Decoding Suicide Decedent Profiles and Signs of Suicidal Intent Using Latent Class Analysis

Yunyu Xiao; Kaiwen Bi; Paul Siu-Fai Yip; Julie Cerel; Timothy T Brown; Yifan Peng; Jyotishman Pathak; J John Mann

doi:10.1001/jamapsychiatry.2024.0171

. 2024 Mar 20;81(6):595–605. doi: 10.1001/jamapsychiatry.2024.0171

Decoding Suicide Decedent Profiles and Signs of Suicidal Intent Using Latent Class Analysis

Yunyu Xiao ^1,^✉, Kaiwen Bi ², Paul Siu-Fai Yip ^2,³, Julie Cerel ⁴, Timothy T Brown ⁵, Yifan Peng ¹, Jyotishman Pathak ¹, J John Mann ^6,^7,⁸

¹Department of Population Health Sciences, Weill Cornell Medicine/NewYork-Presbyterian, New York

²Department of Social Work and Social Administration, The University of Hong Kong, Hong Kong

³Hong Kong Jockey Club Centre for Suicide Research and Prevention, The University of Hong Kong, Hong Kong

⁴College of Social Work, University of Kentucky, Lexington

⁵School of Public Health, University of California, Berkeley

⁶Department of Psychiatry, Columbia University Irving Medical Center, Columbia University, New York, New York

⁷Department of Radiology, Columbia University Irving Medical Center, Columbia University, New York, New York

⁸Division of Molecular Imaging and Neuropathology, New York State Psychiatric Institute, New York

Accepted for Publication: January 7, 2024.

Published Online: March 20, 2024. doi:10.1001/jamapsychiatry.2024.0171

^✉

Corresponding Author: Yunyu Xiao, PhD, Weill Cornell Medicine/NewYork-Presbyterian, Department of Population Health Sciences, DV 306, 425 East 61 St, New York, NY 10065 (yux4008@med.cornell.edu).

Author Contributions: Dr Xiao and Mr Bi had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr Xiao and Mr Bi contributed equally to this work as co–first authors.

Concept and design: Xiao, Yip, Cerel, Mann.

Acquisition, analysis, or interpretation of data: Xiao, Bi, Yip, Brown, Peng, Pathak, Mann.

Drafting of the manuscript: Xiao, Bi, Cerel, Pathak.

Critical review of the manuscript for important intellectual content: Xiao, Bi, Yip, Brown, Peng, Pathak, Mann.

Statistical analysis: Xiao, Bi, Yip, Brown, Peng.

Obtained funding: Xiao, Yip, Peng.

Administrative, technical, or material support: Xiao, Pathak.

Supervision: Xiao, Mann.

Conflict of Interest Disclosures: Dr Mann reported royalties for commercial use of the Columbia-Suicide Severity Rating Scale from the Research Foundation for Mental Hygiene. No other disclosures were reported.

Funding/Support: This study was supported by funding from the Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity (AIM-AHEAD), a program of the National Institutes of Health (1OT2OD032581-02-259; Drs Xiao, Peng, and Pathak), and grants from the National Institute of Mental Health (RF1MH134649), American Foundation for Suicide Prevention (YIG-2-133-22), Mental Health Research Network, Google, Bill & Melinda Gates Foundation, National Institute for Health Care Management, National Institute on Drug Abuse Center for Health Economics of Treatment Interventions for Substance Use Disorder, HCV, and HIV (CHERISH), and Suicide Prevention During COVID-19 (C7151-20G; Dr Yip and Mr Bi).

Role of the Funder/Sponsor: The supporting organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.

^✉

Corresponding author.

PMCID: PMC10955339 PMID: 38506817

This cross-sectional study examines data from the 2003-2020 National Violent Death Reporting System to identify distinct suicide profiles among US suicide decedents and determine associations to signs of suicidal intent and risks.

Key Points

Question

Are there distinct suicide profiles linked to varying signs of suicidal intent and risks?

Findings

This cross-sectional study including 306 800 US suicide decedents from 2003-2020 identified 5 suicide profiles or classes: comorbid mental health and substance use disorders; mental disorders alone; crisis, alcohol-related, and intimate partner problems; physical health problems; and polysubstance use. Class 4 (physical health problems) was the largest, exhibiting minimal explicit suicidal intent, rarely detected psychiatric diagnoses, and minimal psychotropic medication use.

Meaning

Suicide prevention strategies may be more effective when tailored to different suicide profiles because integrated care enhances the detection and treatment of comorbid mental health conditions, substance and alcohol use disorders, and physical health problems.

Abstract

Importance

Suicide rates in the US increased by 35.6% from 2001 to 2021. Given that most individuals die on their first attempt, earlier detection and intervention are crucial. Understanding modifiable risk factors is key to effective prevention strategies.

Objective

To identify distinct suicide profiles or classes, associated signs of suicidal intent, and patterns of modifiable risks for targeted prevention efforts.

Design, Setting, and Participants

This cross-sectional study used data from the 2003-2020 National Violent Death Reporting System Restricted Access Database for 306 800 suicide decedents. Statistical analysis was performed from July 2022 to June 2023.

Exposures

Suicide decedent profiles were determined using latent class analyses of available data on suicide circumstances, toxicology, and methods.

Main Outcomes and Measures

Disclosure of recent intent, suicide note presence, and known psychotropic usage.

Results

Among 306 800 suicide decedents (mean [SD] age, 46.3 [18.4] years; 239 627 males [78.1%] and 67 108 females [21.9%]), 5 profiles or classes were identified. The largest class, class 4 (97 175 [31.7%]), predominantly faced physical health challenges, followed by polysubstance problems in class 5 (58 803 [19.2%]), and crisis, alcohol-related, and intimate partner problems in class 3 (55 367 [18.0%]), mental health problems (class 2, 53 928 [17.6%]), and comorbid mental health and substance use disorders (class 1, 41 527 [13.5%]). Class 4 had the lowest rates of disclosing suicidal intent (13 952 [14.4%]) and leaving a suicide note (24 351 [25.1%]). Adjusting for covariates, compared with class 1, class 4 had the highest odds of not disclosing suicide intent (odds ratio [OR], 2.58; 95% CI, 2.51-2.66) and not leaving a suicide note (OR, 1.45; 95% CI, 1.41-1.49). Class 4 also had the lowest rates of all known psychiatric illnesses and psychotropic medications among all suicide profiles. Class 4 had more older adults (23 794 were aged 55-70 years [24.5%]; 20 100 aged ≥71 years [20.7%]), veterans (22 220 [22.9%]), widows (8633 [8.9%]), individuals with less than high school education (15 690 [16.1%]), and rural residents (23 966 [24.7%]).

Conclusions and Relevance

This study identified 5 distinct suicide profiles, highlighting a need for tailored prevention strategies. Improving the detection and treatment of coexisting mental health conditions, substance and alcohol use disorders, and physical illnesses is paramount. The implementation of means restriction strategies plays a vital role in reducing suicide risks across most of the profiles, reinforcing the need for a multifaceted approach to suicide prevention.

Introduction

Suicide rates in the United States increased by 35.6% from 10.4 to 14.1 per 100 000 per year between 2001 and 2021.¹ This increase is particularly concerning given that up to 79% of suicide decedents die on their first attempt,^2,3,4 underscoring the need for effective early detection and intervention.⁵

The multifaceted nature of suicide, as evidenced by the absence of a single profile,^5,6 challenges traditional prevention efforts that predominantly focus on screening known mental disorders.⁷ Although psychological autopsy studies indicate 70% to 90% of suicides involved a mental disorder,^8,9,10,11,12 only about 20% of suicide decedents with a psychiatric disorder before death had engaged with mental health services.¹³ This suggests a gap in prevention strategies: identifying individuals with mental disorders in mental health settings is insufficient,¹⁴ as that approach fails to reach most of those at risk. Among varying lethality of methods, self-inflicted gunshots have a survival rate of less than 10% compared with a 92% survival rate for drug overdoses, which adds complexity.¹⁵ Suicide decedents using high-lethality methods were less likely to seek psychiatric treatment or have a history of previous nonfatal suicide attempts.¹⁶ Thus, suicide prevention in this high-risk subgroup must begin before the first suicide attempt and outside mental health service settings.

To improve prevention, a better understanding of modifiable risk factors is imperative.^2,5 One approach is identifying “suicide decedent profiles” with a broader spectrum of risk factors: seeking at-risk subgroups that may not manifest traditional warning signs,¹⁷ such as individuals with diagnosed mental health disorders, psychiatric medications, nonfatal suicide attempts, or disclosure of suicidal intent.^17,18 Instead of a 1-size-fits-all approach, identifying heterogeneous suicide decedent profiles may inform prevention strategies tailored for high-risk populations who might die at their first attempt.¹⁹

Previous efforts using this approach have encountered limitations. First, our study, leveraging the National Violent Death Record System (NVDRS), addresses a critical gap in suicide research by focusing on data typically absent in clinical settings. This approach is necessary and innovative, because within 1 month of death, twice as many individuals (40%) had consulted general health care professionals²⁰ compared with mental health professionals.¹³ Second, while it is labor intensive to use psychological autopsies to detect mental disorders, we used toxicology data from medical/legal examiners to assess psychotropic treatment adherence in suicide decedents.^2,21 This information, coupled with suicide method details from autopsy records, offers a more comprehensive understanding of suicide decedent profiles than previous studies.^22,23,24 Third, most US studies lacked national representation, with previous research using NVDRS datasets limited to 17 states.²⁴ Fourth, despite evidence of disparities in suicide deaths and precipitating circumstances across age, race and ethnicity, sex, and geography,^6,22 mapping these characteristics onto suicide profiles remains underexplored.

Our study addresses these gaps by discerning suicide profiles through a comprehensive analysis of variables, including circumstances of death, toxicology, and suicide methods. We analyzed all suicides from 2003 to 2020 in the NVDRS Restricted Access Database. Additionally, we associated suicides with suicide intent disclosure, suicide notes, and psychotropic use while examining sociodemographic differences. Our data-driven approach led us to hypothesize distinct suicide profiles that differ regarding potentially modifiable risk factors and would require different combinations of evidence-based prevention measures in nonpsychiatric and psychiatric clinical settings.

Methods

Study Design, Setting, and Population

The NVDRS is a national surveillance system for violent deaths,²⁵ the largest dataset on US suicide decedents. Beginning in 2003 in 6 states and expanding to 50 states, the District of Columbia, and Puerto Rico by 2018,²⁶ the NVDRS Restricted Access Database compiles information from death certificates, coroner and medical examiner records, and toxicology reports. Abstractors review data to ensure accuracy.²⁷ The system is suitable for identifying suicide profiles (eMethods 1 in Supplement 1). We included all 306 800 suicide decedents at all ages from 2003 to 2020. Statistical analyses were performed from July 2022 to June 2023. Weill Cornell Medicine’s institutional review board exempted this study. We followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional designs.

