Introduction

The challenge of climate change introduces a notable peril to diverse areas of living, involving health, environment, social and economic concerns1,2. It is crucial to underscore and comprehend that climate change transcends a mere environmental concern and concurrently embodies a substantial challenge for public health3. The impacts of climate change are demonstrably inequitable and disproportionately affect vulnerable populations such as children, pregnant women, the elderly and children4. In particular, extreme heat exposure has been associated with poorer adolescents’ health, including diminished cognitive function5, poorer mental health6, and behavioral problems7,8.

Adolescents, compared to adults, have higher outdoor activity levels9 and spend more time participating in strenuous physical activity10 which make them vulnerable to thermal conditions. While theoretically, we argue that a hot environment affects adolescents’ physical and mental health, only a few studies have discussed this11,12,13 and evidences have been largely derived from correlational studies5,6. Most studies often focus on adults and the elderly14,15, but there are limited studies on adolescents, thereby making them underrepresented in global research on the likely impacts of climate change. This highlights the need for more research into how a hot environment affects adolescents’ health.

One particularly salient consequence of climate change is the rise in temperatures, which leads to increased heat stress or negative mental health outcomes16. There has been significant inquiry into the ways heat stress affects physical health, yet its implications for mental wellness are less explored and underreported. Studies suggest that heat stress may adversely affect mental health, manifesting as heightened anxiety and the emergence of mood disorders17. The interplay between the consequences of heat stress on both physical and mental health underscores the necessity for research focusing on adolescents in urban locales susceptible to heat-related impacts.

In urban settings, heat stress is worsened by the dense population, insufficient green areas, and urban heat island phenomena18,19. Moreover, socio-economic inequities, including disparate access to healthcare and cooling resources, may further exacerbate these risks for urban adolescents20. It is thus imperative to comprehend the vulnerabilities and adaptive capacities of adolescents residing in such environments.

Furthermore, there are limited studies on outdoor heat exposure adaptation practices adopted by young people, i.e., what young people usually do to minimize the impact of outdoor heat exposure, the frequency of performing these adaptation practices, and what the behavioral drivers are. In addition, we observed that prior studies often focus only on either the impact of heat exposure on health or adaptation practices (including the drivers), but not both. However, analysing all of the factors and understanding the nuanced relationships in this heat-health nexus is crucial for designing effective public health interventions.

There is no single framework that could comprehensively analyze the complex interaction or situations behind outdoor heat exposure adaptation practices among adolescents. Thus, we combined two main theories or frameworks in this study. First, we followed and adapted the Social-Ecological Model (SEM) framework to understand the drivers of outdoor heat exposure adaptation practices. We than used the RANAS (Risk, Attitude, Norm, Ability, and Self-Regulation) theory to investigate psychological factors or drivers shaping adaptive practices. The SEM framework incorporates individual, interpersonal, community, policy, and environmental dimensions, whereas RANAS emphasizes psychological aspects, including risk perception, attitude, and self-regulation. Collectively, these frameworks provide a comprehensive perspective to elucidate how adolescents react to heat exposure.

To contextualize an urban setting with high heat risk, we explored a case study of Yogyakarta City, Indonesia. In Yogyakarta, elevated ambient temperatures are further intensified by rapid urban development and limited availability of green areas. These result in pronounced heat stress conditions in Yogyakarta City21. Furthermore, Yogyakarta is one of the favourite tourist destinations in Indonesia with many attractions or tourism sites located outdoors. Adolescents in Yogyakarta often engage in outdoor pursuits, thereby heightening their susceptibility to heat stress. However, there exists a dearth of scholarly inquiry into their perceptions, experiences, and adaptations to extreme heat conditions.

By addressing the existing knowledge gaps, this research aims to generate actionable insights to guide evidence-based policies and interventions. The results will aid in the creation of focused public health strategies, bolstering climate resilience for adolescents in urban environments, especially in low- and middle-income countries.

Results

Characteristics of the respondents

The mean temperature and humidity during the data collection were 32.7°C (SD: 1.8) and 57.2% relative humidity (SD: 8.3), respectively (Table 1). Most of the respondents were women (63.3%), with an average age of 20.4 years (SD: 2.1). 35% and 31% of the respondents had a monthly income between “IDR 500,000 ($31) – 1,500,000 ($94)” and “below "IDR 500,000 ($31)”, respectively. Almost 60% of the respondents had a bachelor’s degree, with 28% having completed their education at the senior high school level.

