Lithium's Second Act: The Mood Stabiliser That Might Defend the Aging Brain

#1

A high-profile narrative review argues that lithium — already 75 years old as a psychiatric drug — works by tuning fundamental cell-survival machinery rather than any single neurotransmitter, and that very low (“micro”) doses, especially as lithium orotate, deserve a formal trial as a disease-modifying therapy for mild cognitive impairment (MCI) and Alzheimer disease.

For most of a century, lithium has been filed under “mood stabiliser” — effective, cantankerous, and narrow. This review, led by Gregory Moore and Husseini Manji (figures central to the original 1990s–2000s lithium-neurotrophic story), makes the case that this framing was always too small. Their thesis: lithium doesn’t fix one broken signal; it adjusts the upstream switches that decide whether a stressed neuron repairs itself or dies.

The molecular cast is familiar to anyone tracking neuroprotection. Chronic lithium raises Bcl-2 (an anti-apoptotic protein), boosts BDNF (a growth factor for neurons), and inhibits GSK-3β (an enzyme that, left unchecked, drives tau phosphorylation and amyloid production). Downstream, mitochondria are stabilised, calcium handling improves, and oxidative stress falls. Crucially, these are the same pathways that fail in Alzheimer disease — so the overlap is mechanistic, not coincidental.

The review’s strongest card is human imaging. Spectroscopy studies show lithium raising N-acetylaspartate, a marker of neuronal health, while MRI studies show gray-matter preservation — even reversal of atrophy — in bipolar patients. That lithium can measurably reshape brain structure in living people is the most durable part of the story.

The genuinely new material comes from a 2025 Nature paper (Aron et al.) proposing that amyloid plaques act as negatively charged “ionic sinks,” sequestering lithium and starving nearby tissue of it — a localised deficiency despite normal blood levels. In mice, restoring lithium reversed pathology; a specific salt, lithium orotate, appeared to evade plaque-trapping and reach brain tissue better.

The authors’ conclusion is deliberately concrete: run a proper randomised trial of low-dose lithium orotate in MCI, betting on decades-old neuroprotective biology rather than the shakier “trace-element repletion” idea. If it works, lithium would be that rare thing — a cheap, off-patent, mechanistically grounded brain therapy, with obvious relevance to lower-income countries. The caveats are equally deliberate: the orotate and deficiency findings rest heavily on one paper and await independent replication.

Actionable Insights

This is a hypothesis-generating review, and almost none of it has been prospectively validated in humans at the doses being hyped. The honest take-home messages:

  • The biology supporting lithium’s neuroprotection is robust; the orotate-specific and microdose claims are not. Effect sizes that the review actually cites are mostly preclinical: roughly a 25% increase in new hippocampal neurons in rodents, and a ~3% increase in total gray matter at 4 weeks in a small bipolar imaging study (Moore 2000, Lancet). A ~3% structural change is real but modest, and was measured in patients, not healthy people.
  • Epidemiology suggests a risk reduction, but the review does not quantify it. The Danish drinking-water and registry signals (Kessing 2017) are associative and confounded; treating “lower dementia in high-lithium water regions” as an effect size would be unjustified.
  • Clinical trial reality is mixed. The 2026 LATTICE pilot missed several primary endpoints and salvaged a verbal-memory signal (CVLT-II) as a secondary finding. That is a weak, not a strong, result.
  • Practical magnitude for an individual: unknown. No human RCT has yet shown that 1–5 mg/day elemental lithium (the orotate supplement range) slows cognitive decline. Anyone self-experimenting is acting ahead of the evidence, and lithium — even low-dose — still warrants thyroid and renal awareness.

Source:

  • Open Access Paper: The 25-Year Evolution of Lithium as a Disease-Modifying Agent in Dementia: A Narrative Review, 2026 Jun 10
  • Institutions: Penn State College of Medicine; Alzheimer’s Disease Data Initiative and Gates Ventures (Kirkland, WA); NIMH/NIH; Oxford University; Yale University.
  • Country: United States (with UK co-author affiliation).
  • Journal: JAMA Psychiatry (narrative review).
  • Impact Evaluation: The impact score of this journal is 27.7 (Scopus CiteScore; Clarivate JCR Impact Factor reported ~18 in the latest cycle), therefore this is a High-impact journal.

Related Reading: