REVIEW OF: Chen S et al, PLoS Medicine 2022;19(3):e1003941

STUDY TYPE: Retrospective cohort study

Low-dose lithium has recently been proposed as a potential treatment for dementia, based on a couple of retrospective studies (Gerhard T et al, Br J Psychiatry 2015;207(1):46–51; Kessing LV et al, Bipolar Disord 2010;12(1):87–94). This newest study looks at evidence that standard doses of lithium might also be helpful in preventing the development of dementia.

The study compared the electronic records of 29,618 adults in the UK age 50 years and older without diagnoses of mild cognitive impairment (MCI) or dementia. The 548 patients exposed to lithium had a greater number of cardiovascular risk factors, including antipsychotic use, central vascular disease, diabetes, hyperlipidemia, hypertension, and smoking. Of those prescribed lithium, 73% had a mania/bipolar affective disorder (BPAD) diagnosis, a known risk factor for dementia, while the remaining lithium-exposed patients were diagnosed with depression.

In the lithium-exposed cohort, 9.7% of patients were diagnosed with dementia, compared to 11.2% of the unexposed cohort. The percentage of patients diagnosed with Alzheimer’s disease was significantly lower in the lithium-exposed cohort (6.8%) compared to the unexposed cohort (8.1%), whereas the percentage of vascular dementia was the same in both cohorts (2.6%). Lithium appeared to lower dementia risk for the group taking lithium for one year or less and the group taking it for more than five years (the smaller cohort of patients taking lithium for two to four years saw no benefits, likely due to a lack of statistical power). Documented serum lithium levels were 0.49–0.80 mEq/L, with a mean of 0.61 mEq/L.

Method of diagnosis, dosing, follow-up, and adherence were highly variable or unknown. Some lithium level data were missing, so patients with supratherapeutic levels may have experienced neurotoxic effects, which can mimic dementia.

Carlat Take

We all want something to offer patients with risk factors for dementia or MCI, without peddling false hope. This study suggests that lithium might delay or prevent Alzheimer’s disease in individuals with BPAD, but the real-world conditions are limited. The variables are too numerous to contain even in this careful analysis. However, it is reasonable to discuss possible advantages of lithium in terms of dementia risk with patients already taking lithium, particularly in patients who stick with lithium long term.