For decades, dementia research has fixated on single-target therapies — clearing amyloid plaques or silencing tau tangles — with disappointing clinical results. A sweeping narrative review in JAMA Psychiatry now consolidates 25 years of evidence suggesting that lithium, a drug psychiatry has used since the 1950s, may offer something those targeted therapies cannot: simultaneous engagement of multiple neuronal survival pathways that collectively defend the aging brain.
The review synthesizes mechanistic, neuroimaging, preclinical, and epidemiological data to characterize lithium's neuroprotective profile. At the molecular level, lithium inhibits glycogen synthase kinase-3β (GSK-3β), a kinase critically implicated in tau hyperphosphorylation and amyloid precursor protein processing. Concurrently, it upregulates Bcl-2 anti-apoptotic signaling, amplifies brain-derived neurotrophic factor (BDNF) activity, stabilizes mitochondrial membranes, and attenuates oxidative stress cascades — a convergent multi-target action that no approved Alzheimer's therapy replicates. Crucially, robust neuroprotective effects appear at approximately 0.3 mM serum concentrations, roughly half the threshold used in psychiatric practice, potentially sidestepping the renal and thyroid toxicity concerns that have historically limited lithium's broader application. Human neuroimaging corroborates the biology: proton MRS studies document lithium-associated increases in N-acetylaspartate (a neuronal viability marker), while structural MRI data show preserved or partially restored gray matter volume in hippocampal and corticolimbic regions.
This review matters because it reframes lithium not as a repurposed psychiatric drug but as a biologically coherent neuroprotective agent with a quarter-century of converging evidence. The field should note meaningful limitations: narrative reviews carry inherent selection bias, most neuroimaging studies involved relatively small samples, and translating preclinical neurogenesis findings to humans remains uncertain. Randomized controlled trial data in MCI and Alzheimer's disease specifically remain limited, and long-term safety at sub-therapeutic doses in older adults is incompletely characterized. Still, the mechanistic plausibility here is unusually strong — GSK-3β inhibition alone intersects with both major Alzheimer's proteinopathies. This warrants designation as a potentially paradigm-shifting perspective: not a single breakthrough study, but a rigorous synthesis that could redirect clinical trial priorities toward a low-cost, multi-target intervention.