Methylene Blue (Methylthioninium Chloride)

Identity, and which methylene blue

PubChem gives methylene blue CID 6099, CAS 61-73-4 and molecular formula C16H18ClN3S, and reports a molecular weight of 319.9. The 319.85 figure quoted for the anhydrous chloride is the FDA Global Substance Registration System structure value, 319.8538, rounded, and the approved product's label states 319.86 for the anhydrous form. Three cited sources give three figures and all three are reported here. GSRS holds two separate substance records. UNII 8NAP7826UB is typed as methylene blue anhydrous and carries CAS 61-73-4. UNII T42P99266K is typed simply as methylene blue and carries CAS 7220-79-3 as its primary registry number, the trihydrate. The approved injection's label states that its strength is expressed in terms of trihydrate and writes the formula with a variable water of hydration. Two CAS numbers circulate for this compound and both are correct; they describe different hydration states of the same salt.

The ChEMBL identifier disagrees as well. GSRS lists CHEMBL191083 as the primary ChEMBL code on both UNII records, but UniChem resolves that identifier to InChI=1S/C16H18N3S with a charge of +1, which is the methylene blue cation, formula C16H18N3S, PubChem CID 4139, molecular weight 284.4. The chloride salt this page defines maps instead to CHEMBL405110, which UniChem links to PubChem CID 6099 and DrugBank DB09241, and the trihydrate maps to CHEMBL550495, which links to CID 104827. ChEMBL's own interface returned server errors throughout the session in which this was checked, so the resolution here runs through UniChem and PubChem rather than ChEMBL directly.

The same molecule is registered as a textile and biological stain under Basic Blue 9, Solvent Blue 8 and C.I. 52015, and under several dozen historical trade names. Retail material distinguishes grades — USP, laboratory, technical — and those designations refer to purity specifications rather than to different compounds. What no primary source establishes is the size of the difference. A PubMed search for comparative impurity analyses of methylene blue preparations returns papers about the dye as an adsorption target in water-treatment chemistry, not analyses of material sold for ingestion. That gap is recorded in the unsourced array below rather than in the ledger, which carries only rows with a source.

The one approved indication, and the boxed warning

The approved intravenous product, a 5 mg/mL sterile solution, was cleared under NDA 204630 on 8 April 2016 with an orphan designation and a Type 5 classification covering a new formulation or new manufacturer. Its indication is singular: treatment of paediatric and adult patients with acquired methaemoglobinaemia. The compound had been in clinical use for roughly a century before that under grandfathered status, which is why a 2016 approval date attaches to a molecule that predates the modern regulatory system. No oral methylene blue product carries FDA approval for anything. The European, British, Canadian, Australian and Japanese registers were not searched for this page, so nothing here establishes what is or is not approved outside the United States.

That label carries a boxed warning. It states the product may cause serious or fatal serotonergic syndrome when combined with serotonergic drugs and opioids, and directs avoidance of concomitant SSRIs, SNRIs, MAOIs and opioids. The pharmacology behind it was characterised in vitro. Ramsay and colleagues reported in 2007 that methylene blue is a potent tight-binding inhibitor of purified human monoamine oxidase A, inhibiting MAO B only at much higher concentrations. Delport and colleagues later measured an IC50 of 0.07 microM against MAO A while designing analogues that lose the property. Pharmacologically, this is a monoamine oxidase inhibitor.

Two further label positions belong on the record. The product is contraindicated in glucose-6-phosphate dehydrogenase deficiency because of the risk of haemolytic anaemia, and prescribers are directed to use methods other than pulse oximetry to assess oxygen saturation, since the dye interferes with the reading. Section 2.1 instructs that alternative interventions be considered where methaemoglobinaemia does not resolve after two doses. The label's Overdosage section separately records mild methaemoglobinaemia, up to 7%, among the effects reported after cumulative intravenous doses of 7 mg/kg or more of a methylene blue class product, alongside nausea, vomiting, chest tightness, tremor and electrocardiogram changes lasting 2 to 12 hours.

Sections 8.1 and 8.2 of the same label carry reproductive data. Intra-amniotic injection of a methylene blue class product during the second trimester was associated with neonatal intestinal atresia and fetal death. Methylene blue was administered orally to pregnant rats at 50 to 350 mg/kg per day during organogenesis, with maternal and embryofetal toxicity at all doses and most evident at 200 and 350 mg/kg per day; the label gives 200 mg/kg as roughly 32 times a clinical dose of 1 mg/kg on a body-surface-area basis. In pregnant rabbits at 50, 100 and 150 mg/kg per day, maternal death was observed at 100 mg/kg and spontaneous abortion at all dose levels. The label states there is no information on the presence of methylene blue in human milk and directs that breast-feeding be discontinued for up to eight days after treatment.

