In July 2023, the United Nations Secretary-General announced that the world is entering a “global boiling” stage.1 This assertion is supported by the mounting evidence of climate change that has been visible worldwide. Rising global temperatures continue to break records, leading to more frequent, and intense, heat waves. Glacial melt is accelerating, contributing to rising sea levels and threatening coastal communities. Extreme weather events, such as hurricanes, wildfires, and heavy rainfall are becoming more frequent and severe, causing devastating impacts on ecosystems and human populations.2
Tackling climate change requires a multifaceted approach that involves collaborative efforts across various sectors and levels of society. Strategies to achieve net-zero emissions goal by 20502 primarily involves upstream interventions such as implementing policies to limit emissions from sectors, such as transportation, industry, and agriculture. Furthermore, there is a growing need to develop strategies to better support adaptation, preparedness, and resilience for climate change at the downstream level to co-promote the health of human and natural systems. These complementary efforts can amplify the comprehensive drive toward effective climate action. One approach to contribute to the collective effort is by defining the distinct roles that individuals can play in their daily lives. This can be achieved through an exploration of the interconnected pathways linking climate change, lifestyle factors, and measures of human health.3,4
Though more is to be learned about the directionality between and mechanisms linking climate change and lifestyle factors, and their joint effect on health, some evidence exists indicating their bidirectional relationships. Specifically, physical activity (PA), which is negatively influenced by climate change,5–7 is often cited as a potential strategy that could mitigate climate change as well as generating public health benefits (noted as health co-benefits).4 For instance, one of the strategies for achieving health co-benefits has been investing in public transportation and built environment sectors.8–10 In addition, reducing sedentary behavior(s) (SB) may generate favorable outcomes for climate change mitigation by reducing the use of electricity, minimizing live streaming and freeing up data storage, and decreasing sedentary-based travel.5 Sleep is also known to be associated with climate change; mostly human sleep being affected by changing climate (eg, extreme temperatures, nature disasters).11,12 PA, SB, and sleep are collectively known as movement behaviors and have been shown to be individually and collectively related to many health indicators.13,14
It is the current understanding that lifestyle factors may not have as much impact as addressing fundamental factors such as creating equitable political and economic systems conducive to planetary health11—an ontological and philosophical stance that prone solutions, recognizing the interconnectedness of the health of human and the natural systems, highlights the importance of taking a closed-loop, systems approach to better understand the complex interrelationship(s) between the health of human and natural systems.3 However, given that both climate change and unhealthy lifestyles (eg, physical inactivity, excessive sedentarism, lack of sleep) negatively contribute to human health, either independently or interactively, better understanding of how lifestyle factors interplay with climate change and human health is a public health priority. Clarifying the mechanisms between climate change, lifestyle factors, and health may also be important in developing effective climate change adaptation, preparedness, and resilience strategies.5,12
The need for better understanding of the complex relationships among climate change, lifestyle factors, and health has been highlighted in a recent commentary4 and an umbrella review.9 The latter suggested that the relationship between climate change and health is multidirectional, while the mechanisms linking climate change, 24-hour movement behaviors (ie, PA, SB, sleep), and health is largely unknown.5 A recent commentary calling for action suggested that PA could contribute to solving major challenges such as deteriorating mental health that has been further exacerbated due to the COVID-19 pandemic and climate change.15 Furthermore, another commentary highlighted the importance of studying the role of sleep in strategizing climate change adaptation, mitigation, and resiliency by taking a life-course approach with an equity lens.12 Nevertheless, no review has yet addressed these calls to action. There is currently no comprehensive assessment that examines the extent to which movement behaviors contribute to the health effects of climate change. Likewise, it remains unknown whether climate change indicators, including adaptation measures such as the development of sustainable infrastructure, play a role in the relationship between movement behaviors and health.
The objective of this systematic review was to examine the relationships between climate change, 24-hour movement behaviors, and health. Given the potential bidirectionality between climate change and movement behaviors, this systematic review focused on: (1) clarifying the potential role of 24-hour movement behaviors in mediating/moderating the relationship between climate change and health and (2) clarifying the potential impact of climate change on the relationship between 24-hour movement behaviors and health.
Methods
This systematic review used the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines16 to ensure all methods are reported accurately. The review protocol was registered on PROSPERO (PROSPERO 2020 CRD 42020167386), the international prospective register of systematic reviews (https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=167386).
Eligibility Criteria
To be eligible, a study had to have examined the association between climate change, movement behaviors, and health. Studies needed to: (a) include at least one indicator of climate change as an exposure, mediator, or moderator; (b) include any aspects of 24-hour movement behaviors (ie, PA, SB, or sleep) as an exposure, mediator, or moderator; (c) include a measure(s) of human health as an outcome; (d) reported quantitative associations between climate change, 24-hour movement behaviors, and health; (e) used cross-sectional, case-control, cohort, or intervention study designs; (f) published in a peer-reviewed, refereed, nonpredatory journal17 in 2000 and onward; and (g) published in English language. Intervention studies were considered eligible for inclusion if baseline data included information about the relationship between climate change features, any aspects of 24-hour movement behaviors, and human health. Case studies, reviews, or qualitative studies were excluded as this review was interested in quantifying relationships. Also, studies that did not provide doses (eg, frequency, intensity, amount, type) of movement behaviors were excluded. Finally, studies with an analytical sample size of fewer than 100 participants were deemed ineligible due to the fact that small sample sizes are only adequately powered to detect large effects, whereas the actual effect sizes we expected to see are likely to be small to medium at most, resulting in low statistical power.18,19
For this review, climate change was broadly defined as unusual climatic features and human activity generated impacts directly and indirectly attributable to global warming.20 Specifically, related exposures included greenhouse gases (eg, CO2, O3, N2O, CH4), traffic noise, rising sea levels, poor air quality and air pollution (eg, ultrafine particles with an aerodynamic diameter less than 2.5 micrometers [PM2.5], 10 micrometers [PM10]), patterns and intensities of natural disasters (eg, extreme rain falls, droughts, floods, storm, bushfires), substantial temperature variability, increased severity of seasonality patterns (eg, wet spring, hot summer, snowy winter), barometric pressure, increased frequency and intensity of precipitation, extreme weather events (eg, heat waves, blizzards), allergens (eg, dust, pollen), and disease vectors (eg, ticks, bugs, blackflies, mites, mosquitoes). In addition, climate change adaptation strategies such as the availability of sustainable infrastructure (eg, mixed land use and green or blue space) were also considered as climate change indicators inclusively.
For 24-hour movement behaviors, sleep duration and any type, intensity, or location of PA (eg, sport; active transport; exercise; habitual, occupational, leisure physical activities, moderate-to-vigorous PA, light PA, outdoor PA) and any type of SB (eg, passive travel, screen time, sitting time, sedentary time) were considered. The health measures included a broad range of indicators related to both communicable and noncommunicable diseases, mental health, and injuries.
