Table 1 Summary of the 40 selected articles for the systematic review

Suicide among caregivers of people with dementia

 Hosaka and Sugiyama 2003 [20]

Intervention study

N = 20, all females from 47 to 66 years, 10 caretakers of vascular dementia and 8 caretakers of Alzheimer’s disease

NA

No

To investigate the effects of a 5-week structured group intervention on the immune function of caregivers of dementia patients.

POMS- to measure mood disteubance andGHQ-30- to assess other comorbidity

(1) Pre-intervention mean score for suicidal depression domain in GHQ-30 was 1.10 (SD = 1.48) and post-intervention at 0.85 (SD = 2.85).

(2) Other findings: Depression mean score (POMS) was 16.1 (SD = 23.0) pre-intervention and 12.1 (SD = 12.6) at post-intervention.

(3) 11 out of 20 subjects had no social support.

(1) Small sample size.

(2) Subjects recruited from one centre.

 Shaji et al. 2003 [21]

Qualitative study

N = 17, 76% were females

NA

No

To assess the range of care arrangements, attitudes towards caregiving roles and sources of strain among caregivers of Alzheimer’s disease patients

Use of topic guide

(1) Depressed mood was reported by 16 (94%) of caregivers. Five caregivers had suicidal ideation and one had made an attempt. The caregiver who committed suicide had major depressive disorder.

(2) Sources of caregiver strain: impairment in basic activities such as eating, dressing, bathing, and maintaining personal hygiene; incontinence; behavioral and psychological symptoms of dementia (BPSD).

(3) Those caregivers who received help from others felt less stressed and appeared to be coping better. Better financial status, and other helpful adult women caregivers in the family were clearly helpful factors.

(1) Study did not consider saturation point for subject recruitment.

 Valente et al. 2011 [17]

Cross-sectional study

N = 137, 80.3% females

Convenient sampling

No

To investigate caregivers of dementia patients perceived health and to look into relationships with patients and caregivers’ sociodemographic and clinical data.

BAI, BDI, ZBI, and MBI

(1) 8.8% (n = 12) had death wishes.

(2) Higher burden of care, severity of anxiety symptoms, severity of depression symptoms, emotional burnout, and depersonalization lead to higher odds of emotional problem.

(1) Cross-sectional design, no causal inference.

(2) Non-probability sampling used.

(3) Small sample size.

(4) Sample recruited only from one center.

 O’Dwyer et al. 2012 [22]

Online cross-sectional survey

N = 120, 89.2% female

NA

No

To gather preliminary evidence on suicidal ideation infamily carers of people with dementia.

RMBPC, FCSES, ADKS, SF-12 II, CESDS, BHS, GAI, LOT, ZBS, Brief COPE, DSSI, SBQ-R

(1) 26% of carers had contemplated suicide more than once in the previous year. Only half of these had ever told someone they might commit suicide and almost 30% said they were likely to attempt suicide in the future.

(2) Carers who had contemplated suicide had poorer mental health, lower self-efficacy for community support service use and greater use of dysfunctional coping strategies.

(3) When all factors were controlled, only higher severity of depression predicted presence of suicidal thoughts.

(1) Small sample size.

(2) Cross-sectional study.

(3) Sampling method used was convenient sampling.

(4) Online survey limits generalizability of research findings.

 O’Dwyer et al. 2013 [23]

Qualitative study

N = 9, 4 males and 5 females; 55.6% females

NA

No

To conduct an initial exploration of carers of dementia patients experiences of suicidality and identify factors associated with risk and resilience

A semi-structured interview guide. The questions focused on experiences and challenges of caring, participant approaches to managing stress and maintaining wellbeing, and experiences of suicidal ideation or suicide attempts.

(1) Three themes were identified in the data – ‘experiences of suicidal ideation’, ‘risk factors’ and ‘resilience’. Four of the nine participants had experienced suicidal thoughts and two had made preparations for a suicidal act.

(2) Risk factors included pre-existing mental health problems, physical health conditions, and conflict with other family or care staff.

(3) Factors positively associated with resilience included the use of positive coping strategies, faith, social support and personal characteristics.

(1) Study did not consider saturation point for subject recruitment.

(2) Participants self-selected.

 Lewis, 2015 [18]

Qualitative study

N = 101, 87% females (those who love one with dementia had passed away more than 10 years)

Purposive sampling

No

To discover a substantive theory that identifies the main problem that caregivers of loved ones with dementia face at the end of life and the basic social process by which they resolve that problem.

Grounded theory transcends description of data to conceptualize ideas that are substantive.

(1) Caregivers faced a concern of being trapped in an inescapable role. They felt bound to loved ones emotionally, mentally, and often physically.

