Where SAM-e earns its place: as the second move when a working antidepressant has stalled, and as an alternative to anti-inflammatory pills for arthritic knees or hips if your stomach has had enough of them. Everywhere else — generic mood boost, fatigue, energy — the trial signal is weak. One of the harder-to-judge supplements on the shelf because the older positive trials and the newer skeptical reviews don't quite line up.
Your cells run a constant chain of methylation reactions — tagging DNA, building neurotransmitters, patching membrane lipids — and SAM-e is the molecule that supplies the methyl tag. The body makes it from the amino acid methionine plus a hit of ATP, and most of the time it makes plenty (Bottiglieri 2002). The supplement is a bet that, in particular slumps, the methylation engine runs short and adding more substrate gets things moving — most credibly in mood (where it shows up in catecholamine and membrane-lipid pathways), plausibly in cartilage, and meaningfully in cholestatic liver, where the same molecule sustains the body's main antioxidant.
What it actually does — three claims, ranked by how good the trials are
Depression. The strongest contemporary evidence is for SAM-e added on top of an antidepressant that has worked partially and stalled, not as a stand-alone replacement.
The 2016 Cochrane review of monotherapy trials was unimpressed — eight trials, low to very low quality, no clear separation from placebo and no head-to-head difference from imipramine or escitalopram (Galizia et al. 2016). A recent 49-person monotherapy trial confirmed the null (Sarris et al. 2020). The honest read: as a second move after a partial SSRI response, the signal is small but real; as the first thing you reach for, it probably isn't.
Joints
Older European trials matched SAM-e against ibuprofen, naproxen, and (in a 2004 crossover) celecoxib. By month two, SAM-e at 1200 mg/day reached comparable pain and function scores to the prescription comparator, with milder gut side effects (Najm et al. 2004; Soeken et al. 2002). The catch comes from the direct placebo comparison: a Cochrane review of four placebo-controlled OA trials found a pain difference of about four millimetres on a ten-centimetre scale — too small to feel — and essentially no difference on function (Rutjes et al. 2009). Best interpretation: comparable to a working anti-inflammatory pill, which is itself only modestly better than placebo, with a much slower onset. The trade is acceptable if your stomach can no longer tolerate the daily ibuprofen.
Cholestatic liver
The maddening itch that comes with intrahepatic cholestasis of pregnancy responds to SAM-e at 1600 mg/day, alongside improvement in liver chemistry (Frezza et al. 1990). It's a second-line position behind ursodeoxycholic acid, the standard treatment, and earns its place when the first-line falls short. For alcoholic liver disease the picture is messier — a two-year trial in 123 patients with cirrhosis missed its primary endpoint of death-or-transplant prevention, though a sub-analysis hinted at benefit in less-advanced disease (Mato et al. 1999), and a smaller US trial was flat null (Medici et al. 2011).
How to dose, if you're going to do this
Dosing isn't subtle: the trials that worked used 1200–1600 mg a day, split into morning and afternoon. Lower doses haven't been adequately tested in modern trials and probably under-deliver.
At that dose, expect a yearly cost in the $150–$300 range — cheaper than most brand-name antidepressants out of pocket, well above the $30-a-year supplement bracket. The active ingredient is the same across brands; what matters is the salt form (look for 1,4-butanedisulfonate or tosylate disulfate), the enteric coating, and the blister pack.
When not to take it
The mild stuff — slight jitters, mild gut upset, the occasional run of insomnia — fades after a week or two for most people. The serious end is the same warning that comes with any antidepressant.
Why most people who try it feel nothing
Three patterns account for most of the disappointed-user reports.
Under-dose. Shelf-standard tablets are 200 mg or 400 mg. The trials that found an effect used 1200–1600 mg. People take one tablet a day, feel nothing after two weeks, and conclude the supplement is useless — when in fact they've been taking a quarter of the working dose.
Taken with breakfast. The pharmacokinetic data is clean: food roughly halves the amount absorbed (Aliani et al. 2020). A pill taken with coffee and toast is, in plasma terms, closer to half a pill.
