Abstract
Objective
We conducted a chart review to identify postsecondary students and nonstudents in the same age range who presented to the emergency department following a suicide attempt to (1) compare demographic characteristics and suicide risk factors and (2) determine factors associated with more serious attempts requiring hospitalizations.
Method
The study was conducted in 1 tertiary trauma hospital and 1 community hospital affiliated with McGill University, Montreal, Quebec, Canada, between January 1, 2009, and March 31, 2010. Charts of patients with potential suicide attempts were identified from medical records using ICD-10 codes that indicated traumatic injury, intentional self-harm, poisoning, and psychiatric or perception/cognition disorders and from the emergency department triage file using keywords that indicated suicidality or self-harm at presentation.
Results
In multivariable logistic regression models (odds ratio, 95% CI), students were younger (per 5-year increase: 0.22, 0.12–0.41), less likely to be born in Canada (0.17, 0.06–0.44), and more likely to use less violent methods (laceration, poisoning, other, multiple methods) versus more violent methods (collision, jump, fire burns, firearm, hanging) in their attempt. Fewer students had a history of substance abuse (0.12, 0.02–0.94) but were not different from nonstudents on history of other mental disorders. Less students attempted suicide in the winter/spring (January–April) versus fall (September–December) semester (0.32, 0.11–0.91). Students who attempted suicide were more likely to have family/social support. Those who attempted suicide in the previous year were more likely to require hospitalization for their current suicide attempt.
Conclusions
Knowledge of specific factors associated with suicide attempts in young people can help inform and guide suicide prevention efforts in both academic and community settings. Specific to the findings of this study regarding the method of suicide attempt used, for example, limiting access to dangerous substances or large quantities of medications may help prevent or reduce suicide attempts in this population.
Clinical Points
⊠ In an urban setting, about one third of young adults aged 18–36 years who received medical attention for attempted suicide were postsecondary students; over two thirds of students who attempted suicide requiring medical attention were female.
⊠ Students were more likely to attempt suicide in the fall compared to the winter/spring semester; the 2 most common methods of suicide among students were poisoning and laceration.
⊠ Students who attempted suicide were less likely than nonstudents to have a history of substance abuse and more likely to have family/social support.
⊠ Students who attempted suicide in the previous year were more likely to require hospitalization for their current suicide attempt.
Suicide is the second leading cause of death in North America among people aged 15–34 years following accidental injury.1,2 College years are a particularly vulnerable period for this population because of the pressure to succeed, increased financial burden, and added responsibility resulting from transitioning to adulthood. According to the Spring 2013 National College Health Assessment data for Canadian university and college students,3 most students reported feeling overwhelmed by their studies, 10% seriously considered taking their own lives, and 1% attempted suicide during the past year. Although the rate of attempted suicide in college students has been reported to be half the national rate (7.5/100,000),4 suicide in these young and promising individuals is particularly tragic and has devastating emotional and financial impacts on the family, peers, educational institution, and society. Preexisting or predisposition to mental disorders among students, failure to seek help, and lack of knowledge about available mental health services on campus have all been reported to contribute to suicide attempts in this population.4–7 Among young people, rates of suicidal thoughts but not behavior are higher among females, vary by age and ethnicity,8 and increase with a personal or family history of mental disorders.9 Among those who commit or attempt suicide, depression is the most prevalent mental disorder reported.10
Suicide risk factors among postsecondary (college or university) students in comparison to nonstudents of similar age have not been well characterized because of the absence of comprehensive data on suicide attempts in this population. Substance abuse,11 less family support, and more financial and societal stress10,12,13 as well as Asian versus white and international student status have been found to be factors that increase risk of suicidal behavior in students.14 However, only a small proportion of students who report suicidal behavior have actually attempted suicide.10 Published data15 concerning suicide attempts among postsecondary students have been inconsistent in their definition of an attempt. They have considered inclusion criteria ranging from suicide planning to self-reported attempts not requiring medical attention to attempts requiring hospitalization. The relatively low incidence of suicide attempts among postsecondary students, the nature of self-report data, the selection and participation biases inherit in survey data, and, finally, the wide variation of suicide attempt case definitions have precluded a robust assessment of the characteristics of suicide attempters in this population.4,10,16,17
Among individuals who presented to the emergency department (ED) of 2 adult general hospitals in Montreal, Quebec, Canada, following a suicide attempt, we identified postsecondary students and nonstudents in the same age range and (1) compared demographic characteristics and suicide risk factors between the 2 groups and (2) determined factors associated with more serious attempts requiring hospitalization.
METHOD
Study Design and Sample
A retrospective chart review of potential cases of attempted suicide was conducted. Potential cases were all individuals who presented to the ED of 2 general hospitals (1 tertiary trauma McGill University teaching hospital and 1 community hospital affiliated with McGill University) between January 1, 2009, and March 31, 2010. Study hospitals along with 2 other adult teaching hospitals are located in close proximity to 3 major university campuses in Montreal and a few colleges. Cases of suicide attempts were identified from hospital medical records and the ED nursing triage file. Case selection for chart review involved generating a list of probable ICD-10 revision codes indicating suspicious injury or behavior such as neck injury, poisoning, intentional self-harm, reaction to severe stress, adjustment disorders, and others to identify patients who received at least 1 of these codes as principal or secondary diagnosis or external cause of hospitalization in their hospital record (Table 1). This list included a broad range of codes to maximize the likelihood of capturing all suicide attempts at the expense of reviewing more negative charts. In addition, the electronic ED triage file was used to identify all patients who had an indication of suicidality or self-harm at ED presentation. Paper charts of the identified patients were retrieved, and all medical and nursing notes were reviewed.
Table 1.