(2) Caregivers attempt to resolve this problem through a 5-stage basic social psychological process of rediscovering: (i) missing the past (many caregivers described their years of caregiving as “saying a long goodbye.”), (ii) sacrificing self (caregivers devote themselves to trying to minimize losses and control the “downward spiral” of their loved ones and ultimately sacrifice themselves), (iii) yearning for escape (ambiguity of prognosis toward the end-of-life period led to caregivers guessing when the end was near and seeing “no end in sight, death of care receiver seems to be the only way out), (iv) reclaiming identity (after caregivers reached the point of needing escape, they strategized ways to sustain themselves), and (v) finding joy (as a consequence of reclaiming themselves, caregivers were able to find true joy in their roles).

(1) Lack of diversity in demographic background.

(2) Use of published memoirs as data may skewed the findings.

 Koyama et al. 2017 [8]

Case-control study

N = 104, matched for age and gender, 58.7% females

NA

No

To compare the mental health of dementia caregivers with that of community residents and to clarify factors related to mental health problems in younger and older caregivers.

CESDS, SF-8, NPI, PSMS, LIADLS

(1) Both younger and older caregivers had significantly worse mental QOL than community residents, but were not more depressed.

(2) Sleep problems were significantly more frequent in younger caregivers (39.1%) than in community residents.

(3) Caregivers’ deteriorated mental QOL was associated with patients’ BPSD in younger caregivers and with dementia patients’ instrumental ADL and female gender in older caregivers.

(1) Small sample size.

(2) Cross-sectional design.

(3) Use different questionnaires to assess mental state of older and younger caregivers.

(4) Sample recruited from a single center.

 Joling et al. 2018 [24]

Longitudinal study

N = 192, 70.3% females, divided into three groups (suicidal thought, no suicidal thought, not assess)

NA

No

To explore thoughts of suicide, self-harm and death in dementia caregivers and investigates the characteristics that distinguish them from those without such thoughts.

MINI, CESDS, HADS-A, CRA, PMS, SSCQ

(1) Within 24 months, 76 caregivers reported symptoms of a potential depression and were further assessed for suicidal thoughts.

(2) Nine carers (11.8%, 4.7% of the total sample) reported suicidal thoughts with three of those at multiple points.

(3) Caregivers with suicidal thoughts had more severe depressive and anxious symptoms, had a lower sense of competence and mastery, felt less happy and experienced more health problems, less family support and more feelings of loneliness than caregivers who had not.

(1) thoughts of suicide, thoughts of self-harm and thoughts of death were grouped together.

(2) Age and education differ between the groups of subjects.

 Anderson et al. 2019 [25]

Qualitative study

N = 9 blogs

NA

No

To analyze a sample of blogs written by family caregivers of people with Alzheimer’s disease and related dementia to explore thoughts of suicide and homicide expressed by these caregivers.

Transcripts were analyzed in chronological order by two authors using codes created from the verbatim words used by the bloggers.

(1) Five themes related to thoughts of suicide and homicide by caregivers and people with ADRD were derived from the analysis: (i) end-of-life care (majority of caregivers wrote about what they described as “the long good bye”); (ii) thoughts of death and euthanasia by the person with ADRD (caregivers documented the person with ADRD’s thoughts and reflections on their own impending death); (iii) surrogate decision making (caregivers often wrote about concerns surrounding surrogate decision making); (iv) thoughts of suicide by the caregiver (caregivers expressed thoughts related to their own death); and (v) thoughts of homicide and euthanasia by the caregiver (caregivers in this sample of bloggers also wished for the care recipient’s death).

(1) May missed those who did not have access to internet.

(2) May missed those who referred themselves as “carers” rather than “caregiver” in the blogs.

 Joling et al. 2019 [26]

Longitudinal study

N = 6646 (first wave), 1582 (second wave and third wave); informal caregivers, 70.3% females

NA

No

To compare suicidal thoughts between non-caregivers and informal caregivers of people with a variety of conditions, in a large representative sample, and to identify significant risk factors.

CIDI version 3.0 (suicidal module)

(1) Thirty-six informal caregivers (2.9%) reported suicidal thoughts during the 4 year study period.

(2) The difference between caregivers and non-caregivers (3.0%) was not significant.

(3) Among caregivers, significant risk factors for suicidal thoughts included being unemployed, living without a partner, having lower levels of social support, having a chronic physical disorder, a mood disorder or an anxiety disorder, and having impaired social, physical and emotional functioning. These risk factors were also found in non-caregivers.

(4) No caregiving-related characteristics were associated with suicidal thoughts.

(1) Small number of participants with suicidal thought, no multivariate analysis could be performed.

(2) Participants asked to recall retrospective history of suicidal thought leading to recall bias.

 Kim et al. 2019 [27]

Qualitative study

N = 18, Korean American caregivers, 83.3% females

NA

No

(1) To explore the caregiving experience of KA families of PWD and to understand how KA caregivers of PWD try to fulfill the gaps between their needs and available healthcare services for dementia care in the U.S.