Wrong indication. Bottles market it as a generic mood lift, but the trial evidence is for major depressive disorder at therapeutic doses — the same diagnosis a clinician would prescribe an SSRI for. For ordinary low mood, slept-badly fatigue, or social flatness, the odds of a felt effect are low and the cost of finding out is real.
What else to weigh it against
For depression: SSRIs and SNRIs are first-line, with a vastly larger and more consistent evidence base than SAM-e. Talk therapy, regular aerobic exercise, and bright morning light each have stronger trial support than SAM-e for monotherapy. SAM-e earns its slot specifically as a second move when one of those has worked partially and the response has stalled — and the evidence for that second move is good enough to be worth considering with your prescriber (Papakostas et al. 2010).
For arthritic joints: Ordinary anti-inflammatory pills are the comparator and the established standard, and the topical versions (diclofenac gel) deliver most of the benefit with a fraction of the stomach risk. Duloxetine has moved into the standard kit for chronic knee arthritis. Over a year, the exercise-and-weight-loss combination outperforms most pills for hip and knee arthritis on every endpoint.
For cholestatic itch: ursodeoxycholic acid is the first-line agent across the relevant indications; SAM-e is the second line.
Related territory worth looking into if SAM-e is on your radar: the methylation cofactors (folate and B12) that sit alongside it in the same biochemical loop; the MTHFR gene variants that change how that loop runs; omega-3 fatty acids as a parallel antidepressant add-on; topical anti-inflammatory gels for joints; and ursodeoxycholic acid for cholestatic liver conditions.
Substance and claimed effects
S-adenosyl-L-methionine (SAMe; also AdoMet, ademetionine) is an endogenous sulfonium compound formed from methionine and ATP by methionine adenosyltransferase. It is the body's principal methyl donor: SAMe participates in more than thirty-five known transmethylation reactions involving DNA, histones, phospholipids, catecholamines, melatonin, and creatine (Bottiglieri 2002). After donating its methyl group it becomes S-adenosylhomocysteine, which is hydrolysed to homocysteine and re-enters either the remethylation pathway (via methionine synthase, requiring folate and B12) or the transsulfuration pathway (to cysteine and glutathione). SAMe has been a prescription drug in Italy since 1979, in Spain since 1985, and in Germany since 1989; in the United States and most Commonwealth jurisdictions it is sold over the counter as a dietary supplement.
The claims worth covering, in order of evidence weight: (1) antidepressant effect, both as monotherapy and as adjunct to SSRIs/SNRIs in non-responders; (2) symptomatic relief in osteoarthritis of the knee or hip, presented as an NSAID-equivalent alternative; (3) hepatoprotective effect in cholestatic liver disease (including intrahepatic cholestasis of pregnancy) and a debated benefit in alcoholic liver disease. Tertiary claims (fibromyalgia pain, cognitive decline in mild cognitive impairment) exist but rest on much thinner literature.
Evidence by addressing question
Mechanism
SAMe sits at the hinge of one-carbon metabolism. Once its methyl group is transferred (by any of the SAM-dependent methyltransferases — DNMTs, COMT, PEMT, GNMT, and dozens of others), the resulting S-adenosylhomocysteine product is a feedback inhibitor of those same enzymes, so the SAMe / SAH ratio is the actual regulatory variable for cellular methylation capacity (Bottiglieri 2002). In neuropsychiatry, the load-bearing mechanism is methylation of monoamine precursors and membrane phospholipids: catecholamine and indolamine turnover, phosphatidylcholine synthesis (which affects neuronal membrane fluidity and receptor function), and methylation of biopterin cofactors used in serotonin and dopamine biosynthesis. Patients with severe folate or B12 deficiency, and patients homozygous for the MTHFR C677T variant, show reduced CSF SAMe and a depression-prone phenotype that responds to methyl-donor supplementation. In osteoarthritis, the proposed mechanism is methylation-dependent synthesis of polyamines and stimulation of chondrocyte proteoglycan production, plus mild analgesic action that may overlap with COX inhibition at high doses; the mechanism is plausible but less well evidenced than the depression story. In cholestatic liver disease, SAMe is the substrate for the methylation reactions that produce phosphatidylcholine via the PEMT pathway, and it sustains hepatic glutathione synthesis through the transsulfuration arm — both depleted in alcoholic and cholestatic liver injury (Mato et al. 1999).
