Impact statement
This systematic review is the first to comprehensively map and critically appraise interventions that target eco-anxiety across therapeutic, educational, community and nature-based settings. By synthesising evidence from 22 studies conducted in diverse contexts, it identifies which approaches have shown the most promising findings to date and where existing responses remain limited. For practitioners, policymakers and educators, this review offers a practical overview of intervention types, target groups and outcome measures that can guide the design of more effective, ethically responsible responses to climate-related distress. It highlights recurring components across diverse interventions, including emotional resilience skills, structured opportunities for collective action and community-based delivery. It pinpoints critical gaps, including the lack of culturally diverse evidence, limited community involvement in co-design and the near-absence of system-level or policy-linked initiatives. By framing eco-anxiety as both a mental health concern and a potentially adaptive response to real climate threats, this work supports the development of interventions that protect well-being while enabling constructive engagement with climate action. The review provides a foundation for developing standardised outcome measures, scaling up effective programmes beyond researcher-led pilots and prioritising investment in climate–mental health strategies in low- and middle-income settings. In doing so, it offers timely guidance for health systems, schools and community organisations seeking to respond to the mental health dimensions of the climate crisis in a coordinated and evidence-informed way.
Introduction
Addressing eco-anxiety is a mental health imperative and a critical component of achieving the United Nations Sustainable Development Goals (SDGs), particularly SDG 3 (Good Health and Well-being) and SDG 13 (Climate Action) (United Nations, 2024). Climate change threatens physical and mental health worldwide, and eco-anxiety, defined as a persistent worry and distress about environmental degradation, reflects the psychological dimension of this crisis (Pihkala, Reference Pihkala2020; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021). As the SDGs emphasise integrated approaches to health, education and environmental sustainability, addressing eco-anxiety aligns with global commitments to promote resilience and adaptive capacity in the face of climate-related risks. Recent evidence shows that eco-anxiety is widespread, especially among youth and environmentally engaged populations, and disproportionately affects vulnerable communities, reinforcing the need for equity-focused interventions (Niedzwiedz and Katikireddi, Reference Niedzwiedz and Katikireddi2023; Gallè et al., Reference Gallè, Valeriani, De Giorgi, Grassi, Mazzeo, Napoli and Protano2025; Magruder and Chipman, Reference Magruder and Chipman2025). International agencies, including the World Health Organization (WHO) and the Intergovernmental Panel on Climate Change, have identified mental health as a priority in climate adaptation strategies, underscoring the urgency of developing effective interventions (World Health Organization, 2022; Intergovernmental Panel on Climate Change, 2023).
To support conceptual clarity, this review adopts a working definition of eco-anxiety as climate-related psychological distress that may range from normative, contextually rational concern to more persistent and/or intense worry associated with emotional, cognitive or functional impairment. This framing distinguishes adaptive climate concern, which may motivate engagement and problem-solving, from maladaptive eco-anxiety characterised by sustained distress, rumination or impairment that warrants targeted support (Pihkala, Reference Pihkala2020; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021). This distinction is important for interpreting intervention outcomes, because some approaches may increase emotional activation or concern while simultaneously strengthening agency, coping or engagement; such patterns may reflect heightened salience, adaptive activation or measurement nuance rather than harm (Baudon and Jachens, Reference Baudon and Jachens2021; Becht et al., Reference Becht, Spitzer, Grapsas, van de Wetering, Poorthuis, Smeekes and Thomaes2024). Accordingly, we interpret interventions through three complementary pathways emphasised across the eco-anxiety and climate-mental health literature: (1) emotion regulation and resilience-building (e.g., strengthening distress tolerance and adaptive coping), (2) meaning-making and acceptance-oriented processes (e.g., supporting integration of climate emotions and realistic hope) and (3) agency- and connection-building (e.g., strengthening self-efficacy, collective efficacy and social support to enable constructive engagement) (Pihkala, Reference Pihkala2020; Baudon and Jachens, Reference Baudon and Jachens2021).
Despite this growing recognition of eco-anxiety, intervention research remains limited and difficult to translate into practice. Previous reviews indicate that the majority of research has concentrated on defining eco-anxiety and exploring its psychosocial effects (Baudon and Jachens, Reference Baudon and Jachens2021; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Boluda-Verdú et al., Reference Boluda-Verdú, Senent-Valero, Casas-Escolano, Matijasevich and Pastor-Valero2022), while there has been far less emphasis on systematically analysing evaluated interventions, specifically, detailing what has been implemented, how these interventions are structured and delivered, and the outcomes they produce. For example, Baudon and Jachens (Reference Baudon and Jachens2021) mapped therapeutic and supportive approaches and identified recurring themes (e.g., fostering inner resilience, group support, action-oriented coping, practitioner training and nature connection), but also highlighted that the underpinning evidence was largely conceptual with very few empirical evaluations of intervention effectiveness. Similarly, broader syntheses of climate-related distress interventions report limited methodological rigour and substantial heterogeneity in outcome assessment, restricting comparison across studies and hindering cumulative learning (Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024). Measurement challenges further constrain progress: while several eco-anxiety and climate-distress scales exist, they are inconsistently applied and often lack robust psychometric validation across diverse cultural contexts, limiting confidence in observed intervention effects (Kırımer-Aydınlı et al., Reference Kırımer-Aydınlı, Juaréz-Castelán, Hakim, Gul, Unal, Aguayo-Estremera, Pérez-Fortis, Rojas-Russell and Gallo2025; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025).
Taken together, these limitations point to the need for a synthesis that moves beyond cataloguing intervention types to systematically examine how interventions are designed, evaluated and implemented, as well as the tools used to assess change. In particular, the roles of key interest-holders, such as educators, clinicians, community leaders and policymakers, in shaping and delivering eco-anxiety interventions remain insufficiently examined, limiting understanding of feasibility, scalability and contextual fit. Addressing these gaps is essential for informing more effective and context-sensitive responses. Accordingly, this systematic review synthesises evidence on the design, effectiveness and implementation of interventions aimed at addressing eco-anxiety across diverse populations and settings guided by the following research questions:
-
• What interventions or strategies have been implemented or proposed to address eco-anxiety?