(2) To utilize this needs assessment for developing a community-based, caregiver-centered, and culturally appropriate dementia care education series for the KA community.

Semi-structured interview

(1) Four themes were identified: (i) challenges in finding resources (efforts to search for helpful resources, which were affected by multiple factors such as English proficiency, health insurance, financial status, knowledge of dementia, and attitude towards the illness), (ii) struggling with mental health issues (KA family caregivers’ struggle with several challenges related to dementia symptoms, their own emotions, health management, and family dynamics), (iii) traveling the path of acceptance (several caregivers discussed coming to terms with a new reality and putting efforts into providing the best possible care in the areas like nutrition and diet, exercise and activities, as well as communication strategies with compassion and love), and (iv) finding ways to survive (most caregivers reported self-care strategies such as walking, healthy eating, rest and sleep, medical check-ups, and hobbies).

(1) Recruitment in geographically limited location.

 Rosato et al. 2019 [28]

Cross-sectional study

N = 1,018,000 people aged 25–74 years (130,816 caregivers; 110,467volunteers; and 42,099 engaged in both), not specific to dementia caregiver, 52.3% to 61.2% females

NA

No

(1) To compare the prevalence of poor mental health amongst volunteers and caregivers after adjustment for demographic and socio-economic factors;

(2) To measure the risk of suicide amongst caregivers and volunteers, controlling for baseline health status and possible health selection effects; and

(3) To determine if these prosocial activities reduce suicide risk for those with poor mental health.

Both definite suicides and deaths of undetermined intent were combined to define suicide. Sensitivity analyses were undertaken using just definite suicides.

(1) Intense caregiving was associated with worse mental health and volunteering with better mental health.

(2) For those engaged in both activities, likelihood of poor mental health was determined by caregiving level.

(3) There were 528 suicides during follow-up, with those engaged in both activities having the lowest risk of suicide. Engaging in either volunteering or caregiving was associated with lower suicide risk for those with good mental health at baseline but not for their peers with baseline poor mental health.

(1) Ethnicity background was limited to white.

Suicide among people with dementia

 Lyness et al. 1992 [29]

Cross-sectional study

N = 160, age 60 and above, 73.6% females

NA

No

To describe the psychopathological characteristics of elderly suicide attempters admitted to an inpatient psychiatric unit.

DSM-III

(1) Eighty percent of the attempters had a major depressive syndrome;

(2) Dementia patients did not contribute to suicidal attempts.

(1) Cross sectional design.

(2) Sample recruited from a single center.

 Florio et al. 1997 [30]

Cross-sectional study

N = 683, 66.9% females, elderly patients referred to community-based aging and mental health service

NA

No

To determine whether elderly patients referred to community-based aging and mental health service who judged to be at risk for suicide differed from those persons judged not to be at suicide risk.

DSM-III, self-deisgned questionnaire

(1) Only 8% of subjects with suicidal risk had dementia. Dementia did not contribute to suicidal risk among the elderly.

(1) Self-designed questionnaire was used, except DSM III.

(2) Subjects recruited from single location.

(3) Cross-sectional design of the study.

 Rao et al. 1997 [31]

Cross-sectional study

N = 118, all of whom were a cohort in a pre-existing epidemiological study of dementia who were community residents, 72% females

NA

No

To study the relationship between suicidal thinking and both cognitive impairment and depression.

CAMDEX, GDS, SSI

(1) Those with suicidal thinking showed higher CAMDEX depression scores, weaker strength of the wish to go on living, higher rates of expressing wish to die and higher rates of depressive illness and mixed DAT/multi-infarct dementia as primary psychiatric diagnoses.

(2) No signi®cant associations between suicidal thinking and GDS scores, Alzheimer-type dementia alone, awareness of memory difficulties or severity of dementia.

(1) Small sample size.

(2) Cross-sectional design.

 Rubio et al. 2001 [32]

Retrospective case-control study

N = 28 elderly with completed suicide (case), 56 elderly who died naturally (control), age and gender matched; 39.3% females

NA

No

To determine if Alzheimer’s disease changes are overrepresented in elderly people committing suicide.

A modified Braak scoring system and semiquantitativeassessment of neurofibrillary tangles, amyloid deposition, Lewy bodies, and Lewy-associated neurites.

(1) The brains of individuals who committed suicide had higher modified Braak scores than those of matching control subjects.

(2) The number of neurofibrillary tangles in CA1 was not an independent predictor of suicide status in the statistical analysis, although the distribution was more highly skewed among the cases. Hence, not indicative of dementia as cause of suicide.

(1) Small sample size.

(2) Sample size of cases and controls not the same.

 Draper et al. 2003 [33]

Cross-sectional study

N = 593 residents in 10 nursing home, 73.2% females

NA

No

To determine whether indirect self-destructive behaviors predict mortality in nursing home residents.

HBS, BEHAVE-AD, FASS, RCI, CIRS, EBASD, and the suicide item from the HRDS.

