Low-Dose Lithium Slowed Verbal Memory Decline in Older Adults - A Pilot Trial

Lithium has been used to treat bipolar disorder for more than 70 years. It is inexpensive, well-understood pharmacologically, and available as a simple daily pill. What it has not been tested for - until now - is whether it can slow the cognitive decline that marks the early stages of Alzheimer's disease.

A two-year clinical trial led by researchers at the University of Pittsburgh provides the first prospective evidence that low-dose lithium may do exactly that, at least in one specific domain. The study, published in JAMA Neurology on March 2, 2026, found that adults with mild cognitive impairment who took low-dose lithium showed a slower rate of decline on tests of verbal memory - the ability to remember and recall words and sentences - compared with those who took a placebo. The difference was not definitive by standard statistical thresholds, but the signal was consistent enough to justify a larger trial.

The Scientific Rationale Behind the Trial

The study was led by Dr. Ariel Gildengers, professor of psychiatry at Pitt and a geriatric psychiatrist at UPMC whose earlier research examined lithium's effects in aging patients with bipolar disorder. That prior work found that older adults with bipolar disorder who had taken lithium long-term showed markers of better brain integrity than those who had not - a correlation that pointed toward a possible neuroprotective effect.

"In a prior study, we observed that older adults with bipolar disorder who take lithium long-term tend to show markers of better brain integrity," Gildengers said. "The new question was whether those apparent neuroprotective effects might extend beyond mood disorders - and whether we could test that rigorously in a prospective clinical trial."

The trial enrolled adults aged 60 and older with mild cognitive impairment and randomized them to receive either a low oral dose of lithium or a placebo. Participants were followed annually for two years through detailed cognitive testing, high-resolution brain imaging, and biomarker assessments. The trial was completed in August 2024.

What the Two Years Showed

Participants in the lithium group showed slower decline on a sensitive verbal memory test than those in the placebo group. Verbal memory is one of the first cognitive functions to deteriorate in Alzheimer's disease, making it a meaningful target for an early-stage intervention.

Brain imaging told a more complex story. The hippocampus - a brain region essential for memory formation - shrank over the two-year period in both groups. The overall difference between groups did not reach statistical significance on this measure. However, exploratory analyses of participants who tested positive for amyloid beta - one of the hallmark pathological features of Alzheimer's disease - showed larger protective effects in the lithium group, pointing to a biological signal worth investigating in a properly powered follow-up study.

Importantly, the trial confirmed that low-dose lithium was safe and well tolerated in older adults when carefully monitored. This matters because therapeutic doses of lithium for bipolar disorder can cause serious side effects - kidney damage, thyroid dysfunction, and toxicity - and there had been concern about administering the drug to an aging population. The pilot results alleviate that concern at the lower doses used here.

"The key point is that lithium doesn't restore lost memory," Gildengers emphasized. "What it appears to do - if the signal holds up - is slow deterioration. That distinction matters enormously when you're designing trials and interpreting results."

A Limitation That Shaped the Results

The trial's major structural limitation was a product of its era. When the study was designed nearly a decade ago, blood-based tests for Alzheimer's pathology did not yet exist at clinical scale. Participants were enrolled based on clinical symptoms of mild cognitive impairment alone, without confirming whether they actually had amyloid pathology. Only a subset of participants turned out to be amyloid-positive, which likely diluted the trial's ability to detect effects in the population most likely to benefit from an Alzheimer's-related intervention.

"If we were designing this study today, we would enroll participants based on amyloid status from the start," Gildengers said. "That's exactly what we're planning for next."

The research team is now seeking funding for a larger, more definitive clinical trial that would use blood-based biomarkers to identify amyloid-positive participants at enrollment, recruit enough participants to reach statistical power, and extend the follow-up period beyond two years. Without that larger trial, the current pilot results remain suggestive rather than conclusive.