Characteristics and Risk Factors for Suicide and Deaths Among College Students: A 23-Year Serial Prevalence Study of Data From 8.2 Million Japanese College Students

Related Blog

Characteristics and Risk Factors for Suicide and Deaths Among College Students:

A 23-Year Serial Prevalence Study of Data From 8.2 Million Japanese College Students

Chiyoko Uchida, MDa,*, and Mai Uchida, MDb,c

Suicidality in college students is an international problem that has garnered attention from clinicians, schools, families, and public health institutions. In 2009, the Healthy Minds Study1–3 reported that 7% of US college students have seriously considered suicide, 2% made plans for suicide, and 1% attempted suicide. Another study4 reported that 6% of undergraduates and 4% of graduate students in the United States reported having seriously considered suicide in the previous 12 months. In late adolescence through young adulthood, youths are expected to rapidly form their identities and are exposed to psychosocial risk factors such as academic stress and relationship troubles, which are associated with psychiatric difficulties and suicide.5,6 Furthermore, research has shown that the majority of psychiatric illnesses become fully diagnosable before age 24, an age range including a large portion of college students.7,8 The direct associations between suicide and existing psychiatric diseases including major depressive disorder have been reported in college-aged youths.9,10 Taken together, this research identifies college students as a population vulnerable to suicide.

The issue of suicidality in college students extends beyond the United States. Japan’s national suicide rate has consistently been among the highest in the world, and the number of deaths by suicide reaches approximately 30,000 per year. The high rate of suicide extends to adolescents and youths; suicide is the leading cause of death among Japanese citizens in their teens and twenties.11 As the substantial risk of suicidality persists internationally, colleges have the unique opportunity to provide care by way of connecting students to mental health resources and integrated academic and social activities. Therefore, there is a need to understand the current state of suicide among college students and to identify characteristics that put students at high risk in order to effectively provide preventative care.

To gain a better understanding of the state of affairs surrounding suicide among college students in non-US countries, we conducted a study examining the data on suicide and other causes of death from a large longitudinal dataset collected on Japanese national university students during the 23 years between the 1989-1990 and 2011-2012 academic years. The sample consisted of a total of 8,262,314 college students and 2,889 deaths, including 1,226 suicides. Through analysis of this rich dataset, we aimed to characterize the suicides of college students regarding their psychiatric, academic, and social backgrounds in order to identify the high-risk groups. By clarifying the characteristics of the individuals, backgrounds, and means of the suicides from this large and longitudinal sample, we hope to support the establishment of preventive measures.

METHODS

Subjects

Subjects were college students attending a Japanese national or public college from 1989 to 2012.

Survey

Our group conducted annual surveys of dropouts, holdovers, and deaths among national university students since 1978 (the first survey having been conducted in 1979). National universities were asked in writing to participate in the survey. A questionnaire form was then sent to the appropriate representative and section (the head of the health management center and the staff member in charge of students’ numerical statistics) to collect the responses from each university that agreed to participate in the survey. The survey was completed either on a paper version or an electronic version. Anonymous data are reported for the protection of personal information. The surveys were returned to the laboratory of the first author (C.U.) and were stored electronically.

• Suicide is a leading cause of death for Japanese college students. This study was aimed at identification of risk factors of suicide among college students, which could, in turn, help improve university services to prevent college suicide.
• Men, medical students, and final-year and holdover students were at high risk for suicide. Female medical and dental students had the same risk for suicide as their male counterparts.
• The majority of students who died by suicide did not receive psychiatric diagnoses or university health services, highlighting the low utilization of psychiatric services among students who needed interventions.

