Objectives: Previous research has suggested antiviral properties for lithium, including potential effectiveness against COVID-19 in vitro. This study aimed to investigate the impact of lithium and other psychotropic drugs on the risks of mortality, hospitalization, and ICU admission due to COVID-19 among individuals with bipolar disorder. The primary objective was to assess whether lithium was beneficial in COVID-19-infection in a real-world population.

Methods: Retrospective register study using data from multiple Swedish patient registers, including 39,063 individuals in Sweden with bipolar disorder and prescribed mood stabilizers. Outcomes included COVID-19-associated death, hospitalization, and ICU admission between 11 March 2020 and 10 March 2021. Multivariate logistic regression adjusted for age, sex, and somatic comorbidities was conducted.

Results: Lithium were prescribed to 44.2 % of patients, either as mono- or combination therapy; other mood stabilizers were prescribed to 55.8 %. There were no significant associations between lithium and COVID-19-associated death, hospitalization, or ICU admission. Atypical antipsychotics were associated with increased odds ratios for COVID-19-associated death (OR 1.58 [95 % CI 1.01–2.47]), hospitalization (OR 1.80 [95 % CI 1.49–2.18]), and ICU admission (OR 2.25 [95 % CI 1.33–3.80]). Benzodiazepines were associated with a significant increase in COVID-19-associated death (OR 1.54 [95 % CI 1.01–2.35]) and hospitalization OR 1.26 [95 % CI 1.03–1.53]). In an ad hoc analysis, lithium monotherapy was, however, associated with reduced hospitalizations and ICU admissions.

Conclusions: Our findings weaken the hypothesis that lithium reduces the risk of severe events associated with COVID-19 infection in bipolar disorder.