ABSTRACT
Aim
Children and adolescents in school sports are vulnerable to exertional heat illness, but detailed information on fatal events under school supervision in Japan is limited. This study aimed to characterize the circumstances of fatal heat‐related illness in this setting.
Methods
We conducted a nationwide retrospective observational study using the Japan Sport Council School Accident Search Database. We included deaths among preschool to high school students between April 2005 and March 2024 attributed to heat illness during school‐supervised activities. Demographics, school type, setting, activity at collapse, and first aid at the scene were abstracted.
Results
Of 1151 school‐related deaths, 27 (2.3%) were attributed to heat illness. Median age was 16 years (interquartile range: 15–17); 25 students (93%) were male, and 19 (70%) were high school students. First‐year high school students represented the largest subgroup (n = 8, 30%). Nineteen deaths (70%) occurred outdoors and eight (30%) indoors, most commonly in gymnasiums. Extracurricular sports accounted for 21 deaths (78%), including judo, soccer/futsal, rugby, and baseball. Running preceded collapse in 13 cases (48%). Witnesses were present in 21 cases (78%), yet cooling was documented in 9 cases (33%), while whether cooling was performed could not be determined in 12 cases (44%).
Conclusion
Fatal heat‐related illness in Japanese schools predominantly affects first‐year high school students engaged in strenuous conditioning during extracurricular sports and occurs in outdoor and indoor environments. Limited documentation of cooling despite witnesses indicates an urgent need to strengthen school‐based prevention and emergency response systems, including preparedness for rapid on‐site cooling.
Keywords: adolescent, heat stroke, Japan, schools, sports
A nationwide analysis of 27 heat‐related deaths during school‐supervised activities in Japan showed that cases predominantly involved male high school students, especially first‐year students, and occurred both outdoors (70%) and indoors (30%). Although witnesses were present in 78% of cases, cooling was documented in only 33%, underscoring the importance of preparedness for rapid on‐site cooling.
1. Introduction
Climate change has substantially increased the frequency and intensity of extreme heat events worldwide, making heat‐related illness an urgent public health concern [1]. Exertional heat illness in youth is strongly associated with preventable risk factors such as excessive physical exertion, inadequate rest, and insufficient acclimatization [2]. Consequently, most heat illness in this population is considered preventable; however, heat‐related incidents under school supervision continue to occur. In Japan, more than 17,000 students sought medical care for heat illness during school activities over the past five years [3].
In addition to physical education classes, extracurricular sports activities pose a substantial risk for heat illness, including fatal cases. Heat‐related deaths have been reported predominantly among male high school students and are frequently associated with specific sports such as baseball and judo [4]. Fatal events commonly occur during conditioning activities such as running, and several cases have documented insufficient first aid at the scene [5].
Despite the public health relevance, detailed epidemiological data—such as the exact location, situational context, activity immediately preceding collapse, and first aid measures—remain limited for fatal heat‐related illness under school supervision in Japan. This study aimed to describe the specific circumstances surrounding heat‐related deaths in Japanese schools and to inform more effective, context‐specific prevention strategies.
2. Methods
2.1. Study Design and Data Source
We conducted a nationwide retrospective observational study in accordance with the STROBE guidelines [6], using the School Accident Search Database maintained by the Japan Sport Council (JSC) [7]. The JSC operates a school accident mutual aid system that provides compensation for injuries, disabilities, or deaths occurring under school supervision. Eligible incidents include those during class hours, extracurricular activities, recess periods, and commuting. As of 2024, the coverage rate of this system is 94.8%, capturing nearly all school‐related accidents nationwide.
2.2. Study Population
We reviewed all incidents recorded between April 1, 2005, and March 31, 2024 (fiscal years [FY] 2005–2023). Cases were included if the fatality occurred among children or adolescents enrolled from preschool through high school and was classified as heat illness in the JSC database. Death‐benefit claims require submission of a death certificate or postmortem examination certificate; however, the public database does not disclose detailed clinical information reviewed, whether standardized diagnostic criteria were applied, or how competing causes of death were excluded in individual cases. Therefore, the heat‐illness classification was treated as an administrative classification and was not independently reassessed by the study investigators.
2.3. Variables and Data Extraction
For each case, we extracted variables including academic year, school type, grade level, sex, situational category, sport type, location of occurrence, and contextual details. Narrative fields describing witness presence and first aid measures were independently reviewed by two investigators (S.S. and R.M.). Discrepancies were resolved by a board‐certified emergency physician (Y.I.). Since exact ages were not consistently recorded, age was estimated based on the standard age corresponding to each school grade in the Japanese education system (e.g., 13 years for first‐year junior high school students).