Evidence — depression
The depression literature is the largest body of SAMe trial data, and the most internally divided. The 1994 meta-analysis (Bressa, Acta Neurol Scand) pooled small parenteral and oral studies and reported an effect size of −0.65 versus placebo and approximate equivalence to tricyclic antidepressants (Bressa 1994). Most of those trials were short (2–6 weeks), small (n < 50), used intramuscular or intravenous SAMe (which bypasses the bioavailability problem), and predate modern reporting standards.
The two later landmark trials are both positive but adjunctive. Papakostas et al. (2010) randomised 73 SSRI/SNRI non-responders to 1600 mg/day oral SAMe versus placebo added to their existing antidepressant; HAMD response was 36.1% on SAMe versus 17.6% on placebo, and remission 25.8% versus 11.7% — both clinically meaningful (Papakostas et al. 2010). Sarris et al. (2018) replicated the adjunctive design (n=107, 8 weeks, 800–1600 mg/day on top of SSRI/SNRI) and reported a smaller, statistically borderline benefit (Sarris et al. 2018).
The 2016 Cochrane review (Galizia et al., 8 trials, n=934) judged the evidence quality low to very low and concluded SAMe was no better than placebo and not significantly different from imipramine or escitalopram in head-to-head trials, though confidence intervals were wide (Galizia et al. 2016). A 2020 monotherapy trial (Sarris et al., 8 weeks, 1600 mg/day, n=49) failed to separate SAMe from placebo on the MADRS (Sarris et al. 2020). The most recent meta-analysis (Cuomo et al. 2024, 23 RCTs, n=2234) re-estimated a placebo-controlled standardised mean difference of −0.58 in favour of SAMe — but noted significant heterogeneity, age of many included studies, and that the antidepressant-comparison effect was non-significant (Cuomo et al. 2024). A clinician-oriented review summarises the same trial set: 8 modern double-blind RCTs, mixed but tilting positive, with the adjunctive design carrying the strongest contemporary evidence and the monotherapy design less so (Mischoulon and Fava 2002).
Dose-response: oral 1600 mg/day is the most-studied dose and the dose used in both the Papakostas and Sarris adjunctive trials. Lower doses (400–800 mg/day) have not been adequately tested in modern trials and probably under-dose. Onset is reported as faster than SSRIs (2 weeks vs 4–6) when the drug works.
Evidence — osteoarthritis
The pooled 2002 meta-analysis (Soeken et al., 11 trials) concluded SAMe was equivalent to NSAIDs for OA pain and function with a more favourable side-effect profile (Soeken et al. 2002). Most of the included trials were older European studies versus ibuprofen, indomethacin, piroxicam, or naproxen, typically at SAMe 1200 mg/day for 4–12 weeks. Najm et al. (2004) randomised 61 adults with knee OA to SAMe 1200 mg/day or celecoxib 200 mg/day in a double-blind crossover; celecoxib was significantly better in the first month, but by month two the two were equivalent — i.e., SAMe is slower in onset but reaches comparable effect (Najm et al. 2004).
The 2009 Cochrane review (Rutjes et al., 4 placebo-controlled trials, n=656) reached a sharply more skeptical conclusion: pooled SAMe-versus-placebo standardised mean difference of −0.17 for pain (≈0.4 cm on a 10 cm VAS — not clinically meaningful) and +0.02 for function. Adverse events were 19% on SAMe vs 15% on placebo, not significantly different. The reviewers explicitly recommended against routine use pending better-designed trials (Rutjes et al. 2009). The disconnect with Soeken's meta-analysis is the comparator: SAMe looks good against NSAIDs (which themselves have only modest effect over placebo in OA) and unimpressive against placebo directly.
Evidence — liver disease
The strongest case is for intrahepatic cholestasis of pregnancy (ICP) and primary biliary cholangitis, where SAMe 800–1600 mg/day reduces pruritus and improves biochemical liver markers in placebo-controlled trials (Frezza et al. 1990). In contemporary practice, ursodeoxycholic acid is first-line for ICP and SAMe is positioned as second-line or adjunct when UDCA is insufficient.
