Temporal Fluctuations of Suicide Mortality and Their Attributed Motives in Governmental Databases of Student Suicides in Japan from 2009 to 2025

Abstract

National crude suicide mortality rates (CMRs) in Japan trended downward throughout the 2010s, yet the CMR for the youth population moved in the opposite direction, rising from the mid-2010s onwards. To clarify the risk groups and underlying mechanisms of increasing student suicides, the temporal fluctuations of CMRs of students in middle school, high school, special vocational school, and university were analyzed by joinpoint regression analysis, and identification of high CMR groups were detected by a linear mixed-effect model using government suicide databases. CMRs of female students in all school categories increased from the mid-2010s. CMRs of male students in middle and high schools increased, whereas those in university and special vocational school moved from a decreasing to a stable level. Among the average of CMRs from 2022 to 2025, the highest CMR was observed in part-time high school students, which was approximately fivefold higher than the average of all students overall. The major leading suicide motives for all student subgroups were concerned with underachievement and career-path-associated problems within school-related problems. Family-related problems, such as conflict with parents and parental reprimands, were leading suicide motives for middle school students, but these impacts attenuated with age. Instead, the impacts of psychiatric disorders, including depression and other psychiatric disorders, enhanced with age. Student suicides attributed in the government suicide database to these leading suicide motives showed an upward inflexion, moving from a decreasing or stable pattern to a rising one, beginning in the mid-2010s. Taken together, these patterns indicate that mental health concerns among students have re-emerged as a substantial motive behind student suicide from the mid-2010s.

1. Introduction

Suicide is preventable death that imposes substantial social and public health burdens [1,2,3,4]. In light of the fact that suicide is preventable and exerts considerable societal costs, the World Health Organization (WHO), under its Comprehensive Mental Health Action Plan 2013–2030, set a global objective of reducing suicide mortality by one-third by 2030. Although suicide prevention efforts in many high-income nations have focused predominantly on the working-age and older adult populations [5,6,7,8,9,10,11,12], Japan has taken a distinctive approach since 2007, where the General Principles of Suicide Prevention Policy (GPSPP) has advanced a comprehensive, all-age approach to suicide prevention, a stance that distinguishes Japan from many other countries [13,14].

Although Japan had achieved a reduction in suicide mortality exceeding 30% over the course of the 2010s, the country experienced an atypical upturn in 2020 amid the COVID-19 pandemic, an increase that was internationally unusual. A clear return to the prior downward trajectory was not evident until 2024; in 2025, however, the annual count of suicides dropped sharply from the preceding year and fell below 20,000 for the first time since the inception of national suicide statistics [15]. Although the government and majority of academics have largely attributed the stagnation in suicide reduction during the early 2020s to pandemic-related psychosocial and/or socioeconomic disruptions, several studies have reported the possibility that the deceleration in the downward trend had already begun in the late 2010s [13,16,17,18]. Furthermore, the demographic composition of high-risk groups for suicide has also been reported to have shifted markedly in the 2020s compared with the early 2010s [13,16,17,19,20].

Worldwide, suicide in young populations has become a pressing contemporary public health priority [1,2,21,22,23,24]. Trends in Japan align with this international pattern: suicide mortality in those younger than 30 has been rising in recent years [13,16,17,19,25,26,27,28,29]. However, the increase in youth suicides (<30 years old) in Japan is particularly alarming, because reported statistical results have contradicted two established epidemiological consensuses in suicide statistics [30,31]: age-dependence (suicides increase after adolescence) and sex-dependence (male suicides are more frequent than female suicides) [19,20,28,32].

Among adolescents in Japan, crude annual suicide mortality per 100,000 population (CMR) is now higher in females (8.0) than in males (6.6) [20], deviating from the typical sex-dependent pattern of suicides [21,22,23,33,34,35]. Moreover, among females, the highest CMR is observed in those aged 20–29 years (14.7) [17,19,27,36], contradicting the usual age-dependent pattern in which suicide risk increases with age [21,22,23,33,34,35]. Internationally, countries where young women exhibit the highest CMR among females are exceedingly rare [28,29]. Additionally, suicide is not only the leading cause of death among individuals under 30 in Japan, but also accounts for more than 50% of all deaths in this age group, a proportion which far exceeds that of other OECD countries (20–30%) [28,29]. However, the high-risk suicide groups among young people and their motives in Japan remains to be clarified. Therefore, identifying suicide risk groups and their underlying factors among youth in Japan can provide essential findings for improving suicide prevention programmes.