In quantitative research, moderation and mediation are both statistical concepts used to understand the relationships between variables. Moderation, also called as effect modification, occurs when the relationship between 2 variables (predictor and outcome) is influenced or moderated by a third variable.21,22 In other words, the effect of one variable on another depends on the level of a third variable. Moderation is often explored using interaction terms in regression analysis. For example, in studying the relationship between PA (predictor) and health (outcome), the effect of PA on health may be moderated by the level of exposure to air pollutants (moderating variable or effect modifier). This means that the relationship between PA and health may be stronger or weaker depending on the level of exposure to air pollutants. Mediation occurs when the effect of a predictor variable on an outcome variable is explained by a third variable, known as the mediator.21,22 The mediator variable helps to clarify the mechanism or process through which the predictor variable influences the outcome variable. Mediation is often examined using causal steps or path analysis. For example, in studying the relationship between exposure to greenspace (predictor) and mental health (outcome), PA (mediator) may explain how exposure to greenspace influences mental health. In this case, greenspace may increase mental health indirectly through being physically active in greenspace. It is important to mention that this review did not impose restrictions based on the statistical methods employed to investigate moderation or mediation. Regarding moderation, studies were considered eligible if they incorporated interaction terms in the analysis and/or conducted stratified analyses by different levels of a moderator. For mediation, all statistical techniques (eg, Baron and Kenny, Structural Equation Modeling) were deemed suitable as long as they assessed a hypothesized pathway.
Information Sources and Search Strategy
Literature searches were conducted in the following 7 databases: Ovid MEDLINE, Ovid PsycINFO, EbscoHost SPORTDiscus, Sports Medicine & Education Index, EbscoHost CINAHL, Engineering Village GeoRef, and Ovid EMBASE. Keywords and search strings for each database were developed by a librarian (Ross-White). Searches were restricted by English language and human participants, published in 2000 and onward to only capture the evidence published in the past 25 years. The initial searches were conducted in March 2020 and top-up searches were conducted on April 13, 2023 and April 18, 2024, respectively. The search results for each database were imported into the Clarivate Analytics EndNote X9 then Covidence (www.covidence.org)—a web-based software for screening selected data. A hand search by the primary investigator (Lee) was also conducted on November 5, 2020, July 30, 2023, and on May 10, 2024 to ensure that the most up-to-date, relevant studies post top-up searches were included in the review. Detailed information about search records and strategies are presented in Supplementary Tables S1 and S2 (available online), respectively.
Study Selection
Screening protocol for large-evidence systematic reviews and meta-analysis outlined by Polanin et al23 were followed for the level 1 screening (title and abstract screening). Briefly, it consists of the following 10 steps for the screening of titles and abstracts of identified studies: (1) creating a clear and concise abstract screening tool, (2) ensuring the hierarchical organization of the abstract screening tool, (3) conducting introductory abstract screening, (4) meeting with the screening team on a biweekly basis, (5) minimizing changes to the screening tool, (6) using a text-mining abstract screening application, (7) conducting independent double screening of each study, (8) resolving conflicts, (9) encouraging screening through incentives, and (10) analyzing the process and decisions after the completion of the screening. Any disagreements were resolved through a consensus discussion and if consensus could not be reached, the final inclusion of articles was decided by a third reviewer. In cases where a decision for exclusion or potential inclusion could not be made by the title/abstract, the full text was retrieved. At level 1, reconciling disagreements occurred after every 33.3% of the abstracts screened.23 Double screening of each article was conducted by 7 research assistants at both level 1 and level 2 (full-text screening). Interrater reliability (Cohen
Data Collection Process and Items
Data extraction was conducted by 3 graduate student trainees in Microsoft Excel worksheet developed by the primary investigator (E.-Y. Lee). All data extraction was reviewed by 2 senior authors (E.-Y. Lee and Park). The following information was extracted: the bibliographic information of each study (ie, authors and year of publication); setting and study design; sample characteristics (sample size, mean age, sex [n and % of males and females]); exposure, mediator or moderator, outcome, covariate measurements; and potential associations reported. Discrepancies were resolved through consensus discussion.
Study Risk of Bias Assessment
Joanna Briggs Institute’s critical appraisal tools24 (https://jbi.global/critical-appraisal-tools) were used to assess risk of bias for included studies. Bias was assessed as a judgment (yes, no, unclear, or not applicable) in different domains specific to each study design. Different number of criteria for each study design were judged with either “yes (1 point),” “no (0 points),” or “unclear (0 points).” If “not applicable” was chosen for an item, it was removed from consideration in calculating the percentage. High quality (low risk of bias) was considered when the overall rating was >80%, moderate quality was considered with scores of 41% to 80%, and low quality (high risk of bias) was considered with scores of 0% to 40%. Risk of bias assessment was undertaken by pairs of extractors and discrepancies were addressed through discussion in pairs. A third independent reviewer (Lee) was introduced when discrepancies could not be resolved.
Analysis and Synthesis of Results
All studies that met the eligibility criteria were included in analyses and discussing the overall review findings, regardless of the risk of bias rating. Meta-analyses were planned but not conducted due to the heterogeneity of the data, which could not be meaningfully pooled because they were not homogenous in terms of statistical, clinical, and methodological characteristics. Thus, only pooled results were reported.
Evidence was summarized by the role (mediator/moderator) that either climate change or movement behaviors played on health. Mediator refers to a third variable that explains the relationship between X (exposure) and Y (outcome), indicating “why” of the relationship, while moderator (also known as effect modifier) refers to a variable in which the relationship between X (exposure) and Y (outcome) changes across levels of the moderator, indicating, “how” of the relationship.22 The studies were categorized into those that included a mediator, moderator, or both. A number of studies were first counted for each statistical method used for mediation and moderation separately, or, together. Then a number of observations were counted for difference exposures, mediators/moderators, and outcomes used in each study. For example, if a study made 4 observations with different combinations of exposure, mediator, and outcome, it was counted as one study with 4 observations in evidence synthesis.
Statistical significance of the relationships examined, and the effect of mediation/moderation were then synthesized to determine the extent that climate change or 24-hour movement behaviors played a role in varying health indicators. Specifically, for mediators, total effect (path c), direct effect (path c′), and indirect effect (path ab), and their direction of associations (ie, null, positive, or negative) were reported, if available (Figure 1). For moderators, a main association in the fully adjusted model, interaction effect, and the details of moderation effects and their direction of associations were described, if available. If information was not provided, it was reported as NR (not reported). If the investigation was further stratified by different levels of a moderator, following a significant interaction effect between the exposure and moderator, the observed associations between exposure and outcome at different levels of a moderator, respectively, were reported. If multiple relationships were examined with different variables in one study, directionality of each relationship examined was extracted either as “no association (Ø),” “favorable association (+),” or “unfavorable association (–).”
—Mediation and moderation. Mediation: Total effect (path c), direct effect (path c′), and indirect effect (path ab). Moderation: An arrow pointing from M to Path c/c′.
Citation: Journal of Physical Activity and Health 21, 12; 10.1123/jpah.2023-0637
In addition to the total count for each direction category, percentages were also provided by presenting each direction of association divided by the total number of observations made, except for a NR which was treated as missing data in calculating percentages. Percentages served as a criterion in determining the strength of evidence: 0% to 20%: marginal; 21% to 40%: fair; 41% to 60%: moderate; 61% to 80%: strong; and 81% to 100%: very strong. A marginal level of evidence was considered as “unlikely relevant,” fair level of evidence was considered as “may be relevant,” moderate level of evidence was considered as “likely relevant,” and strong level of evidence was considered as “relevant,” and finally very strong level of evidence was considered as “actionable.”