Measures

Latent class analysis models included theory-informed^28,29,30,31 and evidence-informed^32,33 indicators: circumstances, toxicology-based substance indicators, and suicide-method indicators (see eTable 1 for data dictionary and recoding and eMethods 2 for indicator selection rationale in Supplement 1).

Circumstances included mental health problems, depressed mood, past/current mental illness treatment, suicide-attempt history, recent crisis, financial problems, job-related problems, friend/family member suicide/death, physical health problems, interpersonal violence perpetration/victimhood, criminal/civil legal problems, intimate partner/other relationship problems, alcohol, and other substance-use problems. Responses were binary (0 = no/not available/unknown; 1 = yes). See eMethods 3 and 4 in Supplement 1 for a discussion of no/not available/unknown responses and other recoding.

Toxicology screening included amphetamines, alcohol, opiates, marijuana, and cocaine. We recoded responses into binary variables (0 = no/not available/unknown/missing; 1 = yes) after assessing missing/misclassification (eMethods 4 in Supplement 1). We calculated polysubstance use (0 = none present/unknown/missing, 1 = 1, 2 = 2, 3 = 3, 4 = ≥4 substances).

We grouped suicide methods into 4 categories: 0 indicated firearms; 1, hanging, suffocation, or strangulation; 2, poisoning (suicide by prescription/illicit drugs, alcohol, carbon monoxide, gas, rat poison, or insecticides); and 3, other methods (eg, fire, drowning, hypothermia) or unknown/missing.¹³

Primary outcomes were the absence of suicide intent disclosure and suicide notes. Suicide intent disclosure indicated disclosing suicidal thoughts or plans within the last month to family members, health care workers, friends, intimate partners, neighbors, others, or on social media. Suicide note presence means the decedent left a suicide note or other communication indicating intent but too late for prevention. Additionally, we evaluated the presence of 4 psychotropic drug classes (antidepressant, anticonvulsant, antipsychotic, and benzodiazepine) from toxicology reports. Because not every drug was tested, a decision made by medical/legal examiners, we calculated new variables accounting for whether the drug was tested and the results (response categories: tested–present, tested–not present, not tested, and missing or unknown). We assessed missing values (eTable 1 in Supplement 1).

We included suicide-associated covariates: age, sex, race and ethnicity, education, marital status, military/veteran status, and urban/rural residence.^5,6,34,35 Race and ethnicity descriptions included Hispanic, non-Hispanic American Indian or Alaska Native (hereafter American Indian or Alaska Native), non-Hispanic Asian or Other Pacific Islander (hereafter Asian or Other Pacific Islander), non-Hispanic Black (hereafter Black), non-Hispanic other/unspecified, non-Hispanic 2 or more races, non-Hispanic White (hereafter White), and unknown/missing.³⁶

Statistical Analysis

We conducted latent class analysis using poLCA³⁷ in R version 4.2.1 (R Foundation) to identify profiles. Latent class analysis is a finite mixture modeling approach to identify groups based on similar indicator responses.^23,38,39 We estimated models ranging from 2 to 10 classes, determining the optimal model by fit (eg, Bayesian information criterion, entropy) and interpretability.

Individuals were assigned to classes based on probability,⁴⁰ with class prevalence and membership differences analyzed using analysis of variance, Kruskal-Wallis, or χ² tests. Logistic regression was used to associate class membership with primary outcomes.

There were no missing values for class indicators and outcome variables after recoding. We addressed missing covariate data (0.0%-19.3%) using multiple imputation (mice version 3.14.0)⁴¹ and validated findings against nonimputed data for robustness.

Results

Among 306 800 suicide decedents, 239 627 (78.1%) were male, 67 108 were female (21.9%), 3918 (1.3%) were American Indian or Alaska Native, 7019 (2.3%) were Asian or Other Pacific Islander, 19 986 (6.5%) were Black, 19 354 (6.3%) were Hispanic, 251 861 (82.1%) were White, and 3428 (1.1%) reported more than 2 races (Table 1). There were 51 956 (16.9%) veterans. Regarding marital status, 109 151 (35.6%) were never married, and 99 532 (32.4%) were married or in a civil union or domestic partnership. Most completed at least high school (204 432 [66.6%]) and lived in metropolitan/urban areas (245 042 [79.9%]).

Table 1. Demographic Characteristics^a.

Characteristic	No. (%)
Characteristic	Overall (N = 306 800)	Class 1: mental health and substance problems (n = 41 527 [13.5%])	Class 2: mental health problems (n = 53 928 [17.6%])	Class 3: crisis, alcohol-related, and intimate partner problems (n = 55 367 [18.0%])	Class 4: physical health problems (n = 97 175 [31.7%])	Class 5: polysubstance problems (n = 58 803 [19.2%])
Age, y
≤10	161 (0.1)	10 (0.0)	28 (0.1)	26 (0.0)	93 (0.1)	6 (0.0)
11-15	4926 (1.6)	447 (1.1)	1144 (2.1)	781 (1.4)	2125 (2.2)	429 (0.7)
16-24	37 319 (12.2)	4144 (10.0)	6796 (12.6)	8328 (15.0)	11 345 (11.7)	6706 (11.4)
25-40	80 996 (26.4)	12 031 (29.0)	13 543 (25.1)	19 014 (34.3)	19 238 (19.8)	17 170 (29.2)
41-54	81 037 (26.4)	13 454 (32.4)	14 977 (27.8)	16 431 (29.7)	20 314 (20.9)	15 861 (27.0)
55-70	68 290 (22.3)	9526 (22.9)	12 535 (23.2)	9045 (16.3)	23 794 (24.5)	13 390 (22.8)
≥71	33 894 (11.0)	1915 (4.6)	4905 (9.1)	1740 (3.1)	20 100 (20.7)	5234 (8.9)
Unknown/missing	175 (0.1)	0	0	2 (0.0)	166 (0.2)	7 (0.0)
Sex
Male	239 627 (78.1)	23 962 (57.7)	40 703 (75.5)	49 019 (88.5)	82 191 (84.6)	43 752 (74.4)
Female	67 108 (21.9)	17 563 (42.3)	13 225 (24.5)	6347 (11.5)	14 923 (15.4)	15 050 (25.6)
Unknown/missing	65 (0.0)	2 (0.0)	0	1 (0.0)	61 (0.1)	1 (0.0)
Race and ethnicity
American Indian or Alaska Native^b	3918 (1.3)	415 (1.0)	389 (0.7)	996 (1.8)	1238 (1.3)	880 (1.5)
Asian or Other Pacific Islander^b	7019 (2.3)	686 (1.7)	1178 (2.2)	1187 (2.1)	2541 (2.6)	1427 (2.4)
Black or African American^b	19 986 (6.5)	1682 (4.1)	2764 (5.1)	4422 (8.0)	7014 (7.2)	4104 (7.0)
Hispanic	19 354 (6.3)	2220 (5.3)	2545 (4.7)	4584 (8.3)	5253 (5.4)	4752 (8.1)
Other/unspecified^b	912 (0.3)	68 (0.2)	79 (0.1)	99 (0.2)	491 (0.5)	175 (0.3)
≥2 Races^b	3428 (1.1)	432 (1.0)	625 (1.2)	826 (1.5)	923 (0.9)	622 (1.1)
White^b	251 861 (82.1)	36 010 (86.7)	46 333 (85.9)	43 225 (78.1)	79 485 (81.8)	46 808 (79.6)
Unknown/missing	322 (0.1)	14 (0.0)	15 (0.0)	28 (0.1)	230 (0.2)	35 (0.1)
Military affiliation
Yes	51 956 (16.9)	4301 (10.4)	9113 (16.9)	8563 (15.5)	22 220 (22.9)	7759 (13.2)
No	238 413 (77.7)	35 298 (85.0)	42 310 (78.5)	44 324 (80.1)	68 568 (70.6)	47 913 (81.5)
Unknown/missing	16 431 (5.4)	1928 (4.6)	2505 (4.6)	2480 (4.5)	6387 (6.6)	3131 (5.3)
Marital status
Married, civil union, domestic partnership	99 532 (32.4)	12 152 (29.3)	19416 (36.0)	19 144 (34.6)	32 609 (33.6)	16 211 (27.6)
Divorced	65 066 (21.2)	10 273 (24.7)	10 571 (19.6)	11 040 (19.9)	19 832 (20.4)	13 350 (22.7)
Married, civil union, domestic partnership but separated	7531 (2.5)	1419 (3.4)	1536 (2.8)	2383 (4.3)	982 (1.0)	1211 (2.1)
Never married	109 151 (35.6)	15 034 (36.2)	18 916 (35.1)	20 006 (36.1)	32 248 (33.2)	22 947 (39.0)
Single, not otherwise specified	3968 (1.3)	437 (1.1)	668 (1.2)	940 (1.7)	1075 (1.1)	848 (1.4)
Widowed	17 869 (5.8)	1864 (4.5)	2479 (4.6)	1447 (2.6)	8633 (8.9)	3446 (5.9)
Unknown/missing	3683 (1.2)	348 (0.8)	342 (0.6)	407 (0.7)	1796 (1.8)	790 (1.3)
Education level
High school or above	204 432 (66.6)	29 853 (71.9)	34 165 (63.4)	34 300 (62.0)	62 898 (64.7)	43 216 (73.5)
Below high school	43 044 (14.0)	4486 (10.8)	6363 (11.8)	7774 (14.0)	15 690 (16.1)	8732 (14.8)
Unknown/missing	59 324 (19.3)	7188 (17.3)	13 401 (24.8)	13 293 (24.0)	18 587 (19.1)	6855 (11.7)
Residence^c
Nonmetropolitan	60 143 (19.6)	4905 (13.7)	9511 (17.6)	11382 (20.6)	23 966 (24.7)	9600 (16.3)
Metropolitan	245 042 (79.9)	30 755 (86.0)	44 253 (82.1)	43 792 (79.1)	72 421 (74.5)	48 855 (83.1)
Unknown/missing	1615 (0.5)	94 (0.3)	164 (0.3)	193 (0.3)	788 (0.8)	348 (0.6)

Open in a new tab

^{^a}

For all groups, P < .001 and adjusted P < .001. Test statistic and P value for Kruskal-Wallis tests or χ² tests were used to compare all suicide deaths associated with latent classes of profiles (class 1 vs class 2 vs class 3 vs class 4 vs class 5). Bonferroni-adjusted P < .00625 (0.05/8 tests). Unknown/missing values were excluded when conducting the Kruskal-Wallis test or Pearson χ² test.

^{^b}

Non-Hispanic ethnicity.