2.4. Statistical Analysis
Descriptive statistics were used to summarize the data. Continuous variables are reported as medians with interquartile ranges (IQR). Categorical variables are presented as counts and percentages. Given the small sample size, we did not perform formal hypothesis testing.
2.5. Calculations of Mortality Rates
To calculate mortality rates, we obtained the annual number of students enrolled in junior high and high schools from the School Basic Survey (2005–2023) conducted by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) [8]. The total number of person‐years was determined by summing the annual enrollment numbers over the 19‐year study period. Mortality rates were calculated as the number of deaths per 100,000 person‐years.
2.6. Ethics
This study used anonymized, publicly available data; thus, approval from an institutional review board was not required under Japanese ethical guidelines for medical research involving human subjects. The study was conducted in accordance with the principles of the Declaration of Helsinki.
3. Results
Of the 1151 total deaths recorded during the study period, 27 (2.3%) were attributed to heat illness, with fatal cases identified as recently as FY2023. The number of deaths declined across the study period, from 10 in FY2005–FY2009 and 10 in FY2010–FY2014 to 5 in FY2015–FY2019 and 2 in FY2020–FY2023, corresponding to means of 2.0, 2.0, 1.0, and 0.5 deaths per year, respectively (Figure 1). Fatal cases nevertheless continued to occur throughout the period, with the most recent death recorded in FY2023.
FIGURE 1.
Trends in fatal heat‐related illness during school‐supervised activities in Japan, FY2005–FY2023. Bars show the number of heat‐related deaths (n = 27) grouped by fiscal‐year period. Data were extracted from the Japan Sport Council School Accident Search Database. The final period (FY2020–FY2023) spans four fiscal years, whereas the preceding periods span five; the corresponding mean numbers of deaths per year were 2.0, 2.0, 1.0, and 0.5.
The epidemiological characteristics of the 27 fatal heat‐related illness cases are summarized in Table 1. The median age of students was 16 years (IQR 15–17), and most were male (25 cases, 93%), with only two female students (7%). High school students accounted for the largest proportion of deaths (19 cases, 70%), followed by junior high school students (6 cases, 22%). The crude mortality rate was 0.030 per 100,000 person‐years for high school students, whereas it was 0.009 per 100,000 person‐years for junior high school students. First‐year high school students (10th grade) represented the single largest subgroup (8 cases, 30%).
TABLE 1.
Epidemiological characteristics of fatal heat‐related illness cases in Japanese schools (2005–2023).
| All cases (N = 27) | ||
|---|---|---|
| n | (%) a | |
| Age (years) (median, [IQR]) | 16 [15–17] | |
| Sex | ||
| Male | 25 | (93%) |
| Female | 2 | (7%) |
| Grade | ||
| Nursery school | 1 | (4%) |
| Elementary school | 1 | (4%) |
| 1st year | 1 | (4%) |
| Junior high school | 6 | (22%) |
| 1st year (7th grade) | 3 | (11%) |
| 2nd year (8th grade) | 1 | (4%) |
| 3rd year (9th grade) | 2 | (7%) |
| High school | 19 | (70%) |
| 1st year (10th grade) | 8 | (30%) |
| 2nd year (11th grade) | 6 | (22%) |
| 3rd year (12th grade) | 5 | (19%) |
| Incident location | ||
| Outdoors | 19 | (70%) |
| Schoolyard or sports field | 11 | (41%) |
| Roadway or other public area | 8 | (30%) |
| Indoors | 8 | (30%) |
| Gymnasium | 6 | (22%) |
| Other school facilities | 2 | (7%) |
| By situation | ||
| Extracurricular sports activities | 21 | (78%) |
| On the way home | 2 | (7%) |
| Physical education class | 1 | (4%) |
| Athletic festival | 1 | (4%) |
| Recess | 1 | (4%) |
| During childcare | 1 | (4%) |
| By sport | ||
| Judo | 4 | (15%) |
| Soccer/Futsal | 3 | (11%) |
| Rugby | 3 | (11%) |
| Baseball | 3 | (11%) |
| Handball | 2 | (7%) |
| Kendo | 2 | (7%) |
| Others b | 5 | (19%) |
| Immediate preceding activity c | ||
| Running d | 13 | (48%) |
| Sports practice | 6 | (22%) |
| Training/exercise d | 5 | (19%) |
| Cycling | 2 | (7%) |
| Competition | 2 | (7%) |
| Hiking | 1 | (4%) |
| Others d | 3 | (11%) |
| Witness | ||
| Present | 21 | (78%) |
| Absent | 6 | (22%) |
| First aid c | ||
| Icing/cold therapy | 9 | (33%) |
| Rescue breathing/chest compressions/AED | 4 | (15%) |
| Airway management | 1 | (4%) |
| Unknown e | 2 | (7%) |
| None | 4 | (15%) |
| Not recorded e | 10 | (37%) |
Note: Data are presented as number (percentage) unless otherwise indicated.