Following the work by Chastin et al,25 certainty of evidence was examined using a Bayesian posterior probability. This likelihood P(A|Xn) represents the probability of an association given the evidence X obtained from n observations, assuming a neutral prior probability of the association (P(A) = 0.5). Bayes’ theorem was applied to calculate it, using the evidence probability P(Xn) and the likelihood probability P(Xn|A) based on binomial distributions. The assumption made was no publication bias and that all studies were adequately powered at 80%. This value served as an indicator of the level of uncertainty surrounding the association and the degree of knowledge gained from the current evidence. In cases of highly heterogeneous results, indicating greater uncertainty, P(A|Xn) tends to approach “0.” Conversely, a probability closer to “1” signifies less uncertainty regarding the existence of the association.
Due to the heterogeneity within exposure, mediator/moderator, and outcome measures (eg, having neighborhood greenness vs air pollution as climate change indicators), it was determined a priori that a negative aspect of climate change (eg, low greenness, high air pollution) would be coded as 1 (unfavorable) while a positive aspect of climate change (eg, high greenness, low air pollution) would be coded as 2 (favorable). Similar coding methods were used for movement behaviors (eg, 1 = low PA, high SB, inadequate sleep duration; 2 = high PA, low SB, adequate sleep duration) and health indicators (eg, 1 = poor health, 2 = good health). Finally, in cases where a study explored multiple indicators for each key variable (eg, climate change, movement behaviors, health), such as examining relationships between (1) air pollution, PA, and lung health and (2) air pollution, SB, and lung health, these instances were labeled as observations, each offering distinct evidence.
Results
Study Selection
The overall Preferred Reporting Items for Systematic Reviews and Meta-Analysis flowchart for study selection is described in Figure 2. Of 39,490 documents searched, 7868 duplicates were removed. In total, 31,622 studies were screened at level 1 and 592 studies were screened at level 2. After removing 517 studies for various reasons, and adding 4 additional studies based on expert suggestion, 79 studies were deemed eligible for this review.
—Preferred Reporting Items for Systematic Reviews flow chart.
Citation: Journal of Physical Activity and Health 21, 12; 10.1123/jpah.2023-0637
Study Characteristics and Quality Rating
Descriptive characteristics of the 79 studies, which represent 6,671,791 participants, and 3137 counties, from 25 countries (60% high-income countries and 40% low- and middle-income countries), are described in Table 1 (quality of evidence score: 9/10). The summary of included studies is categorized by different indicators of climate change and movement behaviors, as either mediators or moderators. Of included studies, 36 studies used a prospective cohort study design, 35 were used a cross-sectional study design, and the other 8 were from both cross-sectional and prospective cohort (n = 2), retrospective cohort (n = 4), case-control (n = 1), and ecological study designs (n = 1), respectively. A total of 15 studies employed mediation,26–30,32–34,36–38,41–44 60 employed moderation,45–104 and 4 studies examined both mediation and moderation.31,35,39,40 Of 77 studies that measured PA, 5 studies reported device-measured PA31,45,50,56,68 while the rest used self-reported data. For 4 studies that measured sleep, one study provided device-measured sleep cycle data.44 For 3 studies that provided data on SB, one study provided device-measured SB.45 Overall, studies with device-measured movement behavior data rated lower on the quality of evidence score compared with the overall average (90.4/100 vs 79.6 for device-measured PA, 67.0 for device-measured SB, and 75.0 for device-measured sleep; Table 1).
Summary of Included Studies Categorized By Different Indicators of Climate Change and Movement Behaviors as Mediators and Moderators (N = 79)
| Author | Analytic sample, n | Participants/cohort | Mage, y | Female/women, % | Country | Study designa | Exposureb,c | Outcome | Third variable of interestb,c | Subgroup analysis (yes/no) | ROB score, % |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Mediation models (n = 19)d | |||||||||||
| Climate change and movement behavior as mediators: serial mediation or multimediation models (n = 5) | |||||||||||
| Dzhambov et al26 | 399 | 15–25 y in Plovdiv city | 17.9 | 32 | Bulgaria | 1 | Urban residential greenspace,3 SAVI,3 tree cover density within the 500-m buffer,3 Euclidean distance to the nearest urban greenspace,3 self-reported measures of availability,3 access,3 quality,3 and usage of greenspace3 | Mental health (GHQ-12) | Mediators: PA,1 restorative quality,1 perceived air pollution1 | Yes: age, gender, SES, population density, duration of residence, time spent at home per day, and education levels | 100 |
| Dzhambov et al27 | 720 | Medical university students | 21.0 | 66 | Bulgaria | 1 | NO23 | Mental health | Mediators: annoyance (noise, air pollution, vibration from traffic, construction, neighbors, loud music),1 PA,1 and sleep disturbance1 | No | 88 |
| Dzhambov et al28 | 720 | Medical university students | 21.0 | 66 | Bulgaria | 1 | Green space,3 blue space3 | Mental health | Mediator: perceived neighbor green/blue space,3 noise,1 air pollution,3 annoyance,3 and PA1 | No | 100 |
| Huang et al29 | 11,486 (59.8% hypertensive) | WHO longitudinal cohort SAGE 2007–2010 | 50+ | 53 | China | 3 | NDVI1 | Hypertension | Serial mediators: PA,1 PM2.51 | Yes: urban and rural locality | 90 |
| Guan et al30 | 27,634 | Adults from Shandong Province which is located on the northern coast of China | 50.7 | 52 | China | 1 | Impervious surface area, road density, and annual night light1 | Type 2 diabetes | Mediators: PA,1 NDVI,1 PM2.51 | Y: age, gender, education, and house income | 75 |
| Climate change as a mediator: air pollution (n = 1) | |||||||||||
| Lovinsky-Desir et al31 | 151 | Children living in New York City | 12.5 | 52 | United States | 1 | PA2 | Lung function | Mediator: outdoor pollution1 | No | 100 |
