Abstract
Background:
Suicide is a leading cause of death in persons with schizophrenia and other serious mental illnesses (SMI), however, little is known about the characteristics and circumstances of suicide decedents with SMI in the US compared to those with other or no known mental illness.
Methods:
This study was a retrospective analysis of suicide deaths in individuals aged ≥18 years from the National Violent Death Reporting System, 2003–2017. Odds ratios compared sociodemographic and clinical characteristics, cause of death, precipitating circumstances, and post-mortem toxicology results. All analyses were stratified by gender.
Results:
Of the 174,001 suicide decedents, 8.7% had a known SMI, 33.0% had other mental disorders, and 58.2% had no known mental illness. Relative to persons with other mental disorders, SMI decedents were younger and more likely to have previous suicide attempts and co-occurring drug use. Problems with intimate partners, poor physical health, and recent institutional release were the most common precipitating circumstances for SMI decedents. Firearms were the most common suicide method for males with SMI. Although 67.0% male and 76.0% of female SMI decedents were currently in treatment, toxicology results suggest many were not taking antipsychotic or antidepressant medications at the time of death.
Conclusions:
Persons with SMI are over-represented in suicide deaths. Efforts to improve treatment of co-occurring substance use disorders, continuity of care following hospitalization, medication adherence, and to reduce access to firearms are important suicide prevention strategies.
1. Introduction
Persons with serious mental illness (SMI), such as schizophrenia and bipolar disorder, are at markedly increased risk of death from suicide relative to the general population.(Chesney et al., 2014; Pompili et al., 2013) Different studies using different populations and methods estimate the lifetime risk of suicide for the SMI population to be 5%-13%, with risk being disproportionately higher in younger age groups.(Hayes et al., 2015; Hor and Taylor, 2010) Most population-based studies have been conducted in Scandinavian countries and the United Kingdom, with comparatively fewer cohort studies being conducted in the United States (US). Recent research using 2000-2015 data from the National Center for Health Statistics observed suicide to be a leading cause of death among persons with schizophrenia aged 15-44 years in the US.(Lin et al., 2018)
Static risk factors for suicide in persons with SMI include demographic (e.g., male sex) and clinical characteristics (e.g., prior suicide attempts, substance use).(Cassidy et al., 2017; Fazel et al., 2019a; Hansson et al., 2018; Hettige et al., 2017; Schaffer et al., 2015) Proximal risk factors include recent discharge from psychiatric hospitalization.(Chung et al., 2019; Chung et al., 2017; Olfson et al., 2016) Dynamic factors associated with increased suicide risk include lapses in treatment and medication adherence, (Cassidy et al., 2017; Forsman et al., 2019; Pompili et al., 2013) although suicide decedents with schizophrenia have also been shown to have significantly greater mental health service utilization shortly prior to death compared with decedents with other mental disorders and community controls.(Zaheer et al., 2018) Recent life stressors also play a role in non-fatal and fatal suicide attempts, however, they may be less apparent or recognized in persons with SMI compared to those with other or no mental illness.(Chen and Roberts, 2019; Gallego et al., 2015; Heilä et al., 1999; Malhi et al., 2018)
Differences between persons with schizophrenia with and without suicide attempts have been identified,(Hor and Taylor, 2010) however comparatively less research has examined differences in comparison with other diagnostic psychiatric categories. One study compared characteristics of suicide deaths in Chinese rural counties between decedents with schizophrenia (n=38), other mental disorders (n=150), and no mental illness (n=204).(Lyu and Zhang, 2014) Decedents with schizophrenia were disproportionately female compared to both of the other groups, however, no significant differences in clinical characteristics were observed between decedents with schizophrenia compared to other mental illnesses. An Egyptian study compared characteristics of persons with schizophrenia (n=33) and those with other psychiatric disorders (n=87) with non-fatal suicide attempts.(Abdeen et al., 2019) Individuals with schizophrenia were more likely to have a history of medical illness, significantly longer duration of illness, and to use more violent suicidal means but were less likely to have significant suicidal ideation at the time of suicide attempt. However, data from larger scale, US-based studies are unavailable.
Because the circumstances precipitating death from suicide in the SMI population are less understood, and how they compare to persons with other mental disorder or no mental illness in US, the purpose of this study is to compare the characteristics of suicide decedents with SMI to those other mental disorders and those with no known mental illness using data from the National Violent Death Reporting System (NVDRS). The NVDRS was created by the Centers for Disease Control (CDC) to provide surveillance of violent deaths in the US and contains detailed uniformly coded information on the characteristics of decedents, including demographics, mental health history, various recent stressors, and post-mortem toxicology results. Improved understanding of suicide among persons with SMI in the US will contribute to prevention research and help inform selective and universal suicide prevention initiatives for healthcare providers and policy makers.
2. METHODS
2.1. Data Source
Data were obtained from 37 states and territories (District of Columbia and Puerto Rico) that participated in the NVDRS between 2003 and 2017. The NVDRS has been more fully described elsewhere.(Blair et al., 2016) In brief, trained data abstractors code the manner, cause, and precipitating circumstances of death based on their review of multiple documents, including death certificates and reports from coroners/medical examiners. Additional information is abstracted from law enforcement reports that contain information obtained from open-ended interviews with individuals who were close to the decedent (e.g. family members, friends) and witnesses to the death to assess precipitating factors that may have contributed to the suicide. Based on availability of source documents, the NVDRS has the capability to collect over 600 standardized variables for each death and two narrative summaries based on coroners/medical examiners and law enforcement reports.
2.2. Sample
Serious mental illness (SMI) was defined as whether a decedent was diagnosed with schizophrenia and/or bipolar disorder in NVDRS current mental health diagnosis variables. Additional steps to identify persons with SMI included searching for terms such as “schizoaffective” and “psychosis,” “hallucinations,” “delusions” not associated with intoxication or dementia in narrative summaries based on information from coroner/medical examiner or law enforcement reports. Other mental disorders were defined as depression/dysthymia, anxiety disorder, obsessive-compulsive disorder, post-traumatic stress disorder, attention-deficit/hyperactivity disorder, or eating disorder as coded by NVDRS abstractors. Individuals were classified as having no known mental illness if current mental health diagnosis was coded in NVDRS as “no, not available, or unknown.” The cohort was restricted to persons aged ≥18 years with cause of death classified as intentional self-harm/suicide. Deaths categorized as unintentional or undetermined intent were excluded.
2.3. Variables
Demographic characteristics included age, sex, race/ethnicity, marital status, veteran status, and homelessness status. Mental health characteristics included current mental health diagnosis, depressed at the time of death (not a clinical diagnosis), substance use problems, past mental health or substance abuse treatment, and prior suicide attempt. Precipitating circumstances included: (1) physical health problems; (2) relationship problems (i.e., intimate partner, family, or other); (3) recent release from an institution (e.g., hospital, jail, nursing home); (4) recent violence (i.e., perpetrator and victim); (5) argument or conflict; (6) recent deaths of family/friends; (7) legal problems (criminal and civil); (8) job/financial problems; (9) school - related problem; (10) housing problem; (11) disclosure of suicide intent, and (12) leaving a suicide note. All variables for precipitating circumstances are coded in the NVDRS as binary (“yes” or “no, not available, or unknown”). Suicide methods included: (1) firearms; (2) hanging, strangulation, or suffocation; (3) poisoning due to medicine, alcohol, or drug; (4) poisoning due to gas (e.g., carbon monoxide); (5) drowning, (6) jumping from heights, (7) laceration/sharp object; (8) vehicular-related (e.g., jumping in front of train or from an automobile); (9) fire; (10) other (e.g., electrocution, hypothermia); or (11) ill-defined or unknown.
