Memantine

Memantine is a prescription drug originally approved for moderate to severe Alzheimer's disease. It works on the glutamate system in the brain, dialling down the kind of overactive signalling that drives neuronal stress, intrusive thoughts, and chronic pain amplification. Most people you'll meet taking it aren't using it for dementia, they're using it off-label for OCD, treatment-resistant anxiety, opioid tolerance, or as a long-term neuroprotective tool.

It's not stimulating and it's not sedating. It's slow. The first 2-4 weeks usually feel like nothing, then somewhere around week 6-8 the people it works for notice their head is quieter, the loop they were stuck in has slack in it, or the chronic pain they've been managing is more manageable. If you're expecting an acute effect like caffeine or a benzo, you'll think it's not working and quit too early. The whole compound trades immediacy for tolerability.

Deep-dive

Memantine is an uncompetitive, low-to-moderate affinity NMDA receptor antagonist. NMDA receptors are one of the main glutamate receptors in the brain and they sit at the centre of learning, memory, and synaptic plasticity. The catch is that when they fire too much, calcium pours into neurons and damages them. This is called excitotoxicity and it's involved in Alzheimer's, Huntington's, stroke, chronic pain syndromes, and probably parts of OCD and depression.

The clever thing about memantine is that it only blocks the receptor when it's already overactive. Under normal signalling it gets pushed off the channel quickly enough that learning and memory carry on. Under sustained pathological glutamate flooding, it stays bound and shuts the channel down. This voltage-dependent, fast-off mechanism is what separates it from ketamine and from older NMDA blockers like PCP, which shut everything down indiscriminately and produce dissociation, psychosis, or neurotoxicity. Memantine has minor activity at 5-HT3 serotonin receptors and α4β2 nicotinic acetylcholine receptors, and recent work suggests its neuroprotective effect may run partly through the nicotinic pathway rather than purely through NMDA blockade.

Alzheimer's disease. This is the only approved use. Memantine produces modest improvements in cognition, function, and behavioural symptoms in moderate to severe AD, both as monotherapy and combined with cholinesterase inhibitors like donepezil. It treats symptoms, it doesn't slow the underlying neurodegeneration. If you're considering it for early Alzheimer's or mild cognitive impairment, the evidence is much weaker.

OCD. This is the off-label use with the strongest data. A 2017 randomised, placebo-controlled trial added 20 mg/day of memantine to SSRI-refractory OCD patients and saw a 40.9% reduction in Y-BOCS scores by week 12, with 73% achieving treatment response, against zero improvement in the placebo arm. The benefit took 8 weeks to show up, which fits the slow ramp pattern you see clinically. More recent trials have largely confirmed the augmentation effect, particularly on executive function. The mechanism makes sense: OCD looks like glutamatergic hyperactivity in the cortico-striatal-thalamic loop, and memantine dampens that hyperactivity without flatlining it.

ADHD and executive function. A pilot trial of memantine added to methylphenidate in adults with ADHD found selective improvements in executive function, particularly working memory and inhibitory control, that the stimulant alone wasn't producing. As monotherapy in children, memantine underperformed methylphenidate on standard ADHD measures, so it's not a stimulant replacement, it's a possible add-on for the executive function gaps stimulants don't fill.

Depression. Mixed and mostly underwhelming. The largest placebo-controlled monotherapy trial at 5-20 mg/day was negative. There are positive signals when memantine is combined with an SSRI in geriatric depression with cognitive complaints, but it's not a primary depression drug. If someone is selling it to you as a depression treatment, the evidence isn't there.

Opioid tolerance and chronic pain. This is the underrated use. A 2025 randomised trial gave 10 mg/day to opioid-dependent patients with opioid-induced hyperalgesia and saw significantly increased cold pain tolerance and reduced pain intensity within 2-4 weeks. Meta-analyses of memantine for chronic pain show modest benefit across neuropathic pain, fibromyalgia, and post-surgical pain, with the largest signal in syndromes that involve central sensitisation. The mechanism is that chronic opioid use upregulates NMDA receptors, which is what drives tolerance and the paradoxical worsening of pain on long-term opioids. Blocking NMDA partially reverses this.