| Movement behavior as a mediator: PA (n = 12) | |||||||||||
| Dzhambov32 | 109 | Medical university students in Plovdiv city 2017–2018 | 21.0 | 45 | Bulgaria | 4 | Green/blue space3 | Mental health (GHQ-12) | Mediator: PA1 | No | 89 |
| Gascon et al33 | 958 | ALFA Cohort | 56.5 | 64 | Spain | 1 | Long-term exposure to residential green and blue spaces1 | Mental health (self-reported history of depression and medication with benzodiazepines) | Mediator: PA1 | No | 100 |
| Huang et al34 | 24,845 | Chinese Communities Health Survey 2009 | 45.6 | 49 | China | 1 | NDVI1 | Obesity | Mediator: PA1 | No | 100 |
| James et al35 | 108,630 | Nurses’ Health Study 2000–2008 | 69.0 | 100 | United States | 3 | NDVI1,2 | Nonaccidental all-cause mortality | Mediator: PA1 | No | 90 |
| Mila et al36 | 6039 | APCAPS | 36.2 | 47 | United Kingdom | 3 | Built-up land use1 | Cardiometabolic risk factors | Mediator: PA1 | No | 82 |
| Triebner et al37 | 1069 | RHINESSA | 29.8 | 100 | Norway and Sweden | 3 | Greenness (NDVI)1 | PMS | Mediator: PA1 | No | 64 |
| Wang et al38 | 1029 | Adults in Guangzhou city | 41.2 | 50 | China | 1 | Greenness (NDVI, streetscape greenery)1 | Mental well-being (WHO-5) | Mediator: PA (h/wk)1 | No | 100 |
| Wang et al39 | 20,861 | CLDS | 44.8 | 52 | China | 1 | Air pollution (PM2.5)1 | Depression (CES-D) | Mediator: PA (h/wk)1 | No | 100 |
| Wang et al40 | 24,623 | CHARLS | 67.2 | 51 | China | 3 | Air pollution (PM2.5)1 | Depression (CES-D) | Mediator: PA (h/wk)1 | No | 73 |
| Yang et al41 | 24,845 | Thirty-three Communities Chinese Health Study 2009 | 45.6 | 49 | China | 1 | Greenness (NDVI)2 | DBP, SBP, and hypertension | Mediator: PA1 | No | 100 |
| Yang et al42 | 15,477 | Thirty-three Communities Chinese Health Study 2009 | 45.0 | 47 | China | 1 | Greenness (NDVI, SAVI),2 air pollution (NO2, PM2.5)1 | Blood lipids | Mediator: PA1 | No | 100 |
| Zhang et al43 | 4364 | CHARLS 2011 | 67.5 | 48 | China | 1 | Air quality index1 | Obesity (general, abdominal) | Mediator: PA (active vs inactive)1 | No | 100 |
| Movement behavior as a mediator: sleep (n = 1) | |||||||||||
| Lo et al44 | 4866 | Adults | 49.5 | 31 | Taiwan | 1 | Air pollution (PM2.5, NO2, O3)1 | Cognitive function | Mediator: sleep cycle2 | No | 75 |
| Moderation models (n = 64)d | |||||||||||
| Climate change and movement behavior as moderators (n = 9) | |||||||||||
| Avila-Palencia et al4 | 122 | Healthy adults in 3 European cities 2015–2016 | 35.1 | 55 | Belgium, United Kingdom, and Spain | 3 | PA (≥3 METs),2 SB (≤1.5 METs),2 BC1,2 | Blood pressure | Moderators: PA,2 SB,2 and BC1,2 | No | 67 |
| Coleman et al46 | 403,748 | National Health Interview Surveys 1997–2014 | 45.9 | 55 | United States | 3 | PA,1 PM2.5,1 NDVI1 | Mortality | Moderators: PA,1 PM2.5,1 NDVI1 | Yes: sex, race, age, income, education, marital status, urban/rural, census region, ecoregion | 91 |
| Jiang et al47 | 4537 | CHARLS 2011–2015 | 58.6 | 55 | China | 3 | PM2.51, PA1 | Physical function (grip strength, walking speed, sense of balance, and chair standing tests) | Moderators: PM2.51, PA1 | Yes: age, sex, and urbanicity/rurality | 73 |
| Jiang et al48 | 1326 | Chinese adults (40 and older) | 63.27 | 49 | China | 2 | Air pollution (NO2, O3, Owt)2 | Stroke | Moderators: PM2.52, PA1 | Yes: age, sex, BMI, urbanicity/rurality, smoking status, and medical history | 100 |
| Kubesch et al49 | 50,635 | Danish Diet, Cancer, and Health cohort | 56.7 | 53 | Denmark | 3 | Air pollution (NO2)1 | Myocardial infarction | Moderator: NO21, PA | No | 100 |
| Laeremans et al50 | 122 | PASTA 2015–2016 | 35.0 | 55 | Belgium, United Kingdom, and Spain | 1 | BC,2 PA2 | HRV, retinal vessel diameters, and lung function | Moderators: BC,2 PA2 | No | 88 |
| Li et al51 | 359,153 | UK Biobank participants without diabetes at baseline (40–69 y) | 56.3 | 52.8 | United Kingdom | 3 | PA,1 air pollution (PM2.5, PMcoarse, PM10, NO2)1 | Type 2 diabetes | Moderators: PA,1 air pollution (PM2.5, PMcoarse, PM10, NO2)1 | Yes: genetic risk, PA, PM2.5, PMcoarse, PM10 | 100 |
| Li et al52 | 367,978 | UK Biobank 2006–2010 | 56.3 | 52 | United Kingdom | 2 | Air pollution (PM2.5, PM10, NO2, NOx)1 | Incident chronic kidney disease | Moderators: PA,1 air pollution (PM2.5, PM10, NO2, NOx)1 | No | 80 |
| Zhang et al53 | 359,067 | The Taiwan MJ Cohort 2001–2014 | 39.9 | 51 | Taiwan | 3 | PM2.5,1 PA1 | Systemic inflammation (white blood cell counts) | Moderators: PM2.5, PA (METs h/wk)1 | No | 91 |
| Climate change indicators (multiple) as moderators (n = 1) | |||||||||||
| Gray et al54 | 3137 counties | BRFSS 2009, EQI, and US Census | NA | NA | United States | 6 | PA (physical inactivity)1 | Obesity | Moderator: air quality,1 water quality,1 land quality,1 and environmental quality1 | Yes: sex | 63 |
| Climate change as a moderator: air pollution (n = 11) | |||||||||||
| Andersen et al55 | 52,061 | Danish Diet, Cancer, and Health Cohort 1993–2010 | 56.6 | 53 | Denmark | 3 | PA (sport, cycling, gardening, and walking)1 | Total and cause-specific mortality | Moderator: NO21 | No | 100 |
| Elavsky et al56 | 243 | Women aged 40–60 in high (Moravian-Silesian Region) and low (Southern Bohemian Region) air pollution regions in Czech Republic | 47.8 | 100 | Czech Republic | 3 | Air pollution,1 PA2 | Menopausal symptoms | Moderator: air pollution1 | No | 55 |
| Elliott et al57 | 104,990 | Nurse Health Study 1998–2008 | 63.1 | 100 | United States | 3 | PA1 | CVD (myocardial infarction and stroke) risk, and overall mortality | Moderator: PM2.51 | No | 91 |
| Fisher et al58 | 53,113 | Danish Diet, Cancer, and Health Cohort 1993–2012 | 56.7 | 53 | Denmark | 3 | Transport-related PA1 | Asthma, COPD | Moderator: NO21 | No | 100 |
| Hou et al59 | 39,089 (30.8% with metabolic syndrome) | Henan Rural Cohort study 2015–2017 | 55.6 | 61 | China | 3 | PA1 | Metabolic syndrome | Moderator: air pollution (PM1, PM2.5, PM10, NO2)1 | No | 80 |
| Kim et al60 | 1,259,871 | National Health Insurance Service 2009–2010 | 64.7 | 54 | South Korea | 3 | Air pollution (PM10, PM2.5),1 PA1 | Diabetes | Moderators: air pollution (PM10, PM2.5),1 PA1 | No | 82 |
| Kim et al61 | 1,469,972 | National Health Insurance Service 2009–2012 | 32.3 | 40 | South Korea | 3 | PA (METs)1 | CVD (coronary heart disease and stroke) | Moderator: PM2.5,1 PM101 | Yes: sex, smoking status | 82 |
