Skip to main content
PMC home page
JAMA Network logoLink to JAMA Network
. 2021 May 26;78(8):1–10. doi: 10.1001/jamapsychiatry.2021.0841

Suicide Risk in Medicare Patients With Schizophrenia Across the Life Span

Mark Olfson 1,2,✉, T Scott Stroup 1, Cecilia Huang 3, Melanie M Wall 1,2, Stephen Crystal 3, Tobias Gerhard 3,4
PMCID: PMC8156163  PMID: 34037667

Key Points

Question

How does the risk of suicide in Medicare patients with schizophrenia vary across the life span?

Findings

In this cohort study of 668 836 patients with schizophrenia and Medicare coverage, the risk of suicide was higher compared with the general US population and was highest among those aged 18 to 34 years and lowest among those 65 years and older. In young adults, suicide risk was associated with a recent history of suicide attempts or self-harm, suicide ideation, and substance use disorders.

Meaning

These findings suggest that suicide prevention efforts for individuals with schizophrenia should include a focus on younger adults with suicidal symptoms and substance use disorders.

Abstract

Importance

Although adults with schizophrenia have an increased risk of suicide, sample size limitations of previous research have hindered characterizations of suicide risk across the life span.

Objective

To describe suicide mortality rates and correlates among adults with schizophrenia across the life span and standardized mortality ratios (SMRs) for suicide compared with the general US population.

Design, Setting, and Participants

Five national retrospective longitudinal cohorts of patients with schizophrenia in the Medicare program from January 1, 2007, to December 31, 2016, were identified by age: 18 to 34, 35 to 44, 45 to 54, 55 to 64, and 65 years or older. Death record information was obtained from the National Death Index. The total cohort included 668 836 Medicare patients with schizophrenia, 2 997 308 years of follow-up, and 2218 suicide deaths. Data were analyzed from September 30, 2020, to March 10, 2021.

Main Outcomes and Measures

For each age group, suicide mortality rates per 100 000 person-years and adjusted hazard ratios (aHRs) with 95% CIs of suicide were determined. Suicide SMRs were estimated for the total cohort and by sex and age cohorts standardized to the general US population by age, sex, and race/ethnicity.

Results

The study population of adults 18 years and older included 668 836 Medicare recipients with schizophrenia (52.5% men, 47.5% women). The total suicide rate per 100 000 person-years was 74.00, which is 4.5 times higher than that for the general US population (SMR, 4.54; 95% CI, 4.35-4.73) and included a rate of 88.96 for men and 56.33 for women, which are 3.4 (SMR, 3.39; 95% CI, 3.22-3.57) and 8.2 (SMR, 8.16; 95% CI, 7.60-8.75) times higher, respectively, than the rates for the general US population. Suicide rates were significantly higher for men (aHR, 1.44; 95% CI, 1.29-1.61) and those with depressive (aHR, 1.32; 95% CI, 1.17-1.50), anxiety (aHR, 1.15; 95% CI, 1.02-1.30), drug use (aHR, 1.55; 95% CI, 1.36-1.76), and sleep disorders (aHR, 1.22; 95% CI, 1.07-1.39), suicidal ideation (aHR, 1.41; 95% CI, 1.22-1.63), and suicide attempts or self-injury (aHR, 2.48; 95% CI, 2.06-2.98). The adjusted hazards of suicide were lower for Hispanic patients (aHR, 0.66; 95% CI, 0.54-0.80) or Black patients (aHR, 0.29; 95% CI, 0.24-0.35) than White patients. The suicide rate declined with age, from 141.95 (SMR, 10.19; 95% CI, 9.29-11.18) for patients aged 18 to 34 years to 24.01 (SMR, 1.53; 95% CI, 1.32-1.77) for patients 65 years or older. The corresponding declines per 100 000 person-years were from 153.80 (18-34 years of age) to 34.17 (65 years or older) for men and from 115.70 (18-34 years of age) to 18.66 (65 years or older) for women. In the group aged 18 to 34 years, the adjusted hazards of suicide risk were significantly increased for patients with suicide attempt or self-injury (aHR, 2.57; 95% CI, 18.20-2.04) and with comorbid drug use disorders (aHR, 1.48; 95% CI, 1.17-1.88), but not with comorbid depressive disorders (aHR, 0.99; 95% CI, 0.38-1.26) during the year before the start of follow-up.

Conclusions and Relevance

In this cohort study of adult Medicare patients with schizophrenia, suicide risk was elevated, with the highest absolute and relative risk among young adults. These patterns support suicide prevention efforts with a focus on young adults with schizophrenia, especially those with suicidal symptoms and substance use.

This cohort study describes suicide mortality rates and correlates among adults with schizophrenia across the life span and standardized mortality ratios for suicide compared with the general US population.

Introduction

Suicide is a leading cause of mortality among people with schizophrenia.1 In contrast to the general population, in which suicide risk increases through middle age,2 suicide risk in schizophrenia is high in young adults.3 In a Finnish study,4 the annual suicide rate was roughly 3 times higher during the first 5 years after first hospitalization than after 10 to 16 years. A meta-analysis5 further found that suicide rates were approximately 3 times higher for patients with schizophrenia after illness onset than at any point during their illness. However, few studies have directly evaluated suicide risk of patients with schizophrenia across the life span6,7,8 to inform suicide risk surveillance and prevention by patient age.

Most research on suicide in schizophrenia has focused on young adults, and relatively little is known about suicide risk in older groups.9 A Danish cohort study including 5658 men and 3498 women with schizophrenia recruited as inpatients6 found that the suicide rate per 100 000 person-years for men declined from 883 for patients 29 years or younger to 429 for patients aged 60 to 69 years and increased to 1330 for patients aged 70 to 79 years and to 3230 for patients 80 years or older. No suicide deaths were observed among women with schizophrenia 70 years or older.6 A subsequent Danish study7 reported suicide rates per 100 000 person-years of 218 for men and women aged 50 to 69 years and 107 for men and 68 for women 70 years or older. A study from rural China that followed up 510 patients with schizophrenia for 10 years8 also reported declining suicide risk with advancing patient age.

Beyond uncertainty concerning suicide rates across the life span in schizophrenia, much remains to be learned about suicide risk factors at different ages. Sample size limitations have constrained evaluation of suicide risk factors across the life span. For example, it is not known whether chronic pain conditions,9 other medical comorbidities,10 or attempted suicide, depression, or psychiatric hospital admissions,11,12 all of which are risk factors in the general population, apply to younger, middle-aged, and older adults with schizophrenia.

We describe suicide rates and correlates across the life span in a large cohort of adult patients with schizophrenia and Medicare coverage. Five age groups with schizophrenia in Medicare claims were followed up for suicide risk to estimate suicide rates and standardized mortality ratios (SMRs). Within each age group, we also assessed associations with suicide risk of demographic characteristics, selected clinical diagnoses, and acute health care events before study ascertainment. We hypothesized that suicide rates would be highest among younger adults with schizophrenia and would decline with advancing age but remain elevated among older adults with schizophrenia in relation to the general population.

Methods

Sources of Data

The data source was claims and eligibility information from a 50% random nationally representative sample of Medicare beneficiaries with fee-for-service and Part D prescription drug coverage from January 1, 2007, to December 31, 2016. Medicare is a national public health insurance program that covers people 65 years or older and younger people with disabilities as determined by the Social Security Administration. An estimated 46% of US adults with schizophrenia have Medicare coverage.13 Dates and underlying cause of death information were derived from linkage to the National Death Index. Total resident population and death information were obtained from the 2007-2016 Centers for Disease Control and Prevention WONDER underlying cause of death database.14 The New York State Psychiatric Institute and Rutgers University institutional review boards approved this study and deemed that informed consent was not required for use of retrospective, deidentified data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Cohort Assembly and Follow-up

Medicare beneficiaries were selected with 2 or more outpatient visits within 365 days or 1 or more inpatient schizophrenia diagnoses (International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), code 295.X or International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM), codes F20.X and F25.X). Among these patients, the date of schizophrenia diagnosis was the date on which patients met these criteria and does not necessarily coincide with first lifetime diagnosis of schizophrenia. The start of follow-up (index date) was the first day patients had 12 months of continuous enrollment and met schizophrenia diagnostic criteria. Study patients were 18 years or older on their index date. They were followed up until death, loss of enrollment, age of entering a different age cohort, 5 years from ascertainment, or end of available data, whichever came first. Follow-up was completed on December 31, 2016. In a sensitivity analysis, the sample was restricted to 1 year of follow-up.