(1) Mortality was predicted by older age, male gender, lower level of functioning, lower levels of behavioral disturbance on the BEHAVE-AD, and higher scores on the HBS “passive selfharm” factor-based subscale, which includes refusal to eat, drink, or take medication.

(2) Risk taking, active self-harm, and passive self-harm were postively correlated with behavioral pathology in Alzheimer’s disease.

(1) Lack of direct observational data

(2) Cross-sectional design.

 Heun et al. 2003 [34]

Case-control study

N = 67 Alzheimer’s disease patients, 109 elderly from general population, and 189 siblings; 62–82% females

NA

No

To compare the presence and symptomatology of depression between Alzheimer’s disease patient and age-matched non-demented subjects.

CIDI, MMSE

(1) Lifetime depressive symptoms were significantly more frequent in 76 AD patients than in 109 age-matched elderly from the general population. These 76 AD patients complained more about thinking and concentration disturbances, and less about depressed mood or appetite disturbance than the 298 non-demented participants matched for the lifetime presence of major depression (MD).

(2) In agreement, the 29 patients comorbid for lifetime diagnoses of AD and MD reported less about depressed mood than the 114 age-matched elderly with MD only.

(3) Feelings of worthlessness and suicidal ideas were related to the severity of cognitive decline.

(1) Since demented patients were recruited from a clinical population, there might be an overestimation of the prevalence of depressive symptoms compared to patients with AD in general population.

(2) Analysis was based on retrospectively given information by demented and non-demented elderly.

 Peisah et al. 2007 [35]

Case-control study

N = 143 community-dwelling suicide victims aged 65 years or more and 59 motor vehicle accident victims autopsies; 30.7% females

NA

No

To investigate prevalence of AD-related pathology in older suicide victims.

Senile plaques (diffuse and neuritic) a modified Bielschowsky (Garvey) silver technique, modified Braak score, plaque density was rated using the CERAD criteria, neuropathological rating was performed blind to the subjects’ clinical status.

(1) There were no significant differences in plaque score or neurofibrillary tangle staging between suicide and control groups.

(2) None of the subjects with a history of dementia had neuropathologically confirmed AD.

(1) This retrospective study was limited, by the availability of tissue sections, to hippocampal and neocortical examination.

 Erlangsen et al. 2008 [36]

Dynamic cohort study

N = 2,474,767 (all individuals age 50 + years living in Denmark); 52.8% females

NA

No

To examine the risk of suicide in persons diagnosed with dementia during a hospitalization and its relationship to mood disorders.

Outcome of interest is suicide. Relative risks are calculated based on person days spent in each stratum.

(1) 136 persons who previously had been diagnosed with dementia died by suicide.

(2) Men and women aged 50–69 years with hospital presentations of dementia have a relative suicide risk of 8.5 and 10.8, respectively. Those who are aged 70 or older with dementia have a threefold higher risk than persons with no dementia.

(3) The time shortly after diagnosis is associated with an elevated suicide risk.

(4) The risk among persons with dementia remains significant when controlling for mood disorders.

(5) 26% of the men and 14% of the women who died by suicide died within the first 3 months after being diagnosed, whereas 38% of the men and 41% of the women died more than 3 years after initial dementia diagnosis.

(1) Study only restricted to subjects diagnose with dementia during hospitalization.

 Purandare et al. 2009 [37]

Retrospective case control study

N = 118 dementia patients died by suicide compared with N = 492 age and gender-matched non-dementia patients; 47% females

NA

No

To describe behavioural, clinical and care characteristics of people with dementia who died by suicide.

ICD-10

(1) The most common method of suicide in patients with dementia was self-poisoning, followed by drowning and hanging, the latter being less frequent than in controls.

(2) Significantly fewer suicides occurred within 1 year of diagnosis in patients with dementia.

(3) Patients with dementia were also less likely to have a history of self-harm, psychiatric symptoms and previous psychiatric admissions.

(1) This study is a survey of the clinical circumstances preceding suicide and unable to make causal inference.

(2) The generalisability of findings is limited to patients with dementia in contact with mental health services, most likely patients with dementia with significant BPSD.

(3) The clinicians who provided the information were not masked to patient outcome and this may have affected their response to certain questions.

 Qin, 2011 [38]

Case-control study

N = 21,169 suicides in Denmark over a 17-year period with sex-age-time-matched population controls; 57% females

NA

No

To assess suicide incidence rate ratio (IRR) and population attributable risk (PAR) associated with various psychiatric disorders

ICD-10

(1) Suicide risk was significantly increased for persons with a hospitalized psychiatric disorder and the associated risk varies significantly by diagnosis and by sex and age of subjects. Recurrent depression and borderline personality disorder increase suicide risk the strongest while dementia increases the risk the least for both males and females.

(1) Cross sectional design.