-
• What measurement tools are used to assess the outcomes of interventions aimed at addressing eco-anxiety, and what are their psychometric strengths and limitations?
-
• How effective are these interventions in addressing eco-anxiety, and what aspects of exposure and outcomes do they target?
-
• What barriers and facilitators influence the implementation of intervention delivery and outcomes related to eco-anxiety?
Methodology
This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement (PRISMA) (Page et al., Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann, Mulrow, Shamseer, Tetzlaff, Akl, Brennan, Chou, Glanville, Grimshaw, Hróbjartsson, Lalu, Li, Loder, Mayo-Wilson, McDonald, McGuinness, Stewart, Thomas, Tricco, Welch, Whiting and Moher2021). The PRISMA checklist is provided in Supplementary Table S1. Methods were pre-specified and documented in advance in a protocol that was published on the PROSPERO database for systematic reviews (Registration no.:CRD420251138000).
Search strategy
A comprehensive search of three EBSCOhost databases (Cumulative Index of Nursing and Allied Health Literature [CINAHL] plus, Literature Analysis and Retrieval System Online [MEDLINE] complete databases and PsycINFO), ProQuest Environmental Science Index, Scopus and the Cochrane Central Register of Controlled Trials (CENTRAL). The strategy was developed between the authors and a research librarian trained in conducting systematic reviews. The search strategy was informed by previous studies that examined eco-anxiety and interventions for the treatment of eco-anxiety (Baudon and Jachens, Reference Baudon and Jachens2021; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Boivin et al., Reference Boivin, Gousse-Lessard and Hamann Legris2025) (Supplementary Table S2). The search strategy was tailored for each database to locate original research involving human participants, beginning on the 1st January 2010 and extending until 30th September 2025 (the date the search was conducted). The 2010 start date reflects the period in which eco-anxiety began to be conceptualised as a distinct form of climate-related psychological distress, with emerging theoretical frameworks and measurement approaches that enabled later intervention research (Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Boivin et al., Reference Boivin, Gousse-Lessard and Hamann Legris2025). An additional hand search of the reference lists of included studies was undertaken.
Eligibility criteria
Studies were included in this review if they were published in English or Spanish. The eligibility criteria are based on the Population/problem, Interest, Context, Scope and Time (PICoST) approach (Table 1) (Vargas et al., Reference Vargas, Brooke, Venegas Hargous, Fernández-Rodríguez, Bennett, Jacka and Allender2026). The review focused on interventions explicitly addressing eco-anxiety or closely related climate-related psychological distress in general or at-risk populations. Studies centred on pre-existing mental health diagnoses or current specialist treatment pathways were excluded to maintain a consistent focus on eco-anxiety-specific or climate-distress-specific intervention responses.
Inclusion and exclusion criteria
Table 1. Long description
The table consists of three columns: Category, Inclusion, and Exclusion.
* Population: Inclusion covers individuals of any age with eco-anxiety or climate distress in noninstitutionalized populations. Exclusion covers animals, purely ecological systems, and those with pre-existing mental health diagnoses.
* Intervention(s) or exposure(s): Inclusion covers interventions like nature-based therapy, C B T, mindfulness, and community engagement designed to manage eco-anxiety. Exclusion covers general mental health interventions without a climate link, studies measuring triggers, or mitigation of P T S D.
* Context: Inclusion covers any geographical, cultural, or institutional setting. Exclusion covers settings unrelated to mental health or community engagement, and individuals in patient mental health services.
* Study design: Inclusion covers protocols and empirical studies like R C Ts, quasi-experimental, and longitudinal studies. Exclusion covers editorials, opinion pieces, theses, and reviews.
* Time: Inclusion is from 1st January 2010. Exclusion is publications prior to 2010.
A footnote defines intervention as any support aiming to protect psychosocial well-being or treat mental disorders.
^ Intervention is defined as “any type of local or outside support that aims to protect or promote psychosocial well-being and/or prevent or treat mental disorder” (Cimprich and Ronis Reference Cimprich and Ronis2003).
Study selection
The search results were uploaded to Covidence (Veritas Health Innovation, 2022), where duplicates were removed, and all screening was conducted. Two researchers independently screened all article titles, abstracts and full texts. A third researcher resolved all conflicts at both stages (titles and abstracts, and full text).
Data extraction
A data extraction template was developed to extract data from each article. The extracted information included: (1) general information: title, year, lead author and country; (2) characteristics of included studies: aim of study, study design (3) intervention characteristics: intervention description (activities), study duration, target setting, participant details, target exposure, outcomes, tools used to assess outcomes and interest-holder roles; (4) results: main findings, reported strengths and limitations, implications and gaps for future research. Data extraction and quality assessment were conducted by two authors under the supervision of a second author. All the extracted data were exported from Covidence software (Veritas Health Innovation, 2022) to a Microsoft Excel spreadsheet.
Quality assessment
A quality assessment of included studies was conducted using the Quality Assessment Tool for Studies with Diverse Designs (QuADS), a 13-item tool scored 0–3 per item to support appraisal across heterogeneous methodologies (Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). QuADS does not specify a cut-off for classifying studies as high or low quality; therefore, studies were not excluded based on QuADS scores. Instead, QuADS ratings were used to characterise the overall strength of the evidence base and to qualify interpretation of findings by identifying features likely to limit interpretability (e.g., single-group pre–post designs, small samples, short exposure windows, limited follow-up and weaker reporting of theoretical underpinning or stakeholder involvement). Given heterogeneity of interventions and outcomes, we did not apply any grading system; instead, we summarised confidence narratively based on consistency of effect direction, design strength, sample size, follow-up and key QuADS reporting domains. QuADS was used to support appraisal across diverse study designs, but it does not replace design-specific risk-of-bias assessment or formal certainty grading, and its findings were interpreted accordingly.
Results
A total of 1,259 articles were considered for title and abstract screening after duplicates were removed. Of these, 140 studies were fully assessed against the eligibility criteria, and 118 were excluded for the reasons described in Supplementary Table S3. Finally, 22 articles were included in our systematic review. A PRISMA flow diagram (Figure 1) illustrates the search results and data collection process.
PRISMA diagram of the systematic review process for this review.
Figure 1. Long description
The flowchart is organized into three vertical phases labeled on the left: Identification, Screening, and Included.