Table 1 Descriptive statistics and regression results of variables on each natural environment, policy, interpersonal, and individual–contextual aspect on adaptation practices towards heat exposure.
Full size table

The average scores of physical and mental impacts experienced by the respondents due to heat exposure were 2.8 (SD: 1.7, range: 0–10) and 2.1 (SD: 1.5, range: 0–8), respectively. These values indicate that the respondents experienced mild physical (3 symptoms) and mental (2) symptoms due to heat exposure. The most common physical symptoms were excessive sweating (61%), fatigue (56%), and dehydration (52%). The most common mental symptoms were emotional (54%), sleep disorder (51%), and feeling stress (34%). Furthermore, 49.5% of the respondents mentioned that they were at high risk of experiencing health problems when exposed to heat outdoors.

From a maximum of 11 adaptation practices, the average score was 3.9 (SD: 2.3). The most common adaptation practices were using sunblock (62.6%), followed by bringing drinking water when outdoors (62.2%). The least common practices were avoiding caffeinated drinks (10.3%) and using sunglasses (12.8%). The full descriptive statistics of other variables are shown in Table 1.

The top three “high or good” perceptions or psychological factors (shown by the high mean score relative to the possible maximum score in Table 2) related to outdoor heat exposure were: (1) risk—knowledge related to heat exposure (M = 6.9 from a maximum of 8), (2) self-regulation – commitment (M = 4.3 out of 5), and (3) attitude – affective – 2 (comfortable working or being outdoors when the outside air is very hot) (M = 4.1 out of 5). The “low” perception levels were: (1) Norms—descriptive (M = 2.7 out of 5), (2) Attitude—difficulties (M = 2.8 out of 5), and (3) Norms—injunctive (M = 3.3 out of 5).

Table 2 Descriptive statistics of psychological factors and regression results of psychological factors on adaptation practices towards heat exposure.
Full size table

Regression analysis

Tables 1 and 2 show the 9 variables from 4 SEM aspects that were significantly related to adaptation practices: (1) Feeling about the current airflow, (2) feeling about the current sunlight, (3) feeling comfortable about the current air conditions, (4) acquaintances or relatives ever experienced negative impacts from hot weather, (5) gender, (6) education, (7) physical impact of heat exposure, (8) mental impact of heat exposure, and (9) psychological factor – remembering (part of self-regulation). No variable related to the policy aspect was related to adaptation practices.

All the 9 significant variables were then included in the final regression (Table 3). There were 6 significant variables in the final model: (1) Feeling about the current airflow, (2) feeling about the current sunlight, (3) gender, (4) physical impact of heat exposure, (5) mental impact of heat exposure, and (6) psychological factor – remembering (self-regulation). The most influential variables on adaptation practices were the physical impact of heat exposure, followed by the mental impact of heat exposure, and the feeling about the current sunlight.

Table 3 Final regression of 9 selected variables on adaptation practices.
Full size table

Discussion

In this study, we have compiled extensive data to provide an in-depth analysis of outdoor heat exposure to adolescents and their perceptions in urban Indonesia, focusing on Yogyakarta City as a case study. The collated information was related to the health impacts felt by the young people, their adaptation practices and related psychological factors following the RANAS theory, and environmental data. Furthermore, we adapted the concept of the Socio-ecological model in the analysis. This analysis approach provided us with a wider and nuanced understanding of adolescents’ heat exposure practices in urban areas involving diverse and multi-layer factors ranging from individual determinants to environmental factors.

Our study indicated that young people in urban Yogyakarta experienced mild physical and mental symptoms due to heat exposure. The respondents often felt excessive sweating, fatigue, and dehydration as some of the physical impacts experienced, while being emotional, having a sleep disorder, and feeling stressed were the most common mental symptoms. These physical and mental symptoms are often mentioned in other studies11.