For alcoholic liver disease, the story is more equivocal. Mato et al. (1999) randomised 123 patients with biopsy-confirmed alcoholic cirrhosis to oral SAMe 1200 mg/day or placebo for 2 years; the primary composite of death or liver transplant was not significantly reduced (16% vs 30%, p=0.077), though a Child–Pugh A subgroup analysis suggested benefit (Mato et al. 1999). Medici et al. (2011), a smaller 24-week trial (n=37, 1200 mg/day) in active alcoholic liver disease, found no biochemical or histologic difference between SAMe and placebo at any timepoint (Medici et al. 2011). A subsequent meta-analysis judged the chronic-liver-disease evidence base insufficient to support routine use outside cholestatic indications.
Protocol
Oral dosing for depression: 800–1600 mg/day, divided, taken on an empty stomach (food reduces AUC to ~55%, per Aliani et al. 2020). For osteoarthritis: 1200 mg/day, divided. For liver indications, the trial doses were 1200–1600 mg/day oral or 800 mg/day parenteral. Enteric coating is functionally required; the SAMe salt is hygroscopic and acid-labile, and unprotected formulations have undetectable plasma concentrations. The reference compound (1,4-butanedisulfonate salt with enteric coating) is the formulation used in the Papakostas, Sarris, and Mato trials. The MSI-195 next-generation oral formulation increased bioavailability ~2.8× over the standard commercial product in healthy volunteers, suggesting the typical supplement dose under-delivers SAMe to the systemic circulation by a meaningful margin (Aliani et al. 2020). Onset of effect: for depression typically 1–4 weeks; for OA, 1–2 months (slower than NSAIDs but comparable plateau, per Najm et al. 2004).
Contraindications and interactions
The major safety signal is mania induction in bipolar disorder — SAMe can precipitate hypomanic or manic episodes via the same mechanism as any antidepressant, and reports exist with both parenteral and oral formulations (Mischoulon and Fava 2002). A case report describes manic psychosis on SAMe combined with escitalopram. Serotonin syndrome is a theoretical concern with MAOIs and a documented concern with serotonergic agents; the Papakostas adjunctive trial did not see clinically significant serotonin syndrome with SSRI co-administration but excluded patients on MAOIs. SAMe may interact with levodopa, potentially reducing its efficacy in Parkinson's disease through enhanced peripheral methylation (via COMT, which SAMe substrates). Common adverse effects: GI upset, nausea, headache, insomnia or activation (lower in adjunctive trials, ~10–20% on monotherapy at high dose); the AMSP pharmacovigilance signal is dominated by mood activation rather than physical adverse events (Degner et al. 2004). Hyperhomocysteinaemia is a theoretical concern with prolonged use, particularly in patients with low folate or B12; co-supplementation is conventionally recommended though without firm trial evidence.
Failure modes
Three modal failures. (1) Under-dosing: most over-the-counter products are sold at 200 or 400 mg per tablet; the depression and OA trials used 1200–1600 mg, and lower doses lack RCT support. (2) Taken with food: AUC drops to ~55% of fasting (Aliani et al. 2020). (3) Wrong indication: SAMe is sold as a generic "mood booster" but the evidence is for major depressive disorder at therapeutic doses, not for ordinary low mood. The activation effect (GI, insomnia, anxiety) can be mistaken for the drug "doing something" when it is in fact a side effect; conversely, a true antidepressant response is sometimes dismissed as placebo.
Practicalities
OTC US/UK supplement market price: roughly $0.10–0.20 per 400 mg tablet, or 1600 mg/day ≈ $150–300 per year, putting it cheaper than many supplements at therapeutic dose but well above zero. Shelf stability is poor: SAMe degrades in heat and humidity, so blister-packed products outperform bottled products on label-claim potency at end of shelf life. The salt form matters — tosylate disulfate and 1,4-butanedisulfonate are the stable forms used in trials.
