Low-dose lithium may protect brain networks in early Alzheimer’s disease

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Jul 21 2026

Researchers outline how low-dose lithium could be tested as a carefully monitored approach to preserving cholinergic function and brain networks before Alzheimer’s disease becomes advanced.

Perspective: Revisiting the cognitive potential of lithium in alzheimer’s disease: the role of cholinergic system and brain connectivity. Image Credit: Nanzeeba / Shutterstock

In a new perspective available as an 'Article in Press' in the journal Translational Psychiatry, researchers discuss the cognitive potential of lithium in Alzheimer’s disease (AD).

A recent Nature study reported substantially reduced levels of endogenous lithium in the brain in individuals with AD and mild cognitive impairment (MCI). It also showed that depleting dietary lithium accelerated tau phosphorylation, amyloid β (Aβ) deposition, myelin and synapse loss, and cognitive decline in mouse models.

In transgenic AD mice, lithium orotate replacement treatment ameliorated cognitive deficits and neuropathology, reinvigorating interest in physiological lithium restoration in AD. The authors of the present study speculate that lithium's interactions with the cholinergic system and its effects on brain connectivity may help explain the encouraging preclinical findings reported in AD models.

Cholinergic system and lithium

In 1980, lithium carbonate was shown to inhibit serum cholinesterase activity in individuals with endogenous depression. The cholinergic hypothesis of AD has since been clinically validated, with several cholinesterase inhibitors, such as galantamine, donepezil, and rivastigmine, improving function and cognition in mild to moderate AD. However, lithium's reported effect on serum cholinesterase activity has remained largely overlooked.

Given the known benefit of cholinesterase inhibition in AD and the role of acetylcholine in cognition, the authors posit that lithium's effects on serum cholinesterase may contribute to cognitive protection observed in mouse models. Previously, the authors investigated the effects of lithium supplementation in patients with AD in a pilot study. Although no significant cognitive improvement was observed, several clinical and biochemical effects were noted.

Notably, lithium was associated with substantial increases in red blood cell (RBC) glycine and choline levels. Because some small metabolites show equilibrium between the brain and RBCs, the authors hypothesized that these peripheral increases were accompanied by parallel changes in brain glycine and choline. Preclinical evidence suggests that increased brain choline supports cholinergic tone, whereas glycine modulates glutamatergic signaling and may exert neuroprotective effects.

Effects of lithium on brain connectivity

Functional magnetic resonance imaging (fMRI) studies of the human brain in bipolar disorder (BD) suggest that lithium can alter temporal changes in network-level communication. Lithium monotherapy in patients with BD was associated with transitions in mania-related indices toward healthier patterns. In addition, changes in network metrics were proportional to changes in symptoms.

Another study reported that longer lithium exposure in older adults with BD was associated with greater white matter integrity. While clinical trials have investigated lithium in AD, brain-connectivity outcomes have not been reported. Nevertheless, in studies outside AD, lithium has often been associated with metrics concordant with the preservation of brain structural integrity. For instance, lithium treatment was associated with increased volumes of white and gray matter in healthy volunteers.

Greater white matter volume and regional gray matter volume can support local circuit processing and tract conduction underpinning network organization. Further, in older adults with BD, longer lithium exposure was associated with diffusion measures of white matter integrity that typically worsen in dementia, after accounting for vascular disease burden. Associations of lithium with greater white matter integrity in this group reinforce the notion that lithium may help preserve structural connectivity in older people.

Derisked trial strategies

The pathobiology of AD begins years before dementia. Modern biomarker models define preclinical and prodromal phases with detectable network abnormalities and amyloid, synaptic, and tau dysfunction, but relatively limited irreversible disconnection. At this stage, the authors argue that interventions intended to stabilize brain networks may be most effective. Trials of low-dose lithium in MCI suggest that this approach is tolerable and feasible in older individuals.

As such, future trials should initiate with low doses, defining serum targets below BD ranges, slow titration, thyroid and renal monitoring, hydration counseling, and drug-interaction review. In addition, trials should include systematic evaluation of tremor, gastrointestinal symptoms, confusion, extrapyramidal signs, cardiac conduction abnormalities, and worsening gait. Overall, the goal is meticulous investigation of low- or physiological-dose replacement with target-engagement measures.

Concluding remarks

In sum, the perspective argues that evidence from human brain samples and preclinical models suggests endogenous lithium can modulate brain aging and AD pathology. A deeper understanding of its cognitive effects in AD warrants incorporating its cholinergic properties and ability to alter brain connectivity. The authors suggest that the association between lithium exposure and white matter integrity, observed chiefly outside AD, may partly reflect the maintenance of cholinergic tract integrity. Future studies should evaluate metabolic changes and cholinergic endpoints, with a focus on safety and early intervention.