Heat Exposure and the Risk of Emergency Hospitalization in Germany

Stratified analyses by age, sex, and diagnostic group

Background: Climate change has increased the frequency and intensity of periods of extreme heat, which are associated with elevated mortality. Little is known about the association of emergency hospital admissions with particular risk groups.

Methods: We used anonymized data from the statutory health insurance system in Germany (4.3 million persons; 1.48 million emergency hospital admissions) and temperature data (resolution: 9 x 9 km) to estimate the risk of emergency hospitalization as a function of daily mean temperature during the summer months over the period 2017–2022. The time-series analysis that was used also enabled the calculation of temporally delayed and non-linear associations (distributed lag non-linear model). We conducted subgroup analyses by age, sex, and diagnostic group to identify vulnerable groups.

Results: Our study showed that the risk of emergency hospitalization increased with increasing temperatures. For daily mean temperatures above 17° C, we estimated an additional 7002 emergency admissions over the 6-year study period, with a cumulative overall risk of 1.03 (95% confidence interval: [1.01; 1.04]) compared to days with a mean temperature of 17° C. On days with extreme heat (mean daily temperature, 25° C), the relative risk (RR) of emergency hospitalization was increased for children (RR: 1.06 [1.01; 1.11]) and persons over age 65 (RR: 1.03 [1.01; 1.05]). As for diagnostic groups, the relative risk for emergency hospitalization associated with days of extreme heat was higher among persons with endocrine, nutritional, and metabolic diseases (1.63 [1.50; 1.78]), diseases of the skin and subcutaneous tissue (1.18 [1.07; 1.32]), genitourinary diseases (1.13 [1.03; 1.24]), respiratory diseases (1.09 [1.02; 1.17]), and injuries, poisoning, and other consequences of external causes (1.05 [1.02; 1.09]).

Conclusion: The overall effect of heat on the risk of emergency hospitalization was small, yet in some cases considerable for children, persons aged 65 and older, and certain diagnostic groups. This indicates a need for targeted preventive measures.

Cite this as: Kriit HK, Herrmann A, Norris R, Rocklöv J, De Allegri M: Heat exposure and the risk of emergency hospitalization in Germany: Stratified analyses by age, sex, and diagnostic group. Dtsch Arztebl Int 2025; 122: 709–14. DOI: 10.3238/arztebl.m2025.0186

Rising temperatures together with increased frequency and intensity of extreme heat events are a direct consequence of climate change (1). According to the European State of the Climate Report, the past 3 years have been the warmest on record, with an increasing trend of days with “strong heat stress” and “extreme heat stress” (2). Extreme heat increases mortality globally (3) and is recognized as a leading cause of weather- and climate-related deaths in Europe (2). A comprehensive analysis of heat-related deaths from 1992 to 2021 shows that in hot years, between 3000 and 10 000 heat-related deaths occurred in Germany (4). However, the effects of extreme heat events extend beyond increased mortality. Morbidity also increases, reflected by higher demand on healthcare services (5), especially among vulnerable groups such as older adults, children, persons with chronic diseases, and pregnant women (6). Extreme heat can cause heat-related illness such as heat stroke and aggravate existing illnesses (6, 7). Moreover, extreme heat increases the risk of accidents and decreases the performance of the working population (6, 8, 9). Understanding the health impacts of extreme heat events is important, since most instances of heat-related morbidity and mortality are preventable with improved preparedness (6).

The World Health Organization (WHO) recommends that health care systems be prepared for a higher demand for healthcare services during periods of extreme heat. However, much more research is required into the impact of heat on healthcare services in Europe (10), including Germany (8).

Our study aimed to describe the relationship between mean daily temperature and emergency hospitalizations (EH) in Germany and to determine the scale of risk for different demographic groups and different disease diagnostic categories.

