Low-dose lithium
Lithium at subtherapeutic and trace doses, including lithium orotate
Written by Aaron CuhaReviewed Sep 2026
Also known as: Lithium orotate, Microdose lithium, Trace lithium, Nutritional lithium
The drinking water literature that gets quoted as evidence is not a trial: 15 ecological studies pooled to a regression coefficient of -0.27 for suicide mortality, and the meta-analysis authors label their own result as subject to the ecological fallacy. The randomised trials that exist used lithium carbonate at 150 to 600 mg targeting a blood level, which is a different intervention from the milligram doses of lithium orotate sold in shops.
Overview
Low-dose lithium is where an entire consumer category has been built on top of a study design that cannot support it, and where the actual randomised evidence is for a different dose of a different salt.
The popular case rests on ecological studies. These compare lithium concentrations in the tap water of counties or municipalities against death rates in those same areas. A meta-analysis of 15 such studies found a consistent inverse association with suicide mortality, pooled regression coefficient -0.27 (95 percent CI -0.47 to -0.08, p = 0.006) with heterogeneity of 83.3 percent. The authors' own conclusion begins by naming the problem: this is a synthesis of ecological studies, which are subject to the ecological fallacy and bias. They go on to suggest that randomised community trials of lithium supplementation of the water supply would be the way to test the hypothesis. Nobody has run one. A separate paper in the same tradition reported an inverse correlation between tap water lithium and all-cause mortality across 18 Japanese municipalities covering 1,206,174 people, and paired it with a C. elegans experiment in which a comparably low lithium chloride concentration extended lifespan.
An ecological correlation tells you about places, not people. No individual in those datasets was measured for lithium intake, and no individual outcome was linked to an individual exposure.
The randomised evidence is separate and more interesting, and it is important to be precise about the dose. In 45 people with amnestic mild cognitive impairment, twelve months of lithium carbonate titrated to a subtherapeutic blood level of 0.25 to 0.5 mmol/L significantly lowered cerebrospinal fluid phosphorylated tau and improved two cognitive measures. A follow-up in 61 people over two years found that the placebo group declined while the lithium group remained stable, with better memory and attention at 24 months and a rise in cerebrospinal fluid amyloid-beta 42 at 36 months. Those are real randomised results. They used prescription lithium carbonate, in the 150 to 600 mg range, monitored by blood level. A lithium orotate capsule typically supplies a few milligrams of elemental lithium, roughly two orders of magnitude less, and produces no measurable serum level.
There is one genuinely small-dose randomised trial: 300 micrograms of lithium daily for fifteen months in Alzheimer's disease, in which the treated group did not lose points on the mini mental state examination while controls did. It is a single small study and has not been replicated.
The newest work is mechanistic and is in mice and human brain tissue, not a trial. A 2025 Nature paper reported that endogenous brain lithium is reduced in mild cognitive impairment, that depleting dietary lithium in mice increased amyloid and phospho-tau and accelerated cognitive decline, and that lithium orotate prevented those changes in mouse models. That is a strong reason to run a human trial. It is not a human trial.
Mechanism of action
Lithium inhibits glycogen synthase kinase 3 beta, which is upstream of tau phosphorylation and of amyloid precursor protein processing, and it inhibits inositol monophosphatase, depleting the inositol pool that supports phosphatidylinositol signalling. Both mechanisms are dose-dependent, and the doses at which they have been demonstrated in human brain are the prescription ones, monitored by serum level. The 2025 work adds a different framing: that lithium is a physiologically regulated trace element in the brain whose depletion is an early event in Alzheimer's pathology, and that amyloid sequesters it, further reducing bioavailability. Lithium orotate was used in that work specifically because it binds amyloid less than the carbonate salt. If that framing is right, low-dose lithium would be correcting a deficiency rather than acting as a drug, which would be a different kind of claim and would require a way to measure the deficiency in a living person. No such clinical test is in routine use.
Human evidence
Two quite different bodies of human data that are frequently merged. A large ecological literature correlating tap water lithium with population death and suicide rates, which cannot attribute an outcome to an individual's exposure. And a small set of randomised trials in cognitive impairment, all but one of which used prescription lithium carbonate at doses far above what supplements deliver.
- Meta-analysis of 15 ecological studies: pooled inverse association with suicide mortality, beta -0.27, explicitly labelled by its authors as subject to the ecological fallacy.
- 18 Japanese municipalities, 1,206,174 people: inverse correlation between tap water lithium and all-cause mortality, beta -0.661, p = 0.003. An area-level correlation.