2.4. Statistical Analysis
We first calculated the number and percentage of decedents with SMI, other mental disorders, or no known mental illness stratified by sociodemographic, mental health history, precipitating characteristics, suicide means, and toxicology results. For each characteristic, we then calculated odds ratios (ORs) to compare decedents with SMI to those with other mental disorders and no known mental illness. Multinomial logistic regression was used with SMI decedents as the reference group. To account for large differences in suicide deaths between male and female decedents, analyses were also stratified by sex. To help compensate for the large sample size and number of comparisons, we used 99% confidence intervals (CI) with significance set at p<0.001. Group differences with an ORs of ≥ 1.15 or ≤ 0.85 were considered potentially substantial from a clinical and policy perspective. The study was approved by the Yale University Institutional Review Board.
3.0. Results
3.1. Sample Characteristics
Of the 174,001 suicide decedents, 8.7% (n=15,199) had a known SMI, 33.0% (n=57,478) had other mental disorders, and 58.2% (n=101,324) had no known mental illness. Over three-fourths (77.4%) of the decedents were male (n=134,691). Disproportionately more female than male decedents (13.8% vs. 7.3%, respectively) were diagnosed with SMI (OR=2.08, 99% CI=2.00-2.15). Among decedents with SMI, a majority (79.7%, n=12,116) were diagnosed with a bipolar disorder. Relative to persons with other mental disorders or no known mental illness, males (Table 1) and females (Table 2) with SMI were significantly less likely to have been aged 55 years or older. Relative to decedents with other mental disorders, disproportionately more persons with SMI were African American, never married or single, or homeless. Males with SMI were less likely to be Veterans compared decedents with other or no known mental illness.
Table 1.
Characteristics of Male Suicide Deaths (N=134,691), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2003-2017
| Total % |
Serious Mental Illness (SMI) (n=9,780) % |
Other Mental Illness (MI) (n=40,602) % |
Unknown or No Mental Illness (n=84,309) % |
Odds Ratio (99% CI) SMI compared to MI |
Odds Ratio (99% CI) SMI compared to Unknown or No Mental Illness |
|
|---|---|---|---|---|---|---|
| Sociodemographic | ||||||
| Age (years) | ||||||
| 18-24 (n=15,154) | 11.3% | 13.0% | 10.0% | 11.6% | 1.34(1.23,1.47) | 1.13(1.04,1.23) |
| 25-34 (n=22,899) | 17.0% | 23.5% | 15.1% | 17.2% | 1.73(1.61,1.86) | 1.49(1.39,1.59) |
| 35-44 (n=23,560) | 17.5% | 22.4% | 16.9% | 17.2% | 1.42(1.32,1.52) | 1.38(1.29,1.48) |
| 45-54 (n=27,776) | 20.6% | 22.3% | 22.3% | 19.6% | 1.00(0.93,1.07) | 1.18(1.10,1.26) |
| 55-64 (n=21,689) | 16.1% | 13.5% | 18.2% | 15.4% | 0.70(0.64,0.76) | 0.86(0.79,0.93) |
| 65-74 (n=11,971) | 8.9% | 4.3% | 9.8% | 9.0% | 0.41(0.36,0.47) | 0.45(0.39,0.51) |
| 75-84 (n=8,077) | 6.0% | 0.9% | 5.7% | 6.7% | 0.15(0.11,0.20) | 0.13(0.10,0.17) |
| ≥ 85 (n=3,565) | 2.6% | 0.1% | 2.0% | 3.2% | 0.07,0.03,0.14) | 0.04(0.02,0.09) |
| Race/Ethnicity | ||||||
| White non-Hispanic (n=113,497) | 84.3% | 83.7% | 88.8% | 82.2% | 0.65(0.60,0.71) | 1.12(1.04,1.20) |
| Black non-Hispanic (n=8,344) | 6.2% | 7.2% | 3.9% | 7.2% | 1.92(1.70,2.17) | 1.00(0.90,1.11) |
| Hispanic (n=7,176) | 5.3% | 5.3% | 4.1% | 5.9% | 1.31(1.15,1.49) | 0.90,0.80,1.01) |
| American Indian or Alaskan Native (n=1,532) | 1.1% | 0.7% | 0.7% | 1.4% | 1.00(0.71,1.41) | 0.51(0.37,0.70) |
| Asian American or Pacific Islander (n=2,072) | 1.3% | 1.4% | 1.3% | 1.7% | 1.08(0.84,1.38) | 0.86(0.69,1.09) |
| Other (n=2,015) | 1.5% | 1.6% | 1.2% | 1.6% | 1.37(1.07,1.74) | 0.96(0.77,1.20) |
| Marital Status | ||||||
| Married or Common law (n=47,298) | 34.1% | 23.5% | 37.2% | 33.8% | 0.50(0.47,0.54) | 0.58(0.55,0.62) |
| Divorced, Separated, or Widowed (n=38,992) | 28.1% | 24.3% | 28.1% | 28.6% | 0.80(0.75,0.86) | 0.78(0.73,0.83) |
| Never married/Single (n=46,859) | 36.7% | 51.0% | 34.1% | 36.8% | 2.14(2.02,2.27) | 1.94(1.83,2.05) |
| Unknown (n=1,461) | 1.1% | 1.2% | 0.7% | 1.3% | 1.84(1.38,2.45) | 0.93(0.73,1.20) |
| Military Veteran Status * | ||||||
| Yes (n=30,542) | 22.7% | 14.1% | 23.8% | 23.1% | 0.53(0.49,0.57) | 0.55(0.50,0.59) |
| Homelessness Status * | ||||||
| Yes (n=1,496) | 1.1% | 2.1% | 0.8% | 1.1% | 2.45(1.95,3.07) | 1.92(1.57,2.34) |
| Suicide Means | ||||||
| Firearms (n=75,527) | 56.1% | 36.5% | 52.3% | 60.1% | 0.53(0.50,0.56) | 0.40(0.37,0.42) |
| Hanging or Suffocation (n=34,058) | 25.3% | 30.7% | 26.9% | 23.9% | 1.19(1.12,1.27) | 1.38(1.30,1.46) |
| Poisoning (n=11,318) | 8.4% | 16.3% | 10.8% | 6.3% | 1.61(1.49,1.75) | 2.98(2.75,3.22) |
| Carbon Monoxide or other gas (n=4,208) | 3.1% | 2.6% | 3.3% | 3.1% | 0.79(0.66,0.94) | 0.87(0.73,1.03) |
| Drowning (n=1,089) | 0.8% | 1.4% | 0.8% | 0.7% | 1.79(1.37,2.32) | 1.69(1.35,2.14) |
| Jumping from heights (n=2,464) | 1.8% | 4.7% | 1.8% | 1.5% | 2.73(2.34,3.19) | 2.68(2.34,3.07) |
| Laceration or Sharp Object (n=2,740) | 2.0% | 3.6% | 2.1% | 1.8% | 1.78(1.50,2.10) | 2.03(1.75,2.37) |
| Vehicular-related (n=1,574) | 1.2% | 2.5% | 1.0% | 1.1% | 2.44(1.99,3.01) | 1.76(1.47,2.10) |
| Fire (n=469) | 0.3% | 0.7% | 0.3% | 0.3% | 2.21(1.48,3.30) | 1.82(1.28,2.58) |
| Other (n=216) | 0.2% | 0.3% | 0.2% | 0.1% | 1.44(0.79,2.62) | 1.82(1.04,3.18) |
| Ill-defined or unknown (n=1,024) | 0.8% | 0.9% | 0.6% | 0.8% | 1.45(1.05,2.01) | 0.87(0.65,1.16) |
Alaska, Arizona California, Colorado, Connecticut, Delaware, District of Columbia, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Michigan, Minnesota, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, Puerto Rico
Data available only since August 2013.