Other off-label uses. Memantine has been trialled in binge eating disorder, alcohol use disorder, gambling, schizophrenia, autism, PTSD, and migraine. The evidence ranges from suggestive open-label data to negative placebo-controlled trials. A systematic review of psychiatric off-label uses concluded the data was insufficient to recommend routine use outside OCD and possibly binge eating. None of these uses are wrong to consider, but you should know you're in territory where the upside is plausible and the proof is thin.

Cognitive enhancement in healthy people. There's no evidence memantine improves cognition in people with normal brains. The one systematic review on neuroenhancement with memantine in healthy adults found no consistent benefit. The framing on nootropic forums that memantine resets glutamate tolerance from chronic stimulant use is mechanistically plausible but clinically unproven. If you're cognitively healthy and you're taking it for performance, you're running an experiment with little supporting data.

Women. Most memantine trials, including the OCD and pain studies, have included women throughout, and the clinical effects look comparable between sexes. Pharmacokinetics differ: women have roughly 45% higher plasma exposure than men at the same dose, but this disappears once you adjust for body weight. In practice this means a smaller woman taking 20 mg may be running plasma levels closer to what a larger man gets at 25-30 mg, which is worth knowing if side effects show up at the standard dose. Memantine is renally cleared, so age-related drops in kidney function matter more than sex per se. There's no data supporting use in pregnancy or lactation, and animal studies at high doses showed reduced pup weights and skeletal effects, so it's not recommended in either.

Limits of the evidence. Outside Alzheimer's, the trials are mostly small, short, single-centre, and often from a handful of Iranian research groups for the OCD work. The OCD signal has held up across multiple trials, but most other off-label uses rest on thin foundations. Memantine is well tolerated, which is the main reason it keeps getting tried for new indications. Tolerability isn't the same as efficacy.

Dosage:

Standard target dose: 20 mg/day, split as 10 mg twice daily, or 28 mg/day of the extended-release version once daily. This is the dose used for Alzheimer's, OCD augmentation, and most off-label indications. Doses below 10 mg are subtherapeutic for almost everything

Titration matters. Starting at 20 mg out of the gate causes dizziness, headache, and confusion in a large fraction of people. The standard ramp is 5 mg/day in week 1, 10 mg/day in week 2, 15 mg/day in week 3, 20 mg/day in week 4. Skipping the titration is the single most common reason people abandon memantine. If you're sensitive, slow it down further

Timing: Once you're at the full dose, twice-daily dosing (morning and evening) is the standard, but once-daily 20 mg is similarly tolerated for most people. The XR formulation is once-daily by design. Food doesn't affect absorption, so take it whenever's convenient. Some people get insomnia with an evening dose, in which case shift it earlier

For OCD or chronic pain: Same titration, same target dose. Don't judge response before week 8. Most of the OCD trials show the inflection point between weeks 6 and 12. Quitting at week 4 because nothing's happening is the most common error

For opioid-induced hyperalgesia: Trials have used 10 mg/day and seen benefit within 2 weeks, which is faster than the OCD use case. Higher doses haven't been shown to add benefit for this indication

Renal adjustment: Memantine is cleared by the kidneys. If your creatinine clearance is below 30 mL/min, the maximum dose drops to 10 mg/day (or 14 mg XR). Above 30, no adjustment needed. Worth checking baseline kidney function before starting if you're over 60 or have any history of renal issues

Women at low body weight should consider whether 20 mg is the right ceiling. The 45% higher plasma exposure women see at the same absolute dose mostly washes out with body weight adjustment, so a 55 kg woman is running effectively higher levels than a 90 kg man at 20 mg. Not dangerous, but if you're getting side effects at the standard dose, sitting at 15 mg is reasonable

If you miss several days: Restart at the low end of the titration rather than jumping back to 20 mg. Restarting at the maintenance dose after a gap is what triggers the worst tolerability problems

Here's what you can expect:

Weeks 1-4 are mostly nothing. You're titrating up and the dose is too low to do much. Some people notice mild dizziness or a faint head fog as they ramp, especially in the first few days of each step up. This usually settles within a week.

Weeks 4-8 are where you start to find out whether it's working for you. For OCD, the obsessions feel less sticky, the urge to perform compulsions has more give in it, you can sit with discomfort longer before acting on it. For chronic pain, the baseline intensity drops a notch and you're less reactive to flares. For executive function in ADHD, you can hold things in working memory more reliably and the impulsive jumps between tasks slow down. None of this is dramatic. It's the kind of change you notice retrospectively when you realise you haven't done a thing you used to do.