Study Outcome

The study outcome was death by suicide (ICD-10-CM codes X60-X84 and Y87.0) as the underlying cause of death in the linked National Death Index.15 The total cohort included 668 836 Medicare patients with schizophrenia, 2 997 308 years of follow-up, and 2218 suicide deaths. In a sensitivity analysis, the study outcome also included deaths of undetermined intent (ICD-10-CM codes Y10-Y30) as probable suicide deaths.16

Sociodemographic and Clinical Characteristics

Cohort members were classified according to public records by age (18-34, 35-44, 45-54, 55-64, and ≥65 years), sex, and race/ethnicity, including non-Hispanic White, non-Hispanic Black, Hispanic, Native American or Alaska Native, Asian, Native Hawaiian, other Pacific Islander, more than 1 race, or missing. Variables representing clinical diagnoses contained in final action claims for Medicare services provided in hospitals, physician offices, and other clinical settings were defined as depression, anxiety, alcohol use, drug use, and sleep disorders as well as chronic pain, suicidal ideation, suicide attempt or self-injury, and a modified Elixhauser Comorbidity Index (see eTables 1 and 2 in the Supplement for ICD-9-CM and ICD-10-CM codes),17 excluding alcohol abuse, drug abuse, psychoses, and depression.18 During this period, two 3-level variables were defined based on first-listed diagnoses as mental health, substance use, and nonbehavioral health inpatient admissions and emergency department visits.

Statistical Analysis

Data analysis was performed from September 30, 2020, to March 10, 2021. The data analysis was performed in 4 stages. First, cohort members were partitioned into 5 age groups, and their demographic characteristics, comorbid clinical diagnoses, and acute health service use characteristics were examined during the 1 year before the index date. Second, person-level age-, sex-, and race/ethnicity–adjusted suicide SMRs with 95% CIs were calculated using observed suicide deaths during follow-up. Expected annual suicide rates standardized by age, sex, and race/ethnicity were derived from the general US population using 2007-2016 WONDER data.14 Suicide SMRs were derived for the total cohort and 5 age groups overall and by sex. In these analyses, follow-up was not censored at 5 years after the index date. Third, unadjusted suicide rates per 100 000 person-years with associated 95% CIs were calculated for each age group and the overall combined cohort, stratified by demographic and clinical characteristics. Finally, separate Cox proportional hazards regression models were fit for the 5 age cohorts and the overall combined cohort adjusted for sex, race/ethnicity, comorbid diagnoses, and acute use of health services. In these models, demographic and clinical characteristics were independent variables of interest, with time until suicide or censoring during follow-up as the dependent variable. Because variables for any inpatient admission and any emergency department visit were defined by mental health, substance use, and nonbehavioral health services, separate fully adjusted models were fit with (1) mental health, substance use, and nonbehavioral health service variables and (2) any inpatient admission and any emergency department visit variables.

Rates and adjusted hazard ratios (aHRs) with 95% CIs not overlapping 1.00 were considered statistically significant. All statistical analyses were performed with SAS software, version 9.4 (SAS Institute, Inc).

Results

Background Characteristics

A total of 668 836 Medicare recipients with schizophrenia were included in the analysis (52.5% men, 47.5% women). With increasing age, the share of female Medicare patients increased from 32.4% (aged 18-34 years) to 65.9% (65 years or older) and the share of non-Hispanic White patients increased from 54.5% (aged 18-34 years) to 67.6% (65 years or older) (Table 1). The percentage of patients with depressive disorders (37.9% to 45.7%), chronic pain conditions (27.7% to 55.0%), and nonbehavioral health inpatient admissions (13.0% to 46.1%) and the mean (SD) Elixhauser Comorbidity Index (0.56 [1.18] to 2.10 [2.46]) increased from the youngest to oldest cohort. By contrast, percentages of patients with alcohol use disorders (8.0% to 4.3%), suicidal ideation (16.8% to 3.7%), suicide attempt or self-injury (3.2% to 0.4%), mental health inpatient admissions (41.0% to 20.7%), and mental health emergency department visits (38.4% to 17.9%) in the year before follow-up decreased from the youngest to the oldest groups. The ratio of suicidal ideation to suicide attempt or self-injury increased with age from approximately 5:1 (16.8:3.2) for patients aged 18 to 34 years to 9:1 (3.7:0.4) for patients 65 years or older.

Table 1. Background Demographic and Clinical Characteristics of Adults With Schizophrenia Diagnosis in the Medicare Programa.

Characteristic Patient group
Total (n = 668 836) Age group, y
18-34 (n = 95 195) 35-44 (n = 112 283) 45-54 (n = 166 939) 55-64 (n = 117 860) ≥65 (n = 176 559)
Sex
Female 317 892 (47.5) 30 839 (32.4) 43 857 (39.1) 69 413 (41.6) 57 505 (48.8) 116 278 (65.9)
Male 350 943 (52.5) 64 356 (67.6) 68 425 (60.9) 97 526 (58.4) 60 355 (51.2) 60 281 (34.1)
Race/ethnicityb
White 423 744 (63.4) 51 871 (54.5) 68 743 (61.2) 105 798 (63.4) 77 907 (66.1) 119 425 (67.6)
Hispanic 59 150 (8.8) 10 670 (11.2) 10 536 (9.4) 12 862 (7.7) 8374 (7.1) 16 708 (9.5)
Black 159 903 (23.9) 26 454 (27.8) 27 863 (24.8) 42 962 (25.7) 27 883 (23.7) 34 741 (19.7)
Otherc 22 760 (3.4) 4733 (5.0) 4787 (4.3) 4882 (2.9) 3211 (2.7) 5147 (2.9)
Comorbid diagnosisd
Depressive disorders 277 406 (41.5) 36 104 (37.9) 44 492 (39.6) 66 280 (39.7) 49 909 (42.3) 80 621 (45.7)
Anxiety disorders 170 587 (25.5) 24 744 (26.0) 29 055 (25.9) 40 422 (24.2) 29 985 (25.4) 46 381 (26.3)
Alcohol use disorder 52 758 (7.9) 7615 (8.0) 11 205 (10.0) 16 856 (10.1) 9555 (8.1) 7527 (4.3)
Drug use disorder 129 071 (19.3) 30 400 (31.9) 30 552 (27.2) 38 440 (23.0) 18 762 (15.9) 10 917 (6.2)
Sleep disorders 90 416 (13.5) 11 879 (12.5) 15 834 (14.1) 23 226 (13.9) 16 855 (14.3) 22 622 (12.8)
Chronic pain 288 465 (43.1) 26 389 (27.7) 40 303 (35.9) 68 378 (41.0) 56 343 (47.8) 97 052 (55.0)
Suicidal ideation 64 565 (9.7) 15 995 (16.8) 15 288 (13.6) 17 769 (10.6) 8986 (7.6) 6527 (3.7)
Suicide attempt or self-injury 10 704 (1.6) 3012 (3.2) 2813 (2.5) 2899 (1.7) 1260 (1.1) 720 (0.4)
Elixhauser Comorbidity Index, mean (SD)e 1.25 (2.00) 0.56 (1.18) 0.77 (1.48) 1.02 (1.78) 1.45 (2.18) 2.01 (2.46)
Use of acute health servicesd
Any inpatient admission 321 548 (48.1) 45 236 (47.5) 48 607 (43.3) 72 293 (43.3) 56 086 (47.6) 99 326 (56.3)
Mental health 191 026 (28.6) 38 993 (41.0) 37 594 (33.5) 47 953 (28.7) 29 880 (25.4) 36 606 (20.7)
Substance use 12 753 (1.9) 2171 (2.3) 3375 (3.0) 4372 (2.6) 1916 (1.6) 919 (0.5)
Nonbehavioral health 189 623 (28.4) 12 352 (13.0) 19 599 (17.4) 38 367 (23.0) 37 834 (32.1) 81 471 (46.1)
Any emergency department visit 409 104 (61.2) 61 351 (64.4) 67 626 (60.2) 96 517 (57.8) 70 109 (59.5) 113 501 (64.3)
Mental health 173 919 (26.0) 36 528 (38.4) 34 966 (31.1) 44 108 (26.4) 26 800 (22.7) 31 517 (17.9)
Substance use 18 818 (2.8) 3855 (4.1) 4757 (4.2) 6039 (3.6) 2752 (2.3) 1415 (0.8)
Nonbehavioral health 351 950 (52.6) 48 173 (50.6) 55 932 (49.8) 81 882 (49.0) 61 654 (52.3) 104 309 (59.1)
Open in a new tab
a

