Abstract
Rising global temperatures can lead to heat waves, which in turn can pose health risks to the community. However, a notable gap remains in highlighting the primary contributing factors that amplify heat-health risk among vulnerable populations. This study aims to evaluate the precedence of heat stress contributing factors in urban and rural vulnerable populations living in hot and humid tropical regions. A comparative cross-sectional study was conducted, involving 108 respondents from urban and rural areas in Klang Valley, Malaysia, using a face-to-face interview and a validated questionnaire. Data was analyzed using the principal component analysis, categorizing factors into exposure, sensitivity, and adaptive capacity indicators. In urban areas, five principal components (PCs) explained 64.3% of variability, with primary factors being sensitivity (health morbidity, medicine intake, increased age), adaptive capacity (outdoor occupation type, lack of ceiling, longer residency duration), and exposure (lower ceiling height, increased building age). In rural, five PCs explained 71.5% of variability, with primary factors being exposure (lack of ceiling, high thermal conductivity roof material, increased building age, shorter residency duration), sensitivity (health morbidity, medicine intake, increased age), and adaptive capacity (female, non-smoking, higher BMI). The order of heat-health vulnerability indicators was sensitivity > adaptive capacity > exposure for urban areas, and exposure > sensitivity > adaptive capacity for rural areas. This study demonstrated a different pattern of leading contributors to heat stress between urban and rural vulnerable populations.
Keywords: Heat stress, Contributing factors, Indoor environment, Vulnerable populations, Principal component analysis (PCA)
Subject terms: Climate sciences, Risk factors
Introduction
Global warming is one of the adverse impacts of climate change, defined as an increment in average global temperatures caused by an increase in the greenhouse effect, which is contributed by increased greenhouse gas emissions1. The effects of anthropogenic activities, including industrial processes, agriculture works, waste disposal, and the use of fossil fuel or emission of pollutants, have resulted in elevated surface temperatures, increased extreme weather occurrence, rising sea levels, and fluctuating precipitation patterns. These conditions profoundly impact seven key sectors in Malaysia: public health, agriculture, water resources, biodiversity, forestry, energy, and coastal and marine resources2,3. The continuous increment of global temperature trends has the potential to give rise to extreme heat events, commonly known as heat waves4.
As defined by the Malaysian Meteorology Department5, a daily maximum temperature exceeding 37 °C for three consecutive days is classified as a heat wave. Based on the history of extreme heat occurrence since 2000, the Klang Valley experienced heat waves in 2019 and 20216. Heat waves are prolonged high temperatures that cause significant thermal stress and commonly occur in urban areas with a synergistic effect of urban heat islands (UHI)7,8. Due to the cumulative effect of UHI occurrence, heat waves are more severe in cities than in rural areas7. Greenhouse emissions may propagate heat waves, and the existing UHI may worsen during the heat wave period, increasing the population's heat stress risk8. It is found that the rising surrounding temperature is influenced by higher population density and intensified land use, which are occupied by high-rise and multi-story buildings9.
However, heat waves also adversely impact rural communities, particularly on their health. Previous hospital admission and visit data studies found that rural communities commonly recorded higher hospitalization rates during heat waves than urban communities10 and the most affected individuals are vulnerable groups such as older people and people with low income11. A study of extreme temperature trends across different areas in Klang Valley (2006–2016) found inconsistent results on the annual mean of maximum temperature and the highest value of daily maximum temperature for urban and rural areas12. Some rural areas in Klang Valley also reported the same temperature value as urban areas for the annual mean of maximum temperature, which ranges between 31.3 and 31.6 °C, and the highest daily maximum temperature in rural areas ranges between 35.8 and 36.3 °C12. This result shows that rural areas in Klang Valley also have the potential to experience high temperatures that exceed 37 °C, where heat waves can occur and affect the rural population, especially vulnerable groups.
The direct heat exposure from the environment commonly measured by the ambient and radiant temperature, relative humidity, and air velocity were significantly related to heat stress13,14. However, the intensity of indoor heat exposure may vary depending on other contributing factors, such as building materials, building density, and green spaces14,15. Also, the sensitivity of individuals exposed to heat depends on several factors such as age, gender, income, pre-existing diseases, educational level, exercise, and regular smoking14–18. Since the vulnerability of heat exposure is also characterized by physical, social, economic, and environmental factors, it could vary significantly within a community and over time. In other words, understanding heat health vulnerability requires more than analysing the direct impacts of a hazard; the concerns about the broader individual, environmental, and socioeconomic conditions that limit people and communities from coping with the effects of a hazard shall also be highlighted.
To date, there is an increasing amount of evidence indicating that both individual and environmental factors contribute to the variability in heat stress experienced by individuals19–22, but limited studies were found to emphasize the primary contributing factors influence the heat-health vulnerability, especially among vulnerable population living in urban and rural areas. The current existing heat health mitigation plan prioritizes vulnerable groups but does not sufficiently address area-based concerns. More attention has been given to urban populations subjected to UHI effects12,23, with limited investigations into the unique challenges faced by rural populations in managing and adapting to heat stress. Thus, this study aims to determine the precedence of heat stress contributing factors among vulnerable urban and rural populations. This can enhance the understanding of how specific community characteristics in urban and rural settings may influence heat stress differently. Additionally, it highlights the priority on which factors should be given concern in effectively addressing heat-health adaptation and mitigation responses.
Methods
Study duration
The study was conducted between May to September 2022, coinciding with the Southwest monsoon period in Malaysia. The Southwest monsoon in Malaysia typically spans from May to September and is characterized by low precipitation, less cloud cover, high outgoing long-wave radiation, often featuring a dry period5. This study focused on this period due to its reputation for elevated temperatures, as previously classified as the hottest months of the year (April, May, and June) in prior research6. It is assumed that the temperatures during the Southwest monsoon are indicative of heatwaves conditions. This study involves three phases: Phase I: the pilot study; Phase II: screening; and Phase III: data collection for the main study. A pilot study involving 20 randomly selected participants was conducted in June 2022 to obtain the prevalence of heat stress symptoms, followed by the screening phase in July 2022 and data collection for the main study between July and September 2022.