Weeks 8-12 is where you make the decision. If memantine is going to work, the effect will be clear by then. If you're at full dose, you've stuck with it, and nothing has shifted, it probably isn't your compound. The trials suggest somewhere around half to two-thirds of people get a meaningful response in the indications where it does work, which means a real fraction get nothing and should stop.

With long-term use, the effect tends to be stable rather than escalating. People who respond at month 3 generally still respond at month 12. There's no clear pattern of tolerance or rebound on stopping, but stopping abruptly after long-term use isn't recommended, taper over a few weeks instead.

Side effects & risks:

Dizziness is the most common side effect, hitting roughly 5-7% of people in trials and more during titration. Usually settles. If it persists at the target dose, slow the ramp or sit at a lower dose for longer

Headache in around 5%. Often resolves within 1-2 weeks of reaching steady state. Hydration helps

Constipation in around 5%. Standard fix: more fibre, more water, magnesium if you tolerate it

Confusion or mental fog at higher doses or during rapid titration. This is the one that tends to scare people. It usually means the dose is escalating too fast for you. Drop back a step and re-titrate slower

Insomnia if dosed in the evening. Move the second dose to early afternoon or take the full dose in the morning if you're using XR

Hallucinations are listed as a side effect, mostly observed in patients with severe Alzheimer's where they're hard to separate from the underlying disease. Rare in healthy adults at standard doses

Blood pressure can rise modestly. Hypertension shows up as a 1-10% adverse event in trials. Worth a baseline reading and a recheck if you have any cardiovascular history

MAOI interaction: Combining memantine with monoamine oxidase inhibitors hasn't been formally studied and is generally avoided. If you're on an MAOI for depression, don't add memantine without a psychiatrist signing off

Amantadine, ketamine, dextromethorphan: All NMDA antagonists. Stacking them with memantine compounds the receptor blockade and isn't recommended. This includes recreational ketamine use, which becomes substantially less predictable on memantine

Sodium bicarbonate, carbonic anhydrase inhibitors, anything that alkalises urine: Memantine is excreted unchanged by the kidneys via tubular secretion. Alkalising the urine reduces clearance and raises plasma levels, sometimes substantially. Worth knowing if you take baking soda for performance, are on acetazolamide, or use urinary alkalinisers for kidney stones

Renal impairment: Memantine accumulates if your kidneys aren't clearing it. Dose adjustment required below 30 mL/min creatinine clearance. If you're over 65, a baseline check is worth doing

Pregnancy and breastfeeding: Skip it. Animal studies at high doses showed reduced offspring weight and skeletal effects, and there's no human safety data. Not worth the risk for an off-label use

Long-term safety is well established for the dementia population over multiple years. For off-label use in younger, healthier adults running it for years, the data is thinner. Most likely fine, but the honest answer is we don't have 20-year cohorts of healthy 30-year-olds on memantine

🩸

Blood markers

Creatinine and eGFR, baseline before starting. Memantine is renally cleared and accumulates if your kidneys aren't keeping up. If your eGFR is below 30, the maximum dose drops to 10 mg/day. Recheck annually on long-term use, especially if you're over 60.

Blood pressure, baseline and a recheck at 4-8 weeks once you're at the full dose. Memantine usually doesn't move BP, but mild hypertension shows up as a side effect in a minority. You want a reference point.

Liver enzymes (ALT, AST), baseline if you have any liver history. Memantine isn't significantly hepatically metabolised and hepatotoxicity is rare, but worth a baseline if you're stacking with other compounds that load the liver.

Electrolytes and bicarbonate, useful if you take supplements or medications that alkalise urine, since they raise memantine plasma levels through reduced renal clearance.

For most people on standard doses (20 mg/day) with normal kidney function and no relevant medication interactions, baseline kidney function and BP is enough. The people who actually need the fuller panel are those over 60, anyone with kidney or liver history, anyone running it long-term off-label, and anyone stacking with other neuroactive compounds.

Prescription-only in most countries. Available generically as memantine HCl since 2010, brand names include Namenda (US) and Ebixa (EU).