what do we know, what don’t we know?

MEEFRO April 3, 2023

0 124,547 9 minutes read

The great, late physicist Richard Feynman is reputed to have drolly once said that “Science doesn’t care if you’re happy or you’re not. Science doesn’t care if you believe in it or not. Science just is.” The same can be said of climate change. There is irrefutable evidence that it is happening, whether you care to acknowledge it or not. And climate change shows no mercy, especially to those who have contributed least to its causes but who are likely to face the most devasting impacts. In short, climate change represents the greatest challenge of our times, and understanding its likely impact offers us the best chance of adapting to current and future consequences and minimising impacts on human health.

Climate change impacts health via three main pathways:

First, it can directly affect health through increased occurrence and severity of adverse weather events (e.g. floods, storms, drought, wildfires, heatwaves) and atmospheric pollutants (Cruz et al., 2020; Nori-Sarma et al., 2022; Newbury et al., 2019)
Secondly, climate change can indirectly impact health via the degradation of ecosystems that support life (e.g. food systems), and
Thirdly, via effects on social systems, leading to job losses, rising food prices and food insecurity, conflict, and forced migration and displacement.

More optimistically we know that the availability of green space, especially in urban centres, can mitigate the effects of climate change on mental health. Green spaces are associated with a 1.5 to 3.5°C cooling of urban heat islands, and certain types of street trees can also offer protection from the effects of heat (Saaroni et al., 2018). As shown by a UK Biobank study, higher levels of residential greenness are associated with reduced odds of depression (Sakar et al., 2018), and local-area greenspace is known to be associated with higher levels of wellbeing (Houlden et al., 2018).

How do we make sense however of the ever-expanding evidence base about climate change and mental health to begin to develop targeted solutions? A recent systematic mapping review of research on climate and health highlighted that the literature in this field is growing exponentially (Berrang-Ford et al., 2021), making it increasingly relevant that we have access to robust global assessments of the evidence base. Pim Cuijpers et al., 2023 in the linked paper set about to offer this kind of review, aiming to review the impact of climate events, pollution, and green spaces on mental health from published meta-analyses.

Climate change has both a direct and indirect impact on our mental health and wellbeing.

Methods

This study is an umbrella review of meta-analyses that examined the association between mental health and climate events related to climate change. Umbrella reviews are essentially reviews of reviews, with the original systematic reviews forming the unit of analysis. Here, Cuijpers et al reviewed published meta-analyses that need not have necessarily been reported in a systematic review. The authors followed best practices for undertaking this kind of review and registered the protocol with the Open Science Framework, which enhances the transparency of methods. The authors focused on studies that reported findings about the relationship between discrete and non-overlapping climate-related determinants and mental health, with the aim that they could include as many environmental and climate-related determinants as possible. Their review, therefore, focused on the mental health impacts of climate events that are discrete episodes of extreme weather or climatic conditions, pollution (air, water, and land), and greenspaces, defined as exposure to the natural environment.

The authors searched major bibliographic databases that index medical and psychological literature (PubMed, PsycINFO, and Embase), but didn’t extend the search to include specialist databases that index studies about the human impact on the environment (e.g. GreenFILE), or literature from a broad cross-section of the social sciences (e.g. ASSIA). Care was taken to ensure that the search was sufficiently sensitive to capture evidence that might not have been indexed by terms related to climate change, and the full search was posted online. To increase the certainty of any conclusions the review only included meta-analyses that reported >5 or more comparisons. Further restrictions were that the review was limited to studies in English, and excluded impacts on intelligence, dementia, and cognitive decline. Studies were selected by two independent researchers which reduced the risk of bias and errors.

The authors assessed the quality of included studies with the AMSTAR-2 checklist, which incorporates 16 items related to methodological rigour. This checklist was originally designed for critically appraising systematic reviews of randomised and non-randomised controlled trials and is not designed to produce a total score. However, the authors did not adapt the AMSTAR-2 checklist for use in this research context, where controlled trials are not typical. They also reported a total score based on the arbitrary allocation of points for complete and partial reporting of items, which is not recommended with AMSTAR-2.

Data about outcomes were extracted by one reviewer and a second reviewer checked the validity of this process by independently extracting 25% of the data. High levels of agreement were reached suggesting that data were extracted accurately and appropriately. While the authors were only able to synthesise the findings narratively, they did examine the level of heterogeneity reported in the original meta-analyses using the I²and its 95% confidence intervals.

Results

This umbrella review included 24 meta-analyses. Thirteen meta-analyses focused on climate events; 11 focused on pollution; and there were only two analyses that looked at green space. Based on AMSTAR-2 ratings, confidence in the results of included studies was low, with only three studies with a moderate confidence and none with high confidence. Additionally, the meta-analyses for climate events were highly heterogenous with 27 of 43 (62%) comparisons with I² values ≥75%. The I² statistic describes the percentage of variation across studies that is due to heterogeneity rather than chance, and values of ≥75% suggest high levels of heterogeneity, i.e. the included studies were quite different from one another.