In relation to students’ deaths during each academic year beginning on April 30 (which is 1 month after a typical academic year starts in Japan), the following data were collected: sex; age; year of admission; major (faculty and department); status of holdover; notified cause of death; detailed cause of death; method and place of suicide; status of psychiatric diagnosis; status of diagnosis according to the International Classification of Diseases, Tenth Revision (ICD-10)12; and status of involvement of the university health center. The response psychiatric diagnosis present means that a medical certificate has been issued by a psychiatrist. Any ICD-10 diagnosis entered in the questionnaire form returned from each university was based on the opinion of the student’s primary physician or the psychiatrist in attendance at the university health center. In some cases, when no diagnosis was entered on the form, the diagnosis was made by the author if the disorder could be diagnosed based on data obtained. The involvement of the university health center was determined based on the student’s visit for a mental health consultation. The response visit present does not necessarily mean that the visit took place around the time of the suicide.

The survey also included the numerical count of active students as of April 30 of that year. Respondents reported the counts according to sex, year of admission, faculty, and department.

The number of deaths and mortality by cause were calculated from the active student counts and obtained results of the survey items. The suicide rates were also calculated by sex, major, and academic year.

In this study, we report on the serial annual surveys13 of deaths among the majority of Japanese national university students conducted from 1989 to 2012. Of the 82 national universities, 60 (73.2%) participated in the most recent survey,14 conducted in 2012. In Japanese colleges, students must declare a major prior to enrolling in college. The humanities and science majors are 4-year programs, and medicine, dentistry, and veterinary medicine majors are 6-year programs. The pharmacy major was a 4-year program until 2006, when it became a 6-year program; therefore, pharmacy major students enrolled between 2006 and 2011 were analyzed as 6-year program students. Medicine, dentistry, veterinary medicine, and pharmacy majors do not require a previous undergraduate degree and are considered college majors rather than specialty schools.

Statistical Analysis

We analyzed the rates of the overall mortality per year as well as types of death from 1989 through 2012. To examine the change in mortality rates over the 23 years that were surveyed, mortality rate was compared among 4 periods: 1989 to 1993, 1994 to 1999, 2000 to 2005, and 2006 to 2011. We analyzed the rate of suicide among this population, with a focus on student sex; academic major; sex difference among each academic major; academic year in each major, including holdover status caused by leave of absence or repeating years; psychiatric diagnosis; methods of suicide; and involvement of university health centers. The data were statistically analyzed by χ2 test to examine whether a significant difference existed among the groups. Then residual analysis was done to see which group’s measured value was significantly higher or lower than the expected value. The Φ coefficient was used for effect size of 2 ×— 2 cross tabulation, and Cramer V was used for effect size of other cross tabulation.

Compliance With Ethical Guidelines

This study was approved at the general conference in 1999 to be in compliance with the Ethical Guidelines for Epidemiologic Research issued by the Japanese Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labor, and Welfare.

RESULTS

Participating Universities and Enrolled Students per Year

Table 1 shows the total number of national universities that participated in the survey and the number of enrolled students in the participating universities of each surveyed year from the 1989-1990 academic year to the 2011-2012 academic year. From the initial survey, which was completed by 48 of the 95 national universities (50.53%), to the most recent survey, which was completed by 60 of 82 national universities (73.17%), our surveys were completed by the majority of national universities in Japan throughout the examined 23 years. The highest number of participating universities reached 87 (91.58%) in 1999. In total, the data of 8,262,314 students were analyzed.

Mortality by All Causes

The total number of deaths was 2,889, including 1,226 suicides (42.4%). The detailed numbers of students by major and deaths by cause are presented in Table 2.

The differences of incidence of deaths by cause were statistically significant (χ24 = 1,857.914, P < .01). Ryan’s multiple comparison test showed significantly higher incidences of death by suicide and accidents (P < .01). Men had higher rates than women for deaths by suicide (χ21 = 103.894, P < .01, Φ = 0.003), accidents (χ21 = 141.353, P < .01, Φ = 0.004), and medical illness (χ21 = 48.844, P < .01, Φ = 0.002). There were no differences in rates of murder between the sexes.