 McCarthy et al. 2013 [39]

Retrospective cross-sectional study

N = 281,066 from 137 nursing homes; 3.1% females

NA

No

To assess suicide rates up to 6 months following dischargefrom US Department of Veterans Affairs (VA) nursing homes.

ICD-10 Revised

(1) Suicide rates within 6 months of discharge were 88.0 per 100 000 person-years for men and 89.4 overall.

(2) Dementia not associated with increase hazard ratio of suicide.

(1) Findings were derived from VA data and apply only to veterans who receive care from the VA health system.

(2) Low proportion of women discharged from VA nursing home.

 Borges et al. 2015 [40]

Cross-sectional study

N = 1992,

NA

No

To estimate if dementia and other mental disorders are associated with suicide ideation among the older people controlling for demographic and other suspected risk factors.

DSM-IV, ICD-10, CDRS, GMS–AGECAT Package

(1) Lifetime prevalence of suicide ideation of 13.5% and a 2-week prevalence of 4.2%.

(2) Dementia plays a minor role on suicide ideation after the other variables were taken into account and its effect, if any, could be concentrated among those elders with lower severity scores of dementia.

(1) Cross-sectional design.

(2) Limited set of variables to characterize suicide ideation and no information on the severity and persistency of suicide ideation.

(3) Study cannot differentiate among types of dementia.

 Randall et al. 2014 [41]

Population-based, propensity score–matched analysis

N = 2100 suicide deaths and 8641 attempted suicides. Three control subjects were identified for every case and matched on age, sex, income decile, region of residence, and marital status; 50.7% females

NA

No

To determine the degree of risk during the first year after diagnosis with a mental illness.

The Vital Statistics registry of the province was used to determine cases of suicide. Suicide attempts were determined through analysis of physician claims and hospital admission records. The presence of physician-diagnosed depression, anxiety disorders, substance use, schizophrenia, dementia, and other psychosocial disorders was determined using Manitoba RHA Indicators Atlas 2009.

(1) All disorders, except dementia, were independently related to death. All disorders were related to suicide attempts.

(1) Limited by the diagnostic coding used by physicians and hospitals, such as coding major depressive disorder and bipolar disorder under the same code.

(2) Inability to adjust for factors that are not covered in the administrative data, such as stressful life events, childhood adversity, and similar nonmedical factors could not be measured and adjusted for in the analysis.

(3) Inability to adjust for the severity of the disorder, the treatments used by the patients, and whether patients adhered to the treatments they were prescribed by their physicians.

 Nishida et al. 2015 [42]

Retrospective case-control study

N = 24 posttroke depression deceased subjects, 11 of these had committed suicide, and the other 13 had not; 70.1% females

NA

No

To investigate the neuropathologic characteristics of poststroke depression (PSD) leading to suicide.

DSM-4, ICD-10 Revision, CDRS, immunohistochemistry using antibodies to phosphorylated tau, phosphorylated synuclein, A-amyloid, and B-crystallin

(1) Lesion type, size of stroke, and location of stroke were variable but did not differ significantly between the groups.

(2) Alzheimer disease related pathology stages also did not differ between the groups.

(1) Small sample of patients with PSD, including those who had committed suicide.

(2) Diagnosis and treatment of cases were provided by different neurologists and psychologists at different hospitals.

(3) Unable to evaluate the subjects’ education level, which is a known influence risk for PSD, or other psychobiologic factors such as physical disability, ineffective coping skills, and lack of social resources.

 Matschke et al. 2018 [43]

Retrospective case-control study

N = autopsies of 167 suicide dementia cases compared with age- and sex-matched controls who died of other cause. Each suicide was matched to one control according to sex and age within a range of 3 years; 33.3% females

NA

No

To investigate the prevalence of neurodegenerative changes in the brains of suicides of all ages in comparison with age- and sex-matched controls.

Semiquantitative analysis of neuritic plaques and neurofibrillary tangles visualized with silver stains; quantitative immunohistochemical analysis of β-amyloid load and counts of tau-positive neurofibrillary tangles and neuropil threads

(1) No effect of any parameter associated with the odds of committing suicide. On the contrary, after stratification for age, older suicide victims (over 48 years) showed lower β-amyloid loads when compared to controls in the univariate analysis.

(2) In conclusion, neuropathological characteristics of Alzheimer’s disease and common tauopathies associated with age seem to be of limited relevance for suicides.

(3) However, intact cognition when planning and carrying out complex acts may be of importance in the context of suicide.

(1) No access to any detailed clinical data, especially on the presence of depression, neuropsychiatric illness, or concerning antidepressant medications.

(2) No opportunity to investigate the whole brain but only a defined subset of sections.

 Morgan et al. 2018 [44]

Multiphase study

N = 4124 adults aged 65 years and older with a self-harm episode; 53% females

NA

No

To investigate the incidence of self-harm, subsequent clinical management, prevalence of mental and physical diagnoses, and unnatural-cause mortality risk, including suicide.