Charts Selected and Suicide Attempt Cases Identified Using Hospital Medical Records and Emergency Department Triage Filesa
| Chart Review | ICD-10 Codes and Keywords | Total Charts Reviewed | All Suicide Attempts, n (%) | Suicide Attempts Included, n (%)b |
| Charts and suicide attempt cases | 5,746 | 369 (6.4) | 186 (3.2) | |
| Hospital medical record | ||||
| Intentional self-harm | X60–X84 or Y87.0 | 104 | 89 (85.6) | 41 (39.4) |
| Poisoning/toxic effect of unspecified substances | R78, T36–T65 | 96 | 13 (13.5) | 5 (5.2) |
| Injury (neck, multiple) | S10–S19, T00–T07, T79 | 353 | 6 (1.7) | 2 (0.6) |
| Undetermined intent and sequelae of external causes | Y10–Y34, Y40–Y89 (excludes Y87.0), T90–T98 | 2,357 | 8 (0.3) | 3 (0.1) |
| Cognition perception | R44–R46 | 92 | 28 (30.4) | 10 (10.9) |
| Psychiatric disorders | F10–F19, F25, F30–F39, F43, F53, F60.3c | 921 | 32 (3.5) | 14 (1.4) |
| Emergency department triage file | suicid, attempt, OD, overdo; gun, intox; poison, stab, jump, hang, hung, self, shoot, shot, slash, cut, mutilatd | 1,823 | 193 (10.6) | 111 (6.1) |
ICD-10 codes and number of charts are displayed in a mutually exclusive descending order.
Includes all postsecondary students and nonstudents within the same age range (18–36 y) and excludes homeless individuals. For patients with multiple attempts in the study period, only the last attempt was included.
F10–F19 = mental and behavioral disorders due to psychoactive substance use, F25 = schizoaffective disorders, F30–F39 = mood (affective) disorders, F43 = reaction to severe stress and adjustment disorders, F53 = mental and behavioral disorders associated with the puerperium not elsewhere classified, F60.3 = emotionally unstable personality disorder.
Similar keywords in the French language were also used.
Inclusion/Exclusion Criteria
Individuals who attempted suicide as identified by chart review were eligible for this study if they were postsecondary students or nonstudents in the same age range as the eligible students. Homeless individuals were excluded because they differ substantially from students on sociocultural and mental health disorder risk factors.
Study Procedures
Chart review.
In all Quebec hospitals, patients who arrive to the ED undergo a triage assessment to determine the priority of their needs.18 According to ED management guidelines, triage is conducted by an experienced nurse and follows the Canadian Emergency Department Triage and Acuity Scale19,20; the ED triage nurse is trained to recognize suicidal patients and is present in the ED at all times. A suicide management protocol is followed for the evaluation, treatment, management, and follow-up of suicidal patients. Software accredited by the Ministry of Health, Quebec, Canada, is used for ED triage.18
A suicide attempt was defined as explicit documentation of attempted suicide recorded by at least 2 different health care professionals during that episode of care (ED visit for those discharged from the ED or hospitalization for those admitted to the hospital). The first 20 cases identified were audited by our team to confirm the validity of the identification method. Ethics approval for chart review was obtained from the McGill University Health Centre Ethics Board, Montreal, Quebec, Canada.
Patient Characteristics at Baseline
Patient characteristics assessed from the charts included sociodemographic and sociocultural factors including age, sex, country of birth, ethnicity, employment status, living status (alone, with parent[s], or with spouse/partner/roommate), education level attained, and presence of social/family support; family history of mental disorders and suicidal behavior; suicide methods used in the attempt; history of mental disorders including major depressive episode, bipolar disorder, schizophrenia, panic disorder, generalized anxiety disorder, antisocial personality disorder, borderline personality disorder, and alcohol/drug abuse; previous suicide attempts; comorbidities (eg, cancer, cardiovascular disease, hepatitis, HIV, epilepsy, chronic pain, and others as recorded in the chart); and number of days in the hospital and mortality during that episode of care.
Statistical Analyses
Descriptive statistics (means, medians, and proportions as appropriate) were used to summarize characteristics of students versus nonstudents. Multivariable logistic regression models were used to compare patient characteristics between students and nonstudents. Patient characteristics listed previously were considered in the models if statistically significant in the univariate analyses at the 25% level and were kept in the model if statistically significant at the 5% level.21 The Schwarz information criterion was used to select the final model. Multivariable logistic regression models were used to compare the risk of hospitalization following the attempt in students and nonstudents combined. Similar model selection criteria as those mentioned previously were used. Multivariable logistic regression models were also used to identify factors associated with the risk of being hospitalized for 1 week or more (the median in-hospital days among those hospitalized).
The proportion of patients with missing data was low (less than 5%) for most key variables. Accounting for missing data through multiple imputations did not change the results of the models and will not be discussed further. All analyses were performed using SAS 9.4 for Linux (SAS Institute, Inc, Cary, North Carolina).
RESULTS
A total of 5,746 charts were reviewed, and 369 patients (aged 17–87 years) with suicide attempts were identified (Table 1). Of these, 61 were postsecondary students aged 18–36 years and 125 were nonstudents in the same age range (5 nonstudents were homeless and were excluded). Therefore, 186 individuals were eligible and included and 183 were not used in this study. Table 2 displays patient characteristics of students versus nonstudents. Students’ ages were (median, interquartile range reported in lower and upper quartiles) 22, 20–24 versus 27, 24–31 years. Among students, 70% (vs 59%) were women, 46% (vs 73%) were born in Canada, and 52% (vs 62%) were white. Most students were single, and over one third of them were living with their parents. Poisoning was the method of most attempts for both students and nonstudents followed by laceration. Because of few cases, collision, jump, fire burns, firearm injuries, and hanging/strangulation were grouped into 1 category. About half of the students had a history of mental disorder, mostly depression, and over a quarter of them had a mental illness in the family (all [100%] had a mother with depression) (Table 2).
Table 2.