What the mitochondrial claim actually rests on

Wen and colleagues established the mechanism that gets invoked. Working in cultured neuronal cells in 2011, they demonstrated that methylene blue accepts electrons from NADH and transfers them to cytochrome c, bypassing a blockade at complex I and III, and that a derivative with its redox centre disabled by N-acetylation had no effect on mitochondrial complex activities. Protection was observed at low nanomolar concentrations. The same paper reported attenuation of rotenone-induced dopaminergic loss in a rodent Parkinson model and reduced damage after transient focal cerebral ischaemia. The mechanism is established in cells, at nanomolar concentrations.

Atamna and colleagues supplied the numbers that travel furthest. In human IMR90 fibroblasts, methylene blue extended replicative life span by more than twenty population doublings, increased mitochondrial complex IV by 30%, and enhanced cellular oxygen consumption by 37 to 70%. Every one of those figures comes from tissue culture at nanomolar concentrations. None is a measurement of ATP, none is from an aged animal, and none is from a person. The 37-to-70% figure now circulates as an ATP increase in aged cells, which is a different quantity measured in a different system.

Rodent behaviour has been measured directly. Callaway and colleagues reported in 2004 that rats given 1 mg/kg intraperitoneally after training in a baited holeboard maze made 66% of their probe-trial visits to training-baited holes, against 31% for saline controls, and that brain cytochrome c oxidation was raised 24 hours after injection but not at one or two hours. Group sizes are not stated in the abstract. The finding is a rat finding at 1 mg/kg; the human imaging trials described below administered roughly four times that per kilogram.

What was measured in human brains

Rodriguez and colleagues ran the trial that most cognitive claims trace back to. Twenty-six healthy adults aged 22 to 62 were randomised, double-blind, to a single oral dose or placebo and scanned before and one hour after administration. Reported in Radiology in 2016, the trial found increased response in bilateral insular cortex during a psychomotor vigilance task and in prefrontal, parietal and occipital cortex during a short-term memory task, alongside a 7% increase in correct responses during memory retrieval (P = .01). The dose, per the paper's methods, was 280 mg of USP-grade methylene blue in immediate-release capsules.

Secondary material describes single-digit milligram amounts as the low, mitochondrial-range dose. The trial behind the memory number used 280 mg, roughly 4 mg/kg. A companion analysis from the same group in Brain Imaging and Behavior in 2017 reported that the same administration was associated with a reduction in cerebral blood flow in a task-related network during a visuomotor task, together with stronger resting-state functional connectivity in multiple regions linking perception and memory functions. Cerebral blood flow was reported as reduced, not increased.

Russo and colleagues then measured cerebral oxygen metabolism directly. Fifteen euthymic patients with bipolar I disorder and fifteen age- and gender-matched controls each underwent two MRI sessions after intravenous methylene blue at 0.5 mg/kg or placebo, single-blind and randomised crossover. Global cerebral blood flow and the cerebral metabolic rate of oxygen both decreased after methylene blue across all participants. Regional reductions were larger in patients than controls in the frontal lobe (−14.2 ± 19.5% against 2.3 ± 14.8%) and anterior cingulate (−14.8 ± 23.7% against 2.4 ± 15.7%).

One larger study exists and has not been published. NCT02380573, sponsored by a Texas academic health science centre, enrolled 117 participants across healthy middle age, healthy ageing, mild cognitive impairment and mild Alzheimer's disease, administering 282 mg orally daily against an FD&C Blue No. 2 comparator, and completed in July 2023. Summary results were submitted to ClinicalTrials.gov on 25 July 2024 and posted on 19 September 2024. A PubMed search on that registration number returns no indexed publication as of August 2026, so the largest cognitive dataset has not been peer-reviewed.

Two further randomised human trials used a memory endpoint rather than an imaging one. Telch and colleagues gave 260 mg orally immediately after six five-minute extinction trials to adults with marked claustrophobic fear, 23 assigned to methylene blue and 19 to placebo, and reported that participants with low fear after training showed significantly less fear at one-month follow-up while participants with moderate to high post-training fear tended to fare worse; enhancement of incidental contextual memory was unrelated to fear level. Zoellner and colleagues used the same 260 mg dose after five daily imaginal exposure sessions in 42 patients with chronic post-traumatic stress disorder, randomised against placebo and a waitlist, and reported gains that did not differ from standard prolonged exposure at three-month follow-up.

