‘Wish to die’ is independently associated with cardiovascular mortality in later life. Data from TILDA

There is an established bidirectional relationship between mental and heart health in later life but the link between wish to die (WTD) and cardiovascular mortality is less well‐defined.

The aim of this study therefore is to examine the impact of WTD on all-cause mortality in a large population-based sample of older adults, adjusting for important factors such as cardiovascular disease, mental health and other chronic diseases. Furthermore, we will examine the impact of WTD specifically on cardiovascular mortality in those with and without prior heart disease. Our hypothesis, given the well-established links between heart disease and mental health, was that WTD would be significantly associated with cardiovascular mortality.

| Study design
This is a longitudinal study examining the association between WTD and subsequent mortality over 9-year follow-up in a large population-representative sample of older adults. Individual-level survey data was linked to official death registration data.
The TILDA study design has been outlined previously. 18 Briefly, participants were interviewed at 2-yearly intervals (Waves 1-5). There were three components to data collection: a computer-assisted personal interview carried out by social interviewers in the participants' own home; a self-completion questionnaire completed and returned by the participant; and a comprehensive centre-based health assessment or a modified home-based health assessment carried out by trained research nurses. We analysed data from Waves 1 to 5, collected between 2009 and 2018.
Participants were included in this study if they were aged 50 years or older at Wave 1, and underwent assessment for WTD.
Participants were excluded from participation at Wave 1 if they had a pre-existing diagnosis of dementia.

| Wish to die
At Wave 1, participants were asked: 'In the last month, have you felt that you would rather be dead?' Participants who answered affirmatively were defined as having WTD.

| Mental health
Participants were also asked specifically about a prior diagnosis of depression, anxiety disorder or other psychological/mental illness (psychosis, bipolar affective disorder, schizophrenia).
Participants were also screened for depressive symptoms using the Centre for Epidemiological Studies Depression Scale (CES-D) 19 and for anxiety symptoms using the Hospital Anxiety and Depression Scale (HADS). 20 A CES-D Score ≥ 16 defined clinically significant depressive symptoms, 19 while a HADS-A Score >10 and >14 defined moderate and severe anxiety symptoms, respectively. 20 When adjusting for depression we used a combination of either a prior doctor's diagnosis and/or meeting criteria for clinically signifi-

| Mortality
In order to compile mortality data, death records were obtained for TILDA participants and linked to individual level survey data from the study. 21 Every death in the Republic of Ireland is registered with the General Register Office (GRO), and TILDA was granted approval from the GRO to link respondents to their corresponding death certificate information. The underlying cause of death was operationalised according to WHO definition as 'the disease or injury which initiated the train of morbid events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury'. 22 Death records were available for 775 participants after exclusion criteria were applied and represented a mortality rate of 9% over 9-year follow-up. Average age at death was 75 years.

| Covariates
Further detailed social and biological data were also collected at Wave 1.
Educational attainment and smoking history were obtained by self-report. The CAGE alcohol scale was used to assess for excess alcohol intake. Cardiovascular disease was defined as self-report of angina, congestive cardiac failure or prior myocardial infarction.  Based on the number of the following chronic diseases: eye disease (cataracts, glaucoma or age-related macular degeneration), lung disease, cancer, osteoporosis, liver disease, arthritis, urinary incontinence, Parkinson's disease and diabetes.

| Ethics
The TILDA study was approved by the Faculty of Health Sciences Research Ethics Committee at Trinity College Dublin and all participants gave informed written consent. All experimental procedures adhered to the Declaration of Helsinki.   Table 2 shows data from regression models estimating the hazard ratios for mortality related to WTD. WTD was significantly associated with overall mortality in fully adjusted models with a hazard ratio of 1.41 (95% CI: 1.00-1.99). After exclusion of participants with a history of cardiovascular disease (n = 1132) or depression, anxiety or other psychiatric illness (n = 649) however, this association was attenuated and no longer statistically significant.

| RESULTS
In fully adjusted models, WTD was significantly associated with cardiovascular mortality, with a hazard ratio of 2.14 (95% CI: 1.21-3.79). Findings remained robust after excluding participants with a history of depression, anxiety or other psychiatric illness. However, when participants with a history of cardiovascular disease were excluded, findings were attenuated and no longer significant.
WTD was not associated with an increased risk of death due to noncardiovascular causes.

| DISCUSSION
This study demonstrates that over 3% of older people report WTD, endorsing that during the last month they have had thoughts that they would rather be dead. Participants with WTD had a 4-5-fold increased likelihood of coexisting depression and anxiety, based on a prior doctor's diagnosis and/or meeting criteria for depression/ anxiety on validated scales. In line with prior studies, WTD was also associated with higher burden of chronic medical illness and alcohol misuse. 23,24 After adjusting for covariates, WTD at baseline was associated with a greater than 40% absolute higher likelihood of all-cause mortality at 9-year follow-up. When participants with coexisting cardiovascular disease or psychiatric illnesses were excluded, this independent association was no longer significant. When examined by cause of death, however, WTD was associated with more than double the likelihood of mortality specifically due to cardiovascular disease independently, including when participants with a psychiatric illness were excluded from the analyses. WTD was no longer associated with cardiovascular mortality when participants with a history of heart disease were excluded however. WTD was not associated with noncardiovascular mortality.
A prior study involving a cohort of community-dwelling older people from Australia demonstrated increased risk of death, specifically death due to cardiovascular disease, in participants with SI but did not analyse those reporting WTD. 25 WTD has been shown to increase the risk of 5-year mortality in a cohort of older people attending primary care, however, analysis was adjusted for baseline disability, smoking status and Hamilton Depression Score only and did not examine cases by cause of death. 26 WTD in later life is more prevalent in those with heart disease, 27 and older adults with a history of myocardial infarction are three times more likely to endorse WTD. 28 Death by suicide also correlates strongly with ischaemic heart disease 29 and incident cardiovascular disease modifies the risk of all-cause mortality in older people with depressive symptoms. 30 To our knowledge this is the first study to specifically examine the association between WTD and cardiovascular mortality in later life however.
T A B L E 2 Hazard ratios with 95% confidence intervals for association between wish to die and mortality