Climate Change and Mental Health: A Scoping Review

Climate change is negatively impacting the mental health of populations. This scoping review aims to assess the available literature related to climate change and mental health across the World Health Organisation’s (WHO) five global research priorities for protecting human health from climate change. We conducted a scoping review to identify original research studies related to mental health and climate change using online academic databases. We assessed the quality of studies where appropriate assessment tools were available. We identified 120 original studies published between 2001 and 2020. Most studies were quantitative (n = 67), cross-sectional (n = 42), conducted in high-income countries (n = 87), and concerned with the first of the WHO global research priorities—assessing the mental health risks associated with climate change (n = 101). Several climate-related exposures, including heat, humidity, rainfall, drought, wildfires, and floods were associated with psychological distress, worsened mental health, and higher mortality among people with pre-existing mental health conditions, increased psychiatric hospitalisations, and heightened suicide rates. Few studies (n = 19) addressed the other four global research priorities of protecting health from climate change (effective interventions (n = 8); mitigation and adaptation (n = 7); improving decision-support (n = 3); and cost estimations (n = 1)). While climate change and mental health represents a rapidly growing area of research, it needs to accelerate and broaden in scope to respond with evidence-based mitigation and adaptation strategies.


Introduction
In 2009, a Lancet Commission on Climate Change asserted that "climate change is the biggest global health threat of the 21st century" [1]. In response, the 'Lancet Countdown on health and climate change' has been established as an independent, global monitoring system dedicated to tracking the health dimensions of the impacts of, and the response to, climate change [2]. While the Lancet Countdown includes numerous health indicators, it currently lacks an indicator capable of capturing the impact of climate change on mental health globally. This lack of representation in global climate and health initiatives is of considerable concern, as mental disorders are a leading cause of burden of disease globally and contribute to increased rates of premature mortality [3].

•
Must relate to mental health and climate change across one or more of the domains defined by the Lancet Countdown for Health and Climate Change 2018 [11]: climate change impacts, exposures, and vulnerability; adaptation, planning, and resilience for health; mitigation actions and health co-benefits; finance and economics; and public and political engagement • Must be original research (e.g., cross-sectional or cohort studies) • Publication year must be from 2001 onwards • English language Details of the protocol for this review were registered on PROSPERO (registration number CRD42020161076) and can be accessed at https://www.crd.york.ac.uk/prospero/ display_record.php?ID=CRD42020161076.

Quality Assessment
The same three authors (F.C., S.A., and M.P.) assessed all studies for which appropriate quality assessment tools were available (n = 93). We used the 'Observational Cohort and Cross-Sectional Studies' and 'Case Series Studies' Quality Assessment tools from the National Heart, Lung, and Blood Institute (NHLBI) [12]. The NHLBI tools were robustly developed and include individualised tools for a range of study designs (e.g., cohort, cross-sectional, case series). Qualitative studies were assessed using the Critical Appraisal Skills Programme (CASP) Qualitative Checklist [13]. Ecological studies were evaluated using a scale adapted from Cortes-Ramirez et al. [14], with the "sample size" and "validity of statistical inferences" criteria removed as their specificity was not deemed relevant to providing an overall quality assessment of the studies included in the review. We did not assess the quality of studies employing modelling, case report, time-series, case-crossover, panel, or other unique study designs due to the lack of appropriate quality assessment tools.
Studies were assessed as either 'poor', 'fair, or 'good', guided by a series of questions set by the quality assessment tools. The findings of 'poor' quality studies were excluded from our synthesis of results but can be found listed in Table A1.

Research Framework
Studies were grouped by a member of the team under the five global research priorities for protecting health from climate change proposed by the World Health Organization: assessing the risks; identifying the most effective interventions; guiding health-promoting mitigation and adaptation decisions in other sectors; improving decision-support; and estimating the costs of protecting health from climate change [8]. Further information about the WHO framework can be found in Appendix A.