Data are from US Medicare 2007-2016 files and are expressed as No. (%) of patients.

b

The cohort also includes 3279 persons (0.5%) with unknown race/ethnicity.

c

Denotes Native American or Alaskan, Asian, Native Hawaiian, other Pacific Islander, more than 1 race, or missing race/ethnicity.

d

Measured in the 12 months before baseline.

e

Excludes alcohol abuse, drug abuse, psychosis, and depression.

Suicide Risk Relative to the General US Population

The total suicide rate per 100 000 person-years was 74.00, 4.5 times higher than that of the general US population (SMR, 4.54; 95% CI, 4.35-4.73). The suicide rate per 100 000 person-years declined for patients with Medicare coverage with increasing age, from 153.80 (aged 18-34 years) to 34.17 (65 years or older) for male patients and from 115.70 (aged 18-34 years) to 18.66 (65 years or older) for female patients. The suicide SMR was 10.19 (95% CI, 9.29-11.18) for patients aged 18 to 34 years (rate, 141.95 per 100 000 person-years) and 1.53 (95% CI, 1.32-1.77) for patients 65 years or older (rate, 24.01 per 100 000 person-years) (Table 2). In addition, the SMR was 11.45 for patients aged 18 to 23 years, 10.36 for those aged 24 to 29 years, and 8.34 for those aged 30 to 34 years (eTable 3 in the Supplement). Although unadjusted suicide rates per 100 000 person-years were higher for male patients (153.80 at 18-34 years of age to 34.17 at 65 years or older) than for female patients (115.70 at 18-34 years of age to 18.66 at 65 years or older) across the life span (Figure, panel A), the corresponding SMRs were higher for women than men and declined with advancing age from 21.61 to 4.05 for women aged 18 to 34 years to 65 years and older and from 6.91 to 1.14 for men aged 18 to 34 years to 65 years and older (Figure, panel B). The total and age group–stratified suicide rates per 100 000 person-years from a sensitivity analysis without requiring 12 months of continuous Medicare enrollment (eTable 4 in the Supplement) resembled results from the main analysis (Table 2).

Table 2. Observed and Expected Suicide Rates of Adults With Schizophrenia in the Medicare Programa.

Patient group Total (n = 668 836) Men (n = 350 943) Women (n = 317 892)
Suicide rate per 100 000 person-yearsb SMR (95% CI)c Suicide rate per 100 000 person-yearsb SMR (95% CI)c Suicide rate per 100 000 person-yearsb SMR (95% CI)c
Observed Expected Observed Expected Observed Expected
All 74.00 16.31 4.54 (4.35-4.73) 88.96 26.26 3.39 (3.22-3.57) 56.33 6.90 8.16 (7.60-8.75)
Age group, y
18-34 141.95 13.93 10.19 (9.26-11.18) 153.80 22.26 6.91 (6.19-7.70) 115.70 5.36 21.61 (17.85-25.94)
35-44 100.15 16.41 6.10 (5.58-6.67) 111.11 25.09 4.43 (3.98-4.92) 81.10 7.80 10.40 (8.79-25.94)
45-54 86.26 19.59 4.40 (4.09-4.74) 90.08 29.62 3.04 (2.77-3.34) 80.48 9.85 8.17 (7.21-9.22)
55-64 63.69 17.69 3.60 (3.28-3.95) 69.56 27.78 2.48 (2.19-2.79) 56.66 8.29 6.84 (5.89-7.90)
≥65 24.01 15.69 1.53 (1.32-1.77) 34.17 30.03 1.14 (0.91-1.40) 18.66 4.61 4.05 (3.27-4.97)
Open in a new tab

Abbreviation: SMR, standardized mortality ratio.

a

Data are from US 2007-2016 Medicare files and 2007-2016 National Vital Statistics Mortality multiple cause-of-death files. The total cohort comprised 2218 suicide deaths, including 434 among patients aged 18 to 34 years, 480 among patients aged 35 to 44 years, 691 among patients aged 45 to 54 years, 440 among patients aged 55 to 64 years, and 173 among patients 65 years or older.

b

Because each suicide death was assigned to the age group in which it occurred, person-years were not censored at loss of Medicare fee-for-service enrollment, age of entering a different age cohort, or 5 years from ascertainment, as in the other analyses.

c

Standardized by age, sex, and race/ethnicity.

Figure. Suicide Risk of Adults With Schizophrenia in the Medicare Program by Age Group and Sex.

Figure.

Open in a new tab

Error bars indicate 95% CIs. SMR indicates standardized mortality ratio.

Rates of Suicide by Age Groups

In the overall combined cohort, the suicide rate per 100 000 person-years was 58.2 (95% CI, 53.2-63.2) for women and 98.3 (95% CI, 92.1-104.4) for men. For each sex and racial/ethnic group, unadjusted suicide rates declined with advancing age. For example, the suicide rate per 100,000 person-years declined from 115.6 for women aged 18 to 34 years to 19.1 for women 65 years or older. The rate of suicide was particularly low among Black patients 65 years or older (3.1; 95% CI, 0.0-6.6) (Table 3).

Table 3. Rates of Suicide in Adults With Schizophrenia in the Medicare Programa.