Abbreviations: AED, automated external defibrillator; IQR, interquartile range.
Percentages may not total 100% because of rounding.
“Others” consists of basketball, mountaineering, American football, and track and field sprinting, and track and field throwing (n = 1 each).
Categories are not mutually exclusive; multiple activities or first aid measures were identified for single cases based on the review of case descriptions.
Classification was based on the specific terminology used in the database. “Running” includes cases explicitly recorded as running, jogging, dashing, or sprinting. “Training/exercise” includes cases recorded as training, circuit training, or interval training. Note that activities categorized as “Training/exercise” may inherently involve running elements, but lacked explicit description in the records. “Others” consists of walking home, resting after off‐campus learning, and confinement in a storage unit (n = 1 each).
“Unknown” indicates that first aid was documented as having been provided, but the specific intervention was not identified. “Not recorded” indicates that no information regarding first aid was recorded in the database. Thus, whether cooling was performed could not be determined in 12 cases (44%).
Regarding location, 19 deaths (70%) occurred outdoors, while 8 (30%) occurred indoors; most indoor deaths (6 cases, 22%) took place in gymnasiums. Extracurricular sports activities were the most frequent context (21 cases, 78%), whereas only a minority occurred during regular physical education classes. Among cases with identifiable sports, judo was most common (4 cases, 15%), followed by soccer/futsal, rugby, and baseball (3 cases each, 11%).
Running was the activity immediately preceding collapse in 13 cases (48%). Witnesses were present in 21 cases (78%). First aid measures involving cooling (ice application or other cooling methods) were documented in 9 cases (33%), while no first aid was reported in 4 cases (15%). Whether cooling was performed could not be determined in 12 cases (44%): No first‐aid information was recorded in 10 cases (37%), and first aid was documented in 2 cases (7%), but the specific intervention was not identified.
4. Discussion
This nationwide 19‐year analysis identified the epidemiological characteristics of 27 fatal heat‐related illness cases occurring under school supervision in Japan. Key findings include: (1) 70% of deaths occurred among high school students, particularly first‐year students; (2) approximately 30% of fatalities occurred indoors; and (3) despite the presence of witnesses in most cases, cooling was documented in only one‐third of incidents, and no first aid was administered in 15%. These findings provide a contemporary description of fatal heat‐related illness and documented emergency responses in Japanese school settings.
The concentration of fatalities among first‐year high school students is consistent with the earlier nationwide report [4]. Compared with that report, the present study extends observations through FY2023 and provides more detailed information on event location, activity immediately preceding collapse, witness presence, and documented first aid. The crude mortality rate among high school students was approximately 3.3‐fold higher than that among junior high school students. One possible explanation is an increase in training intensity and volume during the transition to high school; however, these factors were not directly assessed in the present study [9]. The prominence of running as the activity immediately preceding collapse supports the established risk associated with excessive conditioning and suggests that adjustment of conditioning programs and rest periods for newly enrolled students may be warranted.
The finding that 30% of deaths occurred indoors challenges the widespread perception that heat illness is primarily an outdoor hazard. Indoor sports such as judo and kendo are often conducted in poorly ventilated, high‐humidity environments. Evidence indicates that high humidity substantially increases heat illness risk even at moderate ambient temperatures [10]; the substantial proportion of indoor fatalities observed in this study underscores the potential danger of this mechanism. Countermeasures such as continuous wet‐bulb globe temperature (WBGT) monitoring and active ventilation strategies are essential even in indoor athletic facilities [11].
A major concern is the limited documentation of on‐site cooling. Witnesses were present in 78% of cases, while cooling was documented in 33% and no first aid was reported in 15%; however, the available first‐aid documentation was insufficient to determine whether cooling was performed in 12 cases (44%). Therefore, the actual implementation of cooling should be interpreted cautiously.
