Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2023 Feb 22;80(4):287–295. doi: 10.1001/jamapsychiatry.2022.5144

Association of Frailty With Risk of Suicide Attempt in a National Cohort of US Veterans Aged 65 Years or Older

Randall L Kuffel 1,2, Ruth T Morin 1,3, Kenneth E Covinsky 4, W John Boscardin 4, Matthew C Lohman 5, Yixia Li 1,2, Amy L Byers 1,6,✉
PMCID: PMC9947807  PMID: 36811913

This cohort study examines the association of frailty with risk of suicide attempt and suicide death in a nationwide sample of US veterans aged 65 years or older.

Key Points

Question

Is frailty a factor associated with risk for suicide attempts?

Findings

In this cohort study of 2 858 876 US veterans aged 65 years or older, risk of suicide attempt was higher in patients with all levels of frailty (prefrailty and mild, moderate, and severe frailty) compared with those without frailty. Risk of lethal suicide attempt was associated with lower levels of frailty.

Meaning

The findings suggest that additional suicide prevention efforts should be directed toward patients with frailty complications associated with risk for suicide attempts and lethality.

Abstract

Importance

Frailty is associated with reduced physiological reserve, lack of independence, and depression and may be salient for identifying older adults at increased risk of suicide attempt.

Objectives

To examine the association between frailty and risk of suicide attempt and how risk differs based on components of frailty.

Design, Setting, and Participants

This nationwide cohort study integrated databases from the US Department of Veterans Affairs (VA) inpatient and outpatient health care services, Centers for Medicare & Medicaid Services data, and national suicide data. Participants included all US veterans aged 65 years or older who received care at VA medical centers from October 1, 2011, to September 30, 2013. Data were analyzed from April 20, 2021, to May 31, 2022.

Exposures

Frailty, defined based on a validated cumulative-deficit frailty index measured using electronic health data and categorized into 5 levels: nonfrailty, prefrailty, mild frailty, moderate frailty, and severe frailty.

Main Outcomes and Measures

The main outcome was suicide attempts through December 31, 2017, provided by the national Suicide Prevention Applications Network (nonfatal attempts) and Mortality Data Repository (fatal attempts). Frailty level and components of the frailty index (morbidity, function, sensory loss, cognition and mood, and other) were assessed as potential factors associated with suicide attempt.

Results

The study population of 2 858 876 participants included 8955 (0.3%) who attempted suicide over 6 years. Among all participants, the mean (SD) age was 75.4 (8.1) years; 97.7% were men, 2.3% were women, 0.6% were Hispanic, 9.0% were non-Hispanic Black, 87.8% were non-Hispanic White, and 2.6% had other or unknown race and ethnicity. Compared with patients without frailty, risk of suicide attempt was uniformly higher among patients with prefrailty to severe frailty, with adjusted hazard ratios (aHRs) of 1.34 (95% CI, 1.27-1.42; P < .001) for prefrailty, 1.44 (95% CI, 1.35-1.54; P < .001) for mild frailty, 1.48 (95% CI, 1.36-1.60; P < .001) for moderate frailty, and 1.42 (95% CI, 1.29-1.56; P < .001) for severe frailty. Lower levels of frailty were associated with greater risk of lethal suicide attempt (aHR, 1.20 [95% CI, 1.12-1.28] for prefrail veterans). Bipolar disorder (aHR, 2.69; 95% CI, 2.54-2.86), depression (aHR, 1.78; 95% CI, 1.67-1.87), anxiety (aHR, 1.36; 95% CI, 1.28-1.45), chronic pain (aHR, 1.22; 95% CI, 1.15-1.29), use of durable medical equipment (aHR, 1.14; 95% CI, 1.03-1.25), and lung disease (aHR, 1.11; 95% CI, 1.06-1.17) were independently associated with increased risk of suicide attempt.

Conclusions and Relevance

This cohort study found that among US veterans aged 65 years or older, frailty was associated with increased risk of suicide attempts and lower levels of frailty were associated with greater risk of suicide death. Screening and involvement of supportive services across the spectrum of frailty appear to be needed to help reduce risk of suicide attempts.

Introduction

Frailty among older adults is partially defined by reductions in physiological function and reserve leading to adverse outcomes such as death, increased vulnerability to stressors, and loss of independence.1,2 This loss of physiological reserve is often accompanied by a loss of cognitive reserve,3 impairing emotional resilience to stressors. To our knowledge, no study has comprehensively investigated how varying degrees of accumulated health and lifestyle deficits affect risk of late-life suicide. Given that over 9000 US adults aged 65 years or older died by suicide in 2020, up from nearly 6000 in 2008,4 late-life suicide requires evaluation beyond traditional risk factors.5

Prior studies on late-life suicide have lacked breadth and depth to comprehensively examine how frailty influences suicide risk. Most studies have focused on individual risk factors,6 such as depression7,8 or pain9 alone, or have used a limited form of physical and psychiatric multimorbidity to assess suicide risk.10,11,12 However, frailty, captured through an accumulation of deficits, can integrate a wider range of health factors,13,14,15,16 the accumulation of which may be indicative of reduced cognitive reserve3 and reduced emotional resilience.17,18 The potential utility of this type of measure has precedent in the literature, with the claims-based frailty index defined by Medicare data19 predicting mortality better than chronological age alone.20 Through this approach, information on morbidity, function, sensory loss, cognitive deficits, and other relevant conditions could be used to assess suicide risk.21 Although preliminary research suggests an association between physical disability or counts of morbidity and suicidal ideation22 and suicide attempt,23 to our knowledge, studies to date have not investigated an accumulation of deficits captured by a frailty measure and its association with death by suicide in a large clinical cohort.