Results
We identified 120 original research studies that examined mental health in the context of climate change-related exposures ( Figure 1). Sixty-seven studies reported quantitative data, 34 reported qualitative data, and 19 reported a combination of qualitative and quantitative data. The most common study design was cross-sectional (n = 42), followed by case studies (n = 28), modelling studies (n = 12), ecological studies (n = 7) and cohort studies (n = 7). Table A1 provides a summary of the included studies. by case studies (n = 28), modelling studies (n = 12), ecological studies (n = 7) and cohort studies (n = 7). Table A1 provides a summary of the included studies. The geographies which represented the most original research studies were Australia (n = 34), Canada (n = 17), USA (n = 16), China (n = 6), United Kingdom (n = 6), Italy (n = 5), and Bangladesh (n = 5) ( Figure 2). The majority of studies (77%) were carried out in highincome countries, with 12% from upper middle income, 12% from lower middle income and only 3% from low-income countries (as per World Bank classifications). The geographies which represented the most original research studies were Australia (n = 34), Canada (n = 17), USA (n = 16), China (n = 6), United Kingdom (n = 6), Italy (n = 5), and Bangladesh (n = 5) ( Figure 2). The majority of studies (77%) were carried out in high-income countries, with 12% from upper middle income, 12% from lower middle income and only 3% from low-income countries (as per World Bank classifications).
by case studies (n = 28), modelling studies (n = 12), ecological studies (n = 7) and cohort studies (n = 7). Table A1 provides a summary of the included studies. The geographies which represented the most original research studies were Australia (n = 34), Canada (n = 17), USA (n = 16), China (n = 6), United Kingdom (n = 6), Italy (n = 5), and Bangladesh (n = 5) ( Figure 2). The majority of studies (77%) were carried out in highincome countries, with 12% from upper middle income, 12% from lower middle income and only 3% from low-income countries (as per World Bank classifications).  The following sections provide a narrative summary of the original research corr sponding to the WHO framework.

Assessing the Risks
The majority of studies (n = 101 out of 120) examined the mental health risks pose by climate change. The literature linked mental health outcomes to several climate-relate exposures-heat, humidity, rainfall, drought, wildfires, and floods. Other environment exposures, such as tropical cyclones and storms, were not included as no studies we found that explicitly referred to climate change. This section will present studies accor ing to the following three categories (which are not mutually exclusive): (1) environment exposures; (2) mental health outcomes; and (3) specific sub-groups and contexts th might be particularly vulnerable to the impact of climate change on mental health.

Temperature and Humidity
The most commonly examined environmental exposure in relation to health risk w temperature (n = 27), which was often measured alongside other meteorological variabl such as humidity and rainfall. In most studies, an increase in temperature was positive correlated with poor mental health outcomes.
Three studies from Australia found associations between heat and mental health ou comes; however, the evidence was inconsistent. For example, Ding et al., report that a The following sections provide a narrative summary of the original research corresponding to the WHO framework.

Assessing the Risks
The majority of studies (n = 101 out of 120) examined the mental health risks posed by climate change. The literature linked mental health outcomes to several climate-related exposures-heat, humidity, rainfall, drought, wildfires, and floods. Other environmental exposures, such as tropical cyclones and storms, were not included as no studies were found that explicitly referred to climate change. This section will present studies according to the following three categories (which are not mutually exclusive): (1) environmental exposures; (2) mental health outcomes; and (3) specific sub-groups and contexts that might be particularly vulnerable to the impact of climate change on mental health.

Temperature and Humidity
The most commonly examined environmental exposure in relation to health risk was temperature (n = 27), which was often measured alongside other meteorological variables such as humidity and rainfall. In most studies, an increase in temperature was positively correlated with poor mental health outcomes.
Three studies from Australia found associations between heat and mental health outcomes; however, the evidence was inconsistent. For example, Ding et al., report that an increase in temperature and humidity affects psychological distress, independently from pre-existing depression or anxiety [20]. Similarly, another study that looked at data from Adelaide, Australia, found that while there was a positive association between temperatures exceeding the maximum and minimum thresholds (32 • C and 16 • C) and psychiatric presentations to the emergency department, there was a significant decrease in this effect when temperatures were considered extreme [21]. A third study conducted on children in Australia showed a small effect of worsened childhood mental health with an increase in annual average daily maximum temperatures [22].
Several studies from Asia found that fluctuating temperatures influenced mental health and well-being, impacting productivity and livelihoods [23][24][25]. Long-term exposure to high and low temperatures in Taiwan resulted in a 7% increase of major depressive disorder incidence per 1 • C increment in regions with an average annual temperature above the median 23 • C. A higher incidence was found in the older age group (65+ years), with a slightly higher hazard ratio for males in this group, and females in the younger age group (20-64 years) [26].
Temperature also affected mental health and well-being in several studies conducted in North America. Noelke et al. reported that temperatures above daily averages reduced positive emotions like happiness, increased negative emotions like anger and stress, and increased fatigue [27]. Another study found that monthly temperatures above 30 • C increased the probability of mental health difficulties by 0.5%, and ongoing exposure to increasing temperatures (1 • C over five years) was associated with a 2% increase in the prevalence of mental health issues [28]. High ambient temperatures were also found to increase help-seeking for mental health, such as crisis text line usage in young adults in several urban areas in the United States [29]. A study exploring protective factors in the context of extreme seasonal weather in Tennessee, USA found that social cohesion was associated with reduced mental health impacts among low-and moderate-income residents [30].