1. Identification Phase:
- Top-left box: Records identified from ProQuest Agricultural and Environmental Collection (n = 566), PsycINFO (n = 391), Medline (n = 290), Scopus (n = 249), C I N A H L (n = 150), Cochrane Central Register of Controlled Trials (n = 12), and Manually added (n = 9).
- An arrow points right to: Records removed before screening: Duplicate records removed (n = 408).
2. Screening Phase:
- Vertical flow continues to: Records screened (n = 1259).
- An arrow points right to: Records excluded by title-abstract screening (n = 1118).
- Vertical flow continues to: Reports sought for retrieval (n = 141).
- An arrow points right to: Reports not retrieved (n = 1).
- Vertical flow continues to: Reports assessed for eligibility (n = 140).
- An arrow points right to: Reports excluded (n = 118) due to No intervention (n = 73), Wrong publication (n = 22), Wrong outcome (n = 22), and Wrong setting (n = 1).
3. Included Phase:
- Final vertical flow leads to the bottom box: Studies included in the review (n = 22).
Characteristics of included studies
Of the 22 included studies, a subset were protocol or feasibility studies that did not report outcome data. These studies were included to characterise intervention approaches but were not considered when describing effectiveness patterns. Across the 22 included studies, the evidence base is recent and geographically concentrated. Studies were published between 2017 and 2025, with most appearing from 2023 to 2025 (n = 17; 77%) and were conducted predominantly in high-income Western countries. The United States of America (USA) accounted for the largest number of publications (Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Trott, Reference Trott2021; Goldwert et al., Reference Goldwert, Dev, Broos, Broad and Timpano2024; Meixner et al., Reference Meixner, Ciancarelli, Farrell, García, Josek, Kelly, Meister, Soule and Darner2024; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025; Bodas et al., Reference Bodas, Chlebuch and Weisberg2025; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a; Scott et al., Reference Scott, Powell and Lee-Johnson2025) (n = 10; 46%), followed by Australia (Jones et al., Reference Jones, Hine and Marks2017; Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Coates et al., Reference Coates, Brown and Kelly2025; Cosh et al., Reference Cosh, Bartik, Ryan, Jefferys, Fallander, Tully and Lykins2025) (n = 4; 18%), Egypt (Elzohairy et al., Reference Elzohairy, Fatah and Khedr2024; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025) (n = 2; 9%), Sweden (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023, Reference Lindhe, Berg, Andersson and Andersson2024) (n = 2; 9%) and other European countries (Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Eklöf and Klöckner, Reference Eklöf and Klöckner2025; Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025) (n = 4; 18%). Table 2 and Supplementary Table S4 summarise the included studies. These patterns indicate that evaluated eco-anxiety interventions are concentrated within a recent and still developing evidence base and are unevenly distributed across global contexts, with limited representation in low- and middle-income countries.
Outline of the interventions from the included articles (n = 22)
Table 2. Long description
The table is organized into four thematic sections.
1. Educational Interventions: Includes studies by Stevenson et al. 2018, Lawson et al. 2019, Trott 2021, Howard-Jones et al. 2023, Meixner et al. 2024, Bodas et al. 2025, and Eweida et al. 2025. These focus on school curricula, video modules, and citizen science. Key results show increases in knowledge and self-efficacy, though links to behavior change are mixed. Lawson et al. specifically noted intergenerational learning from children to parents.
2. Messaging Interventions: Includes Jones et al. 2017, Asbrand et al. 2025, Benoit et al. 2025, Coates et al. 2025, Ecker et al. 2020, and Goldwert et al. 2024. These examine how framing (proximal vs. distal, high vs. low agency, positive vs. negative) affects psychological distance and concern. High agency and proximal framing generally reduced anxiety and increased engagement.
3. Therapeutic Interventions: Includes Lindhe et al. 2023 and 2024, Eklöf and Klöckner 2025, Epel et al. 2025 a, and Scott et al. 2025. These utilize I C B T, mindfulness, and group psychoeducation. Results indicate significant reductions in depression, stress, and climate distress, with improvements in emotional resilience.
4. Nature-based Hybrid Interventions: Includes Eichholtzer et al. 2024, Elzohairy et al. 2024, Cosh et al. 2025, and Glynn et al. 2025. These combine education with nature immersion or restoration. Findings show increased nature relatedness and climate capability, though some interventions also led to increased eco-anxiety as awareness grew.
Methodological approaches were heterogeneous (Table 2), but randomised control trials (RCTs) (Jones et al., Reference Jones, Hine and Marks2017; Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023; Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025; Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025; Scott et al., Reference Scott, Powell and Lee-Johnson2025) were the most common design (n = 9; 41%), followed by experimental designs without full randomisation, such as pre-post-test treatment-control designs and quasi-experimental approaches (Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Elzohairy et al., Reference Elzohairy, Fatah and Khedr2024; Bodas et al., Reference Bodas, Chlebuch and Weisberg2025; Coates et al., Reference Coates, Brown and Kelly2025) (n = 5; 23%); mixed-methods designs (Trott, Reference Trott2021; Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023; Meixner et al., Reference Meixner, Ciancarelli, Farrell, García, Josek, Kelly, Meister, Soule and Darner2024; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a) (n = 4; 18%), combining quantitative pre/post assessments with qualitative reflections or interviews, often across multiple sites or longitudinally and qualitative research (Lindhe et al., Reference Lindhe, Berg, Andersson and Andersson2024; Eklöf and Klöckner, Reference Eklöf and Klöckner2025) (n = 2; 9%) exploring participant experiences, while observational designs (Goldwert et al., Reference Goldwert, Dev, Broos, Broad and Timpano2024; Cosh et al., Reference Cosh, Bartik, Ryan, Jefferys, Fallander, Tully and Lykins2025) included cross-sectional studies and exploratory pilot studies with single-group pre/post assessments (n = 2; 9%). Studies varied in the outcomes measured and how change was assessed, and the types of inferences that could be made across studies.