Conducting comprehensive investigations into the suicide risk among young people is exceedingly hard, and obtaining actual valid or representative findings is arguably impossible, since adolescence constitutes a heterogeneous population composed of diverse psychosocial and socioeconomic backgrounds [9,27,28,29,37]. Adolescence, as a developmental stage, is characterized by unique psychological structures and drastically transforming social contexts: adolescents face various specific problems, such as demands for independence, educational transitions, and early career decision-making, that collectively shape adolescent-specific vulnerability to psychiatric disorders [38,39]. In other words, adolescents exhibit a unique profile of vulnerability to psychiatric disorders, one that is obviously differentiated from the risks seen in people of working age [13,25,27,28,29]. Based on these backgrounds, this study determined the suicide mortality and the impacts of their motives among students in middle, high, and special vocational schools and university, where social standing is clearly defined. The findings from this study have the potential to generate evidence aligned with the principles of evidence-based policy making (EBPM), which may be particularly valuable for predicting the effectiveness of, and informing future revisions to, school-based suicide prevention programmes currently promoted by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), and the Children and Families Agency (CFA) [40,41,42,43,44].

2. Materials and Methods

2.1. Ethics

Because the present analysis relied exclusively on de-identified, openly accessible governmental records, namely the Suicide Statistics released by the National Police Agency (SSNPA) and the Basic Data on Suicide in the Region (BDSR) by the Ministry of Health, Labour, and Welfare (MHLW), the requirement for a formal ethical review was waived by the Medical Ethics Review Committee of Mie University. Reporting of this observational study followed the Strengthening Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

2.2. Databases

For the period from 2009 to 2025, monthly and annual counts of student suicides in middle school, high school, special vocational school, and university were retrieved from the SSNPA and BDSR [15,45]. Corresponding student enrollment counts for middle school, regular-time high school, part-time high school, special support high school, special vocational school, and university (the latter further stratified by under and over 20 years of age), each disaggregated by sex, were drawn from the School Basic Survey conducted by MEXT [46].

The statutory framework underpinning these statistics operates as follows. The Medical Practitioners Act in Japan requires physicians to notify the National Police Agency (NPA) of any abnormal death within 24 h from discovery. Upon receiving such reports, the NPA is required to conduct a physiological examination to determine the cause of death. Subsequently, police investigators document the personal characteristics of the decedent, including age, sex, and social status, as well as background factors comprising suicide motives, location, time, and residential address. The SSNPA aggregates individual case records collected by local police stations within their respective jurisdictions. Accordingly, the SSNPA and BDSR compile judicial statistics from jurisdiction-level survey data, but not administrative statistics.

Within the SSNPA/BDSR, suicide motives are organized into seven principal domains: health-related, family-related, economic-related, romance-related, employment-related, school-related, and other, each with their own subcategories. Since motives cannot be elicited from the decedent themselves, local police rely on objective sources of evidence, including suicide notes and official documentation such as medical certificates and clinical records, in order to minimize subjective bias as far as possible. Where the decedent had a documented history of treatment for any psychiatric disorder, the attending physician is formally consulted, and the impacts of the psychiatric disorder and/or mental health issue on the suicide is assessed as a background factor.

2.3. Statistical Analyses

Annual and monthly crude suicide mortality rates per 100,000 population (CMR) of students were calculated by dividing the annual/monthly suicide numbers by the population of corresponding groups (student population in each student group) in the same period. Especially, because the CMR for each month reflects the number of days within that month, ranging from 28 to 31 days, a discrepancy of up to 10% exists between months. As this difference is non-negligible in the context of JPRA, all monthly CMR values were standardized to an annual 365-day CMR before conducting the analyses.

Joinpoint regression analysis (JPRA) has been evaluated to be an appropriate statistical method which can detect unknown joinpoints where trends change via fitting the simplest joinpoint model that the trend data allows [25,27,47,48,49]. Based on these statistical backgrounds, to analyze the temporal fluctuations in the CMRs of students in Japan from January 2009 to December 2025, this study adopted the Joinpoint Regression Program ver. 6.0.1. (The National Cancer Institute, Bethesda, MD, USA). A detailed description of “the Methods Used in the Joinpoint Regression Software” is provided in the user manual published by the National Cancer Institute (NCI). The CMRs were derived from the national suicide number of each student group; where standard errors were unavailable, the Constant Variance (homoscedasticity) option was applied. When the Durbin–Watson statistic indicated no significant first-order autocorrelation (DW > 1.5), the uncorrelated error structure was retained as the default setting, whereas when DW < 1.5, the ‘First Order Autocorrelated Errors Estimated from the Data’ option was specified to fit an AR(1) error structure, which accounts for residual serial dependence, including autocorrelation arising from seasonal fluctuation patterns (thereby reducing the risk of spurious joinpoint detection attributable to periodic oscillations in the CMR series). Sensitivity analyses were conducted under both the AR(1) and the uncorrelated-error specifications; results were consistent across both [47,48,49].