Drought and Rainfall
Drought was another frequently studied climate exposure (n = 16). All but one of the studies were from Australia and primarily focussed on rural communities. Almost half of these studies were based on large, longitudinal studies and utilised standard mental health measures such as the Kessler Psychological Distress Scale (K10). A nationally representative cross-sectional survey of Australian residents found that while drought was associated with elevated psychological distress (as measured by the K10) in rural communities, this was not the case for urban dwellers [31].
A longitudinal cohort study showed elevated drought-related stress and psychological distress (as measured by the K10) among farmers [32]. Drought-related stress, which captures worry about the impacts of drought on themselves and their families and communities, was influenced by socio-demographic and community factors (e.g., loss, government compliance pressures, and difficulties accessing mental health support or receiving inappropriate mental health services) that differed from the factors that influenced psychological distress [33]. Factors associated with higher psychological distress included being unemployed and exposure to other adverse life events, while protective factors included financial security, social support, and a sense of community [32,34]. The evidence for differential impacts on gender and age is conflicting [35][36][37].
One multi-country study has looked at rainfall as a climate-related exposure and has reported that the highest prevalence of mood disorders is observed in countries characterised by small variations across monthly rainfall and high levels of rainfall [38].

Wildfire
Two studies from North America examined the impacts of wildfires on mental health. Qualitative interviews of people purposively sampled to include a broad cross-section of backgrounds and experiences from a severe 2014 wildfire season in the North-West Territories of Canada, in the context of climate change, reported how experiences of evacuation and isolation, as well as feelings of fear, stress, and uncertainty, contributed to acute and long-term negative impacts on mental and emotional well-being [39]. Prolonged smoke events were linked to respiratory problems, extended time indoors, and disruptions to livelihood and land-based activities, which negatively affected mental well-being [39].
A household study conducted one year after the Wallow Fire in Arizona, United States, concluded higher solastalgia, a term used to indicate distress caused by environmental change (measured using a scale adapted from [40]), and an adverse financial impact of the fire was associated with clinically significant psychological distress (as measured by the K10). In contrast, a higher family functioning score was associated with less psychological distress [41]. Other countries that experience regular wildfires, such as Australia, referred to as bushfires, had no research exploring mental health impacts in the context of climate change.

Flood
Several studies examined the mental health impacts of flooding. An Australian study qualitatively explored individuals' experiences from rural communities, concluding that the threat of drought and flood are intertwined and contributed to decreased well-being from stress, anxiety, loss, and fear [42]. A cohort study from the UK looking at the long-term impact of flooding found psychological morbidity persisted for at least three years after the flooding event [43].
In addition to the tangible mental health impacts related to exposure to specific climate change-related events, studies have reported psychological impacts of perceived risks associated with climate change [44,45].

Symptom Scales and Screening Tools
Symptom scales were commonly used to examine mental health outcomes. Distress (n = 17) was one of the most investigated outcomes; primarily psychological distress measured using the K10. Some studies attempted to develop specific environmental distress measures, including a scale for solastalgia and a specific 'climate change distress' scale [41,[46][47][48]. Four studies used validated screening tools for mental disorders [39,41,45,46]. Only one study used a structured diagnostic instrument to determine mental disorder diagnoses [49].