NICE Clinical Guideline CG16, ICD-10

(1) Overall incidence of self-harm in older adults aged 65 years and older was 4·1 per 10 000 person-years with stable gender-specific rates observed over the 13-year period.

(2) Prevalence ratio of dementia in self-harm group to controls subsequently after index date not increased

(1) Some hospital-treated cases of selfharm will not have been reported to an individual’s GP and, therefore, will not have been captured.

(2) Studies investigating suicide tend to underestimate because coroners might be reluctant to return a verdict of suicide more frequently in unnatural deaths of older people who might have strong religious affiliations, and levels of stigma surrounding suicide among this age group.

 Zucca et al. 2019 [45]

Retrospective case-control study

N = 35 bvFTD patients and 25 controls; 56.7% females

NA

No

To determine the prevalence of suicidal ideation and attempts in bvFTD patients, evaluating possible risk factors for suicidality.

SSI, MMSE, CDR, FAB, AES-C, HDRS, HARS, PSS, BIS-11, BHS, neuroimaging investigations (brain MRI and 18-FDG PET)

(1) 40% of bvFTD patients had suicidal ideation in comparison to 8% of controls (p = .009). Four bvFTD patients have attempted suicide versus none control (p = .006).

(2) BvFTD patients with suicide risk showed higher levels of anxiety, depression, stress and hopelessness than patients without suicide risk (p < .001).

(3) Patients who attempted suicide were younger, and had a longer disease duration than those with only suicide ideation.

(4) 40% of patients with parkinsonism presented high level of suicide ideation.

(1) Small sample size

(2) Only study on bvFTD, not other types of FTD

(3) bvFTD patients present a lower level of education in respect to controls.

 Ng et al. 2020 [46]

Retrospective case-control study

N = 183 ADAD at-risk individuals (91 mutation carriers and 92 non-carriers); 61.2% females

NA

No

To study the frequency of suicidal ideation and its association with clinical and neurobiological correlates among cognitively intact ADAD at-risk individuals.

Suicide question from the UDS B6, GDS, informant-based NPI-Q, awareness of mutation status, neuropsychological assessments, and biological factors (genetic and cerebrospinal fluid)

(1) Twenty-six (14.20%) ADAD at-risk individuals (13 [14.28%] carriers and 13 [14.13%] non-carriers) had suicidal ideation.

(2) The frequency of suicidal ideation did not differ between carriers and non-carriers.

(3) Suicidal ideation was associated with higher GDS among all ADAD atrisk individuals. When stratified into mutation carrier status, non-carriers with suicidal ideation had higher GDS than carriers.

(4) There was no statistically significant association between suicidal ideation and NPI-Q among ADAD at-risk individuals.

(5) Awareness of mutation status, neuropsychological performances, and cerebrospinal fluid AD biomarkers were not associated with suicidal ideation among carriers and noncarriers.

(1) The single suicide question encompasses three components, which assess different levels of suicidal ideation. This may reduce the specificity in detecting the specific suicidal thought that may lead to an attempt.

(2) GDS is originally designed to detect depressive symptoms among elderly individuals and may thus not be suitable for the younger participants in the DIAN cohort.

(3) Cross-sectional design.

(4) The inclusion of a comparator group, ideally of family members of individuals with sporadic AD, will be ideal to control for stress and caregiver burden, these data are unavailable for this study.

 Ortner et al. 2021 [47]

Retrospective cross-sectional study

N = 157 dementia patients, 55.4% females

NA

No

To evaluate the prevalence of death wishes, suicidal ideation, and suicidal behavior of young and late onset dementia and to identify risk factors for suicidal ideation and behaviour.

CSSRS before diagnosis of dementia, immediately after diagnosis of dementia, and 30 days prior to interview

(1) 28% of the patients expressed suicidal ideation or behavior at some time after the onset of symptoms, and 9% of these within the month prior to the assessment. Two patients had attempted suicide after the onset of dementia.

(2) There were no statistically significant differences between patients with and without suicidal ideations or behavior with regards to demographics or age at onset of dementia.

(3) In patients with advanced dementia, Alzheimer’s disease (rather than frontotemporal lobar degeneration), better cognitive function, more severe psychological, behavioral, and physical symptoms, and a reduced quality of life were associated with the expression of suicidal ideation.

(4) Patients with suicidal ideations in early stage of dementia stop to express them at advanced stages.

(1) Retrospective history from caregivers may lead to recall bias.

(2) Small sample size.

(3) Caregivers were asked to report on sucidal behavior of dementia patients, rather than the patients themselves which may lead to respondent bias.

 Alothman et al. 2022 [48]

population-based case-control study

N = 594 674 patients with 580 159 (97.6%) were controls, 40 live control participants per suicide case were randomly matched on primary care practice and suicide date.