Survey methods regarding suicidal motives in SSNPA/BDSR were revised from January 2022 [50,51]. In 2022, the number of subcategories of suicide motives in the revised SSNPA/BDSR increased from 52 to 75, and the maximum number of permitted suicide motives in each case also increased from 3 to 4. Accordingly, fluctuations in CMRs disaggregated by suicide motive were analyzed using jump-model JPRA, which can detect fluctuations when coding changes cause discontinuities (jump effects), since the SSNPA/BDSR did not provide the formal bridged data with dual coding (between before and after survey methods) [47,52]. In jump-model JPRA, which implements comparability ratio (CR) estimation to quantify the magnitude of the impact of coding revisions, the jump parameter and trend slopes are estimated simultaneously, allowing detection of genuine inflexion points (joinpoints) independently of the coding-induced discontinuity [47,52]. When the joinpoint does not coincide with 2022, the detected joinpoint period can be evaluated to be unaffected by artefacts from revision of the survey method. Sensitivity analyses were conducted as follows. For each suicide motive series, this study determined whether detected joinpoints coincided with 2022. Joinpoints at 2022 would suggest the detected trend change is attributable to the coding revision artefact, whereas joinpoints detected at years other than 2022 indicate genuine epidemiological trend changes unconfounded by the revision. Analyses were repeated, restricting the observation period to 2009–2021 (pre-revision only) to confirm that the direction and approximate timing of trend changes detected in the full-period analyses were consistent with those observed in the pre-revision period alone.

Additionally, following the revision of the survey method of the SSNPA/BDSR in 2022, previous overall suicide data of high school students were disaggregated into regular-time, part-time, and special support high schools (Supplementary Table S1). Therefore, although the observation period was short, differences in the average CMRs among these high school student subgroups between 2022 and 2025 were analyzed by t-test or linear mixed-effect model for repeated measures (LMM) alongside Scheffe’s post hoc test using IBM SPSS Statistics 30.0 for Windows (IBM, Armonk, NY, USA) [16].

3. Results

3.1. Temporal Fluctuation in CMRs of Total Students from 2009 to 2025 and Average CMRs of Students from 2022 to 2025

The SSNPA/BDSR have provided annual and monthly suicide numbers of middle school, high school, university, and special vocational school students since 2009. Therefore, trends for both male and female monthly CMRs of total students in middle school, high school, university, and special vocational school from 2009 to 2025 were determined using JPAR. Trends of CMRs of both male and female students moved from a decreasing to an increasing trend from 2016 (Figure 1 and Supplementary Table S2). Importantly, increasing trends of female CMRs exceeded that of males, showing that the CMR of females became almost equivalent to males in 2025 (Figure 1 and Supplementary Table S1).

The SSNPA/BDSR have published the annual suicide numbers among middle school, high school, special vocational school, and university students from 2009 to 2025. Further, the SSNPA/BDSR have provided annual suicide numbers for high school students, disaggregated by school type (regular-time, part-time, and special support high schools), since 2022. Therefore, in this study, average CMRs among middle school, special vocational school, university, and high school students, including regular-time, part-time, and special support high school students, were analyzed. Both the CMRs of male and female students among middle school, high school, and university increased age-dependently. The CMRs of part-time high school (males < females) students were the highest among the overall student population. The CMRs of male and female part-time high school students were 3.38- (95% CI: 2.98–3.86) and 6.6-fold (95% CI: 5.88–7.40) higher compared to those of regular-time high school students, respectively (Figure 1). CMRs of university male and female students under 20 years of age were smaller than those of high school students (Figure 1).

3.2. Temporal Fluctuation of CMRs of Students

Monthly CMRs of male students in middle school consistently increased from 2009 to 2025, whereas that of females changed from a stable to an increasing trend in 2016. The CMRs of both male and female students in high school moved from stable to increasing trends in 2016 (Figure 2 and Supplementary Table S2). The CMR of female students in special vocational school also moved from a stable to an increasing pattern in 2016, whereas that of males decreased from 2009 to 2024, possibly moving to a non-significant increase after 2024 (Figure 2 and Supplementary Table S2).