Characteristic Patient group
Total (n = 668 836) Age group, y
18-34 (n = 95 195) 35-44 (n = 112 283) 45-54 (n = 166 939) 55-64 (n = 117 860) ≥65 (n = 176 559)
Sex
Female 58.2 (53.2-63.2) 115.6 (92.6-138.7) 81.2 (64.8-97.5) 82.5 (69.9-95.1) 61.5 (49.6-73.4) 19.1 (14.4-23.8)
Male 98.3 (92.1-104.4) 160.8 (142.5-179.2) 114.0 (98.8-129.2) 94.5 (83.4-105.7) 76.1 (63.2-89.0) 38.1 (28.5-47.7)
Race/ethnicity
White 97.9 (92.3-103.5) 186.5 (164.6-208.5) 128.2 (112.1-144.2) 114.3 (102.6-126.1) 88.8 (76.6-101.0) 30.5 (24.6-36.5)
Hispanic 65.0 (52.6-77.4) 89.5 (55.1-123.9) 86.2 (51.7-120.7) 63.4 (37.5-89.4) 68.7 (35.1-102.4) 36.5 (19.1-53.8)
Black 28.4 (23.5-33.3) 73.2 (53.7-92.7) 32.5 (19.5-45.6) 28.4 (19.0-37.9) 13.9 (5.7-22.1) 3.1 (0.0-6.6)
Otherb 105.0 (80.2-129.8) 175.6 (105.4-245.9) 116.3 (59.3-173.3) 142.6 (81.6-203.6) 53.3 (6.6-99.9) 21.2 (0.0-45.1)
Comorbid diagnosisc
Depressive disorders 116.0 (108.3-123.8) 188.3 (161.2-215.3) 158.3 (135.4-181.2) 141.6 (124.4-158.8) 111.7 (93.8-129.6) 38.1 (29.7-46.5)
Anxiety disorders 130.3 (119.4-141.1) 221.2 (184.5-258.0) 156.7 (127.8-185.5) 156.1 (132.3-179.9) 118.4 (93.8-143.0) 43.8 (31.4-56.2)
Alcohol use disorder 176.9 (154.3-199.4) 255.1 (183.6-326.5) 203.9 (150.5-257.4) 185.1 (145.2-224.9) 161.7 (111.0-212.5) 50.6 (17.5-83.7)
Drug use disorders 181.6 (167.3-195.9) 222.8 (190.6-255.0) 192.6 (161.8-223.5) 172.8 (147.7-197.9) 170.5 (133.8-207.2) 75.8 (42.6-109.0)
Sleep disorders 133.4 (118.4-148.4) 211.2 (159.9-262.6) 151.0 (112.8-189.2) 165.0 (132.8-197.2) 119.4 (86.6-152.2) 51.7 (32.6-70.9)
Chronic pain 87.0 (80.4-93.6) 187.6 (155.4-219.9) 124.6 (103.0-146.2) 118.5 (102.9-134.1) 79.0 (64.9-93.1) 25.7 (19.5-32.0)
Suicidal ideation 248.0 (224.1-272.0) 283.8 (232.8-334.8) 230.2 (182.4-278.0) 272.2 (225.3-319.2) 242.0 (178.7-305.4) 130.0 (73.1-187.0)
Suicide attempt or self-injury 535.4 (450.3-620.5) 571.8 (410.1-733.6) 470.4 (312.3-628.5) 566.3 (399.0-733.6) 550.8 (296.3-805.3) 461.9 (141.8-782.0)
Elixhauser Comorbidity Indexd
0 65.9 (61.3-70.5) 134.2 (118.1-150.3) 77.2 (65.5-88.9) 65.9 (57.1-74.6) 48.5 (39.1-57.9) 17.4 (12.2-22.5)
1-2 121.6 (110.1-133.1) 177.5 (141.4-213.6) 164.6 (131.7-197.5) 151.2 (125.8-176.6) 111.4 (85.0-137.9) 39.5 (26.6-52.4)
≥3 83.5 (73.7-93.3) 196.6 (130.5-262.7) 151.1 (108.4-193.9) 123.9 (97.7-150.1) 93.4 (70.9-115.9) 30.2 (21.1-39.2)
Use of acute health servicesc
Any inpatient admission 114.3 (107.1-121.5) 184.0 (160.3-207.6) 165.6 (143.3-187.9) 144.3 (127.6-160.9) 102.9 (86.6-119.2) 34.5 (27.1-41.8)
Mental health 148.5 (138.2-158.9) 189.0 (163.4-214.6) 177.3 (151.4-203.2) 172.0 (150.1-193.9) 138.7 (113.5-163.9) 49.5 (35.6-63.3)
Substance use 227.5 (175.0-280.1) 275.7 (136.2-415.2) 247.6 (139.1-356.1) 214.7 (128.8-300.6) 212.2 (80.7-343.7) 134.2 (0.0-286.1)
Nonbehavioral health 93.5 (84.8-102.3) 222.6 (171.2-274.0) 184.6 (146.3-223.0) 139.9 (116.5-163.2) 87.3 (68.5-106.1) 28.9 (21.3-36.4)
Any emergency department visit 104.4 (98.4-110.5) 174.2 (154.4-194.1) 134.9 (117.9-151.8) 125.5 (112.1-138.8) 90.1 (76.6-103.5) 33.8 (27.0-40.5)
Mental health 147.9 (137.0-158.7) 194.4 (167.5-221.4) 163.2 (137.4-188.9) 168.9 (146.3-191.6) 137.2 (110.6-163.8) 51.3 (35.9-66.6)
Substance use 200.5 (160.0-241.1) 216.3 (123.8-308.8) 271.5 (175.9-367.1) 172.2 (107.2-237.1) 164.0 (67.1-261.0) 119.0 (2.4-235.7)
Nonbehavioral health 99.0 (92.6-105.4) 173.4 (150.9-195.8) 132.5 (113.9-151.2) 122.4 (107.9-136.8) 85.2 (71.1-99.2) 30.8 (24.1-37.6)
Open in a new tab
a

Data are from US 2007-2016 Medicare files and are presented as suicide rate per 100 000 person-years (95% CI).

b

Denotes Native American or Alaskan, Asian, Native Hawaiian, other Pacific Islander, more than 1 race, or missing race/ethnicity.

c

Measured in the 12 months before baseline.

d

Excludes alcohol abuse, drug abuse, psychosis, and depression.

The highest unadjusted suicide rate per 100 000 person-years in the overall combined cohort was among patients with suicide attempts or self-injuries in the year preceding cohort entry (535.4; 95% CI, 450.3-620.5). The 3 next highest suicide rates occurred among patients with prior-year suicidal ideation (248.0; 95% CI, 224.1-272.0), substance use inpatient admissions (227.5; 95% CI, 175.0-280.1), and substance use emergency department visits (200.5; 95% CI, 160.0-241.1). Similar patterns were observed in the 5 age groups: the highest unadjusted suicide rates were for patients with suicide attempts or self-injuries, followed by suicidal ideation and inpatient admissions and emergency department visits for substance use (Table 3). Among patients with suicidal ideation, men had a higher suicide risk than women (aHR, 1.26; 95% CI, 1.03-1.54) (eTable 5 in the Supplement).

Broadening the definition to include deaths of undetermined intent increased the unadjusted suicide rates (eTable 6 in the Supplement). In the total sample, the largest proportionate increase was for Black patients (35.6%, from 28.4 to 38.5 per 100 000 person-years), whereas the smallest increase was for patients with suicide attempts or self-injury (19.7%, from 535.4 to 641.0 per 100 000 person-years) (Table 3 and eTable 6 in the Supplement).

Hazards of Suicide by Age Group

In the fully adjusted Cox proportional hazards regression model of the combined cohort, hazards of suicide were significantly higher for men (aHR, 1.44; 95% CI, 1.29-1.61) and those with depressive (aHR, 1.32; 95% CI, 1.17-1.50), anxiety (aHR, 1.15; 95% CI, 1.02-1.30), drug use (aHR, 1.55; 95% CI, 1.36-1.76), and sleep disorders (aHR, 1.22; 95% CI, 1.07-1.39), suicidal ideation (aHR, 1.41; 95% CI, 1.22-1.63), and suicide attempts or self-injury (aHR, 2.48; 95% CI, 2.06-2.98). The adjusted hazards of suicide were lower for Hispanic (aHR, 0.66; 95% CI, 0.54-0.80) or Black (aHR, 0.29; 95% CI, 0.24-0.35) than White patients. The adjusted hazards of suicide were also significantly associated with mental health (aHR, 1.45; 95% CI, 1.23-1.71) and nonbehavioral (aHR, 1.23; 95% CI, 1.05-1.43) inpatient admissions and emergency department visits (aHR, 1.40; 95% CI, 1,20-1.62) and inversely associated with an Elixhauser Comorbidity Score of 3 or greater (aHR, 0.71; 95% CI, 0.59-0.86) (Table 4). In this model, the adjusted hazard of suicide for patients 65 years or older (reference group, 18-34 years of age) was 0.28 (95% CI, 0.23-0.35) (Table 4 notes).

Table 4. Adjusted Hazard Ratios of Suicide Risk Among Adults With Schizophrenia in the Medicare Programa.