Study participants
A comparative cross-sectional study design was used, and an equal number of respondents from urban and rural areas in the Klang Valley, Malaysia, were recruited using stratified random sampling. The calculation of the sample size using comparing two means formula24 was based on the prevalence of heat stress symptoms reported in the pilot study for both urban and rural areas. A total of 108 respondents were recruited, achieving a 96% response rate. The study participants were selected through a screening checklist based on the inclusive criteria. Those who were classified as vulnerable in any groups (aged 60 years old and above, people with health morbidity, and people with low income) and experiencing any heat stress symptoms for the preceding months (April to June) while residing in their residential areas were included in this study. Conversely, pregnant women and individuals under 13 years old were excluded from this study due to hormonal, blood volume and circulation changes in pregnant women25 as well as the underdevelopment of thermoregulatory mechanisms in children26,27, which have the potential to confound the study results.
Research tools
An adapted questionnaire from the Guidelines on Heat Stress Management at Workplace13 was used to assess heat stress symptoms during Phases I and II. In Phase III, a self-administered questionnaire was used to obtain the participants' sociodemographic background, health status, lifestyle information, and residential information for the target population. The questionnaire underwent content validation, achieving a Scale Content Validity Index (S-CVI) value of 1.0, which is considered acceptable28. A pre-test was conducted during the pilot study to assess test–retest reliability. The intraclass correlation coefficient for the continuous data ranged from 0.78 and 0.99, indicating good reliability (0.75–0.9), and excellent reliability (> 0.9)29. Cohen’s Kappa coefficient for nominal data ranged from 0.89 to 1.00, indicating almost perfect agreement30. For lifestyle information, we adapted International Physical Activity Questionnaire31, and physical activity levels were classified using metabolic equivalent of task (MET) value, obtained from MET score calculation (combination of physical intensity (walking, moderate-intensity, and vigorous-intensity) and hours spend on physical activity by the respondents)32. The relevant questions for each section are provided in Table 1.
Table 1.
Variables included in the PCA analysis.
| Type of factor | Factors | Indicators (for + loadings value) |
|---|---|---|
| Sociodemographic factors | Age | Sensitivity |
| Gender (Female) | Adaptive capacity | |
| Body mass index (BMI) | Adaptive capacity | |
| Household income | Adaptive capacity | |
| Educational level | Adaptive capacity | |
| Health factors | Health morbidity (Yes) | Sensitivity |
| Medicine intake (Yes) | Sensitivity | |
| Lifestyle factor | Smoking (Yes) | Adaptive capacity |
| Daily water intake | Adaptive capacity | |
| Physical activity level | Adaptive capacity | |
| Occupation type (outdoor work) | Adaptive capacity | |
| Residential factors | Building age | Adaptive capacity |
| Building type (multi-story building) | Exposure | |
| Building density | Exposure | |
| Roof material (high thermal conductivity) | Exposure | |
| Wall material (high thermal conductivity) | Exposure | |
| Building size | Adaptive capacity | |
| Green plot ratio | Adaptive capacity | |
| Ceiling availability (Yes) | Adaptive capacity | |
| Ceiling/ roof height | Adaptive capacity | |
| Residency duration | Adaptive capacity |
Ethics approval
This study obtained ethical approval from Ethics Committee for Research Involving Human Subjects, Universiti Putra Malaysia (Reference No: JKEUPM-2022-222). All data collection methods were performed in accordance with the respective ethical guidelines. Written informed consent was obtained from participants and parents or guardians of minors before proceeding with the data collection.
Statistical analysis
The data obtained from the respondents were analyzed using Statistical Package for the Social Sciences (SPSS) version 25. A descriptive analysis was used to get the average and frequency of sociodemographic background, health status, lifestyle information, and residential information. A principal component analysis (PCA) was conducted to determine the factors contributing to heat stress. Data preprocessing, which included the encoding categorical variables and the standardizing numerical variables, was applied before the PCA analysis33. PCA for mixed data was employed, leveraging its powerful technique in interpreting the status of variables across different data types34. Only datasets with a Kaiser–Meyer–Olkin measure (KMO) value of > 0.535 and Barlett’s test of sphericity (BTS) yielding a result of 0.000 (p < 0.001) were selected for interpretation33.
PCA was used for the data reduction by extracting a limited number of principal components (PCs), and varimax rotation was applied to maximize the variances of factor loadings across variables of each factor to enhance the interpretability of the result36. The PC variables with a factor loading of 0.4 and higher were selected as significantly loaded items as recommended for the rotated factor pattern36. Principal components (PCs) with eigenvalue > 1.0 were extracted for the result. A cumulative variance of at least 60% is considered acceptable35. Additionally, PCs with a percentage of more than 10% variance were highlighted as the main contributing factors of heat stress and were further categorized into three heat health vulnerability indicators, which are sensitivity, exposure and adaptability based on the variables' criteria37. Table 1 shows the variables included in the PCA analysis.
Results
Sociodemographic background, health status, lifestyle information and residential information
Table 2 shows the sociodemographic background of the study population from urban and rural areas. A total of 108 Malaysians aged 13 years old and above were recruited in this study; 54 participants were from urban areas, and 54 were from rural areas. In urban areas, the average age of the study population is 53 ± 11.4 years old, with age range of 24 to 74 years old. Most of the respondents from urban areas are female (85.2%). The average body mass index (BMI) is 28.9 ± 6.99 kg/m2, classified as overweight by the Ministry of Health Malaysia38. The highest education level among the study population was recorded for secondary school (68.5%), followed by no formal education (13.0%) and primary school (11.1%). Most respondents are involved in indoor-type of occupations (92.6%) such as cleaning sectors, working from home, office clerk, lorry driver, and housewife. Notably, 90.7% of the respondents were from low-income or B40 groups (which represent the bottom 40% of income in Malaysia, according to the Department of Statistics Malaysia39.
