Ibogaine for Alzheimer's: What the Evidence Shows

Dementia is an umbrella term for acquired, progressive loss of memory, reasoning, language, and independence. Alzheimer's disease (AD) is its most common cause; others include vascular dementia, dementia with Lewy bodies, frontotemporal dementia, and mixed pathology.

AD is defined biologically by extracellular amyloid-beta plaques, intracellular tau tangles, synaptic loss, and chronic neuroinflammation. These changes begin years — often decades — before symptoms appear, and they are largely concentrated in the hippocampus and association cortex.

The Alzheimer's Association's 2026 Alzheimer's Disease Facts and Figures report estimates:

• 7.4 million Americans aged 65 and older are living with Alzheimer's dementia
• About 1 in 9 people aged 65+ (11%) has the disease
• 74% are aged 75 or older, and almost two-thirds are women
• Numbers may reach roughly 13.8 million without a preventive breakthrough
• Deaths attributed to Alzheimer's rose 134% between 2000 and 2024
• Nearly 13 million Americans provide unpaid dementia care, contributing more than 19 billion hours in 2025

Two features of this population matter for everything that follows. First, it is old: the median patient carries accumulated cardiovascular disease, often including atrial fibrillation, heart failure, or conduction abnormalities. Second, decision-making capacity is impaired by definition — the illness itself erodes the ability to weigh a risky choice. Both facts change the risk calculation for any experimental intervention.

No approved therapy cures Alzheimer's or restores lost function. Approved options fall into two categories, and both have real but bounded benefit.

Cholinesterase inhibitors (donepezil, rivastigmine, galantamine) raise synaptic acetylcholine and treat symptoms rather than pathology. The Cochrane review by Birks and Harvey (2018, CD001190) pooled 30 studies and 8,257 participants. At 26 weeks, donepezil 10 mg/day outperformed placebo on the ADAS-Cog by a mean difference of −2.67 points (95% CI −3.31 to −2.02) on a 70-point scale, and on the MMSE by 1.05 points. There was no significant benefit for behavioural symptoms or quality of life, and more participants on donepezil withdrew (24% vs 20%) or reported an adverse event (72% vs 65%).

Memantine, an NMDA-receptor antagonist, is used in moderate-to-severe disease with similarly modest effect sizes.

Anti-amyloid monoclonal antibodies are the first agents to alter the underlying pathology. In CLARITY-AD (van Dyck et al., New England Journal of Medicine, 2023), lecanemab slowed decline on the CDR-Sum of Boxes by 27% over 18 months — an absolute difference of 0.45 points — while ARIA-E (amyloid-related imaging abnormalities with oedema) occurred in 12.5% of treated participants versus 1.7% on placebo. Donanemab produced comparable slowing with higher ARIA rates.

These drugs slow progression. None reverses it. Alongside them, evidence-based care includes management of blood pressure, hearing loss, diabetes, and physical inactivity, plus structured cognitive and social engagement and caregiver support.

The scientific case for looking at ibogaine in neurodegeneration is mechanistic and indirect. It rests on one finding.

He and colleagues (Journal of Neuroscience, 2005) showed that systemic ibogaine increased expression of glial cell line-derived neurotrophic factor (GDNF) in a rat midbrain region including the ventral tegmental area. In dopaminergic-like SH-SY5Y cells, ibogaine phosphorylated the GDNF receptor Ret and downstream ERK1. Critically, the behavioural effect on alcohol drinking was abolished by anti-GDNF neutralising antibodies — establishing GDNF as the mediator in that model.

GDNF and BDNF are growth factors that support neuronal survival and synaptic maintenance. Because Alzheimer's involves progressive synapse and neuron loss, anything that reliably raises neurotrophic signalling is an obvious hypothesis to test.

Ibogaine also has secondary properties cited in this discussion: NMDA-receptor antagonism (the same broad class as memantine), sigma-2 receptor binding, nicotinic acetylcholine receptor block, and serotonin transporter activity.

The chain of reasoning needed to get from these observations to a treatment for Alzheimer's has at least three unproven links:

• Ibogaine raises GDNF in the rodent midbrain dopamine system — not in human hippocampus or cortex, where Alzheimer's does its damage
• It is unknown whether ibogaine raises GDNF in humans at all
• Raising GDNF has already been tested directly in another neurodegenerative disease and failed: Whone et al. (Brain, 2019) infused GDNF straight into the putamen of 41 people with Parkinson's disease and missed the primary endpoint (P = 0.41) despite large changes on PET imaging

A plausible mechanism is a reason to run a study. It is not a result.

The honest summary is short: no controlled trial, no open-label study, no case series, and no published case report has tested ibogaine in Alzheimer's disease or any dementia. The evidence level is not weak. It is absent.