- 45 people with mild cognitive impairment, 12 months of lithium at 0.25 to 0.5 mmol/L serum: cerebrospinal fluid phospho-tau fell, two cognitive measures improved.
- 61 people with mild cognitive impairment, 2 years at the same subtherapeutic serum range: placebo declined, lithium remained stable, with better memory and attention at 24 months.
- Alzheimer's disease, 300 micrograms daily for 15 months: mini mental state examination scores held while controls declined. One small trial, not replicated.
- No randomised trial has tested lithium orotate, the form actually sold, against any endpoint in humans.
What this does not tell you: Ecological studies compare places. Counties with more lithium in their water differ from counties with less in many ways that are not lithium, and the meta-analysis reports 83.3 percent heterogeneity, meaning the individual studies disagree substantially with each other. The randomised cognition trials are small, single-centre, and used a monitored prescription drug at doses roughly two orders of magnitude above a typical supplement. The 300 microgram trial is the only one at a supplement-like dose and it stands alone. Nobody has measured whether any of this changes how long a person lives.
Reading the research record
The label on this literature matters more than usual, so here it is plainly: the drinking water studies are ecological, they are not trials, they cannot establish that giving lithium to a person changes that person's risk, and the researchers who pooled them say so in their own conclusion. Anyone citing them as evidence that low-dose lithium works is citing a design that was chosen because it was cheap and available, not because it answers the question.
The second thing to hold onto is the dose equivocation. The phrase low-dose lithium covers two interventions separated by roughly a hundredfold. Subtherapeutic lithium carbonate at 150 to 600 mg, titrated to a serum level of 0.25 to 0.5 mmol/L, is a prescription drug at a reduced dose with real randomised evidence behind it in mild cognitive impairment, and with real requirements for monitoring kidney and thyroid function. Lithium orotate at a few milligrams of elemental lithium is a supplement that produces no measurable serum level and has been tested in no randomised trial. Marketing copy moves between the two as if they were the same thing.
The 2025 Nature work is the most interesting development in this area in years, and it is worth being precise about what it is: human post-mortem brain tissue plus mouse experiments. If its framing holds, low-dose lithium would be repletion of a deficiency rather than drug action, which would fit this site's usual rule about correcting measured deficits. The missing piece is exactly the thing that rule requires, which is a way to measure the deficiency in a living person and a trial showing that correcting it changes an outcome.
This archive's stated position on supplements is that you take something because a number on your bloodwork is out of range, and that you retest and stop if the number does not move. Applying that rule to this group is uncomfortable, because for most of them there is no deficiency state to correct and therefore no marker that tells you whether the bottle is doing anything. Where a compound here does move a readable number, the number is a surrogate rather than a deficiency, so moving it is not the same as fixing something that was broken.
The evidence, charted
Fig. 1 · evidence composition
4of 6 citations (67%) are in people
Counted from the citation list on this page. The Human count is the same number shown in the badge at the top. Both understate any literature larger than the sources we cite.
Fig. 2 · evidence over time
Evidence spans 5 distinct years, 2011 to 2025, counted from the citation list on this page.
Fig. 3 · legal status at a glance
US
Not approved
UK
Not approved
AU
Not approved
CA
Not approved
Not approved in any of the four jurisdictions shown. A jurisdiction's classification is a regulatory fact, not a verdict on the science; see Legal status below for the full text and any notes.
Key studies & citations
- Review2020
Association between naturally occurring lithium in drinking water and suicide rates: systematic review and meta-analysis of ecological studies.
15 ecological studies pooled. Inverse association between drinking water lithium and total suicide mortality, pooled beta -0.27 (95 percent CI -0.47 to -0.08, p = 0.006, I squared 83.3 percent); male beta -0.26 was not significant (p = 0.08) and female beta -0.13 was (p = 0.03). The authors state that these studies are subject to the ecological fallacy and bias, and suggest randomised community trials of water supplementation as the way to test the hypothesis. No such trial has been run.
British Journal of Psychiatry - Human2011
Low-dose lithium uptake promotes longevity in humans and metazoans.
Two separate pieces of work in one short paper. An observational weighted regression across 18 neighbouring Japanese municipalities totalling 1,206,174 individuals found an inverse correlation between tap water lithium concentration and all-cause mortality (beta -0.661, p = 0.003). Separately, exposing Caenorhabditis elegans to a comparably low lithium chloride concentration extended lifespan (p = 0.047). The human half is an ecological correlation between places, not an intervention in people, and the lifespan half is a worm.