NA

No

To examine the association between a dementia diagnosis and suicide risk in the general population and to identify high-risk subgroups.

multiple linked electronic records from primary care, secondary care, and the Office for National Statistics in England from 2001 to 2019.

(1) Among those who died by suicide, 1.9% had a recorded dementia diagnosis.

(2) There was no overall significant association between a dementia diagnosis and suicide risk.

(3) However, suicide risk was significantly increased in patients diagnosed with dementia before age 65 years, in the first 3 months after diagnosis, and in patients with dementia and psychiatric comorbidity.

(4) In patients younger than 65 years and within 3 months of diagnosis, suicide risk was 6.69 times higher than in patients without dementia.

(1) Diagnosis of dementia not confirmed clinically.

 Barak et al. 2002 [49]

Retrospective case-control study

N = 1551 admission from 1991 to 2000 > 60 years old, divide into suicidal and non-suicidal patients (as controls)

NA

No

To examine the association between dementia and suicidal attempts.

DSM-4

(1) 22% diagnosed with dementia.

(2) 7.4% of all AD patients were admitted immediately following a suicide attempt.

(3) The index group (suicidal patients) differed from controls in Clinical Dementia Rating scores (p = 0.017) and higher frequency of previous suicide attempts (p = 0.022).

(4) Lifetime psychopathology was not associated with higher rates of suicide attempts (p = 0.068).

(5) Higher level of daily functioning and previous suicide attempts are associated with increased suicidal risk.

(1) Recruitment from only one center. Hence, findings not representative of dementia.

 Seyfried et al. 2011 [50]

Case-control study

N = 294,952 dementia patients registered in Department of Veterans Affairs (VA) National Care Patient Database, 2.8% females

NA

No

To compare VA patients with dementia who committed suicide during the study period vs. those who did not by demographic characteristics, medical comorbidity, health care utilization and medication use variables. In addition, to examine the relationship between dementia severity and suicide and the methods used by those who killed themselves.

ICD-9 and ICD-10

(1) 8.17% of dementia patients died by suicide.

(2) Increased suicide risk was associated with white race, depression, a history of inpatient psychiatric hospitalizations, and prescription fills of antidepressants or anxiolytics.

(3) The majority of suicides occurred in those with new dementia diagnoses.

(4) Firearms were the most common suicide method (73%).

(1) Study cohort predorminantly males.

(2) Early dementia cases may not be diagnosed and identified.

 Tu et al. 2016 [51]

Nationawide longitudinal cohort study

N = 1,189 patients aged ≥ 65 years who attempted suicide and 4,756 age- and sex-matched control subjects identified from the Taiwan National Health Insurance Research Database

NA

No

To investigate the risk of developing dementia in elderly people who had attempted suicide.

ICD-9

(1) Geriatric suicide attempt was associated with an increased risk of subsequent dementia.

(2) Both patients aged between 65 and 79 years and patients aged ≥ 80 years who attempted suicide had an increased risk of developing dementia in later life, independent of depression and medical comorbidities.

(1) Only subjects who sought medical help and consultation were enrolled. Hence, subject may not represent the entire geriatric suicidal population.

(2) The study did not include family history, personal lifestyle, environmental factors, and nutrition status.

 Annor et al. 2019 [52]

Cross-sectional study

N = 141,592 persons with dementia in the 2013–2014 Medicare fee-for-service ADRD registry data, 67,706 persons with dementia who died during 2013–2016, 30.8% females

NA

No

To examine the characteristics, precipitants, and risk factors for suicide among persons with dementia.

Georgia Alzheimer’s Disease and Related Dementia (ADRD) registry

(1) Suicide rate among persons with dementia was 9.3/100,000 person-years overall and substantially higher among those diagnosed in the past 12 months.

(2) Common precipitating factors were depressed mood (38.7%) and physical health problems (72.6%).

(3) Being male, dementia diagnosis before age 65, and a recent diagnosis of dementia independently predicted suicide, but not depression or cardiovascular diseases.

(1) Recruitment only from a single state, hence findings may not represent the American dementia population.

(2) Cross-sectional design.

(3) Diagnosis of dementia not confirmed clinically.

(4) Study did not control for all socio-demographic factors.

(5) Those younger than 65 years may have missed as they had not gain Medicare service.

 Choi et al. 2021 [19]

National Insurance Health survey for elderly

N = 528,655, those with dementia = 36,541, those without dementia = 36,541; 1:1 propensity-score matching using sex, age, comorbidities and index year, with follow-up throughout 2013

1:1 propensity-score matching

No

To investigate suicide risk in older adults within 1 year of receiving a diagnosis of dementia.

Mini-Mental State Examination score ≤ 26 and a Clinical Dementia Rating score ≥ 1 or a Global Deterioration Scale score ≥ 3 to identify those with dementia

(1) 46 suicide deaths (0.13%) during the first year after a dementia diagnosis.