Methods

Exposure

We extracted the mean daily temperatures for the period of 2017–2022 from the ERA5-Land dataset, with a resolution of 9 × 9 km, and aggregated them to the federal state level (11, 12). The mean daily temperature refers to the average temperature of a 24-hour period. Since we were interested only in heat effects on EH, the analysis period was restricted to the summer months (May–September). In this study we refer to mean temperature ranges between 17 and 20 °C as “low,” from 21 to 24 °C as “moderate,” and ≥ 25° C as “extreme” heat exposure. In the predictor-specific analyses, heat exposure from the 95th (21° C) and the 97th (22° C) percentile was defined as “moderate” and from the 99th (25° C) percentile as “extreme.”

Outcome

Daily counts of EH for each of the 16 federal states of Germany were obtained from the research database of the Institute for Applied Health Research Berlin GmbH (InGef). This research database contains the anonymized longitudinal healthcare claims data of approximately 8.8 million persons with statutory health insurance from over 50 different providers. The InGef database sampling strategy ensures the representativeness by age and sex (13), so it has hospitalization rates similar to those in the overall German population. It is less representative, however, with regard to insurance coverage (e.g., private insurance is excluded). The data for the inpatient sector comprise date of admission, diagnosis on admission, and type of hospitalization (e.g., planned admission and emergency hospitalization). All diagnoses are coded using the German modification of the International Classification of Diseases version 10 (ICD-10-GM). In this study, a sample of 4.3 million persons was available for analysis.

Statistical analysis

We used a distributed lag non-linear model (DLNM), a special form of time-series analysis. This model permits analysis of the delayed and non-linear relationship between heat and EH over the 6-year period for each federal state in Germany. The method is well established and was developed specifically to capture non-linear effects, e.g., a steeper risk increase for hospitalizations at higher temperatures. Moreover, delayed effects can be taken into account, e.g., hospitalization a few days after the heat exposure (14). In our model we also accounted for the effects of the day of the week and for long-term and seasonal trends (14). Temperature effect was modeled using a combination of functions to capture the non-linear risk for EH above the reference value and delayed effects more than 6 days after exposure (14). We defined the average mean temperature of 17° C, corresponding to the 80th percentile, as the reference temperature. Additionally, we estimated the effect of days with moderate and extreme heat exposure on EH over a 6-day period (15), because we found no effects with a lag time exceeding 6 days. To assess the model fit for each region, we examined standard model diagnoses such as autocorrelation and residual patterns (eSupplement Figure S3).

Moreover, we used a multivariate random-effect meta-analysis (16) to pool the state-specific estimates (eSupplement Figures S1, S2) and to calculate an average relative risk (RR) estimate for EH in relation to the mean temperature range for Germany.

Furthermore, we determined the attributable number (AN) and attributable fraction (AF) of EH to the range of mean temperature for each state, using distributed lag non-linear models (17). We assumed no heat effects on EH if negative effects of heat were estimated for a specific state (18).

Finally, we conducted a subgroup analysis by gender, age group (children 0–17 years, adults 18–64 years, and older adults 65 +), and ICD-10-GM group.

We conducted all analyses with RStudio software (version 4.1.1), using the packages splines, dlnm, and mixmeta. ChatGPT (version GPT-5) was used to enhance the readability of the language in the methods section, and all changes suggested by artificial intelligence were carefully reviewed by the authors.

Ethics statement

The anonymization of the InGef research database ensures that individual patients, health insurance companies, and service providers can no longer be identified. Separate approval by an institutional review board or medical ethics committee was thus not required. We followed the general principles outlined in the Helsinki Declaration.

Results

Descriptive

The mean daily temperature for the summer months over the 6-year study period was 17° C, corresponding to the 80th percentile of the yearly mean temperature. The 99th, 97th, and 95th percentiles of mean daily temperature corresponded to 25° C, 22° C, and 21° C respectively. (See Appendix Table 1 for daily temperature and mean daily EH in each 16 federal states). In total, 1.48 million EH were included in the analysis, of which older adults accounted for 48%, adults 45%, and children 7% (Table). There was an almost equal distribution of men and women: 48% and 52% respectively. The largest proportion of diagnosis-specific codes was accounted for by symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified (R00–R99), followed by injury, poisoning and certain other consequences of external causes (S00–T98).

