Toward a Unified Framework for Climate and Health Research: Gaps in Exposure Metrics and Health Outcomes

Abstract

As the health impacts of climate change become increasingly evident, research on climate-related exposures and health outcomes has grown. However, a disproportionate focus on certain exposures and health outcomes persists, and substantial inconsistencies remain in the definitions and measurement methods used across studies, hindering the integration of findings and the development of comprehensive public health strategies. To address this gap, we assessed epidemiological literature, identifying uneven research distribution and highlighting areas requiring further investigation. We conducted a literature search in PubMed, the Cochrane Library, and Embase using key terms related to climate change and health. A total of 611 original studies on human populations, published in English from January 2000 to June 2024, were included. Significant heterogeneity was evident in the definitions of climate-related exposure metrics. The focus areas of climate-related exposure were narrow, primarily limited to temperature and temperature-related hazards (n = 526), precipitation (n = 88), and air pollution (n = 39), while research on extreme weather events was scarce. Similarly, the scope of health outcomes studied was largely confined to all-cause mortality (n = 249), cardiovascular disease (n = 159), and communicable diseases (n = 148). In contrast, relatively few studies have addressed the impacts of climate change on neurodegenerative diseases (n = 8), endocrine disorders (n = 4), or hematologic diseases (n = 3), indicating a critical need for broader investigation into these underrepresented health outcomes. Future research should prioritize the transparent reporting of both context-specific and standardized exposure definitions to improve cross-study comparability and broaden its scope to encompass more diverse climate-related exposures and health outcomes.

Graphical Abstract

INTRODUCTION

Climate change poses an urgent and substantial risk to both global stability and the health of populations worldwide. In 2024, the average global temperature reached an unprecedented 15.10°C, making it the hottest year on record.1 Consequently, the health impacts of climate change are escalating dramatically. Between 2000 and 2019, approximately 5.08 million annual deaths (9.43% of global deaths) were attributed to non-optimal temperatures.2 Heat-related deaths among older adults surged by 167% in 2023 compared with the 1990s, far exceeding the 65% increase previously estimated.3 Extreme weather events such as storms, floods, and cyclones have been associated with elevated mortality rates, with effects persisting for up to six weeks.4

In response to these growing threats, the World Health Organization hosted the first Health Day at the 28th Conference of the Parties (COP) in 2023,5 calling for global action. Because climate change is not a regional issue but a worldwide challenge, concerted international efforts are needed to effectively address and mitigate its health risks. COP29 in 2024 also included a Health Day, engaging efforts by the COP coalition to permanently embed health commitments within climate action frameworks. Comprehensive evidence on the relationship between climate change and health is now urgently required to facilitate more precise and impactful actions.

Although research on climate change and health has increased significantly,3,6 inconsistent terminology and exposure definitions pose challenges to integrating findings effectively.7 Furthermore, the health outcomes explored tend to focus on a narrow range, with insufficient attention paid to the broader health impacts of climate change.

Therefore, we conducted a mapping review to characterize the range of climate-related exposure metrics and health outcomes examined in epidemiologic studies on climate change. The primary research questions of this review were 1) how climate-related exposures are defined and scoped in climate epidemiology, and 2) how health outcomes are defined and scoped in climate epidemiology. These questions cannot be reformulated into the closed-ended questions required for a conventional systematic review or meta-analysis. For this reason, we adopted a systematic mapping review approach, which is specifically designed to describe, classify, and quantify research landscapes rather than to synthesize effect estimates.8,9

DATA SOURCE

A literature search was performed in PubMed, the Cochrane Library, and Embase using key terms such as “climate change,” “mortality,” and “morbidity.” Detailed search strategies, the selection process, and selected articles are provided in the Supplementary Materials (Supplementary Tables 1, 2, 3, Supplementary Fig. 1). Inclusion criteria encompassed original epidemiological research involving human populations. Studies were excluded if they were not original, did not focus on human subjects, or were not published in English. Eligible studies were published between January 2000 and June 2024. Titles and abstracts were screened independently by two authors (SS and DH), who also reviewed and classified the full text of the selected articles.