A search of ClinicalTrials.gov in September 2026 returned nine registered ibogaine studies. Their indications are opioid use disorder, opioid withdrawal, methadone detoxification, alcoholism, PTSD and traumatic brain injury, and personality research. None names Alzheimer's disease, dementia, or mild cognitive impairment.

A PubMed search combining ibogaine with Alzheimer's or dementia returns five records, none of which is a study of ibogaine in dementia. The closest is Currais et al. (Journal of Ethnopharmacology, 2014), who screened West African medicinal plants in cell-based assays modelling Alzheimer's-relevant neurotoxicity. The protective compound they isolated was voacamine from Voacanga africana — a different iboga-type alkaloid, active in cell culture at EC50 values ≤3.4 µM. That is a petri-dish finding about a different molecule.

No published preclinical study has given ibogaine to an amyloid or tau transgenic animal.

The nearest human data come from a different condition. Cherian et al. (Nature Medicine, 2024) reported the MISTIC protocol — magnesium plus ibogaine — in 30 male Special Operations veterans with predominantly mild traumatic brain injury. This open-label, uncontrolled study found improvement in disability (Cohen's d = 2.20 at one month), PTSD, depression, and anxiety, alongside neuropsychological gains in processing speed (d = 0.97–1.34), executive function, visual and verbal memory, and sustained attention, with no decline in any domain.

Those findings are genuinely interesting. They also involve young-to-middle-aged men with injury-related, non-progressive cognitive complaints, no control group, and no biomarker measurement. They say nothing about a degenerative disease in a 78-year-old.

Two details about the field are worth stating plainly, because they shape how claims travel.

The Nature Medicine veterans paper discloses that several co-authors are shareholders in the clinic that delivered the treatment, and that two Stanford investigators are named on a patent application (no. 18/467,343) "related to the use of ibogaine to treat disorders associated with brain aging." A patent application is a commercial claim staked on a hypothesis. It is not evidence that the hypothesis is correct, and it is frequently misrepresented in marketing as though it were.

The overall ibogaine evidence base is far thinner than public coverage suggests. Sharma et al. (Journal of Psychopharmacology, 2026) conducted an updated scoping review of human ibogaine research and identified only three randomised controlled trials in total across all indications: a pilot in 20 cocaine-dependent adults, an ascending-dose safety study of noribogaine in 36 healthy volunteers, and a crossover trial in 27 opioid-dependent patients that found dose-dependent QTc prolongation with non-significant reductions in withdrawal.

Their conclusion is the most accurate one-sentence statement of the evidence available: "The available evidence is confined to case reports, observational analyses, and small early-phase or proof-of-concept studies" and "clinical use cannot be recommended without confirmation from larger, well-controlled trials."

If that is the state of the evidence in the conditions ibogaine has actually been studied for, the state of the evidence in dementia — where nothing has been studied — cannot be stronger.

The most active line of iboga-related neuroscience is not ibogaine itself. It is the effort to keep the molecule's plasticity effects while discarding its cardiac liability.

Cameron and colleagues (Nature, 2020) used function-oriented synthesis to produce tabernanthalog (TBG), a water-soluble, single-step analogue of ibogaine. In rodents it promoted structural neural plasticity, reduced alcohol- and heroin-seeking, and produced antidepressant-like effects, without hallucinogenic head-twitch behaviour and without potent hERG channel inhibition — the mechanism behind ibogaine's arrhythmia risk.

This matters for anyone reading about "ibogaine for Alzheimer's." Chemists who study the compound most closely have concluded that its therapeutic pharmacophore can be separated from its dangers, and their work has moved toward analogues. Several groups, including at the University of Texas Medical Branch, are now developing iboga-derived "psychoplastogens" as candidate neurotherapeutics.

Important qualifications:

• Tabernanthalog and related ibogalogs have been tested in rodents, in addiction, depression, and pain models — not in dementia models
• None has entered a clinical trial for Alzheimer's disease or any dementia
• Tabernanthalog is not an approved, prescribable, or legally obtainable medicine
• Analogue data do not transfer to ibogaine; the whole point of the analogue programme is that the parent compound is too dangerous

For comparison, psilocybin already has what ibogaine lacks in this space: a registered clinical trial in people with mild cognitive impairment or early Alzheimer's disease (NCT04123314). Even that study targets depression in this population, not cognition or disease progression.

Despite the absence of research, a market exists. Medical-tourism directories list clinics in Mexico advertising "ibogaine therapy for Alzheimer's disease" and "ibogaine therapy for dementia," with facilities named in Rosarito, Tijuana, Cancún, Playa del Carmen, and Puerto Vallarta. Quoted prices on one such directory range from roughly $4,500 to $13,000, including packages described as "Alzheimer's neuroplasticity microdosing" at $4,500–$6,500.