Table

Emergency hospitalizations (EH): overall and stratified by gender, age group, and diagnostic category

eTable

Heat-associated emergency hospitalizations (EH) in the federal states of Germany

Relationship between temperature and EH

For the state-specific pooled estimates, the overall cumulative risk for EH increased with rising mean temperature (Figure 1). The overall cumulative risk for EH was 1.03 [95% confidence interval 1.01; 1.04] in relation to the temperature range above the reference value. The heterogeneity of the meta-analysis was very low (I2 = 1%), suggesting the absence of any substantial regional differences. In absolute terms, we estimated for this sample 7002 EH in the summer months of the period 2017–2022 that were attributable to mean daily temperatures above the 80th percentile. This corresponds to 0.47% of all EH in this sample (n = 1 481 706).

Figure 1

Pooled cumulative risk estimates for emergency hospitalization (EH) in relation to temperature

For both men and women, the risk of EH increased with rising temperature (Figure 2). Overlapping confidence intervals indicate the absence of any significant difference in risk between the sexes. For adults the risk increased for days with low to moderate temperatures and declined with extreme temperatures. For children and older adults, however, the risk of EH increased sharply for moderate and extreme temperatures (Figure 2). On days with temperatures over the 95th, 97th and 99th percentiles, predictor-specific analysis suggested an increased risk of EH on the same day for men, children and older adults (eSupplement Figure S3: B, E, F). A somewhat delayed EH risk was seen after 2–3 days (lag 2–3) among woman and adults (eSupplement Figure S5: C, D).

Figure 2

Overall cumulative association of temperature and EH: a) in men and women; b) in older adults, adults, and children. Note the different scales for relative risk (RR).

Diagnosis-specific analysis

On days of extreme heat (25 °C mean daily temperature, against the reference value of 17 °C), the relative risk (RR) of EH due to endocrine, nutritional, and metabolic diseases increased by 63% (RR 1.63 [1.50; 1.78]). The figure for skin and subcutaneous tissue diseases was 18% (1.18 [1.07; 1.32]), for genitourinary system diseases 13% (1.13 [1.03; 1.24]), for respiratory diseases 9% (1.09 [1.02; 1.17]), and for injuries, poisoning, and certain other external causes 5% (1.05 [1.02; 1.09]) (Figure 3: a–e). Predictor-specific analysis suggested an immediate risk increase (lag 0–1) for all of these diagnostic groups. However, a somewhat delayed (lag 3–5) increase in relative risk was also observed for the following diagnostic groups: injuries, burns and poisoning, genitourinary disease, skin and subcutaneous tissue disease, and factors influencing health status and contact with health services (eSupplement Figure S6).

Figure 3

Relative risk (RR) for diagnosis-specific EH by mean temperature

Decreased risk of EH with increasing mean daily temperature was observed for cardiovascular disease (I00–I99) (eSupplement Figure S4: C). Furthermore, positive associations were observed for infectious diseases (A00–B99), mental illness (F00–F99), and pregnancy, childbirth and puerperium (O00–O99) in relation to low and moderate heat exposure, but not for extreme heat (eSupplement Figure S4: A, B, D). No other diagnostic groups revealed statistically significant results.

Discussion

This is the first nationwide study in Germany to demonstrate the association between heat exposure and EH across all age groups and in different diagnostic categories. The strength of our study lays in the use of data from a health insurance database that is representative in terms of age and sex distribution. The data span a period of 6 years and granted us the ability to perform a nationwide analysis of EH. In contrast, previous studies were confined to individual regions (19, 20), investigated only planned and unplanned hospitalizations (21, 22), or examined only heat-related diagnostic codes (18). Our study contributes valuable evidence for preparedness planning in health care systems, e.g., designing climate change adaptation strategies and evaluating their efficacy (8).