Alternatives
For depression: SSRIs/SNRIs (first-line, larger evidence base, NHS/USPSTF-supported), psychotherapy, exercise, light therapy — each with a stronger evidence base than SAMe. SAMe's niche is augmentation of partial responders to SSRIs, or as a second- or third-line option where conventional antidepressants have failed or been refused. For OA: NSAIDs (the comparator and the established standard), topical NSAIDs (similar efficacy with lower GI risk), duloxetine, intra-articular corticosteroid injections, exercise / weight loss / physical therapy. For cholestatic liver disease: UDCA (first-line for ICP and PBC).
History
SAMe was discovered in 1952 (Cantoni) as the "active methionine" intermediate in methylation reactions. Italian clinical investigation began in the 1970s, leading to its 1979 approval as a prescription drug for depression and cholestasis. The substance entered the US OTC supplement market in 1999 under DSHEA after a Newsweek cover story; the supplement-industry surge predates most of the placebo-controlled evidence.
Out of scope
Forward pointers: methylation cycle nutrition (folate, B12), MTHFR variants, homocysteine testing, omega-3 augmentation of antidepressants. Excluded from this entry: SAMe for fibromyalgia (small mixed trials), Alzheimer's / MCI (early-stage Phase II only), HIV-related neuropathy (one trial, niche).
Credibility range
The optimist case
SAMe is a rare example of a methylation-cycle intermediate with decades of European clinical use as a prescription antidepressant, plausible neurochemical mechanism (catecholamine and phospholipid methylation), and at least two replicated placebo-controlled trials demonstrating benefit as adjunct to SSRIs in non-responders (Papakostas et al. 2010; Sarris et al. 2018). For osteoarthritis the comparison-trial evidence puts it at parity with celecoxib and NSAIDs (Najm et al. 2004; Soeken et al. 2002), with a meaningfully cleaner side-effect profile than chronic NSAID use (no GI bleed risk, no cardiovascular signal). For cholestatic liver disease the mechanism (PEMT pathway, glutathione synthesis) is well-grounded and the symptomatic effect on pruritus is real. The 1994 meta-analysis effect size of −0.65 vs placebo in depression is comparable to modern SSRI trial effects. Onset is faster than SSRIs. Side effects are mild, transient, and well-tolerated at therapeutic dose. Combined picture: a real, mild-to-moderate intervention with several legitimate indications, regulated as a drug in much of Europe, that the US has misclassified as a "supplement."
The skeptic case
The Cochrane reviews are the strongest skeptic exhibit. For depression, Galizia et al. (2016) graded evidence as low to very low quality with no significant placebo-controlled benefit pooled across 8 trials (Galizia et al. 2016). For OA, Rutjes et al. (2009) found a clinically trivial placebo-controlled effect (0.4 cm on a 10 cm VAS) (Rutjes et al. 2009). The 2020 monotherapy RCT was null (Sarris et al. 2020). Bioavailability is genuinely poor: even the enteric-coated reference formulation delivers a fraction of an equivalent intravenous dose, and the next-generation MSI-195 was needed to bring oral AUC to a respectable level (Aliani et al. 2020). Many older "positive" trials used parenteral SAMe, which is not what consumers buy. The Mato alcoholic cirrhosis trial missed its primary endpoint (Mato et al. 1999) and Medici was flat null (Medici et al. 2011). Commercial interests (the supplement industry, the original European patent holder) have driven a publication landscape that skews positive. The activation profile and mania-induction risk impose a real safety floor on otherwise-marginal benefit.
The author's call
SAMe is a real-but-modest agent whose evidence has weakened over time as trial design has tightened. The strongest contemporary use is as adjunctive therapy in SSRI non-responders, where Papakostas's effect is clinically meaningful and Sarris's adjunctive trial replicates it directionally. As monotherapy for depression, the modern trial signal is weak; the older positive trials are dominated by parenteral administration, which is not what an OTC consumer accesses. For osteoarthritis, the honest summary is "comparable to NSAIDs in head-to-head trials, mildly better than placebo, with a slow onset" — a reasonable NSAID-alternative for the gut-intolerant patient, not a first-line choice. For cholestatic liver disease the symptomatic benefit on pruritus is real but UDCA is first-line. Meta scores reflect a modest mood effect, a modest short-term-health effect, and a low-to-moderate evidence rating. The honest framing is not "supplement-industry hype" nor "European miracle drug" but "a real but unspectacular tool with three plausible indications, real interactions, and an under-appreciated dose / formulation / cost burden at therapeutic levels."