Interventions/strategies implemented to address eco-anxiety
Across the 22 included studies, interventions did not cluster into a single modality. Instead, they spanned therapeutic, educational, messaging/communication and nature/community-based approaches, with several interventions combining emotional coping elements with engagement or agency-building components. Educational interventions were most common (n = 7; 32%) (Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Trott, Reference Trott2021; Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023; Meixner et al., Reference Meixner, Ciancarelli, Farrell, García, Josek, Kelly, Meister, Soule and Darner2024; Bodas et al., Reference Bodas, Chlebuch and Weisberg2025; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025), followed by messaging/communication approaches (n = 6; 27%) (Jones et al., Reference Jones, Hine and Marks2017; Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Goldwert et al., Reference Goldwert, Dev, Broos, Broad and Timpano2024; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025; Coates et al., Reference Coates, Brown and Kelly2025), therapeutic interventions (n = 5; 23%) (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023, Reference Lindhe, Berg, Andersson and Andersson2024; Eklöf and Klöckner, Reference Eklöf and Klöckner2025; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a; Scott et al., Reference Scott, Powell and Lee-Johnson2025) and nature based−/hybrid interventions (n = 4; 18%) (Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Elzohairy et al., Reference Elzohairy, Fatah and Khedr2024; Cosh et al., Reference Cosh, Bartik, Ryan, Jefferys, Fallander, Tully and Lykins2025; Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025).
Across these categories, intervention aims can be summarised into three recurring “action logics”:
The first focused on distress reduction and coping support, targeting eco-anxiety symptoms through psychological skills or structured support (Elzohairy et al., Reference Elzohairy, Fatah and Khedr2024; Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023, Reference Lindhe, Berg, Andersson and Andersson2024; Eklöf and Klöckner, Reference Eklöf and Klöckner2025; Epel et al., Reference Epel, White, Brownell, Rodin, Hollis, Diefenbach, Alegria, Fromer, Czajkowski, Bacon, Revenson, Ruiz and Maibach2025b). The second focused on capability and agency building, aiming to strengthen self-efficacy, hope and engagement (Jones et al., Reference Jones, Hine and Marks2017; Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023; Meixner et al., Reference Meixner, Ciancarelli, Farrell, García, Josek, Kelly, Meister, Soule and Darner2024; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025). The third focused on connection-based resilience, using group, community or nature-based approaches to enhance belonging and social support (Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Cosh et al., Reference Cosh, Bartik, Ryan, Jefferys, Fallander, Tully and Lykins2025; Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025; Scott et al., Reference Scott, Powell and Lee-Johnson2025). This functional clustering reflects variation in intervention priorities, with some interventions focused primarily on symptom reduction and others targeting engagement, capability or agency.
Measurement tools used to assess intervention outcomes
Outcome measurement was highly heterogeneous across studies and included three broad categories of constructs: (1) distress or impairment-focused measures, including eco-anxiety/climate-distress specific scales and general mental health measures; (2) concern or emotional activation measures and (3) agency-, engagement-, hope- or self-efficacy-related measures. Eco-anxiety and climate distress were assessed using instruments such as the Climate Anxiety Scale and Climate Change Anxiety Scale, alongside more general measures such as the Depression Anxiety Stress Scale (DASS). Social dimensions, such as perceived support or community connectedness, were less commonly measured. Variation in outcome measures and inconsistent use of eco-anxiety-specific instruments limited comparability across studies and informed the decision not to formally group intervention effects (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023; Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Elzohairy et al., Reference Elzohairy, Fatah and Khedr2024; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025; Coates et al., Reference Coates, Brown and Kelly2025; Cosh et al., Reference Cosh, Bartik, Ryan, Jefferys, Fallander, Tully and Lykins2025; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a; Scott et al., Reference Scott, Powell and Lee-Johnson2025).
Beyond symptom-focused outcomes, studies frequently assessed constructs aligned with four hypothesised mechanisms of change: (1) Hope/agency and emotional constructs (e.g., hope scales, agency ratings, subjective distress) (Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Trott, Reference Trott2021; Goldwert et al., Reference Goldwert, Dev, Broos, Broad and Timpano2024; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025). (2) Knowledge/cognitive outcomes (climate knowledge/competence, module assessments) (Stevenson et al., Reference Stevenson, Peterson and Bondell2018; Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Meixner et al., Reference Meixner, Ciancarelli, Farrell, García, Josek, Kelly, Meister, Soule and Darner2024; Bodas et al., Reference Bodas, Chlebuch and Weisberg2025). (3) Behavioural/self-efficacy and engagement outcomes (pro-environmental intentions/actions, self-efficacy, policy support/behavioural engagement indices) (Jones et al., Reference Jones, Hine and Marks2017; Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Goldwert et al., Reference Goldwert, Dev, Broos, Broad and Timpano2024; Coates et al., Reference Coates, Brown and Kelly2025; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025; Scott et al., Reference Scott, Powell and Lee-Johnson2025). (4) Social dimensions were less commonly measured directly (e.g., perceived social support/community connectedness), despite frequent discussion of peer/group mechanisms (Lindhe et al., Reference Lindhe, Berg, Andersson and Andersson2024; Eklöf and Klöckner, Reference Eklöf and Klöckner2025; Scott et al., Reference Scott, Powell and Lee-Johnson2025). Variation in measurement and inconsistent use of eco-anxiety-specific instruments limited cross-study comparability and informed the decision not to formally group intervention effects.
Intervention effectiveness in addressing eco-anxiety and related climate distress outcomes
Of the 22 included studies, some were protocol or feasibility studies that did not report outcome data. These studies were included to describe intervention approaches but were not used to inform effectiveness findings. Across intervention types, approaches targeting coping, emotional regulation and resilience were most consistently associated with improvements in distress-related outcomes compared with interventions focused on knowledge or capability. The included studies differed in intervention type, outcome measures and follow-up periods. For this reason, meta-analysis was not conducted, and effectiveness patterns were summarised descriptively.
Distress reduction interventions were most consistently associated with improvements in distress-related outcomes. Therapeutic approaches, particularly CBT/ACT-informed or mindfulness/acceptance-based programmes, most consistently reported reductions in climate distress and related mental health outcomes (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a). For example, tailored internet-delivered CBT reduced depression, stress and climate distress and improved quality of life in a pilot RCT (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023) and a psychosocial climate resilience course reported significant reductions in climate distress/anxiety and stress alongside increased self-efficacy, with improvements maintained at 5-month follow-up (Lindhe et al., Reference Lindhe, Bengtsson, Byggeth, Engström, Lundin, Ludvigsson, Aminoff, Berg and Andersson2023; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a).