Our primary objective was to investigate the association of frailty with risk of suicide attempt in a large national cohort of patients aged 65 years or older and examine whether this association was independent of demographic factors and other morbidity. Our secondary objective was to assess which frailty deficits are associated with fatal or nonfatal suicide attempts. We conducted a longitudinal cohort study in 2 858 876 million veterans aged 65 years or older who received their health care at the Veterans Health Administration (VHA). As the largest integrated health care system in the US, the VHA is uniquely positioned to test our hypothesis that an accumulation of health burden is associated with increased risk of late-life suicide. By using a frailty index metric defined by physical, neurological, and psychological deficits (ie, diagnoses) recently developed using VHA data,15 this study examined the utility of a frailty index in the context of late-life suicide risk assessment.

Methods

Data and Participants

We conducted a longitudinal cohort study of all veterans aged 65 years or older who used VHA services across all US states from October 1, 2011, to September 30, 2013 (study baseline), with follow-up through December 31, 2017. Data were analyzed from April 20, 2021, to May 31, 2022. To create this cohort, we used electronic health data (EHD) and linked the following EHD national databases (each having unique identifiers for linkage): (1) National Patient Care Database (NPCD), including all Veterans Affairs (VA) inpatient and outpatient services; (2) Centers for Medicare & Medicaid Services (CMS), including medical claims and diagnoses; (3) Suicide Prevention Applications Network (SPAN), including information on nonfatal suicide attempts24,25,26; and (4) Mortality Data Repository (MDR), including cause-specific death information (primary underlying cause and date of death), drawn largely from the Centers for Disease Control and Prevention (CDC) National Death Index (NDI).27,28 All patients in this study were included in NPCD and CMS. If a patient died or attempted suicide, they would also be included in the MDR or SPAN, as these data sources are outcome specific. The time of the sampling (2012-2013) was based on the VA’s initiation of tracking suicide attempts and coincides with the VA’s agreement to receive cause-specific mortality NDI data from the CDC in 2012. Participant consent for the study was waived by the institutional review board of the University of California, San Francisco, as the investigation involved secondary data analysis. This study was approved by the San Francisco Veterans Affairs Health Care System and the University of California, San Francisco, institutional review boards. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Measures

Frailty

We defined frailty based on a frailty index using a cumulative deficit method16 for VHA data developed by Orkaby et al.15 Our frailty index included 31 variables (eTable 1 in Supplement 1)15 related to 5 health domains: morbidity (eg, diabetes, kidney disease), functional status (eg, arthritis, use of durable medical equipment and resources [eg, motorized wheelchair, home-based practitioner care]), cognition and mood (eg, dementia, depression), sensory loss (eg, blindness, hearing impairment), and other conditions common among older adults (eg, chronic pain, weight loss in past year). These variables were gathered using International Classification of Diseases, Ninth Revision (ICD-9) and Healthcare Common Procedure Coding System codes in medical records at or within 2 years prior to the baseline index date. Finally, we assigned frailty scores to each individual by summing the total number of deficits (variables) acquired and dividing by the total number of possible deficits (31). This method produced scores ranging from 0 to 1. We defined frailty as a 5-level categorical variable, per Orkaby et al15: nonfrailty (frailty index ≤0.1), prefrailty (frailty index >0.1 to 0.2), mild frailty (frailty index >0.2 to 0.3), moderate frailty (frailty index >0.3 to 0.4), and severe frailty (frailty index >0.4).29,30

Suicide and Suicide Attempt

We derived nonfatal and fatal suicide attempt information from the SPAN and MDR databases at the follow-up time (through December 31, 2017). Nonfatal suicide attempt data were derived from SPAN and identified using the CDC’s Self-Directed Violence Classification System.31 SPAN data are compiled using information (from suicide prevention coordinators and clinicians) on individual-level case records for suicide events based on reports in VA medical facilities.24,25,26 We defined death by suicide from MDR with International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes X60-X84 and Y87.0. Corresponding dates of death for these ICD-10 codes were included in the VA MDR database. Those who had a documented nonfatal suicide attempt and subsequently died by suicide were assigned to the fatal attempt group, with lethality of the attempt taking precedence over whether the first attempt was fatal or nonfatal.

Other Measures

Sociodemographics (age, sex, race and ethnicity, and US region information) were obtained using NPCD and/or CMS database records. Age was defined at baseline and grouped into 4 age categories: 65-69, 70-74, 75-79, and 80 or more years. Race and ethnicity included Hispanic, non-Hispanic Black, non-Hispanic White, and other (American Indian or Alaska Native, Asian, and Native Hawaiian) or unknown (declined to answer); race and ethnicity were included in the analysis to describe the sample and adjust for confounding in statistical models. Region corresponded to geographic area in the US: West, Southwest, Midwest, Northeast, and Southeast. Psychiatric morbidity included substance use disorders (SUDs) and posttraumatic stress disorder (PTSD).

Statistical Analysis

We described the sample using bivariate analyses of baseline characteristics stratified by suicide attempt outcomes. Characteristics were summarized using means and SDs or frequencies and proportions and were compared by suicide attempt status using t tests for continuous variables and χ2 tests for categorical variables. Cumulative incidence of any suicide attempt based on follow-up time was plotted by severity of frailty. We examined the association of frailty severity with risk of suicide attempt over time in multivariable models. Fine-Gray proportional hazards regression was used to examine time to outcome event at follow-up, with follow-up time as a time scale, while accounting for competing risk of death from other causes and censoring at the end of the study. We also assessed whether individual items of our frailty index were associated with risk of any suicide attempt (fatal or nonfatal). Models were first adjusted for demographic factors (age, sex, race and ethnicity, and region) and then with additional adjustments for SUDs and PTSD (fully adjusted model).