Patient Sociodemographic and Clinical Characteristics in Postsecondary Students and Nonstudents of the Same Age Rangea
| Characteristic | Students | Nonstudents | Total |
| Total patients | 61 (33) | 125 (67) | 186 (100) |
| Female | 43 (70) | 74 (59) | 117 (63) |
| Age, median (IQR), y | 22 (20–24) | 27 (24–31) | 25 (21–29) |
| Citizenship | |||
| Canadian born | 28 (46) | 91 (73) | 119 (64) |
| Canadian citizens/permanent residentsb | 47 (77) | 113 (90) | 160 (86) |
| Student visa | 13 (21) | 0 | 13 (7) |
| Ethnicity | |||
| White | 32 (52) | 77 (62) | 109 (59) |
| Asian | 13 (21) | 9 (7) | 22 (12) |
| Arab | 6 (10) | 8 (6) | 14 (8) |
| Black | 4 (7) | 10 (8) | 14 (8) |
| American Indian | 0 | 11 (9) | 11 (6) |
| Latin American | 2 (3) | 5 (4) | 7 (4) |
| Other | 4 (7) | 5 (4) | 9 (5) |
| Level of education achieved | |||
| Students (current level) | |||
| College (2 y) | 19 (31) | … | 19 (10) |
| University | 42 (69) | … | 42 (23) |
| Nonstudents (level achieved) | |||
| < High school | … | 62 (50) | 62 (33) |
| High school | … | 35 (28) | 35 (19) |
| College | … | 7 (6) | 7 (4) |
| University | … | 16 (13) | 16 (9) |
| Other (professional diploma) | … | 5 (4) | 5 (3) |
| Employed | |||
| Full-time | … | 45 (36) | 45 (24) |
| Part-time | … | 10 (8) | 10 (5) |
| Self-employed | … | 2 (2) | 2 (1) |
| Unemployed | … | 66 (53) | 66 (35) |
| Other | … | 2 (2) | 2 (1) |
| Marital status | |||
| Single | 44 (72) | 69 (55) | 113 (61) |
| Married | 3 (5) | 7 (6) | 10 (5) |
| Divorced/separated | 5 (8) | 14 (11) | 19 (10) |
| Widowed | 0 | 1 (1) | 1 (0.5) |
| Engaged/cohabitating | 6 (10) | 24 (19) | 30 (16) |
| Unknown | 3 (5) | 10 (8) | 13 (7) |
| Children | |||
| 0 | 54 (89) | 85 (68) | 139 (75) |
| ≥ 1 | 1 (2) | 28 (22) | 29 (16) |
| Unknown | 6 (10) | 12 (10) | 18 (10) |
| Living status | |||
| With 1 or both parents | 23 (38) | 28 (22) | 51 (27) |
| With roommate/spouse or partner | 25 (41) | 60 (48) | 85 (46) |
| Alone | 13 (21) | 37 (30) | 50 (27) |
| History of any mental disorder | 33 (54) | 88 (70) | 121 (65) |
| Axis I | 30 (49) | 83 (66) | 113 (61) |
| Depression | 27 (44) | 41 (33) | 68 (37) |
| Substance abuse | 4 (7) | 34 (27) | 38 (20) |
| Schizophrenia | 0 | 13 (10) | 13 (7) |
| Bipolar disorder | 1 (2) | 5 (4) | 6 (3) |
| Other | 4 (7) | 15 (12) | 19 (10) |
| Axis II | 12 (20) | 24 (19) | 36 (19) |
| Borderline personality disorder | 8 (13) | 15 (12) | 23 (12) |
| Unspecified disorders | 1 (2) | 4 (3) | 5 (3) |
| Attempt in past year | 17 (28) | 25 (20) | 42 (23) |
| Any mental disorder in family | 16 (26) | 46 (37) | 62 (33) |
| Mother/father | 15 (25) | 40 (32) | 55 (30) |
| Depression in family | 13 (21) | 21 (17) | 34 (18) |
| Mother | 13 (21) | 17 (14) | 30 (16) |
| Suicide in family | 1 (2) | 9 (7) | 10 (5) |
| Comorbidity | 16 (26) | 26 (21) | 42 (23) |
| Cancer | 1 (2) | 2 (2) | 3 (2) |
| Respiratory | 3 (5) | 8 (6) | 11 (6) |
| Chronic pain | 3 (5) | 1 (1) | 4 (2) |
| Diabetes | 0 | 2 (2) | 2 (1) |
| Hepatitis | 0 | 1 (1) | 1 (0.5) |
| Thyroid | 1 (2) | 5 (4) | 6 (3) |
| HIV | 1 (2) | 0 | 1 (0.5) |
| Inflammatory bowel disease | 3 (5) | 0 | 3 (2) |
| Other (eg, epilepsy, cardiovascular disease, brain injury) | 5 (8) | 8 (6) | 13 (7) |
| Died in hospital | 0 | 2 (2) | 2 (1) |
| Method used | |||
| Poisoning | 42 (69) | 60 (48) | 102 (55) |
| Laceration | 17 (28) | 22 (18) | 39 (21) |
| Collision/jump/fire burns/firearm | 5 (8) | 24 (19) | 29 (16) |
| Hanging/strangulation | 5 (8) | 14 (11) | 19 (10) |
| Other | 1 (2) | 9 (7) | 10 (5) |
| Social/family support | 33 (54) | 27 (22) | 60 (32) |
| Admission to hospital at presentation | 21 (34) | 59 (47) | 80 (43) |
| In-hospital days, total (median) [IQR] | 300 (1) [0–5] | 1,126 (1) [0–6] | 1,426 (1) [0–6] |
| 0 | 24 (39) | 33 (26) | 57 (31) |
| 1–2 | 16 (26) | 44 (35) | 60 (32) |
| 3–9 | 11 (18) | 28 (22) | 39 (21) |
| ≥ 10 | 10 (16) | 20 (16) | 30 (16) |
| Time of attempt | |||
| September–December 2009 | 21 (34) | 29 (23) | 50 (27) |
| January–April 2009 | 15 (25) | 33 (26) | 48 (26) |
| May–August 2009 | 12 (20) | 37 (30) | 49 (26) |
| January–March 2010 | 13 (21) | 26 (21) | 39 (21) |
Data are presented as n (%) unless otherwise specified.
bPermanent residents are mostly immigrants who have not yet obtained the Canadian citizenship, but have all civil rights as Canadian citizens.
Abbreviation: IQR = interquartile range.
In multivariable logistic regression models (odds ratio [OR], 95% CI), students were younger (per 5-year increase: 0.22, 0.12–0.41), less likely to be born in Canada (0.17, 0.06–0.44), and more likely to use less violent methods (laceration, poisoning, other, multiple methods) versus more violent methods (collision, jump, fire burns, firearm, hanging) in their attempt. Fewer students had a history of substance abuse (0.12, 0.02–0.94) but were not different from nonstudents on history of other mental disorders. Less students attempted suicide in the winter/spring (January–April) versus fall (September–December) semester (0.32, 0.11–0.91) (Table 3).
Table 3.
