This article comes from my Monday night livestream, where I work through a piece of research with whoever shows up to chat. My mom sent me the paper that started this one, which tells you how far it traveled into the popular press. The finding is worth your attention, so let me walk through what it actually says and where the evidence stands.
A quick frame before we start. This is education, not medical advice. Do not start lithium without talking to your clinician.
What does the new lithium and Alzheimer's research say?
When we talk about brain changes in Alzheimer's, we usually focus on sugar metabolism, oxidative stress, and tissue wear. Minerals and electrolytes get far less airtime. A 2025 paper in Nature (Aron et al., 2025) puts lithium near the center of the mechanism.
The core observation: your brain uses tiny amounts of lithium as part of its normal chemistry. Lithium functions here as a trace element doing ordinary work in healthy tissue, completely separate from its role as a psychiatric drug at gram-level doses. In Alzheimer's pathology, that trace lithium gets pulled out of circulation. Amyloid plaques sequester it, and that sequestration shows up early in the disease process, in both human tissue and mouse models (Aron et al., 2025).
To test whether lithium loss is a cause or a bystander, the researchers artificially lowered lithium in Alzheimer's-prone genetic mice. The mice developed progressive features resembling Alzheimer's. Then the researchers put lithium back into the water. Memory function returned, the mice ran their mazes better, and amyloid plaque dropped (Aron et al., 2025).
That second step matters. Removing lithium produced the pathology; restoring it reversed measurable parts of the pathology in the animal. That is a stronger design than a correlation.
How does lithium affect amyloid and tau? The GSK-3β mechanism
The proposed mechanism runs through an enzyme called GSK-3β (glycogen synthase kinase 3 beta). When GSK-3β is hyperactive, it drives amyloid plaque buildup and tau tangle formation, the two structural hallmarks of Alzheimer's.
Give lithium orotate to these mice and GSK-3β activity drops (Aron et al., 2025). The overproduction calms down. You also see a soothing of overactive glial cells.
The glial piece is what most coverage skips. Amyloid participates in the brain's innate immune system, helping with basic defense against microbes. You need amyloid at some level. Glial cells are also immune actors in the brain, especially in inflammatory signaling. When lithium gets sequestered into amyloid, the glial cells appear to over-activate, and that inflammatory environment pushes amyloid production higher. Restoring lithium brings that immune-style activation back down. The picture looks like an oxidative-stress and neuroinflammation loop that lithium helps regulate.
Confidence levels matter here. The GSK-3β mechanism and the glial inflammation findings are well-characterized in the mouse work. The assumption that the same mechanism operates in humans rests on what we already know about lithium's biochemistry, but that assumption stays extrapolation until human trials confirm it.
Why lithium orotate instead of lithium carbonate?
The researchers reintroduced lithium using lithium orotate, and that detail matters.
Lithium carbonate is the psychiatric workhorse. It is one of the better-understood psychiatric medications because clinicians have used it for decades, but it requires high doses, in the hundreds of milligrams, with close blood monitoring because the toxic range sits close to the therapeutic range.
Lithium orotate penetrates the brain with good affinity at a much lower effective dose. In the mouse study, the dosing was genuinely low compared to all prior lithium work. Convert the mouse dose to human body mass and you land in the range of the supplemental lithium orotate already on the shelf, somewhere in the single-digit milligram range.
A separate thread in the psychiatric literature supports the low-dose idea. Researchers have documented cases using single-digit milligram doses of lithium, far below the standard psychiatric dose, with reported effects on mood and related symptoms. Low-dose lithium showing broad regulatory effects at those levels is an observation with some support in the literature. The Alzheimer's mechanism is the new part.
How strong is the human evidence so far?
The human data on lithium for cognitive decline are early but more developed than most people expect.
- A 12-month randomized controlled trial in amnestic mild cognitive impairment (Forlenza et al., 2011) reported disease-modifying signals.
- Extended follow-up work on lithium in this population (Forlenza et al., 2019) tracked the longer arc.
- A microdose lithium study (Nunes et al., 2013) reported stabilized cognitive impairment in Alzheimer's patients.
Population data add context. A Danish study (Kessing et al., 2017) linked lithium levels in drinking water to dementia incidence.
This is also part of how lithium's psychoactive properties came to attention. Regions with naturally lithium-rich water have shown correlations with mood and lower depression rates in the literature, going back decades. The obvious next study follows people who bring lithium on board and checks whether lithium-rich mineral regions overlap with low-dementia populations.
Honest confidence summary: well-established that lithium engages GSK-3β and dampens neuroinflammation; emerging that low-dose lithium modifies cognitive decline in humans, with a handful of small RCTs and population studies pointing the same direction; extrapolation that supplemental lithium orotate prevents Alzheimer's in people. The trials to settle that last question have not been run.
