Association of Lithium Treatment With the Risk of Osteoporosis in Patients With Bipolar Disorder

IMPORTANCE: Osteoporosis, a systemic skeletal disorder associated with substantial morbidity and mortality, may be particularly common among individuals with bipolar disorder. Lithium, a first-line mood-stabilizing treatment for bipolar disorder, may have bone-protecting properties. OBJECTIVE: To evaluate if treatment with lithium is associated with a decrease in risk of osteoporosis among patients with bipolar disorder. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study included 22 912 adults from the Danish Psychiatric Central Research Register who received an initial diagnosis of bipolar disorder in the period from January 1, 1996, to January 1, 2019. For each patient with bipolar disorder, 5 age- and sex-matched individuals were randomly selected from the general population as reference individuals. Individuals with bipolar disorder prior to January 1, 1996, those with a diagnosis of schizophrenia or schizoaffective disorder prior to being diagnosed with bipolar disorder, and those with osteoporosis prior to the index date were excluded. Of the 114 560 reference individuals included, 300 were diagnosed with bipolar disorder during follow-up and were censored from the reference group from the date of diagnosis forward. For patients with bipolar disorder, treatment periods with lithium, antipsychotics, valproate, and lamotrigine were identified. Analyses were performed between January 2021 and January 2022. EXPOSURES: Bipolar disorder and treatment with lithium, antipsychotics, valproate, and lamotrigine. MAIN OUTCOMES AND MEASURES: The primary outcome was osteoporosis, identified via hospital diagnoses and prescribed medications. First, incidence of osteoporosis was compared between patients with bipolar disorder and reference individuals (earliest start of follow-up at age 40 years) using Cox regression. Subsequently, incidence of osteoporosis for patients receiving treatment with lithium, antipsychotics, valproate, and lamotrigine, respectively, was compared with that of patients who were not treated with these medications. RESULTS: A total of 22 912 patients with bipolar disorder (median [IQR] age, 50.4 [41.2-61.0] years; 12 967 [56.6%] women) and 114 560 reference individuals (median [IQR] age, 50.4 [41.2-61.0] years; 64 835 [56.6%] women) were followed up for 1 213 695 person-years (median [IQR], 7.68 [3.72-13.24] years). The incidence of osteoporosis per 1000 person-years was 8.70 (95% CI, 8.28-9.14) among patients and 7.90 (95% CI