Changes in Mortality by Cause Over 23 Years

Figure 1A shows the overall mortality rate per year, and Figure 1B shows mortality by cause per year over the 23 years. A χ2 test revealed statistically significant differences among the periods (χ23 = 41.182, P < .01, Cramer V = 0.002). A residual analysis showed that overall mortality was significantly higher for the periods 1989 to 1993 and 1994 to 1999 (P < .01) and significantly lower for the period 2006 to 2011 (P < .01), indicating a decreasing trend in overall mortality.

A χ2 test showed statistically significant differences of suicide rates among the periods (χ23 = 40.279, P < .01, Cramer V = 0.002). A residual analysis showed that the incidence of suicide was significantly lower for the periods 1989 to 1993 and 2000 to 2005 (P < .01) and significantly higher for the period 2006 to 2011 (P < .01), indicating a recent increase in suicide.

A χ2 test revealed statistically significant differences of accidental deaths among the periods (χ23 = 172.905, P < .01, Cramer V = 0.004). A residual analysis showed that the incidence of accidental death was significantly higher for the periods 1989 to 1993 and 1994 to 1999 (P < .01) and significantly lower for the periods 2000 to 2005 and 2006 to 2011 (P < .01), indicating a decrease in accidental death.

Suicide

Academic majors. Table 3 shows the total number of suicides per academic major. A χ2 test revealed statistically significant differences among the 6 academic majors (χ25 = 18.606, P < .01, Cramer V = 0.001). A residual analysis showed that medical students were significantly more likely to die by suicide (P < .01), and students majoring in humanities were significantly less likely to die by suicide (P < .05).

Sexes among each academic major. As presented in Figure 2, significant differences were found between the male and female suicide rates among humanities (χ21 = 75.555, P < .01, Φ = 0.004) and science (χ21 = 38.650, P < .01, Φ = 0.003) students. A residual analysis showed that men were significantly more likely to die by suicide in both humanities and science majors (P < .01). There were no significant differences between the likelihood of suicide between male and female students among medical and dental students. The data on sex of veterinary and pharmacy students could not be statistically tested due to the small sample size.

Academic year in each major. Among humanities students, a χ2 test revealed statistically significant differences between academic years (χ24 = 325.624, P < .01, Cramer V = 0.009). Similarly, among science students, a χ2 test revealed statistically significant differences among academic years (χ24 = 381.839, P < .01, Cramer V = 0.009). For both majors, a residual analysis showed that students in their fourth year and those in their fifth year or higher because of repeating academic years or taking leaves of absence were significantly more likely to die by suicide (P < .01), and first- and second-year students were significantly less likely to die by suicide (P < .01) (Figure 3A). For both majors, the suicide rate was highest in students in their fifth year or higher.

Among medical students, a χ2 test revealed statistically significant differences among academic years (χ26 = 42.783, P < .01, Cramer V = 0.009). A residual analysis showed that students in their seventh year or higher because of repeating academic years or taking leaves of absence were significantly more likely to die by suicide (P < .01) and first-year students were significantly less likely to die by suicide (P < .05) (Figure 3B).

The data on academic years of dentistry, veterinary, and pharmacy students could not be statistically tested due to the presence of cells with no sample.

Psychiatric diagnosis. Of the students who died by suicide, 201 (16.4%) had a psychiatric diagnosis. The remaining 1,025 students (83.6%) did not. The categories of psychiatric diagnosis for these 201 students are listed in Table 4. When the rates of each psychiatric diagnostic category were compared, a χ2 test revealed statistically significant differences among the diagnoses (χ29 = 415.667, P < .01). Ryan’s multiple comparison test showed a higher rate of mood disorders than other diagnoses (P < .01). Psychotic disorders and stress-related disorders (P < .01) were the second most common disorders.

Method of suicide. Table 5 lists the number of students who died by suicide and their methods. A χ2 test revealed statistical differences among the methods of suicide (χ210 = 2,273.197, P < .01). Ryan’s multiple comparison test showed that hanging was used significantly more frequently than other methods (P < .01). The second most common method was jumping (P < .01).