Boldface indicates statistical significance (p < .001).
Table 2.
Characteristics of Female Suicide Deaths (N=39,310), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2003-2017
| Total % |
Serious Mental Illness (SMI) (n=5,419) % |
Other Mental Illness (MI) (n=16,876) % |
Unknown or No Mental Illness (n=17,015) % |
Odds Ratio (99% CI) SMI compared to MI |
Odds Ratio (99% CI) SMI compared to Unknown or No Mental Illness |
|
|---|---|---|---|---|---|---|
| Sociodemographic | ||||||
| Age (years) | ||||||
| 18-24 (n=3,262) | 8.3% | 7.9% | 7.1% | 9.6% | 1.12(0.96,1.30) | 0.81(0.70,0.94) |
| 25-34 (n=5,947) | 15.1% | 18.0% | 13.4% | 16.0% | 1.42(1.27,1.58) | 1.15(1.03,1.28) |
| 35-44 (n=7,765) | 19.8% | 23.6% | 19.2% | 19.1% | 1.30(1.18,1.43) | 1.31(1.19,1.45) |
| 45-54 (n=10,167) | 25.9% | 27.6% | 27.3% | 23.8% | 1.01(0.92,1.11) | 1.22(1.11,1.33) |
| 55-64 (n=7,231) | 18.4% | 17.1% | 20.1% | 17.1% | 0.82(0.74,0.91) | 1.00(0.90,1.11) |
| 65-74 (n=3,073) | 7.8% | 4.7% | 8.5% | 8.1% | 0.53(0.45,0.64) | 0.56(0.47,0.68) |
| 75-84 (n=1,350) | 3.4% | 0.9% | 3.3% | 4.4% | 0.28(0.19,0.41) | 0.21(0.14,0.30) |
| ≥ 85 (n=515) | 1.3% | 0.1% | 1.0% | 2.0% | 0.13(0.05,0.34) | 0.06(0.02,0.17) |
| Race/Ethnicity | ||||||
| White non-Hispanic (n=33,563) | 85.4% | 85.9% | 87.7% | 82.9% | 0.85(0.76,0.96) | 1.25(1.12,1.40) |
| Black non-Hispanic (n=1,923) | 4.9% | 6.0% | 3.6% | 5.8% | 1.68(1.40,2.01) | 1.03(0.87,1.22) |
| Hispanic (n=1,672) | 4.3% | 4.1% | 3.9% | 4.6% | 1.06(0.87,1.30) | 0.88(0.87,1.08) |
| American Indian or Alaskan Native (n=479) | 1.2% | 0.7% | 0.9% | 1.7% | 0.81(0.52,1.29) | 0.42(0.27,0.65) |
| Asian American or Pacific Islander (n=1,065) | 2.5% | 2.0% | 2.5% | 3.2% | 0.78(0.59,1.04) | 0.61(0.46,0.80) |
| Other (n=586) | 1.5% | 1.3% | 1.4% | 1.6% | 0.98(0.69,1.38) | 0.82(0.58,1.15) |
| Marital Status | ||||||
| Married or Common law (n=13,588) | 33.2% | 28.7% | 36.3% | 31.5% | 0.68(0.62,0.74) | 0.83(0.76,0.91) |
| Divorced, Separated, or Widowed (n=15,122) | 37.0% | 38.5% | 36.9% | 36.5% | 1.04(0.96,1.13) | 1.04(0.95,1.12) |
| Never married or Single (n=10,194) | 28.8% | 32.0% | 26.0% | 30.6% | 1.46(1.34,1.60) | 1.20(1.10,1.31) |
| Unknown (n=381) | 1.0% | 0.7% | 0.7% | 1.3% | 1.04(0.65,1.66) | 0.56(0.36,0.88) |
| Military Veteran Status | ||||||
| Yes (n=1,159) | 2.8% | 3.0% | 2.9% | 3.0% | 1.07(0.84,1.35) | 1.01(0.80,1.28) |
| Homelessness Status | ||||||
| Yes (n=311) | 0.8% | 1.0% | 0.6% | 1.0% | 1.83(1.18,2.84) | 1.07(0.71,1.60) |
| Suicide Means | ||||||
| Firearms (n=11,990) | 30.5% | 20.5% | 28.1% | 36.0% | 0.67(0.60,0.73) | 0.47(0.43,0.51) |
| Hanging or Suffocation (n=8,962) | 22.8% | 21.2% | 22.6% | 25.5% | 0.88(0.80,0.97) | 0.84(0.77,0.92) |
| Poisoning (n=13,585) | 34.6% | 44.2% | 37.3% | 26.8% | 1.33(1.23,1.44) | 2.18(2.01,2.36) |
| Carbon Monoxide or other gas (n=1,185) | 3.0% | 2.1% | 3.0% | 3.3% | 0.71(0.54,0.93) | 0.67(0.52,0.87) |
| Drowning (n=680) | 1.7% | 1.8% | 1.8% | 1.7% | 0.98(0.73,1.33) | 1.00(0.75,1.35) |
| Jumping from heights (n=929) | 2.4% | 3.6% | 2.2% | 2.1% | 1.71(1.36,2.15) | 1.54(1.25,1.92) |
| Laceration or Sharp Object (n=655) | 1.7% | 2.2% | 1.6% | 1.5% | 1.37(1.03,1.81) | 1.50(1.14,1.98) |
| Vehicular-related (n=577) | 1.5% | 1.9% | 1.1% | 1.7% | 1.76(1.29,2.42) | 0.89(0.67,1.18) |
| Fire (n=237) | 0.6% | 0.8% | 0.5% | 0.6% | 1.65(1.02,2.66) | 1.26(0.80,1.96) |
| Other (n=89) | 0.2% | 0.2% | 0.2% | 0.2% | 0.73(0.28,1.90) | 0.67(0.26,1.68) |
| Ill-defined or unknown (n=417) | 1.1% | 1.1% | 0.8% | 1.3% | 1.55(1.04,2.31) | 0.76(0.53,1.08) |
Alaska, Arizona California, Colorado, Connecticut, Delaware, District of Columbia, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Michigan, Minnesota, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, Puerto Rico
Data available only since August 2013.
Boldface indicates statistical significance (p < .001).
