Lithium at doses far below psychiatric levels is being studied for Alzheimer's disease. A 2025 Nature study found that brain lithium falls early in Alzheimer's and that replacing it with lithium orotate reversed pathology and memory loss in mice. Population studies link higher lithium in drinking water and lithium prescriptions to less dementia, and a small Brazilian trial reported slower decline. But a 2026 pilot randomized trial in 80 older adults with mild cognitive impairment missed all six of its primary outcomes over two years. The human case is unresolved.
Summary drafted from this entry’s sources; editorial review pending.
01Key questions
Does lithium prevent Alzheimer's?
Not proven. In mice, lithium orotate at low doses reversed Alzheimer's-like pathology and memory loss (Aron et al., Nature 2025). In people, a 2026 pilot trial in mild cognitive impairment missed all six primary outcomes (Gildengers et al., JAMA Neurology 2026).13
What did the 2026 trial show?
In 80 adults with mild cognitive impairment randomized to low-dose lithium carbonate or placebo for two years, none of six primary outcomes reached significance; verbal recall leaned toward lithium (P = .05) and brain volume loss was equal. The trial was a small pilot and used carbonate rather than orotate, so it does not settle the question (NCT03185208).33
What is the population evidence?
A Danish nationwide study linked the highest drinking-water lithium levels to lower dementia rates, though the relationship was not linear (Kessing et al., JAMA Psychiatry 2017), and a UK records study found lithium prescription associated with lower dementia risk (Chen et al., PLoS Medicine 2022).56
Is low-dose lithium safe?
Trial doses were far below those used in bipolar disorder, where lithium affects the kidneys and thyroid and needs blood monitoring. In the 2026 trial, diarrhoea, tiredness and tremor were more common on lithium, while kidney changes were equally common on placebo (Gildengers et al., 2026). Over-the-counter orotate has no human dementia data.3
Lithium is best known as a psychiatric drug for bipolar disorder, taken at doses that need blood monitoring. A separate idea has been building for 25 years: that much lower amounts of lithium, closer to what some drinking water contains, protect the ageing brain. A 2025 Nature paper turned that idea into a headline by showing that lithium is the one metal depleted early in the Alzheimer's brain. The first randomized trial of low-dose lithium in mild cognitive impairment (MCI), published in 2026, then missed all of its primary outcomes. Both facts belong in the same sentence.
02The 2025 Nature paper: lithium as a trace nutrient for the brain
Aron and colleagues (Yankner lab, Harvard) measured dozens of metals in post-mortem brain tissue. Lithium was the only one significantly reduced in people with MCI, and in Alzheimer's disease it was further trapped by amyloid plaques. In mice, cutting dietary lithium so that cortical lithium fell by about half sped up amyloid and phospho-tau build-up, inflamed microglia, destroyed synapses and myelin and accelerated memory loss, partly through the enzyme GSK3β. Replacing it with lithium orotate, a salt that binds less to amyloid, prevented the pathology and memory loss in Alzheimer's mouse models and in normally ageing mice (PMID 40770094).
That is strong mechanistic work, but it is mouse work. The authors themselves describe replacement with amyloid-evading salts as a "potential approach", and a 2026 review notes that the repletion hypothesis "awaits independent replication" (PMID 42268602).
03Human evidence
| Study | Type | Finding |
|---|---|---|
| Gildengers et al., JAMA Neurology 2026 | Pilot RCT, 80 adults 60+ with MCI, 2 years, lithium carbonate vs placebo | None of 6 coprimary outcomes was significant. Verbal recall declined 0.73 vs 1.42 points a year (difference 0.69, 95% CI 0.01 to 1.37, P = .05); hippocampal and cortical volume fell equally in both groups. More diarrhoea, tiredness and tremor on lithium (PMID 41770546) |
| Forlenza et al., Br J Psychiatry 2019 | RCT, 61 older adults with amnestic MCI, 2 years blinded plus 2 years follow-up | Placebo group declined cognitively and functionally while the lithium group stayed stable; better memory and attention scores at 24 months and a rise in CSF amyloid-β42 at 36 months (PMID 30947755) |
| Kessing et al., JAMA Psychiatry 2017 | Danish nationwide case-control, 73,731 dementia cases | Drinking water above 15 µg/L lithium linked to 17% lower dementia incidence versus 2-5 µg/L, but the curve was non-linear (5-10 µg/L was linked to higher incidence) and regional confounding could not be excluded (PMID 28832877) |
| Chen et al., PLoS Medicine 2022 | UK retrospective cohort, 29,618 patients aged 50+ (548 on lithium) | Lithium use associated with lower dementia risk (HR 0.56, 95% CI 0.40 to 0.78); small exposed group, mostly people with mood disorders (PMID 35298477) |
Swipe table →
04Reading the null trial honestly
The Pittsburgh trial was a feasibility pilot, so "no significant coprimary outcome" partly reflects its size: 80 people cannot reliably detect a modest effect over two years. The memory difference pointed the right way and sat right on the significance line. But a pilot is also where a large effect would have shown up, and it did not. Brain-volume measures, the most objective outcomes, showed nothing. The honest reading is that low-dose lithium carbonate has not been shown to slow cognitive decline in MCI, and that a properly powered trial is needed. Proposals now favour lithium orotate, following the mouse data, which no human trial has yet tested for this purpose (PMID 42268602).