Between male and female students, there were statistically significant differences for hanging (χ21 = 5.812, P < .05) and overdose (χ21 = 6.469, P < .05). Hanging was significantly more frequent for men than for women (P < .05), whereas overdose was significantly more frequent for women than for men (P < .05).

Involvement of university health centers. Table 6 shows the number of students who died by suicide according to the involvement status of university health centers. Of the students who died by suicide, 196 (16%) had involvement with the university health center and 70.7% did not.

DISCUSSION

This study used serial prevalence data from the majority of Japanese national colleges to examine the characteristics and backgrounds of deaths and suicides. Our results documented that while overall mortality and accidental deaths have decreased over the 23 years surveyed, college suicide has increased. Since 1996, suicide has been the leading cause of death among university students. Men were significantly more prone to commit suicide than women, except for students majoring in medicine or dentistry. In medicine and dentistry, women and men were equally likely to commit suicide. Medical students were most vulnerable to suicide among all majors. A higher rate of suicidal death was found in students who were held back, who were in their fifth year or higher in 4-year programs, and who were in their seventh year or higher in 6-year programs, suggesting that holdover students are at increased risk for suicide. Only a minority of the students who committed suicide received a psychiatric diagnosis (the majority of which were mood disorders, psychotic disorders, and stress-related disorders) or had involvement with university health centers, findings that confirm the low utilization of psychiatric care. Hanging and jumping were the most common methods of suicide.

Decreased Deaths by Accidents, Increased Suicide

Our results documented that overall mortality and accidental deaths of college students have decreased, while suicides have increased and have been the leading cause of death since 1996. This is consistent with epidemiologic reports11,15 published through the Japanese police department on the overall Japanese youth population that document a sharp decline of accidental deaths after 1993 and an increase in suicidal deaths.

Men Have Higher Risk for Suicide Than Women

Our results documented that men were more likely than women to commit suicide. This is consistent with a US study16 reporting that the suicide rate of female undergraduate students is half the rate of male students. Our previous reports13,14,17 on leaves of absence, dropouts, and holdovers among Japanese college students also showed that all of these events occurred more frequently for male than for female students. Several factors may contribute to the lower rate of suicide and fewer leaves of absence, dropouts, and holdovers in female students. In a male-dominant society such as Japan, expectations for female students to achieve higher education are not high, and females attend college at approximately 70% the rate of males, thus suggesting that women who reach college are particularly resilient. As another factor, female students receive mental health consultations at a higher rate than male students, allowing females to access help prior to contemplating or attempting suicide.18,19 This indicates that college mental health programs could focus more outreach efforts on male students.

Suicide in Medical Students: High Suicide Risk, Comparable Risk Between Males and Females, and High Rates Among Holdover Students

The medical student population showed a significantly higher risk for suicide compared to those of other majors. Also, unlike other majors, female students were as prone to suicide as male students were. The high suicide rate among medical students and the equal rates of suicide among male and female students are consistent with the international literature.20-22

Across all majors, we found that medical students with a holdover status had the highest rate of suicide. As previously reported in our work13,14,17 on leaves of absence, dropouts, and holdovers among college students, these 3 events occurred much less frequently for students in 6-year programs (medicine, dentistry, and veterinary medicine majors) than for those in 4-year programs (humanities and science majors). We hypothesized that the lower incidences of leaves of absence, dropouts, and holdovers among the students in the 6-year program majors are due to having a clearer curriculum and an expectation for their professional careers in these majors. While these characteristics are helpful to most students, for the students who have deviated from the originally set course by being held back, the clear curriculum and expectations could present as stressors that lead to psychiatric disturbances. This may be one factor that contributed to our findings of the high suicide rate among holdover students in Japanese medical schools.

