Microdose Lithium for Longevity & Brain Aging: The Evidence

Lithium is having a longevity moment, and a genuinely interesting one. It's not a trendy peptide or an exotic plant extract — it's a basic element, the same one used for decades to treat bipolar disorder, now being floated as a low-dose "brain-aging" intervention. A high-profile 2025 study reignited the idea by linking lithium deficiency to Alzheimer's disease in mice. But this is exactly the kind of story where the gap between a striking animal result and a safe, proven human intervention is wide — and where a drug's narrow safety window makes "just try some" genuinely risky advice. This page walks through it honestly. For the wider map of what's earned its place versus what's hype, start with our pillar on longevity medicine: what's proven vs hyped.

What "microdose lithium" means

The word "lithium" covers a huge dose range, and conflating those ranges is the single biggest source of confusion. Therapeutic lithium for bipolar disorder is prescribed at hundreds of milligrams a day (of lithium carbonate) and titrated against blood levels, because the window between an effective dose and a toxic one is famously narrow. Microdose lithium — usually lithium orotate or low-dose carbonate sold as a supplement, often around 1–5 mg of elemental lithium — is orders of magnitude lower, in the range of what people get naturally from drinking water and food. Trace lithium is the amount in tap water, which varies by region. The longevity conversation is mostly about the microdose and trace ranges, and almost all the suggestive human data come from those low exposures — not from prescription doses.

The 2025 Nature study: real, important — and in mice

The study that reignited interest was published in Nature in 2025 (Aron and colleagues), and its central claim is striking: that lithium deficiency may be an early driver of Alzheimer's disease¹. The researchers reported that lithium is normally present in the brain, that amyloid pathology sequesters it (lowering available lithium), that depleting lithium in mice accelerated Alzheimer's-like pathology and memory loss, and that replacing it — notably with lithium orotate, a form that evades amyloid capture — rescued those changes at very low doses. It's a genuinely important mechanistic paper, and it gives the microdose-lithium idea its most serious scientific footing to date.

Now the honest flags, because the headlines dropped them. First, this is mouse work plus human tissue/epidemiological correlation, not a human treatment trial. The rescue experiments — the part that actually shows giving lithium back helps — were done in mice. Second, what it demonstrates is a healthspan/brain-aging signal (less Alzheimer's-like pathology), not a lifespan extension and not a demonstration that microdose lithium prevents dementia in people. "Correcting a lithium deficiency reversed Alzheimer's-like changes in mice" is a real and exciting result; "microdose lithium will keep your brain young" is an extrapolation across a species barrier that, in neurodegeneration research especially, has swallowed many promising mouse findings whole.

The supporting human signals: epidemiology, not trials

Lithium's longevity case doesn't rest only on the 2025 mouse study — there's a thread of human data, but it's observational. Population studies have repeatedly found that higher trace lithium in drinking water tracks with lower all-cause mortality and lower rates of suicide and some neurodegenerative outcomes³, and a multi-organism study reported that low-dose lithium uptake was associated with longevity in humans and extended lifespan in the worm C. elegans². These are intriguing and biologically coherent — lithium inhibits GSK-3β and promotes autophagy, mechanisms tied to several catalogued hallmarks of aging⁹.

But they're ecological and observational associations, the weakest rung on the evidence ladder: regions differ in countless ways besides their water lithium, and "places with more lithium in the water have lower mortality" cannot establish that swallowing a lithium supplement will make you live longer. (It's worth noting that a 2023 paper claiming lithium extends human lifespan from UK Biobank data was subsequently retracted — a reminder of how fragile this literature can be.) Association is not causation, and no randomized trial has tested whether microdose lithium extends human life.

What the human cognition trials actually show

Closer to a real test, lithium has been studied in small human cognition trials — with mixed, modest results. A pilot study reported that microdose lithium stabilized cognitive impairment in Alzheimer's patients over a year⁴, building on the same group's mouse work showing chronic microdose lithium prevented memory loss in an Alzheimer's mouse model⁵. More rigorously, a randomized controlled trial of low-dose lithium in older adults with amnestic mild cognitive impairment reported some disease-modifying and biomarker signals over long-term treatment⁶ — a genuine RCT, but small, in a specific at-risk population, using a clinical (not micro-) dose, and measuring cognitive and biomarker endpoints rather than lifespan. At standard psychiatric doses, large cohort studies have also linked lithium treatment to lower dementia risk in people with bipolar disorder⁷ — again observational, and in a clinical population.

Put together, the human picture is: suggestive epidemiology, a couple of small or specialized cognition trials with modest signals, and zero trials of microdose lithium for longevity or dementia prevention in healthy people. Promising direction, thin proof. We apply this same mechanism-versus-outcome lens across the field in our best longevity supplements, rated by evidence roundup, and the same caution about a single striking study in our hallmarks of aging explained primer.

The safety window — the part you can't ignore

This is where lithium differs from a benign vitamin, and it deserves a hard stop. Lithium is one of the classic narrow-therapeutic-index drugs: at psychiatric doses it requires regular blood-level monitoring precisely because the gap between therapeutic and toxic concentrations is small, and toxicity affects the kidneys, thyroid, and nervous system⁸. Long-term clinical lithium use is associated with risks to renal and thyroid function that require ongoing surveillance.

Microdose lithium (1–5 mg elemental) is far below those thresholds and is generally considered low-risk at those amounts — but "generally low-risk" is not "proven safe to self-dose indefinitely." Supplements are unregulated for dose accuracy, lithium interacts with common drugs (NSAIDs, ACE inhibitors, diuretics) that can raise blood levels into dangerous territory, and people with kidney disease, thyroid disease, or on those medications can be pushed toward toxicity by amounts that would be harmless in others. The honest bottom line on safety: microdose lithium is probably low-risk for most healthy people at genuinely tiny doses, but it is the one item in this category where casual, unmonitored self-experimentation can actually hurt you — and that warrants a real conversation with a clinician, not a purchase based on a mouse study.

The grade

Microdose lithium earns a C on longevity: a serious 2025 mechanistic paper, coherent epidemiology, and a couple of small human cognition signals — but the lifespan/dementia-prevention evidence in humans is observational or absent, the strongest causal data are in mice, and unlike most supplements it carries a real, dose-dependent safety tail. It's a genuinely interesting research direction held back from a higher grade by the same two words that recur across this site: no outcomes. We rate it against neighbors like metformin for longevity, where a repurposed drug also has mechanism and epidemiology but awaits its definitive human trial.

The bottom line

Microdose lithium is one of the more scientifically credible items in the longevity-supplement aisle and simultaneously one of the most overstated. The 2025 Nature study linking lithium deficiency to Alzheimer's is real and important — but it's mouse rescue data plus human correlation, a brain-healthspan signal, not a human lifespan result. The supporting human evidence is observational epidemiology (lower mortality where water lithium is higher) plus a few small cognition trials with modest signals, and a high-profile human-lifespan claim was retracted. Layered on top is the one thing that sets lithium apart: a narrow safety window and real drug interactions that make unmonitored self-dosing genuinely riskier than for an ordinary supplement. If you're interested in microdose lithium, treat it as a promising hypothesis to discuss with a clinician — not a proven anti-aging intervention. For where supplements like this fit alongside programs with real clinical oversight, see our graded best longevity clinics hub.