We conducted sensitivity analyses to assess whether associations between frailty and risk of suicide remained independent of other prominent morbidity indexes, such as the Charlson Comorbidity Index.32 We also calculated E-values for fully adjusted model results to estimate the minimum strength of an unmeasured confounder needed to fully explain away the association. Larger E-values indicate lower likelihood of unmeasured confounding. All statistical tests for models were 2-tailed, with P < .05 defining statistical significance. Analyses were performed using SAS, version 9.4 (SAS Institute), and Stata/MP, version 16.1 (64-bit) (StataCorp LLC).

Results

Descriptive Characteristics

Table 1 presents descriptive characteristics of 2 858 876 veterans aged 65 years or older by suicide attempt outcome. The sample’s mean (SD) age was 75.4 (8.1) years; 2.3% were female, 97.7% were male, 0.6% were Hispanic, 9.0% were non-Hispanic Black, 87.8% were non-Hispanic White, and 2.6% had other or unknown race and ethnicity. Of all veterans, 16.6% had an SUD diagnosis and 6.8% had a PTSD diagnosis. A total of 37.0% of the sample had no frailty, 30.0% had prefrailty, 17.1% had mild frailty, 9.1% had moderate frailty, and 6.8% had severe frailty. Demographic characteristics were relatively even across frailty categories, and prevalence of individual frailty items increased with greater levels of frailty (eTable 2 in Supplement 1). A total of 8955 older veterans in this cohort (0.3%) had a documented suicide attempt during the study period, with 5497 (0.2%) having died by suicide at follow-up. The greatest percentage of any suicide attempts was among those with mild and moderate frailty (0.4%), and the smallest percentage was among those without frailty (0.2%).

Table 1. Baseline Characteristics by Suicide Attempt Outcomes.

Characteristic No. (%) (N = 2 858 876)a
No suicide attempt (n = 2 849 921) Any suicide attempt (n = 8955) Death by suicideb (n = 5497)
Age, y
Mean (SD) 75.4 (8.1) 74.0 (7.7) 75.7 (7.8)
65-69 934 280 (32.8) 3576 (39.9) 1625 (29.6)
70-74 484 311 (17.0) 1556 (17.34) 964 (17.5)
75-79 493 739 (17.3) 1442 (16.1) 1058 (19.3)
≥80 937 591 (32.9) 2381 (26.6) 1850 (33.7)
Sex
Female 66 881 (2.3) 133 (1.5) 28 (0.5)
Male 2 783 040 (97.7) 8822 (98.5) 5469 (99.5)
Race and ethnicity
Hispanic 16 170 (0.6) 39 (0.4) 19 (0.4)
Non-Hispanic Black 256 868 (9.0) 458 (5.1) 152 (2.8)
Non-Hispanic White 2 502 626 (87.8) 8280 (92.5) 5243 (95.4)
Other or unknownc 74 257 (2.6) 178 (2.0) 83 (1.5)
Region
West 462 993 (16.2) 1920 (21.4) 1120 (20.4)
Southwest 317 621 (11.1) 1116 (12.5) 738 (13.4)
Midwest 684 057 (24.0) 1872 (20.9) 1097 (20.0)
Northeast 471 967 (16.6) 1117 (12.5) 667 (12.1)
Southeast 870 686 (30.6) 2782 (31.1) 1839 (33.5)
Substance use disorder 471 875 (16.6) 2803 (31.3) 1396 (25.4)
Posttraumatic stress disorder 192 278 (6.8) 1382 (15.4) 443 (8.1)
Mental health service 503 278 (17.7) 3990 (44.6) 1325 (24.1)
Open in a new tab
a

P values for differences among baseline characteristics comparing no suicide attempt and any suicide attempt were all significant at P < .001 based on t tests and χ2 tests.

b

Death by suicide is a subgroup of any suicide attempt.

c

Other includes American Indian or Alaska Native, Asian, and Native Hawaiian.

Frailty and Risk of Suicide Attempt

Figure 1A presents cumulative incidence of any suicide attempt plotted over time by frailty severity. Veterans with moderate frailty had the greatest cumulative incidence of any suicide attempt over time, followed by those with mild frailty, those with severe frailty, those with prefrailty, and those without frailty. Figure 1B presents cumulative incidence of fatal suicide attempts by frailty severity. Veterans with prefrailty had the greatest cumulative incidence of fatal suicide attempt over time, followed by those with mild frailty, those with moderate frailty, those without frailty, and those with severe frailty.

Figure 1. Cumulative Incidence of Any Suicide Attempt and Death by Suicide Across Frailty Categories.

Figure 1.

Open in a new tab

After adjusting for SUDs and PTSD (fully adjusted model) (Table 2), results were attenuated for any suicide attempt and for death by suicide compared with the model adjusted for baseline demographics (Table 2). Fully adjusted hazard ratios (aHRs) for risk of any suicide attempt were 1.34 (95% CI, 1.27-1.42; P < .001) for veterans with prefrailty, 1.44 (95% CI, 1.35-1.54; P < .001) for veterans with mild frailty, and 1.48 (95% CI, 1.36-1.60; P < .001) for veterans with moderate frailty, with a slight decrease among veterans with severe frailty (1.42; 95% CI, 1.29-1.56; P < .001), compared with those without frailty. In the fully adjusted analysis of the association of frailty with risk of death by suicide, the aHRs were attenuated, with the highest being for veterans with prefrailty (1.20; 95% CI, 1.12-1.28), and there was no association (aHR, 0.92; 95% CI, 0.81-1.04) for veterans with severe frailty compared with veterans without frailty.

Table 2. Frailty Index and Risk of Suicide Among 2 858 876 US Veterans Aged 65 Years or Oldera.