High Suicide Rate in Final-Year Students in 4-Year Programs, and Students Who Needed to Repeat Academic Years in All Majors

Our findings confirmed a higher risk for suicide in students in their final year of 4-year programs and in holdover students in all majors. The last year of college comes with increased stress related to preparing oneself for the future, such as completing a graduation thesis and seeking a career path. This may contribute to the high suicide rate and warrants special attention. Our results also suggest that particular attention should be given to holdover students. We have previously reported that holdover students are also at high risk for isolating themselves or lapsing into student apathy.17,23 Studies from the United Kingdom24,25 and the United States16 also suggest higher risks in students in later academic years. Hence, support for this cohort of students is critical. Establishment of a system that reaches out to students with insufficient credits and the creation of support groups may be beneficial.

Low Rates of Diagnosis and Treatment of Psychiatric Disorders

Our study documented that the most common psychiatric diagnoses observed in the students who died by suicide were mood disorders, followed by psychotic disorders and stress-related disorders. However, 84% of the students who died by suicide were never diagnosed with a psychiatric disorder. While there may be a minority of suicides that were not related to any psychiatric disorder, it is likely that the majority of these students had an undiagnosed psychiatric disorder that did not receive appropriate medical attention. Our results also showed that a university health center was involved with less than 20% of the students who committed suicide. This highlights the low utilization of services among college students who needed intervention and support and suggests the need for improved presence and accessibility of intra-university services.

Methods of Suicide

Our results indicated that hanging, jumping, and gas (such as carbon monoxide and hydrogen sulfide) poisoning, in this order, were the most common methods of suicide among college students. Comparing this result to the method of suicide in other countries, it is noteworthy that the use of firearms is not included in the Japanese college population, likely due to the strict gun control laws in Japan. In the United States, a high rate of gun suicide has been reported, especially among male students,26 and reduced access to firearms on college campuses has been reported to lead to lower suicide rates in students.16,27

The changes in method of suicide over the years observed in our study also appear to reflect the cluster suicides following a suicide that had been portrayed in the media. Our survey showed that in 2005, unlike in previous years, suicide by gas poisoning was the second most common method.13,14,28 This timing followed a period when Japanese media covered suicides of people who obtained “suicide companions” over the internet and used charcoal briquettes in closed cars to inhale lethal amounts of poisonous gas such as carbon monoxide. Our results show that there was a sharp increase in suicides by gas poisoning following this period of media coverage, indicating that college age youth may be vulnerable to media reports and suggesting that media reports exclude detailed accounts of methods of suicide.

Suggestions for Suicide Prevention in Colleges

Previous works on college suicides have recommended the following prevention actions. Furr et al29 suggested that in suicide prevention education, priority should be given to (1) warning signs of suicide, (2) supportive actions that should be taken by a student who recognizes the warning signs in others, and (3) information on the presence of an intra-university mental health center. King et al30 highlight the importance of encouraging students that they can prevent the suicide of a friend by taking action such as advising the friend to seek help. Joffe31 reported on the efficacy of a university program tested at the University of Illinois. The program, which began in 1984, was based on a mandatory approach in which students who have indicated a suicidal thought or attempted suicide must have 4 consultation or assessment sessions provided by a university psychiatric specialist. After the implementation of the program, the suicide rate dropped by 45.3%.

Taking the above successful interventions in consideration along with our reported results, we have the following suggestions as potential interventions that may reduce the suicide rate among college students.

First, in light of the extremely low rates of health center involvement and psychiatric diagnosis among the students who died by suicide, it is clear that interventions aimed at improving psychoeducation and access to care are necessary. This measure can be implemented by informing students and staff of the presence of university mental health services through websites, brochures, and presentations at new-student orientations; in lectures; and during the new-student physical examination and psychiatric interview upon the students’ entry in the college (new-student physical examinations are currently mandatory in Japan, and many universities include a psychiatric interview in this assessment). Additionally, university health centers should be made more accessible to students. Because deep-seated prejudice against psychiatric disorders persists, particularly in Eastern Asian countries such as Japan, it is important that students view the center as a safe place where their privacy will be protected. Architectural designs to eliminate student fears of being seen by other students when entering the health center for psychiatric care may be helpful. Another approach is to provide students with psychoeducation about suicide risk factors and psychiatric disorders so that they can identify these risk factors in themselves and their peers and intervene when necessary.

