← Back
NeuroTrials.ai
Neurology Clinical Trial Database

SAMe-AD

A phase 2, randomized, multicenter, double-blind, placebo-controlled trial of S-adenosyl methionine in participants with mild cognitive impairment or dementia due to Alzheimer's disease

Share Share Copied! Compare

Year of Publication: 2026

Authors: Holper S, Barnham KJ, Churilov L, ..., Yassi N

Journal: Alzheimer's & Dementia

Citation: Holper S, et al. Alzheimer's Dement. 2026;22:e71381.

Link: https://doi.org/10.1002/alz.71381

Clinical Question

Does oral S-adenosyl methionine (SAMe) 400 mg daily for 6 months reduce plasma p-tau217 in patients with MCI or dementia due to Alzheimer's disease?

Bottom Line

SAMe 400 mg/day for 180 days did not reduce plasma p-tau217 or any other AD biomarker, and did not improve cognition compared to placebo, although it was safe and well tolerated. SAMe monotherapy at this dose and duration is unlikely to be disease-modifying in AD.

Major Points

  • First randomized, double-blind, placebo-controlled trial of single-agent SAMe in Alzheimer's disease.
  • No reduction in plasma p-tau217 with SAMe vs placebo (mean % change +53.22 vs +25.34; standardized mean difference 37.58, 95% CI -32.61 to 107.76; p=0.288).
  • No significant differences in other plasma biomarkers (p-tau181, GFAP, NfL) or in cognitive endpoints (RBANS, MMSE, MoCA, CVLT, DSST, CDR).
  • SAMe 400 mg daily was safe and well tolerated over 6 months.
  • Authors conclude SAMe monotherapy at this dose and duration is unlikely to be disease-modifying; alternative dosing, longer treatment, or more stable formulations may be worth exploring.

Design

Study Type: Phase 2 randomized, double-blind, placebo-controlled trial

Randomization: 1

Blinding: Double-blind (participants, investigators, outcome assessors); blinded until database lock

Allocation: 1:1, central randomization via independent statistician, stratified by age at enrollment (<70 vs ≥70 years), schedule housed in REDCap eCRF

Follow-up Duration: 180 days of treatment plus follow-up telephone call 14 days post-endpoint visit

Centers: 4

Countries: Australia

Sample Size: 63

Analyzed: 57

Analysis: Principal stratum approach within the estimand framework: primary efficacy analysis included participants with ≥80% study drug adherence and available 180-day primary outcome sample; ANCOVA with treatment arm as independent variable and baseline p-tau217 as covariate; G-computation for standardized mean differences with 95% CIs

Power Calculation: Total target n=60 (30/arm); n=52 (26/arm, allowing for 4 per-arm attrition) estimated to yield 80% power to detect a large effect (Cohen d ≥0.8) at two-tailed alpha 0.05; no prior literature available for effect size estimation

Registration: ACTRN12620000506998 (Australian New Zealand Clinical Trials Registry)

Inclusion Criteria

  • Community-dwelling adults aged ≥60 years
  • MCI or dementia due to Alzheimer's disease per NIA-AA 2011 criteria
  • MMSE score ≥18
  • Reliable study partner with regular (≥3x weekly) contact to oversee study drug administration and assist with assessments (e.g., CDR)
  • If on AD-symptomatic medications (cholinesterase inhibitors or memantine), dose stable ≥8 weeks prior to screening

Exclusion Criteria

  • History of bipolar affective disorder
  • Concurrent use of antidepressant medication
  • Medical condition other than AD that may contribute to cognitive impairment
  • Stroke or transient ischemic attack in the past 12 months
  • Clinically significant unstable psychiatric condition in the past 6 months
  • Inability to swallow oral medications
  • Other medical conditions limiting life expectancy to <6 months
  • Participation in another interventional clinical trial
  • Intake of another investigational drug within 4 weeks or 5 half-lives (whichever longer)

Baseline Characteristics

CharacteristicSAMe (n=31)Placebo (n=26)
Age, years (median, IQR)76 (70, 81)77 (72, 80)
Female sex (n, %)16 (51.61%)13 (50%)
Years of education (median, IQR)15 (11, 15)15 (12, 16)
≥1 cardiovascular disease or risk factor (n, %)23 (74.19%)19 (73.08%)
AD dementia (n, %)16 (51.61%)14 (53.85%)
MCI (n, %)15 (48.39%)12 (46.15%)
Use of AD medications (n, %)14 (45.16%)11 (42.31%)
MMSE (median, IQR)24 (23, 25)24.5 (22, 28)
MoCA (median, IQR)19 (17, 22)20 (18, 22)
RBANS (mean, SD)65.68 (12.17)72.46 (15.38)
GDS-15 (median, IQR)2 (1, 4)2 (1, 3)
CDR global (median, IQR)0.5 (0.5, 1)0.5 (0.5, 1)
Baseline p-tau217, pg/mL (mean, SD)0.79 (0.54)0.69 (0.32)

Arms

FieldSAMeControl
N3126
InterventionOral S-adenosyl methionine 400 mg once daily (initially over-encapsulated 400 mg commercial product; later switched mid-trial without unblinding to 2 x 200 mg soft-gel capsules of liquid-source SAMe), adjunct to AD standard of careMatched placebo capsule once daily (later 2 capsules daily under second formulation), adjunct to AD standard of care
Duration180 days180 days

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Percentage change in plasma p-tau217 concentration from baseline to day 180PrimaryMean % change +25.34 (SD 94.83)Mean % change +53.22 (SD 159.19)Standardized mean difference 37.580.288
Change in total scaled RBANS score (cognition)SecondaryNo significant difference between SAMe and placebo
Percentage change in plasma p-tau181SecondaryNo significant difference between SAMe and placebo
Percentage change in plasma GFAPSecondaryNo significant difference between SAMe and placebo
Percentage change in plasma neurofilament light chain (NfL)SecondaryNo significant difference between SAMe and placebo
Other cognitive measures (MMSE, MoCA, CVLT-3, Digit Symbol Substitution Test) and CDR functional scaleSecondaryNo significant differences reported between treatment arms
Proportion of participants with one or more serious adverse events (primary safety estimand; population = all who received ≥1 dose)SafetySAMe 400 mg daily for 180 days was safe and well tolerated; no significant safety differences vs placebo

Subgroup Analysis

Prespecified subgroup analyses (including APOE ε4 carrier status and homocysteine level [elevated defined as ≥12 µmol/L]) were performed to further characterize treatment effect; no signal of efficacy reported in source excerpt.

Criticisms

  • Small sample size (n=57 analyzed) powered only to detect a large effect (Cohen d ≥0.8); modest disease-modifying effects could be missed.
  • Short 6-month treatment duration may be insufficient to detect biomarker or cognitive changes in a slowly progressive disease.
  • Mid-trial switch in investigational product formulation (over-encapsulated commercial product → liquid-source soft-gel capsules) may have introduced bioavailability variability.
  • Single-country (Australia), 4-site recruitment limits generalizability.
  • Wide confidence interval around primary effect estimate (-32.61 to 107.76) reflects substantial imprecision.

Funding

National Health and Medical Research Council; Alzheimer's Association

Based on: SAMe-AD (Alzheimer's & Dementia, 2026)

Authors: Holper S, Barnham KJ, Churilov L, ..., Yassi N

Citation: Holper S, et al. Alzheimer's Dement. 2026;22:e71381.

Content summarized and formatted by NeuroTrials.ai.