Outcome, model HR (95% CI)
Nonfrailty Prefrailty P value Mild frailty P value Moderate frailty P value Severe frailty P value
Participants, No. 1 057 961 857 132 NA 490 013 NA 258 739 NA 195 031 NA
Any suicide attempt
Adjusted for demographicsb 1 [Reference] 1.49 (1.41-1.58) <.001 1.69 (1.58-1.80) <.001 1.78 (1.64-1.93) <.001 1.78 (1.62-1.95) <.001
Fully adjustedc 1 [Reference] 1.34 (1.27-1.42) <.001 1.44 (1.35-1.54) <.001 1.48 (1.36-1.60) <.001 1.42 (1.29-1.56) <.001
Death by suicide
Adjusted for demographicsb 1 [Reference] 1.26 (1.17-1.34) <.001 1.21 (1.11-1.31) <.001 1.18 (1.06-1.30) <.001 1.03 (0.91-1.17) .62
Fully adjustedc 1 [Reference] 1.20 (1.12-1.28) <.001 1.12 (1.03-1.22) .005 1.07 (0.97-1.19) .17 0.92 (0.81-1.04) .18
Open in a new tab

Abbreviations: HR, hazard ratio; NA, not applicable.

a

Frailty was defined as a 5-level categorical variable, per Orkaby et al15: nonfrailty (frailty index ≤0.1), prefrailty (frailty index >0.1 to 0.2), mild frailty (frailty index >0.2 to 0.3), moderate frailty (frailty index >0.3 to 0.4), and severe frailty (frailty index >0.4).29,30

b

Adjusted for age group (65-69, 70-74, 75-79, and ≥80 years), race and ethnicity, sex, and US region.

c

Adjusted for age group (65-69, 70-74, 75-79, and ≥80 years), race and ethnicity, sex, US region, substance use disorder, and posttraumatic stress disorder.

Individual Frailty Variables and Risk of Suicide Attempt

Some items of the frailty index were associated with increased risk of any suicide attempt in fully adjusted models (Figure 2). Statistically significant individual frailty items were, in the morbidity domain, lung disease (aHR, 1.11; 95% CI, 1.06-1.17); function domain, use of durable medical equipment and resources (aHR, 1.14; 95% CI, 1.03-1.25); and other frailty item domain, chronic pain (aHR, 1.22; 95% CI, 1.15-1.29). Most items in the cognition and mood domain were associated with an increased risk for suicide attempt, including anxiety (aHR, 1.36; 95% CI, 1.28-1.45), depression (aHR, 1.78; 95% CI, 1.67-1.87), and bipolar disorder (aHR, 2.69; 95% CI, 2.54-2.86). No items in the sensory loss domain were associated with increased risk of any suicide attempt.

Figure 2. Frailty Item and Risk of Any Fatal or Nonfatal Suicide Attempt.

Figure 2.

Open in a new tab

Hazard ratios (HRs) are based on a single multivariable model including all frailty variables and adjusted for sociodemographics, substance use disorders, and posttraumatic stress disorder. Frailty variables are presented in ascending order of point estimates within each domain. Markers indicate HRs, with horizontal lines indicating 95% CIs.

Additionally, some frailty conditions were associated with lower risk of any suicide attempt in the fully adjusted models (Figure 2). For example, dementia diagnosis was associated with a lower risk for any suicide attempt compared with no dementia diagnosis (aHR, 0.79; 95% CI, 0.73-0.85). Similarly, incontinence (aHR, 0.84; 95% CI, 0.76-0.92) and heart failure (aHR, 0.86; 95% CI, 0.80-0.93) were associated with a lower risk of suicide attempt. Diagnosis of failure to thrive was also associated with a lower risk of suicide attempt (aHR, 0.66; 95% CI, 0.51-0.87).

In sensitivity analyses that added the Charlson Comorbidity Index in the fully adjusted model, results remained similar and robust. Moreover, with E-values greater than 2 for all fully adjusted HRs for any attempt, unmeasured confounding was unlikely to diminish strength of the results.

Discussion

In this cohort study of 2 858 876 million US veterans aged 65 years or older, mild, moderate, and severe frailty were associated with more than 40% increased risk of any suicide attempt, with older veterans with moderate frailty having the highest risk (48%) compared with veterans without frailty. We further discovered that results were attenuated for risk of death by suicide. Bipolar disorder, depression, anxiety, chronic pain, use of durable medical equipment and resources, and lung disease were all associated with increased risk of any suicide attempt. In contrast, incontinence, heart failure, diabetes, kidney disease, coronary artery disease, hypertension, dementia, fatigue, peripheral vascular disease or intermittent claudication, hearing impairment, and failure to thrive were associated with a lower risk of suicide attempt. To our knowledge, this is the first study to assess the association of frailty overall and as individual components with risk of suicide in late life.

Prior studies have examined associations between disability and suicidal ideation.33,34 In a small study of 88 community-dwelling older adults,22 traditional measures of physical function35,36 (weight loss, slow gait speed, and reduced grip strength) were associated with increased suicidal ideation independent of depression severity. In contrast, our method of capturing frailty using EHD and a cumulative deficit model allowed us to assess 31 components of frailty among approximately 3 million adults aged 65 years or older and study outcomes of suicide attempt and death by suicide. One study from Australia23 investigating multimorbidity followed 38 000 older men for 16 years to examine the association between the sum of individual health systems affected by diseases (eg, circulatory system diseases, endocrinological diseases) and suicide attempts. They found that individuals with more than 5 health systems affected by disease had an 11-fold greater risk of death by suicide compared with those with 2 or fewer health systems affected by disease. In comparison, our granular frailty measures suggested that the individuals with the most severe frailty may be at reduced risk of suicide death. Thus, our measure of frailty may indicate different mechanisms underlying an individual’s lethality of attempt, which may be influenced by their current state of frailty and expected declines in function, capacity, and desire to die by suicide.