We examined climate-related exposure indicators and identified their limitations based on the retrieved literature. Climate-related exposures were defined as exposure variables used in individual studies that represent environmental variables or events influenced by climate change. To further understand the scientific evidence on the health impacts of climate change, we referred to Chapter 7: Health, Wellbeing and the Changing Structure of Communities from the Sixth Assessment Report (AR6) of the United Nations Intergovernmental Panel on Climate Change.10 Climate-sensitive health outcomes were defined as diseases or health conditions whose incidence, severity, or progression may be affected by climate-related exposures, as summarized in Chapter 7 of the AR6. Confidence levels assigned to these health outcomes were informed by Table Box 7.2.1.10

To assess the research landscape beyond the exposures and diseases explicitly included in the AR6, we analyzed and visualized the association between climate change exposures and health outcomes using a Sankey diagram, drawing on the results of the literature search. The health impacts of climate change discussed in the studies were divided into short-term (< 1 month) and long-term (≥ 1 month). Separate Sankey diagrams for mortality and morbidity were produced to account for variations in fatality rates across diseases and to reflect differences in exposures, health outcomes, and research priorities between the two domains. The diagrams were constructed using the number of studies retrieved in our search, with each study contributing counts to the respective categories of exposure, duration of impacts, and health outcomes. If a single study investigated multiple exposures, durations, or health outcomes, each was counted separately. For example, if one study examined the effects of temperature and humidity on all-cause mortality, cardiovascular mortality, and respiratory mortality within a short-term period (a maximum lag of 14 days), it was categorized as having two exposures, one duration of impact, and three health outcomes.

CONCEPTUAL FRAMEWORK OF THE HEALTH IMPACTS OF CLIMATE CHANGE AND HUMAN ACTIVITIES

Fig. 1 is a modified version of the framework proposed by Helldén et al.11 for understanding the health impacts of climate change. In this updated framework, we elaborated on the direct and indirect effects of climate change in greater detail and expanded the range of health outcomes to include additional diseases. We provided human interventions that can be implemented at each pathway linking climate change to health impacts. The direct effects of climate change on health arise from immediate environmental disruptions caused by climate-related hazards. These include temperature-related hazards, such as heatwaves and cold spells, as well as changes in humidity, precipitation patterns, and the frequency and intensity of extreme weather events, including tropical cyclones, floods, droughts, and wildfires. Indirect effects can emerge from complex and interrelated mechanisms. Climate change intensifies air pollution and pollen emission, increases exposure to heavy metals and chemicals, disrupts agricultural yields, and alters vector ecology, leading to shifts in the spread of infectious diseases. It also drives migration and displacement, which amplifies health vulnerabilities. These pathways contribute to a wide range of health outcomes, including communicable diseases such as malaria, dengue fever, diarrheal diseases, and respiratory tract infections (RTIs), as well as non-communicable diseases such as mental health disorders.

Fig. 1. Schematic diagram of health impacts of climate change.

GHG = greenhouse gas, AFOLU = agriculture, forestry, and other land use.

REGIONAL DISTRIBUTION OF CLIMATE-HEALTH RESEARCH

We classified all retrieved publications by country and visualized their geographic distribution using a choropleth world map (Fig. 2). In total, studies originated from 96 countries between 2003 and 2024.

Fig. 2. Global distribution of studies identified in the literature search on climate change and health (2003–2024).

At the continental level, Asia accounted for the largest share of publications (n = 318), followed by Europe (n = 210), North America (n = 118), Africa (n = 83), South America (n = 48), Oceania (n = 7), and other regions (n = 31).

At the country level, research output was highly concentrated in a small number of countries. China contributed the largest number of publications (n = 142), followed by the United States (n = 88), Republic of Korea (= South Korea) (n = 42), Australia (n = 30), and the United Kingdom (n = 28). Other major contributors included Brazil (n = 24), Japan (n = 23), Spain (n = 20), and Taiwan (n = 20). In contrast, numerous countries in Africa, South America, and small island states contributed only one or two studies.