Table 2.
Sociodemographic background of the study population (N = 108).
| Sociodemographic characteristics | Urban (n = 54) | Rural (n = 54) |
|---|---|---|
| Mean (SD) | ||
| Continuous variables | ||
| Age (years) | 52.8 (11.44) | 49.4 (16.1) |
| Body mass index, BMI (kg/m2) | 28.9 (6.99) | 27.2 (6.42) |
| Duration of residency (years) | 16.6 (6.34) | 30.4 (6.52) |
| Categorical variables | Frequency, f (%) | |
| Age groups | ||
| < 60 years old | 39 (72.2) | 38 (70.4) |
| ≥ 60 years old | 15 (27.8) | 16 (29.6) |
| Gender | ||
| Male | 8 (14.8) | 18 (33.3) |
| Female | 46 (85.2) | 36 (66.7) |
| Educational level | ||
| No formal education | 7 (13.0) | 1 (1.9) |
| Primary school | 6 (11.1) | 16 (29.6) |
| Secondary school | 37 (68.5) | 27 (50.0) |
| Tertiary education (diploma/undergraduate/graduate) | 4 (7.4) | 10 (18.5) |
| Household income (Ringgit Malaysia, RM) | ||
| B40 (Below RM4851) | 49 (90.7) | 48 (88.9) |
| M40 (RM4851-RM10,970) | 5 (9.3) | 6 (11.1) |
| Occupation type | ||
| Indoor work | 50 (92.6) | 47 (87.0) |
| Outdoor work | 4 (7.4) | 7 (13.0) |
For rural areas, the average age of the study population is 49 ± 16.1 years old, with an age range of 13 to 76 years old. The majority of participants from rural areas are female (66.7%). The average body mass index (BMI) of respondents from rural areas is 27.2 ± 6.42 kg/m2, which also falls in the overweight category. Most of the respondents from rural areas had the highest education level of secondary school (50.0%), followed by primary school (29.6%), and tertiary education (18.5%). Most of them are involved in indoor-type occupations (87.0%) compared to outdoor-type (13.0%), where the typical outdoor jobs among them are agriculture-based, such as farmer, gardener, and landscaper. It was recorded that 88.9% of the respondents were categorized as low-income or B40 groups.
Table 3 shows the health status and lifestyle information of the study population. Based on the result, 46.3% of the urban area respondents and 53.7% of the rural areas respondents have health morbidities such as hypertension, diabetes, obesity, respiratory diseases, cardiovascular diseases, kidney diseases, and skin diseases. More rural respondents (48.1%) were taking medicine regularly as prescribed by medical practitioners compared to the urban respondents (38.9%). The average daily water intake for urban and rural respondents was within the recommended amount by the Ministry of Health Malaysia40, ranging between 1.5 and 2.0 L per day. The calculated MET value categorizes most urban (53.7%) and rural (46.3%) respondents being at a moderate level of physical activity. Only 3.7% of urban and 16.7% of rural respondents smoke. None of the respondents is regular alcohol drinkers.
Table 3.
Health status and lifestyle information of the study population (N = 108).
| Health status and lifestyle variables | Urban (n = 54) | Rural (n = 54) |
|---|---|---|
| Mean (SD) | ||
| Continuous variables | ||
| Daily water intake (litre) | 1.7 (0.8) | 1.6 (0.7) |
| Categorical variables | Frequency, f (%) | |
| Health morbidity | ||
| Yes | 25 (46.3) | 29 (53.7) |
| No | 29 (53.7) | 25 (46.3) |
| Medicine intake (prescribed by a medical practitioner) | ||
| Yes | 21 (38.9) | 26 (48.1) |
| No | 33 (61.1) | 28 (51.9) |
| Physical activity level (based on MET value) | ||
| Low | 10 (18.5) | 8 (14.8) |
| Moderate | 29 (53.7) | 25 (46.3) |
| High | 15 (27.8) | 21 (38.9) |
| Smoking | ||
| Yes | 2 (3.7) | 9 (16.7) |
| No | 52 (96.3) | 45 (83.3) |
| Regular alcohol intake (at least once daily) | ||
| Yes | 0 (0.0) | 0 (0.0) |
| No | 54 (100.0) | 54 (100.0) |
Table 4 shows the residential information of the study population. Most urban respondents (83.3%) live in multi-story buildings, staying for almost 17 ± 6.34 years. In contrast, most rural respondents (61.1%) live in landed houses, staying for nearly 30 ± 6.52 years. For the residential information, the average building age (year) in urban areas is 26.41 ± 10.18 and 29.70 ± 10.29 for rural. The building size (m2) is relatively lower in urban areas (80.67 ± 25.37) compared to rural areas (165.82 ± 73.6). Most urban (88.9%) and rural (61.1%) houses equipped with ceilings, with an average of 3.51 ± 0.63 m and 4.26 ± 1.38 m in height, respectively. The wall material for urban residential buildings is primarily concrete (83.3%), whereas in rural areas primarily bricks with cement plaster (63.0%). Both urban and rural houses mainly used concrete as roof material. Building density and green plot ratios recorded in urban and rural areas showed that urban areas are denser with buildings and lower with green spaces than rural areas.
Table 4.