3.2. Suicide Means
The most common methods of suicide among male (Table 1) SMI decedents were firearms (36.5%) and hanging/suffocation (30.7%) followed by self-poisoning (16.3%). Among females (Table 2) with SMI, self-poisoning was the most prevalent (44.2%) followed by hanging/suffocation (21.2%), and firearms (20.5%). Relative to persons with other or no known mental illness, males and females with SMI were significantly less likely to use firearms, but more likely to die from self-poisoning. Among males with SMI, odds for suicide deaths from hanging/suffocation, jumping from heights, and vehicular-related were significantly greater than males with other mental disorders or no known mental illness. For females with SMI, suicide deaths from jumping were significantly greater than for females with other disorders or no known mental illness.
3.3. Clinical Characteristics
Most SMI decedents were currently receiving treatment (66.8% of males, 75.8% of females). However, persons with SMI (36.0% of males, 38.5% of females) were significantly less likely than decedents with other mental disorders to be described as having been depressed at the time of death (48.5% of males, 45.7% of females) (Figures 1 and 2). Conversely, females with SMI, but not males, were more likely than persons with no known mental illness to be depressed. Relative to decedents with other mental illness, males and females with SMI were equally likely to have current alcohol use, but significantly more likely than persons with no known mental illness. Alternatively, odds of other substance use problems were significantly greater for males and females with SMI compared to individuals with other or no known mental illness.
Fig. 1.
Risk Estimates of Clinical Characteristics Among Male Suicide Deaths (N=134,691), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2003-2017
Fig. 2.
Risk Estimates of Clinical Characteristics Among Female Suicide Deaths (N=39,310), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2003-2017
Over one-third of males (36.8%) and one-half of females (52.9%) with SMI had prior suicide attempts with odds of past attempts being significantly higher relative to persons with other or no known mental illness (Figures 1 and 2). More than 1 in 4 (28%) of persons with SMI disclosed their suicidal intent. No difference in disclosure of intent was observed compared to decedents with other mental illness (29.3% of males, 27.0% of females), but persons with SMI were more likely than persons with no mental illness to have expressed their desires to selfharm (24.8% of males, 23.5% of females).
3.4. Precipitating Circumstances
Of the recent circumstances that preceded death from suicide (Figures 1 and 2), problems with intimate partners was the most commonly documented, observed in roughly 25% of all deaths. Among male decedents, intimate partner problems were less likely in those with SMI (23.9%) compared to those with other mental illness (28.0%) or no known mental illness (32.7%). Among female decedents, significant differences in intimate partner problems were only observed between those with SMI (22.2%) relative to no known mental illness (27.9%). The second most commonly observed problem was poor physical health, observed in approximately 22% of decedents. Males and females with SMI were significantly less likely than persons with other mental disorders or no known mental illness to have physical health problems as a contributing factor.
The third most frequently observed precipitating circumstance was recent job problem for males (13.1%) and recent argument or conflict for females (10.5%). Male decedents with SMI (11.2%) were significantly less likely than those with other mental disorders, but equally likely as those with no known mental illness, to have job problems a contributing factor. Among females, recent argument was reported for 9.7% of decedents with SMI or other mental illness, which was significantly lower than those with no known mental illness (11.5%). Financial problems were the fourth most common precipitant for both males and females.
Compared to males with no known mental illness (4.2%), males with SMI (2.6%) and other mental illness (2.2%) were significantly less likely to have recent violent behavior. Among females, those with SMI (0.8%) and other mental disorders (0.7%) were also less likely than those with no known mental illness (1.5%) to have recent violent behavior. Conversely, females with SMI (1.3%) were significantly more likely than those with other mental illness (0.9%) to have recently been a victim of violence.
3.5. Post-Mortem Toxicology
Of the 82,783 suicide deaths for whom toxicology information was available, both males (Table 3) and females with SMI (Table 4) were more likely than persons with other disorders or no known mental illness to test positive for anticonvulsant and antipsychotic medications. Although antipsychotics were more likely to be detected in persons with SMI, only 30.9% of males and 41.0% of females with SMI tested positive for antipsychotic medications. Antidepressants were detected in 45.7% of males with SMI and 65.9% of females with SMI, which was equivalent to decedents with other mental disorders, but significantly higher compared to those with no known mental illness (18.9% of males, 44.1% of females). A similar trend was observed for benzodiazepines, which were detected in 31.5% of male and 53.3% of female decedents with SMI, as compared to 19.5% of males and 39.0% of females with no known mental illness.
Table 3.
Toxicology Screening and Results of Male Suicide Deaths (N=82,783), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2013-2017
| Substance Type | Total % |
Serious Mental Illness (SMI) (n=5,479) % |
Other Mental Illness (MI) (n=23,437) % |
Unknown or No Mental Illness (n=53,967) % |
Odds Ratio (99% CI) SMI compared to MI |
Odds Ratio (99% CI) SMI compared to Unknown or No Mental Illness |
|---|---|---|---|---|---|---|
| Alcohol | ||||||
| Tested (n=41,532) | 50.2% | 58.5% | 58.0% | 46.7% | 1.06(0.98,1.14) | 1.63(1.51,1.75) |
| Positive Result (n=17,143) | 40.8% | 35.7% | 42.1% | 42.3% | 0.78(0.70,0.86) | 0.74(0.67,0.82) |
| Amphetamines | ||||||
| Tested (n=28,039) | 34.0% | 41.5% | 37.2% | 31.8% | 1.19(1.10,1.29) | 1.54(1.43,1.66) |
| Positive Result (n=2,838) | 10.1% | 14.5% | 8.9% | 10.2% | 1.73(1.44,2.07) | 1.49(1.26,1.76) |
| Anticonvulsants | ||||||
| Tested (n=15,237) | 18.9% | 25.7% | 20.9% | 16.9% | 1.30(1.19,1.43) | 1.70(1.56,1.85) |
| Positive Result (n=1,658) | 11.1% | 27.6% | 13.7% | 8.4% | 2.51(2.07,30.3) | 4.81(3.99,5.79) |
| Antidepressants | ||||||
| Tested (n=18,968) | 23.0% | 32.2% | 29.6% | 19.3% | 1.15(1.05,1.25) | 1.96(1.81,2.12) |
| Positive Result (n=5,905) | 31.7% | 45.7% | 47.3% | 18.9% | 0.93(0.81,1.07) | 3.76(3.27,4.33) |
| Antipsychotics | ||||||
| Tested (n=15,828) | 19.3% | 27.4% | 22.0% | 18.1% | 1.37(1.25,1.49) | 1.85(1.70,2.01) |
| Positive Result (n=1,228) | 8.0% | 30.9% | 8.5% | 3.5% | 4.62(3.81,5.59) | 12.31(10.02,15.11) |
| Barbiturates | ||||||
| Tested (n=22,709) | 27.4% | 34.3% | 30.3% | 25.5% | 1.20(1.10,1.30) | 1.52(1.41,1.65) |
| Positive Result (n=447) | 2.0% | 2.4% | 2.3% | 1.8% | 1.04(0.67,1.61) | 1.33(0.87,2.04) |
| Benzodiazepines | ||||||
| Tested (n=28,103) | 33.9% | 43.1% | 39.2% | 30.9% | 1.16(1.08,1.26) | 1.70(1.58,1.83) |
| Positive Result (n=6,817) | 24.5% | 31.5% | 32.3% | 19.5% | 0.95(0.84,1.08) | 1.90(1.68,2.16) |
| Cocaine | ||||||
| Tested (n=29,097) | 35.1% | 44.2% | 38.7% | 32.8% | 1.21(1.12,1.31) | 1.56(1.45,1.68) |
| Positive Result (n=2,101) | 7.2% | 6.8% | 6.1% | 7.8% | 1.16(0.92,1.48) | 0.87(0.70,1.08) |
| Marijuana | ||||||
| Tested (n=24,653) | 29.7% | 36.1% | 32.3% | 28.4% | 1.19(1.10,1.29) | 1.44(1.33,1.55) |
| Positive Result (n=5,775) | 23.5% | 29.3% | 23.7% | 24.3% | 1.34(1.16,1.55) | 1.30(1.13,1.49) |
| Opiates | ||||||
| Tested (n=30,793) | 37.2% | 45.2% | 41.6% | 34.4% | 1.16(1.07,1.25) | 1.57(1.46,1.69) |
| Positive Result (7,057) | 23.0% | 20.1% | 25.0% | 22.3% | 0.75(0.65,0.87) | 0.88(0.77,1.01) |
Table 4.