| Luo et al62 | 336,545 | UK Biobank prospective cohort (aged 40–69 y) | 57 | NR | United Kingdom | 3 | PA1 | Type 2 diabetes, mood disorder, and death | Moderator: air pollution (PM2.5) | No | 100 |
| Oudin et al63 | 3036 | Swedish National Stroke Register (Risk-stroke) and the Malmö and Lund Stroke Registers | Birth year 1923–1965 | 42 | Sweden | 5 | PA1 | Hospital admissions for ischemic stroke | Moderator: NOx1 | No | 82 |
| Sun et al64 | 66,820 | Elderly Health Service Cohort 1998–2011 | 71.9 | 66 | Hong Kong | 3 | PA (METs, type)1 | Death risk | Moderator: PM2.51 | No | 90 |
| Wang, et al40 | 24,623 | CHARLS | 67.2 | 51 | China | 3 | Air pollution (PM2.5)1 | Depression (CES-D) | Moderator: air pollution (PM2.5)1 | No | 73 |
| Climate change as a moderator: Greenspace (n = 1) | |||||||||||
| Chen et al65 | 23,732 | National Prospective Cohort and 14-y Longitudinal Dynamic Cohort Study 2013–2018 | 6–18 y | NA | China | 3 | PA,1 air pollution (PM1, PM2.5, PM10, SO2, NO2, CO, and O3)1 | Overweight, obesity | Moderator: NDVI1 | No | 90 |
| Climate change as a moderator: seasonality (n = 1) | |||||||||||
| Lovinsky-Desir et al31 | 151 | Children living in New York City | 12.5 | 52 | United States | 1 | PA2 | Lung function | Moderator: season1 | No | 100 |
| Movement behaviors (multiple) as moderators (n = 2) | |||||||||||
| Li et al52 | 354,897 | UK Biobank 2006–2010 | 56.7 | 53 | United Kingdom | 3 | Air pollution (PM2.5, PM10, NO2, NOx)1 | Major depressive disorder | Moderators: healthy lifestyle factors1 including PA, SB (television viewing time), and sleep duration | No | 73 |
| Tang et al66 | 1102 | College students in Guangzhou city | 18.3 | 0.3 | China | 4 | Temperature (outdoor thermal indices)1 | Perceived thermal comfort and cognitive ability/efficiency | Moderators: PA,1 SB1 | No | 89 |
| Movement behavior as a moderator: PA (n = 38) | |||||||||||
| Ao et al67 | 36,562 (11% with diabetes) | China multiethnic cohort (50–79 y) | 60.5 | 59 | China | 1 | Air pollution (PM10, PM2.5)1 | Type 2 diabetes | Moderator: PA1 | No | 100 |
| Chen et al68 | 530 | Older adults (≥65 y) | 70.2 | 58 | Taiwan | 1 | Air pollution (PM2.5)1 | Skeletal muscle mass, body fat | Moderator: PA2 | No | 88 |
| Coogan et al69 | 43,003 | Black Women’s Health Study 1995–2011 | 38.6 | 100 | United States | 3 | NO21 | Type 2 diabetes | Moderator: PA1 | No | 90 |
| Endes et al70 | 2823 | Swiss Cohort Study on Air Pollution and Lung and Health Diseases 1991–2003 | 63.5 | 51 | Switzerland | 2 | Air pollution (PM10, PM2.5, NO2, PNC)1 | Arterial stiffness | Moderator: PA1 | No | 89 |
| Eze et al71 | 3769 | Swiss Cohort Study on Air Pollution and Lung and Heart Diseases 1991–2002 | 43.6 | 46 | Switzerland | 2 | 10-y mean residential PM101 and NO21 | Metabolic syndrome | Moderator: PA1 | No | 82 |
| Gandini et al72 | 74,989 | NHIS (2001–2008) | 35+ | 53 | Italy | 3 | Air pollution (PM2.5, NO2)1 | Incident hospital admissions | Moderator: PA1 | No | 100 |
| Gao et al73 | 2048 | Chinese children | 9 | 48 | Hong Kong | 1 | Air pollution (PM10, NO2, SO2, O3)1,2 | Cardiorespiratory fitness: VO2max | Moderator: PA1 | Yes: gender | 100 |
| Guo et al74 | 6486 | Adults | 61.1 | 54 | China | 1 | Air pollution (PM2.5, PM10, O3, NO2, SO2, CO)1 | Metabolic equivalent | Moderator: PA1 | No | 100 |
| Hou et al75 | 31,282 | Henan Rural Cohort study 2015–2017 | 55.2 | 60 | China | 3 | Air pollution (PM1, PM2.5, PM10, NO2)1 | CVD | Moderator: PA1 | No | 90 |
| James et al35 | 108,630 | Nurses’ Health Study 2000–2008 | 69.0 | 100 | United States | 3 | NDVI1,2 | Nonaccidental all-cause mortality | Moderator: PA1 | No | 90 |
| Ju et al76 | 21,944 | China Family Panel Study 2014 and 2016 | 48.5 | 51 | China | 1 | Air pollution (mean PM 2.5 concentrations, ground surface ozone concentrations)1 | Mental health (self-reported depression) | Moderator: PA1 | Yes: age, sex, income, and chronic conditions | 100 |
| Kim et al77 | 545 | The Korean Elderly Environmental Panel study 2008–2010 | 70.6 | 74 | South Korea | 3 | Air pollution (PM2.5, NO2, O3, CO, SO2)2 | Liver enzyme concentrations | Moderator: PA1 | Yes: alcohol use | 90 |
| Kim et al78 | 100,867 | 2012 Korean Community Health Survey 2012 | 47.8 | 49 | South Korea | 1 | Air pollution (PM10, O3)1 | Cancer risk | Moderator: PA1 | Yes: age, sex, smoking status, and alcohol use | 82 |
| Laffan79 | 4277 | MENE spanning 2012–2015 | NR | 52 | United Kingdom | 1 | Air pollution (PM2.5)1 | Subjective well-being | Moderator: PA1 | Yes: frequency of visits to the outdoors | 75 |
| Lamichhane et al80 | 1264 | Adults (20–85 y) | 57.9 | 54 | South Korea | 1 | Air pollution (PM2.5, PM10, NO2)1 | Lung function, COPD | Moderator: PA1 | No | 100 |
| Lao et al81 | 147,908 | The Taiwan MJ Cohort 2001–2014 | 38.3 | 50 | Taiwan | 3 | Air pollution (PM2.5)1 | Type 2 diabetes | Moderator: PA1 | No | 82 |
| Li et al82 | 32,963 | Pregnant women | 28.0 | 100 | China | 1 | Air pollution (PM2.5, PM10, NO2, CO)1,2 | Variation of blood glucose level | Moderator: PA1 | No | 100 |
| Li et al83 | 39,207 | Henan Rural Cohort Study 2015 (included hypertensive participants) | 55.6 | 61 | China | 1 | Air pollution (PM10, PM2.5, NO2)1 | DBP, SBP, mean arterial pressure, and pulse pressure | Moderator: outdoor PA1 | No | 100 |
| Lin et al84 | 45,625 | WHO Study of Global Ageing and Adult Health | 58.3 | 57 | China, Ghana, India, Mexico, Russia, and South Africa | 1 | Air pollution (PM2.5) 1 | Stroke | Moderator: PA (work, transport, and recreational/leisure-time activities)1 | No | 100 |
| McConnell et al85 | 3535 | Children with no history of asthma in southern California | 9–16 | 46 | United States | 3 | Air pollution1 | Asthma | Moderator: PA (sport)1 | No | 73 |
| Park et al86 | 1454 | South Korean adults aged 50 y and older | 67.5 | 62 | South Korea | 1 | Air pollution1 | Depression | Moderator: PA1 | Yes: intensity of PA, gender, and age groups (<65 y and ≥65 y) | 88 |
| Puett et al87 | 66,250 | Nurses’ Health Study 1992–2002 | 62.4 | 100 | United States | 3 | Air pollution (PM2.5, PM10)1 | Mortality, CHD | Moderator: PA (METs hr/wk)1 | No | 91 |