Residential information of the study population (N = 108).
| Residential Information | Urban (n = 54) | Rural (n = 54) |
|---|---|---|
| Mean (SD) | ||
| Continuous variables | ||
| Building age (year) | 26.41 (10.18) | 29.70 (10.29) |
| Building size (m2) | 80.67 (25.37) | 165.82 (73.6) |
| Ceiling/ roof height (m) | 3.51 (0.63) | 4.26 (1.38) |
| Building density (%) | 23.15 (3.14) | 16.87 (5.22) |
| Green plot ratio (%) | 18.73 (8.16) | 47.66 (14.20) |
| Residency duration (year) | 16.6 (6.34) | 30.4 (6.52) |
| Categorical variables | Frequency, f (%) | |
| Building type | ||
| Landed house | 9 (16.7) | 33 (61.1) |
| Multi-story building | 45 (83.3) | 21 (38.9) |
| Wall material | ||
| Brick with plaster | 9 (16.7) | 34 (63.0) |
| Concrete | 45 (83.3) | 20 (37.0) |
| Roof material | ||
| Ceramic/clay roof tiles | 2 (3.7) | 14 (25.9) |
| Concrete roof tiles/concrete slab | 49 (90.7) | 22 (40.7) |
| Zinc | 3 (5.6) | 18 (33.3) |
| Ceiling availability | ||
| Yes | 48 (88.9) | 33 (61.1) |
| No | 6 (11.1) | 21 (38.9) |
Green plot ratio (%) = percentage of green spaces in 16000 m2 land area.
Building density (%) = percentage of building density in 16000 m2 land area.
Heat stress contributing factors
Table 5 shows the result of PCA for the urban vulnerable population. Based on PCA analysis for urban areas, 14 factors were identified as contributing factors with significant loadings value (≥ 0.4), further grouped into five principal components (PCs). The five PCs with eigenvalue > 1.0 explain 64.3% of the variability from the original contributing factors. The components comprised PC1 (health morbidity [yes], medicine intake [yes], and increased age), PC2 (occupation type [outdoor], ceiling availability [no], and longer residency duration), PC3 (lower ceiling height and increased building age), PC4 (lower BMI, higher educational level and gender [male]), and PC5 (smoking [yes], lower daily water intake, and lower physical activity level).
Table 5.
Principal component analysis (PCA) for urban vulnerable population (n = 54).
| Components | |||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | |
| Eigenvalue | 2.575 | 2.112 | 1.606 | 1.372 | 1.34 |
| % of Variances | 18.396 | 15.087 | 11.469 | 9.801 | 9.572 |
| Cumulative % | 18.396 | 33.483 | 44.951 | 54.752 | 64.324 |
| Rotated factor pattern: Varimax rotation method | |||||
| Variables | |||||
| Health morbidity | 0.855 | −0.047 | 0.274 | −0.002 | 0.025 |
| Medicine intake | 0.848 | 0.002 | 0.26 | 0.054 | 0.062 |
| Age | 0.675 | 0.133 | −0.258 | −0.152 | 0.001 |
| Occupation type | 0.029 | 0.828 | −0.081 | 0.118 | −0.097 |
| Ceiling availability | 0.111 | −0.742 | −0.057 | 0.325 | −0.02 |
| Residency duration | 0.198 | 0.683 | 0.268 | 0.183 | −0.12 |
| Ceiling height | −0.253 | −0.017 | −0.749 | −0.132 | 0.332 |
| Building age | 0.018 | 0.105 | 0.712 | −0.219 | 0.211 |
| Body mass index (BMI) | 0.021 | 0.068 | 0.058 | −0.757 | 0.146 |
| Educational level | −0.459 | 0.041 | 0.381 | 0.522 | 0.07 |
| Gender | 0.066 | -0.407 | 0.287 | −0.516 | −0.174 |
| Smoking | 0.03 | -0.015 | −0.045 | 0.305 | 0.816 |
| Daily water intake (liter) | −0.188 | 0.072 | −0.208 | 0.284 | −0.594 |
| Physical activity level | 0.125 | 0.107 | 0.184 | 0.316 | −0.485 |
Variable is significantly loaded at ≥ 0.4.
Extraction Method: Principal Component Analysis (Eigenvalue > 1.0).
Rotation Method: Varimax with Kaiser Normalization.
KMO value: 0.504, Bartlett’s test: p < 0.001.
Table 6 shows the result of PCA for the rural vulnerable population. Based on PCA analysis for rural areas, 15 factors were identified as contributing factors with significant loading values ≥ 0.4), further grouped into five PCs. The five PCs with eigenvalue > 1.0 explain 71.5% of the variability from the original contributing factors. The components comprised PC1 (ceiling availability [no], higher thermal conductivity roof material, increased building age, shorter residency duration), PC2 (health morbidity [yes], medicine intake [yes], and increased age), PC3 (gender [female], higher BMI, and smoking [no]), PC4 (increased household income, higher educational level, lower daily water intake), and PC5 (higher physical activity level and occupation type [outdoor]).
Table 6.
Principal component analysis (PCA) for rural vulnerable population (n = 54).
| Components | |||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | |
| Eigenvalue | 3.839 | 2.538 | 1.966 | 1.318 | 1.063 |
| % of Variances | 25.594 | 16.922 | 13.105 | 8.789 | 7.085 |
| Cumulative % | 25.594 | 42.516 | 55.621 | 64.41 | 71.494 |
| Rotated factor pattern: Varimax rotation method | |||||
| Variables | |||||
| Ceiling availability | −0.942 | −0.138 | 0.046 | −0.055 | −0.041 |
| Roof material | 0.936 | −0.103 | 0.088 | −0.072 | −0.056 |
| Building age | 0.915 | 0.162 | −0.155 | 0.147 | −0.003 |
| Residency duration (years) | −0.626 | 0.001 | 0.157 | −0.064 | −0.177 |
| Health morbidity | 0.062 | 0.928 | 0.003 | −0.046 | −0.103 |
| Medicine intake | 0.173 | 0.925 | 0.163 | −0.006 | −0.07 |
| Age | 0.469 | 0.526 | 0.088 | −0.131 | 0.223 |
| Gender | −0.065 | 0.297 | 0.825 | 0.141 | −0.145 |
| Body mass index (BMI) | 0.167 | 0.042 | 0.728 | −0.14 | −0.043 |
| Smoking | 0.181 | 0.047 | −0.627 | −0.329 | 0.04 |
| Household income | 0.032 | −0.003 | 0.135 | 0.758 | 0.076 |
| Educational level | 0.086 | −0.244 | 0.158 | 0.757 | −0.279 |
| Daily water intake (litre) | −0.083 | −0.129 | 0.441 | −0.567 | −0.069 |
| Physical activity level | −0.187 | −0.046 | −0.009 | 0.072 | 0.868 |
| Occupation type | 0.25 | −0.075 | −0.331 | -0.211 | 0.587 |
Variable is significantly loaded at ≥ 0.4.
