news.
32. CIRS: SAMe disulfate tosylate considered a novel food (2026) — regulatory consultancy report.
33. NutraIngredients: FSSAI orders removal of SAMe supplements (2023) — trade news.
34. Abbott: Heptral (ademetionine) — manufacturer product page.
35. Otsuka Holdings: group companies; Pharmavite — company ownership records.
36. NutraIngredients: Lesaffre completes acquisition of Gnosis (2018); Lesaffre — trade news; company site.
37. Gnosis by Lesaffre: Adonat sponsored content (2018) — paid company marketing.
38. NutraIngredients: Gnosis spotlights new SAMe data (2024) — trade news.
39. ConsumerLab: 30% of SAM-e supplements fail testing (2011); NutraIngredients on 2003 tests — private testing company; trade news.
40. Iruela LM et al. Toxic interaction of SAMe and clomipramine. Am J Psychiatry 1993 — case report.
41. Olsufka W, Abraham MA. Treatment-emergent hypomania possibly associated with supplements. Ment Health Clin 2017 — case report; no conflicts.
42. MSD Manual Professional: SAMe (updated July 2025) — reference text published by Merck & Co.
43. Cochrane commercial sponsorship policy (2014) — Cochrane (UK charity).

Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.