Characteristic Patient group
Total (n = 668 836)b Age group, y
18-34 (n = 95 195) 35-44 (n = 112 283) 45-54 (n = 166 939) 55-64 (n = 117 860) ≥65 (n = 176 559)
Sex
Female 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference]
Male 1.44 (1.29-1.61) 1.56 (1.23-1.99) 1.50 (1.17-1.93) 1.27 (1.04-1.55) 1.27 (0.97-1.65) 2.02 (1.40-2.90)
Race/ethnicity
White 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference]
Hispanic 0.66 (0.54-0.80) 0.52 (0.34-0.77) 0.69 (0.44-1.00) 0.56 (0.37-0.86) 0.77 (0.46-1.28) 1.14 (0.68-1.92)
Black 0.29 (0.24-0.35) 0.45 (0.33-0.60) 0.27 (0.18-0.42) 0.26 (0.18-0.37) 0.16 (0.09-0.30) 0.11 (0.04-0.35)
Otherc 1.08 (0.84-1.38) 1.04 (0.69-1.59) 1.08 (0.65-1.79) 1.43 (0.92-0.22) 0.66 (0.27-1.60) 0.73 (0.23-2.30)
Comorbid diagnosisd
Depressive disorders 1.32 (1.17-1.50) 0.99 (0.38-1.26) 1.49 (1.14-1.95) 1.23 (1.03-1.64) 1.70 (1.25-2.32) 1.58 (1.05-2.40)
Anxiety disorders 1.15 (1.02-1.30) 1.23 (0.97-1.55) 1.04 (0.80-1.35) 1.12 (0.90-1.40) 1.11 (0.82-1.50) 1.27 (0.84-1.90)
Alcohol use disorder 1.04 (0.89-1.23) 1.07 (0.77-1.50) 0.98 (0.70-1.38) 1.08 (0.81-1.44) 1.09 (0.72-1.64) 0.67 (0.29-1.52)
Drug use disorder 1.55 (1.36-1.76) 1.48 (1.17-1.88) 1.60 (1.21-2.11) 1.42 (1.12-1.80) 2.03 (1.46-2.83) 1.66 (0.92-2.98)
Sleep disorders 1.22 (1.07-1.39) 1.14 (0.86-1.51) 1.08 (0.80-1.45) 1.29 (1.02-1.64) 1.20 (0.87-1.67) 1.43 (0.91-2.24)
Chronic pain 0.99 (0.88-1.11) 1.11 (0.87-1.40) 0.92 (0.71-1.18) 1.10 (0.89-1.35) 0.89 (0.67-1.19) 0.76 (0.52-1.11)
Suicidal ideation 1.41 (1.22-1.63) 1.54 (1.16-2.04) 1.14 (0.84-1.56) 1.54 (1.18-2.00) 1.50 (1.03-2.19) 1.93 (1.06-3.51)
Suicide attempt or self-injury 2.48 (2.06-2.98) 2.57 (1.82-3.63) 2.04 (1.37-3.02) 2.46 (1.74-3.47) 3.07 (1.82-5.17) 6.01 (2.67-13.53)
Elixhauser Comorbidity Indexe
0 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference]
1-2 0.92 (0.80-1.08) 0.83 (0.63-1.11) 0.91 (0.66-1.24) 0.86 (0.65-1.14) 1.12 (0.73-1.71) 1.65 (0.86-3.18)
≥3 0.71 (0.59-0.86) 0.97 (0.71-1.32) 0.69 (0.46-1.05) 0.57 (0.40-0.81) 0.83 (0.50-1.38) 1.19 (0.56-2.49)
Use of acute health servicesd
Any inpatient admissionf 1.47 (1.24-1.74) 1.05 (0.78-1.42) 2.19 (1.53-3.15) 1.70 (1.23-2.36) 1.34 (0.79-2.26) 1.64 (0.69-3.93)
Mental health 1.45 (1.23-1.71) 0.97 (0.71-1.32) 1.94 (1.36-2.76) 1.66 (1.22-2.27) 1.42 (0.91-2.23) 1.10 (0.58-2.09)
Substance use 1.05 (0.79-1.39) 1.12 (0.63-2.00) 0.88 (0.51-1.50) 1.10 (0.67-1.83) 1.20 (0.54-2.65) 1.32 (0.28-6.35)
Nonbehavioral health 1.23 (1.05-1.43) 1.24 (0.91-1.70) 1.57 (1.14-2.16) 1.29 (0.98-1.70) 0.97 (0.65-1.43) 0.68 (0.38-1.24)
Any emergency department visitf 1.40 (1.20-1.62) 1.27 (0.96-1.68) 1.35 (0.97-1.87) 1.54 (1.17-2.03) 1.16 (0.80-1.67) 2.01 (1.18-3.45)
Mental health 1.14 (0.98-1.32) 1.17 (0.88-1.55) 0.90 (0.65-1.24) 1.24 (0.99-1.64) 1.18 (0.80-1.75) 1.21 (0.69-2.13)
Substance use 0.78 (0.60-1.00) 0.67 (0.41-1.11) 1.28 (0.80-2.02) 0.62 (0.38-1.00) 0.59 (0.28-1.28) 1.26 (0.30-5.25)
Nonbehavioral health 1.10 (0.97-1.25) 1.02 (0.80-1.29) 1.11 (0.84-1.47) 1.22 (0.96-1.54) 1.00 (0.72-1.39) 1.30 (0.81-2.08)
Open in a new tab

Abbreviation: aHR, adjusted hazard ratio.

a

Data are from US 2007-2016 Medicare files and are presented as aHR (95% CI). Adjusted HRs are from Cox proportional hazards regression models that control for all variables listed.

b

In the total model, the aHRs for the age groups were 1.00 for 18 to 34 years; 0.76 (95% CI, 0.65-0.88) for 35 to 44 years; 0.76 (95% CI, 0.66-0.82) for 45 to 54 years; 0.65 (95% CI, 0.55-0.77) 55 to 64 years; and 0.28 (95% CI, 0.23-0.35) for 65 years or older.

c

Denotes Native American or Alaskan, Asian, Native Hawaiian, other Pacific Islander, more than 1 race, or missing race/ethnicity.

d

Measured in the 12 months before baseline.

e

Excludes alcohol abuse, drug abuse, psychosis, and depression.

f

Results are shown from separate fully adjusted models without mental health, substance use, or nonbehavioral health acute service variables.

The suicide aHRs for each age group were lower for Black than White patients, ranging from 0.11 (95% CI, 0.04-0.35) for patients 65 years or older to 0.45 (95% CI, 0.33-0.60) for patients aged 18 to 34 years. Among patients aged 45 to 54 years, the aHRs were lower for Hispanic compared with White patients (aHR, 0.56; 95% CI, 0.37-0.86). Men had higher aHRs than women, especially among those 65 years or older (aHR, 2.02; 95% CI, 1.40-2.90). Across the age groups, the aHRs were higher for patients with than without a prior-year suicide attempt/self-injury, ranging from 2.04 (95% CI 1.37-3.02) for patients aged 35 to 44 years to 6.01 (95% CI, 2.67-13.53) for patients 65 years or older. For patients with prior-year suicidal ideation, the aHRs ranged from 1.14 (95% CI, 0.84-1.56) for patients aged 35 to 44 years to 1.93 (95% CI, 1.06-3.51) for the oldest group. The aHRs for patients with prior-year depressive disorders were significantly increased in all but the youngest group (aHR, 0.99; 95% CI, 0.38-1.26), whereas the aHRs for patients with prior-year drug use disorders were increased for all but the oldest group (aHR, 1.66; 95% CI, 0.92-2.98).

Suicide hazards for patients aged 35 to 44 years (aHR, 2.19; 95% CI, 1.53-3.15) and 45 to 54 years (aHR, 1.70; 95% CI, 1.23-2.36) were associated with prior-year inpatient admission. For these 2 age groups, mental health inpatient admissions were also associated with increased hazards (aHRs, 1.94 [95% CI, 1.36-2.76] and 1.66 [95% CI, 1.22-2.27], respectively), as were nonbehavioral admissions for patients aged 35 to 44 years (aHR, 1.57; 95% CI, 1.14-2.16) and emergency department visits for patients aged 45 to 54 years (aHR, 1.54; 95% CI, 11.7-2.03) and 65 years or older (aHR, 2.01; 95% CI, 1.18-3.45). A sensitivity analysis with 1-year follow-up yielded generally similar results. However, the adjusted suicide hazards were notably higher in the combined cohort for depressive disorder (aHR, 1.32; 95% CI, 1.17-1.50), any emergency department visit (aHR, 1.40; 95% CI, 1.20-1.62), and any inpatient admission (aHR, 1.47; 95% CI, 1.24-1.74) (Table 4 and eTable 7 in the Supplement).