Overall, research output was concentrated in high-income countries, particularly in East Asia, Europe, and North America. In contrast, low-income countries were represented by relatively few studies.

EXPOSURES IN CLIMATE CHANGE AND HEALTH RESEARCH

Climate-health research employs a wide variety of exposure metrics, which are often inconsistently defined or operationalized, limiting their comparability across studies (Table 1). One of the most commonly used variables is temperature, which is frequently assessed using daily mean, minimum, or maximum values.12,13 To reflect human-perceived temperature more accurately, some studies use composite metrics like apparent temperature or the heat index.14,15 With over 20 documented algorithms for calculating heat index values, inconsistent use of these terms and formulas can lead to divergent risk assessments under similar climatic conditions.16

Table 1. Commonly used climate-related exposure indicators in climate change research.

Type	Exposure category	Exposure variables	Description
Direct impacts	Temperature	Mean temperature	Daily and monthly average of measured temperature
		Apparent temperature	The comprehensive metric that accounts for multiple environmental factors, including ambient temperature, relative humidity, wind speed, and solar radiation
		Heat index	The index that combines air temperature and relative humidity to estimate how hot it feels to humans
Direct impacts	Temperature-related hazards	Heatwaves	Periods of more than two consecutive days with temperatures exceeding a threshold (e.g., ≥ 98th percentile of daily average temperature)
		Cold spells	Periods of more than two consecutive days with temperatures below a threshold (e.g., < 5th percentile of daily average temperature)
		Temperature variability	The difference between the maximal and minimal temperatures within specific times (e.g., diurnal temperature range, temperature change between neighboring days)
Direct impacts	Humidity	Relative humidity	The amount of moisture in the air as a percentage of the maximum amount the air can hold at a given temperature
		Specific humidity	The mass of water vapor in a given mass of air
Direct impacts	Precipitation	Heavy rainfall	Precipitation exceeding a threshold over periods from hours to several days (e.g. ≥ 90th percentile of daily precipitation)
		Accumulated precipitation	Daily, weekly or monthly total precipitation
Direct impacts	Tropical cyclones	Single event of tropical cyclone	Single events in a specific area
		Maximum wind speed	When the maximum wind speed is above the ≥ 17.5 m/s (34 knots, 63 km/hr, 39 mph)
Direct impacts	Floods	Single event of floods	Flood events in a specific area as defined by the country’s national weather service or disaster management agency
		Area covered by water	Evaluate the extent of land covered by water by analyzing satellite photographs
Direct impacts	Droughts	Single event of drought	Drought events in a specific area
		Precipitation-based drought	A metric that assesses drought conditions by comparing precipitation (e.g., Standardized Precipitation Index, Standardized Precipitation-Evapotranspiration Index)
		Drought index	An index that integrates multiple weather variables, such as precipitation, soil moisture, and stream flow, to assess drought (e.g., United States Drought Monitor, Evaporative Demand Drought Index)
Direct impacts	Wildfires	PM10, PM2.5	Increased level of daily average PM10, PM2.5 from the wildfire
Indirect impacts	Air pollution	Ozone	Daily maximum 8 hr average concentration
		PM10, PM2.5	Daily, weekly and monthly average of measured PM10 and PM2.5
		NO2	Daily, weekly and monthly average of measured NO2

PM = particulate matter, NO₂ = nitrogen dioxide.

To examine the health impacts of extreme temperatures, studies commonly define heatwaves and cold spells based on thresholds sustained over consecutive days.17,18,19,20 Variability arises in the choice of thresholds (e.g., ≥ 90th or ≥ 95th percentile for heatwaves, < 3rd or < 5th percentile for cold spells) and the number of required consecutive days (e.g., ≥ 2, ≥ 3, ≥ 4, or ≥ 5 days). Some studies use fixed thresholds, such as defining heatwaves as days with daily maximum temperatures above 30°C and cold spells as days below 30°C.21,22

To consider short-term temperature variability, diurnal temperature ranges and temperature changes between neighboring days can be used.23,24,25 Distinguishing between intra-day and inter-day variability remains challenging, and the diverse methods used to calculate inter-day changes add complexity to cross-study comparisons.