The CMR of total male university students (younger plus older than 20 years) moved from a decreasing to an unchanging trend in 2016, whereas that of females changed from a decreasing to an increasing trend (Figure 2 and Supplementary Table S2). Fluctuation patterns in the CMRs of both male and female university students over 20 years old showed the same pattern as total university students; however, the CMRs of university students for both males and females under 20 years old moved from a stable to an increasing trend in 2016 (Figure 2 and Supplementary Table S2).

3.3. Fluctuations in Student Suicides Disaggregated by Suicide Motives

The SSNPA provids annual suicide numbers disaggregated by suicidal motives. However, the SSNPA revised the methods for the survey of suicidal motives in January 2022 [50,51]. Details in the revised SSNPA include the number of subcategories for suicide motives in the survey master sheet, which increased from 52 to 75, and the number of listable suicide motives per case, which increased from 3 to 4. Due to this revision, direct comparison of the impacts of suicide motives across before and after revision periods is no longer feasible [50,51].

Major leading suicide motives among junior high school, high school, university, and special vocational school students from 2022 to 2025 are represented in Figure 3. Temporal fluctuations in CMRs disaggregated by suicide motives were determined using jump-model JPRA, which can detect fluctuations induced by coding changes caused by discontinuities (jump effects) [47,52].

3.3.1. Fluctuation in the CMRs of Middle School Students Disaggregated by Major Leading Suicide Motives

Major leading suicide motives for middle school students included health-related motives (depression and other psychiatric disorders [mainly anxiety disorders, excluding depression and schizophrenia]), school-related motives (underachievement, career-path-associated problems, entrance-examination-associated problems, conflict with peers), and family-related motives (conflict with parents and reprimand from parents) (Figure 3 and Figure 4).

Among middle school students, the CMRs of females attributed in the SSNPA to major leading suicide motives increased from approximately the mid-2010s, whereas the CMRs of males attributed to major leading suicide motives, except for depression, increased (Figure 4 and Supplementary Table S3). Importantly, trends of the CMRs of females attributed to depression, other psychiatric disorders, and conflict with peers were predominant compared to males, resulting in the fact that, during the 2020s, the CMRs of female students attributed to these three motives were markedly higher compared to males (Figure 4 and Supplementary Table S3). Notably, these increasing CMRs had already begun to increase before the outbreak of the COVID-19 pandemic.

3.3.2. Fluctuation in the CMRs of High School Students Disaggregated by Major Leading Suicide Motives

Major leading suicide motives for high school students were identical to middle school students, including depression, other psychiatric disorders, underachievement, career-path-associated problems, entrance-examination-associated problems, conflict with peers, conflict with parents, and reprimands from parents.

Among high school students, the CMRs of females attributed in the SSNPA to major leading suicide motives, except for reprimands from parents, increased, whereas the CMRs of males attributed to major leading suicide motives increased from the mid-2010s (Figure 5 and Supplementary Table S4). Trends of the CMRs of females attributed to underachievement and conflict with peers were also more predominant compared to males, resulting in their totals being almost equivalent between males and females in the early 2020s (Figure 5). None of the disaggregated CMRs of males were obviously higher than females. Compared to middle school students, the impact of family-related motives (conflict with parents and reprimands from parents) on the CMRs of high school students decreased, but conversely the impact of health-related motives (depression and other psychiatric disorders) and school-related motives (underachievement and conflict with peers) became dominant (Figure 3, Figure 4 and Figure 5).

3.3.3. Fluctuation in the CMRs of University Students Disaggregated by Major Leading Suicide Motives

Major leading suicide motives for university students comprised health-related motives (depression and other psychiatric disorders), school-related motives (underachievement, career-path-associated problems and conflict with peers), heartbreak, unsuccessful job search, and social isolation. However, family-related motives and entrance-examination-associated problems, which were major leading suicide motives for middle and high school students, were no longer the major leading motives for university students; instead, heartbreak, an unsuccessful job search, and social isolation have become the predominant motives (Figure 6). Nonetheless, the impacts of depression, other psychiatric disorders, underachievement, and career-path-associated problems were greater than the other four motives (Figure 6).