Discussion

Compared with the general US population, adult Medicare patients with schizophrenia had a 4.5-fold increased risk of dying of suicide. The increased risk was distributed across men and women and concentrated among younger adults, with declining risk with each passing decade. Among young adults, suicide risk in a multivariable model was higher for men than women, higher for White than Black or Hispanic patients, and higher for patients with a history of drug use disorder, suicidal ideation, or suicide attempt or self-injury than patients without such histories. This level of suicide risk elevation resembles results from a population-based Canadian study19 that reported a relative suicide risk of 5.96 for patients with schizophrenia aged 40 to 59 years. Several prior studies limited to schizophrenia cohorts ascertained as inpatients reported higher suicide SMRs6,20 and suicide rates6,7,8 than in the present population-based study.

Some possible explanations for the decline in risk with advancing age include functional recovery, learned adaptations to living with symptoms, stabilization of symptoms, or even gradual improvement among some older adults with schizophrenia.21 Although substantial cognitive22 and social23 deficits are prevalent in older adults with schizophrenia, modest improvements in general psychopathology also occur.24 Consistent with an age-related improvement in psychopathology, a substantially smaller proportion of the oldest group had prior-year mental health hospital admissions, mental health emergency department visits, clinical diagnoses of suicidal ideation, or suicide attempt or self-injury compared with the youngest group. Comorbid drug use disorders, which were associated with suicide risk in this cohort and in prior research,25 were also less common in the oldest group than in the youngest group. However, comorbid depression, which is related to suicide risk in schizophrenia,26 was diagnosed in nearly half of patients in the oldest group.

A healthy survivor bias may have contributed to the relatively low suicide risk in older adult Medicare patients with schizophrenia. Older people with schizophrenia represent a group who has survived high mortality risks from other natural and nonnatural causes earlier in their lives.27 Health behaviors that contribute to their survival might also contribute to their lower risk of suicide later in life. In the older groups, there may also be a depletion of patients who are most susceptible28 to dying of suicide, because those with the strongest predisposition may have already died of suicide.

In the youngest group, the risk of suicide was significantly associated with prior-year comorbid drug use disorders, suicidal ideation, and suicide attempt or self-harm. A substantial proportion of this group had prior-year clinical diagnoses of drug use disorders and suicidal ideation. Similar findings have been reported for risk of nonfatal suicidal attempts in patients with first-episode schizophrenia.29 Consistent with evidence that depressed mood is linked to nonfatal self-harm in patients with first-episode disease,30 we found that the young adult Medicare-insured patients with schizophrenia and comorbid depressive disorders had a high unadjusted rate of suicide. However, comorbid depressive disorders were not associated with suicide risk in the youngest adults in the fully adjusted model.

In the general population, physical pain is a risk factor for suicidal behavior,31 and some theories of suicide assign a central role to physical (or psychological) pain.32 However, chronic pain conditions were not independently associated with suicide risk. Because patients with schizophrenia appear to have a higher pain threshold and pain tolerance than people without schizophrenia,33 the effects of painful physical conditions on suicide risk may be attenuated in patients with schizophrenia.

Within the general population, a larger number of physical illnesses have also been modestly associated with increased suicide risk in older adults.34 However, a psychological autopsy study of decedents without schizophrenia 60 years or older found similar physical illness severity between those who died of suicide and those with other sudden causes of death.35 In the present analysis, higher comorbidity burden as measured by the Elixhauser Comorbidity Index was associated with higher suicide risk in unadjusted analyses, but not after controlling for potentially confounding demographic and clinical characteristics.

Suicide risk was higher among male than female Medicare patients with schizophrenia. However, this sex difference was smaller than the corresponding difference in the general population. A similar pattern has been reported from Finnish6 and Danish7 hospital registries. Outside schizophrenia, sex differences in the prevalence of several risk factors (or the strength of their association with suicide), such as substance use,36 suicidal intent,37 unemployment,38 and single marital status,39 have been posited as explanations for the sex difference in suicide risk. The basis of the relatively narrow sex gap in risk among adults with schizophrenia is not known.

Among Medicare patients with schizophrenia, suicide risk was lower among Black and Hispanic patients than White patients. This finding is consistent with prior research indicating that White individuals with schizophrenia are at greater risk for suicide.29 A similar risk pattern exists in the general population,14 including wide differences in suicide risk among older Black and White adults.40 In the general population, the reasons for the lower risk of suicide among Black vs White individuals are not well understood. Although cultural explanations have stressed the role of religion in lowering suicide acceptability,41 suicide acceptability does not appear to be lower among young Black people vs young White people.42

Limitations

This study has several limitations. First, we have no means of validating the accuracy of schizophrenia diagnoses in Medicare claims data, although we used a validated case identification algorithm,43 and no means of validating suicide as a cause of death, although suicide in death certificates has been found to have a sensitivity of 90% with information from hospital, autopsy, law enforcement, and medical examiner records as the criterion standard.44 It was also not possible to validate the accuracy of the suicidal ideation,45 chronic pain, or the other mental disorder codes in the claims data. Second, because the analysis included only Medicare recipients who were enrolled for at least 1 year and with at least 1 inpatient or at least 2 outpatient schizophrenia codes, selection bias is possible owing to exclusion of patients with shorter Medicare enrollment, private insurance, no insurance, or only Medicaid coverage. The group 65 years or older may also be more broadly representative than younger groups. Third, information was not available concerning several potentially important suicide risk factors such as access to firearms,46 family history of suicide, proximal stressful life events, lifetime history of nonfatal suicide attempts, and age of disorder onset. Shorter illness length may be associated with suicide risk to a greater extent than younger age.29 Fourth, comparisons with the general population were only possible for demographic characteristics. Finally, the data capture suicide mortality patterns from January 1, 2007, to December 31, 2016, and these patterns may have changed since that time, particularly since the onset of the COVID-19 pandemic.

Conclusions

The findings of this cohort study suggest that suicide risk in patients with schizophrenia is highest among young adults and declines over subsequent decades. Among male patients with schizophrenia 65 years or older, suicide risk resembles expected rates in the general US population. Although much remains unknown about why suicide risk declines in schizophrenia across the life span, several risk factors—including suicidal ideation, suicide attempt or self-injury, comorbid drug use diagnoses, and mental health inpatient admissions and emergency department visits—also decline with advancing age. Nevertheless, high suicide risk in young adults with schizophrenia supports clinical vigilance in this patient population, with particular attention to patients with prior suicide attempts or self-injury, suicidal ideation, and comorbid substance use disorders. Suicide prevention in schizophrenia might include expanding access to clozapine,47,48,49 treating comorbid drug use disorders, increasing availability of early psychosis detection programs,50 suicide risk screening in inpatient settings,51 and using cognitive behavioral therapy to help patients with schizophrenia resist command hallucinations52 and reduce their suicidal symptoms.53

Supplement.

eTable 1. Definition of Variables

eTable 2. ICD-9-CM and ICD-10-CM Coding Algorithms for Elixhauser Comorbidities

eTable 3. Observed and Expected Suicide Rates of Young Adults With Schizophrenia in the Medicare Program, Total Cohort and by Age and Sex Groups

eTable 4. Sample Size, Person-Years at Risk, Number of Suicide Deaths, and Unadjusted Rates of Suicide, Total and by Age Group, Excluding Requirement for 1 Year of Medicare Enrollment

eTable 5. Adjusted Hazards of Suicide Risk Among Adults With Schizophrenia and Suicidal Ideation in the Medicare Program

eTable 6. Rates of Broadly Defined Suicide per 100 000 Person-Years of Adults With Schizophrenia in the Medicare Program Overall and by Age Group

eTable 7. Adjusted Hazards of Suicide Risk Among Adults With Schizophrenia in the Medicare Program by Age Group With 1 Year of Follow-up