The mechanism advertised is almost always GDNF stimulation — the rodent finding described earlier, presented as though it were a demonstrated clinical effect in humans with dementia.

Some marketing is comparatively restrained. The same directory page states that ibogaine therapy for Alzheimer's is "not classified as an absolute cure, but rather as an incredibly powerful therapeutic tool." That is still an efficacy claim with no data behind it.

What these programmes have not produced, individually or collectively, is any published outcome data:

• No peer-reviewed case report of ibogaine in a person with dementia
• No case series, registry, or prospective cohort
• No cognitive, functional, imaging, or fluid-biomarker outcomes
• No published adverse-event reporting for this population

Separately, one Canadian provider has publicised a "neuroregenerative" programme aimed at Parkinson's disease, multiple sclerosis, and TBI. It does not list dementia as an indication, and it has published no outcome data either.

Marketing volume is not evidence volume. Families comparing clinics should recognise that they are comparing sales pages, not results.

Ibogaine blocks the hERG potassium channel, delaying cardiac repolarisation and prolonging the QT interval. This can precipitate torsades de pointes, a ventricular arrhythmia that can be fatal. Deaths have been documented in the ibogaine literature since the 1990s.

Brunt (Addiction, 2026) is explicit that these events occur at therapeutic doses and in people without pre-existing cardiac disease, that CYP2D6 metabolic variability drives large differences in exposure, and that any administration requires CYP2D6 genotyping and rigorous cardiovascular monitoring.

The dementia population is arguably the least suitable population for a hERG-blocking drug:

• Age itself. With 74% of patients aged 75 or older, structural heart disease, atrial fibrillation, conduction disease, and reduced renal and hepatic clearance are common — often undiagnosed.
• The first-line Alzheimer's drugs are themselves QT-liable. Zhang et al. (Frontiers in Pharmacology, 2024) analysed FDA adverse-event data from 2004–2022 and identified 557 cases of QT prolongation or torsades de pointes associated with cholinesterase inhibitors, with the strongest signal for donepezil, concentrated in elderly women. Adding a potent hERG blocker to donepezil stacks two risks.
• Antipsychotics used for behavioural symptoms of dementia carry an FDA boxed warning for increased mortality in elderly patients with dementia-related psychosis, and several prolong QT.
• Memantine plus ibogaine means two NMDA antagonists; the combination has never been studied.
• Vomiting is routine with ibogaine, causing potassium loss. Hypokalaemia has been present in reviewed fatalities, and older adults on diuretics start closer to that threshold.
• Ataxia and unsteadiness can persist 24–36 hours in a population already at high risk of falls and hip fracture.
• Delirium risk. Ibogaine produces a prolonged oneirogenic state. Older adults with cognitive impairment are highly prone to delirium and agitation, which are themselves associated with accelerated decline. No data describe how a person with dementia tolerates this experience.

Responsible supervised administration anywhere involves 12-lead ECG with QTc measurement, magnesium and potassium correction, echocardiography, full medication review, liver function testing, continuous telemetry, and advanced cardiac life support capability. None of this removes the risk — it narrows it.

Ibogaine is a Schedule I controlled substance in the United States, meaning it has no accepted medical use there and cannot be lawfully prescribed. It is not approved for dementia in any country. Treatment therefore takes place in unregulated or loosely regulated settings abroad, at the patient's own expense, with no insurance coverage and no regulatory recourse.

Dementia raises a consent problem that does not arise in addiction contexts. Alzheimer's disease progressively removes the capacity to understand, retain, and weigh information about a decision. A family member consenting on a relative's behalf to a Schedule I drug with a documented fatality signal, no efficacy evidence, and no published safety data in this age group is making a decision that no research ethics committee would currently approve.

Recent state ibogaine research funding — including Texas's $50 million initiative and Arizona's programme at the Barrow Neurological Institute — targets opioid use disorder, PTSD, and traumatic brain injury. Neither names Alzheimer's disease or dementia among its indications.

The gaps that would need to close before ibogaine could be considered a candidate therapy:

• No preclinical data. No study has given ibogaine to an animal model of amyloid or tau pathology.
• No human neurotrophic data. Whether ibogaine raises GDNF or BDNF in people has never been measured.
• No biomarker outcomes. No amyloid PET, tau PET, or plasma p-tau data exist for ibogaine.
• No pharmacokinetics in older adults, where clearance and cardiac reserve differ most.
• A failed proof of concept for the mechanism itself. Direct GDNF delivery did not produce clinical benefit in Parkinson's disease.

A published preclinical study in a dementia model, followed by a Phase 1 safety and pharmacokinetic study in older adults, would be the minimum credible starting point. Neither exists.

This page is educational information and is not medical advice. Anyone considering treatment for cognitive decline should discuss options with a neurologist or geriatrician.