Second, preventative measures to combat the high risk for suicide among students who need to repeat academic years should be considered. We suggest the establishment of a system that reaches out to students with insufficient credits identified at the administrative level. A large portion of holdover students withdraw from their college environment and often have little connection with others,17 increasing their risk for student apathy and loneliness, as well as depression and suicide, and making it more difficult for them to take initiative to seek help. Hence, colleges may better help holdover students if they would reach out to them rather than putting on them the burden of seeking care. We advise colleges to establish an administrative method of identifying students who have not registered for necessary classes or have insufficient credits or long absences so that those students can be individually contacted and their parents can be notified. Additionally, it is important for faculty and staff members to be specifically informed of the high risk of suicide among holdover students. Faculty and staff members who observe the students at close proximity and on a regular basis are in the best position to identify and support high-risk students before administrators or treaters become aware of a problem.

It is also advisable to allocate support staff, such as a facilitator, teaching assistant, or mentor, to the group of holdover students. Academic and social support for this high-risk group could help these students remain connected to the college environment as well as decrease academic stress.

Another approach can be to provide groups for students who need to repeat academic years due to leave of absence or holdovers. Students who take leaves of absence or repeat years in school often report a lack of both strong peer relationships and places on campus where they feel comfortable.17 Therefore, providing these students with a community and meeting space may reduce their resistance to attending college and can prevent them from withdrawal and isolation.

At a systematic level, increasing the flexibility of the university system to avoid unnecessary holdovers could be helpful. For instance, circumstances where a student is forced to stay in the same academic year only to obtain a single credit that they have failed could be considered unnecessary and impractical. Reducing the stringency of credit transfer rules for students who change majors or transfer schools could also help such students academically as well as psychiatrically. Through these approaches, the number of unnecessary holdover students could be minimized, leading to improved mental health among the students.

Third, we need to give special attention to the high risk of suicide among medical students. The high suicide rate among medical students has been discussed internationally.20 Along with the influence of mental stress associated with medical school and the medical profession, the knowledge of and accessibility to lethal substances are suggested to affect the high rates of suicide among physicians and medical students.32 Studies21,33 have suggested that having individual guidance and a role model from early on in the program could alleviate psychological stress for students. Taking these reports and our results into consideration, we suggest that medical school faculty, administrative staff, and students themselves be educated about the high risk for mental distress and suicide among medical students and that this risk can be ameliorated by appropriate intervention, including psychiatric treatment.

It also may be beneficial for medical schools to provide individual or small-group guidance and mentoring sessions to support the individual development of future physicians. The experience of receiving this personal guidance can also provide additional opportunities for students to learn how to seek help as well as how to be an effective provider.

Limitations

Although the method of collecting surveys through the university mental health centers garnered us access to a large sample, it prevented us from conducting individual assessments of students. The sample consisted of students attending Japanese national universities; hence, our study provides valuable data on a unique population with few previous reports but may limit the generalization of results to the international population. Additionally, the relatively small effect sizes also limit the generalization of our conclusions.

CONCLUSION

Despite these limitations, this study confirmed that suicide is a major concern among college students who received little intervention prior to the event and identified men, medical students, and final-year and holdover students as high-risk groups for suicide, suggesting that preventive efforts should target these groups.

Submitted: March 11, 2016; accepted August 5, 2016.

Author contributions: Dr C. Uchida confirms that she had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Both Dr C. Uchida and Dr M. Uchida contributed substantially to the study concept and design and to the acquisition, analysis, and interpretation of the work. Both authors were responsible for drafting and revising the work. Both authors gave their final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Potential conflicts of interest: The authors have nothing to disclose.