Furthermore, we found that if young people experienced physical and mental symptoms, they were more likely to adopt heat adaptation practices. The relationships between how physical and mental symptoms were associated with heat adaptation by young people are complex and multifaceted. Based on the protection motivation theory28, an individual’s intention and decision to engage in health protective behaviors (i.e. adaptation practices towards heat) are influenced by their perceived threat severity and vulnerability. The personal experiences of heat-related health symptoms could modulate threat appraisal and motivate people to perform heat adaptation practices, but on the other hand, heat adaptation practices mitigate the health symptoms that occur, i.e., a bi-directional relationship. Therefore, this reciprocal and interesting relationship warrants further investigation.

To the best of our knowledge, there is no study that specifically investigates heat adaptation practices among adolescents who are outdoors. In our case, young people had low engagement in adaptation practices during outdoor heat exposure, as shown by the average score of 3.9 out of 11. They tend to be unconcerned about the threat of outdoor heat. One possible reason could be due to their lack of knowledge about climate change29. This suggests that there is a need for better education and program to raise young people’s understanding and awareness of the types of heat adaptation practices they can adopt, e.g., using digital tools and interactive methods to engage young people and enhance their knowledge about climate adaptation30.

Almost half of the respondents perceived that they had a high likelihood of experiencing health problems when exposed to heat outdoors. However, this perception of vulnerability (part of risk) was not sufficient to induce adaptation practices (not significant in Table 2). This finding was contrary to another study that reported that higher risk perception was associated with increased adaptation behaviors31,32. Nonetheless, we argue that vulnerability perception is important in triggering heat adaptation practices among youth, but the ‘influence’ is lower than other psychological factors, such as ‘self-regulation – remembering’.

The general knowledge about heat exposure was good (M = 6.9 from a maximum of 8). However, this knowledge was not significantly associated with the adoption of adaptation practices. This highlights the gap between knowledge and practice33. Also, as shown by our regression analysis, other psychological factors besides knowledge were responsible for the youth’s adaptation practices.

The psychological factor ‘remembering” was significantly related to adaptation practices. Based on the mean score of health knowledge (6.9 from 0–8) in our case study, it appears that the youths were already aware of the health aspects related to heat exposure, but they often forgot to engage in adaptation practices. We have not further investigated the reasons behind this. However, when we performed a correlational analysis between all psychological factors, the psychological factor ‘action control’ (whether they always check all the things needed to avoid heat exposure before doing outdoor activities) had the strongest correlation with ‘remembering’. This suggests that if adolescents had with them any material or equipment for the adaptation practices, e.g., hat, sunblock, umbrella, drinking bottles, etc., they were more likely to remember to adopt heat adaptation practices by using these when outdoors. Hence, we recommend that young people have all these items in a place that is easily visible when leaving the room to increase attentional salience and promote the likelihood of performing adaptation practices.

Two psychological factors under norms had lower scores compared to others, i.e., descriptive norm (M = 2.7 from 5) and injunctive norm (M = 3.3 from 5). These indicate that there was no social norm or social pressure related to heat exposure adaptation practice in society, i.e., people perceived that not performing adaptation practice was normal. In fact, social pressure or norms are one of the most powerful drivers of health behavior34,35. While this was not statistically significant, we argue that efforts should be made to improve the current norm related to outdoor heat exposure adaptation practice, e.g., by issuing local policies related to the prevalent weather36.

Two variables related to natural environments were significantly associated with the adaptation: “Feeling about the current airflow” and “Feeling about the current sunlight”. This indicates that perceptions about outdoor situations would induce heat adaptation. We may term this as ‘situational heat adaptation practice’, meaning that the person’s action is influenced by their circumstances37. However, this situational practice should be seen with caution, especially for those who spend a lot of time outdoors. For example, people may go to work or other activities in the morning when the ambient temperature is ‘comfortable’ and prepare nothing for the hot temperature at midday. In essence, we want to caution and highlight that people should be prepared for heat adaptation practice if they want to spend a lot of time outdoors, even when the current or anticipated ambient temperature is not too hot.

We observed that young women adapt more often than young men, a finding that is also reported in other studies12,38. This could be because women have higher risk perceptions towards climate or the environment and are more likely to practice short-term coping mechanisms39,40. In our case, young women tended to have significant (p-value < 0.05) and higher psychological levels than men in almost all RANAS psychological factors. These mean that young women were more ‘mentally’ prepared to engage in heat adaptation when outdoors than young men. Therefore, we should consider potential gender-related differences when formulating new heat adaptation strategies or interventions.