| Raichlen et al88 | 35,562 | UK Biobank 2006–2010 | 64.98 | 50 | United Kingdom | 3 | Air pollution (NO, NO2, PM10, PM2.5, PM2.5–10, PM absorbance)1 | All-cause dementia | Moderator: PA (accelerometry)2 | No | 73 |
| Raza et al89 | 2221 | Västerbotten Intervention Program 1990–2013 | 52 (stroke), 53 (IHD) | 23% (stroke), 47% (IHD) | Sweden | 3 | Air pollution (PM2.5)1 | Recurrence of ischemic heart disease and stroke | Moderator: PA (active transport, LTPA)1 | No | 90 |
| Roswall et al90 | 44,438 | Danish Diet, Cancer, and Health cohort 1993–2015 | 57.3 | 51 | Denmark | 3 | Nighttime road traffic noise1 | Redemption of sleep medication | Moderator: PA (LTPA)1 | No | 91 |
| Tallon et al91 | 3377 | NSHAP 2005–2011 | 72.38 | 55 | United States | 3 | Air pollution (PM2.5, NO2)1 | Cognitive function | Moderator: PA1 | No | 64 |
| Thiering et al92 | 837 | German birth cohorts (LISAplus and GINIplus) | 15.2 | 49 | Germany | 3 | Residential long-term air pollution, greenness1 | Insulin resistance | Moderator: PA,1 outdoor time1 | No | 82 |
| Tu et al93 | 31,162 | Henan Rural Cohort Study 2015–2017 | 55.9 | 61 | China | 1 | Air pollution (PM1, PM2.5, PM10, NO2)1 | High 10-y ASCVD risk | Moderator: PA (METs h/wk)1 | No | 100 |
| Vencloviene et al94 | 7077 | International HAPIEE study 2006–2008 | 60.5 | 55 | Lithuania | 1 | Metrology (air temperature, atmospheric pressure, relative humidity, wind speed, and North Atlantic oscillation indices)1 | SBP, DBP | Moderator: PA (LTPA)1 | No | 100 |
| Wang et al39 | 20,861 | CLDS | 44.8 | 52 | China | 1 | Air pollution (PM2.5)1 | Depression (CES-D) | Moderator: PA (h/wk)1 | No | 100 |
| Wu et al95 | 35,334 | CNSSCH 2014 | 13.4 | 50 | China | 1 | Air pollution (PM1, PM1–2.5, PM2.5, SO2, NO2)2 | Forced vital capacity | Moderator: PA (h/d)1 | No | 100 |
| Wu et al96 | 39,168 | UK Biobank 2010 | 64.1 | 47 | United Kingdom | 1 | Air pollution (PM2.5, PM10)1 | Depressive disorder | Moderator: PA (METs h/wk)1 | No | 100 |
| Xu et al97 | 70,668 | Chinese adults in 5 provinces in Southwest China (CMEC survey) | 52.2 | 60 | China | 1 | Air pollution (PM1, PM2.5, PM10, O3, and NO2)1 | Insomnia | Moderator: PA (total PA and domain-specific PA including occupation, transportation, housework, and leisure-time exercise)1 | Yes: age, sex, ethnicity, BMI, and smoking status | 75 |
| Yang et al98 | 15,477 | Thirty-three Communities Chinese Health Study 2009 | 45.0 | 47 | China | 1 | Air pollution (PM1, PM2.5, PM10, SO2, NO2, O3)1 | Metabolic syndrome | Moderator: PA (regular exercise)1 | No | 100 |
| Yu et al99 | 469 | Adolescents (14–18 y) 2016 | 16.3 | 43 | Indonesia | 1 | Air pollution: PM2.51 | Fasting plasma glucose | Moderator: PA (moderate/high vs low)1 | No | 100 |
| Yu et al100 | 1090 | Latino Study on Aging 1998–2007 | 70.5 | 59 | United States | 3 | Air pollution: O31 | Type 2 diabetes | Moderator: outdoor LTPA1 | No | 100 |
| Zhang et al101 | 4544 | KORA 2004–2008 | 60.1 | 42 | Germany | 1 | Air pollution (PM10, PMcoarse, PM2.5, PM2.5, NO2)1 | SBP | Moderator: PA (h/wk)1 | No | 100 |
| Zhang et al102 | 123,045 | KSHS | 39.4 | 40 | South Korea | 3 | Air pollution (PM2.5, PM10)1 | Onset of Depression | Moderator: PA (times/week)1 | No | 73 |
| Movement behavior as a moderator: sleep (n = 1) | |||||||||||
| Shi et al103 | 2043 | Adults without hearing or language impairment | 62.8 | 49 | China | 1 | Air pollution (PM2.5)1 | Cardiac conduction abnormalities | Moderator: sleep1 | No | 100 |
Abbreviations: ASCVD, atherosclerotic cardiovascular diseases; ALFA, ALzheimer and FAmilies; APCAPS, Andhra Pradesh Children and Parent Study; BC, black carbon; BMI, body mass index; BRFSS, Behavioral Risk Factor Surveillance System; CES, Center for Epidemiologic Studies Depression Scale; CHARLS, China Health and Retirement Longitudinal Study; CLDS, China Labor-force Dynamics Survey; CMEC, China Multi-Ethnic Cohort; CNSSCH, Chinese National Survey on Students’ Constitution and Health; COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; EQI, Environmental Quality Index; GHQ, General Health Questionnaire; GINIplus, The German Infant Study on the Influence of Nutrition Intervention plus Air pollution and Genetics on Allergy Development; HAPIEE, Health, Alcohol and Psychosocial Factors in Eastern Europe; HRV, heart rate variability; IHD, ischemic heart disease; KORA, German population-based Cooperative Health Research in the Region of Augsburg; KSHS, Kangbuk Samsung Health Study; LISAplus, Influence of Life-style factors on Development of the Immune System and Allergies in East and West Germany plus Air Pollution and Genetics on Allergy Development; LTPA, leisure-time PA; MENE, Monitor of Engagement with the Natural Environment survey; MET, metabolic equivalent of task; NA, not available; NDVI, normalized difference vegetation index; NHIS, National Health Interview Survey; NR, not reported; NSHAP, National Social Life, Health, and Aging Project; NO, nitric oxides; NO2, nitrogen dioxide; O3, ozone; Owt, weighted average of a combined oxidation of NO2 and O3; PA, physical activity; PASTA, Physical Activity through Sustainable Transport Approaches; PM1, ultrafine particles with an aerodynamic diameter less than 1 micrometers; PM2.5, ultrafine particles with an aerodynamic diameter less than 2.5 micrometers; PM10, ultrafine particles with an aerodynamic diameter less than 10 micrometers; PMS, premenstrual syndrome; PNC, particle number concentration; RHINESSA, Respiratory Health in Northern Europe, Spain and Australia; ROB, risk of bias; SES, socioeconomic status; SAGE, Study on Global Ageing and Adult Health; SAVI, soil adjusted vegetation index; SB, sedentary behavior; SBP, systolic blood pressure; SO2, sulfur dioxide; VO2max, maximum oxygen consumption; WHO, World Health Organization.
aStudy design coding: 1= cross-sectional; 2 = RCS: retrospective cohort study; 3 = PCS: prospective cohort study; 4 = both cross-sectional and prospective cohort design; 5 = case control; 6 = ecological study. bMeasurement method used for movement behaviors coding: 1 = self-reported, 2 = device measured. cMeasurement nature of climate change indicators coding: 1 = long-term exposure, 2 = short-term exposure, 3 = unknown/not reported. dFour studies provided both mediation and moderation models.