Click here for additional data file. (219.3KB, pdf)

References

  • 1.Olfson M, Gerhard T, Huang C, Crystal S, Stroup TS. Premature mortality among adults with schizophrenia in the United States. JAMA Psychiatry. 2015;72(12):1172-1181. doi: 10.1001/jamapsychiatry.2015.1737 [DOI] [PubMed] [Google Scholar]
  • 2.Hedegaard H, Curtin SC, Warner M. Increase in Suicide Mortality in the United States, 1999–2018. National Center for Health Statistics; 2020. NCHS data brief 362. [PubMed] [Google Scholar]
  • 3.Nordentoft M, Laursen TM, Agerbo E, Qin P, Høyer EH, Mortensen PB. Change in suicide rates for patients with schizophrenia in Denmark, 1981-97: nested case-control study. BMJ. 2004;329(7460):261. doi: 10.1136/bmj.38133.622488.63 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Heilä H, Haukka J, Suvisaari J, Lönnqvist J. Mortality among patients with schizophrenia and reduced psychiatric hospital care. Psychol Med. 2005;35(5):725-732. doi: 10.1017/S0033291704004118 [DOI] [PubMed] [Google Scholar]
  • 5.Palmer BA, Pankratz VS, Bostwick JM. The lifetime risk of suicide in schizophrenia: a reexamination. Arch Gen Psychiatry. 2005;62(3):247-253. doi: 10.1001/archpsyc.62.3.247 [DOI] [PubMed] [Google Scholar]
  • 6.Mortensen PB, Juel K. Mortality and causes of death in first admitted schizophrenic patients. Br J Psychiatry. 1993;163:183-189. doi: 10.1192/bjp.163.2.183 [DOI] [PubMed] [Google Scholar]
  • 7.Erlangsen A, Eaton WW, Mortensen PB, Conwell Y. Schizophrenia—a predictor of suicide during the second half of life? Schizophr Res. 2012;134(2-3):111-117. doi: 10.1016/j.schres.2011.09.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ran MS, Chan CL, Chen EY, et al. Mortality of geriatric and younger patients with schizophrenia in the community. Suicide Life Threat Behav. 2008;38(2):143-151. doi: 10.1521/suli.2008.38.2.143 [DOI] [PubMed] [Google Scholar]
  • 9.Erlangsen A, Stenager E, Conwell Y. Physical diseases as predictors of suicide in older adults: a nationwide, register-based cohort study. Soc Psychiatry Psychiatr Epidemiol. 2015;50(9):1427-1439. doi: 10.1007/s00127-015-1051-0 [DOI] [PubMed] [Google Scholar]
  • 10.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]
  • 11.Franklin JC, Ribeiro JD, Fox KR, et al. Risk factors for suicidal thoughts and behaviors: a meta-analysis of 50 years of research. Psychol Bull. 2017;143(2):187-232. doi: 10.1037/bul0000084 [DOI] [PubMed] [Google Scholar]
  • 12.Fazel S, Runeson B. Suicide. N Engl J Med. 2020;382(3):266-274. doi: 10.1056/NEJMra1902944 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Khaykin E, Eaton WW, Ford DE, Anthony CB, Daumit GL. Health insurance coverage among persons with schizophrenia in the United States. Psychiatr Serv. 2010;61(8):830-834. doi: 10.1176/ps.2010.61.8.830 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.CDC WONDER. Multiple cause of death, 1999-2018. Reviewed March 11, 2021. Accessed February 25, 2021. https://wonder.cdc.gov/mcd-icd10.html
  • 15.Xu J, Murphy SL, Kochanek KD, Bastian B, Arias E. Deaths: final data for 2016. Natl Vital Stat Rep. 2018;67(5):1-76. [PubMed] [Google Scholar]
  • 16.Chung DT, Ryan CJ, Hadzi-Pavlovic D, Singh SP, Stanton C, Large MM. Suicide rates after discharge from psychiatric facilities: a systematic review and meta-analysis. JAMA Psychiatry. 2017;74(7):694-702. doi: 10.1001/jamapsychiatry.2017.1044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Agency for Healthcare Policy and Research . Elixhauser comorbidity software refined for ICD-10-CM. Modified October 23, 2020. Accessed December 22, 2020. https://www.hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/comorbidity_icd10.jsp
  • 18.Goldman ML, Mangurian C, Corbeil T, et al. Medical comorbid diagnoses among adult psychiatric inpatients. Gen Hosp Psychiatry. 2020;66:16-23. doi: 10.1016/j.genhosppsych.2020.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kredentser MS, Martens PJ, Chochinov HM, Prior HJ. Cause and rate of death in people with schizophrenia across the lifespan: a population-based study in Manitoba, Canada. J Clin Psychiatry. 2014;75(2):154-161. doi: 10.4088/JCP.13m08711 [DOI] [PubMed] [Google Scholar]
  • 20.Nordentoft M, Wahlbeck K, Hällgren J, et al. Excess mortality, causes of death and life expectancy in 270,770 patients with recent onset of mental disorders in Denmark, Finland and Sweden. PLoS One. 2013;8(1):e55176. doi: 10.1371/journal.pone.0055176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Shepherd S, Depp CA, Harris G, Halpain M, Palinkas LA, Jeste DV. Perspectives on schizophrenia over the lifespan: a qualitative study. Schizophr Bull. 2012;38(2):295-303. doi: 10.1093/schbul/sbq075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Irani F, Kalkstein S, Moberg EA, Moberg PJ. Neuropsychological performance in older patients with schizophrenia: a meta-analysis of cross-sectional and longitudinal studies. Schizophr Bull. 2011;37(6):1318-1326. doi: 10.1093/schbul/sbq057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Meesters PD, Stek ML, Comijs HC, et al. Social functioning among older community-dwelling patients with schizophrenia: a review. Am J Geriatr Psychiatry. 2010;18(10):862-878. doi: 10.1097/JGP.0b013e3181e446ff [DOI] [PubMed] [Google Scholar]
  • 24.Jeste DV, Twamley EW, Eyler Zorrilla LT, Golshan S, Patterson TL, Palmer BW. Aging and outcome in schizophrenia. Acta Psychiatr Scand. 2003;107(5):336-343. doi: 10.1034/j.1600-0447.2003.01434.x [DOI] [PubMed] [Google Scholar]
  • 25.Østergaard MLD, Nordentoft M, Hjorthøj C. Associations between substance use disorders and suicide or suicide attempts in people with mental illness: a Danish nation-wide, prospective, register-based study of patients diagnosed with schizophrenia, bipolar disorder, unipolar depression or personality disorder. Addiction. 2017;112(7):1250-1259. doi: 10.1111/add.13788 [DOI] [PubMed] [Google Scholar]
  • 26.Hor K, Taylor M. Suicide and schizophrenia: a systematic review of rates and risk factors. J Psychopharmacol. 2010;24(4)(suppl):81-90. doi: 10.1177/1359786810385490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Laursen TM, Munk-Olsen T, Nordentoft M, Mortensen PB. Increased mortality among patients admitted with major psychiatric disorders: a register-based study comparing mortality in unipolar depressive disorder, bipolar affective disorder, schizoaffective disorder, and schizophrenia. J Clin Psychiatry. 2007;68(6):899-907. doi: 10.4088/JCP.v68n0612 [DOI] [PubMed] [Google Scholar]
  • 28.Moride Y, Abenhaim L. Evidence of the depletion of susceptibles effect in non-experimental pharmacoepidemiologic research [published correction appears in J Clin Epidemiol. 2004;57(1):111]. J Clin Epidemiol. 1994;47(7):731-737. [DOI] [PubMed] [Google Scholar]