Educational and community approaches were more likely to be associated with favourable wellbeing-related outcomes when they incorporated emotional engagement, agency or peer support. Educational interventions showed mixed effects on eco-anxiety and worry outcomes, but a consistent pattern was that interventions embedding emotional engagement, agency or peer support tended to align with more favourable wellbeing-related outcomes than knowledge-only approaches (Trott, Reference Trott2021; Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023; Eweida et al., Reference Eweida, OʼConnell, Zoromba, Mahmoud, Altheeb, Selim and Atta2025). Some curricula reported slight decreases or no change in anxiety while increasing self-efficacy (Howard-Jones et al., Reference Howard-Jones, Warren-Lee and Aldred2023), whereas other media-based content increased environmental worry (Bodas et al., Reference Bodas, Chlebuch and Weisberg2025), indicating information alone may increase concern without providing coping or agency pathways.
Messaging/framing interventions produced mixed and sometimes null effects on eco-anxiety. Messaging studies varied widely (agency framing; imagery valence; issue framing) and yielded inconsistent effects on anxiety outcomes (Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025; Coates et al., Reference Coates, Brown and Kelly2025). Agency-promoting framing reduced state anxiety in one experimental study (Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025), whereas imagery and framing manipulations did not consistently change climate anxiety in others (Ecker et al., Reference Ecker, Butler, Cook, Hurlstone, Kurz and Lewandowsky2020; Coates et al., Reference Coates, Brown and Kelly2025). Brief messaging appeared more reliable for shifting proximal constructs such as hope/agency than for altering distress-related outcomes within short exposure windows (Benoit et al., Reference Benoit, Lowe, Thomas, Amsalem and Martin2025).
Some interventions increased eco-anxiety alongside increased capability or engagement. A subset of interventions reported increases in eco-anxiety alongside improvements in capability or engagement-related outcomes (Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025). For instance, a nature-based intervention increased climate capability while also increasing eco-anxiety scores, with change scores correlated (Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025).
Barriers and facilitators that influence implementation
Implementation reporting was uneven, but cross-cutting constraints were evident. Many interventions were researcher-led and time-bound, with common challenges related to scalability, cultural relevance and sustainability beyond pilot contexts (Eichholtzer et al., Reference Eichholtzer, Driscoll, Patrick, Galletta and Lawson2024; Epel et al., Reference Epel, Mishra, Ekman, Ogunseitan, Fromer, Kho, Grialou and Goldin2025a; Scott et al., Reference Scott, Powell and Lee-Johnson2025). Where facilitators were described, they frequently involved social and participatory features. Group formats and peer support reduced isolation and validated emotions in qualitative accounts (Lindhe et al., Reference Lindhe, Berg, Andersson and Andersson2024; Eklöf and Klöckner, Reference Eklöf and Klöckner2025), and structured opportunities for action/agency were emphasised as important for sustaining engagement (Trott, Reference Trott2021; Asbrand et al., Reference Asbrand, Spirkl, Reese, Spangenberg, Shibata and Dippel2025). However, explicit reporting of the roles of educators, clinicians, community leaders and policymakers as co-designers or delivery partners remained limited, consistent with lower QuADS ratings for stakeholder involvement across studies.
Quality assessment
QuADS appraisal indicated moderate overall methodological quality across the included studies, with most scoring relatively well on the appropriateness of methods to address study aims and on the clarity of analysis, but weaker reporting in domains critical for interpretability and transferability. Overall, 18 studies (82%) averaged between 2 and 2.8 (out of 3) across QuADS items, while 4 studies (18%) averaged below 2. Two recurrent limitations were (i) limited reporting of theoretical or conceptual underpinnings and (ii) limited consideration of stakeholder involvement in research design or conduct, with only 11 studies scoring ≥2 for theoretical underpinning and 7 studies scoring ≥2 for stakeholder involvement. These gaps limit the extent to which mechanisms of change, generalisability and implementation readiness can be assessed across the included studies. QuADS findings were considered when interpreting effectiveness patterns, particularly where study design, sample size, reporting or follow-up limited confidence in the findings. Confidence was highest where improvements were observed in studies using stronger study designs (e.g., controlled trials) with clearer reporting, and more cautious where findings were derived from small samples, short-duration interventions or designs with limited follow-up. Consistent with these quality patterns, evidence about longer-term impact, cultural adaptation and co-designed delivery remains comparatively underdeveloped, reinforcing the need for more rigorous, participatory and longer-term evaluations. Individual study quality scores are provided in Supplementary Table S5.
Discussion
This systematic review synthesised how interventions explicitly addressing eco-anxiety are designed, evaluated, implemented and measured across diverse populations and settings. Across 22 included studies, most published between 2023 and 2025 and conducted predominantly in high-income countries, particularly the United States and Australia, we identified a recent but methodologically mixed evidence base spanning therapeutic, educational, messaging and nature- or community-based approaches. Interventions were commonly researcher-led, time-bound and delivered within single settings, with few leveraging broader community partnerships or more embedded system-level approaches. Positive findings within the current evidence base were most consistently reported for interventions targeting coping, emotional regulation and resilience, whereas educational and messaging approaches showed more variable results and, in some cases, were associated with increased emotional engagement or concern (Baudon and Jachens, Reference Baudon and Jachens2021; Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024). Outcome measurement was highly heterogeneous, and most studies provided limited follow-up, restricting confidence in cross-study comparability and longer-term effects (Kırımer-Aydınlı et al., Reference Kırımer-Aydınlı, Juaréz-Castelán, Hakim, Gul, Unal, Aguayo-Estremera, Pérez-Fortis, Rojas-Russell and Gallo2025; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025). Taken together, these findings indicate that the evidence base is promising but still preliminary, and that intervention effects should be interpreted in relation to intervention aims and outcome type rather than assuming symptom reduction as the only relevant endpoint (Pihkala, Reference Pihkala2020; Baudon and Jachens, Reference Baudon and Jachens2021).