Among university students, the CMRs of females attributed in the SSNPA to underachievement increased from the mid-2010s, whereas that of males decreased during the 2010s but moved to an increasing pattern from the late 2010s (Figure 6 and Supplementary Table S5). CMRs attributed to depression for both males and females decreased until the mid-2010s, but changed to an increasing trend from the mid-2010s (Figure 6 and Supplementary Table S5). CMRs attributed to other psychiatric disorders moved from a stable to an increasing pattern from the late 2010s (Figure 6 and Supplementary Table S5). The CMRs of females attributed to underachievement and career-path-associated problems moved to an increasing trend from the mid-2010s. Contrarily, the CRM of males attributed to career-path-associated problems consistently decreased, whereas that attributed to underachievement changed from a stable to an increasing trend in the late 2010s (Figure 6 and Supplementary Table S5). The CMRs of both males and females attributed to conflict with peers and social isolation changed to an increasing pattern from the late 2010s (Figure 6 and Supplementary Table S5). The CMR of males attributed to heartbreak increased from the late 2010s, and the CMR of females attributed to an unsuccessful job search increased from the mid-2010s (Figure 6 and Supplementary Table S5).

3.3.4. Fluctuations in the CMR of Special Vocational School Students Disaggregated by Major Leading Suicide Motives

Major leading suicide motives for special vocational school students were identical to university students, consisting of depression, other psychiatric disorders, underachievement, career-path-associated problems, conflict with peers, heartbreak, an unsuccessful job search, and social isolation.

Among female special vocational school students, all CMRs attributed to major leading motives increased during the 2020s, whereas those of males attributed to major leading motives, except for career-path-associated problems and an unsuccessful search job, also increased during the 2020s. The CMRs of males attributed to career-path-associated problems and an unsuccessful search job decreased from the mid-2010s (Figure 7 and Supplementary Table S6). Importantly, the CMR of females attributed to other psychiatric disorders changed to an increasing trend in 2016, whereas the CMRs of females attributed to other major leading motives moved to an increasing pattern around approximately 2018 (Figure 7 and Supplementary Table S6). The CMRs of males attributed to depression and underachievement began to increase from around 2020, and those attributed to social isolation also increased from 2018 (Figure 7 and Supplementary Table S6). As a result, one can observe that increasing CMRs among special vocational school students seem to increase at a slower rate than among university students.

4. Discussion

Our analysis brought to light a number of features of student suicide in present-day Japan that depart markedly from conclusions previously drawn from general suicide statistics in the governmental database. First, while the national CMR in Japan declined throughout the 2010s, the CMR specific to students moved in the opposite direction from the mid-2010s. Second, beginning in the mid-2010s, the upward shift in the CMR was steeper among female than among male students; by 2022, 2024, and 2025, respectively, the CMRs for female students in middle school, high school, and in the under-20 university subgroup had overtaken the corresponding male rates. Third, the highest CMR among student categories was observed in part-time high school students, exceeding the average for all students and for all high school students by roughly three- to six-fold. Fourth, the dominant motives identified for student suicides overall lay in the school-related domain; specifically, academic underperformance and career-related concerns. Family-centred difficulties (conflict with parents and parental reprimand), which had figured prominently among middle school cases, attenuated with age, whereas the relative contribution of psychiatric issues, depression, and other psychiatric disorders increased with age. Of particular note, mental health concerns now stand as priority targets for suicide prevention among students, since impacts of mental health disorders have increased drastically as factors for suicide among female students in Japan.

4.1. Features of Student CMRs

A long-standing epidemiological observation is that, considered across all age groups, suicide mortality among males exceeds that among females. In East Asian populations, including Japan, this male-to-female ratio in suicide mortality has been reported as lower than the corresponding ratio in Western settings. The male-to-female ratio is known to widen with increasing age, and adolescents are no exception to this pattern [28,29,53]. Gender ratio (males/females) increases with age, and adolescents are no exception [21,28,29,53]. The widening of this sex difference with age reflects a steep rise in male suicide rates during the late teens against a more modest concurrent rise among females [21,23]. Against this backdrop, the trajectory observed among students in Japan represents a marked deviation from the global sex-dependent features of suicides [21]. Indeed, the gender ratios of student CMRs in middle school, high school, and university (<20 years) became equal in 2022, 2024, and 2025, respectively; subsequently, the CMRs of these females groups were higher than male equivalents. Put differently, although suicide rates have been climbing in both male and female student populations in present-day Japan, the accelerating rise observed in female students since the mid-2010s constitutes a particularly pressing priority for adolescent-focused suicide prevention efforts. Collectively, these results suggest that the rise in student suicides, more marked among female students than males, appeared first in middle school cohorts and then spread outward to high school, university, and special vocational school students. The continuing escalation observed among female students at every educational stage indicates that this trajectory is unlikely to reverse without intervention.