Hospital Admissions
Alternatively, many studies utilised routinely collected administrative data. Psychiatric hospital admissions associated with temperature and heat waves were frequently studied (n = 15). These studies demonstrate a positive association, particularly with elevated temperatures, with susceptibility varying in terms of demographic variables (e.g., older people shown to be at higher risk) [50,51] and type of disorder (e.g., organic disorders such as dementia) [52]. The relationship between hospital admissions and heat has been examined for a range of mental and neurological disorders, including dementia, mood disorders, anxiety disorders, schizophrenia, bipolar disorder, somatoform disorders, and disorders of psychological development [52][53][54].

Mortality
Mortality has also been found to be influenced by high ambient temperatures for people living with mental illness and neurological conditions. A study of health outcome data from Adelaide, South Australia, for 1993-2006, has demonstrated that mortality attributed to mental and behavioural disorders increased during heat waves in the 65to 74-years age group and in persons with psychosis [52,55]. Another European study supports this finding with increased mortality risk for people with psychiatric disorders during heat waves from 2000 to 2008 in Rome and Stockholm, particularly for older people (75+) and women. A study from England found that primary care patients with psychosis, dementia, and alcohol or substance misuse experienced significantly higher heatrelated mortality above regionally defined temperature thresholds. This effect was more marked for younger patients and those with a primary diagnosis of alcohol or substance misuse [56,57]. Nitschke et al., found a small increase in mental health-related mortality in people aged 65-74 years during heat waves in metropolitan Adelaide [58]. Projections of mortality under different climate change scenarios in China also estimate increasing trends in heat-related excess mortality for mental disorders but a decreasing trend in cold-related excess mortality [59].

Self-Harm
Temperature has also been associated with self-harm and suicide rates [60][61][62][63][64]. Data from Finland found that temperature variability explained more than 60% of the total suicide variance over several decades [62]. Using data from the US and Mexico, suicide rates were found to increase by 0.7% and 3.1%, respectively, for a 1 • C increase in monthly average temperature, with unmitigated climate change projected to result in a combined 21,770 (95% CI 8950-39,260) additional suicides by 2050 [64].

Burden of Disease
Studies attempting to quantify the burden of mental disorders attributable to climate change are sparse. One study from South Korea has estimated the burden of disease related to climate change for a range of conditions and climate-related measures, including PTSD, as an outcome of disasters [65]. Unfortunately, this study was limited in several ways, including a lack of availability of input data.

Pre-existing Mental Illness
We identified a body of research related to populations and contexts that are anticipated to be more vulnerable to climate change and its mental health impacts. People taking certain psychotropic medications (including hypnotics, anxiolytics, and antipsychotics) are at increased risk of heatstroke and death as a result of high temperatures, possibly due to disruptions in thermoregulation triggered by some psychotropic medications [56,66]. One study from Australia showed that people living with obsessive-compulsive disorder (OCD) experience obsessions and compulsions directly aligned with climate change concerns [67].

Youth
According to the results of a national survey of Australian children aged 6-11, higher temperatures may impact children's mental health [22]. Youth affected by drought may also experience cumulative mental health impacts [68].

Indigenous People
Qualitative studies reporting the unique mental health impacts of climate change on Inuit communities in Canada have described a loss of place-based solace, land-based activities such as hunting, and cultural identity due to changing weather and local landscapes [69,70]. Changes in sea ice stability and surface area are of particular cultural concern for Inuit [71]. Studies among these communities report that climate and environmental changes increase family stress, enhance the possibility of increased drug and alcohol use, amplify previous traumas and mental health stressors, and may increase the risk of suicidal ideation [72]. Indigenous communities from Alaska have also expressed significant concern for climate change and its impacts on local sustenance resources, such as berries [73]. Mental health, physical health, traditional/western education, access to country food and store-bought foods, access to financial resources, social networks, and connection to Inuit identity emerged as key components of Inuit adaptive capacity in the face of climate change [74]. Studies conducted with Aboriginal and Torres Strait Islander peoples from Australia also highlight the environmental impacts of climate change on emotional wellbeing, including increased community distress from deteriorating the connection to country [75,76]. Heat also appeared to be associated with suicide incidence in Australia's Indigenous populations; however, other socio-demographic factors may play a more critical role than meteorological factors [77].