  • 29.Cassidy RM, Yang F, Kapczinski F, Passos IC. Risk factors for suicidality in patients with schizophrenia: a systematic review, meta-analysis, and meta-regression of 96 studies. Schizophr Bull. 2018;44(4):787-797. doi: 10.1093/schbul/sbx131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Challis S, Nielssen O, Harris A, Large M. Systematic meta-analysis of the risk factors for deliberate self-harm before and after treatment for first-episode psychosis. Acta Psychiatr Scand. 2013;127(6):442-454. doi: 10.1111/acps.12074 [DOI] [PubMed] [Google Scholar]
  • 31.Calati R, Laglaoui Bakhiyi C, Artero S, Ilgen M, Courtet P. The impact of physical pain on suicidal thoughts and behaviors: meta-analyses. J Psychiatr Res. 2015;71:16-32. doi: 10.1016/j.jpsychires.2015.09.004 [DOI] [PubMed] [Google Scholar]
  • 32.Klonsky ED, May AM. The three-step theory (3ST): a new theory of suicide rooted in “ideation-to-action” framework. Int J Cogn Ther. 2015;8(2):114-129. doi: 10.1521/ijct.2015.8.2.114 [DOI] [Google Scholar]
  • 33.Stubbs B, Thompson T, Acaster S, Vancampfort D, Gaughran F, Correll CU. Decreased pain sensitivity among people with schizophrenia: a meta-analysis of experimental pain induction studies. Pain. 2015;156(11):2121-2131. doi: 10.1097/j.pain.0000000000000304 [DOI] [PubMed] [Google Scholar]
  • 34.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]
  • 35.De Leo D, Draper BM, Snowdon J, Kõlves K. Suicides in older adults: a case-control psychological autopsy study in Australia. J Psychiatr Res. 2013;47(7):980-988. doi: 10.1016/j.jpsychires.2013.02.009 [DOI] [PubMed] [Google Scholar]
  • 36.Bohnert KM, Ilgen MA, Louzon S, McCarthy JF, Katz IR. Substance use disorders and the risk of suicide mortality among men and women in the US Veterans Health Administration. Addiction. 2017;112(7):1193-1201. doi: 10.1111/add.13774 [DOI] [PubMed] [Google Scholar]
  • 37.Freeman A, Mergl R, Kohls E, et al. A cross-national study on gender differences in suicide intent. BMC Psychiatry. 2017;17(1):234. doi: 10.1186/s12888-017-1398-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Milner A, Page A, LaMontagne AD. Cause and effect in studies on unemployment, mental health and suicide: a meta-analytic and conceptual review. Psychol Med. 2014;44(5):909-917. doi: 10.1017/S0033291713001621 [DOI] [PubMed] [Google Scholar]
  • 39.Qin P, Agerbo E, Westergård-Nielsen N, Eriksson T, Mortensen PB. Gender differences in risk factors for suicide in Denmark. Br J Psychiatry. 2000;177:546-550. doi: 10.1192/bjp.177.6.546 [DOI] [PubMed] [Google Scholar]
  • 40.Garlow SJ, Purselle D, Heninger M. Ethnic differences in patterns of suicide across the life cycle. Am J Psychiatry. 2005;162(2):319-323. doi: 10.1176/appi.ajp.162.2.319 [DOI] [PubMed] [Google Scholar]
  • 41.Stack S, Kposowa AJ. The effect of survivalism--self-expressionism culture on black male suicide acceptability: a cross-national analysis. Soc Sci Med. 2011;72(7):1211-1218. doi: 10.1016/j.socscimed.2011.01.028 [DOI] [PubMed] [Google Scholar]
  • 42.Joe S, Romer D, Jamieson PE. Suicide acceptability is related to suicide planning in U.S. adolescents and young adults. Suicide Life Threat Behav. 2007;37(2):165-178. doi: 10.1521/suli.2007.37.2.165 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Lurie N, Popkin M, Dysken M, Moscovice I, Finch M. Accuracy of diagnoses of schizophrenia in Medicaid claims. Hosp Community Psychiatry. 1992;43(1):69-71. doi: 10.1176/ps.43.1.69 [DOI] [PubMed] [Google Scholar]
  • 44.Moyer LA, Boyle CA, Pollock DA. Validity of death certificates for injury-related causes of death. Am J Epidemiol. 1989;130(5):1024-1032. doi: 10.1093/oxfordjournals.aje.a115403 [DOI] [PubMed] [Google Scholar]
  • 45.Claassen CA, Larkin GL. Occult suicidality in an emergency department population. Br J Psychiatry. 2005;186(4):352-353. doi: 10.1192/bjp.186.4.352 [DOI] [PubMed] [Google Scholar]
  • 46.Anglemyer A, Horvath T, Rutherford G. The accessibility of firearms and risk for suicide and homicide victimization among household members: a systematic review and meta-analysis. Ann Intern Med. 2014;160(2):101-110. doi: 10.7326/M13-1301 [DOI] [PubMed] [Google Scholar]
  • 47.Kelly DL, Freudenreich O, Sayer MA, Love RC. Addressing barriers to clozapine underutilization: a national effort. Psychiatr Serv. 2018;69(2):224-227. doi: 10.1176/appi.ps.201700162 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Meltzer HY, Alphs L, Green AI, et al. ; International Suicide Prevention Trial Study Group . Clozapine treatment for suicidality in schizophrenia: International Suicide Prevention Trial (InterSePT). Arch Gen Psychiatry. 2003;60(1):82-91. doi: 10.1001/archpsyc.60.1.82 [DOI] [PubMed] [Google Scholar]
  • 49.Tiihonen J, Lönnqvist J, Wahlbeck K, et al. 11-year follow-up of mortality in patients with schizophrenia: a population-based cohort study (FIN11 study). Lancet. 2009;374(9690):620-627. doi: 10.1016/S0140-6736(09)60742-X [DOI] [PubMed] [Google Scholar]
  • 50.Melle I, Johannesen JO, Friis S, et al. Early detection of the first episode of schizophrenia and suicidal behavior. Am J Psychiatry. 2006;163(5):800-804. doi: 10.1176/ajp.2006.163.5.800 [DOI] [PubMed] [Google Scholar]
  • 51.Quinlivan L, Cooper J, Steeg S, et al. Scales for predicting risk following self-harm: an observational study in 32 hospitals in England. BMJ Open. 2014;4(5):e004732. doi: 10.1136/bmjopen-2013-004732 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Birchwood M, Michail M, Meaden A, et al. Cognitive behaviour therapy to prevent harmful compliance with command hallucinations (COMMAND): a randomised controlled trial. Lancet Psychiatry. 2014;1(1):23-33. doi: 10.1016/S2215-0366(14)70247-0 [DOI] [PubMed] [Google Scholar]
  • 53.Bateman K, Hansen L, Turkington D, Kingdon D. Cognitive behavioral therapy reduces suicidal ideation in schizophrenia: results from a randomized controlled trial. Suicide Life Threat Behav. 2007;37(3):284-290. doi: 10.1521/suli.2007.37.3.284 [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.

eTable 1. Definition of Variables

eTable 2. ICD-9-CM and ICD-10-CM Coding Algorithms for Elixhauser Comorbidities

eTable 3. Observed and Expected Suicide Rates of Young Adults With Schizophrenia in the Medicare Program, Total Cohort and by Age and Sex Groups

eTable 4. Sample Size, Person-Years at Risk, Number of Suicide Deaths, and Unadjusted Rates of Suicide, Total and by Age Group, Excluding Requirement for 1 Year of Medicare Enrollment

eTable 5. Adjusted Hazards of Suicide Risk Among Adults With Schizophrenia and Suicidal Ideation in the Medicare Program

eTable 6. Rates of Broadly Defined Suicide per 100 000 Person-Years of Adults With Schizophrenia in the Medicare Program Overall and by Age Group

eTable 7. Adjusted Hazards of Suicide Risk Among Adults With Schizophrenia in the Medicare Program by Age Group With 1 Year of Follow-up

Click here for additional data file. (219.3KB, pdf)

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

RESOURCES