A major finding of this review is the geographic concentration of the available evidence in high-income Western settings. Although eco-anxiety is globally relevant, most evaluated interventions were conducted in the United States, Australia or Europe, with very limited representation from low- and middle-income countries or regions experiencing severe climate-related disruption (Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022). This limits the generalisability of the current evidence base. Eco-anxiety measured in high-income student or general population samples may differ meaningfully from climate-related distress in settings where environmental change is closely linked to drought, displacement, livelihood disruption or constrained access to mental health care (Berry et al., Reference Berry, Waite, Dear, Capon and Murray2018; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Boluda-Verdú et al., Reference Boluda-Verdú, Senent-Valero, Casas-Escolano, Matijasevich and Pastor-Valero2022; IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022). Cultural variation may also shape how climate-related emotions are expressed, interpreted and supported, with implications for both measurement validity and intervention acceptability (Berry et al., Reference Berry, Waite, Dear, Capon and Murray2018; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Kırımer-Aydınlı et al., Reference Kırımer-Aydınlı, Juaréz-Castelán, Hakim, Gul, Unal, Aguayo-Estremera, Pérez-Fortis, Rojas-Russell and Gallo2025; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025). The current literature therefore provides only partial insight into how eco-anxiety interventions may need to be adapted for highly climate-affected and resource-constrained contexts. In these settings, delivery feasibility is also likely to depend on scalable, low-intensity and contextually grounded approaches, such as group-based, community-led, task-shared or digitally supported formats, alongside equitable research partnerships that prioritise under-represented regions (Inter-Agency Standing Committee, 2007; IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022; World Health Organization, 2022).
Our review also reinforces the importance of conceptual clarity in how eco-anxiety is framed. The included interventions reflected both clinical and non-clinical understandings of eco-anxiety, consistent with broader literature distinguishing maladaptive eco-anxiety from more adaptive climate concern (Pihkala, Reference Pihkala2020; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021). Therapeutic approaches such as CBT-, ACT- and mindfulness-informed programmes were more often associated with improvements in distress-related outcomes, suggesting that some forms of eco-anxiety may benefit from structured psychological support (World Health Organization, 2019; Fordham et al., Reference Fordham, Sugavanam, Edwards, Stallard, Howard, das Nair, Copsey, Lee, Howick, Hemming and Lamb2021). At the same time, several interventions were designed not to reduce climate concern itself, but to strengthen knowledge, hope, agency, resilience or engagement (Baudon and Jachens, Reference Baudon and Jachens2021; Martin et al., Reference Martin, Roswell and Cosma2024). In this context, increases in eco-anxiety or climate concern do not necessarily indicate harm. Rather, they may reflect heightened salience, emotional activation or deeper engagement with climate realities, particularly when accompanied by improvements in self-efficacy or capability-related outcomes (Becht et al., Reference Becht, Spitzer, Grapsas, van de Wetering, Poorthuis, Smeekes and Thomaes2024; Marinkovic Chavez et al., Reference Marinkovic Chavez, Quinn, Gibbs, Block, Leppold, Stanley and Vella-Brodrick2024). This distinction is important because it suggests that intervention effects should be interpreted according to whether the measured outcome reflects impairment, emotional concern or adaptive engagement, and whether the intervention was intended to reduce distress, strengthen coping or foster constructive action (Pihkala, Reference Pihkala2020; Baudon and Jachens, Reference Baudon and Jachens2021). At the same time, the reviewed studies also indicate that eco-anxiety does not arise only at the level of individual cognition or emotion; it is shaped by broader environmental and social conditions. This suggests that individual-level interventions may be beneficial but are unlikely to be sufficient if disconnected from opportunities for collective efficacy, social support or credible pathways for action (Lawson et al., Reference Lawson, Stevenson, Nils, Carrier, Renee and Seekamp2019; Glynn et al., Reference Glynn, Dwyer, Casey, Sharma, Harrold, OʼConnor, Harrold and Glynn2025).
Another important finding concerns implementation. Across the included studies, interventions were often researcher-led, bounded to specific settings and evaluated over relatively short timeframes. Reporting on feasibility, scalability, cultural adaptation and sustainability was uneven (World Health Organization, 2022; Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024). Where facilitators were described, these commonly included peer support, group-based formats and opportunities for action or agency (Baudon and Jachens, Reference Baudon and Jachens2021; Patrick et al., Reference Patrick, Noy, Henderson-Wilson, Hayward, Allender, Gunasiri, Bruges, Snell and Capetola2023). In contrast, there was limited explicit reporting of the roles of educators, clinicians, community leaders or policymakers in intervention co-design or delivery, consistent with weaker reporting on stakeholder involvement in the QuADS appraisal (Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). These patterns suggest that, although a range of intervention approaches is emerging, many have not yet been evaluated in forms that are implementation-ready or readily transferable beyond pilot or research contexts (World Health Organization, 2022; Patrick et al., Reference Patrick, Noy, Henderson-Wilson, Hayward, Allender, Gunasiri, Bruges, Snell and Capetola2023). In conjunction with broader guidance emphasising community assets, workforce capacity and coordination across services, these findings indicate that future intervention research should pay greater attention to how eco-anxiety responses can be delivered feasibly and sustainably beyond small-scale, researcher-led pilots(Black et al., Reference Black, Jamrozik, Khan, Peralta, Havenga, Merritt and Kelley2024; World Health Organization, 2024; Lee-Johnson et al., Reference Lee-Johnson, Scott and Powell2025).
Our review further highlights that confidence in the evidence base remains constrained by limitations in study design and reporting. Although QuADS was appropriate for appraisal of this heterogeneous literature, the included studies varied substantially in methodological strength, and no formal certainty grading framework was applied (Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). Confidence was greater where findings were observed in controlled studies with clearer reporting and some follow-up, and lower where evidence was derived from small samples, single-group designs, short exposure windows or protocol-only studies (Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025). Two recurrent weaknesses across studies were limited reporting of theoretical or conceptual underpinnings and limited stakeholder involvement in design or conduct (Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). These features do not diminish the early positive signals emerging in the field; however, they warrant caution in drawing conclusions regarding mechanisms of change, implementation readiness and comparative effectiveness (Baudon and Jachens, Reference Baudon and Jachens2021; Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024). Overall, the review supports a narrative of emerging promise rather than settled evidence.