Extraction Method: Principal Component Analysis (Eigenvalue > 1.0).
Rotation Method: Varimax with Kaiser Normalization.
KMO value: 0.648, Bartlett’s test: p < 0.001.
Urban and rural areas have three PCs with more than 10% variances, highlighted as the primary contributing factors to heat stress. For urban, PC1 (health morbidity, medicine intake, and increased age) can be classified as sensitivity, PC2 (outdoor occupation type, lack of ceiling, and longer residency duration) as adaptive capacity, and PC3 (lower ceiling height and increased building age) as exposure. For rural areas, PC1 (lack of ceiling, higher thermal conductivity roof material, increased building age, and shorter residency duration) can be classified as exposure, PC2 (health morbidity, medicine intake, and increased age) as sensitivity, and PC3 (female, non-smoking, and increased BMI) as adaptive capacity. To summarize, the pattern of heat-health vulnerability indicators in urban areas in decreasing order according to the percentage of variances was sensitivity > adaptive capacity > exposure. In contrast, heat-health vulnerability indicators in rural areas were exposure > sensitivity > adaptive capacity.
Discussion
Based on the results, different patterns of heat stress-contributing factors were discovered between urban and rural areas in this study. This finding corresponded to a previous study that agreed heat stress-contributing factors varied in different areas41. A comparative study on the risk factors of heat related illnesses between urban and rural areas found that the urban population is influenced by low education levels, poverty, living in old building structures, and mobile homes with poor insulation. In contrast, risk factors in rural areas include elderly, outdoor workers in agricultural sectors, mobile homes with poor insulation, and developed land42. Another study indicated that urban populations are more susceptible to heat stress due to high heat-prone areas, while rural populations influenced by poor health status, poverty, and challenges in accessing healthcare related to geographic and finances43. While other studies classified heat stress contributing factors into several groups, commonly individual, environmental, and occupational44,45, our study categorizes these factors into three heat health vulnerability indicators, which are sensitivity, exposure, and adaptive capacity to enhance the appropriate mitigation measures during extreme heat events.
This study revealed that sensitivity is the most impactful indicator influencing heat-health vulnerability in urban areas (PC1), while it ranks as a second most important indicator in rural areas (PC2). Both areas highlighted similar factors, including the presence of health morbidity, medicine intake, and older age. Individuals with health morbidities such as diabetes, obesity, hypertension, respiratory disease, and cardiovascular disease have physiological deficiencies for acclimatization46, which potentially influence the relationship between heat exposure and adverse health impacts47. Additionally, medication consumption can interfere with thermoregulation, as anticholinergics can lead to dehydration48. Increased age has been linked with reduced sweat output, which is associated with a decrease in epidermal blood flow during heat exposure, thereby reducing the body’s ability for thermoregulation48. Previous studies have also agreed that poor health status and elderly are factors related to sensitivity, increasing heat-related health issues among urban populations49–51, and rural populations52,53.
Exposure was found to be the most impactful indicator in rural areas (PC1), while it falls in third priority in urban areas (PC3). Most of the factors listed in PCs for both areas are related to residential characteristics. However, increased building age is found to be a similar factor in both areas. Other highlighted factors in rural areas include higher thermal conductivity roof material, absence of ceiling installation, and reduced residency duration, while lower ceiling height is highlighted as exposure factor in urban areas. Other studies have also outlined residential characteristics as exposure indicator in urban53, and rural areas54. Most of the previous studies highlighted direct sources of heat such as land surface temperature50,54, heatwave occurrence55, and high ambient temperature56 as exposure indicator. Nonetheless, some studies have also classified contributing factors such as housing characteristics as exposure indicator in urban53 and rural areas52. Previous studies have agreed that the outdoor-indoor temperature varies depending on the building characteristics57,58.
Adaptive capacity is another indicator ranked as the second most impactful indicator in urban areas (PC2) and third (PC3) in rural areas. However, distinct factors were observed in both PCs. Factors influencing heat stress in urban areas consist of outdoor occupation type, absence of ceiling installation, and longer residency duration, all of which may expose individuals to higher heat exposure. However, long-term exposure to hot conditions may lead to acclimatization, contributing to a better adaptive mechanism59,60. In contrast, rural areas highlighted gender (female), higher BMI, and non-smoking as components of adaptive capacity. Females may tolerate heat more efficiently than males due to a broader range of their resting core temperature, although they have a slower sweating response compared to males61. Although it is common for higher BMI to increase insulation and heat retention, it may also serve as a reservoir for fluid and electrolytes, aiding in thermoregulation and maintaining hydration status during heat exposure62. While other studies have outlined residential and environmental-related factors such as building type63 and green spaces or vegetation index64,65 as components of adaptive capacity, our study proposes acclimatization-induced factors (urban) and individual factors (rural) as components of adaptive capacity.
Our study addresses previous research gaps in several ways. Firstly, this study emphasizes comparing the patterns of heat stress contributing factors between urban and rural areas. Secondly, our study focuses on vulnerable populations to enhance understanding of heat health vulnerability, as these groups are often linked to high morbidity and mortality rates. Although prior studies have provided information on heat stress contributing factors, limited emphasis has been placed on highlighting the precedence of these factors. Furthermore, existing mitigation plans lack area-based concerns. Therefore, our findings are essential for providing baseline information to address appropriate heat mitigation measures based on priority and specific areas or populations.