Funding/support: Dr C. Uchida reports grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan during the conduct of the study under grant numbers 17500459 and 23500794. She also has received a university management grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Role of the sponsor: All of the listed grants partially supported the execution of the surveys, as well as storing, managing, and analyzing of the data.

Acknowledgments: The authors thank the Japanese Association for College Mental Health, the Mental Health Study Group of the Japanese National University Council of Health Administration for their collaboration and the University Health Centers that participated in this study for their collaboration. The supporters had no role in the design, analysis, interpretation, or publication of this study. The authors express their deepest appreciation to Junko Nakajima, MD, Ishizaki Hospital, for her work on initiating this study; Shuntaro Watanabe, PhD, Osaka University of Comprehensive Children Education, and Hiroshi Sato, PhD, Kansai University, for their help with the analysis and interpretation of the data; Masafumi Nomura, MD, Hospital Bando for his help with the organization of data collection; and Hilary Woodworth, BA, Massachusetts General Hospital; and Alexandra Saunders, BA, Massachusetts General Hospital for assisting with the preparation of the manuscript. Drs Nakajima, Watanabe, and Sato and Mss Woodworth and Saunders have no conflicts of interest to declare.

REFERENCES

1. Golberstein E, Eisenberg D, Gollust SE. Perceived stigma and help-seeking behavior: longitudinal evidence from the healthy minds study. Psychiatr Serv. 2009;60(9):1254-1256. PubMed doi:10.1176/ps.2009.60.9.1254

2. Cranford JA, Eisenberg D, Serras AM. Substance use behaviors, mental health problems, and use of mental health services in a probability sample of college students. Addict Behav. 2009;34(2):134-145. PubMed doi:10.1016/j.addbeh.2008.09.004

3. Zivin K, Eisenberg D, Gollust SE, et al. Persistence of mental health problems and needs in a college student population. J Affect Disord. 2009;117(3):180-185. PubMed doi:10.1016/j.jad.2009.01.001

4. Drum DJ, Brownson C, Denmark AB, et al. New data on the nature of suicidal crises in college students: shifting the paradigm. Prof Psychol Res Pr. 2009;40(3):213-222. doi:10.1037/a0014465

5. Khan A, Hamdan AR, Ahmad R, et al. Problem-solving coping and social support as mediators of academic stress and suicidal ideation among Malaysian and Indian adolescents. Community Ment Health J. 2016;52(2):245-250. PubMed doi:10.1007/s10597-015-9937-6

6. Wilburn VR, Smith DE. Stress, self-esteem, and suicidal ideation in late adolescents. Adolescence. 2005;40(157):33-45. PubMed

7. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62(6):593-602. PubMed doi:10.1001/archpsyc.62.6.593

8. Kessler RC, Chiu WT, Demler O, et al. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62(6):617-627. PubMed doi:10.1001/archpsyc.62.6.617

9. 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. PubMed doi:10.1002/da.20321

10. Pfeffer CR. Suicide in mood disordered children and adolescents. Child Adolesc Psychiatr Clin N Am. 2002;11(3):639-647, x [x.]. PubMed doi:10.1016/S1056-4993(02)00012-3

11. The Ministry of Health LaWoJ. White paper on suicide prevention. 2016. http://www.mhlw.go.jp/wp/hakusyo/jisatsu/16/dl/1-03.pdf.

12. World Health Organization. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, Switzerland: World Health Organization; 1992.

13. Uchida C. Survey of Taking-off and Dropping-out Students—28th Report. http://hotai1.htc.nagoya-u.ac.jp/~kondo/hakusho/hakusho2005.pdf.

14. Uchida C. Survey of taking-off, dropping- out and repeating academic-years students, 34th report. The 35th Meeting Report of the Japanese Association of College Mental Health.2013:35-51. http://ci.nii.ac.jp/naid/40020142077.