Results of Mediation
The overall summary of mediation with different indicators of climate change and movement behaviors are described in Table 2. Of 19 studies that employed mediation, 16 studies examined PA as a mediator,26,29,30,32–43 2 studies (including one nonmutually exclusive study) examined sleep as a mediator,27,44 one study examined climate change indicators (O3, NO2) as mediators,31 and 2 studies27,28 (including one nonmutually exclusive study) examined PA, sleep, and climate change indicators (restorative quality) as comediators in the multiple-mediator model.
Overall Summary of Mediation (n = 19 Studies)
| Exposure | Mediator | Outcome | No. of studies reported | No. of observations made | No. of c path reported | c path†,‡ | No. of c′ path reported | c′ path†,‡ | No. of ab path reported | ab path†,‡ | Uncertainty of evidenceß |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Air pollution/urbanization | PA | Varying health measures | 4 | 7 | 7 | Ø = 0% − = 100% + = 0% | 3 | Ø = 0% − = 100% + = 0% | 7 | Ø = 29% − = 0% + = 71% | 0.9999 |
| Green/blue space, mixed land use | PA | Varying health measures | 11§ | 52 | 31 | Ø = 58% − = 0% + = 42% | 23 | Ø = 44% − = 0% + = 56% | 52 | Ø = 65% − = 2% + = 33% | <0.0001 |
| Green/blue space | Perceived green/bluespace, restorative quality, PA, and sleep | Mental health | 3§ | 39 | 12 | Ø = 50% − = 17% + = 33% | 12 | Ø = 100% − = 0% + = 0% | 39 | Ø = 62% − = 0% + = 38% | <0.0001 |
| Air pollution | Sleep | Mental health/cognitive function | 2§ | 12 | 3 | Ø = 33% − = 67% + = 0% | 3 | Ø = 100% − = 0% + = 0% | 12 | Ø = 29% − = 0% + = 71% | <0.0001 |
| Overall summary of the association between climate change and health with movement behavior as a mediator | |||||||||||
| Climate change indicators | Movement behavior (95% of PA) | Varying health measures | 19 | 110 | 66 | Ø = 37% − = 45% + = 18% | 41 | Ø = 61% − = 39% + = 0% | 110 | Ø = 64% − = 3% + = 33% | <0.0001 |
| Exposure | Mediator | Outcome | No. of studies reported | No. of observations made | No. of c path reported | c path†,‡ | No. of c′ path reported | c′ path†,‡ | No. of ab path reported | ab path†,‡ | Uncertainty of evidenceß |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Outdoor PA | Air pollution | Lung function | 1 | 2 | 2 | Ø = 0% − = 100% + = 0% | 2 | Ø = 100% − = 0% + = 0% | NR¶ | NR¶ | NA |
Abbreviations: NA, not applicable; NR = not reported; PA, physical activity; path c, total effect; path c′, direct effect; path ab, indirect effect.
†Direction: Ø = null association; − = unfavorable association; + = favorable association. ‡Strength of evidence: 0% to 20% = marginal (unlikely relevant); 21% to 40%: fair (may be relevant); 41% to 60%: moderate (likely relevant); 61% to 80%: strong (relevant); 81% to 100%: very strong (actionable). §Two nonmutually exclusive studies provided both single and multimediator models, separately. ¶Indirect effect was not reported but reduction from path c (total effect) to path c′ (direct effect) was reported, which indicates the role of mediator. ßUncertainty of evidence is indicated by calculating the posterior Bayesian probability of association: P(A|Xn) = “0” indicates the greatest uncertainty and “1” indicates the least uncertainty of the evidence.
A total of 7 observations were reported from 4 studies30,39,40,43 examining the relationships between the environmental characteristics that are unfavorable to climate change (ie, air pollution, urbanization) with PA as a mediator. All reported observations indicated negative total (path c) and direct (path c′) associations between climate change indicators and health measures (100%). Of these, 71% of the observations indicated that PA favorably mediated the association between climate change indicators and health measures, with very low uncertainty of evidence.
Of 16 studies that examined PA as a mediator,26,29,30,32–43,79,86 52 independent observations from 11 studies26,29,32–38,41,42 reported on the mediating role of PA in the relationship between greenspace/bluespace and mixed land use factors and health measures. Of 31 observations reporting the total effect (c path), 18 observations showed null associations (58%), 13 showed favorable associations (42%), and none of the observations showed unfavorable associations (0%), indicating generally a positive association between the environmental characteristics that are favorable to climate change and health measures. Of 23 direct effects (c′ path) reported, 10 observations were null (44%), 13 were favorable (56%), and none were unfavorable after the inclusion of PA as a mediator. For the indirect effect (path ab), 34 observations reported not having any indirect effect (65%), 17 observations showed a favorable indirect effect (33%), and one observation showed an unfavorable indirect effect (2%). The uncertainty of evidence was high.
Three studies26–28 examined serial or parallel mediation models between climate change and mental health with PA, sleep, and varying environmental indicators, yielding a total of 39 observations. Of these, 62% of the reported observations showed null associations while 38% of the reported observations showed a favorable impact of mediators on the relationship between climate change indicators and mental health. Specifically, in serial mediation models examining the pathways between green/bluespace and mental health,28 higher greenness and bluespace within a 300-m buffer were associated with positive mental health via higher perceived green/bluespace, increased restorative quality, and increased PA. The uncertainty of evidence was high.
Two studies27,44 examined sleep as a mediator on the relationship between air pollution and health measures, with a total of 12 observations. Of 3 total effects (c path) reported, 67% showed a negative association between unhealthy sleep cycle/duration and health measures while 33% of the associations were null. Of these, 100% of the 3 reported observations became null when a mediator was introduced to the model (c′ path). At the end, of 12 observations made, healthy sleep cycle/duration showed a favorable indirect effect (ab path) on the relationship in 71% of the observations. The uncertainty of evidence was high.
In total, when movement behaviors (95% of studies had PA as a mediator) were examined as a mediator with or without other mediators, 33% of the observations indicated that movement behaviors may favorably mediate the relationship between climate change and health. The uncertainty of evidence was high.
One study31 examined the relationship between outdoor PA and lung function with O3 and NO2 as mediators among 151 children (Mage=12.5 y) in New York City. A positive association was found between outdoor moderate-to-vigorous PA and lung function during warmer months. Though the magnitude of the association was reduced after introducing air pollution as a mediator, neither NO2 nor PM2.5 mediated the association. The uncertainty of evidence was high.
Only the mediation models with a fair (may be relevant) or higher level of evidence and possessing >50% of the reported findings based on ≥5 studies were further synthesized. As a result, a potential mediating role of PA on the relationship between climate change and health is described in Figure 3. PA may favorably mediate the relationship between climate change and health measures (38%) and this role is enhanced when combined with other mediators, such as neighborhood green spaces and a sense of community. The uncertainty of evidence was high.