However, it is important to address the study’s limitations. While acknowledging that a heatwave is defined by prolonged abnormally high temperatures, this study does not equate temperatures recorded during the Southwest monsoon with heat waves. Instead, we assumed that the sampling periods could reflect heat wave conditions based on the climate data mentioned in the previous study, rather than conducting heat exposure monitoring to confirm the existence of extreme heat or heat wave occurrence. Future research endeavors could explore the specific quantification of heatwave events during the Southwest monsoon period to further refine the understanding of heat stress dynamics in this region and enhance the validity of these findings. Additionally, it is worth noting that data collection was primarily conducted on weekdays, which may have limited the randomization of respondents, particularly concerning working adults and young adults attending school.
Conclusion
This study demonstrates different patterns of primary heat stress contributors among vulnerable populations in urban and rural areas. It sheds light on the unique challenges faced by both urban and rural vulnerable populations. The findings not only enhance our understanding of how specific community characteristics in urban and rural settings may influence heat stress differently, but also highlight the factors that should be prioritized to effectively address appropriate heat health adaptation and mitigation responses, reducing the adverse impacts of excessive heat exposure on vulnerable populations.
Acknowledgements
This work was funded by the Higher Education, Malaysia (MOHE) under the Fundamental Research Grant Scheme (FRGS/1/2020/SKK06/UPM/02/1).
Author contributions
VH initiated the project. SNM and NSAMS involved in data collection and data analysis. SNM prepared the manuscript, supervised and edited by FLL, VH, KK, and AMA. All authors reviewed the manuscript.
Data availability
The datasets used and/or analyzed in this study available from the corresponding author on reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
The original online version of this Article was revised: The original version of this Article contained an error in the Acknowledgements section. It now reads: "This work was funded by the Higher Education, Malaysia (MOHE) under the Fundamental Research Grant Scheme (FRGS/1/2020/SKK06/UPM/02/1)".
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Change history
10/14/2024
A Correction to this paper has been published: 10.1038/s41598-024-75132-7
References
- 1.Sivaramanan, S. Global Warming and Climate Change Causes, Impacts, and Mitigation (Central Environmental Authority, Sri Lanka, 2015). 10.13140/RG.2.1.4889.7128. [Google Scholar]
- 2.Tang, K. H. D. Climate change in Malaysia: Trends, contributors, impacts, mitigation, and adaptations. Sci. Total Environ.650, 1858–1871. 10.1016/j.scitotenv.2018.09.316 (2019). [DOI] [PubMed] [Google Scholar]
- 3.Rahman, H. A. Global climate change and its effects on human habitat and environment in Malaysia. Malays. J. Environ. Manag.10, 17–32 (2009). [Google Scholar]
- 4.Marx, W., Haunschild, R. & Bornmann, L. Heat waves: A hot topic in climate change research. Theor. Appl. Climatol.146, 781–800. 10.1007/s00704-021-03758-y (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Malaysian Meterological Department (MetMalaysia). Weather Phenomena-Heat wave. Malaysian Meterological Department (MetMalaysia). https://www.met.gov.my/en/pendidikan/fenomena-cuaca/#Heat%20Wave (2024).
- 6.Muhammad, M. K. I. et al. Heatwaves in peninsular Malaysia: A spatiotemporal analysis. Sci. Rep.14, 4255. 10.1038/s41598-024-53960-x (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ngarambe, J., Nganyiyimana, J., Kim, I., Santamouris, M. & Yun, G. Y. Synergies between urban heat island and heat waves in Seoul: The role of wind speed and land use characteristics. PLoS ONE15, 0243571. 10.1371/journal.pone.0243571 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chew, L. W., Liu, X., Li, X.-X. & Norford, L. K. Interaction between heat wave and urban heat island: A case study in a tropical coastal city. Singapore Atmos. Res.247, 105134. 10.1016/j.atmosres.2020.105134 (2020). [Google Scholar]
- 9.Elsayed, I. S. M. Mitigation of the urban heat island of the city of Kuala Lumpur, Malaysia. Middle East J. Sci. Res.11, 1602–1613. 10.5829/idosi.mejsr.2012.11.11.1590 (2012). [Google Scholar]
- 10.Bishop-Williams, K. E., Berke, O., Pearl, D. L. & Kelton, D. F. A spatial analysis of heat stress related emergency room visits in rural Southern Ontario during heat waves. BMC Emerg.15, 2–9. 10.1186/s12873-015-0043-4 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Anita, A. R., Tan, H. S., Fatimah, A. F., Netto, E. & Muhamad, H. J. Public health impacts of heat waves: A review. Int. J. Public Health Clin. Sci.5, 2289 (2018). [Google Scholar]
- 12.Yatim, A. N. M. et al. Observed trends in extreme temperature over the Klang Valley, Malaysia. Adv. Atmos. Sci.36, 1355–1370. 10.1007/s00376-019-9075-0 (2019). [Google Scholar]
- 13.Department of Occupational, Safety and Health Malaysia (DOSH). Guideline on heat stress management at workplace. Department of Occupational, Safety and Health, Malaysia.https://www.dosh.gov.my/index.php/legislation/guidelines/industrial-hygiene-1/2017-guidelines-heat-stress-management-at-workplace/file (2016).