Humidity modifies the perception and physiological effects of temperature. Most studies use relative humidity because it is widely available in meteorological data.26 Specific humidity and dew point offer more stable, physically meaningful alternatives, but are less frequently used due to lower accessibility and higher computational requirements.27,28

Precipitation exposure is assessed using daily, weekly, or monthly cumulative rainfall levels.13,29,30 However, definitions of “extreme precipitation” vary: some rely on percentile thresholds,31,32 while others use absolute amounts, and the selected reference period may differ substantially.13,29,30 These variations may reflect regional differences in climate characteristics and adaptation capacity. In some regions, cumulative rainfall over several consecutive days drives health risks, whereas in others, short-duration, high-intensity rainfall events can trigger flooding. In addition, regional differences in water management systems and sewerage coverage may lead to the use of different thresholds for defining extreme precipitation across regions. These findings highlight that precipitation exposure metrics may need to be tailored to local conditions, and universal standardization may not always be appropriate for capturing precipitation-related health risks.

Cyclone exposure is frequently defined based on maximum sustained wind speed, following wind speed thresholds established by the World Meteorological Organization (WMO), which classify tropical storms (> 34 knots) and typhoons or hurricanes (> 64 knots).33,34 However, implementation of these classifications may vary across regional meteorological agencies. The averaging period used to determine sustained winds differs; while the WMO standard and the Japan Meteorological Agency utilize a 10-minute average,35 the U.S. Joint Typhoon Warning Center employs a 1-minute average.36

Floods are sometimes identified through administrative records,37 while other studies use satellite-derived water coverage estimates.38,39

Some research focused on drought events,40 while others used various drought indices for assessment. The Standardized Precipitation Index, based solely on precipitation anomalies, is commonly used due to its simplicity.41 The more advanced Standardized Precipitation-Evapotranspiration Index incorporates potential evapotranspiration, offering better reflect with climate variability.42 Because no single index can fully consider the diverse characteristics of drought, many meteorologists and climatologists now use multiple indices and consider the unique features of each. For example, the U.S. Drought Monitor combines five key indicators, including streamflow and soil moisture, to classify drought into five levels, excluding normal conditions.43

Wildfire exposure assessment of wildfire is a complex task due to the diverse and dynamic nature of wildfire events. Current research on the health impacts of wildfires primarily focuses on evaluating the effects of wildfire-related particulate matter (PM).44,45 Levels for PM_2.5 (airborne particles < 2.5 µm in diameter) are typically obtained from ground-based monitoring stations, satellite remote sensing, or air quality models.

CLIMATE-SENSITIVE HEALTH OUTCOMES

The AR6 classifies climate-sensitive health outcomes based on the current body of evidence and assigns confidence levels to each health outcome.10 The health outcomes within the green circle are climate-sensitive diseases while the outcomes within the gray circle are diseases that are not climate-sensitive according to the AR6 (Fig. 3).

Fig. 3. Climate-sensitive health outcomes and underexplored disease categories. Climate-sensitive health outcomes are defined according to Chapter 7: Health, Wellbeing and the Changing Structure of Communities from the Sixth Assessment Report of the United Nations Intergovernmental Panel on Climate Change, with confidence levels assigned based on Table Box 7.2.1 of the same chapter. Injuries, exposure to chemical contaminants, and maternal, foetal and neonatal health are categorized as climate-sensitive health, but no assessment of the global burden of climate-related risk has been made in Table Box 7.2.1.

^aThe impacts of climate change on diarrhoeal diseases and respiratory tract infections were assessed as a mix of negative and positive effects, though these impacts are not evenly distributed. All other health outcomes assigned to confidence levels were evaluated as having negative impacts from climate change.