^{^c}

A metropolitan area was defined as scoring 3 or less on the Rural-Urban Continuum Code and a nonmetropolitan area, 4 or more.³⁵

Latent Classes of Suicide Decedents

A 5-class model (Figure 1) provided the best fit (eTable 2 and eFigure 1 in Supplement 1) and interpretability (eTable 3 and eFigures 2-10 in Supplement 1). Class 1 (mental health and substance problems, n = 41 527; 13.5%) had high percentages of mental health problems (40 979 [98.7%]), depressed mood (16 981 [40.9%]), past (41 409 [99.7%]) and current (35 217 [84.8%]) mental health treatment, and prior suicide attempts (16 811 [40.5%]). Almost everyone (41 031 [98.8%]) in class 1 had illegal substances detected on toxicology, with 45.8% (n = 19 029) of these decedents exhibiting 4 or more substances. Class 1 had the highest suicide proportion by poisoning (20 829 [50.2%]).

Class 2 (mental health problems, 53 928 [17.6%]) had comparably high known prior mental health problems (53 305 [98.8%]) but fewer substance problems than class 1 (80.5% [43 390] of class 2 had no substances at death, and <1% [427] had ≥2 substances). Class 2 had the highest proportion of suicides by hanging (18 887 [35.0%]). Class 3 (crisis, alcohol-related, and intimate-partner problems, 55 367 [18.0%]) manifested the most recent crisis, alcohol-related, and/or intimate partner problems.

Class 4 (physical health problems, 97 175 [31.7%]) was the largest class of suicide decedents and was characterized by the most physical health problems (23 647 [24.3%]) and firearm suicide deaths (61 674 [63.5%]). Class 4 had the fewest known mental health problems (13 033 [13.4%]), depressed mood (20 710 [21.3%]), positive amphetamine test results (44 [0.05%]), positive alcohol test results (2755 [2.8%]), positive opiate test results (717 [0.7%]), positive marijuana test results (229 [0.2%]), positive cocaine test results (67 [0.1%]), and prior suicide attempts (5561 [5.7%]), and less than 1% had known history of (668 [0.7%]) or current (10 [0.01%]) mental illness treatment.

Class 5 (polysubstance problems, n = 58 803, 19.2%) reported the most substance misuse (43 578 [74.1%] with ≥2 substances at death) and alcohol misuse (23 298 [39.6%]). Classes 1 and 5 differed: class 1 had a higher frequency of known mental health issues and treatment (40 979 [98.7%] and 41 409 [99.7%], respectively) compared with lower rates in class 5 (14 421 [24.5%] and 1697 [2.9%], respectively). Class 1 primarily attempted suicide by poisoning (20 829 [50.2%]), while class 5 mostly used firearms (20 220 [34.4%]).

Shifts in class proportions were observed (Figure 2). Class 5 saw the largest increase, rising by 194% from 2010 to 2015 ([8.4%] to 5075 [24.7%]). Class 4 increased by 31.2% in 2016 to 2020 (7223 [27.9%] to 14 198 [36.6%]). In contrast, class 2 and class 3 decreased 49.8% (888 [24.5%] to 4769 [12.3%], 2003-2020) and 36.7% (800 [22.1%] to 5436 [14.0%], 2003-2020), respectively. Between 2011 and 2017, class 1 increased by 67.0%, from 1206 (9.7%) to 4821 (16.2%).

Associations With Signs of Suicide Intent

Known suicide intent disclosure and not leaving a suicide note varied across classes (Table 2). Class 1 (12 217 [29.4%]) and class 2 (15 708 [29.1%]) showed the highest rates of suicide intent disclosure, whereas class 4 (13 952 [14.4%]) had the lowest. Suicide notes ranged from the highest in class 1 (14 706 [35.4%]) and class 5 (18 445 [31.4%]) to the lowest in class 4 (24 351 [25.1%]). Psychotropic drugs detected via toxicology also varied, with class 1 showing the highest rates for antidepressants (17 551 [50.5%]), anticonvulsants (4366 [19.1%]), antipsychotics (3693 [16.2%]), and benzodiazepines (10 930 [41.5%]). In contrast, class 4 had the lowest rates of psychotropics detected: antidepressants (824 [2.3%]), antipsychotics (39 [0.3%]), and benzodiazepines (534 [3.5%]). Adjusting for covariates (Table 3), compared with class 1, the odds of not disclosing suicide intent were highest in class 4 (OR, 2.58; 95% CI, 2.51-2.66), and suicide note absence was more common in class 4 (OR, 1.45; 95% CI, 1.41-1.49).

Table 2. Distribution of Signs of Suicide Intent by 5 Suicide Decedent Profiles.

Characteristic	No. (%)
Characteristic	Class 1: mental health and substance problems (n = 41 527 [13.5%])	Class 2: mental health problems (n = 53 928 [17.6%])	Class 3: crisis, alcohol-related, and intimate partner problems (n = 55 367 [18.0%])	Class 4: physical health problems (n = 97 175 [31.7%])	Class 5: polysubstance problems (n = 58 803 [19.2%])
Recent suicide disclosure^a
No	29 310 (70.6)	38 220 (70.9)	40 138 (72.5)	83 223 (85.6)	46 626 (79.3)
Yes	12 217 (29.4)	15 708 (29.1)	15 229 (27.5)	13 952 (14.4)	12 177 (20.7)
Suicide note^a
No	26 821 (64.6)	37 329 (69.2)	39 228 (70.9)	72 824 (74.9)	40 358 (68.6)
Yes	14 706 (35.4)	16 599 (30.8)	16 139 (29.1)	24 351 (25.1)	18 445 (31.4)
Psychiatric medication in toxicology testing^b
Antidepressants
Present	17 551 (50.5)	3288 (10.3)	819 (2.4)	824 (2.3)	8642 (21.1)
Not present	9939 (28.6)	9821 (30.7)	14 123 (41.3)	12 526 (35.3)	18 022 (44)
Not tested	7259 (20.9)	18 885 (59.0)	19 235 (56.3)	22 181 (62.4)	14 327 (35)
Anticonvulsants
Present	4366 (19.1)	129 (1.0)	120 (0.8)	132 (0.9)	3000 (9.8)
Not present	10 626 (46.4)	4254 (33.1)	5825 (38.2)	5190 (36.7)	13 802 (45)
Not tested	7893 (34.5)	8477 (65.9)	9303 (61.0)	8802 (62.3)	13 839 (45.2)
Antipsychotics
Present	3693 (16.2)	114 (0.9)	38 (0.2)	39 (0.3)	1406 (4.7)
Not present	11 367 (49.7)	4480 (34.6)	6101 (39.9)	5243 (37)	14 787 (49.1)
Not tested	7801 (34.1)	8347 (64.5)	9142 (59.8)	8894 (62.7)	13 937 (46.3)
Benzodiazepines
Present	10 930 (41.5)	729 (5.3)	769 (4.7)	534 (3.5)	10 134 (27.9)
Not present	13 535 (51.4)	7177 (52.4)	10 436 (63.4)	9334 (61.5)	22 143 (60.9)
Not tested	1889 (7.2)	5778 (42.2)	5245 (31.9)	5300 (34.9)	4066 (11.2)

Open in a new tab

^{^a}

No indicates no/not available/unknown as originally provided by the National Violent Death Record System.

^{^b}

Data were combined for whether the specific substance was tested (tested, not tested, unknown) and the results of the toxicology tests. Percentages were calculated with all suicide decedents. Column totals do not sum to 100% owing to missing data (coded as unknown for whether the substance was tested).

Table 3. Multivariable Logistic Regression Models With Suicide Intent Disclosure and Suicide Note Leaving as Outcomes.

	Odds ratio (95% CI)
	No recent suicide intent disclosure	No suicide note
Class
1: Mental health and substance problems	1 [Reference]	1 [Reference]
2: Mental health problems	1.04 (1.01-1.07)	1.15 (1.12-1.18)
3: Crisis, alcohol-related, and intimate partner problems	1.16 (1.12-1.19)	1.10 (1.07-1.13)
4: Physical health problems	2.58 (2.51-2.66)	1.45 (1.41-1.49)
5: Polysubstance problems	1.62 (1.57-1.66)	1.09 (1.06-1.12)
Age, y
≤10	1 [Reference]	1 [Reference]
11-15	1.22 (0.82-1.80)	0.32 (0.19-0.54)
16-24	0.98 (0.66-1.44)	0.48 (0.29-0.80)
25-40	1.03 (0.70-1.52)	0.61 (0.37-1.01)
41-54	1.22 (0.83-1.79)	0.50 (0.30-0.83)
55-70	1.39 (0.94-2.05)	0.45 (0.27-0.75)
≥71	1.04 (0.71-1.53)	0.46 (0.28-0.76)
Sex
Male	1 [Reference]	1 [Reference]
Female	1.08 (1.06-1.10)	0.72 (0.71-0.74)
Education level
High school or above	1 [Reference]	1 [Reference]
Below high school	0.87 (0.84-0.89)	1.50 (1.45-1.55)
Race and ethnicity
American Indian or Alaska Native^a	0.86 (0.80-0.93)	1.70 (1.56-1.85)
Asian or Other Pacific Islander^a	1.31 (1.23-1.40)	1.05 (1.0003-1.11)
Black or African American^a	1.27 (1.22-1.32)	1.88 (1.81-1.95)
Hispanic	1.02 (0.99-1.06)	1.39 (1.34-1.44)
Other/unspecified^a	1.88 (1.53-2.30)	1.70 (1.44-2.01)
≥2 Races^a	0.83 (0.77-0.90)	0.99 (0.92-1.06)
White^a	1 [Reference]	1 [Reference]
Marital status
Married, civil union, domestic partnership	1 [Reference]	1 [Reference]
Divorced	0.98 (0.96-1.003)	0.81 (0.79-0.83)
Married, civil union, domestic partnership but separated	0.80 (0.76-0.84)	0.63 (0.60-0.66)
Never married	1.22 (1.19-1.25)	0.85 (0.83-0.87)
Single, not otherwise specified	1.17 (1.09-1.27)	0.90 (0.84-0.97)
Widowed	1.03 (0.99-1.08)	0.83 (0.80-0.86)
Military affiliation
Yes	0.96 (0.93-0.98)	0.93 (0.91-0.95)
No	1 [Reference]	1 [Reference]
Residence
Nonmetropolitan	1.01 (0.99-1.03)	1.28 (1.25-1.30)
Metropolitan	1 [Reference]	1 [Reference]

Open in a new tab

^{^a}

Non-Hispanic ethnicity.

Demographic Disparities

Compared with White individuals, Black individuals had higher odds of nondisclosure of intent (OR, 1.27; 95% CI, 1.22-1.32) and no note (OR, 1.88; 95% CI, 1.81-1.95). Females were less likely to disclose intent (OR, 1.08; 95% CI, 1.06-1.10) and more likely to leave a note (OR, 0.72; 95% CI, 0.71-0.74).

Discussion

We identified 5 suicide profiles, each requiring tailored prevention strategies. Class 1 had comorbid mental and substance use disorders. Class 2 was dominated by mental disorders. Class 3 experienced crisis, alcohol-related, and intimate partner problems; class 4 displayed physical health problems; and class 5 displayed substance-use problems. We found class 4 was least likely to disclose suicide intent, leave suicide notes as intent indicators, and have known mental illness or psychotropics detected by toxicology.