Our secondary analyses on individual deficits and their association with risk of suicide attempt further expands the literature. To our knowledge, our study was the first to demonstrate that use of durable medical equipment and resources was associated with risk of suicide attempt. This variable is defined by wheelchair use and at-home nursing care, providing greater nuance for understanding the association between disability and suicide risk.37,38 Additionally, we demonstrated that prevalent dementia diagnoses were associated with reduced risk of suicide attempt. This measure likely reflects past diagnoses of dementia recorded more remotely. As the literature indicates, recent dementia diagnosis is associated with increased risk of suicide attempt.39 This finding may highlight that time-dependent systems mediate the pathway between deficit accumulation and suicidal behavior.

Implications

In this study, increasing frailty was associated with increased risk of any suicide attempt. Frailty is often associated with exhaustion and low physical activity1 and with poor social functioning40 and support.41 Conditions such as prolonged grief42 and social isolation and loneliness43 may foster feelings of disconnectedness or thwarted belongingness. Further, an accumulation of deficits, such as chronic pain,38,44 may be associated with loss of independence, increases in a patient’s perceived burdensomeness, and ultimately, increases in suicide risk.45 In the context of declining health and dwindling social support, suicidal ideation can increase.46 Understanding the individualized factors that may contribute to increased suicide risk will be helpful in developing targeted psychosocial interventions (eg, reducing feelings of burdensomeness, reducing social isolation, and improving emotion regulation) to reduce the risk.

The multitude of age-related medical and life challenges associated with late-life suicide may require more encompassing treatment approaches informed by geroscience, and frailty and late-life depression may prove to be achievable targets. Late-life depression is a major risk factor for suicidal behavior47 and is synergistic with frailty.48 Older adults with depression with a high risk of mortality related to frailty often have depressive symptoms resistant to antidepressants, and patients who reach depression remission often continue to experience frailty symptoms.49 With new evidence linking the depressed-frail phenotype to mitochondrial dysfunction,50 both novel bioenergetic therapeutic approaches and aerobic exercise behavioral interventions51 may be associated with improved depression and frailty together52 and, ultimately, with reduced suicide risk.

Future directions may focus on frailty as a collective construct rather than its components. As depression among older adults often manifests with fewer mood-related symptoms and more somatic symptoms (eg, fatigue, psychomotor slowness),53 physicians and researchers not sensitive to these potential differences in older adults may miss important cues for suicide risk. Instead, capturing data points indicative of reduced physiological and cognitive reserve (eg, deficit accumulation, reduced independence) may be a salient way to evaluate suicide risk. The frailty gestalt captures these distinct and sometimes counteracting aspects of suicide risk (eg, loss of physical independence and fatigue may prevent an individual from carrying out a suicide plan) that would not otherwise be captured by mood or physical health symptoms alone. As each patient has unique clinical and life experience informing suicidal behavior, person-centered care approaches are needed to reduce late-life suicide risk.

Understanding the perceived burdensomeness, desire to die by suicide, and capacity to attempt suicide among patients with frailty is critical for connecting patients with appropriate services.45 For example, a physician meeting an older adult with mild frailty who has chronic pain may guide conversations to assess the patient’s perspective on their health status to determine if suicide risk screening is necessary. Moreover, the same physician meeting an older adult with moderate frailty who has been diagnosed with failure to thrive may instead assess suicide risk by discussing general well-being and quality of life given that the patient may lack physiological or cognitive reserve for a lethal suicide attempt. Future directions could augment the frailty index using deficit-specific weighting to produce a tool more indicative of patient distress and predictive of suicide risk.

Our findings further demonstrated that older adults with severe frailty were at similar risk for death by suicide as those without frailty, while adults with prefrailty and mild frailty were at increased risk. This association between frailty and death by suicide may indicate an inverted U-shaped curve association in which older adults at extreme ends of physiological reserve are less likely to carry out lethal suicide attempts. This phenomenon may be explained by willingness to die, capacity to attempt suicide, and access to lethal means. An individual with mounting health problems (prefrailty and mild frailty) may be more willing to attempt suicide and die by suicide while they retain independence and access to lethal means. This could also be explained in part by the presence of mood disorders, mild cognitive impairment or mild dementia, or both and their interaction with mild frailty. Perhaps at greater levels of frailty, individuals require more caregiver support, and these caregivers, especially of patients with dementia, may discuss and implement removal of firearms,54,55 reducing risk of lethal suicide attempt.

Strengths and Limitations

Our study has important strengths. Use of VA and Medicare data sources facilitated analysis of comprehensive morbidity data for patients aged 65 years or older. Additionally, our large, longitudinal study over 6 years captured more suicide attempts than other comparable studies and thus determined precise estimates of suicide attempt risk for individual frailty components. To our knowledge, no other study has investigated the association of frailty with suicide attempts, including death by suicide, at this level of detail.