Limited follow-up across the included studies also restricts confidence in the durability of intervention effects. This is particularly relevant because eco-anxiety and climate-related distress arise in the context of an ongoing and evolving stressor rather than a discrete event (Berry et al., Reference Berry, Waite, Dear, Capon and Murray2018; World Health Organization, 2022). Short-term reductions in distress may not necessarily translate into sustained adaptation, and some interventions may require repeated support, social reinforcement or institutional embedding to maintain effects over time (Sciberras and Fernando, Reference Sciberras and Fernando2022; World Health Organization, 2022). The short duration of most evaluations therefore limits what can currently be concluded about longer-term effectiveness. Future intervention studies would benefit from longer follow-up periods, repeated outcome assessment and designs capable of examining whether observed changes are maintained, attenuated or shift over time (Williamson et al., Reference Williamson, Altman, Bagley, Barnes, Blazeby, Brookes, Clarke, Gargon, Gorst, Harman, Kirkham, McNair, Prinsen, Schmitt, Terwee and Young2017; Berry et al., Reference Berry, Waite, Dear, Capon and Murray2018). Read more broadly, these findings also support the value of evaluation designs that can detect delayed, nonlinear or context-dependent effects, particularly where interventions seek not only to reduce distress but also to foster resilience, agency or sustained engagement over time (Meadows, Reference Meadows2008; Berry et al., Reference Berry, Waite, Dear, Capon and Murray2018).
This review has several strengths. It addresses a timely and underdeveloped area of research using a transparent and reproducible systematic review design, guided by PRISMA and supported by an inclusive search strategy developed with a research librarian (Page et al., Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann, Mulrow, Shamseer, Tetzlaff, Akl, Brennan, Chou, Glanville, Grimshaw, Hróbjartsson, Lalu, Li, Loder, Mayo-Wilson, McDonald, McGuinness, Stewart, Thomas, Tricco, Welch, Whiting and Moher2021). The review was also designed to capture heterogeneity in terminology and intervention format, allowing inclusion of a broad range of approaches across therapeutic, educational, community and nature-based domains (Baudon and Jachens, Reference Baudon and Jachens2021; Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021). A further contribution is the identification of extensive measurement heterogeneity, including inconsistent use and limited validation of eco-anxiety-specific tools, which complicates cross-study comparison (Kırımer-Aydınlı et al., Reference Kırımer-Aydınlı, Juaréz-Castelán, Hakim, Gul, Unal, Aguayo-Estremera, Pérez-Fortis, Rojas-Russell and Gallo2025; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025). The review also highlights implementation-related gaps, particularly the limited use of participatory design, culturally grounded delivery models and longer-term evaluations, which shape what can realistically be sustained beyond research pilots (World Health Organization, 2022; Patrick et al., Reference Patrick, Noy, Henderson-Wilson, Hayward, Allender, Gunasiri, Bruges, Snell and Capetola2023).
Several limitations should be acknowledged. Although the search strategy used a broad range of climate-related and eco-anxiety terms, studies using less common terminology for climate-related distress or weather-related anxiety may not have been captured (Coffey et al., Reference Coffey, Bhullar, Durkin, Islam and Usher2021; Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024). The review was limited to English- and Spanish-language publications, which may have contributed to under-representation of evidence from regions where other languages are more commonly used. The exclusion of studies centred on pre-existing mental health diagnoses or specialist treatment settings also narrows the direct applicability of the findings to clinical populations. In addition, the inclusion of protocol and feasibility studies was useful for characterising the intervention landscape, but these studies did not contribute outcome data in the same way as completed empirical evaluations. Finally, heterogeneity in design, intervention type and outcome measurement precluded meta-analysis and limits the strength of comparative conclusions (Baudon and Jachens, Reference Baudon and Jachens2021; Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021; Xue et al., Reference Xue, Massazza, Akhter-Khan, Wray, Husain and Lawrance2024).
The findings nevertheless suggest several implications for practice. Current evidence indicates that interventions addressing eco-anxiety may be most useful when outcomes are aligned with intervention intent – for example, targeting distress reduction where impairment is present, while also recognising that some interventions may appropriately prioritise agency, hope, connection or constructive engagement (Pihkala, Reference Pihkala2020; Martin et al., Reference Martin, Roswell and Cosma2024). The review also suggests that interventions delivered in group or community contexts, or those that incorporate opportunities for agency and social support, may offer advantages beyond individual-level coping alone, although these pathways remain under-evaluated (Baudon and Jachens, Reference Baudon and Jachens2021; Patrick et al., Reference Patrick, Noy, Henderson-Wilson, Hayward, Allender, Gunasiri, Bruges, Snell and Capetola2023). In highly climate-affected or low-resource settings, future programme development will likely need to consider culturally appropriate and scalable delivery formats, alongside partnerships that support local adaptation and implementation feasibility (IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022; World Health Organization, 2022).
Future research should prioritise three areas. First, greater consistency in measurement is needed, including clearer distinction between distress or impairment outcomes, emotional concern and agency- or engagement-related outcomes, as well as stronger psychometric validation across settings and populations (Kırımer-Aydınlı et al., Reference Kırımer-Aydınlı, Juaréz-Castelán, Hakim, Gul, Unal, Aguayo-Estremera, Pérez-Fortis, Rojas-Russell and Gallo2025; Ramsay et al., Reference Ramsay, Williams, Marks and Morgan2025). Second, the field would benefit from more rigorous and longer-term evaluations, including controlled studies with repeated follow-up and clearer reporting of theoretical pathways and stakeholder roles (Williamson et al., Reference Williamson, Altman, Bagley, Barnes, Blazeby, Brookes, Clarke, Gargon, Gorst, Harman, Kirkham, McNair, Prinsen, Schmitt, Terwee and Young2017; Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). Third, there is a clear need for participatory and contextually grounded intervention research in low- and middle-income countries, climate-vulnerable regions and communities disproportionately affected by environmental change, including Indigenous-led and community-governed partnerships where appropriate (IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022; World Health Organization, 2022; Patrick et al., Reference Patrick, Noy, Henderson-Wilson, Hayward, Allender, Gunasiri, Bruges, Snell and Capetola2023). Addressing these priorities will be essential for moving from a fragmented early evidence base towards more robust, implementation-ready and globally relevant responses to eco-anxiety.