Health effects

Heat exposure was seen to increase the risk of EH across all age groups. Adults below the age of 65 years were at higher risk only with low and moderate heat exposure, as demonstrated by a study in Australia (9). This may be due to workplace precautions in extreme temperatures (e.g., ceasing work [23]), or adults may change their behavior to reduce heat exposure during extreme heatwaves. In line with previous studies, children (24) and persons aged over 65 years (7) faced a higher risk in relation to extreme heat. Both physiological and behavioral attributes can contribute to this increased risk. Children, for example, have a less favorable ratio of body surface area to body mass (25), which hampers heat dissipation, and spend more time outdoors (26), while a factor in older adults is decreased capacity for thermoregulation due to poorer cutaneous perfusion and reduced sweating (6). Our findings also confirm the pattern identified in previous studies for diagnostic categories, where an increased risk of hospitalization due to metabolic diseases (21, 27), respiratory diseases (7, 21), skin and subcutaneous tissue diseases (21, 27), kidney and urinary tract diseases (21, 27, 28), and trauma (21, 27) in relation to heat exposure has been reported. Volume depletion (ICD E86) and other disorders of fluid, electrolyte and acid-base balance (ICD E87) form part of the group of endocrine, nutritional and metabolic diseases (ICD E00–E90) and may partially explain the high increase in EH at extreme temperatures in this diagnosis group. Another possible factor in the dramatic rise in ICD E00–E90 are diabetes-related diagnoses, as persons with diabetes are known to be at greater risk of suffering from heat (27), especially if their blood glucose levels are poorly controlled (29, 30). We observed a slightly decreased risk for EH due to cardiovascular disease in relation to heat, also confirming the findings of previous studies (31, 32). It has already been hypothesized that heat may exacerbate acute cardiovascular conditions and lead to out-of-hospital deaths before the patient can be delivered to the hospital (33).

This study suggests that, to use public health and healthcare resources efficiently, prevention measures should be targeted at particularly vulnerable groups, as EH are not increasing at the same pace across all age and disease groups. EH can best be prevented in the prehospital setting, either at the public health level or in primary care. For instance, children could be protected more effectively by the introduction of heat-protective measures at day-care facilities, kindergartens, and schools, as well as by the provision of targeted educational materials for families (34). To protect older adults, nursing homes and retirement homes should take efficacious steps to reduce heat exposure (35, 36). At the primary care level, high-risk patients could be identified and educated about the health risks of heat before and during the summer. For example, at times of high temperature persons with diabetes should be informed about protective measures, including the storage of insulin and the management of treatment (37).

Limitations

In the exposure data, we did not capture the urban heat island effects; therefore, a possible exposure misclassification might lead to underestimation of the effects of heat on EH. The metaestimate suggests minor regional differences, which might also be attributable to crude exposure measurement being unable to capture variations in smaller geographical units. Moreover, we concentrated on air temperature to the exclusion of humidity or other measures, because the additional effects of humidity are negligible in large population-based studies (38). Also, we did not take account of air pollution as a possible effect modifier.

In two regions (Brandenburg, Mecklenburg–Western Pomerania; eSupplement Figure S1) we observed a negative association between heat and EH, These results need to be interpreted with caution owing to the low population density in these regions and the low regional representativeness for Brandenburg in the InGef database. Heat exposure could also be lower due to the lesser population density and predominantly rural structure of these federal states.

Due to the short observation period, the crude measurement of exposure, and the small sample size for some of the diagnostic groups, uncertainty remains regarding the effects of heat exposure on EH. It would be desirable for future studies to feature larger sample sizes and more specific diagnostic groups or for analyses to be broken down by pre-existing conditions.

Conclusion

Although the relation between heat and EH in Germany was slight overall, yet remarkable for certain diagnostic groups, the relative risks were higher, in some cases much higher, for certain diagnostic groups, e.g., endocrine, nutritional and metabolic diseases. Children and older adults were at the greatest risk of EH. These findings should be used to tailor prevention measures to high-risk groups at the level of public health and in healthcare.

Funding

This study was supported through state funds approved by the State Parliament of Baden–Württemberg for the Innovation Campus Health + Life Science Alliance Heidelberg Mannheim.