This study also has limitations. First, the study population consisted mostly of White, male veterans; thus, our findings may be less generalizable to racial and ethnic minority individuals, females, nonveterans, and irregular users of VHA services. Second, our use of secondary databases for outcome status may have resulted in misclassification of those attempting suicide as those who did not attempt suicide when patients were instead lost to follow-up. With nonfatal suicide attempt data derived from SPAN, non–medically serious suicide attempts may have never been documented in the internal VHA suicide event case management and tracking system and subsequently may not have been captured in our study. Thus, the group not attempting suicide may have included veterans with less serious nonfatal suicide attempts, attenuating the HRs. However, as our fatal attempt outcome data were derived from the MDR and NDI, we likely captured all fatal suicide attempt outcomes and deaths, minimizing loss to follow-up. Third, because we defined frailty status at baseline, we did not capture dynamic changes in frailty with age and suicide risk. However, frailty is a largely absorptive process, meaning individuals are unlikely to transition from frailty to nonfrailty.56,57,58 Fourth, although we had complete capture of VA and Medicare information, we did not know whether patients received care beyond these services. Further, we did not know whether patients with comorbid medical disorders received diagnosis codes for all deficits. These limitations may reduce data available for classifying frailty status but would likely underestimate our results (toward the null) instead of overestimate. In addition, such missing information is likely negligible given the high rate of Medicare enrollment for patients aged 65 years or older and our 2-year history period used to define frailty status. Finally, our study was unable to account for differential (weighted) influence of individual conditions due to use of a validated yet unweighted cumulative deficit frailty index.

Conclusions

This cohort study of 2 858 876 million US veterans aged 65 years or older demonstrated that frailty was associated with an increased risk of suicide attempt, with less robust estimates at higher levels of frailty for fatal attempts. While diagnoses such as bipolar disorder, depression, and anxiety are well-documented factors associated with suicide,59 we also found that chronic pain, use of durable medical equipment and resources, and lung disease were independently associated with risk for suicide. Future studies may replicate our findings and investigate potential underlying mechanisms to better understand the role of frailty as a factor associated with late-life suicide and a candidate for intervention and prevention efforts.

Supplement 1.

eTable 1. Components of the Frailty Index

eTable 2. Baseline Characteristics by Frailty Index Category

eReferences

Click here for additional data file. (201.6KB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (13.9KB, pdf)

