People with severe mental illness, such as bipolar disorder, schizophrenia, and severe depression, die approximately 10-20 years earlier compared to the general population (Walker et al, 2015). Excess mortality is defined as the number of deaths during a period of time that are above the estimated number of deaths expected for that time period. Having bipolar disorder is associated with a two-fold increase in all-cause mortality compared to the general population (Chesney et al, 2014).

The causes of excess mortality in bipolar disorder are still largely unknown. The current study by Paljarvi and colleagues, recently published in BMJ Mental Health, aims to shed light on this issue and better understand where resource allocation should be targeted in early mortality prevention, raising awareness about safety planning and multidisciplinary approaches.

Methods

The authors conducted a cohort study using Finnish nationwide population databases to investigate the increased premature mortality between 2004 to 2018 among people with bipolar disorder and aged 15-64 years old. They broke down the specific causes of excess mortality into somatic (e.g., alcohol related causes, cardiovascular disease, cancer) and external (accidents, suicides, violence, and events of underdetermined intent) causes. Absolute mortality rates and standardised mortality ratios (SMRs) with 95% confidence intervals (CIs) were calculated from the cause of death register by Statistics Finland.

SMR is a ratio of the observed number of deaths in a population over a set-period of time to the number of deaths that would be expected over the same set-time period, if the population had the same age-specific rates as the standard population. A ratio >1.0 indicates there were more than expected deaths in the study population (in this study, bipolar disorder). An absolute mortality rate is the absolute number of deaths per year.

The authors also analysed the differences in mortality between age groups, they categorised age at death into 10-year intervals (5–24, 25–34, 35–44, 45–54 and 55–64 years).

The authors followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (von Elm et al, 2007), a checklist of 22 items specific for observational studies, that helps authors present their work in a more rigorous and high-quality way.

Results

In total, 47,018 individuals with bipolar disorder (corresponding to 377,386 person years) were included in the analysis (mean age: 38 years old; 57% women). Overall, there were 3,300 deaths (7%), mean age: 50 years; 65% men. In the comparative Finnish general population without bipolar disorder (corresponding to 52,144,411 person-years) there were 141,536 deaths.

Absolute mortality rates:

• 61% of total deaths in bipolar disorder (2,027/3,300) were somatic deaths:
  • Alcohol-related causes were the leading cause of death 29% (595/2,027);
  • Cardiovascular disease (CVD) was the second highest cause 27% (552/2,027) followed by cancer 22% (442/2,027); respiratory diseases 4% (78/2,027); diabetes mellitus 2% (41/2,027); and deaths due to substance use-related mental or behavioural disorders 1% (23/2,027).
  • 51% (1,043/2,027) of observed excess deaths in bipolar disorder were due to somatic causes:
    • Alcohol-related causes 40% (414/1,043)
    • CVD 26% (267/1,043)
    • Cancer 10% (100/1,043)
• 39% (1,273/3,300) of deaths were due to external causes:
  • Suicide was the most common cause of external cause deaths 58% (740/1,273).
  • Accidental deaths were responsible for 40% (509/1,273) of external cause of deaths.
  • 83% (1,061/1,273) of observed excess deaths in bipolar disorder were due to external causes – of them, 61% were due to suicide (651/1,061).

Standardised Mortality Ratios (SMRs):

• All-cause mortality was nearly three times higher in bipolar disorder (SMR: 2.76, CI 2.67 to 2.85)
• Mortality due to somatic causes was over two times higher in bipolar disorder (SMR: 2.06, CI 1.97 to 2.15)
  • Alcohol related causes of death were responsible for the highest relative excess mortality in the somatic causes – more than three times higher compared to the general population (SMR 3.28, CI 3.02 to 3.55)
• Mortality due to external causes was six times higher in bipolar disorder (SMR: 6.01, CI 5.98 to 6.34)
  • Suicide was the leading cause of excess mortality by external causes – over eight times higher than the general population (SMR 8.30, CI 7.71 to 8.90)

Age

The age group with the highest burden of somatic and external causes of deaths was the 45–65-year-old group, representing 81% deaths due to somatic causes and 49% deaths due to external causes.

Conclusions

This study highlights two keys points:

1. The awareness of the importance of suicide as a preventable cause of mortality in people with bipolar disorder.
2. The effect of alcohol-related causes, causing more excess mortality than cardiovascular disease and cancer among people with bipolar disorder.

These causes of deaths are potentially preventable as they could be avoided through public health and primary prevention interventions. While external causes were responsible for most of the excess deaths in the younger age groups, there was a high number of excess deaths caused by suicide across all ages, highlighting this as an important area to target preventative interventions.