Toxicology Screening and Results of Female Suicide Deaths (N=24,327), by Type of Mental Health Diagnosis, from 37 NVDRS States and Territories,* 2013-2017
| Substance Type | Total % |
Serious Mental Illness (SMI) (n=3,024) % |
Other Mental Illness (MI) (n=9,586) % |
Unknown or No Mental Illness (n=11,747) % |
Odds Ratio (99% CI) SMI compared to MI |
Odds Ratio (99% CI) SMI compared to Unknown or No Mental Illness |
|---|---|---|---|---|---|---|
| Alcohol | ||||||
| Tested (n=13,557) | 55.7% | 61.7% | 62.5% | 48.6% | 0.96(0.86,1.08) | 1.70(1.53,1.90) |
| Positive Result (n=4,889 | 36.2% | 30.9% | 36.1% | 38.1% | 0.79(0.68,0.92) | 0.73(0.63,0.74) |
| Amphetamines | ||||||
| Tested (n=9,712) | 39.9% | 46.1% | 43.9% | 35.0% | 1.09(0.98,1.22) | 1.58(1.42,1.76) |
| Positive Result (n=1,0787) | 11.2% | 14.8% | 9.8% | 11.4% | 1.59(1.25,2.01) | 1.34(1.06,1.69) |
| Anticonvulsants | ||||||
| Tested (n=6,563) | 27.0% | 35.7% | 30.0% | 22.2% | 1.30(1.16,1.45) | 1.94(1.74,2.18) |
| Positive Result (n=1,855) | 28.4% | 46.6% | 27.8% | 21.4% | 2.26(1.87,2.74) | 3.20(2.62,3.91) |
| Antidepressants | ||||||
| Tested (n=8,934) | 36.7% | 47.4% | 44.6% | 27.5% | 1.12(1.00,1.25) | 2.38(2.13,2.65) |
| Positive Result (n=5,227) | 58.8% | 65.9% | 67.4% | 44.1% | 0.93(0.79,1.10) | 2.45(2.07,2.91) |
| Antipsychotics | ||||||
| Tested (n=6,504) | 36.9% | 37.9% | 30.1% | 21.4% | 1.42(1.27,1.58) | 2.24(2.00,2.51) |
| Positive Result (n=1,202) | 18.6% | 41.0% | 16.3% | 11.2% | 3.57(2.92,4.38) | 5.49(4.38,6.88) |
| Barbiturates | ||||||
| Tested (n=8,353) | 34.3% | 39.7% | 38.1% | 29.8% | 1.07(0.96,1.20) | 1.55(1.39,1.73) |
| Positive Result (n=450) | 5.4% | 5.5% | 6.4% | 4.5% | 0.86(0.59,1.25) | 1.25(0.65,1.84) |
| Benzodiazepines | ||||||
| Tested (n=11,268) | 46.3% | 55.2% | 52.6% | 38.9% | 1.11(1.00,1.24) | 1.93(1.74,2.15) |
| Positive Result (n=5,165) | 46.0% | 53.3% | 50.0% | 39.0% | 1.13(0.98,1.31) | 1.77(1.52,2.05) |
| Cocaine | ||||||
| Tested (n=10,425) | 42.9% | 49.8% | 47.7% | 37.2% | 1.09(0.98,1.21) | 1.68(1.51,1.86) |
| Positive Result (n=573) | 5.5% | 6.0% | 4.7% | 6.3% | 1.31(0.94,1.83) | 0.96(0.70,1.33) |
| Marijuana | ||||||
| Tested (n=8,258) | 33.9% | 37.4% | 36.8% | 30.8% | 1.03(0.92,1.15) | 1.34(1.20,1.50) |
| Positive Result (n=1,410) | 17.2% | 21.0% | 15.7% | 17.4% | 1.43(1.14,1.78) | 1.27(1.12,1.58) |
| Opiates | ||||||
| Tested (n=11,605) | 47.7% | 54.4% | 53.3% | 41.3% | 1.04(0.94,1.16) | 1.69(1.52,1.88) |
| Positive Result (4,317) | 37.4% | 35.7% | 39.0% | 36.3% | 0.87(0.75,1.04) | 0.98(0.84,1.14) |
Toxicology results also revealed persons with SMI were significantly less likely than those with other or no known mental illness to test positive for alcohol. Conversely, higher odds were observed for positive results for amphetamines in males and females with SMI compared to those with other mental disorders. Males with SMI were less likely to test positive for cocaine relative to those with no known mental illness. Males and females with SMI were more likely to test positive for marijuana relative to decedents with other mental disorders. Opiates were detected in significantly fewer males with SMI relative to persons with other mental disorders, whereas no significant differences in positive testing for opiates were observed in females.
4. Discussion
The current study used national US data to examine the characteristics and precipitating circumstances of nearly 175,000 suicide deaths to compare individuals with serious mental illness (SMI) to decedents with other mental disorders and those with no known mental illness. These results suggest that individuals with SMI, particularly bipolar disorder, are over-represented in suicide deaths. Despite representing approximately 4% of the US population,(Latzman et al., 2019) persons with SMI account for over twice the percentage (8.7%) of deaths from suicide, as captured in the National Violent Death Reporting System (NVDRS).
4.1. Strengths and limitations
To our knowledge, this is the largest study to examine the circumstances surrounding suicide among the SMI population in the US. The NVDRS is the only US-based system that uniformly collects and links information about a range of characteristics of persons who died from suicide, including demographic information, mental health and substance use history, a range of recent life events, and post-mortem toxicology results. Because of its national scope and duration of data collection, the NVDRS has sufficient statistical power to make between-group comparisons regarding suicide mortality suicide.