- 14.Nayak, S. G. et al. Development of a heat vulnerability index for New York State. Public Health161, 127–137. 10.1016/j.puhe.2017.09.006 (2018). [DOI] [PubMed] [Google Scholar]
- 15.Arifwidodo, S. D. & Chandrasiri, O. Urban heat stress and human health in Bangkok, Thailand. Environ. Res.185, 109398. 10.1016/j.envres.2020.109398 (2020). [DOI] [PubMed] [Google Scholar]
- 16.Min, J. Y., Lee, H. S., Choi, Y. S. & Min, K. B. Association between income levels and prevalence of heat- and cold-related illnesses in Korean adults. BMC Public Health21, 1264. 10.1186/s12889-021-11227-4 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Anderson, G. B. et al. Heat-related emergency hospitalizations for respiratory diseases in the medicare population. Am. J. Respir. Crit. Care Med.187, 1098–1103. 10.1164/rccm.201211-1969oc (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Centers for Disease Control and Prevention (CDC). Protecting vulnerable groups from extreme heat. Centers for Disease Control and Prevention. https://www.cdc.gov/disasters/extremeheat/specificgroups.html (2017).
- 19.Garami, A. & Szekely, M. Body temperature: It’s regulation in the framework of energy balance. Temperature1, 28–29 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Leon, L. R. & Bouchama, A. Heat stroke. Compr. Physiol.5, 611–647. 10.1002/cphy.c140017 (2015). [DOI] [PubMed] [Google Scholar]
- 21.Hanna, E. G. & Tait, P. W. Limitations to thermoregulation and acclimatization challenge human adaptation to global warming. Int. J. Environ. Res. Public Health12, 8034–8074. 10.3390/ijerph120708034 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gomez, C. R. Role of heat shock proteins in aging and chronic inflammatory diseases. GeroScience43, 2515–2532. 10.1007/s11357-021-00394-2 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Shaharuddin, A., Noorazuan, M. H., Takeuchi, W. & Noraziah, A. The effects of urban heat islands on human comfort: A case of Klang Valley Malaysia. Glob. J. Adv. Pure Appl. Sci.02, 01–08 (2014). [Google Scholar]
- 24.Rosner, B. Fundamentals of Biostatistics (5th ed.). USA: Cengage Learning. https://www.cengageasia.com (2015).
- 25.Baharav, Y. et al. The impact of extreme heat exposure on pregnant people and neonates: A state of the science review. J. Midwifery Women’s Health68, 324–332. 10.1111/jmwh.13502 (2023). [DOI] [PubMed] [Google Scholar]
- 26.Tsuzuki, K. Effects of heat exposure on the thermoregulatory responses of young children. J. Therm. Biol.113, 103507. 10.1016/j.jtherbio.2023.103507 (2023). [DOI] [PubMed] [Google Scholar]
- 27.Notley, S. R., Akerman, A. P., Meade, R. D., McGarr, G. W. & Kenny, G. P. Exercise thermoregulation in prepubertal children: A brief methodological review. Med. Sci. Sports Exerc.52, 2412–2422. 10.1249/MSS.0000000000002391 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Polit, D. F. & Beck, C. T. The content validity index: Are you sure you know what’s being reported? Critique and recommendations. Res. Nurs. Health29, 489–497. 10.1002/nur.20147 (2006). [DOI] [PubMed] [Google Scholar]
- 29.Koo, T. K. & Li, M. Y. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J. Chiropr. Med.15, 155–163. 10.1016/j.jcm.2016.02.012 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McHugh, M. L. Interrater reliability: The kappa statistic. Biochem. Med.22, 276–282 (2012). [PMC free article] [PubMed] [Google Scholar]
- 31.International Physical Activity Questionnaire (IPAQ). https://youthrex.com/wp-content/uploads/2019/10/IPAQ-TM.pdf (2002).
- 32.Lee, P. H., Macfarlane, D. J., Lam, T. H. & Stewart, S. M. Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. Int. J. Behav. Nutr. Phys. Act.8, 115. 10.1186/1479-5868-8-115 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wu, R. M. X. et al. An FSV analysis approach to verify the robustness of the triple-correlation analysis theoretical framework. Sci. Rep.13, 9621. 10.1038/s41598-023-35900-3 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kalantan, Z. I. & Alqahtani, N. A. A study of principal component analysis for mixed data. Int. J. Adv. Appl. Sci.69, 99–104. 10.21833/ijaas.2019.12.012 (2019). [Google Scholar]
- 35.Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E. & Tatham, R. L. Multivariate Data Analysis 6th edn. (Pearson University Press, 2006). [Google Scholar]
- 36.University of California, Los Angeles. A practical introduction to factor analysis: Exploratory factor analysis. Statistical Consulting Group. https://stats.oarc.ucla.edu/spss/seminars/introduction-to-factor-analysis/a-practical-introduction-to-factor-analysis/ (2020).
- 37.Xiang, Z., Qin, H., He, B., Han, G. & Chen, M. Heat vulnerability caused by physical and social conditions in a mountainous megacity of Chongqing, China. Sustain. Cities Soc.80, 103792. 10.1016/j.scs.2022.103792 (2022). [Google Scholar]
- 38.Ministry of Health of Malaysia (MOH). Body Mass Index. Ministry of Health of Malaysia.http://www.myhealth.gov.my/en/bmi/ (2023).
- 39.Department of Statistics Malaysia (DOSM). Household Gross Income. Department of Statistics Malaysia.https://www.dosm.gov.my/v1/index.php?r=column/cthemeByCat&cat=120&bul_id=TU00TmRhQ1N5TUxHVWN0T2VjbXJYZz09&menu_id=amVoWU54UTl0a21NWmdhMjFMMWcyZz09 (2020).
- 40.Ministry of Health of Malaysia (MOH). Malaysian Dietary Guidelines. Ministry of Health of Malaysia.https://www2.moh.gov.my/moh/images/gallery/Garispanduan/diet/km11.pdf (2017).