(2) Older adults with dementia had an increased risk of suicide death compared to those without dementia (AHR 2.57; 95% confidence interval [CI] 1.49–4.44).

(3) Older adults with Alzheimer disease (AHR 2.50; 95% CI 1.41–4.44) or other/unspecified dementia (AHR 4.32; 95% CI 2.04–9.15) had an increased risk of suicide death compared to those without dementia.

(4) Patients with dementia but without other mental disorders (AHR 1.96; 95% CI 1.02–3.77) and patients with dementia and other mental disorders (AHR 3.22; 95% CI 1.78–5.83) had an increased risk of suicide death compared to patients without dementia.

(5) Patients with dementia and schizophrenia (AHR 8.73; 95% CI 2.57–29.71), mood disorders (AHR 2.84; 95% CI 1.23–6.53) or anxiety or somatoform disorders (AHR 3.53; 95% CI 1.73–7.21), respectively, had an increased risk of suicide death compared to patients with those conditions but without dementia.

(1) This study examined only elderly patients in South Korea, a population with a substantially higher suicide rate than the global population. Caution must be exercised when generalizing the results to populations with dissimilar backgrounds.

 Günak et al. 2021 [53]

Nationwide longitudinal cohort study

N = 147 595: 21 085 patients with MCI, 63 255 with dementia, and 63 255 in the propensity-matched comparison group; 2.9% females

NA

No

To examine the association between diagnoses of MCI and dementia and suicide attempt and explore potential psychiatric moderators and to assess whether the association differs based on recency of diagnosis.

ICD-9, ICD-10

(1) 0.7% of patients with MCI and 0.6% of patients with dementia attempted suicide during follow-up, compared with 0.4% of patients without MCI or dementia.

(2) After adjustment for demographic details and medical and psychiatric comorbidities, risk of suicide attempt was consistently highest for patients with a recent MCI or dementia diagnosis.

(3) Risk associated with prior diagnosis was not significant.

(4) No psychiatric comorbidity moderated the association between MCI or dementia and suicide attempt.

(1) Predorminant by male patients.

(2) Study did not include potential risk factors such as social isolation, including loneliness, and assessment of brain injury.

(3) Study did not focus on course or stage of illness itself.

 Holmstrand et al. 2021 [54]

Cohort study

N = 1223 people with dementia from 8 European countries

NA

No

(1) To investigate the occurrence of suicidal ideation in older persons with dementia living at home, proxy-reported by their informal caregivers, in eight European countries.

(2) To investigate factors associated with suicidal ideation, such as demographics, physical and mental health, type and stage of dementia, QoL, and psychotropic medication, and changes in suicidal ideation over time using 3-month follow-up data

Primary diagnosis of dementia and a Standardized Mini-Mental State Examination (SMMSE) score of < 24. Neuropsychiatric Inventory Questionnaire (NPI-Q); the SMMSE; the Cornell Scale for Depression in Dementia (CSDD); the Charlson Comorbidity Index (CCI); and the Quality of Life in Alzheimer’s Disease (QoL-AD) scale

(1) The occurrence of suicidal ideation in the participating countries varied between 6 and 24%.

(2) Factors significantly (p < 0.0018) associated with suicidal ideation using bivariate analysis were: nationality, depressive symptoms, delusions, hallucinations, agitation, anxiety, apathy, disinhibition, irritability, night-time behaviour disturbances, anxiolytics and anti-dementia medication.

(3) In multivariate regression analysis, country of origin, moderate stage of the dementia, depressive and delusional symptoms, and anti-dementia medication were significantly associated with suicidal ideation (p < 0.05).

(4) Over time, suicidal ideation decreased from severe to mild or became absent in 54% of the persons with dementia.

(1) participants in the study were a very specific group of individuals on the margin of care, as they were deemed to require nursing home care within 6 months.

(2) CSDD was not originally designed for use as a diagnostic instrument for depression in persons with dementia.

 Schmutte et al. 2022 [55]

Nationwide retrospective longitudinal cohort

N = 2,667,987 older adults with newly diagnosed dementia, 62.7% females

NA

No

examined the risk of suicide in the first year following ADRD diagnosis relative to the general geriatric population.

ICD-10

(1) The suicide rate for the ADRD cohort was 26.42 per 100,000 person-years.

(2) The overall standardized mortality ratio (SMR) for suicide was 1.53 with the highest risk among adults aged 65–74 years and the first 90 days following ADRD diagnosis.

(3) Rural residence and recent mental health, substance use, or chronic pain conditions were associated with increased suicide risk.

(1) Diagnosis of dementia, not validated.

(2) Study did not include potential risk factors such as lifetime history of self-harm, proximal stressful life events, social disconnection (e.g., marital status, loneliness), and access to lethal means.

(3) Other causes of death may complicate detection of and contribute to underestimated counts of suicide deaths in older adults, particularly in the ≥ 75 years of age group.