Acknowledgments

We would like to thank Christofer Åström, Johan Nilsson Sommar, and Prasad Liyanage for their guidance on the methodology applied. We are also grateful to Marion Ludwig and Stella Dafka for providing expert insights into the health and climate data sources utilized.

Conflict of interest statement

AH is a member of KLUG e.V. and spokesperson of the Climate Change and Health section of the German Society for General Practice and Family Medicine (DEGAM). AH has received institutional support for a project on adaptation to climate change in the health care system from the innovation fund of the German Federal Joint Committee(G-BA) and personal payments for training courses on the topic of heat exposure and health from the Institute for Continuing Education of Primary Care Physicians (Institut für Hausärztliche Fortbildung, IHF e.V.).

JR receives support from the Alexander von Humboldt Foundation.

HKK, MDA, and RN declare that no conflict of interest exists.

Manuscript received on 27 March 2025, revised version accepted on 15 October 2025

Corresponding author
Dr. Hedi Katre Kriit

References

1. IPCC: Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, H. Lee and J. Romero (eds.)]. IPCC, Geneva, Switzerland 2023; 184 pp.

2. Adverse effects of climate change on health Copernicus EU2023 www.climate.copernicus.eu/esotc/2023/extreme-weather-and-human-health#ed53e998-1cd9-403e-83e2-11b5fd1b4eff (last accessed on 23 July 2024).

3. Zhao Q, Guo Y, Ye T, et al.: Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: A three-stage modelling study. Lancet Planet Health 2021; 5: e415–e25

4. Winklmayr C, Muthers S, Niemann H, Mücke HG, an der Heiden M: Heat-related mortality in Germany from 1992 to 2021. Dtsch Arztebl Int 2022; 119: 451–7

5. Mason H, King JC, Peden AE, Frankling RC: Systematic review of the impact of heatwaves on health service demand in Australia. BMC Health Serv Res 2022; 22: 960

6. Ebi KL, Capon A, Berry P, et al.: Hot weather and heat extremes: Health risks. Lancet 2021; 398 (10301): 698–708

7. Bunker A, Wildenhain J, Vandenbergh A, et al.: Effects of air temperature on climate-sensitive mortality and morbidity outcomes in the elderly; a systematic review and meta-analysis of epidemiological evidence. EBioMedicine 2016; 6: 258–68

8. Winklmayr C, Matthies-Wiesler F, Muthers S, et al.: Heat in Germany: Health risks and preventive measures. J Health Monit 2023; 8 (Suppl 4): 3–32.

9. Varghese BM, Hansen A, Nitschke M, et al.: Heatwave and work-related injuries and illnesses in Adelaide, Australia: A case-crossover analysis using the Excess Heat Factor (EHF) as a universal heatwave index. Int Arch Occup Environ Health 2019; 92: 263–72

10. Heat and health in the WHO European Region: Updated evidence for effective prevention. Copenhagen WHO Regional Office for Europe 2021.

11. Copernicus Climate Change Service (C3S) Climate Data Store (CDS): ERA5-Land hourly data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS) 2019.

12. Muñoz-Sabater J, Dutra E, Agustí-Panareda A, et al.: ERA5-Land: A state-of-the-art global reanalysis dataset for land applications. Earth Syst Sci Data 2021;13: 4349–83

13. Ludwig M, Enders D, Basedow F, Walker J, Jacob J: Sampling strategy, characteristics and representativeness of the InGef research database. Public Health 2022; 206: 57–62

14. Gasparrini A, Armstrong B, Kenward MG: Distributed lag non-linear models. Stat Med 2010; 29: 2224–34

15. Gasparrini A, Armstrong B: Reducing and meta-analysing estimates from distributed lag non-linear models. BMC Med Res Methodol 2013; 13: 1

16. Gasparrini A, Armstrong B, Kenward MG: Multivariate meta-analysis for non-linear and other multi-parameter associations. Stat Med 2012; 31: 3821–39