Low-and Middle-Income Countries
A study from Ethiopia, a low-income country that has a high dependency on the local environment to meet basic human and animal needs, demonstrated that seasonal environmental changes related to water security expose populations to significant emotional distress [78]. Small island developing states in the Pacific Ocean have been deemed to be particularly at risk and vulnerable to the impacts of climate change. Low-lying villages in the Solomon Islands in the South Pacific report that sea-level rise causes fear and worry on a personal and community level [79]. Research from Tuvalu, another small Pacific Island experiencing the threat of sea-level rise, describes the individual experiences of distress in the face of climate change and underscores the necessity to provide access to culturally informed social and mental health services in the region [80]. Three studies from Bangladesh have reported negative effects on emotional well-being as a result of climate-induced immobility [81][82][83].

Identifying the Most Effective Interventions
There were eight studies related to interventions; however, these were primarily exploratory and qualitative and focused on health professionals and community settings, and did not provide clear support for any one specific intervention.
One study looked at the health and social service responses to the long-term mental health impacts of a flood event. It concluded that sustained recovery interventions rooted in local knowledge and interdisciplinary action were required and that there are unintended consequences related to psychosocial interventions that can incite complex emotions and impact psychosocial recovery [84].
A study from rural Australia has described how a community development model, incorporating elements of health promotion, education, and early intervention, was accepted and considered effective in helping communities build capacity and resilience in the face of chronic drought-related hardship [85]. Unfortunately, this study did not measure the intervention's impact on the mental health status of participants. Another Australian study that focused on rural health service managers found that 90% of respondents perceived climate change as likely to impact mental health and highlighted the important role of rural health services in education and advocacy on climate change's health impacts [86]. A case study of health promotion practices that address climate change in Victorian (Australia) healthcare settings found that community gardens improved social connectedness and mental and physical health [87]. Results of a study from Canada revealed the need for improved medical education on climate change and health, including mental health, suggesting that a 3-h education activity would be useful and would allow family physicians to use this knowledge in their daily practice, notably through prevention and counselling [88].
Two studies explored the relationship between pro-environmental behaviour and mental health in terms of awareness of health risks and potential health co-benefits. Results from a survey in China indicated that residents' health-risk perception is positively affected by climate-change information. Additionally, perceiving climate change as detrimental to mental health was shown to influence a residents' attitude and intention to take environmental action more than physical health-risk perception [89]. A study of UK households found that while individual-level pro-environmental attitudes were negatively associated with mental health, household pro-environmental behaviours and attitudes of other household members were associated with positive outcomes, pointing towards a role of social capital and the household level as an important target for interventions [90].
Finally, a study exploring health perceptions in climate-driven migrants found that emphasis was placed on the importance of mental health, indicating that mental health interventions could be effective in this population [91].

Guiding Health-Promoting Mitigation and Adaptation Decisions in Other Sectors
A small number of studies (n = 7) examined how sectors outside of health can mitigate and adapt to reduce climate change impacts on mental health. Three of these studies were from Inuit communities. Social and environmental factors that may protect mental health and wellbeing were highlighted, including being on the land and connecting to Inuit culture, as well as how Inuit-led monitoring of environmental conditions can guide climate change adaptation that considers intangible losses and damages to well-being and ways of living [92][93][94]. A study from Nepal revealed similar themes [95], while a 15-year prospective cohort study of Hurricane Katrina survivors (New Orleans, LA, USA) has concluded that mitigation efforts should prevent trauma exposure through investments in climate resilience and eliminating evacuation impediments [96].
Barriers to implementing strategies to adapt to combat the agricultural impacts of climate change have resulted in negative mental health outcomes among subsistence farmers in Burkina Faso [97]. Barriers comprised financial and time constraints, material and labour shortages, and inaccessible information; however, awareness of climate change was also reported to be limited.
Finally, an exploratory study surveying people affected by widespread public safety power shutoffs (PSPS) to reduce the risk of wildfires in California found that while people were mostly supportive of PSPS as an adaptation measure, it was associated with poorer physical and mental health [98].