Among communicable diseases, the impacts of climate change on malaria, dengue, and salmonellosis are considered negative, based on robust evidence (high confidence).10 These diseases are strongly linked to climate-driven changes in temperature, humidity, and precipitation, which influence vector ecology and pathogen transmission.46,47 The AR6 indicates that climate change may have both positive and negative impacts on diarrhoeal diseases,10 which may partly reflect differences in pathogen-specific responses to climate change. A meta-analysis reported a positive association between ambient temperature and both all-cause diarrhoea and bacterial diarrhoea, whereas no significant association was observed for viral diarrhoea.48 The scientific evidence evaluated in the AR6 indicates that climate change may have both positive and negative impacts on RTIs, with a medium level of confidence.10 RTIs can be caused by a wide range of pathogens, including bacteria, viruses, and fungi. In addition, incidence of RTIs is influenced by multiple factors, such as comorbidities, age, vaccination status, and healthcare access, which makes it difficult to isolate the independent effects of climate change on RTIs.49 The association between climate change and respiratory infections may also vary by pathogen type. For example, a case-crossover study examining temperature and pneumonia reported that extremely low temperatures were associated with a greater increase in influenza-related pneumonia than bacterial pneumonia.50 Furthermore, the climatic drivers of influenza transmission appear to differ across climate zones. In temperate regions, low humidity has been consistently associated with influenza outbreaks,51,52,53 whereas in tropical and subtropical regions, influenza epidemics often occur during warm and humid seasons.54

Non-communicable diseases, including respiratory diseases, cardiovascular diseases, cancer, and diabetes, are increasingly recognized as climate-sensitive, at medium to high confidence levels.10 Climate change is also likely to increase the risk of several malignancies. Proposed pathways for increased cancer risk include greater exposure to multiple carcinogens,55,56 but epidemiological evidence from population-level studies remains insufficient.10 Individuals with diabetes are particularly vulnerable to heat-related morbidity and mortality, with medium confidence according to the AR6.10 Recent studies have also reported that cold exposure is associated with increased mortality and morbidity among individuals with diabetes.20,57 Impaired thermoregulatory responses in patients with diabetes may contribute to this vulnerability. In particular, diabetes-related dysfunction in cutaneous vasodilation and vasoconstriction can limit the body’s ability to regulate heat exchange, thereby increasing susceptibility to adverse health outcomes during temperature extremes.58

Several domains remain underexplored or insufficiently quantified. While some studies report associations between prenatal heat or air pollution exposure and adverse birth outcomes,59 comprehensive and geographically representative data on maternal, fetal, and neonatal health impacts are still lacking. Disruptions to healthcare access during extreme events, such as floods or wildfires, can further compound these risks, particularly in settings with fragile health systems.60 However, AR6 does not assign a formal confidence level to this category, reflecting a need for more consistent and globally representative evidence.10

Mental health is also increasingly recognized as sensitive to climate change, with stressors ranging from direct trauma following natural disasters to chronic anxiety related to climate uncertainty and displacement. These effects are rarely integrated into health surveillance systems, and the existing literature remains fragmented. Although the psychological impacts of acute disasters are relatively well documented, the long-term and cumulative mental health burden of climate change remains poorly quantified and is currently classified with low confidence.10

Another underappreciated domain involves altered exposure to environmental contaminants. Climate change may modulate human exposure to heavy metals, persistent organic pollutants, and other hazardous substances through complex environmental pathways. For instance, rising ocean temperatures can elevate methylmercury concentrations in fish,61 while altered rainfall patterns and soil chemistry may affect arsenic or cadmium mobility in drinking water. Although plausible, these pathways have yet to be systematically studied in human populations, and current evidence remains largely ecological or theoretical.

These gaps extend to broader disease categories such as autoimmune disorders, neurodegenerative diseases, and skin conditions. For example, heat stress, UV radiation, and air pollution may exacerbate systemic lupus or multiple sclerosis, yet epidemiologic data are less explored.