The results of the meta-analyses for climate events presented a mixed picture. A meta-analysis of six studies reported a non-significant standardised mean difference before and after climate events and psychological distress (0.32, 95% CI -0.06 to 0.70). There was insufficient before and after data to reliably report whether climate events were associated with mental health problems in general. Eight meta-analyses examined the pooled prevalence of mental disorders and climate events, but many of these comparisons were erroneously based on different cut-off values for mental health problems. Of the analyses that relied on diagnostic interviews, there was evidence that the prevalence of PTSD after floods was 16% (95% CI 0.11 to 0.21). Heat events were also shown to be significantly associated with mental health, with one meta-analysis showing a significant risk ratio for the daily suicide rates and an increase in temperature of 7.1°C.

Increased particulate matter pollution (PM_2.5) was associated with a significant pooled prevalence of depression (OR 1.07, 95% CI 1.01 to 1.12). Other vehicle-related pollutants that were significantly associated with depression were NO₂ and CO. Sulphur dioxide, which typically comes from burning coal for electricity generation, was also found to be significantly associated with depression. Other significant associations of note were between air pollution and the risk of autism spectrum disorders (four studies), and air pollution and the risk of suicide (one study).

The link between green spaces and mental health was based on the results of only two studies. One analysis showed a 6% reduction in the odds of depression for a 30% increase in neighbourhood green space; the converse was true for an increase in urbanisation. Experimental evidence showed that short-term exposure to natural environments was significantly associated with a reduction in depression (SMD = 0.30, 95% CI 0.10 to 0.50).

Mental health difficulties are associated with pollution, yet we cannot draw definitive conclusions based on this review.

Conclusions

The authors concluded that they had found:

reasonable evidence for an association between climate events and mental health and some association between pollution and mental disorders.

They also reported a signal that exposure to higher levels of green space is beneficial for mental health. However, overall, the quality of the evidence reviewed was low, making it difficult to draw definitive conclusions. In their summation the authors make a plea for high-quality research to not only confirm the associations reported, but to better quantify the size of climate-related impacts on mental health.

Exposure to green spaces could be beneficial for mental health.

Strengths and limitations

This was a well-conducted umbrella review that drew on robust and transparent methods to identify, select, appraise and report quantitative estimates of the relationship between a broad range of climate-related determinants and mental health. The range of evidence captured is impressive, pointing to the utility and benefits of an umbrella reviews as a way of making signature statements about an ever-expanding body of knowledge.

The broad scope of the review is also however a possible limitation. This review focused on not only the direct and indirect impacts of climate-related determinants on mental health, but also examined solutions, with an attempt to summarise the evidence about the benefits of green space on mental health. The evidence base for the impact of green space interventions and infrastructure on mental health is growing and is probably worth reviewing in its own right. Additionally, in this umbrella review, we were not given clear descriptions and definitions of what green space is, which is key to understanding likely health benefits. Green space can be measured in terms of proximity, amount, and quality (e.g. biodiversity); and there is increasing evidence that engagement and connection with greenspace are possibly more important for mental health than exposure (Coventry et al., 2021). As the authors acknowledged, limiting the review to studies with ≥5 meta-analyses was an arbitrary decision. However, this pragmatic approach did preclude the inclusion of meta-analyses based on fewer studies but derived from large population datasets, where sample sizes are large and certainty of conclusions is therefore also high (e.g. Noordzij et al., 2020).

Targeted research in specific climate events and the direct impact on mental health will shed light on potential solutions.

Targeted research on specific climate events and their direct impact on mental health will shed light on potential solutions.

Implications for practice

The conclusions of this review are not sufficiently compelling to warrant a change in practice or policy about the management of climate-related determinants for promoting mental health. The evidence base is too dominated by low-quality and cross-sectional studies to prompt investment in targeted solutions. However, this review has fulfilled an important service to the research and policy community by signalling the need for robust assessment and quantification of the mental health impacts of climate change. Too few reviews in this space focus on mental health. Indeed, even the Lancet Countdown only tangentially mentions mental health among the 43 indicators of the relationship between health and climate change (Romanello et al., 2022). Moreover, from a science communication perspective, this review reinforces the message that framing climate change in terms of impacts on health, and in particular mental health, is a vital way to stimulate engagement and activism in the climate emergency (Roque et al., 2020). There is now a much-needed call for higher quality research that addresses the longitudinal impact of climate change on mental health, and for that research to unpick mechanisms of impact so we can develop effective adaptation and mitigation solutions for the benefit of specific vulnerable populations.

This review has fulfilled an important service to the research and policy community by signalling the need for robust assessment and quantification of the mental health impacts of climate change.

Statement of interests

Peter Coventry is part-funded by the NIHR Yorkshire and Humber Applied Research Collaboration. The views expressed are those of the author, and not necessarily those of the NIHR or the Department of Health and Social Care.