Characteristics That Differed Between Postsecondary Students Versus Nonstudents
| Characteristic | Odds Ratio (95% CI)a |
| Age (per 5-y increase) | 0.22 (0.12–0.41) |
| Born in Canada | 0.17 (0.06–0.44) |
| History of any mental disorder | |
| No mental disorder | 1 (reference) |
| Depression | 1.37 (0.52–3.64) |
| Substance abuse | 0.12 (0.02–0.94) |
| Other | 0.26 (0.06–1.21) |
| Method used | |
| Collision/jump/fire burns/ firearm/hangingb | 1 (reference) |
| Laceration | 5.43 (1.16–25.37) |
| Poisoning | 4.10 (1.16–14.47) |
| Other | 14.00 (0.62–315.79) |
| ≥ 2 methods used | 9.46 (1.87–47.74) |
| Time of attempt | |
| September–December | 1 (reference) |
| January–April | 0.32 (0.11–0.91) |
| May–August | 0.58 (0.17–1.96) |
| Social/family support | 3.23 (1.33–7.81) |
The model used logistic regression. All characteristics listed in Table 1 were tried for inclusion. Those significant at the .05 level were kept.
This category was chosen as the reference category to compare less violent to more violent methods.
In total, 80 patients (43%) were hospitalized (Table 4): 34% of students and 47% of nonstudents (Table 2). The risk (OR, 95% CI) of making an attempt that required hospitalization increased with age (for every 5-year increase: 1.45, 1.01–2.07). Individuals living with someone other than their parents (roommate/spouse/partner) were less likely to attempt a suicide requiring hospitalization than those living with 1 or both parents (0.32, 0.14–0.76); the risk of requiring hospitalization was not different between those living alone and those living with 1 or both parents. Those who used poisoning and laceration methods were less likely to require hospitalization compared to those who used more violent methods (collision, jump, fire burns, firearm, hanging). In addition, those who attempted suicide in the previous year were more than 3 times as likely to require hospitalization as those who did not. The risk of being hospitalized did not differ between students and nonstudents (Table 5).
Table 4.
Demographic and Clinical Characteristics in Hospitalized Versus Nonhospitalized Students and Nonstudents Combineda
| Characteristic | Nonhospitalized | Hospitalized | Total |
| Total patients | 106 (57) | 80 (43) | 186 (100) |
| Female | 72 (68) | 45 (56) | 117 (63) |
| Age, median (IQR), y | 24 (21–28) | 26 (22–30) | 25 (21–29) |
| Citizenship | |||
| Canadian born | 66 (62) | 53 (66) | 119 (64) |
| Canadian citizens/permanent residentsb | 90 (85) | 70 (88) | 160 (86) |
| Student visa | 8 (8) | 5 (6) | 13 (7) |
| Ethnicity | |||
| White | 61 (58) | 48 (60) | 109 (59) |
| Asian | 15 (14) | 7 (9) | 22 (12) |
| Arab | 5 (5) | 9 (11) | 14 (8) |
| Black | 9 (8) | 5 (6) | 14 (8) |
| American Indian | 6 (6) | 5 (6) | 11 (6) |
| Latin American | 3 (3) | 4 (5) | 7 (4) |
| Other | 7 (7) | 2 (3) | 9 (5) |
| Level of education (current level) | |||
| College (2 y) | 12 (11) | 7 (9) | 19 (10) |
| University | 28 (26) | 14 (18) | 42 (23) |
| Nonstudents (level achieved) | |||
| < High school | 35 (33) | 27 (34) | 62 (33) |
| High school | 18 (17) | 17 (21) | 35 (19) |
| College | 2 (2) | 5 (6) | 7 (4) |
| University | 7 (7) | 9 (11) | 16 (9) |
| Other (professional diploma) | 4 (4) | 1 (1) | 5 (3) |
| Employed | |||
| Full-time | 23 (22) | 22 (28) | 45 (24) |
| Part-time | 6 (6) | 4 (5) | 10 (5) |
| Self-employed | 0 | 2 (3) | 2 (1) |
| Unemployed | 36 (34) | 30 (38) | 66 (35) |
| Student | 40 (38) | 21(26) | 61 (33) |
| Other | 1 (1) | 1 (1) | 2 (1) |
| Marital status | |||
| Single | 60 (57) | 53 (66) | 113 (61) |
| Married | 8 (8) | 2 (3) | 10 (5) |
| Divorced/separated | 10 (9) | 9 (11) | 19 (10) |
| Widowed | 1 (1) | 0 | 1 (0.5) |
| Engaged/cohabitating | 19 (18) | 11 (14) | 30 (16) |
| Unknown | 8 (8) | 5 (6) | 13 (7) |
| Children | |||
| 0 | 78 (74) | 61 (76) | 139 (75) |
| ≥1 | 18 (17) | 11 (14) | 29 (16) |
| Unknown | 10 (9) | 8 (10) | 18 (10) |
| Living status | |||
| With 1 or both parents | 25 (24) | 26 (33) | 51 (27) |
| With roommate/spouse | 60 (57) | 25 (31) | 85 (46) |
| Alone | 21 (20) | 29 (36) | 50 (27) |
| History of any mental disorder | 64 (60) | 57 (71) | 121 (65) |
| Axis I disorder | 59 (56) | 54 (68) | 113 (61) |
| Depression | 38 (36) | 30 (38) | 68 (37) |
| Substance abuse | 23 (22) | 15 (19) | 38 (20) |
| Schizophrenia | 2 (2) | 11 (14) | 13 (7) |
| Bipolar disorder | 3 (3) | 3 (4) | 6 (3) |
| Other | 10 (9) | 9 (11) | 19 (10) |
| Axis II disorder | 20 (19) | 16 (20) | 36 (19) |
| Borderline personality disorder | 14 (13) | 9 (11) | 23 (12) |
| Unspecified | 3 (3) | 2 (3) | 5 (3) |
| Attempt in past year | 17 (16) | 25 (31) | 42 (23) |
| Any mental disorder in family | 31 (29) | 31 (39) | 62 (33) |
| Mother/father | 30 (28) | 25 (31) | 55 (30) |
| Depression in family | 16 (15) | 18 (23) | 34 (18) |
| Mother | 15 (14) | 15 (19) | 30 (16) |
| Suicide in family | 3 (3) | 7 (9) | 10 (5) |
| Comorbidity | 21 (20) | 21 (26) | 42 (23) |
| Cancer | 1 (1) | 2 (3) | 3 (2) |
| Respiratory | 8 (8) | 3 (4) | 11 (6) |
| Chronic pain | 3 (3) | 1 (1) | 4 (2) |
| Diabetes | 0 | 2 (3) | 2 (1) |
| Hepatitis | 0 | 1(1) | 1 (0.5) |
| Thyroid | 1 (1) | 5 (6) | 6 (3) |
| HIV | 0 | 1 (1) | 1 (0.5) |
| Inflammatory bowel disease | 3 (3) | 0 | 3 (2) |
| Other (eg, epilepsy, cardiovascular disease, brain injury) | 6 (6) | 7 (9) | 13 (7) |
| Died in hospital | 0 | 2 (3) | 2 (1) |
| Method used | |||
| Poisoning | 68 (64) | 34 (43) | 102 (55) |
| Laceration | 27 (25) | 12 (15) | 39 (21) |
| Collision/jump/fire burns/firearm | 6 (6) | 23 (29) | 29 (16) |
| Hanging/strangulation | 7 (7) | 12 (15) | 19 (10) |
| Other | 5 (5) | 5 (6) | 10 (5) |
| Social/family support | 36 (34) | 24 (30) | 60 (32) |
| In-hospital days, total (median) [IQR] | 55 (0) [0–1] | 1,371 (7) [4–18] | 1,426 (1) [0–6] |
| 0 | 56 (53) | 1 (1)c | 57 (31) |
| 1–2 | 50 (47) | 10 (13) | 60 (32) |
| 3–9 | … | 39 (49) | 39 (21) |
| ≥ 10 | … | 30 (38) | 30 (16) |
| Time of attempt | |||
| September–December 2009 | 25 (24) | 25 (31) | 50 (27) |
| January–April 2009 | 29 (27) | 19 (24) | 48 (26) |
| May–August 2009 | 26 (25) | 23 (29) | 49 (26) |
| January–March 2010 | 26 (25) | 13 (16) | 39 (21) |
Data are presented as n (%) unless otherwise specified.