The review addresses a timely question at what appears to be a critical juncture in the emerging science of addressing the mental health impacts of climate change. A strength of the review is the design, which allowed the inclusion of a wide diversity of studies. This review was constructed to adapt to the expected heterogeneity and variation in terminology through an inclusive suite of search terms and a broad set of databases. Our methods are transparent and reproducible, and we use the PRISMA process (Page et al., Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann, Mulrow, Shamseer, Tetzlaff, Akl, Brennan, Chou, Glanville, Grimshaw, Hróbjartsson, Lalu, Li, Loder, Mayo-Wilson, McDonald, McGuinness, Stewart, Thomas, Tricco, Welch, Whiting and Moher2021). We used a clear set of inclusion and exclusion criteria, and deliberately included protocol, feasibility and completed intervention studies to capture a broader representation of intervention approaches in this emerging field. Protocol and feasibility studies contributed to mapping the intervention landscape, whereas completed studies reporting outcome data informed the effectiveness synthesis. We used established approaches to assess study quality, using the Quality Assessment Tool (QuADS), which was appropriate to multiple study designs included in this review (Harrison et al., Reference Harrison, Walton, Manias, Wilson, Girgis, Chin, Leone, Seale, Smith and Chauhan2021). A further unique contribution of this review is clarifying the extent of measurement heterogeneity, including limited consistency and validation of eco-anxiety outcome tools, which constrains cross-study comparability. It also surfaces under-examined implementation constraints and enablers, particularly scalability, cultural relevance and limited stakeholder participation in design and delivery, that shape what can realistically be sustained beyond research pilots.
Although a broad set of climate-related and eco-anxiety search terms was used, studies employing less common terminology for climate-related distress or adverse weather–related anxiety may not have been captured. Rigorous dual screening at title, abstract and full-text stages reduced an initial pool of over 1,600 records to 22 eligible studies, excluding those in which changes in eco-anxiety were reported only as unintended outcomes of interventions targeting other aims. In addition, the exclusion of studies centred on pre-existing mental health diagnoses or specialist treatment contexts narrows the direct applicability of the review to clinical populations and specialist mental health services. The review was further limited to English- and Spanish-language publications, introducing potential language bias and likely under-representation of evidence from regions where research is commonly published in other languages. Substantial heterogeneity in study designs and outcome measures precluded meta-analysis. Finally, limited follow-up across included studies restricts confidence in durability of effects, and the concentration of evidence in high-income Western settings limits generalisability to highly climate-affected and low-resource contexts.
This review highlights several important implications for practice in addressing the mental health impacts of climate change and adverse weather events. Critically, there are emerging directions that indicate it is possible to develop effective, context-sensitive approaches for supporting populations affected by these environmental stressors, particularly when mental health is integrated into climate preparedness and response and delivered through community-based, culturally informed models (World Health Organization, 2022). Strengthening climate-mental health responses will likely require pairing individual coping support with community-delivered and structurally informed strategies, particularly in highly climate-affected regions where feasibility and relevance may differ from high-income university or general population samples.
However, the current evidence base is fragmented, with wide variation in intervention design, delivery, evaluation methods and reporting standards, challenging the identification and scale-up of best practice. National and international bodies, such as the WHO, should provide authoritative guidance, convene collaborative learning platforms and support capacity-building for climate-linked mental health and psychosocial support (Pan American Health Organization, 2023). Finally, while our initial aim was to examine the roles of different actors engaged in mitigating the mental health impacts of climate hazards, the included studies largely positioned participants as passive recipients rather than partners. Future interventions should adopt participatory approaches that empower affected communities and embed their perspectives throughout design, implementation and evaluation, consistent with established guidance on community mobilisation and multi-sector coordination (Inter-Agency Standing Committee, 2007).
Future research should deliberately move towards consistency in outcomes, represented by a core outcome set for climate-related mental health research to harmonise measurement across trials and evaluations (Williamson et al., Reference Williamson, Altman, Bagley, Barnes, Blazeby, Brookes, Clarke, Gargon, Gorst, Harman, Kirkham, McNair, Prinsen, Schmitt, Terwee and Young2017). To address the current high-income geographic skew, future studies should prioritise participatory, community-led trials and implementation studies in low- and middle-income countries and highly climate-affected regions, where distress may be shaped by displacement, livelihood disruption and limited service access (IPCC, Reference Pörtner, Roberts, Tignor, Poloczanska, Mintenbeck, Alegría, Craig, Langsdorf, Löschke, Möller, Okem and Rama2022). This work should be conducted with engagement of Indigenous-led and community-governed research partnerships (e.g., co-design, data sovereignty arrangements and shared decision-making), with funding directed to local partners and workforce development to support culturally safe delivery and context-specific models of care. The medium-term goal of this research endeavour would be to collate and align this research programme to best-practice approaches that are transparent and shared across regions represented by significant investment in practice-based evidence through multi-country pragmatic trials and open-access monitoring and evaluation frameworks.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2026.10313.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/gmh.2026.10313.
Acknowledgements
Microsoft M365 Copilot was used only to support text editing, grammar refinement and clarity improvements. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. No generative AI tools were used to develop the ideas, arguments, interpretations or substantive content of this manuscript.
Author contributions
C.V.: conceptualisation, methodology, data curation, investigation, formal analysis, visualisation, project administration, writing – original draft, reviewing and editing; C.M.: conceptualisation, methodology, data curation, investigation, writing – review & editing; Y.C.: conceptualisation, methodology, validation, writing – review & editing; S.A.: conceptualisation, methodology, investigation, validation, supervision, writing – review & editing. All authors approved the final version of the manuscript for publication.
Competing interests
The authors declare no conflicts of interest.
Data availability statement
Data sharing associated with this study is not applicable; no data were collected or generated as part of this systematic review.
Ethics and approvals
All authors declare to adhere to the publishing ethics of Global Mental Health. Ethical approval for this study is not applicable; no primary data was collected nor secondary data analysed as part of this systematic review.
Funding statement
This work did not receive any financial support.
Open access