This study has several limitations. First, mental health diagnoses cannot be validated and are not based on standardized or structured assessments. All data in the NVDRS are based on abstractors coding of information from reports by coroner/medical examiners and law enforcement, which are prone to inaccurate or incomplete data.(Cheung et al., 2015) Furthermore, these reports are based on unstandardized interviews with informants that are not performed by trained researchers or clinicians; therefore, informant and interviewer bias and misattribution of precipitating circumstances are possible.(Kaplan et al., 2017) Second, NVDRS data abstractors’ selection of “no” for current mental illness and precipitating circumstances is not distinguishable from “unknown” or “unavailable.” Thus, it is likely that these data contain false negatives for mental illness and other contributing factors. Third, NVDRS data do not contain information about several clinically important factors, including SMI decedents’ clinical status (e.g., active psychosis versus period of recovery) and the degree of the treatment participation which influence risk of suicidal behavior.(Lopez-Morinigo et al., 2019) Fourth, toxicology results were not available for a substantial proportion of decedents and large variations exist in toxicology practices.(Stone et al., 2017) Finally, data for this study are limited to the 37 US states and territories that participated in the NVDRS and may not be nationally representative or generalizable to other countries.
4.2. Similarities and differences between decedents with SMI and other or no known mental illness
Despite some demographic and clinical similarities between suicide decedents with and without SMI (e.g., predominately Caucasian and male),(CDC, 2019) these results revealed a number of important differences among suicide decedents with SMI relative to those with other or no known mental illness. First, disproportionately more females with SMI died from suicide relative to males with SMI. This finding may reflect the higher prevalence of SMI in females.(Latzman et al., 2019) Second, in both males and females, decedents with SMI were more likely to be younger and African American compared to those with other mental disorders, which is consistent with other research.(Hayes et al., 2015; Hor and Taylor, 2010; Lopez-Morinigo et al., 2014) Decedents with SMI were as likely as persons with other mental disorders to have problems with alcohol use, but significantly more likely to have other substance use problems. This is consistent with also other studies. (Østergaard et al., 2017; Yoon et al., 2011)
Firearms were the most common suicide method among males with SMI, whereas self-poisoning accounted for a majority of suicide deaths in females with SMI. These findings correspond with national US statistics for suicide methods by gender.(CDC, 2019) However, this observation differs from older US research and analyses from other countries for which hanging or jumping from heights were the most common methods for persons with SMI.(De Hert et al., 2001; Kreyenbuhl et al., 2002; Lopez-Morinigo et al., 2014; Nah et al., 2017; Westermeyer et al., 1991) The reduced odds of suicide deaths from firearms among persons with SMI relative to persons with other or no known mental illness has been observed previously(Swanson et al., 2016) and may be associated with federal and state laws prohibiting sales and possession of firearms among persons who have been hospitalized, institutionalized, or adjudicated for mental illness.
Problems with intimate relationships was the most frequently observed precipitating life stressor for decedents with SMI, other mental disorders, and no known mental illness. Problems with physical health were also commonly recognized, but appeared to play less of a role among decedents with SMI, despite considerable evidence of excess physical morbidity in this population.(De Hert et al., 2011) SMI decedents were significantly more likely than persons with other mental disorders to have been released within the past 30 days from psychiatric hospitalization. The first months following hospital discharge is one of critical suicide risk,(Chung et al., 2019; Chung et al., 2017) especially for persons with schizophrenia and bipolar disorder.(Olfson et al., 2016)
Results showing a high prevalence of prior suicide attempts among SMI decedents, even relative to those with other mental disorders, is consistent with prior research.(Cassidy et al., 2017; Gallego et al., 2015; Hansson et al., 2018; Pompili et al., 2013) Prior suicide attempt is one of the most powerful known clinical risk factor for suicide mortality,(Franklin et al., 2016; Ribeiro et al., 2016) including among persons with SMI.(Fazel et al., 2019b)
4.3. Medications among SMI decedents
Post-mortem toxicology revealed antipsychotic medication was detected in a minority of decedents with SMI (30.9% of males, 41.0% of females). Antipsychotic use reduces risk of suicide and all-cause mortality,(Reutfors et al., 2013; Taipale et al., 2020) whereas partial and non-adherence to antipsychotic medication are associated with a six- to twelve-fold increased risk of suicide.(Forsman et al., 2019) Anticonvulsants have been shown to reduce psychopathology in persons with SMI (Tseng et al., 2016) and to reduce suicide risk in persons with bipolar disorder (Smith and Cipriani, 2017; Song et al., 2017) but were detected in only 27.6% of men and 46.6% of females with SMI. Gender differences in positive toxicology for antipsychotic and anticonvulsant medications may be associated with higher likelihood of medication adherence in females with bipolar and psychotic disorders.(Higashi et al., 2013; Leclerc et al., 2013)
Roughly one-half of males and two-thirds of females with SMI tested positive for antidepressant medication. Depressive symptoms during first episode psychosis is associated with greater risk of suicide, particularly within two years (McGinty et al., 2018) and antidepressants reduce the risk of suicide in persons with schizophrenia.(Tiihonen et al., 2012) Alternatively, benzodiazepine use in persons with schizophrenia is associated with a two- to four-fold increase in risk of suicide death (Fontanella et al., 2016; Stroup et al., 2019; Tiihonen et al., 2012) and was detected in 31.1% of males and 53.1% of females with SMI.
5. Conclusions
The results from this study confirm a marked excess of suicide deaths in persons with schizophrenia, bipolar disorder, and other serious mental illnesses that is evident in early and middle adulthood. Although the current analysis of data from the NVDRS sheds some light on environmental precursors to suicide, further research is needed to achieve a better understanding of the interplay between neurological and environmental factors that may heighten suicide risk in persons with mental illness.(Malhi et al., 2020a; Malhi et al., 2020b) The current findings support efforts to improve treatment adherence to psychotropic medications that reduce suicide risk and concerns about pharmacotherapy that may heighten risk. Concerted efforts should be made to improve the access and quality of services to treat co-occurring substance use, particularly drugs of abuse, and continuity of care shortly following hospital discharge. Other strategies, such as legislation that reduces access to handguns, may also help reduce suicide in this population.(Anestis and Anestis, 2015; Anestis et al., 2019; Anestis et al., 2017) Recovery-oriented care, which focuses on fostering empowerment and hope, is associated with personal recovery in persons with serious mental illness (Loos et al., 2017; Thomas et al., 2018) and may contribute to suicide prevention efforts.
Acknowledgments
Role of funding source
Dr. Schmutte acknowledges support from the National Institute of Mental Health, NIH (grant R01MH107452-02S1). The supporters had no role in the design, analysis, interpretation, or publication in this study.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of interest
The authors declare that we have no conflict of interest.