Permanent residents are mostly immigrants who have not yet obtained the Canadian citizenship, but have all civil rights as Canadian citizens.
This person died shortly after admission.
Abbreviation: IQR = interquartile range.
Table 5.
Factors Associated With Suicide Attempts Requiring Hospitalization Among Students and Nonstudents Combined
| Factor | Odds Ratio (95% CI)a |
| Age (per 5-y increase) | 1.45 (1.01–2.07) |
| Living status | |
| Lives with 1 or both parents | 1 (reference) |
| Lives alone | 0.93 (0.36–2.38) |
| Lives with roommate spouse or partner | 0.32 (0.14–0.76) |
| Attempt in past year | 3.75 (1.64–8.55) |
| Method used | |
| Collision/jump/fire burns/ firearm/hanging | 1 (reference) |
| Laceration | 0.13 (0.04–0.41) |
| Poisoning | 0.17 (0.07–0.43) |
| Other | 0.25 (0.04–1.71) |
| ≥ 2 methods used | 0.22 (0.06–0.75) |
The model used logistic regression. All characteristics listed in Table 4 were tried for inclusion. Those significant at the .05 level were kept. “Student” was not a significant variable and was removed.
Study patients spent 1,371 days in the hospital and 55 days in the ED (Table 4). Among the 80 individuals who were hospitalized, 45 were hospitalized for 1 week or more. Individuals who attempted suicide in the previous year were more than 5 times as likely to be hospitalized for 1 week or more compared to individuals who did not, while those who used poisoning or laceration versus more violent methods of suicide (collision, jump, fire burns, firearm, hanging) were less likely to be hospitalized for 1 week or more (Table 6).
Table 6.
Factors Associated With a Longer Hospitalization (≥ 1 Week)a
| Hospitalized |
Hospitalized Versus Emergency Department Odds Ratio (95% CI) |
||||
| Factor | Emergency Department | < 7 d | ≥ 7 d | < 7 d | ≥ 7 d |
| Total number of patients, n (%) | 106 (57) | 35 (19) | 45 (24) | ||
| Living status, n (%) | |||||
| Live with 1 or both parents | 25 (24) | 11 (31) | 15 (33) | 1 (reference) | 1 (reference) |
| Live alone | 21 (20) | 13 (37) | 16 (36) | 1.25 (0.44–3.58) | 1.20 (0.40, 3.59) |
| Live with roommate spouse or partner | 60 (57) | 11 (31) | 14 (31) | 0.34 (0.12–0.95) | 0.52 (0.19, 1.43) |
| Attempt in past year, n (%) | 17 (16) | 9 (26) | 16 (36) | 2.30 (0.84–6.30) | 5.44 (2.01, 14.71) |
| Method used, n (%) | |||||
| Collision/jump/fire burns/firearm/hanging | 12 (11) | 9 (26) | 22 (49) | 1 (reference) | 1 (reference) |
| Laceration | 22 (21) | 4 (11) | 7 (16) | 0.30 (0.06–1.44) | 0.08 (0.02–0.32) |
| Poisoning | 55 (52) | 18 (51) | 9 (20) | 0.58 (0.18–1.87) | 0.06 (0.02–0.20) |
| Other | 4 (4) | 2 (6) | 1 (2) | 1.10 (0.13–9.30) | 0.14 (0.01–1.75) |
| ≥ 2 methods used | 13 (12) | 2 (6) | 6 (13) | 0.25 (0.04–1.55) | 0.17 (0.05–0.65) |
DISCUSSION
Our study compared postsecondary students to nonstudents of similar age who attempted suicide and presented to the ED of 1 general tertiary trauma hospital and 1 general community hospital in Montreal between January 2009 and March 2010. About one third of our study subjects were students, and the majority of them were female. Students were younger than nonstudents. Students were less likely to be born in Canada and to have a history of substance abuse and more likely to have family/social support. Students also were more likely to have used laceration or poisoning versus more violent methods of suicide and to have attempted suicide during the fall versus winter/spring semester.
Several authors have found that college years are particularly stressful for students6,22–25 and that elevated stress and vulnerability to mental illness are manifest by high rates of suicidal ideation and behavior.4,26,27 The actual rate of attempted suicide among postsecondary students is not well known because of the difficulty in tracking students who have attempted suicide16 and because of the inherent reporting and selection biases in studies using survey data.4 In a web-based survey of college students,4 1% of undergraduates and 0.3% of graduates reported having attempted suicide in the past year. Among those, 19% of undergraduates and 28% of graduates required medical attention. Our study included only suicide attempts that required medical attention and were treated in either an ED or hospital setting in an urban area. Therefore, direct comparison of our results to those from survey studies should be done with caution.