We note some limitations of our study and provide some recommendations for future studies. First, the physical and mental symptoms mentioned by respondents may not be due to the hot environment only, but other factors, e.g., lack of rest time that results in fatigue or sleep disorder, or personal problems resulting in emotional distress. However, we argue that a hot environment exacerbates those symptoms41,42. Second, our cross-sectional study did not capture changes in the youth’s perceptions or heat exposure adaptation over time or changes in outdoor temperatures or seasons. Also, the significant associations do not necessarily imply a causal relationship. A prospective cohort study can reveal these relationships, including the relationship between all multi-level aspects in the SEM, which can either be direct, indirect, moderation, or bi-directional influence. For example, local climate policies may influence someone’s perception through the creation of norms in society, or climate-related knowledge influences the adaptation practice via risk perception. These complex relationships warrant further investigations. Third, our study did not include social aspects as part of the analysis. While some studies have analysed the impact of climate change or hot outdoor environments on adolescents’ physical and mental health, very little is known about the impact of hot outdoor environments on their social well-being, as ‘health’ encompasses physical, mental, and social well-being aspects43. This would provide us with a more comprehensive understanding of the impact of hot outdoor environments on adolescents’ health. Fourth, we did not clarify answers and findings with the respondents, although it was worth knowing how their perceptions affect their daily heat adaptation or obstacles to their adaptation. Thus, a qualitative study may be conducted to better understand the perceptions and adaptation practices of the youth. It is also important to involve the affected urban adolescents as stakeholders in heat or climate adaptation planning44. Lastly, this study was conducted in one location. Local settings, culture, and norms could influence the youth’s heat adaptation practices. Thus, we need more local studies on this topic to enrich our understanding of adolescents’ heat exposure adaptation practices.

Conclusion

Our findings highlight that although adolescents in Yogyakarta are aware of the health risks of outdoor heat exposure, their adaptive practices for heat exposure prevention are still relatively low, especially among young males. Having poor physical and mental experiences related to heat exposure have the strongest associations with outdoor heat adaptation practices. Microclimatic perceptions, especially feelings about the ambient airflow and sunlight, were associated with the adolescents’ decision to perform heat-exposure adaptation practices outdoors. Among all psychological factors, only the self-regulation ‘remembering’ factor was significantly related to adaptation practices, indicating that young people need a reminder to prepare items for adaptation practices before going outdoors. To reduce the health impacts of outdoor heat exposure, targeted education and awareness programs for adolescents are required, emphasizing the importance of preparation and visible reminders. In addition, local policies should promote heat adaptation as a social norm. Given the complexity of behavioral changes in adolescents, further research and local interventions are essential to enhance climate resilience, improve heat adaptation practices, and enrich our scientific knowledge of adolescents’ heat exposure adaptation practices.

Methods

Data collection

We conducted a cross-sectional study in Yogyakarta City, Indonesia, during October–November 2024. Yogyakarta City (hereafter called “Yogya”) is the capital of the Special Region of Yogyakarta Province with a population of 414,000. The estimated proportion of adolescents aged 15 to 24 years represents 15% of the total population22.

Based on the statistical assumption, we interviewed 439 adolescents who met two main criteria: aged between 19 and 24 years, living in Yogyakarta and the surrounding areas (but still in Yogyakarta Province), and and was doing outdoor activities at that time. We conducted the interview outdoors. The interview was conducted by 6 trained enumerators using KoBoToolbox v2021.2.423 on a smartphone. We divided the data collection into three times of the day: morning, mid-day, and afternoon, with almost equal respondents in those three times of the day.

Participation was voluntary and all participants gave their informed consent prior to the interview. This study was approved by the Medical and Health Research Ethics Committee Faculty of the Faculty of Medicine, Public Health, and Nursing, Gadjah Mada University (Number KE/FK/1496/EC/2024). The research procedure was in accordance with the Declaration of Helsinki.

The interview’s questionnaire consisted: (1) environmental conditions (interview time, ambient temperature, and humidity), (2) respondent’s characteristics (gender, age, education, etc.), (3) ecological conditions (green open space facilities, access to health care facilities, environmental policy, etc.), (4) microclimatic perception and preference (regarding ambient temperature, humidity, airflow, sunbeam, and comfort), (5) impact of heat exposure on physical and mental health, (6) the respondent’s mitigation and adaptation practices, and (7) the related psychological factors. We used the Likert scale for most of the psychological-related questions. Items, statements, or questions that were used in the statistical analysis can be found in Tables 1 and 2.