CRITICAL GAPS IN CLIMATE CHANGE EXPOSURE AND HEALTH OUTCOMES

We constructed Sankey diagrams based on the number of studies retrieved from our literature search (Fig. 4). In terms of climate-related exposures, in mortality-focused research, temperature and temperature-related hazards were the most frequently studied climate-related exposures, followed by air pollution and precipitation. Drought, humidity, tropical cyclones, floods, and wildfires are comparatively underrepresented. In morbidity-focused research, temperature-related hazards remained the most frequently studied exposure, followed by precipitation and humidity. Air pollution, floods, droughts, and tropical cyclones received less attention. Mortality studies more frequently investigated air pollution, while morbidity studies tended to focus on precipitation and humidity. The predominance of temperature-related exposures may reflect the geographic concentration of studies in high-income, mid- to high-latitude regions. As a result, other climate-related exposures that are more relevant in tropical or low-income settings may be comparatively underrepresented in the current evidence base.

Fig. 4. Health impacts of climate-related exposures on mortality and morbidity outcomes. (A) Depicts the health effects of climate-related exposures on mortality outcomes, while (B) illustrates their impacts on morbidity outcomes. The numbers represent the count of studies retrieved in our literature search. Studies are categorized into short-term (< 1 month) and long-term (≥ 1 month) based on the duration of the impact.

Regarding the duration of impacts, short-term impacts are overwhelmingly more frequently studied compared with long-term impacts. In mortality research, the focus is predominantly on short-term impacts of climate-related exposures, with few studies examining long-term effects. A similar pattern is evident in morbidity research, in which short-term studies are more prevalent. The prevalence of short-term studies shows that most research focuses on acute, day-to-week associations between weather variables and health outcomes. While these investigations offer important insights into acute risks, they may miss the long-term and cumulative implications of climate change.

In mortality-focused research, all-cause mortality was the most frequently studied health outcome, followed by cardiovascular diseases and respiratory diseases. Communicable diseases, mental health, and diabetes were less commonly explored. In morbidity-focused research, communicable diseases were the most frequently studied, reflecting their broad and non-lethal impacts. This was followed by cardiovascular diseases, mental health, and genitourinary diseases. Less studied outcomes included injuries, maternal, fetal, and neonatal health, and digestive tract diseases. Health outcomes that are under-researched in both the domains of mortality and morbidity are as follows: neurodegenerative diseases, digestive tract diseases, cancer, endocrine diseases, skin diseases, malnutrition, hematologic diseases, and sensory organ diseases.

CHALLENGES AND DIRECTIONS FOR FUTURE RESEARCH

Through this analysis, we identified several challenges facing research on this topic and proposed directions to address them (Fig. 5). Regarding exposure, there is considerable variation in terminology and exposure metrics across studies, with substantial heterogeneity in the operational definitions of climate-related exposures. For example, the duration and threshold used to define heatwaves and cold spells vary across studies. For cyclones, meteorological definitions are not uniformly standardized across regions. In the case of drought, a wide range of indices is used, which may reflect differences in available monitoring infrastructure and scientific objectives across regions. Such differences in exposure definitions may reflect regional climatic conditions, adaptation capacity, and sociocultural contexts. Even when locally tailored exposure metrics are appropriate, discussion of standardized definitions is necessary to enable cross-study comparison. Future research may benefit from reporting both context-specific and standardized exposure definitions whenever feasible.

Fig. 5. Challenges in current research and directions for future studies.

The interconnected nature of climate-related exposures demands an integrated approach to research. Current single-exposure studies often overlook the compounded and synergistic effects on health outcomes. Future research should use multi-pollutant models. Creating a unified climate anomalies index that integrates variables such as temperature, humidity, and precipitation could significantly improve our understanding of the combined effects of multiple climatic factors. Standardized metrics would facilitate comprehensive assessments of the health impacts of climate change.