Improving Decision-Support
Very few studies (n = 3) reported issues that could support improved decision-making around mental health and climate change. Between 2010 and 2012, the WHO Division of Pacific Technical Support led a regional climate change and health vulnerability assessment and adaptation planning project in collaboration with health sector partners in 13 Pacific Island countries. Mental health was identified as a high-priority climate-sensitive health risk in eight out of the 13 countries [99]. However, another study found that policy documents on climate change from 12 countries made little or no reference to nine vulnerable groups, including people with mental illness [100].
A study designed to develop and validate indices of adaptation to flooding found that people who perceived a risk of flooding in their home in the next five years adopted more preventative behaviours and adaptation behaviours than those who perceived little or no risk at all. Additionally, people who felt more adverse effects of flooding on their physical or mental health tended to adopt more adaptive behaviours than those who felt little or no adverse effects on their health [101].

Estimating the Costs of Protecting Health from Climate Change
We did not find any studies that assessed the costs associated with climate change impacts on mental health. However, one study from the United Kingdom found that a social prescribing service for mental health-a service model which supports people to access healthcare resources and psychosocial support, considered to hold potential for reduced cost and carbon-footprint-was associated with reduced financial costs but an increased carbon footprint per patient. However, none of the differences between groups reached statistical significance [102].

Quality of Studies
Overall, the studies identified in this review were of sound (fair to good) quality. Out of the 93 studies that were quality assessed, 75% were rated good, 23% fair and 2% poor. It is important to note that we found significant variability in how environmental exposures were measured. Among the quantitative studies, we identified two types of approaches to measure exposure to climate-sensitive environmental hazards: (i) self-reported exposures using questionnaires and interviews (note that all qualitative studies used self-reported exposures) and (ii) interpolated exposures using external sources of data.
Fifty-two studies relied on self-reported information to assign a climate change-related exposure to study participants. We found large variability across the survey questions used to capture self-reported exposures, as no standardised questionnaire exists (in contrast to self-reported mental health outcomes) to measure subjective experience regarding climatesensitive environmental hazards. Specifically, each study in our selected sample used a different question or metric to capture the environmental exposure. Such variability makes the evaluation of the robustness of used tools and the comparison between studies difficult. To the best of our knowledge, we did not identify any self-reported tool regarding the exposures of interest that were validated beforehand.
In addition, 48 studies relied on interpolated exposures using various types of available meteorological data sources. Some studies used meteorological data (mostly precipitation and temperature) from monitoring stations in the area of interest, often retrieved from the national or regional bureau of meteorology or data service centres. Other studies relied on remote sensing data from various satellite products with reanalysis, including the PRISM Climate Group [103] or the Berkeley Earth Surface Temperature Dataset [104]. While some studies used the meteorological variable directly in their analysis, such as daily maximum temperature, others further calculated indices or metrics to define weather events such as drought and heat waves. Most studies used the finest spatial resolution that was available while some used large-scale measurements when assigning exposure to meteorological variables, which may lead to important misclassification biases and impact effect estimates of interest. Indeed, assigning the same exposure to temperature, for example, on a given day to all individuals in a country or region is unlikely to accurately represent individuals' exposure given well-documented local variations in microclimates [105,106]. We also found differences in the temporal scales that were used across studies. While most studies relied on daily estimates to capture acute effects on mental health outcomes, other studies used monthly or annual estimates to capture chronic impacts, describing various mechanisms such as economic insecurity or migration. Finally, some studies provided limited details regarding data sources and how indices were calculated, making it difficult to understand how exposures were estimated and hindering reproducibility efforts.