The mouse-to-human translation problem
Whenever you hear "mice," apply a discount. The average translation rate from a rodent brain or aging finding to a human drug effect is low, on the order of roughly one in ten.
That discount shrinks under specific conditions. When the pathology mechanism is shared between species, when the animal model is a well-studied analog, and when the model is genetically engineered to match the human mechanism with knockout genes, confidence in translation goes up. The Alzheimer's mice here are genetically engineered and the GSK-3β mechanism is shared with humans, which raises confidence above that floor. Certainty is a different matter.
Should you add lithium to your brain-health strategy?
I am not recommending lithium for everyone. For people with significant family history or elevated Alzheimer's risk, it earns a place in the conversation alongside other tools we already use for diseases of aging.
Start with what we know works:
- Manage your APOE risk. If you carry APOE4/4, higher dietary starch and sugar will likely accelerate problems. Tighten that. With APOE2/3 status you have more flexibility.
- Fix sleep and apnea. Sleep quality and untreated apnea are major drivers of cognitive decline. This is part of why I point people toward sleep optimization consistently.
- Address vascular and metabolic factors. Strategic fasting and metabolic work matter for the aging brain.
- Support tissue and processing speed. Omega-3s and compounds like citicoline (CDP-choline) support myelination and remyelination. Alzheimer's involves real tissue loss and declining processing speed, and both respond to targeted support.
- Get a baseline. A blood panel for metabolic and genetic drivers, plus a QEEG brain map, tells you where your specific brain stands.
Lithium fits into that pharmacopeia for the biohacker who already thinks this way. A few milligrams of lithium orotate sits in a different safety category than high-dose carbonate, because the toxicity profile is gentler and the kidney burden appears lower. I am going to keep reading the research to understand the mechanism better, and I would treat lithium orotate as a tool to test and track rather than a blanket recommendation.
Common questions from the stream
Is the lithium in mineral spring water enough to matter? Some natural spring waters carry lithium at roughly 0.17 milligrams per liter. That is plausibly in the relevant range. Naturally lithium-rich springs exist in Oregon, Georgia, and Texas, and the historical mood-correlation data come from exactly these hard-water regions.
Does lithium show up on a QEEG? It should. Lithium's half-life runs roughly 18 to 36 hours. The rough rule of thumb for a psychoactive substance clearing the EEG is about five half-lives, so figure close to a week before lithium stops showing up on a map. Lithium orotate clears the serum fast, within about 24 hours, while brain concentrations stay elevated longer because the orotate carrier helps brain entry. An easy self-experiment: map on lithium, wash out, map again. The effect is acute enough that you can probably see the difference.
Does lithium affect sodium? Likely yes, partly by formulation. Carbonate and salt forms compete with sodium more directly. Lithium functions as an ion in solution, the same way sodium does, so it can shift the electrical and ionic balance regardless of form.
Any link to autism or ADHD symptoms? Lithium acts as a mood stabilizer even without depression present, and it sometimes gets used for ADHD when there is heavy mood involvement, often adult ADHD with depression. The mechanism there looks like the broad inflammatory and regulatory effect, rather than the executive-function or social circuits. For autism, the literature is thin. A few cell-culture and rat-model studies show GSK-3β and tau phosphorylation effects shared with the Alzheimer's pathway, but I would not call the autism evidence strong. Most of what exists is theory about drinking-water exposure, not treatment data. If you want the circuit-level view of social and sensory processing, I covered that in biohacking sensory and social processing.
Does it help seizures? Some reviews describe lithium having both proconvulsant and anticonvulsant actions depending on dose, with neuroprotective and neurotrophic effects reported across several neurological conditions. The exact mechanism for each of these effects is still largely unknown. My read is an energy-metabolism and inflammatory pathway, since glutamate activation and inflammation are pro-seizure and lithium pushes toward the inhibitory, soothing side. This is dose-specific and not well understood. Deeper seizure questions belong with a neurologist.
Where this goes next
Lithium has been on the map as a healthy trace mineral for decades, well past its role as a mood drug. The Aron et al. Nature paper (Aron et al., 2025) ties lithium to basic immune and inflammatory processes that contribute to Alzheimer's, and shows a measurable reversal of plaque burden in an animal model at a low dose.
The work worth watching next: cohort and retrospective studies following people who add low-dose lithium, and a geographic analysis of whether lithium-rich mineral regions overlap with low-dementia populations. Until those run, the human prevention claim is a promising hypothesis with preliminary support, and that is worth tracking carefully.
If you want to track your own brain over time, a QEEG baseline is the place to start. You can read more about what it shows in the brain mapping guide. I read every comment on the stream, so bring questions and tell me what you want covered next.