In terms of outcomes, most research continues to focus on well-known health outcomes, such as all-cause mortality and cardiovascular and respiratory diseases. This concentration may reflect the population-level disease burden associated with these diseases. According to the Global Burden of Disease Study 2023, ischemic heart disease ranked as the leading cause of disability-adjusted life years (DALYs) globally, followed by neonatal disorders and stroke.62 Lower respiratory infections and chronic obstructive pulmonary disease ranked fourth and seventh, respectively.62 However, the prioritization of research outcomes does not always correspond with the global burden profile. For example, neonatal disorders represent the second leading contributor to global DALYs, yet few climate–health studies have examined neonatal or perinatal outcomes. This indicates that current research remains heavily concentrated on a narrow range of well-established outcomes, potentially overlooking other significant health threats. Therefore, it is crucial to move beyond the traditional focus on a few major diseases and incorporate a broader spectrum of health outcomes to understand the diverse risks posed by a changing climate.

A major challenge in current climate–health research lies in moving beyond short-term assessments to a more comprehensive understanding of long-term impacts. Climate change is fundamentally about long-term alterations in temperature, precipitation patterns, and the frequency of extreme weather events, yet “climate change and health” studies focused on acute meteorological conditions rather than cumulative shifts in climate. Consequently, most epidemiological research does not address the attribution of health outcomes to climate change itself, concentrating instead on short-term exposures. However, the limited evidence on long-term impacts reflects not only research priorities but also substantial structural constraints. Health datasets with continuous follow-up spanning several decades are often unavailable, limiting the ability to evaluate cumulative exposures over extended time horizons. Moreover, over long periods, non-climatic determinants—including ageing population, socioeconomic development, urbanization, and adaptation measures such as air conditioning and heating—may exert considerable influence on health outcomes. These factors complicate efforts to isolate the contribution of anthropogenic climate change from broader population-level transformations. Nevertheless, advancing long-term investigations remain essential for understanding impacts of climate–health relationships. Incorporating historical baselines and long-term climate anomalies to help clarify the contribution of anthropogenic climate change relative to natural climate variability and societal transformations.

STRENGTHS AND LIMITATIONS

This study has several notable strengths that distinguish it from prior reviews in the field of climate change and health. First, we quantified exposure–outcome combinations across 611 epidemiologic studies, allowing us to identify structural patterns and imbalances in the existing evidence base. This approach allowed us to present novel evidence on how exposures and health outcomes have been disproportionately addressed in the existing literature. Second, we integrated the AR6 classifications of climate-sensitive health outcomes into our analytical framework, thereby linking epidemiologic research practices with the current state of consensus. Third, we used Sankey diagrams to clearly illustrate research disproportions in the existing climate–health literature.

At the same time, limitations should be acknowledged. First, this review was restricted to English-language publications, which may have resulted in the underrepresentation of studies conducted in regions where climate-related health impacts are highly relevant, but where research is more commonly published in local languages. This restriction may contribute to geographic and socioeconomic bias in the mapped research landscape and could partially influence the observed concentration of studies in high-income countries. Second, our assessment of research disproportions relied on study counts rather than effect estimates. Consequently, this review describes the distribution and scope of research activity but does not evaluate the strength, direction, or magnitude of climate–health associations. Therefore, areas with fewer studies are not necessarily associated with weaker evidence, and highly studied exposure–outcome combinations may still have inconsistent or uncertain effect estimates.

CONCLUSION

Climate-health research remains hindered by inconsistent exposure metrics, a narrow focus on health outcomes, and an overemphasis on short-term risks. To improve cross-study comparability, researchers should document both context-specific and standardized exposure definitions in parallel, acknowledging regional and sociocultural variation. Beyond exposure, the field should expand into underexplored health outcomes. A unified research framework that integrates these elements will strengthen evidence synthesis, improve policy relevance, and support equitable adaptation strategies across diverse health systems.

SUPPLEMENTARY MATERIALS

Supplementary Table 1

Search strategies for each database.

Supplementary Table 2

Selected articles on mortality-related health outcomes.

Supplementary Table 3

Selected articles on morbidity-related health outcomes.

Supplementary Fig. 1

Flow diagram of study selection.