References

  • 1.Fried LP, Tangen CM, Walston J, et al. ; Cardiovascular Health Study Collaborative Research Group . Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146-M156. doi: 10.1093/gerona/56.3.M146 [DOI] [PubMed] [Google Scholar]
  • 2.Morley JE, Vellas B, van Kan GA, et al. Frailty consensus: a call to action. J Am Med Dir Assoc. 2013;14(6):392-397. doi: 10.1016/j.jamda.2013.03.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Sardella A, Catalano A, Lenzo V, et al. Association between cognitive reserve dimensions and frailty among older adults: a structured narrative review. Geriatr Gerontol Int. 2020;20(11):1005-1023. doi: 10.1111/ggi.14040 [DOI] [PubMed] [Google Scholar]
  • 4.National Center for Injury Prevention and Control. WISQARS—Web-Based Injury Statistics Query and Reporting System. Centers for Disease Control and Prevention. 2005. Accessed Jan 22, 2021. https://www.cdc.gov/injury/wisqars
  • 5.Conwell Y, Van Orden K, Caine ED. Suicide in older adults. Psychiatr Clin North Am. 2011;34(2):451-468, ix. doi: 10.1016/j.psc.2011.02.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Conejero I, Olié E, Courtet P, Calati R. Suicide in older adults: current perspectives. Clin Interv Aging. 2018;13:691-699. doi: 10.2147/CIA.S130670 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Szanto K, Galfalvy H, Kenneally L, Almasi R, Dombrovski AY. Predictors of serious suicidal behavior in late-life depression. Eur Neuropsychopharmacol. 2020;40:85-98. doi: 10.1016/j.euroneuro.2020.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Vannoy SD, Duberstein P, Cukrowicz K, Lin E, Fan MY, Unützer J. The relationship between suicide ideation and late-life depression. Am J Geriatr Psychiatry. 2007;15(12):1024-1033. doi: 10.1097/JGP.0b013e3180cc2bf1 [DOI] [PubMed] [Google Scholar]
  • 9.Petrosky E, Harpaz R, Fowler KA, et al. Chronic pain among suicide decedents, 2003 to 2014: findings from the National Violent Death Reporting System. Ann Intern Med. 2018;169(7):448-455. doi: 10.7326/M18-0830 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Ahmedani BK, Peterson EL, Hu Y, et al. Major physical health conditions and risk of suicide. Am J Prev Med. 2017;53(3):308-315. doi: 10.1016/j.amepre.2017.04.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kavalidou K, Smith DJ, O’Connor RC. The role of physical and mental health multimorbidity in suicidal ideation. J Affect Disord. 2017;209:80-85. doi: 10.1016/j.jad.2016.11.026 [DOI] [PubMed] [Google Scholar]
  • 12.Stickley A, Koyanagi A, Ueda M, Inoue Y, Waldman K, Oh H. Physical multimorbidity and suicidal behavior in the general population in the United States. J Affect Disord. 2020;260:604-609. doi: 10.1016/j.jad.2019.09.042 [DOI] [PubMed] [Google Scholar]
  • 13.Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381(9868):752-762. doi: 10.1016/S0140-6736(12)62167-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kim DH, Schneeweiss S. Measuring frailty using claims data for pharmacoepidemiologic studies of mortality in older adults: evidence and recommendations. Pharmacoepidemiol Drug Saf. 2014;23(9):891-901. doi: 10.1002/pds.3674 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Orkaby AR, Nussbaum L, Ho YL, et al. The burden of frailty among US veterans and its association with mortality, 2002-2012. J Gerontol A Biol Sci Med Sci. 2019;74(8):1257-1264. doi: 10.1093/gerona/gly232 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Searle SD, Mitnitski A, Gahbauer EA, Gill TM, Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24. doi: 10.1186/1471-2318-8-24 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Baumeister RF. Suicide as escape from self. Psychol Rev. 1990;97(1):90-113. doi: 10.1037/0033-295X.97.1.90 [DOI] [PubMed] [Google Scholar]
  • 18.Freitag S, Schmidt S. Psychosocial correlates of frailty in older adults. Geriatrics (Basel). 2016;1(4):26. doi: 10.3390/geriatrics1040026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kim DH, Schneeweiss S, Glynn RJ, Lipsitz LA, Rockwood K, Avorn J. Measuring frailty in Medicare data: development and validation of a claims-based frailty index. J Gerontol A Biol Sci Med Sci. 2018;73(7):980-987. doi: 10.1093/gerona/glx229 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Mitnitski AB, Mogilner AJ, Rockwood K. Accumulation of deficits as a proxy measure of aging. ScientificWorldJournal. 2001;1:323-336. doi: 10.1100/tsw.2001.58 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kim DH, Glynn RJ, Avorn J, et al. Validation of a claims-based frailty index against physical performance and adverse health outcomes in the Health and Retirement Study. J Gerontol A Biol Sci Med Sci. 2019;74(8):1271-1276. doi: 10.1093/gerona/gly197 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bickford D, Morin RT, Woodworth C, et al. The relationship of frailty and disability with suicidal ideation in late life depression. Aging Ment Health. 2021;25(3):439-444. doi: 10.1080/13607863.2019.1698514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Almeida OP, McCaul K, Hankey GJ, Yeap BB, Golledge J, Flicker L. Suicide in older men: the Health in Men Cohort Study (HIMS). Prev Med. 2016;93:33-38. doi: 10.1016/j.ypmed.2016.09.022 [DOI] [PubMed] [Google Scholar]
  • 24.Hoffmire C, Stephens B, Morley S, Thompson C, Kemp J, Bossarte RM. VA Suicide Prevention Applications Network: a national health care system–based suicide event tracking system. Public Health Rep. 2016;131(6):816-821. doi: 10.1177/0033354916670133 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Kemp J, Bossarte RM. Surveillance of suicide and suicide attempts among veterans: addressing a national imperative. Am J Public Health. 2012;102(Suppl 1)(suppl 1):e4-e5. doi: 10.2105/AJPH.2012.300652 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kemp J, Bossarte R. Suicide Data Report, 2012. Suicide Prevention Resource Center. Accessed March 26, 2021. https://www.sprc.org/resources-programs/suicide-data-report-2012
  • 27.Office of Mental Health and Suicide Prevention . 2019 National Veteran Suicide Prevention Annual Report. US Department of Veterans Affairs. 2019. Accessed March 26, 2021. https://www.mentalhealth.va.gov/docs/data-sheets/2019/2019_National_Veteran_Suicide_Prevention_Annual_Report_508.pdf
  • 28.Anderson RN, Miniño AM, Hoyert DL, Rosenberg HM. Comparability of cause of death between ICD-9 and ICD-10: preliminary estimates. Natl Vital Stat Rep. 2001;49(2):1-32. [PubMed] [Google Scholar]
  • 29.Orkaby AR, Hshieh TT, Gaziano JM, Djousse L, Driver JA. Comparison of two frailty indices in the Physicians’ Health Study. Arch Gerontol Geriatr. 2017;71:21-27. doi: 10.1016/j.archger.2017.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Pajewski NM, Williamson JD, Applegate WB, et al. ; SPRINT Study Research Group . Characterizing frailty status in the Systolic Blood Pressure Intervention Trial. J Gerontol A Biol Sci Med Sci. 2016;71(5):649-655. doi: 10.1093/gerona/glv228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Crosby AE, Ortega L, Melanson C. Self-Directed Violence Surveillance: Uniform Definitions and Recommended Data Elements. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. 2011. Accessed March 26, 2021. https://www.cdc.gov/suicide/pdf/self-directed-violence-a.pdf