Class 4 emerged as the most concerning group: the largest (31.7%) and increasing the most (31.2% increase since 2016). With the strikingly lowest rate of diagnosed mental disorders and toxicology-detected psychotropic medications, class 4 comes the closest to having an “invisible suicide decedent risk profile.” Several factors may explain this. Men’s suicide rate is 4 times higher than women’s, and men are poorer at help-seeking. Physical illnesses can cause physical pain and psychological distress, which can lead to feelings of helplessness and heighten suicide risk that may be overlooked in medical settings.⁴² Suicidal patients may prioritize physical over mental health care because they consulted non–mental health care professionals twice as often as mental health professionals in the 30 days before death by suicide. Therefore, nonpsychiatric health care settings, especially primary care and emergency departments, are crucial for suicide risk screening.⁴³

The composition of class 4 (predominantly older adults, non-Hispanic Asian American individuals, and veterans) reflects broader suicide trends in these populations.⁴⁴ Systemic barriers such as stigma and social isolation may impede the recognition and treatment of mental health issues.¹⁰ Suicide decedents in class 4 were more likely to die on the first attempt because they mostly used firearms.⁵ The contrast in psychotropic medication usage between class 4 and class 1 is noteworthy. In class 4, only 2.3% of individuals were using antidepressants, compared with 50.5% in class 1. This discrepancy raises concerns about the potential undertreatment of major depression, given its link to nearly half of US suicide deaths.⁵ This highlights the urgent need to enhance the diagnosis and treatment of major depression in health care settings.⁵ Future research is needed to understand why class 4 has grown as a proportion of suicide decedents over time because that may identify other suicide prevention targets.

In contrast, while both classes 1 and 2 experienced high rates of mental health problems, they exhibited distinct challenges. Class 1 faced the additional challenge of comorbid substance use. A significant majority of class 1 members (84.8%-99.7%) had received or were receiving mental illness treatment.⁴⁵ Notably, treatment rates escalated between 2011 and 2015, paralleling the increase in opioid-involved overdose deaths, which surged from 22 784 in 2011 to 68 630 in 2020.⁴⁶ The increase in prescription and illegal opioids, particularly fentanyl, and its association with poisoning-related suicides in class 1 necessitates further investigation. Toxicology findings indicate that nearly half of the suicide decedents in class 1 were not taking psychotropic medications at the time of death, indicating possible nonadherence or cessation of treatment before suicide.⁴⁷ In contrast, class 2, with minimal substance use disorders and less mental health treatment than class 1, experienced a significant decrease in suicide rates, suggesting improvements in mental health care, but limited resources for managing substance use disorders were underlying the trends in these 2 classes of suicide.

Class 5, marked by widespread polysubstance problems, grew the fastest. Toxicology reports revealed one-fifth of class 5 decedents had no current psychotropic medication at the time of death, and only 3% had a treatment history. This class also recorded the highest rates of positive test results for substances such as amphetamine (12.5%), alcohol (39.6%), marijuana (19.5%), and cocaine (7.9%). Challenges like stigma and health complications from multiple substance dependencies likely hinder recovery efforts, leading to lower treatment engagement and adherence and a hesitancy to reveal suicidal tendencies.⁴⁸ As indicated by the trends in classes 1 and 2, systemic barriers⁴⁹ to accessing substance abuse treatment, and potentially the absence of culturally tailored and gender-specific programs, may have contributed to the suicide risk in classes 1 and 5.

Following prior studies, we found class 3 represented younger males struggling with alcohol misuse amid economic challenges⁵⁰ and separated or single males facing intimate partner relationship difficulties.⁵¹ Prevention should extend beyond immediate, visible aspects of these crises to comprehensively address less apparent but underlying causes often overlooked in research and intervention strategies.

Demographic Differences

The demographic characteristics of class 4, comprising an aging, rural, racial and ethnic minority, widowed, veteran, and less educated population, suggest that shared factors (eg, limited access to education/health care, social isolation, financial difficulties, trauma, low health literacy) may have contributed to neglected suicide risks and increase physical-illness–related suicide rates observed from 2016 to 2020.^6,44 The post-2011 increase of class 5 may be partially explained by societal changes affecting this high school–educated, single/never-married, 25- to 54-year-old male demographic segment, including economic stress from uneven recovery after the Great Recession, increased substance availability,⁵² and absence of protective marriage.⁵³ A higher proportion of divorced individuals (24.7%) and females (42.3%) in class 1 suggests that relationship breakdowns are significant stressors, as divorce is often associated with psychological and substance use problems.⁵⁴ The observed greater reluctance among females to disclose suicidal intent was unexpected, presenting an area for further investigation. Class 3, with a notable majority of males (88.5%), aligns with previous findings linking males to heightened crisis, alcohol-related issues, and intimate partner problems.^3,5,17

Research and Policy Implications to Improve Suicide Prevention

Our findings reinforce the need for tailored approaches that recognize the diverse needs of the suicide decedent profile. The 5 suicide profiles based on postmortem data can also inform clinicians to screen for each suicide risk subtype for personalized suicide prevention strategies. For class 4, integrated care coordinating physical and mental health services is crucial, especially in primary care/internal medicine settings,⁵⁵ where a significant proportion of patients experience at least 1 (60%) or 2 (40%) chronic diseases,⁵⁶ including diabetes (37.3 million people), Alzheimer disease (5.8 million people), and cancer (1.6 million people).

Class 1 is characterized by significant comorbid mental health and substance use problems, calling for a holistic approach that integrates psychiatric and substance abuse treatment,⁵⁷ medication management,⁵⁸ and community-based rehabilitation to foster recovery and social reintegration.⁵⁹ Class 2 might benefit from comprehensive mental health services,⁶⁰ including psychoeducation tailored to individuals and families.⁶¹ Improved medication adherence is potentially helpful but not the only option. Beyond medication, therapies such as cognitive-behavioral therapy⁶² and dialectical behavior therapy⁶³ can be effective for suicidal behaviors, supporting risk reduction and enhanced coping mechanisms, particularly for those with depression and borderline personality disorder.⁶²

For class 3, interventions like safety planning, crisis intervention,⁶⁴ alcohol treatment programs,⁶⁵ and family/relationship counseling are imperative.⁶⁶ Class 5, characterized by polysubstance issues, could benefit from mental health screenings, substance use treatment,⁵² harm reduction,⁶⁷ and peer support.⁶⁸ Both classes 3 and 5 may also gain from treatments addressing the overlap of suicide risks and substance use, as well as firearm control measures.

Given the increasing diversity in the US, clinical interventions should be culturally relevant and sensitive, accommodating different racial and ethnic, sex, age, and educational backgrounds. Such inclusivity ensures that interventions are effective and resonate with diverse groups.^{6,22,26,69,70}

Strengths and Limitations

This study harnessed comprehensive suicide data to identify 5 suicide profiles, each informing unique prevention strategies. This study has limitations. First, the NVDRS data revealed geographic variations due to decentralized coroner systems and diverse law enforcement practices. Combined with variable experience levels among data abstractors and reliance on second-hand information, these variations may introduce data incompleteness. To mitigate this, we applied fixed effects and county-year interactions to control for location and time-specific variations. Our conclusions remained robust.

Second, data coding, particularly in ambiguous categories like no/not available/unknown, may affect the accuracy of our findings. This is particularly true for toxicology data, where underreporting is common. We tackled this by examining drug testing details and cross-validating our findings through multiple imputations and direct data comparison.

Third, the potential for misclassifying suicide deaths and underreporting previous suicide attempts poses a risk of overestimating the incidence of first-attempt suicides. This challenge is amplified by reliance on medical examiner and law enforcement reports, which may not always capture the full history of attempts.¹ Our high rate of first-attempt decedents (60%-94%) aligns with other NVDRS studies (79%)³³ but contrasts with lower rates reported in some psychological autopsy studies (10%-40%).^8,10,71,72

Fourth, the high frequency of firearm-related deaths in the US limits the generalizability of our findings to countries with lower firearm availability.^5,6,73 With the highest per capita gun ownership (45%) in the developed world,⁷⁴ the US presents a unique context for firearm suicide by making firearms accessible to many at-risk individuals.

Finally, there is a risk of biased screening related to demographics.⁷⁵ For example, Black and Hispanic individuals may receive less comprehensive investigations than White individuals, affecting data reliability.⁷⁵ Given possible unreliable data, we suggest interpreting our findings as part of a broader approach to guide suicide prevention.

Despite these limitations, the NVDRS remains the only national US dataset that comprehensively captures circumstances, toxicology, and suicide methods.¹³ We found internally consistent results, despite dataset weaknesses, offering a nuanced understanding of suicide.

Conclusions

This study identified 5 distinct suicide profiles, underscoring the complexity of suicide and illustrating its multifaceted nature. A notable finding is the rise of an invisible risk profile characterized by physical health problems, highlighting psychiatric underdiagnosis and subsequent undertreatment. These findings highlight the need for tailored suicide prevention strategies that address the specific needs of each profile to the extent they are identifiable, or prevent impulsive thoughts from being acted on, moving away from a 1-size-fits-all approach. Improving the detection and treatment of coexisting mental health conditions, substance and alcohol use disorders, and physical illnesses is paramount. Moreover, the implementation of means restriction strategies plays a vital role in reducing suicide risks across most of the profiles, reinforcing the need for a multifaceted approach to suicide prevention.

Supplement 1.

eMethods 1. National Violent Death Reporting System (NVDRS) Data: Unique Value and Fit for the Current Study

eMethods 2. Rationale Behind the Selection of Indicators in the National Violent Death Reporting System (NVDRS) to Identify Suicide Death Profiles using Latent Class Analysis

eMethods 3. Reliability and Possible Misclassification Scenarios in the National Violent Death Reporting System (NVDRS) Data

eMethods 4. Strategies to Deal with Missing Data and Misclassification Error

eMethods 5. Addressing the Implications of Unbalanced Panel Data on Study Outcomes in the National Violent Death Reporting System (NVDRS) Data

eFigure 1. Scree Plot of Bayesian Information Criterion (BIC) Values Across Ten Latent Class Analysis Solutions

eFigure 2. Two-Class Latent Class Solution

eFigure 3. Three-Class Latent Class Solution

eFigure 4. Four-Class Latent Class Solution

eFigure 5. Five-Class Latent Class Solution

eFigure 6. Six-Class Latent Class Solution

eFigure 7. Seven-Class Latent Class Solution

eFigure 8. Eight-Class Latent Class Solution

eFigure 9. Nine-Class Latent Class Solution

eFigure 10. Ten-Class Latent Class Solution

eTable 1. Indicator Information including Definition, Original Response Categories, and Recoding Rules

eTable 2. Model Fit Statistics of Latent Class Analysis Models

eTable 3. Distribution of Precipitating Circumstances by Five Suicide Decedent Profiles in the U.S., 2003-2020 (N=306 800)

eTable 4. Results of Analysis without Missing Data Imputation

eTable 5. Class Prevalence by Year

eTable 6. Results of Analysis without Missing Data Imputation Additionally controlling for Fixed Effects of Year and State

eReferences

jamapsychiatry-e240171-s001.pdf^{(3.2MB, pdf)}

Supplement 2.