Notably, the CMR observed in part-time high school students stood out as markedly elevated relative to other student categories. Recent national statistics have documented a substantial expansion in enrolment at part-time high schools [46]. Over 80% of middle school students who had been unable to attend regular middle school owing to school refusal or their own disabilities have subsequently enrolled in part-time high schools [46,54,55]. Part-time high schools confer their graduation qualifications principally through correspondence-style coursework and formal assessments, with attendance obligations that are considerably less stringent than those imposed in regular-time high schools. This more flexible structure positions part-time high school as a viable route for obtaining a high school diploma, and, by extension, a pathway to university admission, for those who, owing to school refusal or their own disabilities, find the conventional educational route inaccessible [56]. Given the demographic profile of those increasingly choosing this route, the possibility that students with poor school adaptation or with elevated suicide risk are disproportionately concentrated within part-time high schools warrants serious consideration.

Most part-time high schools are administratively attached to full-time high schools, and the same mental health professionals, school counsellors, and school caseworkers are typically expected to cover both school formats. Dedicated specialists for part-time high schools therefore remain in short supply [56,57,58,59,60]. Many part-time high schools, moreover, mandate only around 10–20 days of in-person attendance per year, leaving teaching staff and school mental health professionals with little opportunity to gain a substantive understanding of each student’s wellbeing [56,60]. Given that part-time high schools combine a high concentration of at-risk students with markedly limited in-person attendance [55,56], the kind of mental-health-professional-led support system used effectively in middle schools and in regular-time high schools cannot be expected to function comparably well in this setting. Accordingly, the initial step should fall to teachers themselves, who must take responsibility for investigating and gaining a substantive grasp of each student’s situation; once high-risk students have been identified, collaboration with educational psychological staff must result in the development of appropriate support programmes for each individual high-risk student.

4.2. Features of Suicide Motive for Student Suicides

Characterizing the temporal, age-related, and sex-related dimensions of prevalent suicide motives among students serves a dual purpose: it enables suicide prevention efforts to be deployed more efficiently, and it supports the planning of interventions tailored to the psychosocial transitions that students undergo. In our middle school sample, family-centered difficulties, particularly regarding conflict with parents and parental reprimand, emerged as the dominant suicide motives, yet the salience of these family-related factors progressively diminished in high school students and was no longer prominent among university and special vocational school students. By the high school stage, school-centred difficulties such as academic underperformance, career path concerns, problems linked to entrance examinations, and peer conflict had displaced family-related issues as the leading concerns. Such an age-related reorganization is consistent with the well-described life cycle pattern in which psychosocial dependence on the parental dyad gradually attenuates in line with individual psychological maturation [61,62,63]. Consistent with this developmental view, the contribution of peer conflict was more prominent in high school students than in middle school students. By contrast, in university and special vocational school students, the relative weight of entrance examination concerns and of peer conflict declined, while underperformance and career-related concerns became more salient. For middle and high school students, gaining admission to a higher educational institution, i.e., passing an entrance examination, constitutes a pivotal event tied directly to their future career trajectory. In university and special vocational school students, however, whose primary objective is to acquire the skills, knowledge, and qualifications required for employment, the entrance examination is no longer a major event; rather, the career path itself has become a major event for them. This can also be fully interpreted when considering the life cycle [37,64]. For students who died by suicide across middle school, high school, special vocational school, and university combined, two motives—academic underperformance and career-related concerns—were consistently among the leading recorded factors. The school system bears responsibility for equipping young people with the knowledge and competencies they will need to construct economically and emotionally fulfilling adult lives, and for verifying that each student is in fact acquiring those competencies. It is widely understood, however, that the very process of evaluating individual academic outcomes can itself constitute a substantial stressor for students. Yet underperformance or career-related concerns alone are unlikely to be sufficient to precipitate suicide; instead, the more plausible scenario is that these school-related difficulties arise out of a confluence of factors, including inadequate parenting, psychiatric morbidity, problematic interpersonal relationships within the community, and others, that interact with one another [65,66,67].

Beyond these age-related shifts in dominant motives, our analysis revealed several distinctive features of mental health-related suicide motives that may help to account for the recent rise in student suicide in Japan. First, the upward shift in CMR has been steeper in female than in male students. Second, this pattern, in which the female CMR has come to exceed the male CMR, first emerged in middle school students and then propagated, age-dependently, to high school and university student populations. Third, within the middle school group, the relative contribution of other psychiatric disorders (predominantly anxiety) outweighed that of depression; with advancing age, the relative weight of other psychiatric disorders declined while that of depression rose, such that in university and special vocational school students the two became roughly comparable, with depression exerting a marginally larger contribution. Collectively, these observations are highly suggestive of a role of internalizing disorders in the rising CMR among Japanese students, and especially among female students. In short, the evidence assembled here is sufficient to motivate an attempt, informed by the existing literature, to interpret and intervene in the mechanisms underlying the increase in Japanese student suicide through the lens of internalizing disorders.