References
- Abdeen MS, Shaker NM, Elrassas HH, Hashim MA, Abo Zeid MY, 2019. Characteristics of the schizophrenia suicide attempts in comparison with the suicide attempts with other diagnosed psychiatric disorders: An Egyptian study. Int. J. Soc. Psychiatry 65(5), 368–377. [DOI] [PubMed] [Google Scholar]
- Anestis MD, Anestis JC, 2015. Suicide Rates and State Laws Regulating Access and Exposure to Handguns. Am. J. Public Health 105(10), 2049–2058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Anestis MD, Houtsma C, Daruwala SE, Butterworth SE, 2019. Firearm legislation and statewide suicide rates: The moderating role of household firearm ownership levels. Behav. Sci. Law 37(3), 270–280. [DOI] [PubMed] [Google Scholar]
- Anestis MD, Selby EA, Butterworth SE, 2017. Rising longitudinal trajectories in suicide rates: The role of firearm suicide rates and firearm legislation. Prev. Med 100, 159–166. [DOI] [PubMed] [Google Scholar]
- Blair JM, Fowler KA, Jack SP, Crosby AE, 2016. The national violent death reporting system: overview and future directions. Inj. Prev 22(Suppl 1), i6–i11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cassidy RM, Yang F, Kapczinski F, Passos IC, 2017. Risk factors for suicidality in patients with schizophrenia: a systematic review, meta-analysis, and meta-regression of 96 studies. Schizophr. Bull 44(4), 787–797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- CDC, 2019. Health, United States, 2018., National Center for Health Statistics, Centers for Disease Control, Hyattsville, MD. . [Google Scholar]
- Chen T, Roberts K, 2019. Negative Life Events and Suicide in the National Violent Death Reporting System. Archives of Suicide Research, 1–15. [DOI] [PubMed] [Google Scholar]
- Chesney E, Goodwin GM, Fazel S, 2014. Risks of all-cause and suicide mortality in mental disorders: a meta-review. World Psychiatry 13(2), 153–160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cheung G, Merry S, Sundram F, 2015. Medical examiner and coroner reports: uses and limitations in the epidemiology and prevention of late-life suicide. Int. J. Geriatr. Psychiatry 30(8), 781–792. [DOI] [PubMed] [Google Scholar]
- Chung D, Hadzi-Pavlovic D, Wang M, Swaraj S, Olfson M, Large M, 2019. Meta-analysis of suicide rates in the first week and the first month after psychiatric hospitalisation. BMJ open 9(3), e023883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chung DT, Ryan CJ, Hadzi-Pavlovic D, Singh SP, Stanton C, Large MM, 2017. Suicide rates after discharge from psychiatric facilities: a systematic review and meta-analysis. JAMA psychiatry, 694–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- De Hert M, Cornell CU, Bobes J, Cetkovich-Bakmas M, Cohen D, Asai I, Detraux J, Gautam S, MÖLLER HJ, Ndetei DM, 2011. Physical illness in patients with severe mental disorders. I. Prevalence, impact of medications and disparities in health care. World psychiatry 10(1), 52–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- De Hert M, McKenzie K, Peuskens J, 2001. Risk factors for suicide in young people suffering from schizophrenia: a long-term follow-up study. Schizophr. Res. 47(2), 127–134. [DOI] [PubMed] [Google Scholar]
- Fazel S, Wolf A, Larsson H, Mallett S, Fanshawe TR, 2019a. The prediction of suicide in severe mental illness: development and validation of a clinical prediction rule (OxMIS). Translational psychiatry 9(1), 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fazel S, Wolf A, Larsson H, Mallett S, Fanshawe TR, 2019b. The prediction of suicide in severe mental illness: development and validation of a clinical prediction rule (OxMIS). Translational psychiatry 9(1), 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fontanella CA, Campo JV, Phillips GS, Hiance-Steelesmith DL, Sweeney HA, Tam K, Lehrer D, Klein R, Hurst M, 2016. Benzodiazepine use and risk of mortality among patients with schizophrenia: a retrospective longitudinal study. The Journal of clinical psychiatry 77(5), 661–667. [DOI] [PubMed] [Google Scholar]
- Forsman J, Taipale H, Masterman T, Tiihonen J, Tanskanen A, 2019. Adherence to psychotropic medication in completed suicide in Sweden 2006–2013: a forensic-toxicological matched case-control study. Eur. J. Clin. Pharmacol, 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Franklin JC, Ribeiro JD, Fox KR, Bentley KH, Kleiman EM, Huang X, Musacchio KM, Jaroszewski AC, Chang BP, Nock MK, 2016. Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychol. Bull 143(2), 187–232. [DOI] [PubMed] [Google Scholar]
- Gallego JA, Rachamallu V, Yuen EY, Fink S, Duque LM, Kane JM, 2015. Predictors of suicide attempts in 3.322 patients with affective disorders and schizophrenia spectrum disorders. Psychiatry Res. 228(3), 791–796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hansson C, Joas E, Pålsson E, Hawton K, Runeson B, Landén M, 2018. Risk factors for suicide in bipolar disorder: a cohort study of 12 850 patients. Acta Psychiatr. Scand 138(5), 456–463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hayes J, Miles J, Walters K, King M, Osborn D, 2015. A systematic review and meta-analysis of premature mortality in bipolar affective disorder. Acta Psychiatr. Scand 131(6), 417–425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heilä H, Heikkinen ME, Isometsä ET, Henriksson MM, Marttunen MJ, Lönnqvist JK, 1999. Life events and completed suicide in schizophrenia: a comparison of suicide victims with and without schizophrenia. Schizophr. Bull 25(3), 519–531. [DOI] [PubMed] [Google Scholar]
- Hettige NC, Bani-Fatemi A, Sakinofsky I, De Luca V, 2017. A biopsychosocial evaluation of the risk for suicide in schizophrenia. CNS Spectrums 23(4), 253–263. [DOI] [PubMed] [Google Scholar]
- Higashi K, Medic G, Littlewood KJ, Diez T, Granström O, De Hert M, 2013. Medication adherence in schizophrenia: factors influencing adherence and consequences of nonadherence, a systematic literature review. Therapeutic advances in psychopharmacology 3(4), 200–218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hor K, Taylor M, 2010. Suicide and schizophrenia: a systematic review of rates and risk factors. Journal of psychopharmacology 24(4_suppl), 81–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kaplan MS, Caetano R, Giesbrecht N, Huguet N, Kerr WC, McFarland BH, Nolte KB, 2017. The National Violent Death Reporting System: use of the restricted access database and recommendations for the system’s improvement. Am. J. Prev. Med 53(1), 130–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kreyenbuhl JA, Kelly DL, Conley RR, 2002. Circumstances of suicide among individuals with schizophrenia. Schizophr. Res 58(2-3), 253–261. [DOI] [PubMed] [Google Scholar]
- Latzman NE, Ringeisen H, Forman–Hoffman VL, Munoz B, Miller S, Hedden SL, 2019. Trends in mental health service use by age among adults with serious mental illness. Ann. Epidemiol 30, 71–73. [DOI] [PubMed] [Google Scholar]
- Leclerc E, Mansur RB, Brietzke E, 2013. Determinants of adherence to treatment in bipolar disorder: a comprehensive review. J. Affect. Disord 149(1-3), 247–252. [DOI] [PubMed] [Google Scholar]
- Lin J-J, Liang F-W, Li C-Y, Lu T-H, 2018. Leading causes of death among decedents with mention of schizophrenia on the death certificates in the United States. Schizophr. Res 197, 116–123. [DOI] [PubMed] [Google Scholar]