Stakeholder and incentive map
- Supplement industry — the US OTC market for SAMe is several hundred million dollars annually, dominated by Nature Made, Doctor's Best, Jarrow. Incentive to publish and amplify positive trial findings; minimal incentive to fund head-to-head trials against modern SSRIs at therapeutic doses.
- European prescription holders — Abbott / Knoll held the original ademetionine patent; the substance is genericised in Europe and still actively prescribed in Italy, Spain, Russia, and Germany. Established clinical custom predisposes prescribers toward the positive interpretation.
- Integrative psychiatry — figures like Mischoulon (MGH), Papakostas (MGH), Sarris (Western Sydney) have built academic programs around nutraceuticals; their trial work is the most rigorous available but the authors are not neutral on whether the substance "works."
- Mainstream psychiatry — generally skeptical, citing Cochrane and the formulation/bioavailability concerns. Sees SAMe as a second-tier augmentation strategy at best.
- NSAID-alternative community — OA patients with GI intolerance and the practitioners who serve them have a real incentive to find a milder analgesic; SAMe and glucosamine occupy this niche.
Population variability
Folate / B12 status matters: deficient patients have lower endogenous SAMe and may respond more robustly to supplementation. MTHFR C677T homozygotes (≈10% of European-ancestry populations) have a documented depression-prone phenotype and may benefit disproportionately, though no SAMe trial has prospectively stratified on genotype. Bipolar-spectrum patients are at elevated risk of switch; the integrative-psychiatry literature recommends mood-stabiliser co-administration and starting at 100–200 mg/day. Patients on MAOIs are excluded from essentially all SAMe trials and should not combine. Parkinson's patients on levodopa may experience reduced therapeutic effect via enhanced peripheral O-methylation. Older patients (≥65) have not been preferentially studied; the OA trials skewed middle-aged. SAMe has been used in pregnancy specifically for intrahepatic cholestasis with safety data spanning multiple trials and decades; for non-cholestatic indications, pregnancy data are thin and the conventional caution applies.
Knowledge gaps
(1) A modern, adequately powered (n > 300), placebo-controlled monotherapy depression trial at 1600 mg/day for 12 weeks with an enteric-coated reference formulation does not exist; Sarris 2020 at n=49 underpowers any interpretation. (2) The MSI-195 / next-generation high-bioavailability formulations have not been tested in efficacy trials, so it is unknown whether the 2.8× AUC translates to a better clinical effect. (3) No head-to-head OA trial against SSRI augmentation (e.g., duloxetine), which would be the relevant modern comparator. (4) No prospective MTHFR-genotype-stratified trial. (5) The mania-switch absolute risk in unipolar depressives without bipolar family history is unquantified — the signal is real in bipolar patients but the base rate in screened unipolar populations may be very low. (6) Long-term (≥1 year) homocysteine and cardiovascular outcomes on chronic SAMe at therapeutic dose are not characterised.
Scope and the brief. The topic brief named mood, joint health, and liver function. The article covers all three honestly: mood gets the lead (strongest contemporary evidence, adjunctive design), joints get a parity-with-NSAIDs treatment with the Cochrane caveat exposed, and liver function is scoped specifically to the cholestatic indication where the evidence is strongest, with the more equivocal alcoholic-liver-disease story handled in a single honest sentence rather than padded out.
Action choice (decide, not do). The mood indication's strongest evidence is adjunctive, which structurally requires a prescriber already in the loop. The bipolar-switch and MAOI/levodopa interaction signals further push it out of the "self-administered supplement" bucket. decide reflects the honest reading.