Although internalizing disorders have not been recognized as a formal diagnostic category within ICD-10, ICD-11, or DSM-5 [68], the term is nonetheless well entrenched as a higher-order construct in the fields of child and adolescent psychiatry and developmental psychology [69]. The construct encompasses depressive symptoms, anxiety, obsessive–compulsive features, hypochondriacal presentations, and suicidal behaviour, and has been linked to an overcontrolled behavioural pattern [70]. The onset age is around 15 years of age [71,72,73], and is consistently prevalent in females than in males [69]. It is well-known that individuals with internalizing symptoms have a high suicide risk, and being female further increases suicide risk [74,75]. National data from the MHLW Patient Survey have documented a recent rise in the prevalence of depressive and anxiety disorders in Japanese adolescents [13,25]. Recent work analyzing social media content has further suggested that internalizing processes may manifest more prominently in young women than in young men. In particular, locutions such as ‘life is painful’ or ‘life is unbearable’ have been treated as markers of a diminished ability to form autonomous social ties, and tend to co-occur with loneliness, anxiety, low self-worth, hopelessness, and anger [76,77,78,79,80]. Posts containing such expressions on Japanese social media platforms have shown a sustained increase among females; the frequency among young women is reported to exceed that among young men by more than fivefold [77]. Taken together, these lines of prior work offer convergent symptom-based and epidemiological support for a link between the rising burden of internalizing disorders and the rise in suicide among Japanese adolescent females [13,25].

The development of internalizing disorders is thought to reflect a gene–environment interaction in which adverse childhood experiences constitute a major environmental contributor [81]. More recent work has highlighted parenting factors, including negative parental mood, reluctance to express emotion, only-child status, an unfulfilling maternal relationship, and over-restrictive paternal behaviour, as playing important roles in the pathomechanisms underlying internalizing disorders [78,79,80,82,83,84,85,86,87,88,89,90]. The age-dependent reorganization of leading suicide motives observed across student groups can be situated within a developmental psychopathology framework [38,39,91]. In middle school students, whose age precedes the typical onset of internalizing disorders, the heightened CMR linked to conflict with parents and parental reprimands reflects the central role that the parent–child relationship plays as the principal source of psychosocial strain at this developmental stage [90,92,93]. As students move up to higher educational levels, the relative contribution of family-related motives diminishes; in its place, peer conflict, that is, the demands of interpersonal communication with peers, together with concerns surrounding education and career trajectory, becomes an increasingly salient factor of suicide risk [64]. Running in parallel, the age-dependent rise in CMR attributed to psychiatric disorders, most notably depression and anxiety disorders, is congruent with the developmental course of internalizing disorders, which are widely viewed as the core expressions of adolescent internalizing psychopathology and which characteristically rise in prevalence throughout mid-to-late adolescence [71,72,73,94]. These findings may be consistent with the hypothesis that psychological internalization contributes to recent increases in student suicides, but the present registry-based analysis cannot directly determine the mechanism.

4.3. Proposal for Measures to PreventRising Student Suicides

Under the Fourth GPSPP, prefectural boards of education have begun rolling out suicide prevention curricula based on locally developed instructional materials, with provision extending to all students. The principal aims of these curricula are to strengthen resilience (psychological recovery), to enable the identification of peers at elevated risk of suicide, and to teach students how to seek support when they are in a high-risk situation [95]. The validity of these objectives is supported by the results of this study.

The deployment of school counsellors in Japanese schools has been progressively expanded since 2008, such that virtually all schools are now staffed with at least one counsellor [57,58]. Depression-related student suicides followed a downward trajectory from 2009 to the mid-2010s, but turned upward thereafter. This pattern raises the possibility that school counselling on its own, or student-directed counselling on its own, may not be sufficient to stem the rising tide of suicide among students with psychiatric disorders [96,97,98]. In the Japanese setting, school counsellors are most commonly clinical psychologists who deliver one-on-one counselling to students and provide indirect support to teachers. Because most are engaged on a non-full-time basis, with their on-site hours typically capped at around four hours per week, the time available for case conceptualisation, interdisciplinary case meetings, or coordination with external medical providers is severely constrained [56,59]. Even within their own schools, counsellors encounter obstacles to building collaborative relationships owing to fragmented schedules, ambiguous role definitions, and only sporadic chances for sustained interaction with teachers [56,57,58,59,60]. Under such conditions, it is difficult for school counsellors to coordinate the care needed by students with complex psychiatric presentations, students whose situations frequently call for ongoing monitoring, pharmacological management, and specialized interventions that extend beyond what school-based services can offer [56,57,58,59,60].