- Loos S, Clarke E, Jordan H, Puschner B, Fiorillo A, Luciano M, Ivánka T, Magyar E, Krogsgaard-Bording M, Østermark-Sørensen H, Rössler W, Kawohl W, Mayer B, Slade M, Group CS, 2017. Recovery and decision-making involvement in people with severe mental illness from six countries: a prospective observational study. BMC Psychiatry 17(1), 38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lopez-Morinigo J-D, Di Forti M, Ajnakina O, Wiffen BD, Morgan K, Doody GA, Jones PB, Ayesa-Arriola R, Canal-Rivero M, Crespo-Facorro B, Murray RM, Dazzan P, Morgan C, Dutta R, David AS, 2019. Insight and risk of suicidal behaviour in two first-episode psychosis cohorts: Effects of previous suicide attempts and depression. Schizophr. Res 204, 80–89. [DOI] [PubMed] [Google Scholar]
- Lopez-Morinigo J-D, Fernandes AC, Chang C-K, Hayes RD, Broadbent M, Stewart R, David AS, Dutta R, 2014. Suicide completion in secondary mental healthcare: a comparison study between schizophrenia spectrum disorders and all other diagnoses. BMC Psychiatry 14(1), 213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lyu J, Zhang J, 2014. Characteristics of schizophrenia suicides compared with suicides by other diagnosed psychiatric disorders and those without a psychiatric disorder. Schizophr. Res 155(1), 59–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Malhi GS, Bell E, Mannie Z, Das P, 2020a. Attempting suicide changes the brain? SAGE Publications Sage UK: London, England. [DOI] [PubMed] [Google Scholar]
- Malhi GS, Das P, Outhred T, Bryant RA, Calhoun V, Mann JJ, 2020b. Default mode dysfunction underpins suicidal activity in mood disorders. Psychol. Med 50(7), 1214–1223. [DOI] [PubMed] [Google Scholar]
- Malhi GS, Outhred T, Das P, Morris G, Hamilton A, Mannie Z, 2018. Modeling suicide in bipolar disorders. Bipolar disorders 20(4), 334–348. [DOI] [PubMed] [Google Scholar]
- McGinty J, Haque MS, Upthegrove R, 2018. Depression during first episode psychosis and subsequent suicide risk: a systematic review and meta-analysis of longitudinal studies. Schizophr. Res 195, 58–66. [DOI] [PubMed] [Google Scholar]
- Nah G, Choi S, Kim H, Lee J-Y, Kim J-M, Shin I-S, Yoon J-S, Kim S-W, 2017. Characteristics of patients who died by suicide in a community mental health center. Korean Journal of Schizophrenia Research 20(2), 55–60. [Google Scholar]
- NCSL.org, 2018. National Conference of State Legislatures (1/5/2018). Possession of Firearms by People with Mental Illness. Last accessed 1/28/2020. Website available at: https://www.ncsl.org/research/civil-and-criminal-justice/possession-of-a-firearm-by-the-mentally-ill.aspx. [Google Scholar]
- Olfson M, Wall M, Wang S, Crystal S, Liu S-M, Gerhard T, Blanco C, 2016. Short-term suicide risk after psychiatric hospital discharge. JAMA psychiatry 73(11), 1119–1126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Østergaard ML, Nordentoft M, Hjorthøj C, 2017. 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 112(7), 1250–1259. [DOI] [PubMed] [Google Scholar]
- Pompili M, Gonda X, Serafini G, Innamorati M, Sher L, Amore M, Rihmer Z, Girardi P, 2013. Epidemiology of suicide in bipolar disorders: a systematic review of the literature. Bipolar disorders 15(5), 457–490. [DOI] [PubMed] [Google Scholar]
- Reutfors J, Bahmanyar S, Jönsson EG, Brandt L, Bodén R, Ekbom A, Ösby U, 2013. Medication and suicide risk in schizophrenia: A nested case–control study. Schizophr. Res 150(2-3), 416–420. [DOI] [PubMed] [Google Scholar]
- Ribeiro J, Franklin J, Fox KR, Bentley K, Kleiman EM, Chang B, Nock MK, 2016. Self-injurious thoughts and behaviors as risk factors for future suicide ideation, attempts, and death: a meta-analysis of longitudinal studies. Psychol. Med 46(2), 225–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schaffer A, Isometsä ET, Tondo L, H Moreno D, Turecki G, Reis C, Cassidy F, Sinyor M, Azorin JM, Kessing LV, 2015. International Society for Bipolar Disorders Task Force on Suicide: meta-analyses and meta-regression of correlates of suicide attempts and suicide deaths in bipolar disorder. Bipolar disorders 17(1), 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith KA, Cipriani A, 2017. Lithium and suicide in mood disorders: updated meta-review of the scientific literature. Bipolar disorders 19(7), 575–586. [DOI] [PubMed] [Google Scholar]
- Song J, Sjölander A, Joas E, Bergen SE, Runeson B, Larsson H, Landén M, Lichtenstein P, 2017. Suicidal behavior during lithium and valproate treatment: a within-individual 8-year prospective study of 50,000 patients with bipolar disorder. Am. J. Psychiatry 174(8), 795–802. [DOI] [PubMed] [Google Scholar]
- Stone DM, Holland KM, Bartholow B, Logan JE, , LiKamWa McIntosh W, Trudeau A, Rockett IR, 2017. Deciphering suicide and other manners of death associated with drug intoxication: a Centers for Disease Control and Prevention consultation meeting summary. Am. J. Public Health 107(8), 1233–1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stroup TS, Gerhard T, Crystal S, Huang C, Tan Z, Wall MM, Mathai C, Olfson M, 2019. Comparative Effectiveness of Adjunctive Psychotropic Medications in Patients With Schizophrenia. JAMA Psychiatry 76(5), 508–515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Swanson JW, Easter MM, Robertson AG, Swartz MS, Alanis-Hirsch K, Moseley D, Dion C, Petrila J, 2016. Gun violence, mental illness, and laws that prohibit gun possession: evidence from two Florida counties. Health Aff. (Millwood) 35(6), 1067–1075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Taipale H, Tanskanen A, Mehtälä J, Vattulainen P, Correll CU, Tiihonen J, 2020. 20-year follow-up study of physical morbidity and mortality in relationship to antipsychotic treatment in a nationwide cohort of 62,250 patients with schizophrenia (FIN20). World Psychiatry 19(1), 61–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thomas EC, Despeaux KE, Drapalski AL, Bennett M, 2018. Person-oriented recovery of individuals with serious mental illnesses: A review and meta-analysis of longitudinal findings. Psychiatr. Serv 69(3), 259–267. [DOI] [PubMed] [Google Scholar]
- Tiihonen J, Suokas JT, Suvisaari JM, Haukka J, Korhonen P, 2012. Polypharmacy with antipsychotics, antidepressants, or benzodiazepines and mortality in schizophrenia. Arch. Gen. Psychiatry 69(5), 476–483. [DOI] [PubMed] [Google Scholar]
- Tseng P-T, Chen Y-W, Chung W, Tu K-Y, Wang H-Y, Wu C-K, Lin P-Y, 2016. Significant Effect of Valproate Augmentation Therapy in Patients With Schizophrenia: A Meta-analysis Study. Medicine 95(4), e2475–e2475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Westermeyer JF, Harrow M, Marengo JT, 1991. Risk for suicide in schizophrenia and other psychotic and nonpsychotic disorders. J. Nerv. Ment. Dis [DOI] [PubMed] [Google Scholar]
- Yoon Y-H, Chen CM, Yi H.-y., Moss HB, 2011. Effect of comorbid alcohol and drug use disorders on premature death among unipolar and bipolar disorder decedents in the United States, 1999 to 2006. Compr. Psychiatry 52(5), 453–464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zaheer J, Jacob B, de Oliveira C, Rudoler D, Juda A, Kurdyak P, 2018. Service utilization and suicide among people with schizophrenia spectrum disorders. Schizophr. Res 202, 347–353. [DOI] [PubMed] [Google Scholar]