Data sharing statement

jamapsychiatry-e240171-s002.pdf^{(14.8KB, pdf)}

References

1.Centers for Disease Control and Prevention . Suicide increases in 2021 after two years of decline. Published September 30, 2022. Accessed June 4, 2023. https://www.cdc.gov/nchs/pressroom/nchs_press_releases/2022/20220930.htm
2.Mann JJ, Apter A, Bertolote J, et al. Suicide prevention strategies: a systematic review. JAMA. 2005;294(16):2064-2074. doi: 10.1001/jama.294.16.2064 [DOI] [PubMed] [Google Scholar]
3.Zalsman G, Hawton K, Wasserman D, et al. Suicide prevention strategies revisited: 10-year systematic review. Lancet Psychiatry. 2016;3(7):646-659. doi: 10.1016/S2215-0366(16)30030-X [DOI] [PubMed] [Google Scholar]
4.Jordan JT, McNiel DE. Characteristics of persons who die on their first suicide attempt: results from the National Violent Death Reporting System. Psychol Med. 2020;50(8):1390-1397. doi: 10.1017/S0033291719001375 [DOI] [PubMed] [Google Scholar]
5.Mann JJ, Michel CA, Auerbach RP. Improving suicide prevention through evidence-based strategies: a systematic review. Am J Psychiatry. 2021;178(7):611-624. doi: 10.1176/appi.ajp.2020.20060864 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Xiao Y, Cerel J, Mann JJ. Temporal trends in suicidal ideation and attempts among US adolescents by sex and race/ethnicity, 1991-2019. JAMA Netw Open. 2021;4(6):e2113513-e2113513. doi: 10.1001/jamanetworkopen.2021.13513 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.García de la Garza Á, Blanco C, Olfson M, Wall MM. Identification of suicide attempt risk factors in a national US survey using machine learning. JAMA Psychiatry. 2021;78(4):398-406. doi: 10.1001/jamapsychiatry.2020.4165 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Favril L, Yu R, Uyar A, Sharpe M, Fazel S. Risk factors for suicide in adults: systematic review and meta-analysis of psychological autopsy studies. Evid Based Ment Health. 2022;25(4):148-155. doi: 10.1136/ebmental-2022-300549 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Cavanagh JTO, Carson AJ, Sharpe M, Lawrie SM. Psychological autopsy studies of suicide: a systematic review. Psychol Med. 2003;33(3):395-405. doi: 10.1017/S0033291702006943 [DOI] [PubMed] [Google Scholar]
10.Bostwick JM, Pabbati C, Geske JR, McKean AJ. Suicide attempt as a risk factor for completed suicide: even more lethal than we knew. Am J Psychiatry. 2016;173(11):1094-1100. doi: 10.1176/appi.ajp.2016.15070854 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.McMahon EM, Greiner BA, Corcoran P, et al. Psychosocial and psychiatric factors preceding death by suicide: a case-control psychological autopsy study involving multiple data sources. Suicide Life Threat Behav. 2022;52(5):1037-1047. doi: 10.1111/sltb.12900 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Ross V, Kõlves K, De Leo D. Beyond psychopathology: a case-control psychological autopsy study of young adult males. Int J Soc Psychiatry. 2017;63(2):151-160. doi: 10.1177/0020764016688041 [DOI] [PubMed] [Google Scholar]
13.Liu GS, Nguyen BL, Lyons BH, et al. Surveillance for violent deaths: National Violent Death Reporting System, 48 states, the District of Columbia, and Puerto Rico, 2020. MMWR Surveill Summ. 2023;72(5):1-38. doi: 10.15585/mmwr.ss7205a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Xiao Y, Zhang Z, Yip P, Cerel J, Mann JJ. Trends in suicide deaths with and without warning from 2003-2019: identifying the missing link for effective suicide prevention. 2022 Annual meeting of American Public Health Association APHA. Accessed June 22, 2023. https://apha.confex.com/apha/2022/meetingapp.cgi/Paper/518592
15.Cai Z, Junus A, Chang Q, Yip PSF. The lethality of suicide methods: a systematic review and meta-analysis. J Affect Disord. 2022;300:121-129. doi: 10.1016/j.jad.2021.12.054 [DOI] [PubMed] [Google Scholar]
16.Bond AE, Bandel SL, Rodriguez TR, Anestis JC, Anestis MD. Mental health treatment seeking and history of suicidal thoughts among suicide decedents by mechanism, 2003-2018. JAMA Netw Open. 2022;5(3):e222101. doi: 10.1001/jamanetworkopen.2022.2101 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Wang S, Dang Y, Sun Z, et al. An NLP approach to identify SDoH-related circumstance and suicide crisis from death investigation narratives. J Am Med Inform Assoc. 2023;30(8):1408-1417. doi: 10.1093/jamia/ocad068 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Rodway C, Tham SG, Turnbull P, Kapur N, Appleby L. Suicide in children and young people: can it happen without warning? J Affect Disord. 2020;275:307-310. doi: 10.1016/j.jad.2020.06.069 [DOI] [PubMed] [Google Scholar]
19.Yip PSF, Caine E, Yousuf S, Chang SS, Wu KCC, Chen YY. Means restriction for suicide prevention. Lancet. 2012;379(9834):2393-2399. doi: 10.1016/S0140-6736(12)60521-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Hochman E, Shelef L, Mann JJ, et al. Primary health care utilization prior to suicide: a retrospective case-control study among active-duty military personnel. J Clin Psychiatry. 2014;75(8):e817-e823. doi: 10.4088/JCP.13m08823 [DOI] [PubMed] [Google Scholar]
21.Mann JJ. Can knowledge of genetic and environmental causal factors of fatal and nonfatal suicidal behavior be translated into better prevention? Am J Psychiatry. 2021;178(11):994-997. doi: 10.1176/appi.ajp.2021.21090913 [DOI] [PubMed] [Google Scholar]
22.Xiao Y, Romanelli M, Lindsey MA. A latent class analysis of health lifestyles and suicidal behaviors among US adolescents. J Affect Disord. 2019;255:116-126. doi: 10.1016/j.jad.2019.05.031 [DOI] [PubMed] [Google Scholar]
23.Grant RW, McCloskey J, Hatfield M, et al. Use of latent class analysis and k-means clustering to identify complex patient profiles. JAMA Netw Open. 2020;3(12):e2029068-e2029068. doi: 10.1001/jamanetworkopen.2020.29068 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Logan J, Hall J, Karch D. Suicide categories by patterns of known risk factors: a latent class analysis. Arch Gen Psychiatry. 2011;68(9):935-941. doi: 10.1001/archgenpsychiatry.2011.85 [DOI] [PubMed] [Google Scholar]
25.Paulozzi LJ, Mercy J, Frazier L Jr, Annest JL; Centers for Disease Control and Prevention . CDC’s National Violent Death Reporting System: background and methodology. Inj Prev. 2004;10(1):47-52. doi: 10.1136/ip.2003.003434 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Wilson RF, Liu G, Lyons BH, et al. Surveillance for Violent Deaths: National Violent Death Reporting System, 42 states, the District of Columbia, and Puerto Rico, 2019. MMWR Surveill Summ. 2022;71(6):1-40. doi: 10.15585/mmwr.ss7106a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Centers for Disease Control and Prevention, National Center for Injury Prevention and Control . National Violent Death Reporting System web coding manual version 6. Accessed February 13, 2024. https://www.cdc.gov/violenceprevention/datasources/nvdrs/resources.html
28.Lindsey MA, Xiao Y. Depression, trauma, and suicide among adolescent and young adult males. Men’s Health Equity. Accessed February 13, 2024. https://www.taylorfrancis.com/chapters/edit/10.4324/9781315167428-18/depression-trauma-suicide-among-adolescent-young-adult-males-michael-lindsey-yunyu-xiao
29.Mann JJ. Neurobiology of suicidal behaviour. Nat Rev Neurosci. 2003;4(10):819-828. doi: 10.1038/nrn1220 [DOI] [PubMed] [Google Scholar]
30.Joiner T. Why People Die by Suicide. Harvard University Press; 2009. [Google Scholar]
31.Klonsky ED, May AM. The three-step theory (3ST): a new theory of suicide rooted in the “ideation-to-action” framework. Int J Cogn Ther. 2015;8(2):114-129. doi: 10.1521/ijct.2015.8.2.114 [DOI] [Google Scholar]
32.Xiao Y, Hinrichs R, Johnson N, et al. Suicide prevention among college students before and during the COVID-19 pandemic: protocol for a systematic review and meta-analysis. JMIR Res Protoc. 2021;10(5):e26948. doi: 10.2196/26948 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.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]
34.Kposowa AJ. Marital status and suicide in the National Longitudinal Mortality Study. J Epidemiol Community Health. 2000;54(4):254-261. doi: 10.1136/jech.54.4.254 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Hawley JE, Sun T, Chism DD, et al. Assessment of regional variability in COVID-19 outcomes among patients with cancer in the United States. JAMA Netw Open. 2022;5(1):e2142046. doi: 10.1001/jamanetworkopen.2021.42046 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Management and Budget Office, Federal Register . Revisions to the standards for the classification of federal data on race and ethnicity. Published October 30, 1997. Accessed February 13, 2024. https://www.federalregister.gov/d/97-28653
37.Linzer DA, Lewis JB. poLCA: an R package for polytomous variable latent class analysis. J Stat Softw. 2011;42:1-29. doi: 10.18637/jss.v042.i10 [DOI] [Google Scholar]
38.Laska MN, Pasch KE, Lust K, Story M, Ehlinger E. Latent class analysis of lifestyle characteristics and health risk behaviors among college youth. Prev Sci. 2009;10(4):376-386. doi: 10.1007/s11121-009-0140-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Mori M, Krumholz HM, Allore HG. Using latent class analysis to identify hidden clinical phenotypes. JAMA. 2020;324(7):700-701. doi: 10.1001/jama.2020.2278 [DOI] [PubMed] [Google Scholar]
40.Collins LM, Lanza ST. Latent Class and Latent Transition Analysis: With Applications in the Social, Behavioral, and Health Sciences. John Wiley & Sons; 2009. doi: 10.1002/9780470567333 [DOI] [Google Scholar]
41.van Buuren S, Groothuis-Oudshoorn K. mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(3). doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]
42.Xiao Y, Brown TT. Moving suicide prevention upstream by understanding the effect of flourishing on suicidal ideation in midlife: an instrumental variable approach. Sci Rep. 2023;13(1):1320. doi: 10.1038/s41598-023-28568-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Geale K, Henriksson M, Jokinen J, Schmitt-Egenolf M. Association of skin psoriasis and somatic comorbidity with the development of psychiatric illness in a nationwide swedish study. JAMA Dermatol. 2020;156(7):795-804. doi: 10.1001/jamadermatol.2020.1398 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Centers for Disease Control and Prevention . Disparities in suicide. Published May 9, 2023. Accessed November 6, 2023. https://www.cdc.gov/suicide/facts/disparities-in-suicide.html
45.Gibbons RD, Brown CH, Hur K, Davis J, Mann JJ. Suicidal thoughts and behavior with antidepressant treatment: reanalysis of the randomized placebo-controlled studies of fluoxetine and venlafaxine. Arch Gen Psychiatry. 2012;69(6):580-587. doi: 10.1001/archgenpsychiatry.2011.2048 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.National Institute on Drug Abuse . Drug overdose death rates. Published June 30, 2023. Accessed November 6, 2023. https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates
47.Miller M, Swanson SA, Azrael D, Pate V, Stürmer T. Antidepressant dose, age, and the risk of deliberate self-harm. JAMA Intern Med. 2014;174(6):899-909. doi: 10.1001/jamainternmed.2014.1053 [DOI] [PubMed] [Google Scholar]
48.Clement S, Schauman O, Graham T, et al. What is the impact of mental health-related stigma on help-seeking? a systematic review of quantitative and qualitative studies. Psychol Med. 2015;45(1):11-27. doi: 10.1017/S0033291714000129 [DOI] [PubMed] [Google Scholar]
49.Priester MA, Browne T, Iachini A, Clone S, DeHart D, Seay KD. Treatment access barriers and disparities among individuals with co-occurring mental health and substance use disorders: an integrative literature review. J Subst Abuse Treat. 2016;61:47-59. doi: 10.1016/j.jsat.2015.09.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Abdou R, Cassells D, Berrill J, Hanly J. An empirical investigation of the relationship between business performance and suicide in the US. Soc Sci Med. 2020;264:113256. doi: 10.1016/j.socscimed.2020.113256 [DOI] [PubMed] [Google Scholar]
51.Kazan D, Calear AL, Batterham PJ. The impact of intimate partner relationships on suicidal thoughts and behaviours: a systematic review. J Affect Disord. 2016;190:585-598. doi: 10.1016/j.jad.2015.11.003 [DOI] [PubMed] [Google Scholar]
52.Cicero TJ, Ellis MS, Kasper ZA. Polysubstance use: a broader understanding of substance use during the opioid crisis. Am J Public Health. 2020;110(2):244-250. doi: 10.2105/AJPH.2019.305412 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Kyung-Sook W, SangSoo S, Sangjin S, Young-Jeon S. Marital status integration and suicide: a meta-analysis and meta-regression. Soc Sci Med. 2018;197:116-126. doi: 10.1016/j.socscimed.2017.11.053 [DOI] [PubMed] [Google Scholar]
54.Yip PSF, Yousuf S, Chan CH, Yung T, Wu KCC. The roles of culture and gender in the relationship between divorce and suicide risk: a meta-analysis. Soc Sci Med. 2015;128:87-94. doi: 10.1016/j.socscimed.2014.12.034 [DOI] [PubMed] [Google Scholar]
55.Spottswood M, Lim CT, Davydow D, Huang H. Improving suicide prevention in primary care for differing levels of behavioral health integration: a review. Front Med (Lausanne). 2022;9:892205. doi: 10.3389/fmed.2022.892205 [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Centers for Disease Control and Prevention . Chronic diseases in America. Published December 13, 2022. Accessed November 7, 2023. https://www.cdc.gov/chronicdisease/resources/infographic/chronic-diseases.htm
57.Substance Abuse and Mental Health Services Administration (SAMHSA) . Integrated treatment for co-occurring disorders evidence-based practices (EBP) kit. Published February 5, 2021. Accessed October 27, 2023. https://www.samhsa.gov/resource/ebp/integrated-treatment-co-occurring-disorders-evidence-based-practices-ebp-kit
58.Minkoff K, Covell NH. Recommendations for integrated systems and services for people with co-occurring mental health and substance use conditions. Psychiatr Serv. 2022;73(6):686-689. doi: 10.1176/appi.ps.202000839 [DOI] [PubMed] [Google Scholar]
59.Treatment models and settings for people with co-occurring disorders. In: Substance Use Disorder Treatment for People With Co-Occurring Disorders. Substance Abuse and Mental Health Services Administration; 2020:chap 7. [PubMed] [Google Scholar]
60.Ee C, Lake J, Firth J, et al. An integrative collaborative care model for people with mental illness and physical comorbidities. Int J Ment Health Syst. 2020;14(1):83. doi: 10.1186/s13033-020-00410-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Higgins A, Murphy R, Downes C, et al. Factors impacting the implementation of a psychoeducation intervention within the mental health system: a multisite study using the consolidation framework for implementation research. BMC Health Serv Res. 2020;20(1):1023. doi: 10.1186/s12913-020-05852-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Brown GK, Ten Have T, Henriques GR, Xie SX, Hollander JE, Beck AT. Cognitive therapy for the prevention of suicide attempts: a randomized controlled trial. JAMA. 2005;294(5):563-570. doi: 10.1001/jama.294.5.563 [DOI] [PubMed] [Google Scholar]
63.Mehlum L, Tørmoen AJ, Ramberg M, et al. Dialectical behavior therapy for adolescents with repeated suicidal and self-harming behavior: a randomized trial. J Am Acad Child Adolesc Psychiatry. 2014;53(10):1082-1091. doi: 10.1016/j.jaac.2014.07.003 [DOI] [PubMed] [Google Scholar]
64.Gould MS, Pisani A, Gallo C, et al. Crisis text-line interventions: evaluation of texters’ perceptions of effectiveness. Suicide Life Threat Behav. 2022;52(3):583-595. doi: 10.1111/sltb.12873 [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Conner KR, Bagge CL, Goldston DB, Ilgen MA. Alcohol and suicidal behavior: what is known and what can be done. Am J Prev Med. 2014;47(3)(suppl 2):S204-S208. doi: 10.1016/j.amepre.2014.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Stanley AR, Aguilar T, Holland KM, Orpinas P. Precipitating circumstances associated with intimate partner problem-related suicides. Am J Prev Med. 2023;65(3):385-394. doi: 10.1016/j.amepre.2023.03.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
67.National Institute on Drug Abuse . Priority Scientific Area No. 2: develop and test novel prevention, treatment, harm reduction, and recovery support strategies. Accessed October 27, 2023. https://nida.nih.gov/about-nida/2022-2026-strategic-plan/priority-area-2
68.Connor JP, Gullo MJ, White A, Kelly AB. Polysubstance use: diagnostic challenges, patterns of use and health. Curr Opin Psychiatry. 2014;27(4):269-275. doi: 10.1097/YCO.0000000000000069 [DOI] [PubMed] [Google Scholar]
69.Bridge JA, Horowitz LM, Fontanella CA, et al. Age-related racial disparity in suicide rates among US youths from 2001 through 2015. JAMA Pediatr. 2018;172(7):697-699. doi: 10.1001/jamapediatrics.2018.0399 [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Ruch DA, Heck KM, Sheftall AH, et al. Characteristics and precipitating circumstances of suicide among children aged 5 to 11 years in the United States, 2013-2017. JAMA Netw Open. 2021;4(7):e2115683. doi: 10.1001/jamanetworkopen.2021.15683 [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Innamorati M, Pompili M, Masotti V, et al. Completed versus attempted suicide in psychiatric patients: a psychological autopsy study. J Psychiatr Pract. 2008;14(4):216-224. doi: 10.1097/01.pra.0000327311.04153.01 [DOI] [PubMed] [Google Scholar]
72.Buchman-Schmitt JM, Chu C, Michaels MS, et al. The role of stressful life events preceding death by suicide: evidence from two samples of suicide decedents. Psychiatry Res. 2017;256:345-352. doi: 10.1016/j.psychres.2017.06.078 [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Mann JJ, Michel CA. Prevention of firearm suicide in the United States: what works and what is possible. Am J Psychiatry. 2016;173(10):969-979. doi: 10.1176/appi.ajp.2016.16010069 [DOI] [PubMed] [Google Scholar]
74.Gallup . Guns. Published 2022. Accessed October 27, 2023. https://news.gallup.com/poll/1645/Guns.aspx
75.Mezuk B, Kalesnikava VA, Kim J, Ko TM, Collins C. Not discussed: inequalities in narrative text data for suicide deaths in the National Violent Death Reporting System. PLoS One. 2021;16(7):e0254417. doi: 10.1371/journal.pone.0254417 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplement 1.