15. National Police Agency. Traffic Accident Statistics. 2013. http://www8.cao.go.jp/koutu/chou-ken/h24/pdf/2-1-1.pdf.

16. Silverman MM, Meyer PM, Sloane F, et al. The Big Ten Student Suicide Study: a 10-year study of suicides on midwestern university campuses. Suicide Life Threat Behav. 1997;27(3):285-303. PubMed

17. Uchida C. Apathetic and withdrawing students in Japanese universities—with regard to Hikikomori and student apathy. J Med Dent Sci. 2010;57(1):95-108. PubMed

18. Eisenberg D, Hunt J, Speer N, et al. Mental health service utilization among college students in the United States. J Nerv Ment Dis. 2011;199(5):301-308. PubMed doi:10.1097/NMD.0b013e3182175123

19. Yorgason JB, Linville D, Zitzman B. Mental health among college students: do those who need services know about and use them? J Am Coll Health. 2008;57(2):173-181. PubMed doi:10.3200/JACH.57.2.173-182

20. Goldman ML, Shah RN, Bernstein CA. Depression and suicide among physician trainees: recommendations for a national response. JAMA Psychiatry. 2015;72(5):411-412. PubMed doi:10.1001/jamapsychiatry.2014.3050

21. Pepitone-Arreola-Rockwell F, Rockwell D, Core N. Fifty-two medical student suicides. Am J Psychiatry. 1981;138(2):198-201. PubMed doi:10.1176/ajp.138.2.198

22. Millan LR, Rossi E, De Marco OL. Suicide among medical students [in Portuguese]. Rev Hosp Clin Fac Med Sao Paulo. 1990;45(3):145-149. PubMed

23. Walters PA. Student Apathy. 2nd ed. New York, NY: Appleton-Century-Crofts; 1971.

24. Centers for Disease Control and Prevention. Youth Risk Behavior Surveillance. Morbidity and Mortality Weekly Report. 1995;46(SS-6).

25. Collins IP, Paykel ES. Suicide amongst Cambridge University students 1970-1996. Soc Psychiatry Psychiatr Epidemiol. 2000;35(3):128-132. PubMed doi:10.1007/s001270050195

26. Schwartz AJ. Four eras of study of college student suicide in the United States: 1920-2004. J Am Coll Health. 2006;54(6):353-366. PubMed doi:10.3200/JACH.54.6.353-366

27. Schwartz AJ. College student suicide in the United States: 1990-1991 through 2003-2004. J Am Coll Health. 2006;54(6):341-352. PubMed doi:10.3200/JACH.54.6.341-352

28. Uchida C. Suicide among Japanese University students-from the results of a 21 year survey, a clue to prevent suicide among university students [in Japanese]. Psychiatria et Neurologia Japonica. 2010;112(6):543-560.

29. Furr SR, Westefeld JS, McConnell GN, et al. Suicide and depression among college students: a decade later. Prof Psychol Res Pr. 2001;32(1):97-100. doi:10.1037/0735-7028.32.1.97

30. King KA, Vidourek RA, Strader JL. University students’ perceived self-efficacy in identifying suicidal warning signs and helping suicidal friends find campus intervention resources. Suicide Life Threat Behav. 2008;38(5):608-617. PubMed doi:10.1521/suli.2008.38.5.608

31. Joffe P. An empirically supported program to prevent suicide in a college student population. Suicide Life Threat Behav. 2008;38(1):87-103. PubMed doi:10.1521/suli.2008.38.1.87

32. Epstein LC, Thomas CB, Shaffer JW, et al. Clinical prediction of physician suicide based on medical student data. J Nerv Ment Dis. 1973;156(1):19-29. PubMed doi:10.1097/00005053-197301000-00002

33. Gerber LA. The search for clinical role models as a way of coping with clerkship stress. J Med Educ. 1979;54(8):659-661. PubMed