Of the postsecondary students in our study, 70% were female. Although this proportion was not statistically different from that among nonstudents (59%), it was statistically higher than the reported proportion of females (58%) enrolled in postsecondary education in Quebec in 2009.28,29 Further examination of our data did not provide additional insight that could explain the higher rate of female students who attempted suicide compared to that among enrolled postsecondary students. Published data regarding sex differences among postsecondary students concerning suicidal thoughts and behavior are not consistent. While some studies reported a higher rate of suicidal thoughts but not behavior among female students,8 others reported no difference on suicidal thoughts30 or attempts in the previous year.31 However, these studies were based on student surveys, and participation was much higher among females versus males, which may have biased the results. For example, in Garlow et al,31 72% of participants were female, while the student population of the university was only 56% female. Interestingly, in that study,31 the number of females (n = 95) who reported having attempted suicide in the previous year was much higher than that of males (n = 25), which corroborate the numbers found in our study. Further studies are needed to examine sex differences in suicide attempts among postsecondary students.
In our study, students were more likely to attempt suicide in the fall compared to the winter/spring semester. This finding may be explained perhaps by the higher level of stress at the time of enrollment. However, this assumption cannot be verified in our data, as the year of enrollment was not indicated in the charts. Another study32 has reported a higher risk of suicidal ideation in the summer compared to both fall and winter/spring semesters. However, results from that study32 cannot be directly compared to ours, as very few suicidal ideations translate into actual attempts.10 In addition, it is possible that the rate of suicide attempts among students treated in our study hospitals during summer was underestimated because some students may have moved from the region during their summer vacation.
In our study, students were less likely to be born in Canada compared to nonstudents, and about 21% of them were international compared to an 11% proportion of international students registered in Quebec colleges and universities in 2009, a proportion that varies between universities and is much higher in Montreal compared to universities in other Quebec regions.29 Although we cannot determine from our sample if the risk of suicide attempt in the postsecondary student population in general differed between those born in Canada and those born elsewhere, students not born in Canada are more likely to be immigrants and therefore more likely to have financial and societal stress; students with more financial and societal stress10,12,13 as well as international students are at increased risk of suicidal behavior.14
An association between maternal depression and suicidal behavior among college students has been reported.33 In our study, maternal depression was not different between students and nonstudents, although the possibility of a trend toward a higher proportion in students could not be excluded. History of depression did not differ between students and nonstudents in our study, nonetheless about half of students had a history of depression. The association between depression and suicidal ideation is well known.34–38 However, less than 25% of young adults with a mental illness seek treatment.4,10,26,31,39 Efforts to promote mental health literacy, which is associated with decreased stigma and increased help-seeking behavior, may help decrease the risk of suicide attempts in this population.
In our study, students were less likely to have a history of substance abuse and more likely to have family/social support than nonstudents. A lower risk of drug use disorder and nicotine dependence among college students compared to peers not attending college has been reported.39 Students with no family/social support were found to be more likely to drop out of school and not enter college in some studies.12,22,40,41 In addition, emotional, informational, and tangible social support was associated with less suicidal thoughts and behaviors among college students, while negative social exchanges were associated with increased suicidal thoughts and behaviors.13 Friendship also has been found to play a role in suicidality in some ethnic groups, although this relationship remains poorly understood.42
Suicide attempt in the prior year was associated with suicide attempts that required hospitalization in our study. Although other studies did not assess the risk of suicide attempts requiring hospitalization versus those treated in the ED, many studies9,43,44 reported a higher risk of reattempts among those with prior suicide attempts. Older age and living with 1 or both parents versus with a partner/roommate/spouse were also associated with suicide attempts requiring hospitalization in our study; these findings require further investigation in future studies.
Knowledge of specific factors associated with suicide attempts in young people can help inform and guide suicide prevention efforts in both academic and community settings. Specific to the findings of this study regarding the method of suicide attempt used, for example, limiting access to dangerous substances or large quantity of medications may help prevent or reduce suicide attempts in this population.
Our study has several strengths over previous studies. We used both medical records and ED triage files to identify suicide attempts. In Quebec, triage nurses are trained to recognize suicide cases at presentation, and an ED suicidal patient management protocol is in place in all hospitals. Patients presenting with signs of mental illness receive a comprehensive suicide evaluation and a detailed psychiatric evaluation that are recorded in the charts.18 Moreover, missing data were minimal in our study, not exceeding 5% for most key variables. In addition, methods of data imputation were used to address this issue in the analyses. Therefore, we believe that our suicide identification method has captured most suicide attempts that presented to the ED of the study hospitals during the study period. The inclusion of all students who presented to the ED of the study hospitals for suicide attempts limited selection and reporting bias in our study in comparison to published studies4,16,33 that have been based on surveys. Using our method of data collection, we were able to assess and compare these factors between students and nonstudents in a reliable way.
Our study also has some limitations. First, data obtained from hospital paper-based chart review may suffer from missing information; however, this was addressed previously and did not affect the final results. Second, the study was based on data from a convenience sample of 2 hospitals located in proximity to numerous universities and colleges. Although the proportion of students may be higher in our study hospitals, it is likely that characteristics of students and nonstudents do not differ greatly between these hospitals and the remaining hospitals in Quebec. Finally, this study is limited to young adults who sought help at the ED and excludes those in the community who do not seek help after a suicide attempt.
In conclusion, in this study, one third of all individuals aged 18 to 36 years who attempted suicide and were treated in 2 general hospitals in Montreal were postsecondary students. Over half of the students were not born in Canada, half were depressed at the time of the attempt, and one fifth were international. Compared to nonstudents, students were less likely to have a history of substance abuse and more likely to have family/social support. They were more likely to attempt suicide in the fall versus winter/spring semester and more likely to use poisoning and laceration compared to more violent methods (collision, jump, fire burns, firearm, hanging) of suicide.
Potential conflict of interest:
Dr Renaud has received grant/research support from Standard Life Centre for Breakthroughs in Teen Depression and Suicide Prevention and Canadian Institutes of Health Research. Dr Lesage has received grant/research support from Quebec’s Suicide Research Network funded by the Quebec Health Research Fund. Drs Rahme, Low, Lamarre, Turecki, Bonin, Morin, Szkrumelak, and Singh; Mss Daneau and Yung; and Mr Habel report no conflicts of interest related to the subject of this article.