Discussion
This scoping review is the first to explore the existing original research literature that investigates climate change and mental health using WHO's global research priorities as a framework [8]. We identified 120 original studies published between 2001 and 2020 that specifically referenced climate change impacts, adaptation, mitigation, and other interventions relevant to mental health. Most studies were quantitative, used a crosssectional design, were conducted in high-income countries, and were concerned with assessing the mental health risks associated with climate change-related exposures.
The literature consistently points to the negative associations that climate changerelated events have with individuals' and communities' mental health. Climate changerelated events were shown to be associated with psychological distress, worsened mental health (particularly among people with pre-existing mental health conditions), increased psychiatric hospitalisations, higher mortality among people with mental illness, and heightened suicide rates. These results are largely in line with the detrimental impacts that climate change has been shown to have on many other aspects of health [2].
While the overwhelming focus on the mental health risks of climate change is understandable and has generated insight and traction for advocating the importance of considering mental health within climate change, only four studies have focused on protective factors or the coping mechanisms people and communities have used when responding to the detrimental mental health impacts of climate change [26,38,90,105]. An in-depth understanding of which factors constitute 'resilience' in the face of climate change would contribute to a more nuanced picture of the associations between climate change and mental health [107] and could be useful when devising programs and policies for those most heavily impacted by climate change.
Additionally, the focus on characterising and quantifying the linkages between climate change and mental health has not been matched by applied research on addressing and reducing these associated mental health risks [8]. Future work should investigate more systematically the possible effectiveness, scalability, and cost-effectiveness of interventions and policies aimed at safeguarding mental health in the face of climate change, including non-mental health interventions. Various health professionals, including nurses [108], emergency department clinicians [109,110], clinical psychologists, and psychiatrists [111,112], have already outlined ways to address mental health problems related to climate change within clinical practice and broader policy, highlighting the strong willingness to act in the medical community. There is also a growing movement of climate advocacy in the health sector, recognising the responsibility to protect the health of current and future generations and calling for meaningful government action [113].
The current review points to some important avenues for future research. As mentioned above, one key research priority is to dissect the association between mental health and climate change by investigating possible mediators and moderators as well as risk, vulnerability, and resilience factors in different populations. A systems-thinking approach, as advocated by Berry et al., might be particularly suitable for this endeavour [114]. This will be aided by more precise operationalization and conceptualization of environmental and mental health variables in the climate change space, another important research priority. Recent work to devise precise measurement instruments for constructs such as climate anxiety [115] represents an important step in this direction. Additionally, in order to move beyond the assessment of risk, other important research priorities relate to understanding how climate change adaptation and mitigation strategies might impact mental health and wellbeing [98] and how engagement in different types of pro-environmental behaviours such as climate activism might impact mental health [116]. A more comprehensive discussion of future directions in climate change and mental health research and practice will be covered in a separate paper by the authors.
The current review has a number of implications for policymakers and intergovernmental organizations such as the World Health Organization. While WHO is increasingly directing its attention to the health impacts of climate change, mental health has generally remained absent from these conversations. For example, as part of the Health and Climate Change Country Profile Project WHO 2019-2020 cycle, WHO reported that only three (i.e., Antigua and Barbuda, United Arab Emirates, and Solomon Islands) out of the 10 recently reviewed countries included mental health and well-being outcomes as part of their climate-sensitive health outcomes monitoring systems [117]. This is in stark contrast with other conditions such as vector-borne and waterborne disease, for which monitoring systems were in place for nine out of the 10 recently reviewed countries. The current review highlights the need for monitoring systems to systematically include mental health as part of their indicators in order to ensure precise monitoring and availability of data at the national level. Additionally, it indicates the importance of including climate change as an important determinant of mental health in policy documentation and action plans, such as in the future extension of the WHO Mental Health Action Plan 2013-2020 [118] to 2030. Importantly, future work might want to explore whether existing WHO documentation and intervention plans for humanitarian settings such as the mhGAP-HIG [119], the Building Back Better framework [120], and scalable psychological interventions [121] are applicable in the context of climate change or whether more specific adaptations are needed. Finally, WHO's recent work on the integration of mental health and psychosocial support (MHPSS) within disaster risk reduction efforts [122] might represent an important opportunity to integrate climate change within MHPSS.
Palinkas et al. have reviewed the preparedness and response strategies from a mental health services perspective across three areas: (1) acute and extreme weather events; (2) sub-acute or long-term events; and (3) the prospect of long-term and permanent changes, such as an uninhabitable physical environment [122,123]. It is these long-term impacts, in particular, which may demand a shift in attention, as we found only one study which examined the mental health impacts of more general and permanent environmental degradation [124].
A substantial limitation in the surveyed literature concerns the under-representation of research from low-and middle-income countries. The majority of studies identified in this review were conducted in Australia, Canada, and the US. However, evidence indicates that low-and middle-income countries are more at risk of climate changerelated events and have fewer resources to respond to these stressors due to pre-existing vulnerabilities [125,126]. For example, according to the Climate Change Vulnerability Index, the five most vulnerable countries in the face of climate change globally in 2016 were Central African Republic, Democratic Republic of Congo, Haiti, Liberia, and South Sudan, all low-income countries considered at "extreme risk" in the face of climate change [127]. These vulnerable regions and populations are among the least responsible for the greenhouse gases that are warming the planet. The concept of climate justice highlights this double inequality of climate change, where there is an inverse distribution between risk and responsibility [128].
A possible limitation of the current study is that the search strategy used means that studies were included only if the authors explicitly linked the climate-related stressor they were considering and climate change. Indeed, a large amount of literature exists concerning the association between disasters and mental health [129] (without explicitly linking an increase in frequency and severity to climate change) which is likely to add value to the field. The geographical origin of the literature reviewed in the current paper is skewed towards high-income countries and unrepresentative of many regions of the world. The reasons for the relative abundance of research in countries such as Australia is worth exploring. Perhaps one reason for an under-representation of research from certain regions is that we only assessed studies in English. The decision to not explore non-English studies was related to our limited resources and capacity combined with the anticipation this extended search would yield limited studies.
It is also important to acknowledge that existing quality assessment tools are largely subjective and not perfectly adapted to the designs of studies included in this review and we were not able to assess the quality of studies employing modelling, case report, time-series, case-crossover, panel, or other unique study designs. Overall, our quality assessments highlight a need for improvement in future research-notably concerning improved measures of environmental exposures and mental health outcomes. For example, assessing the sensitivity and specificity of self-reported tools, for a given context and a given environmental hazard exposure, would be beneficial for assessing the reliability of such tools and allowing comparison across populations and studies.
The increasing number of studies on climate change and mental health parallels the growing traction in the broader field of climate change and health. Since 2007, the academic literature on this topic has tripled and media reports have increased by 78% [130]. However, mental health remains in a secondary position to physical health concerns, reflecting the state of global mental health more generally-despite research advances, the substantial burden of disease attributable to mental disorders has not improved, the quality of mental health services is routinely worse than services for physical health and government investment remains extremely limited [131]. This neglect is apparent in our findings and results in a continued lack of progress in improving the mental health of populations. The inattention to climate change impacts on mental health found in our scoping review echoes that noted by the Lancet Countdown report in 2018 [130]. This report noted that of the 16 national health adaptation strategies reviewed, mental health was the least considered climate-sensitive health outcome (mentioned by five out of 16 national health adaptation strategies).