Evidence rating. Rated 2 with deliberation. Multiple RCTs exist but the 2009 (osteoarthritis, Rutjes) and 2016 (depression, Galizia) Cochrane reviews independently graded the literature unfavourably, and the most recent monotherapy depression RCT (Sarris 2020) was null. The positive 2024 Cuomo meta-analysis is real but rests on heterogeneous and largely older trials. 3 would overclaim; 1 would ignore the Papakostas adjunctive replication.
Controversy rating. Rated 2 rather than 3. The disagreement between integrative psychiatry and the Cochrane camp is real but methodological, not ideological — closer to evidence-quality dispute than paradigm fight.
No dream narrative. Computed overall score is approximately 15, well below the obligatory threshold of 40. The entry's honest hook is clarity — accurate positioning of a modest tool — rather than aspiration or relief. Forcing a transformed-life projection would ring false here.
Excluded scope and why.
- Fibromyalgia. Small mixed trial set; effect size unclear; folded into "weaker on most other claims" in the dek rather than given a dedicated section.
- Mild cognitive impairment / Alzheimer's. Active Phase II only, no usable signal yet; would warrant its own entry if a clear trial result lands.
- Children and adolescents. No paediatric trial data; not a population the article addresses.
- Hep-C cirrhosis AFP-lowering. Specific Phase II indication, not on-axis for a general-audience entry.
Separate-entry candidates surfaced by the write. MTHFR variants and the broader methylation-cofactor story (folate / B12 / B6) keep coming up — likely warrants its own entry in supplements or screening. Ursodeoxycholic acid for cholestatic conditions is large enough to deserve its own entry. Adjunctive strategies in SSRI partial-responders (omega-3, lithium, T3, SAM-e) could become a single comparator entry in mental.
Future-link candidates. Once written: MTHFR variant and methylation status, folate / B12 sufficiency, topical NSAIDs for joint pain, ursodeoxycholic acid, omega-3 EPA/DHA augmentation, SSRI plateau strategies. The related field on meta is left empty pending those entries.
Tone calibration. The entry sits close to the wellness-influencer edge precisely because the supplement market overpromises. Voice was deliberately positioned as a clear-eyed friend who would tell you when a supplement was working, when it wasn't, and why most users don't get the dose right.
SAM-e (S-Adenosylmethionine)
Daily pill, divided dose, on an empty stomach to preserve bioavailability (AUC drops to ~55% with food per Aliani 2020). No injections, no monitoring required at standard use. Trivial behavioral burden once routine.
Therapeutic doses are 1200–1600 mg/day; OTC pricing on enteric-coated 400 mg tablets runs $0.10–0.20/tablet, so $150–300/year at depression or OA dose. Sits in the minor-cost band, higher than most generic supplements but well below brand-name prescription antidepressants without insurance.
Two replicated placebo-controlled adjunctive trials show response rates roughly double placebo when added to SSRIs/SNRIs in non-responders (Papakostas 36.1% vs 17.6%; Sarris directionally similar). Monotherapy effect is weaker — the 2016 Cochrane review found no significant separation, but the older Bressa meta-analysis (mostly parenteral) reported effect size −0.65 vs placebo, and the 2024 Cuomo meta-analysis estimated −0.58. Meaningful but not dominant — a clear named effect.
Najm 2004 (n=61 cross-over) showed SAMe 1200 mg/day reached celecoxib-equivalent pain relief in knee OA by month two; Soeken 2002 meta-analysis pooled NSAID-comparison trials at parity. The 2009 Rutjes Cochrane review found a clinically trivial placebo-controlled effect (~0.4 cm on 10 cm VAS) — so the headline is honest as 'real but small, slow onset.' Pruritus relief in cholestatic liver disease adds a second short-term symptomatic effect (Frezza 1990).
Multiple RCTs across three indications, but the contemporary high-quality reviews are unfriendly: 2016 Cochrane (depression, Galizia) graded evidence low to very low; 2009 Cochrane (OA, Rutjes) found placebo-controlled effect non-significant. Two positive adjunctive depression RCTs (Papakostas 2010, Sarris 2018) and a recent positive meta-analysis (Cuomo 2024) keep the case alive but not strong. The reference enteric-coated formulation's bioavailability is genuinely poor (Aliani 2020), which weighs against translating older parenteral-trial effects to modern oral practice.