Funding/support:
This work was supported by the Quebec Network on Suicide, Mood Disorders and Related Disorders (RQSHA). RQSHA is funded by the Fonds de recherche du Québec-Santé and Société et culture. Dr Turecki is the director, Dr Lesage is the codirector, and Drs Rahme, Low, Lamarre, Bonin, and Renaud and Ms Daneau are members of RQSHA.
Role of the sponsor:
RQSHA had no direct role in designing or conducting the study or in writing or reviewing the manuscript.
Acknowledgment:
The authors thank Jivko Jelev, MD, and Elena Molchynska, MD (Research Institute of the McGill University Health Centre, Montreal, Canada), for chart review and data collection. Drs Jelev and Molchynska report no conflicts of interest.
References
- 1.Ranking and number of deaths for the 10 leading causes of death by age group, Canada, 2009. Statcan website. www.statcan.gc.ca/pub/84-215-x/2012001/table-tableau/tbl003-eng.htm. Updated October 3, 2013. Accessed June 17, 2015.
- 2.10 leading causes of death by age group, United States. 2013. CDC website. http://www.cdc.gov/injury/images/lc-charts/leading_causes_of_death_by_age_group_2013-a.gif. Accessed June 17, 2015.
- 3.Farabaugh A, Bitran S, Nyer M, et al. Depression and suicidal ideation in college students. Psychopathology. 2012;45(4):228–234. doi: 10.1159/000331598. [DOI] [PubMed] [Google Scholar]
- 4.Drum D, Brownson C, Burton A, et al. New data on the nature of suicidal crises in college students: shifting the paradigm. American College Health Association, 2006. Prof Psychol Res Pr. 2009;40(3):213–222. [Google Scholar]
- 5.Gallagher R. Survey of Counseling Center Directors. Arlington, VA: International Association of Counseling Services; http://www.sprc.org/collegesanduniversities/research/national-survey-counseling-center-directors . Accessed June 17, 2015. [Google Scholar]
- 6.Benton S, Robertson J, Tseng W, et al. Changes in counseling center client problems across 13 years. Prof Psychol Res Pr. 2003;34(1):66–72. [Google Scholar]
- 7.Masuda A, Boone M. Mental health stigma, self-concealment, and help-seeking attitudes among Asian American and European American college students with no help-seeking experience. Int J Adv Couns. 2011;33(4):266–279. [Google Scholar]
- 8.Crosby AE, Han B, Ortega LA, et al. Centers for Disease Control and Prevention (CDC). Suicidal thoughts and behaviors among adults aged ≥18 years: United States, 2008–2009. MMWR Surveill Summ. 2011;60(13):1–22. [PubMed] [Google Scholar]
- 9.Johnson JG, Cohen P, Gould MS, et al. Childhood adversities, interpersonal difficulties, and risk for suicide attempts during late adolescence and early adulthood. Arch Gen Psychiatry. 2002;59(8):741–749. doi: 10.1001/archpsyc.59.8.741. [DOI] [PubMed] [Google Scholar]
- 10.Kisch J, Leino EV, Silverman MM. Aspects of suicidal behavior, depression, and treatment in college students: results from the spring 2000 national college health assessment survey. Suicide Life Threat Behav. 2005;35(1):3–13. doi: 10.1521/suli.35.1.3.59263. [DOI] [PubMed] [Google Scholar]
- 11.Brener ND, Hassan SS, Barrios LC. Suicidal ideation among college students in the United States. J Consult Clin Psychol. 1999;67(6):1004–1008. doi: 10.1037//0022-006x.67.6.1004. [DOI] [PubMed] [Google Scholar]
- 12.Maimon D, Browning CR, Brooks-Gunn J. Collective efficacy, family attachment, and urban adolescent suicide attempts. J Health Soc Behav. 2010;51(3):307–324. doi: 10.1177/0022146510377878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hirsch JK, Barton AL. Positive social support, negative social exchanges, and suicidal behavior in college students. J Am Coll Health. 2011;59(5):393–398. doi: 10.1080/07448481.2010.515635. [DOI] [PubMed] [Google Scholar]
- 14.Wong SS, Sugimoto-Matsuda JJ, Chang JY, et al. Ethnic differences in risk factors for suicide among American high school students, 2009: the vulnerability of multiracial and Pacific Islander adolescents. Arch Suicide Res. 2012;16(2):159–173. doi: 10.1080/13811118.2012.667334. [DOI] [PubMed] [Google Scholar]
- 15.Miller EJ, Chung H. A literature review of studies of depression and treatment outcomes among US college students since 1990. Psychiatr Serv. 2009;60(9):1257–1260. doi: 10.1176/ps.2009.60.9.1257. [DOI] [PubMed] [Google Scholar]
- 16.Haas A, Hendin H, Mann J. Suicide in college students. Am Behav Sci. 2003;46(9):1224–1240. [Google Scholar]
- 17.Skala K, Kapusta ND, Schlaff G, et al. Suicidal ideation and temperament: an investigation among college students. J Affect Disord. 2012;141(2-3):399–405. doi: 10.1016/j.jad.2012.03.010. [DOI] [PubMed] [Google Scholar]
- 18.Association québecoise d’établissement de santé et de services sociaux. guide de gestion de l’urgence. Sante et Services Sociaux Quebec website. http://publications.msss.gouv.qc.ca/acrobat/f/documentation/2006/06-905-01.pdf. Updated September 2006. Accessed June 17, 2015.