Internationally, exertional heat stroke (EHS) remains a leading cause of sport‐related fatalities in the United States, particularly among secondary‐school and collegiate football athletes during practice and conditioning [12, 13]. Stearns et al. reported that 94% of fatal EHS cases in U.S. secondary‐school sports occurred in American football [13]. In contrast, fatalities in our Japanese dataset occurred across a broader range of sports and settings, with running preceding collapse in 48% and 30% of deaths occurring indoors. However, direct comparison is limited by differences in study populations and settings [13]. U.S. military and athletic organizations have adopted structured prevention strategies, including WBGT‐guided work‐rest ratios and mandated heat‐acclimatization periods [12, 14]. International recommendations emphasize rapid whole‐body cooling, particularly cold‐water immersion (CWI), to reduce core temperature promptly [15, 16, 17]. Previous Japanese research reported no use of CWI among fatal high‐school EHS cases, which differs from current international recommendations emphasizing rapid whole‐body cooling [5].
Notably, fatalities occurred even in cases in which cooling was documented; however, the available data do not allow assessment of the timing, modality, or effectiveness of cooling. Furthermore, although the number of deaths per period declined over the 19 years (Figure 1), fatal cases continued to occur after the 2019 guideline update. These observations alone do not allow assessment of the effectiveness or implementation of the guideline update [18, 19].
This study has several limitations. First, the overall small sample size limits the precision of the findings. Second, as compensation‐based data, detailed clinical information such as rectal temperature or specific WBGT measurements was unavailable. Specifically, the lack of documented body temperature data precluded the application of the Japanese Association for Acute Medicine (JAAM) heat‐illness severity classification. Third, in 12 cases (44%), the available first‐aid documentation was insufficient to determine whether cooling was performed, limiting the completeness of the analysis. Fourth, reliance on administrative classification introduces the possibility of outcome misclassification. Because detailed clinical and autopsy information was unavailable, we could not independently verify the diagnosis of heat illness or definitively exclude competing causes of sudden collapse, such as fatal arrhythmias or cerebrovascular events. In addition, because month and season of occurrence were unavailable, we could not assess whether seasonal or contextual factors influenced diagnostic classification. Fifth, the database lacks denominator data on the background population participating in each sport, precluding the calculation of sport‐specific incidence or relative risks. Finally, the small number of fatal cases after 2019 precluded meaningful assessment of the effectiveness of the guideline update. In addition, Figure 1 presents crude case counts that were not adjusted for the concurrent decline in the number of enrolled students in Japan, and the final period spans four rather than five fiscal years; the apparent decrease should therefore not be interpreted as evidence of improved prevention.
5. Conclusion
Fatal heat‐related illness in Japanese schools is characterized by its concentration among first‐year high school students, occurrence in both outdoor and indoor environments, and limited documentation of first aid, particularly cooling. These findings provide essential evidence to guide targeted, actionable measures by school administrators, boards of education, and school physicians, including risk‐aware training plans for first‐year students, active management of indoor heat stress, and preparation for rapid on‐site cooling.
Funding
This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology.
Ethics Statement
Ethical approval was not required for this study in accordance with national guidelines, as the data used were anonymized and publicly available from the Japan Sport Council database.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
We would like to thank the Japan Sport Council for making the School Accident Search Database publicly available for research purposes.
Data Availability Statement
The data that support the findings of this study are available from the Japan Sport Council School Accident Search Database (https://www.jpnsport.go.jp/anzen/).
References
- 1. Ebi K. L., Capon A., Berry P., et al., “Hot Weather and Heat Extremes: Health Risks,” Lancet 398, no. 10301 (2021): 698–708, 10.1016/S0140-6736(21)01208-3. [DOI] [PubMed] [Google Scholar]
- 2. Council on Sports M, Fitness; Council on School Health . “Policy Statement‐Climatic Heat Stress and Exercising Children and Adolescents,” Pediatrics 128, no. 3 (2011): e741–e747, 10.1542/peds.2011-1664. [DOI] [PubMed] [Google Scholar]
- 3. Japan Sport Council , “Heat‐Illness Prevention Information and Materials Site [in Japanese],” (2025) accessed October 3, 2026, https://www.jpnsport.go.jp/anzen/heat/tabid/3056/Default.aspx.