The Alzheimer's programme, and which molecule each trial gave

The programme began with methylene blue itself. Wischik and colleagues reported an exploratory double-blind, randomised, placebo-controlled dose-finding trial of methylthioninium at 69, 138 and 228 mg per day in 321 subjects with mild or moderate Alzheimer's disease, registered as NCT00515333, with change on the ADAS-cog at 24 weeks as the primary outcome and a regional cerebral blood flow sub-study in 135 subjects. At 24 weeks the reported benefits were at 138 mg per day: a treatment effect of −5.42 ADAS-cog units in moderate subjects (corrected p = 0.047), and 1.97% on regional cerebral blood flow in mild subjects (corrected p < 0.001). Delivery of the highest dose was impaired by dose-dependent dissolution and absorption limitations. Baddeley and colleagues, describing the programme's chemistry, name that formulation as a chloride salt, methylthioninium chloride, methylene blue.

The two phase 3 trials that followed administered something else. Both gave leuco-methylthioninium bis(hydromethanesulfonate) — LMTM, later renamed hydromethylthionine mesylate — a stable reduced form of the methylthioninium moiety formulated as a different salt. Gauthier and colleagues reported the first in The Lancet in 2016: 891 participants with mild to moderate Alzheimer's disease at 115 centres across 16 countries, randomised 3:3:4 to 75 mg twice daily, 125 mg twice daily, or a 4 mg twice-daily control. Prespecified primary analyses showed no benefit on either co-primary outcome.

Wilcock and colleagues published the second in 2018. Eight hundred participants with mild Alzheimer's disease were randomised to 100 mg or 4 mg twice daily; before unblinding, the statistical analysis plan was revised to compare non-randomised monotherapy subgroups of 79 and 76 against 396 and 297, and it is those revised comparisons that reached significance. The paper's own conclusion states that a further suitably randomised trial is required to test the hypothesis. LUCIDITY, reported by Wischik and colleagues in 2026 in 598 amyloid-PET-positive participants, could not demonstrate significant differences on its co-primary clinical endpoints at 52 weeks, though several secondary and delayed-start outcomes did separate and are logged in the ledger.

The 4 mg control arm in LUCIDITY was methylthioninium chloride, methylene blue, given twice weekly and intended as an inert urinary colourant to preserve the blind; the investigators concluded it had therapeutic activity of its own. A population pharmacokinetic analysis in the British Journal of Clinical Pharmacology in 2026, built on 7,784 plasma measurements from 710 participants, described U-shaped clearance over 24 months producing plasma concentrations at 12 months three times (mesylate) and five times (chloride) above linear-model predictions, and concluded that comparable urinary discolouration cannot be maintained without therapeutic activity. Schelter and colleagues then compared the 16 mg/day arm against propensity-score-matched placebo controls from the Critical Path for Alzheimer's Disease database, matched natural-history controls from ADNI, and a meta-analysis of placebo arms, rather than against a randomised placebo group. Authorship across these is substantially the sponsor's.

The rest of the human record

Dicko and colleagues ran a phase 2, single-blind randomised controlled trial in Bamako, Mali, enrolling 80 G6PD-normal male participants aged 5 to 50 with asymptomatic Plasmodium falciparum gametocytaemia. Twenty received dihydroartemisinin-piperaquine plus 15 mg/kg per day of oral methylene blue for three days. Among those infectious at baseline, that arm showed a median 100% within-person reduction in mosquito infectivity at day 2 by membrane feeding, against −6.0% for dihydroartemisinin-piperaquine alone (p<0.0001). Eleven participants in the arm were infectious at baseline and contributed to the comparison. Haemoglobin changes were similar to controls.

Psychiatry has two small trials thirty years apart, and both share a design problem. Naylor and colleagues ran a two-year double-blind crossover in 31 bipolar patients maintained on lithium, comparing 300 mg per day against 15 mg per day; seventeen completed, and depression ratings were lower in the higher-dose year. Alda and colleagues repeated the structure in 2017 with 37 participants on lamotrigine, comparing 195 mg against 15 mg, reporting improvement on the Montgomery-Åsberg and Hamilton depression scales (P = 0.02 and 0.05) and the Hamilton anxiety scale (P = 0.02), with no significant cognitive effect. In each, the comparator was a low dose of the same compound rather than an inert placebo.