Conclusions
The current scoping review has shown that climate change and mental health represents a rapidly growing area of research; however, it is underdeveloped and will need to accelerate and broaden in scope to respond with evidence-based mitigation and adaptation strategies. Most of the research has been devoted to assessing the mental health risks due to climate change, with less applied research investigating practical issues such as identifying the most effective interventions and policies to safeguard mental health in the face of climate change. Most of the research conducted to date is quantitative, cross-sectional, and conducted in high-income countries-therefore somewhat limiting the interpretability and the generalisability of the findings. Despite these limitations, the existing evidence overwhelmingly points to a negative association between climate change and mental health. Future research should aim at producing more robust and methodologically sound evidence on the link between mental health and climate change while also bearing in mind the importance of investigating what interventions, policies, and decision-making mechanisms can be put in place to mitigate the mental health impacts of climate change. This should represent a key priority in mental health research as, just like with physical health, climate change represents the biggest global threat to mental health of the 21st century [1]. Improved evaluation of current climate-related health risks, rather than a primary focus on risks over very long timeframes b.
Identification of vulnerable populations and life stages c.
Quantification of the fraction of morbidity and mortality attributable to climate hazards, and to climate change d.
Better assessment of neglected climate-mental health linkages (II) Identifying the most effective interventions a. Systematic reviews of the evidence base for interventions b.
Methodological research to improve analytical tools for cost-effectiveness analysis (III) Guiding health-promoting mitigation and adaption decisions in other sectors a. Improved methods for assessment of the health implications of decisions in other sectors b.
Health implications of climate change mitigation: energy and transport sectors c.
Health implications of climate change adaptation: water, food and agriculture sectors d.
Improved integration of climate change mitigation, adaptation, and health through "settings-based" research (IV) Improving decision-support a.
Research to improve vulnerability and adaptation assessments b.
Improvement of operational predictions c.
Improved understanding of decision-making processes (V) Estimating the costs of protecting health from climate change a. Definition of harmonized methods to estimate costs and benefits b.
Assessment of the health costs of inaction and the costs of adaptation c. Improved economic assessment of the health co-benefits of climate change mitigation