- 4. Hatori Y., “Heat Stroke in Schools,” Journal of the Japan Medical Association 56, no. 3 (2013): 179–185. [Google Scholar]
- 5. Yamanaka M. S., Hosokawa Y., Ayusawa M., et al., “Epidemiology of Sports‐Related Fatalities During Organized School Sports in Japanese High Schools Between 2009 and 2018,” PLoS One 16, no. 8 (2021): e0256383, 10.1371/journal.pone.0256383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. von Elm E., Altman D. G., Egger M., Pocock S. J., Gøtzsche P. C., and Vandenbroucke J. P., “The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies,” Journal of Clinical Epidemiology 61, no. 4 (2008): 344–349, 10.1016/j.jclinepi.2007.11.008. [DOI] [PubMed] [Google Scholar]
- 7. Japan Sport Council . Disaster Mutual Aid Benefit Status in Fiscal Year 2024: Preliminary Report [in Japanese] (2024). accessed October 3, 2026, https://www.jpnsport.go.jp/anzen/kankobutuichiran/tabid/3064/Default.aspx. [Google Scholar]
- 8. Ministry of Education Culture, Sports, Science, and Technology , School Basic Survey (Long‐Term Statistics) [in Japanese] (2025). accessed October 3, 2026, https://www.mext.go.jp/b_menu/toukei/chousa01/kihon/1267995.htm.
- 9. Coris E. E., Ramirez A. M., and Van Durme D. J., “Heat Illness in Athletes: The Dangerous Combination of Heat, Humidity and Exercise,” Sports Medicine 34, no. 1 (2004): 9–16, 10.2165/00007256-200434010-00002. [DOI] [PubMed] [Google Scholar]
- 10. Bouchama A. and Knochel J. P., “Heat Stroke,” New England Journal of Medicine 346, no. 25 (2002): 1978–1988, 10.1056/NEJMra011089. [DOI] [PubMed] [Google Scholar]
- 11. Hampo C. C., Schinasi L. H., and Hoque S., “Surviving Indoor Heat Stress in United States: A Comprehensive Review Exploring the Impact of Overheating on the Thermal Comfort, Health, and Social Economic Factors of Occupants,” Heliyon 10, no. 3 (2024): e25801, 10.1016/j.heliyon.2024.e25801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Anderson S. A., Eichner E. R., Bennett S., et al., “Preventing Exertional Heat Stroke in Football: Time for a Paradigm Shift,” Sports Health 17, no. 3 (2025): 484–490, 10.1177/19417381241260045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Stearns R. L., Kucera K. L., Hosokawa Y., et al., “Fatal Exertional Heat Stroke Trends in Secondary School Sports From 1982 Through 2022,” Sports Health 17, no. 5 (2025): 951–957, 10.1177/19417381241298293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Periard J. D., DeGroot D., and Jay O., “Exertional Heat Stroke in Sport and the Military: Epidemiology and Mitigation,” Experimental Physiology 107, no. 10 (2022): 1111–1121, 10.1113/EP090686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. American College of Sports Medicine; Armstrong L. E., Casa D. J., et al., “American College of Sports Medicine Position Stand. Exertional Heat Illness During Training and Competition,” Medicine and Science in Sports and Exercise 39, no. 3 (2007): 556–572, 10.1249/MSS.0b013e31802fa199. [DOI] [PubMed] [Google Scholar]
- 16. Casa D. J., McDermott B. P., Lee E. C., et al., “Cold Water Immersion: The Gold Standard for Exertional Heatstroke Treatment,” Exercise and Sport Sciences Reviews 35, no. 3 (2007): 141–149, 10.1097/jes.0b013e3180a02bec. [DOI] [PubMed] [Google Scholar]
- 17. Filep E. M., Murata Y., Endres B. D., Kim G., Stearns R. L., and Casa D. J., “Exertional Heat Stroke, Modality Cooling Rate, and Survival Outcomes: A Systematic Review,” Medicina (Kaunas, Lithuania) 56, no. 11 (2020): 589, 10.3390/medicina56110589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Hosokawa Y., Nagata T., and Hasegawa M., “Inconsistency in the Standard of Care‐Toward Evidence‐Based Management of Exertional Heat Stroke,” Frontiers in Physiology 10 (2019): 108, 10.3389/fphys.2019.00108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Japan Sport Association . A Guidebook for the Prevention of Heat Stroke During Sports Activities, 5th ed. [in Japanese] (2019). [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the Japan Sport Council School Accident Search Database (https://www.jpnsport.go.jp/anzen/).