eMethods 1. National Violent Death Reporting System (NVDRS) Data: Unique Value and Fit for the Current Study

eMethods 2. Rationale Behind the Selection of Indicators in the National Violent Death Reporting System (NVDRS) to Identify Suicide Death Profiles using Latent Class Analysis

eMethods 3. Reliability and Possible Misclassification Scenarios in the National Violent Death Reporting System (NVDRS) Data

eMethods 4. Strategies to Deal with Missing Data and Misclassification Error

eMethods 5. Addressing the Implications of Unbalanced Panel Data on Study Outcomes in the National Violent Death Reporting System (NVDRS) Data

eFigure 1. Scree Plot of Bayesian Information Criterion (BIC) Values Across Ten Latent Class Analysis Solutions

eFigure 2. Two-Class Latent Class Solution

eFigure 3. Three-Class Latent Class Solution

eFigure 4. Four-Class Latent Class Solution

eFigure 5. Five-Class Latent Class Solution

eFigure 6. Six-Class Latent Class Solution

eFigure 7. Seven-Class Latent Class Solution

eFigure 8. Eight-Class Latent Class Solution

eFigure 9. Nine-Class Latent Class Solution

eFigure 10. Ten-Class Latent Class Solution

eTable 1. Indicator Information including Definition, Original Response Categories, and Recoding Rules

eTable 2. Model Fit Statistics of Latent Class Analysis Models

eTable 3. Distribution of Precipitating Circumstances by Five Suicide Decedent Profiles in the U.S., 2003-2020 (N=306 800)

eTable 4. Results of Analysis without Missing Data Imputation

eTable 5. Class Prevalence by Year

eTable 6. Results of Analysis without Missing Data Imputation Additionally controlling for Fixed Effects of Year and State

eReferences

jamapsychiatry-e240171-s001.pdf^{(3.2MB, pdf)}

Supplement 2.