Strengths and limitations

This study was carried out by an international team of renowned researchers, with expertise using population datasets. They have published widely on ways of developing scalable and evidence-based risk assessment methods that can be applied across different populations. One strength of this study is the large sample size, which allowed the authors to identify cause-specific mortality estimates, instead of previously reported all-cause mortality. As acknowledged by the authors, excess mortality estimates are affected by the general population mortality rates, so future work should replicate these findings in different populations. Another important issue is how to predict risk in routine care and stratify patients in order to tailor the intervention.

Implications for practice

Now that it is known that bipolar disorder is associated with specific causes of death, this research poses the vital questions:

• How can we enhance how we assess risk in people with bipolar disorder?
• What can we do to prevent suicide in this population of patients?

Assessing risk is crucial in bipolar disorder. The excess mortality, as shown in this work, highlights that everyone with bipolar disorder is at risk and thus safety planning as part of routine clinical care should be built into our practice. For example, in routine clinical practice, it is common that patients are given a cardiovascular risk score, indicating that everyone should have a basic level of safety planning. However, whilst assessing risk in everyone with bipolar is important, it would also be useful to be able to predict those at particularly high risk, which could help clinicians stratify and focus their management (prevention). For example, using risk prediction models to identify those at highest risk could help clinicians identify people who require extra resource. Current suicide risk is assessed by clinical judgement, but this is not transparent and likely inconsistent across mental health services.

Risk prediction tools including older previous tools (symptom checklists) were not designed to predict what happens going forward to the patient and are not recommended by NICE. The Oxford Mental Illness and Suicide tool (OxMIS) is a risk assessment tool designed to assess the risk of suicide in patients with schizophrenia-spectrum disorders or bipolar disorder, it provides a 12-month suicide risk (probability score) (Sariaslan et al, 2023). OxMIS provides a transparent consistent way of estimating risk in severe mental illness (including bipolar disorder). Ideally a risk score and recommendations for prevention could be set, and include a set of measures proven to be effective in suicide prevention.

Suicide prevention in people with bipolar disorder

• Safety planning
• Medication optimisation, e.g., ensuring Lithium is at the correct dose. Evidence suggests that among people with bipolar disorder Lithium is an effective suicide prevention strategy
• Enhanced follow-up, bringing forward the care planning meetings and having a multidisciplinary approach when addressing their needs
• More detailed assessment of their other psychosocial needs and working with carers/family to manage their risk
• Psychoeducation for patients and enhanced awareness for mental health care professionals, to enhance understanding and provide support to reduce risk factors for suicide (e.g., substance misuse).

In summary, the study by Paljarvi and colleagues has identified the age-specific causes of excess deaths, highlighting the contribution of external causes in those aged 15-64 years old. Targeting these specific causes for preventative interventions are crucial to reduce the mortality gap between the general population in bipolar disorder and should become top priorities in clinical research.

Statement of interest

Andrea Cipriani and Anneka Tomlinson work in the Department of Psychiatry, University of Oxford where the senior author (Professor Seena Fazel) of this manuscript also works; however, this blog was drafted independently by the two authors.

Links

Primary paper

Paljärvi T, Herttua K, Taipale H, et al. Cause-specific excess mortality after first diagnosis of bipolar disorder: population-based cohort study. BMJ Mental Health 2023;26:e300700. http://dx.doi.org/10.1136/bmjment-2023-300700

Other references

Walker ER, McGee RE, Druss BG. (2015) Mortality in mental disorders and global disease burden implications: a systematic review and meta- analysis. JAMA Psychiatry 2015;72:334–41.

Chesney E, Goodwin GM, Fazel S. (2014) Risks of all‐cause and suicide mortality in mental disorders: a meta‐review. World Psychiatry 2014;13:153‐60

von Elm E, Altman DG, Egger M, et al. (2007) Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ 2007;335:806–8.

Sariaslan, A., Fanshawe, T., Pitkänen, J. et al. (2023) Predicting suicide risk in 137,112 people with severe mental illness in Finland: external validation of the Oxford Mental Illness and Suicide tool (OxMIS). Transl Psychiatry 13, 126 (2023). https://doi.org/10.1038/s41398-023-02422-5

Photo credits

• Photo by Steve Johnson on Unsplash
• Photo by Kenny Eliason on Unsplash
• Photo by Scott Graham on Unsplash
• Photo by Jungwoo Hong on Unsplash
• Photo by Rebecca Peterson-Hall on Unsplash
• Photo by Mike Enerio on Unsplash
• Photo by LinkedIn Sales Solutions on Unsplash