For the questionnaire, we adapted the Socio-Ecological Model (SEM) framework (Fig. 1)24 in combination with the RANAS theory at the individual level25. The SEM framework argues that individual behavior is influenced by the interplay of multi-level aspects, i.e., individual, interpersonal, community, policy, and natural environment. Items, statements, or questions from the surveys were then grouped into those five aspects. Examples of individual factors are age, gender, education, and the impacts of heat exposure experienced by the person. Examples of interpersonal factors are the negative impacts related to heat exposure that is experienced by the respondent’s relatives or peers, the frequency of talking about adaptation practices toward heat exposure with peers, and the frequency of hearing about the impact of heat exposure on health. An example of community factors are the social norms. To measure the policy factors, we collected people’s perceptions about government policies to create a cooler environment and whether the government provided information about the impact of exposure to a hot environment. For the natural environment, we included questions on microclimate perceptions, temperature, and humidity.

Fig. 1
Fig. 1
Full size image

An adapted socio-ecological model of adaptation towards indoor heat exposure. The values indicate the number of variables used in the statistical analysis. The ‘community’ aspect-related variables are analysed under the individual–psychological determinants.

We measured ambient temperature and humidity, each time before an interview was conducted with each respondent, using the Air quality monitor HT9600 particle counter, Digital humidity & temperature meter GM1360A, and AZ 87,786 WBGT logger.

Data analysis

Data from KoboToolbox was transferred to Microsoft Excel for cleaning, and afterwards to IBM SPSS version 27 for statistical analysis. All data are presented as mean (M) and standard deviation (SD).

For the impacts of heat exposure, we enquired about the physical and mental impacts experienced by the person. There were 10 and 8 items related to physical and mental impacts, respectively. The questions that we asked are: “Have you ever experienced symptoms related to heat exposure?” and “How do you feel when the weather is very hot?”. Examples of physical impacts are fatigue, dehydration, excessive sweating, heat stroke, etc. Examples of mental impacts are stress, sleep disturbances, confusion, anxiety, etc. We summed up all items, and the level of physical and mental impacts experienced by the respondent was determined from the number of items mentioned (i.e. more items relate to a higher level of impact).

For the microclimate perceptions, we enquired about their perceptions of the present temperature, humidity, wind flow, and overall comfort with the environment situation. We used 5–7 Likert scale answer options.

We analysed the norms-related variables (‘community’ aspect or layer in the SEM model, Fig. 1) at the individual aspect because these norms-related variables are included in the RANAS theory at the individual aspect. According to the RANAS theory, the behavioral determinants can be divided into contextual (age, income, education, etc.) and psychological factors (perceptions of risk, attitude, norms, etc.). We adopted the approach of existing RANAS studies that analysed first the contextual factors, and significant contextual factors were included in the final regression analysis using psychological factors26,27.

For the dependent variable in the regression analysis, we used a question about the adaptation practice that the respondent did at the time of the interview. In this study, we define adaptation as an action taken to manage and minimize the impact of heat exposure. There were 11 items related to heat adaptation practices, e.g., Use head protection, sunglasses, sunscreen, wear appropriate clothing (light, loose, and light-colored), bring drinking water when outdoors, consume sufficient drinking water to prevent dehydration, limit activities under direct sunlight, etc. We summed up all actions that were mentioned by the respondent, and the engagement level in adaptation practice towards heat exposure was determined from the number of actions mentioned by the respondent (i.e. more actions related to higher engagement in adaptation practices).

We performed the stepwise multivariate regression analysis using all variables. First, we entered all variables in each aspect and regressed them on adaptation practices, as the dependent variable. For example, we entered 11 variables in the natural environment aspect as independent variables and regressed them on adaptation practices. Significant variables (p-value < 0.05) in this regression were then included in the final regression. We repeated the same procedures for other aspects of SEM. The procedure is shown in Fig. 2.

Fig. 2
Fig. 2
Full size image

Procedure of the analysis. Parallelogram boxes are the independent variables, and an oval box is the dependent variable.