Data sharing statement

jamapsychiatry-e240171-s002.pdf^{(14.8KB, pdf)}

[yoi240006r1] 1.Centers for Disease Control and Prevention . Suicide increases in 2021 after two years of decline. Published September 30, 2022. Accessed June 4, 2023. https://www.cdc.gov/nchs/pressroom/nchs_press_releases/2022/20220930.htm

[yoi240006r2] 2.Mann JJ, Apter A, Bertolote J, et al. Suicide prevention strategies: a systematic review. JAMA. 2005;294(16):2064-2074. doi: 10.1001/jama.294.16.2064 [DOI] [PubMed] [Google Scholar]

[yoi240006r3] 3.Zalsman G, Hawton K, Wasserman D, et al. Suicide prevention strategies revisited: 10-year systematic review. Lancet Psychiatry. 2016;3(7):646-659. doi: 10.1016/S2215-0366(16)30030-X [DOI] [PubMed] [Google Scholar]

[yoi240006r4] 4.Jordan JT, McNiel DE. Characteristics of persons who die on their first suicide attempt: results from the National Violent Death Reporting System. Psychol Med. 2020;50(8):1390-1397. doi: 10.1017/S0033291719001375 [DOI] [PubMed] [Google Scholar]

[yoi240006r5] 5.Mann JJ, Michel CA, Auerbach RP. Improving suicide prevention through evidence-based strategies: a systematic review. Am J Psychiatry. 2021;178(7):611-624. doi: 10.1176/appi.ajp.2020.20060864 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r6] 6.Xiao Y, Cerel J, Mann JJ. Temporal trends in suicidal ideation and attempts among US adolescents by sex and race/ethnicity, 1991-2019. JAMA Netw Open. 2021;4(6):e2113513-e2113513. doi: 10.1001/jamanetworkopen.2021.13513 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r7] 7.García de la Garza Á, Blanco C, Olfson M, Wall MM. Identification of suicide attempt risk factors in a national US survey using machine learning. JAMA Psychiatry. 2021;78(4):398-406. doi: 10.1001/jamapsychiatry.2020.4165 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r8] 8.Favril L, Yu R, Uyar A, Sharpe M, Fazel S. Risk factors for suicide in adults: systematic review and meta-analysis of psychological autopsy studies. Evid Based Ment Health. 2022;25(4):148-155. doi: 10.1136/ebmental-2022-300549 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r9] 9.Cavanagh JTO, Carson AJ, Sharpe M, Lawrie SM. Psychological autopsy studies of suicide: a systematic review. Psychol Med. 2003;33(3):395-405. doi: 10.1017/S0033291702006943 [DOI] [PubMed] [Google Scholar]

[yoi240006r10] 10.Bostwick JM, Pabbati C, Geske JR, McKean AJ. Suicide attempt as a risk factor for completed suicide: even more lethal than we knew. Am J Psychiatry. 2016;173(11):1094-1100. doi: 10.1176/appi.ajp.2016.15070854 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r11] 11.McMahon EM, Greiner BA, Corcoran P, et al. Psychosocial and psychiatric factors preceding death by suicide: a case-control psychological autopsy study involving multiple data sources. Suicide Life Threat Behav. 2022;52(5):1037-1047. doi: 10.1111/sltb.12900 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r12] 12.Ross V, Kõlves K, De Leo D. Beyond psychopathology: a case-control psychological autopsy study of young adult males. Int J Soc Psychiatry. 2017;63(2):151-160. doi: 10.1177/0020764016688041 [DOI] [PubMed] [Google Scholar]

[yoi240006r13] 13.Liu GS, Nguyen BL, Lyons BH, et al. Surveillance for violent deaths: National Violent Death Reporting System, 48 states, the District of Columbia, and Puerto Rico, 2020. MMWR Surveill Summ. 2023;72(5):1-38. doi: 10.15585/mmwr.ss7205a1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r14] 14.Xiao Y, Zhang Z, Yip P, Cerel J, Mann JJ. Trends in suicide deaths with and without warning from 2003-2019: identifying the missing link for effective suicide prevention. 2022 Annual meeting of American Public Health Association APHA. Accessed June 22, 2023. https://apha.confex.com/apha/2022/meetingapp.cgi/Paper/518592

[yoi240006r15] 15.Cai Z, Junus A, Chang Q, Yip PSF. The lethality of suicide methods: a systematic review and meta-analysis. J Affect Disord. 2022;300:121-129. doi: 10.1016/j.jad.2021.12.054 [DOI] [PubMed] [Google Scholar]

[yoi240006r16] 16.Bond AE, Bandel SL, Rodriguez TR, Anestis JC, Anestis MD. Mental health treatment seeking and history of suicidal thoughts among suicide decedents by mechanism, 2003-2018. JAMA Netw Open. 2022;5(3):e222101. doi: 10.1001/jamanetworkopen.2022.2101 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r17] 17.Wang S, Dang Y, Sun Z, et al. An NLP approach to identify SDoH-related circumstance and suicide crisis from death investigation narratives. J Am Med Inform Assoc. 2023;30(8):1408-1417. doi: 10.1093/jamia/ocad068 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r18] 18.Rodway C, Tham SG, Turnbull P, Kapur N, Appleby L. Suicide in children and young people: can it happen without warning? J Affect Disord. 2020;275:307-310. doi: 10.1016/j.jad.2020.06.069 [DOI] [PubMed] [Google Scholar]

[yoi240006r19] 19.Yip PSF, Caine E, Yousuf S, Chang SS, Wu KCC, Chen YY. Means restriction for suicide prevention. Lancet. 2012;379(9834):2393-2399. doi: 10.1016/S0140-6736(12)60521-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r20] 20.Hochman E, Shelef L, Mann JJ, et al. Primary health care utilization prior to suicide: a retrospective case-control study among active-duty military personnel. J Clin Psychiatry. 2014;75(8):e817-e823. doi: 10.4088/JCP.13m08823 [DOI] [PubMed] [Google Scholar]

[yoi240006r21] 21.Mann JJ. Can knowledge of genetic and environmental causal factors of fatal and nonfatal suicidal behavior be translated into better prevention? Am J Psychiatry. 2021;178(11):994-997. doi: 10.1176/appi.ajp.2021.21090913 [DOI] [PubMed] [Google Scholar]

[yoi240006r22] 22.Xiao Y, Romanelli M, Lindsey MA. A latent class analysis of health lifestyles and suicidal behaviors among US adolescents. J Affect Disord. 2019;255:116-126. doi: 10.1016/j.jad.2019.05.031 [DOI] [PubMed] [Google Scholar]

[yoi240006r23] 23.Grant RW, McCloskey J, Hatfield M, et al. Use of latent class analysis and k-means clustering to identify complex patient profiles. JAMA Netw Open. 2020;3(12):e2029068-e2029068. doi: 10.1001/jamanetworkopen.2020.29068 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r24] 24.Logan J, Hall J, Karch D. Suicide categories by patterns of known risk factors: a latent class analysis. Arch Gen Psychiatry. 2011;68(9):935-941. doi: 10.1001/archgenpsychiatry.2011.85 [DOI] [PubMed] [Google Scholar]

[yoi240006r25] 25.Paulozzi LJ, Mercy J, Frazier L Jr, Annest JL; Centers for Disease Control and Prevention . CDC’s National Violent Death Reporting System: background and methodology. Inj Prev. 2004;10(1):47-52. doi: 10.1136/ip.2003.003434 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r26] 26.Wilson RF, Liu G, Lyons BH, et al. Surveillance for Violent Deaths: National Violent Death Reporting System, 42 states, the District of Columbia, and Puerto Rico, 2019. MMWR Surveill Summ. 2022;71(6):1-40. doi: 10.15585/mmwr.ss7106a1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r27] 27.Centers for Disease Control and Prevention, National Center for Injury Prevention and Control . National Violent Death Reporting System web coding manual version 6. Accessed February 13, 2024. https://www.cdc.gov/violenceprevention/datasources/nvdrs/resources.html

[yoi240006r28] 28.Lindsey MA, Xiao Y. Depression, trauma, and suicide among adolescent and young adult males. Men’s Health Equity. Accessed February 13, 2024. https://www.taylorfrancis.com/chapters/edit/10.4324/9781315167428-18/depression-trauma-suicide-among-adolescent-young-adult-males-michael-lindsey-yunyu-xiao

[yoi240006r29] 29.Mann JJ. Neurobiology of suicidal behaviour. Nat Rev Neurosci. 2003;4(10):819-828. doi: 10.1038/nrn1220 [DOI] [PubMed] [Google Scholar]

[yoi240006r30] 30.Joiner T. Why People Die by Suicide. Harvard University Press; 2009. [Google Scholar]

[yoi240006r31] 31.Klonsky ED, May AM. The three-step theory (3ST): a new theory of suicide rooted in the “ideation-to-action” framework. Int J Cogn Ther. 2015;8(2):114-129. doi: 10.1521/ijct.2015.8.2.114 [DOI] [Google Scholar]

[yoi240006r32] 32.Xiao Y, Hinrichs R, Johnson N, et al. Suicide prevention among college students before and during the COVID-19 pandemic: protocol for a systematic review and meta-analysis. JMIR Res Protoc. 2021;10(5):e26948. doi: 10.2196/26948 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r33] 33.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]

[yoi240006r34] 34.Kposowa AJ. Marital status and suicide in the National Longitudinal Mortality Study. J Epidemiol Community Health. 2000;54(4):254-261. doi: 10.1136/jech.54.4.254 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r35] 35.Hawley JE, Sun T, Chism DD, et al. Assessment of regional variability in COVID-19 outcomes among patients with cancer in the United States. JAMA Netw Open. 2022;5(1):e2142046. doi: 10.1001/jamanetworkopen.2021.42046 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r36] 36.Management and Budget Office, Federal Register . Revisions to the standards for the classification of federal data on race and ethnicity. Published October 30, 1997. Accessed February 13, 2024. https://www.federalregister.gov/d/97-28653

[yoi240006r37] 37.Linzer DA, Lewis JB. poLCA: an R package for polytomous variable latent class analysis. J Stat Softw. 2011;42:1-29. doi: 10.18637/jss.v042.i10 [DOI] [Google Scholar]

[yoi240006r38] 38.Laska MN, Pasch KE, Lust K, Story M, Ehlinger E. Latent class analysis of lifestyle characteristics and health risk behaviors among college youth. Prev Sci. 2009;10(4):376-386. doi: 10.1007/s11121-009-0140-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r39] 39.Mori M, Krumholz HM, Allore HG. Using latent class analysis to identify hidden clinical phenotypes. JAMA. 2020;324(7):700-701. doi: 10.1001/jama.2020.2278 [DOI] [PubMed] [Google Scholar]

[yoi240006r40] 40.Collins LM, Lanza ST. Latent Class and Latent Transition Analysis: With Applications in the Social, Behavioral, and Health Sciences. John Wiley & Sons; 2009. doi: 10.1002/9780470567333 [DOI] [Google Scholar]

[yoi240006r41] 41.van Buuren S, Groothuis-Oudshoorn K. mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(3). doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]