- 41.Zhang, X. et al. Assessment of regional health vulnerability to extreme heat China, 2019. China CDC Weekly3, 490–494. 10.46234/ccdcw2021.126 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kovach, M. M., Konrad, C. E. & Fuhrmann, C. M. Area-level risk factors for heat-related illness in rural and urban locations across North Carolina, USA. Appl. Geogr.60, 175–183. 10.1016/j.apgeog.2015.03.012 (2015). [Google Scholar]
- 43.Lopez-Bueno, J. A. et al. Analysis of vulnerability to heat in rural and urban areas in Spain: What factors explain heat’s geographic behavior?. Environ. Res.207, 112213. 10.1016/j.envres.2021.112213 (2022). [DOI] [PubMed] [Google Scholar]
- 44.Nunfam, V. F. et al. The nexus between social impacts and adaptation strategies of workers to occupational heat stress: A conceptual framework. Int. J. Biometeorol.10.1007/s00484-019-01775-1 (2019). [DOI] [PubMed] [Google Scholar]
- 45.Nassiri, P. et al. Application of universal thermal climate index (UTCI) for assessment of occupational heat stress in open-pit mines. Ind. Health55, 437–443 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Kenny, G. P., Yardley, J., Brown, C., Sigal, R. J. & Jay, O. Heat stress in older individuals and patients with common chronic diseases. Can. Med. Assoc.182, 1053–1060. 10.1503/cmaj.081050 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Kuras, E. R. et al. Opportunities and challenges for personal heat exposure research. Environ. Health Perspect.125, 085001. 10.1289/ehp556 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Kovats, R. S. & Hajat, S. Heat stress and public health: A critical review. Annu. Rev. Public Health29, 41–55. 10.1146/annurev.publhealth.29.020907.0908 (2008). [DOI] [PubMed] [Google Scholar]
- 49.Wolf, T. & McGregor, G. The development of a heat wave vulnerability index for London, United Kingdom. Weather Clim. Extremes.1, 59–68 (2013). [Google Scholar]
- 50.Zhang, W., McManus, P. & Duncan, E. A raster-based subdividing indicator to map urban heat vulnerability: A case study in Sydney, Australia. Int. J. Environ. Res. Public Health15, 2516. 10.3390/ijerph15112516 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Mallen, E., Stone, B. & Lanza, K. A methodological assessment of extreme heat mortality modeling and heat vulnerability mapping in Dallas, Texas. Urban Clim.30, 100528 (2019). [Google Scholar]
- 52.Pradyumna, A., Bendapudi, R., Zade, D., D’Souza, M., Tasgaonkar, P. Managing the Increasing Heat Stress in Rural Areas in Handbook of Climate Change Resilience (ed. Filho, W. L.) 207–228 10.1007/978-3-319-93336-8_46 (2019).
- 53.Hu, M., Zhang, K., Nguyen, Q. C., Tasdizen, T. & Krusche, K. U. A Multistate study on housing factors influential to heat-related illness in the United States. Int. J. Environ. Res. Public Health19, 15762. 10.3390/ijerph192315762 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Cai, Z., Tang, Y., Chen, K. & Han, G. F. Assessing the heat vulnerability of different local climate zones in the old areas of a Chinese Megacity. Sustainability11, 15. 10.3390/su11072032 (2019). [Google Scholar]
- 55.He, C. et al. Exploring the mechanisms of heat wave vulnerability at the urban scale based on the application of big data and artificial societies. Environ. Int.127, 573–583. 10.1016/j.envint.2019.01.057 (2019). [DOI] [PubMed] [Google Scholar]
- 56.Chow, W. T. L., Chuang, W. C. & Gober, P. Vulnerability to extreme heat in Metropolitan Phoenix: Spatial, temporal, and demographic dimensions. Prof. Geogr.64, 286–302. 10.1080/00330124.2011.600225 (2012). [Google Scholar]
- 57.Franck, U. et al. Heat stress in urban areas: Indoor and outdoor temperatures in different urban structure types and subjectively reported well-being during a heat wave in the city of Leipzig. Meteorol. Z.22, 167–177. 10.1127/0941-2948/2013/0384 (2013). [Google Scholar]
- 58.Quinn, A. et al. Predicting indoor heat exposure risk during extreme heat events. Sci. Total Environ.490, 686–693. 10.1016/j.scitotenv.2014.05.039 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Tochihara, Y. et al. How humans adapt to hot climates learned from the recent research on tropical indigenes. J. Physiol. Anthropol.41, 27. 10.1186/s40101-022-00302-3 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Wijayanto, T., Toramoto, S., Wakabayashi, H. & Tochihara, Y. Effects of duration of stay in temperate area on thermoregulatory responses to passive heat exposure in tropical south-east Asian males residing in Japan. J. Physiol. Anthropol.31, 25. 10.1186/1880-6805-31-25 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.McGlynn, M. L., Collins, C., Hailes, W., Ruby, B. & Slivka, D. Heat acclimation in females does not limit aerobic exercise training outcomes. Int. J. Env. Res. Public Health19, 5554. 10.3390/ijerph19095554 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Foster, J., Hodder, S. G., Lloyd, A. B. & Havenith, G. Individual responses to heat stress: implications for hyperthermia and physical work capacity. Front. Physiol.11, 541483. 10.3389/fphys.2020.541483 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Dongo, K., Kablan, A. K. M. & Kouame, F. K. Mapping urban residents’ vulnerability to heat in Abidjan, Cote d’Ivoire. Clim. Dev.10, 600–613. 10.1080/17565529.2018.1450213 (2018). [Google Scholar]
- 64.Weber, S., Sadoff, N., Zell, E. & de Sherbinin, A. Policy-relevant indicators for mapping the vulnerability of urban populations to extreme heat events: A case study of Philadelphia. Appl. Geogr.63, 231–243. 10.1016/j.apgeog.2015.07.006 (2015). [Google Scholar]
- 65.Maragno, D., Fontana, M. D. & Musco, F. Mapping heat stress vulnerability and risk assessment at the neighborhood scale to drive urban adaptation planning. Sustainability12, 1056 (2020). [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed in this study available from the corresponding author on reasonable request.