  • 32.Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47(11):1245-1251. doi: 10.1016/0895-4356(94)90129-5 [DOI] [PubMed] [Google Scholar]
  • 33.Fässberg MM, Cheung G, Canetto SS, et al. A systematic review of physical illness, functional disability, and suicidal behaviour among older adults. Aging Ment Health. 2016;20(2):166-194. doi: 10.1080/13607863.2015.1083945 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lutz J, Fiske A. Functional disability and suicidal behavior in middle-aged and older adults: a systematic critical review. J Affect Disord. 2018;227:260-271. doi: 10.1016/j.jad.2017.10.043 [DOI] [PubMed] [Google Scholar]
  • 35.Innes E. Handgrip strength testing: a review of the literature. Aust Occup Ther J. 1999;46(3):120-140. doi: 10.1046/j.1440-1630.1999.00182.x [DOI] [Google Scholar]
  • 36.Mitrushina M, Boone KB, Razani J, D’Elia LF. Handbook of Normative Data for Neuropsychological Assessment. 2nd ed. Oxford University Press; 2005. [Google Scholar]
  • 37.Marlow NM, Xie Z, Tanner R, Jo A, Kirby AV. Association between disability and suicide-related outcomes among US adults. Am J Prev Med. 2021;61(6):852-862. doi: 10.1016/j.amepre.2021.05.035 [DOI] [PubMed] [Google Scholar]
  • 38.Wilson KG, Kowal J, Caird SM, Castillo D, McWilliams LA, Heenan A. Self-perceived burden, perceived burdensomeness, and suicidal ideation in patients with chronic pain. Can J Pain. 2017;1(1):127-136. doi: 10.1080/24740527.2017.1368009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Günak MM, Barnes DE, Yaffe K, Li Y, Byers AL. Risk of suicide attempt in patients with recent diagnosis of mild cognitive impairment or dementia. JAMA Psychiatry. 2021;78(6):659-666. doi: 10.1001/jamapsychiatry.2021.0150 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Hoogendijk EO, Suanet B, Dent E, Deeg DJ, Aartsen MJ. Adverse effects of frailty on social functioning in older adults: results from the Longitudinal Aging Study Amsterdam. Maturitas. 2016;83:45-50. doi: 10.1016/j.maturitas.2015.09.002 [DOI] [PubMed] [Google Scholar]
  • 41.Penninx BW, van Tilburg T, Kriegsman DM, Boeke AJ, Deeg DJ, van Eijk JT. Social network, social support, and loneliness in older persons with different chronic diseases. J Aging Health. 1999;11(2):151-168. doi: 10.1177/089826439901100202 [DOI] [PubMed] [Google Scholar]
  • 42.Lutz J, Van Orden KA. Sadness and worry in older adults: differentiating psychiatric illness from normative distress. Med Clin North Am. 2020;104(5):843-854. doi: 10.1016/j.mcna.2020.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Mehrabi F, Béland F. Effects of social isolation, loneliness and frailty on health outcomes and their possible mediators and moderators in community-dwelling older adults: a scoping review. Arch Gerontol Geriatr. 2020;90:104119. doi: 10.1016/j.archger.2020.104119 [DOI] [PubMed] [Google Scholar]
  • 44.Racine M. Chronic pain and suicide risk: a comprehensive review. Prog Neuropsychopharmacol Biol Psychiatry. 2018;87(Pt B):269-280. . [DOI] [PubMed]
  • 45.Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, Joiner TE Jr. The interpersonal theory of suicide. Psychol Rev. 2010;117(2):575-600. doi: 10.1037/a0018697 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Préville M, Hébert R, Boyer R, Bravo G, Seguin M. Physical health and mental disorder in elderly suicide: a case-control study. Aging Ment Health. 2005;9(6):576-584. doi: 10.1080/13607860500192973 [DOI] [PubMed] [Google Scholar]
  • 47.Van Orden K, Conwell Y. Suicides in late life. Curr Psychiatry Rep. 2011;13(3):234-241. doi: 10.1007/s11920-011-0193-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Soysal P, Veronese N, Thompson T, et al. Relationship between depression and frailty in older adults: a systematic review and meta-analysis. Ageing Res Rev. 2017;36:78-87. doi: 10.1016/j.arr.2017.03.005 [DOI] [PubMed] [Google Scholar]
  • 49.Brown PJ, Ciarleglio A, Roose SP, et al. Frailty and depression in late life: a high-risk comorbidity with distinctive clinical presentation and poor antidepressant response. J Gerontol A Biol Sci Med Sci. 2022;77(5):1055-1062. doi: 10.1093/gerona/glab338 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Ampo E, Mendes-Silva AP, Goncalves V, Bartley JM, Kuchel GA, Diniz BS. Increased levels of circulating cell-free mtDNA in the plasma of subjects with late-life depression and frailty: a preliminary study. Am J Geriatr Psychiatry. 2022;30(3):332-337. doi: 10.1016/j.jagp.2021.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Izquierdo M, Merchant RA, Morley JE, et al. International exercise recommendations in older adults (ICFSR): expert consensus guidelines. J Nutr Health Aging. 2021;25(7):824-853. doi: 10.1007/s12603-021-1665-8 [DOI] [PubMed] [Google Scholar]
  • 52.Brown PJ. Evidence for a geroscience approach to late life depression: bioenergetics and the frail-depressed. Am J Geriatr Psychiatry. 2022;30(3):338-341. doi: 10.1016/j.jagp.2021.11.003 [DOI] [PubMed] [Google Scholar]
  • 53.Hegeman JM, de Waal MW, Comijs HC, Kok RM, van der Mast RC. Depression in later life: a more somatic presentation? J Affect Disord. 2015;170:196-202. doi: 10.1016/j.jad.2014.08.032 [DOI] [PubMed] [Google Scholar]
  • 54.Betz ME, Azrael D, Johnson RL, et al. Views on firearm safety among caregivers of people with Alzheimer disease and related dementias. JAMA Netw Open. 2020;3(7):e207756. doi: 10.1001/jamanetworkopen.2020.7756 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Polzer ER, Nearing KA, Knoepke CE, Matlock DD, Betz ME. Firearm access and dementia: a qualitative study of reported behavioral disturbances and responses. J Am Geriatr Soc. 2022;70(2):439-448. doi: 10.1111/jgs.17496 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Gill TM, Gahbauer EA, Allore HG, Han L. Transitions between frailty states among community-living older persons. Arch Intern Med. 2006;166(4):418-423. doi: 10.1001/archinte.166.4.418 [DOI] [PubMed] [Google Scholar]
  • 57.O’Caoimh R, Galluzzo L, Rodríguez-Laso Á, et al. ; Work Package 5 of the Joint Action ADVANTAGE . Transitions and trajectories in frailty states over time: a systematic review of the European Joint Action ADVANTAGE. Ann Ist Super Sanita. 2018;54(3):246-252. [DOI] [PubMed] [Google Scholar]
  • 58.Rogers NT, Marshall A, Roberts CH, Demakakos P, Steptoe A, Scholes S. Physical activity and trajectories of frailty among older adults: evidence from the English Longitudinal Study of Ageing. PLoS One. 2017;12(2):e0170878. doi: 10.1371/journal.pone.0170878 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Beghi M, Butera E, Cerri CG, et al. Suicidal behaviour in older age: A systematic review of risk factors associated to suicide attempts and completed suicides. Neurosci Biobehav Rev. 2021;127:193-211. doi: 10.1016/j.neubiorev.2021.04.011 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eTable 1. Components of the Frailty Index

eTable 2. Baseline Characteristics by Frailty Index Category

eReferences

Click here for additional data file. (201.6KB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (13.9KB, pdf)

Articles from JAMA Psychiatry are provided here courtesy of American Medical Association

RESOURCES