05Safety
At psychiatric doses lithium has a narrow therapeutic window and can affect the kidneys and thyroid; toxicity is a real risk in older people, especially with dehydration or interacting drugs such as diuretics and NSAIDs. The low doses in MCI trials were tolerated, with creatinine rises equally common on placebo, but diarrhoea, tiredness and tremor were more frequent on lithium (PMID 41770546). Over-the-counter lithium orotate is sold as a supplement with no human dementia data and no quality control comparable to prescription products.
06Where it sits on the Heretic scale
Early. Striking mouse mechanism, supportive population data and one small positive RCT, offset by a null pilot RCT. What would move it up: a powered, multi-year trial showing slower decline on cognitive and imaging outcomes. See also Shingles vaccination and dementia and 40 Hz gamma stimulation for other non-drug-company dementia bets, and Autophagy for lithium's role as an autophagy inducer.
What studies and practitioners report
Amounts as reported in the cited sources: trial doses with the population studied, and practitioner-reported ranges marked as such. Not medical advice.
- Trial serum 0.25–0.5 mEq/L · lithium carbonate, titrated to a low serum level, 2 years older adults with amnestic MCI, n=61 PMID 30947755
Further reading: detailed protocols
- ClinicalTrials.gov: low-dose lithium in MCI (NCT03185208) clinicaltrials.gov
07Connections
08Sources
- Aron L, et al. (2025). "Lithium deficiency and the onset of Alzheimer's disease." Nature 645:712-721. PMID 40770094 . DOI 10.1038/s41586-025-09335-xPubMed 40770094 · DOI 10.1038/s41586-025-09335-x
- Moore GJ, et al. (2026). "The 25-Year Evolution of Lithium as a Disease-Modifying Agent in Dementia: A Narrative Review." JAMA Psychiatry 83(9):976-986. PMID 42268602 . DOI 10.1001/jamapsychiatry.2026.1296PubMed 42268602 · DOI 10.1001/jamapsychiatry.2026.1296
- Gildengers AG, et al. (2026). "Low-Dose Lithium for Mild Cognitive Impairment: A Pilot Randomized Clinical Trial." JAMA Neurol 83(4):310-319. PMID 41770546 . DOI 10.1001/jamaneurol.2026.0072 . Trial NCT03185208PubMed 41770546 · DOI 10.1001/jamaneurol.2026.0072 · NCT03185208
- Forlenza OV, et al. (2019). "Clinical and biological effects of long-term lithium treatment in older adults with amnestic mild cognitive impairment: randomised clinical trial." Br J Psychiatry 215(5):668-674. PMID 30947755 . DOI 10.1192/bjp.2019.76PubMed 30947755 · DOI 10.1192/bjp.2019.76
- Kessing LV, et al. (2017). "Association of Lithium in Drinking Water With the Incidence of Dementia." JAMA Psychiatry 74(10):1005-1010. PMID 28832877 . DOI 10.1001/jamapsychiatry.2017.2362PubMed 28832877 · DOI 10.1001/jamapsychiatry.2017.2362
- Chen S, et al. (2022). "Association between lithium use and the incidence of dementia and its subtypes: A retrospective cohort study." PLoS Med . PMID 35298477 . DOI 10.1371/journal.pmed.1003941PubMed 35298477 · DOI 10.1371/journal.pmed.1003941
09Entry history
Latest content update.
Heretic informs and never prescribes. Where amounts appear, they describe what published studies used. Talk to a qualified clinician before acting on anything here.