- 19.Bullard MJ, Unger B, Spence J, et al. CTAS National Working Group. Revisions to the Canadian Emergency Department Triage and Acuity Scale (CTAS) adult guidelines. CJEM. 2008;10(2):136–151. doi: 10.1017/s1481803500009854. [DOI] [PubMed] [Google Scholar]
- 20.Beveridge R, Ducharme J, Janes L, et al. Reliability of the Canadian Emergency Department Triage and Acuity Scale: interrater agreement. Ann Emerg Med. 1999;34(2):155–159. doi: 10.1016/s0196-0644(99)70223-4. [DOI] [PubMed] [Google Scholar]
- 21.Hosmer D, Lemeshow S. Applied Survival Analysis: Regression Modeling of Time-to-Event Data. New York, NY: Wiley-Interscience; 1999. [Google Scholar]
- 22.Galambos NL, Barker ET, Krahn HJ. Depression, self-esteem, and anger in emerging adulthood: seven-year trajectories. Dev Psychol. 2006;42(2):350–365. doi: 10.1037/0012-1649.42.2.350. [DOI] [PubMed] [Google Scholar]
- 23.Hofstra MB, Van der Ende J, Verhulst FC. Pathways of self-reported problem behaviors from adolescence into adulthood. Am J Psychiatry. 2002;159(3):401–407. doi: 10.1176/appi.ajp.159.3.401. [DOI] [PubMed] [Google Scholar]
- 24.Hooven C, Herting JR, Snedker KA. Long-term outcomes for the promoting CARE suicide prevention program. Am J Health Behav. 2010;34(6):721–736. doi: 10.5993/ajhb.34.6.8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shea RH. On the edge on campus: the state of college students’ mental health continues to decline. what’s the solution? US News World Rep. 2002;132(5):56–57. [PubMed] [Google Scholar]
- 26.Westefeld JS, Homaifar B, Spotts J, et al. Perceptions concerning college student suicide: data from four universities. Suicide Life Threat Behav. 2005;35(6):640–645. doi: 10.1521/suli.2005.35.6.640. [DOI] [PubMed] [Google Scholar]
- 27.Zisook S, Downs N, Moutier C, et al. College students and suicide risk: prevention and the role of academic psychiatry. Acad Psychiatry. 2012;36(1):1–6. doi: 10.1176/appi.ap.10110155. [DOI] [PubMed] [Google Scholar]
- 28.Chatel-DeRepentigny J, Montmarquette C, Vaillancourt F. Les étudiants internationaux au Québec: état des lieux, impacts économiques et politiques publiques. Cirano website. http://www.cirano.qc.ca/pdf/publication/2011s-71.pdf. Updated November 2011. Accessed June 17, 2015.
- 29. Effectif universitaire, selon le sexe, le type d’inscription, et le type de programme, Canada et provinces, 1998–1999, 2003–2004 et 2008–2009. Statistique Canada website. http://www.statcan.gc.ca/pub/81-582-x/2010004/tbl/tbld1.5-fra.htm. Updated December 13, 2010. Accessed June 17, 2015.
- 30.Tupler LA, Hong JY, Gibori R, et al. Suicidal ideation and sex differences in relation to 18 major psychiatric disorders in college and university students: anonymous web-based assessment. J Nerv Ment Dis. 2015;203(4):269–278. doi: 10.1097/NMD.0000000000000277. [DOI] [PubMed] [Google Scholar]
- 31.Garlow SJ, Rosenberg J, Moore JD, et al. Depression, desperation, and suicidal ideation in college students: results from the American Foundation for Suicide Prevention College Screening Project at Emory University. Depress Anxiety. 2008;25(6):482–488. doi: 10.1002/da.20321. [DOI] [PubMed] [Google Scholar]
- 32.Van Orden KA, Witte TK, James LM, et al. Suicidal ideation in college students varies across semesters: the mediating role of belongingness. Suicide Life Threat Behav. 2008;38(4):427–435. doi: 10.1521/suli.2008.38.4.427. [DOI] [PubMed] [Google Scholar]
- 33.Wilcox HC, Arria AM, Caldeira KM, et al. Prevalence and predictors of persistent suicide ideation, plans, and attempts during college. J Affect Disord. 2010;127(1–3):287–294. doi: 10.1016/j.jad.2010.04.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hirsch JK, Visser PL, Chang EC, et al. Race and ethnic differences in hope and hopelessness as moderators of the association between depressive symptoms and suicidal behavior. J Am Coll Health. 2012;60(2):115–125. doi: 10.1080/07448481.2011.567402. [DOI] [PubMed] [Google Scholar]
- 35.Rice K, Leever B, Christopher J, et al. Perfectionism, stress and social (dis)connection: a short-term study of hopelessness, depression, and academic adjustment among honours students. J Couns Psychol. 2006;53(4):524–534. [Google Scholar]
- 36.Lewinsohn PM, Rohde P, Seeley JR, et al. Gender differences in suicide attempts from adolescence to young adulthood. J Am Acad Child Adolesc Psychiatry. 2001;40(4):427–434. doi: 10.1097/00004583-200104000-00011. [DOI] [PubMed] [Google Scholar]
- 37.Harrington R, Pickles A, Aglan A, et al. Early adult outcomes of adolescents who deliberately poisoned themselves. J Am Acad Child Adolesc Psychiatry. 2006;45(3):337–345. doi: 10.1097/01.chi.0000194564.78536.f2. [DOI] [PubMed] [Google Scholar]
- 38.Schulenberg JE, Zarrett N. Mental health during emerging adulthood: Continuity and discontinuity in courses, causes, and functions. In: Arnett JJ, Tanner JL, editors. Emerging Adults in America: Coming of Age in the 21st Century. Washington, DC: American Psychological Association; 2006. pp. 135–172. [Google Scholar]
- 39.Blanco C, Okuda M, Wright C, et al. Mental health of college students and their non–college-attending peers: results from the National Epidemiologic Study on Alcohol and Related Conditions. Arch Gen Psychiatry. 2008;65(12):1429–1437. doi: 10.1001/archpsyc.65.12.1429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Borowsky IW, Ireland M, Resnick MD. Adolescent suicide attempts: risks and protectors. Pediatrics. 2001;107(3):485–493. doi: 10.1542/peds.107.3.485. [DOI] [PubMed] [Google Scholar]
- 41.van Wel F, ter Bogt T, Raaijmakers Q. Changes in the parental bond and the well-being of adolescents and young adults. Adolescence. 2002;37(146):317–333. [PubMed] [Google Scholar]
- 42.Winterrowd E, Canetto SS, Chavez EL. Friendship factors and suicidality: common and unique patterns in Mexican American and European American youth. Suicide Life Threat Behav. 2011;41(1):50–65. doi: 10.1111/j.1943-278X.2010.00001.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Fergusson DM, Horwood LJ, Ridder EM, et al. Suicidal behavior in adolescence and subsequent mental health outcomes in young adulthood. Psychol Med. 2005;35(7):983–993. doi: 10.1017/s0033291704004167. [DOI] [PubMed] [Google Scholar]
- 44.Fergusson DM, Woodward LJ, Horwood LJ. Risk factors and life processes associated with the onset of suicidal behavior during adolescence and early adulthood. Psychol Med. 2000;30(1):23–39. doi: 10.1017/s003329179900135x. [DOI] [PubMed] [Google Scholar]