The internalizing disorders whose contribution to student suicide has been increasing, as documented here, fall within the category of treatable adolescent psychiatric conditions. Combined treatment with cognitive behavioural therapy and medication using selective serotonin transporter inhibitors has been reported to deliver a more substantial benefit than either modality administered alone [99]. In light of this, prefectural and municipal boards of education will need to invest school counsellors with the authority to set up collaborative arrangements with psychiatric medical institutions as needed, and to put in place an operating environment that allows psychiatric care, including pharmacotherapy where indicated, to be delivered promptly to those students who require it [100,101,102]. In a number of European countries, school counsellors function as formal liaisons with psychiatric services, yet no equivalent institutional role exists in Japan. The practical consequence is that general teachers, who have not received specialized training, end up performing duties analogous to those of European school counsellors [103,104]. Beyond linkages with psychiatric services, partnerships extending to parents, caregivers, teachers, and administrative bodies such as the prefectural Child Welfare Centre are equally important. The collaborative framework must therefore be expanded systemically as the situation requires, rather than depending on ad hoc “personal” arrangements between individuals.

4.4. Limitations

Direct comparability analysis between before and after the revision of the survey method of the SSNPA was not feasible, since the SSNPA did not provide formal bridge data derived from dual coding of cases under both the pre- and post-revision schemes. Consequently, we were constrained to adopt a jump-model JPRA alone, in which the effect of the coding revision is parameterized as an additive level shift (as jump parameter) at the changepoint year. This formulation assumes that the magnitude of the revision’s influence remains constant throughout the post-2022 period. Should the effect of the coding change have varied over time rather than remaining stable, the model would be unable to fully capture such temporal heterogeneity, underscoring the need for ongoing monitoring as additional data become available. Notwithstanding this limitation, the validity of the jump-model JPRA was supported through sensitivity analyses, including verification of the assumed changepoint year and a reproducibility check using data restricted to the pre-2022 period, both of which yielded results consistent with the primary findings. Nevertheless, because the data underlying this study are registry-based, the observed temporal trends cannot be directly linked to the psychological processes that motivated individual deaths. Although the suicide motives attributed in the SSNPA are officially determined through meticulous investigation, suicide is a multifactorial phenomenon, and the attributed motives may not necessarily identify the entire psychological, social, and clinical pathways leading to individual suicide. In other words, elucidating such mechanisms would require analyses using collected data derived from systematic psychological autopsy studies.

A further limitation of this study is the relatively short observation period of four years for analyzing high-risk groups among students, which is insufficient for robust analysis. Despite this constraint, the CMR of students in part-time high schools was markedly higher compared to all other groups, a finding that possibly remained consistent. However, the comparative results across the remaining groups should be interpreted with caution, as the brevity of the observation period introduces the possibility of estimation error and may compromise the stability of trend estimates for those subgroups. Continued surveillance with an extended observation window is therefore indispensable to verify and refine these comparative findings.

5. Conclusions

The present analysis documents a sustained rise in suicide mortality among Japanese students in middle school, high school, special vocational school, and university beginning in the mid-2010s. Of particular note, the CMRs among female students in middle and high school surpassed that of their male counterparts during the 2020s. Across student categories, the principal recorded motives for suicide underwent a developmentally patterned shift, beginning with family-related issues such as conflict with parents and parental reprimands in the youngest cohort and giving way, in accordance with the life cycle, to school-related concerns such as academic underperformance, problems associated with career trajectories, and conflict with peers. Increases in suicides attributed to depression and other psychiatric disorders were considerably more pronounced among female students than among male students. Concurrent with these trends, suicides attributed to social isolation and to conflict with peers began to rise from the late 2010s or early 2020s, signalling the emergence of socially struggling students as an additional risk group. Taken together, our observations point to a rising trend in student CMRs that took hold in the mid-2010s, with the steepest increases concentrated in female student populations and the CMRs of part-time high school students reaching exceptionally high levels. The leading recorded motives identified here vary as a function of age, spanning family, school, and mental health domains. Accommodating these age-related shifts in the design of suicide prevention programmes offers substantial promise for improving their effectiveness. Dedicated cohort investigations that permit fine-grained longitudinal monitoring of individual psychological trajectories and evolving suicide motives are likewise needed to clarify the psychological mechanisms underpinning these age-dependent shifts.