17. Gasparrini A, Leone M: Attributable risk from distributed lag models. BMC Med Res Methodol 2014; 14: 55

18. Adélaïde L, Chanel O, Pascal M: Health effects from heat waves in France: An economic evaluation. Eur J Health Econ 2022; 23: 119–31

19. Steul KS, Latasch L, Jung HG, Heudorf U: Health impact of the heatwave of 2015: Hospital admissions in Frankfurt/Main, Germany. Gesundheitswesen 2018; 80: 353–9

20. Steul K, Jung HG, Heudorf U; Heat-related morbidity: Real-time surveillance via rescue service operations data from the interdisciplinary care capacity proof system (IVENA). Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2019; 62: 589–98

21. Alho AM, Oliveira AP, Viegas S, Nogueira P: Effect of heatwaves on daily hospital admissions in Portugal, 2000–18: An observational study. Lancet Planet Health 2024; 8: e318–e26

22. Karlsson M, Ziebarth NR: Population health effects and health-related costs of extreme temperatures: Comprehensive evidence from Germany. J Environ Econ Manage 2018; 91: 93–117

23. Rameezdeen R, Elmualim A: The impact of heat waves on occurrence and severity of construction accidents. Int J Environ Res Public Health 2017; 14: 70

24. Sheffield PE, Herrera MT, Kinnee EJ, Clougherty JE: Not so little differences: Variation in hot weather risk to young children in New York City. Public Health 2018; 161: 119–26

25. Tsuzuki-Hayakawa K, Tochihara Y, Ohnaka T: Thermoregulation during heat exposure of young children compared to their mothers. Eur J Appl Physiol Occup Physiol 1995; 72: 12–7

26. Agency USEP: Exposure factors handbook: 2011 edition. National Center for Environmental Assessment, Washington, DC; 2011.

27. Xu Z, FitzGerald G, Guo Y, Jalaludin B, Tong S: Assessing heatwave impacts on cause-specific emergency department visits in urban and rural communities of Queensland, Australia. Environ Res 2019; 168: 414–9

28. Lee WS, Kim WS, Lim YH, Hong YC: High temperatures and kidney disease morbidity: A systematic review and meta-analysis. J Prev Med Public Health 2019; 52: 1–13

29. Yardley JE, Stapleton JM, Sigal RJ, Kenny GP: Do heat events pose a greater health risk for individuals with type 2 diabetes? Diabetes Technol Ther 2013; 15: 520–9

30. Kenny GP, Sigal RJ, McGinn R: Body temperature regulation in diabetes. Temperature (Austin) 2016; 3: 119–45

31. Gronlund CJ, Zanobetti A, Schwartz JD, Wellenius GA, O’Neill MS: Heat, heat waves, and hospital admissions among the elderly in the United States, 1992–2006. Environ Health Perspect 2014; 122: 1187–92

32. Turner LR, Barnett AG, Connell D, Tong S: Ambient temperature and cardiorespiratory morbidity: A systematic review and meta-analysis. Epidemiology 2012; 23: 594–606

33. O‘Neill MS, Ebi KL: Temperature extremes and health: Impacts of climate variability and change in the United States. J Occup Environ Med 2009; 51: 13–25

34. Sheffield PE, Landrigan PJ: Global climate change and children‘s health: Threats and strategies for prevention. Environ Health Perspect 2011; 119: 291–8

35. Klenk J, Becker C, Rapp K: Heat-related mortality in residents of nursing homes. Age Ageing 2010; 39: 245–52

36. Sera F, Hashizume M, Honda Y, et al.: Air conditioning and heat-related mortality: A multi-country longitudinal study. Epidemiology 2020; 31: 779–87

37. Herrmann A, Haefeli WE, Lindemann U, Rapp K, Roigk P, Becker C: Epidemiologie und Prävention hitzebedingter Gesundheitsschäden älterer Menschen. Z Gerontol Geriatr 2019; 52: 487–502

38. Baldwin JW, Benmarhnia T, Ebi KL, Jay O, Lutsko NJ, Vanos JK: Humidity’s role in heat-related health outcomes: A heated debate. Environ Health Perspect 2023; 131: 55001