[yoi240006r42] 42.Xiao Y, Brown TT. Moving suicide prevention upstream by understanding the effect of flourishing on suicidal ideation in midlife: an instrumental variable approach. Sci Rep. 2023;13(1):1320. doi: 10.1038/s41598-023-28568-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r43] 43.Geale K, Henriksson M, Jokinen J, Schmitt-Egenolf M. Association of skin psoriasis and somatic comorbidity with the development of psychiatric illness in a nationwide swedish study. JAMA Dermatol. 2020;156(7):795-804. doi: 10.1001/jamadermatol.2020.1398 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r44] 44.Centers for Disease Control and Prevention . Disparities in suicide. Published May 9, 2023. Accessed November 6, 2023. https://www.cdc.gov/suicide/facts/disparities-in-suicide.html

[yoi240006r45] 45.Gibbons RD, Brown CH, Hur K, Davis J, Mann JJ. Suicidal thoughts and behavior with antidepressant treatment: reanalysis of the randomized placebo-controlled studies of fluoxetine and venlafaxine. Arch Gen Psychiatry. 2012;69(6):580-587. doi: 10.1001/archgenpsychiatry.2011.2048 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r46] 46.National Institute on Drug Abuse . Drug overdose death rates. Published June 30, 2023. Accessed November 6, 2023. https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates

[yoi240006r47] 47.Miller M, Swanson SA, Azrael D, Pate V, Stürmer T. Antidepressant dose, age, and the risk of deliberate self-harm. JAMA Intern Med. 2014;174(6):899-909. doi: 10.1001/jamainternmed.2014.1053 [DOI] [PubMed] [Google Scholar]

[yoi240006r48] 48.Clement S, Schauman O, Graham T, et al. What is the impact of mental health-related stigma on help-seeking? a systematic review of quantitative and qualitative studies. Psychol Med. 2015;45(1):11-27. doi: 10.1017/S0033291714000129 [DOI] [PubMed] [Google Scholar]

[yoi240006r49] 49.Priester MA, Browne T, Iachini A, Clone S, DeHart D, Seay KD. Treatment access barriers and disparities among individuals with co-occurring mental health and substance use disorders: an integrative literature review. J Subst Abuse Treat. 2016;61:47-59. doi: 10.1016/j.jsat.2015.09.006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r50] 50.Abdou R, Cassells D, Berrill J, Hanly J. An empirical investigation of the relationship between business performance and suicide in the US. Soc Sci Med. 2020;264:113256. doi: 10.1016/j.socscimed.2020.113256 [DOI] [PubMed] [Google Scholar]

[yoi240006r51] 51.Kazan D, Calear AL, Batterham PJ. The impact of intimate partner relationships on suicidal thoughts and behaviours: a systematic review. J Affect Disord. 2016;190:585-598. doi: 10.1016/j.jad.2015.11.003 [DOI] [PubMed] [Google Scholar]

[yoi240006r52] 52.Cicero TJ, Ellis MS, Kasper ZA. Polysubstance use: a broader understanding of substance use during the opioid crisis. Am J Public Health. 2020;110(2):244-250. doi: 10.2105/AJPH.2019.305412 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r53] 53.Kyung-Sook W, SangSoo S, Sangjin S, Young-Jeon S. Marital status integration and suicide: a meta-analysis and meta-regression. Soc Sci Med. 2018;197:116-126. doi: 10.1016/j.socscimed.2017.11.053 [DOI] [PubMed] [Google Scholar]

[yoi240006r54] 54.Yip PSF, Yousuf S, Chan CH, Yung T, Wu KCC. The roles of culture and gender in the relationship between divorce and suicide risk: a meta-analysis. Soc Sci Med. 2015;128:87-94. doi: 10.1016/j.socscimed.2014.12.034 [DOI] [PubMed] [Google Scholar]

[yoi240006r55] 55.Spottswood M, Lim CT, Davydow D, Huang H. Improving suicide prevention in primary care for differing levels of behavioral health integration: a review. Front Med (Lausanne). 2022;9:892205. doi: 10.3389/fmed.2022.892205 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r56] 56.Centers for Disease Control and Prevention . Chronic diseases in America. Published December 13, 2022. Accessed November 7, 2023. https://www.cdc.gov/chronicdisease/resources/infographic/chronic-diseases.htm

[yoi240006r57] 57.Substance Abuse and Mental Health Services Administration (SAMHSA) . Integrated treatment for co-occurring disorders evidence-based practices (EBP) kit. Published February 5, 2021. Accessed October 27, 2023. https://www.samhsa.gov/resource/ebp/integrated-treatment-co-occurring-disorders-evidence-based-practices-ebp-kit

[yoi240006r58] 58.Minkoff K, Covell NH. Recommendations for integrated systems and services for people with co-occurring mental health and substance use conditions. Psychiatr Serv. 2022;73(6):686-689. doi: 10.1176/appi.ps.202000839 [DOI] [PubMed] [Google Scholar]

[yoi240006r59] 59.Treatment models and settings for people with co-occurring disorders. In: Substance Use Disorder Treatment for People With Co-Occurring Disorders. Substance Abuse and Mental Health Services Administration; 2020:chap 7. [PubMed] [Google Scholar]

[yoi240006r60] 60.Ee C, Lake J, Firth J, et al. An integrative collaborative care model for people with mental illness and physical comorbidities. Int J Ment Health Syst. 2020;14(1):83. doi: 10.1186/s13033-020-00410-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r61] 61.Higgins A, Murphy R, Downes C, et al. Factors impacting the implementation of a psychoeducation intervention within the mental health system: a multisite study using the consolidation framework for implementation research. BMC Health Serv Res. 2020;20(1):1023. doi: 10.1186/s12913-020-05852-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r62] 62.Brown GK, Ten Have T, Henriques GR, Xie SX, Hollander JE, Beck AT. Cognitive therapy for the prevention of suicide attempts: a randomized controlled trial. JAMA. 2005;294(5):563-570. doi: 10.1001/jama.294.5.563 [DOI] [PubMed] [Google Scholar]

[yoi240006r63] 63.Mehlum L, Tørmoen AJ, Ramberg M, et al. Dialectical behavior therapy for adolescents with repeated suicidal and self-harming behavior: a randomized trial. J Am Acad Child Adolesc Psychiatry. 2014;53(10):1082-1091. doi: 10.1016/j.jaac.2014.07.003 [DOI] [PubMed] [Google Scholar]

[yoi240006r64] 64.Gould MS, Pisani A, Gallo C, et al. Crisis text-line interventions: evaluation of texters’ perceptions of effectiveness. Suicide Life Threat Behav. 2022;52(3):583-595. doi: 10.1111/sltb.12873 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r65] 65.Conner KR, Bagge CL, Goldston DB, Ilgen MA. Alcohol and suicidal behavior: what is known and what can be done. Am J Prev Med. 2014;47(3)(suppl 2):S204-S208. doi: 10.1016/j.amepre.2014.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r66] 66.Stanley AR, Aguilar T, Holland KM, Orpinas P. Precipitating circumstances associated with intimate partner problem-related suicides. Am J Prev Med. 2023;65(3):385-394. doi: 10.1016/j.amepre.2023.03.011 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r67] 67.National Institute on Drug Abuse . Priority Scientific Area No. 2: develop and test novel prevention, treatment, harm reduction, and recovery support strategies. Accessed October 27, 2023. https://nida.nih.gov/about-nida/2022-2026-strategic-plan/priority-area-2

[yoi240006r68] 68.Connor JP, Gullo MJ, White A, Kelly AB. Polysubstance use: diagnostic challenges, patterns of use and health. Curr Opin Psychiatry. 2014;27(4):269-275. doi: 10.1097/YCO.0000000000000069 [DOI] [PubMed] [Google Scholar]

[yoi240006r69] 69.Bridge JA, Horowitz LM, Fontanella CA, et al. Age-related racial disparity in suicide rates among US youths from 2001 through 2015. JAMA Pediatr. 2018;172(7):697-699. doi: 10.1001/jamapediatrics.2018.0399 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r70] 70.Ruch DA, Heck KM, Sheftall AH, et al. Characteristics and precipitating circumstances of suicide among children aged 5 to 11 years in the United States, 2013-2017. JAMA Netw Open. 2021;4(7):e2115683. doi: 10.1001/jamanetworkopen.2021.15683 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r71] 71.Innamorati M, Pompili M, Masotti V, et al. Completed versus attempted suicide in psychiatric patients: a psychological autopsy study. J Psychiatr Pract. 2008;14(4):216-224. doi: 10.1097/01.pra.0000327311.04153.01 [DOI] [PubMed] [Google Scholar]

[yoi240006r72] 72.Buchman-Schmitt JM, Chu C, Michaels MS, et al. The role of stressful life events preceding death by suicide: evidence from two samples of suicide decedents. Psychiatry Res. 2017;256:345-352. doi: 10.1016/j.psychres.2017.06.078 [DOI] [PMC free article] [PubMed] [Google Scholar]

[yoi240006r73] 73.Mann JJ, Michel CA. Prevention of firearm suicide in the United States: what works and what is possible. Am J Psychiatry. 2016;173(10):969-979. doi: 10.1176/appi.ajp.2016.16010069 [DOI] [PubMed] [Google Scholar]

[yoi240006r74] 74.Gallup . Guns. Published 2022. Accessed October 27, 2023. https://news.gallup.com/poll/1645/Guns.aspx

[yoi240006r75] 75.Mezuk B, Kalesnikava VA, Kim J, Ko TM, Collins C. Not discussed: inequalities in narrative text data for suicide deaths in the National Violent Death Reporting System. PLoS One. 2021;16(7):e0254417. doi: 10.1371/journal.pone.0254417 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Decoding Suicide Decedent Profiles and Signs of Suicidal Intent Using Latent Class Analysis

Yunyu Xiao, PhD

Kaiwen Bi, BA

Paul Siu-Fai Yip, PhD

Julie Cerel, PhD

Timothy T Brown, PhD

Yifan Peng, PhD

Jyotishman Pathak, PhD

J John Mann, MD

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Exposures

Main Outcomes and Measures

Results

Conclusions and Relevance

Introduction

Methods

Study Design, Setting, and Population

Measures

Statistical Analysis

Results

Table 1. Demographic Characteristicsa.

Latent Classes of Suicide Decedents

Figure 1. Distribution of Precipitating Circumstances by 5 Suicide Decedent Profiles in the US, 2003-2020 (N = 306 800).

Figure 2. Temporal Changes of Identified Latent Classes of Suicide Decedent Profiles in the US, 2003-2020.

Associations With Signs of Suicide Intent

Table 2. Distribution of Signs of Suicide Intent by 5 Suicide Decedent Profiles.

Table 3. Multivariable Logistic Regression Models With Suicide Intent Disclosure and Suicide Note Leaving as Outcomes.

Demographic Disparities

Discussion

Demographic Differences

Research and Policy Implications to Improve Suicide Prevention

Strengths and Limitations

Conclusions

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Table 1. Demographic Characteristics^a.