Levin and colleagues randomised 56 cardiac surgery patients with vasoplegic syndrome to 1.5 mg/kg intravenously or placebo and reported no deaths in the treated arm against six of 28 in controls. The 56 were identified by five criteria from 638 consecutive cases across five collaborating institutions in Buenos Aires, and PubMed indexes the paper as a multicentre study. It is a 2004 trial with 28 patients per arm and an unusually large effect. ClinicalTrials.gov returned 339 studies naming methylene blue as an intervention, and 400 by free-text term, when its API was queried on 18 August 2026; the conditions those studies most often list are septic shock, sentinel lymph node mapping, periodontitis and cancer surgery, which are dye and photodynamic uses rather than metabolic ones.

Two randomised trials in elderly surgical patients test a higher intravenous dose than any imaging study. Deng and colleagues randomised 248 patients undergoing major non-cardiac surgery, open-label, to 2 mg/kg intravenously after anaesthetic induction or saline, and reported postoperative delirium in 7.3% against 24.2% (OR 0.24, 95% CI 0.11 to 0.53, p < 0.001) and early postoperative cognitive dysfunction at day 7 in 16.1% against 40.2% (OR 0.30, 95% CI 0.16 to 0.57, p < 0.001). Zhang and colleagues randomised 217 elderly joint replacement patients to the same 2 mg/kg and reported delirium in 8.3% against 20.4% (hazard ratio 0.39, 95% CI 0.173 to 0.858, p = 0.024), with postoperative fever more common in the treated group, 16.5% against 7.4% (p = 0.039).

Pharmacokinetics that do not agree

Two published human studies give incompatible half-lives. Peter and colleagues measured whole-blood concentrations by HPLC in seven volunteers after 100 mg intravenously and 100 mg orally, estimating a terminal half-life of 5.25 hours; oral area under the curve was 9 nmol/min/mL against 137 for the intravenous route, a difference they attributed to organ distribution. Repici and colleagues, studying a delayed-release oral tablet in 22 healthy volunteers after bowel preparation, reported a median time to peak of 16 hours and half-lives of 14 to 27 hours after 200 mg and 6 to 26 hours after 400 mg. Formulation differs between the two, and so does route.

The later study also reported an absolute oral bioavailability of 139.19 ± 52.00%, calculated as the dose-corrected ratio of oral to intravenous exposure. A bioavailability above 100% is not physically meaningful for an unchanged drug and points to an assay or modelling artefact — plausibly the redox interconversion between methylene blue and its leuco form, which any assay has to resolve. Peak blood concentration did not rise proportionally with dose. The single half-life figure quoted on secondary pages is the intravenous one, applied without the route it belongs to.

Excretion is partial. Peter and colleagues recovered 18% of an oral dose and 28% of an intravenous dose in urine as methylene blue plus its leuco form. Urinary discolouration was reported after both routes in the trials logged above. Dicko and colleagues excluded blue urine from their adverse-event comparison for that reason; the LUCIDITY investigators used a low-dose methylthioninium chloride arm specifically to match it; and the San Antonio cognitive trial used a dye and a urinary analgesic to mask it.

What is not known

No trial has tested a sustained single-digit milligram oral dose of methylene blue against a cognitive endpoint in healthy adults. The published oral human record does reach single-digit milligrams: the LUCIDITY control arm gave methylthioninium chloride 4 mg twice weekly to part of 598 participants. At the other end it runs to 15 mg/kg per day for three days in the Bamako malaria trial, which in a 50 kg participant is roughly 750 mg a day. Intravenous exposures in the trials logged here run from 0.5 mg/kg to 2 mg/kg. What is missing is not a low dose but a low daily oral dose measured for cognition. No oral formulation holds FDA approval for any indication, and no non-US register was searched for this page, so nothing here establishes what is approved elsewhere. Long-term oral exposure is characterised mainly within the tau-aggregation programme, whose two phase 3 trials administered a different salt in a different oxidation state; only the phase 2 and the LUCIDITY control arm speak directly to methylthioninium chloride. Pharmacokinetics remain unsettled: two human studies give half-lives differing by a factor of four, and one reports an absolute bioavailability above 100%, which cannot be literally true of an unchanged drug. The largest cognitive dataset, 117 participants completed in July 2023, posted summary results to a registry in September 2024 but has not been published or peer-reviewed. No study has established a dose below which the MAO-A inhibition that drives the boxed warning ceases to be relevant. Sex differences are unexamined in humans, and the one rodent lifespan result was female-only and did not touch median survival. The reproductive data on the label are animal data plus case reports of intra-amniotic exposure; no controlled human pregnancy or lactation study exists, the label states there is no information on methylene blue in human milk, and the approved product is contraindicated in G6PD deficiency.

Questions