Higher Risk of Premature Death Found in Bipolar II Disorder
Key Clinical Summary
- Patients with bipolar II disorder had a 62% higher risk of all-cause mortality compared with matched controls (adjusted hazard ratio [AHR], 1.62).
- Mortality risk was elevated for both natural causes (AHR, 1.37) and markedly for unnatural causes, including suicide (AHR, 4.46).
- Bipolar II disorder was associated with higher all-cause and natural-cause mortality than bipolar I disorder, challenging assumptions of lower severity.
Bipolar II disorder (BD-II) is associated with significantly increased mortality, according to a large population-based cohort study published in JAMA Network Open. Using nationwide data from Taiwan, investigators found that patients with BD-II face elevated risk of death compared with matched controls, unaffected siblings, and individuals with bipolar I disorder (BD-I).
Study Findings
This retrospective cohort study analyzed data from Taiwan’s National Health Insurance Database spanning 2000 to 2022. The cohort included 11,427 individuals with BD-II (mean age 39.6 years; 61.9% female) matched to 45,708 controls without BD-II.
Over a mean follow-up of 7.3 years, 1089 patients with BD-II died compared with 1879 controls, corresponding to mortality rates of 13.6 versus 5.6 deaths per 1000 person-years and showing a significantly increased risk of all-cause mortality (adjusted hazard ratio [AHR], 1.62; 95% CI, 1.47–1.78).
Cause-specific analyses revealed elevated mortality from natural causes (AHR, 1.37; 95% CI, 1.23–1.52), including circulatory, respiratory, digestive, and mental health–related conditions. Unnatural causes showed a more pronounced increase (AHR, 4.46; 95% CI, 3.53–5.64), driven by suicide (AHR, 6.16), unintentional injury (AHR, 2.81), and homicide (AHR, 6.63).
Within-family analyses comparing patients with BD-II to unaffected siblings demonstrated persistent excess risk for all-cause mortality (AHR, 1.31) and unnatural deaths (AHR, 2.05), suggesting that these associations are not fully explained by shared genetic or environmental factors.
Notably, when compared with BD-I, BD-II was associated with higher all-cause mortality (AHR, 1.24) and natural-cause mortality (AHR, 1.45), while risks for unnatural death were similar between subtypes.
Clinical Implications
These findings challenge the perception of BD-II as a milder bipolar subtype and highlight its substantial mortality burden. Clinicians should recognize that BD-II carries elevated risks across multiple domains, including cardiovascular, metabolic, and psychiatric causes of death.
The markedly increased risk of unnatural death, particularly suicide, underscores the need for proactive risk assessment and targeted suicide prevention strategies. Screening for comorbid psychiatric conditions, including substance use and anxiety disorders, remains critical, although the study suggests that these comorbidities do not fully explain the mortality gap.
From a medical standpoint, excess natural-cause mortality likely reflects a combination of biological and behavioral factors, including systemic inflammation, metabolic dysfunction, and reduced engagement in physical health care. Patients with BD-II often experience prolonged depressive episodes, which may contribute to sedentary behavior, obesity, and cardiovascular risk.
Integrated care models that address both psychiatric and physical health are therefore essential. Early diagnosis and consistent monitoring may mitigate long-term risks, particularly given the documented delays in BD-II diagnosis.
Expert Commentary
“Collectively, these findings suggest the need for strengthened surveillance, targeted prevention, and proactive comprehensive psychiatric care, with emphasis on early treatment of BD-II,” concluded Chih-Wei Hsu, MD, Department of Psychiatry, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taiwan, and study co-authors.
Reference
Chih-Wei H, Yang-Chieh BC, Edward Chia-Cheng L, et al. All-cause and cause-specific mortality in patients with bipolar II disorder. JAMA Netw Open. Published online April 7, 2026. doi:10.1001/jamanetworkopen.2026.5535