European Journal of Nutrition - Human2011
Disease-modifying properties of long-term lithium treatment for amnestic mild cognitive impairment: randomised controlled trial.
45 participants with amnestic mild cognitive impairment randomised to lithium (titrated to 0.25 to 0.5 mmol/L serum, a subtherapeutic range) or placebo for 12 months. Lithium significantly lowered cerebrospinal fluid phosphorylated tau (p = 0.03) and improved the cognitive subscale of the Alzheimer's Disease Assessment Scale and attention tasks. Tolerability was good and adherence 91 percent. Registered as NCT01055392. Note the dose: prescription lithium carbonate monitored by blood level, not a trace-dose supplement.
British Journal of Psychiatry - Human2019
Clinical and biological effects of long-term lithium treatment in older adults with amnestic mild cognitive impairment: randomised clinical trial.
61 community-dwelling older adults with mild cognitive impairment randomised to subtherapeutic lithium (0.25 to 0.5 mEq/L) or placebo for 2 years double blind, followed a further 24 months single blind. The placebo group declined cognitively and functionally while the lithium group remained stable; lithium was associated with better memory and attention at 24 months and a significant rise in cerebrospinal fluid amyloid-beta 42 at 36 months. Same registration, NCT01055392.
British Journal of Psychiatry - Human2013
Microdose lithium treatment stabilized cognitive impairment in patients with Alzheimer's disease.
The only randomised trial located at a dose comparable to supplement products: 300 micrograms of lithium once daily for 15 months in Alzheimer's disease patients. The treated group showed no decrease in mini mental state examination performance while controls declined, with differences appearing from three months and increasing. A single small study, not replicated, and the report gives limited methodological detail.
Current Alzheimer Research - Animal2025
Lithium deficiency and the onset of Alzheimer's disease.
Endogenous lithium was the only metal significantly reduced in the brains of individuals with mild cognitive impairment, with bioavailability further reduced in Alzheimer's disease by amyloid sequestration. In mice, reducing cortical lithium by about 50 percent through diet increased amyloid-beta deposition and phospho-tau, activated microglia, caused loss of synapses, axons and myelin, and accelerated cognitive decline, partly through GSK3 beta. Lithium orotate replacement prevented pathology and memory loss in mouse models and ageing wild-type mice. Human post-mortem tissue plus mouse experiments; no human intervention. NIH funded.
Nature
Frequently asked questions
Does lithium in drinking water reduce suicide?
The studies that report this are ecological: they compare lithium levels in an area's water with that area's death rates, without measuring any individual's intake or outcome. Pooled across 15 of them the association is consistent, beta -0.27 for total suicide mortality. The meta-analysis authors state that the design is subject to the ecological fallacy and call for randomised community trials, which have not been run. It is a hypothesis with population-level support, not a demonstrated effect.
Is lithium orotate the same as the lithium in the studies?
Not in the randomised cognition trials. Those used prescription lithium carbonate at 150 to 600 mg titrated to a serum level of 0.25 to 0.5 mmol/L. A lithium orotate supplement typically supplies a few milligrams of elemental lithium, roughly two orders of magnitude less, and produces no measurable serum level. One small randomised trial did use 300 micrograms daily in Alzheimer's disease and reported stabilised mini mental state examination scores over 15 months; it has not been replicated.
What did the 2025 Nature paper actually show?
That endogenous lithium was the only metal significantly reduced in the brains of people with mild cognitive impairment, that amyloid sequesters it further in Alzheimer's disease, and that depleting dietary lithium in mice by about half drove amyloid, phospho-tau, microglial activation, synapse and myelin loss and faster cognitive decline. Lithium orotate replacement prevented that in mouse models. It is human tissue plus mouse experiments. No person was given lithium in that study.
Is low-dose lithium safe to take without monitoring?
Prescription lithium requires serum level, kidney and thyroid monitoring because its therapeutic window is narrow and long-term use affects both organs. Supplement-dose orotate is far below that range and produces no measurable level, which is both the reason it is sold without monitoring and the reason there is no way to tell whether you are getting a dose that does anything. This site does not give dosing advice; the point is that the absence of monitoring and the absence of a measurable effect are the same fact.
Does it extend lifespan?
No human trial has measured lifespan. The lifespan claim traces to a 2011 paper that paired an ecological correlation across 18 Japanese municipalities with a C. elegans experiment in which low-dose lithium chloride extended worm lifespan. A correlation between places and a result in a roundworm.