—A potential mediating role of physical activity on the relationship between different indicators of climate change and health measures. †Direction: Ø, null association; –, unfavorable association; +, favorable association; NR, not reported. ‡Strength of evidence: 0% to 20%, marginal (unlikely relevant); 21% to 40%, fair (may be relevant); 41% to 60%, moderate (likely relevant); 61% to 80%, strong (relevant); 81% to 100%, very strong (actionable).
Citation: Journal of Physical Activity and Health 21, 12; 10.1123/jpah.2023-0637
Results of Moderation
The overall summary of moderation with different indicators of climate change and movement behaviors are described in Table 3. Of 64 studies employing moderation,31,35,39,40,45–96,98–104 47 nonmutually exclusive studies examined movement behaviors as moderators (PA [n = 46],35,39,45–48,50,52,53,60,61,66–77,80–96,98–102,104 sleep [n = 2],52,103 SB [n = 3]45,52,66) on the relationship between indicators of climate change and health measures. Of 296 observations made from those studies that examined one or two of the 24-hour movement behaviors as a moderator, 26% of the studies indicated that 24-hour movement behaviors favorably modified the negative association between climate change and health. A majority of the evidence was driven by observations that examined PA as a moderator (90%), with 45% of the observations indicating the negative association between climate change and health measures. The uncertainty of evidence was high.
Overall Summary of Moderation (n = 64 Studies)
| Exposure | Moderator | Outcome | No. of studies reported | No. of observations made | Main association reported | Main association†,‡ | No. of interaction reported | Effect modification†,‡ | Uncertainty of evidenceß |
|---|---|---|---|---|---|---|---|---|---|
| Climate change indicators | PA | Varying health measures | 46§ | 274 | 260‡ | Ø = 27% − = 47% + = 26% | 112§ | Ø = 53% − = 19% + = 28% | <0.0001 |
| Climate change indicators | SB | Varying health measures | 3§ | 10 | 6 | Ø = 67% − = 33% + = 0% | 10§ | Ø = 60% − = 30% + = 10% | 0.0039 |
| Air pollution | Sleep duration | Varying health measures | 2§ | 8 | 8 | Ø = 67% − = % + = 33% | 8§ | Ø = 88% − = 0% + = 13% | 0.0002 |
| Overall summary of the association between climate change and health with movement behavior as moderator | |||||||||
| Climate change indicators | Movement behavior (90% of PA) | Varying health measures | 47 | 296 | 281† | Ø = 30% − = 45% + = 25% | 118§ | Ø = 56% − = 18% + = 26% | <0.001 |
| Exposure | Moderator | Outcome | No. of studies reported | No. of observations made | Main association reported | Main association†,‡ | No. of interactions reported | Effect modification†,‡ | Uncertainty of evidenceß |
|---|---|---|---|---|---|---|---|---|---|
| PA | Air pollution | Varying health measures | 20¶ | 138 | 117 | Ø = 30% − = 25% + = 45% | 34 | Ø = 77% − = 17% + = 6% | <0.0001 |
| PA | Greenspace/environmental quality | Varying health measures | 3¶ | 6 | 6 | Ø = 16% − = 0% + = 83% | 2 | Ø = 20% − = 0% + = 80% | 0.9846 |
| Outdoor PA | Temperature (warm) | Lung function | 1 | 9 | 9 | Ø = 100% − = 0% + = 0% | 8 | Ø =75% − = 25% + = 0% | 0.0039 |
| Overall summary of the association between PA and health with climate change as moderator | |||||||||
| PA | Climate change indicators (86% of air pollution) | Varying health measures | 22 | 153 | 132 | Ø = 34% − = 22% + = 44% | 36 | Ø = 75% − = 17% + = 8% | <0.0001 |
Abbreviations: PA, physical activity; SB, sedentary behavior.
†Direction: Ø = null association; − = unfavorable association; + = favorable association; NR = not reported. ‡Strength of evidence: 0% to 20% = marginal (unlikely relevant); 21% to 40%: fair (may be relevant); 41% to 60%: moderate (likely relevant); 61% to 80%: strong (relevant); 81% to 100%: very strong (actionable). §Four nonmutually exclusive studies examined more than 2 movement behaviors as effect modifiers. ¶One nonmutually exclusive studies examined both air pollution and normalized difference vegetation index as effect modifiers. ßUncertainty of evidence is indicated by calculating the posterior Bayesian probability of association: P(A|Xn) = “0” indicates the greatest uncertainty and “1” indicates the least uncertainty of the evidence.
A total of 22 nonmutually exclusive studies examined climate change indicators as moderators on the relationship between PA and varying health measures.31,39,40,46,49–51,53–65,67,104 Of these, 20 studies used air pollutants (PM2.5, PM10, NO2, black carbon, ozone), 3 studies used greenspace/environmental quality, and one study used meteorological parameters to indicate climate change. Of 153 observations made, 132 reported observations indicated 34% null, 22% negative, and 44% positive main associations between PA and health measures. Of these, climate change indicators did not modify the association in 75% of the observations. The uncertainty of evidence was high.
Only the moderation models with a fair (may be relevant), or higher level of evidence, and possessing >50% of the reported findings based on ≥5 studies were further synthesized. A potential moderating role of PA on the association between climate change and health is described in Figure 4. In general, the potential role of PA and its direction on the association between climate change indicators and health measures are largely inconclusive (19% unfavorable and 25% favorable with 56% null). The uncertainty of evidence was high.
—A potential moderating role of physical activity on the relationship between different indicators of climate change and health measures. †Direction: Ø, null association; –, unfavorable association; +, favorable association; NR, not reported. ‡Strength of evidence: 0% to 20%, marginal (unlikely relevant); 21% to 40%, fair (may be relevant); 41% to 60%, moderate (likely relevant); 61% to 80%, strong (relevant); 81% to 100%, very strong (actionable).
Citation: Journal of Physical Activity and Health 21, 12; 10.1123/jpah.2023-0637
Discussion
A potential interplay between climate change and 24-hour movement behaviors, especially PA and sleep, and their combined influence on human health is emerging. However, mechanisms linking climate change, 24-hour movement behaviors, and health remain less understood. This systematic review synthesized the evidence on the potential mechanisms between varying indicators of climate change, 24-hour movement behaviors, and health among 6.6 million individuals from 25 countries with varying income levels. Our overall findings indicate that PA may play both mediating and moderating roles in the association between certain indicators of climate change (ie, air pollutants and green/blue spaces) and health outcomes. However, it is important to note that the strength of evidence ranged from fair to moderate. Conversely, indicators of climate change, primarily air pollution, do not appear to modify or mediate the association between PA and health measures. Furthermore, based on the posterior Bayesian probability of association, the results were, in general, highly heterogeneous and uncertain.
Building upon previous reviews5,6 suggesting bidirectional associations between climate change and PA, our review further investigated the potential intermediatory role of PA on the relationship between the indicators of climate change and health measures. Though our review found the potential mediating role of PA in the relationship between climate change and health measures, an interesting narrative emerged when the interplay of PA with other upstream factors was considered. Specifically, when PA was combined with factors like low air pollution, increased greenspace, higher tree cover density, enhanced social cohesion, or improved restorative quality, it appeared to amplify the potential mediating impact of PA.26–28 Such potential role of PA in mitigating the adverse health measures related to climate change suggests that PA could serve as a complementary strategy alongside broader upstream efforts to address both climate change and public health challenges. Therefore, PA promotion efforts could be combined with upstream efforts at community, regional, and national governmental levels to increase greenspace, reduce air pollution, and foster healthier communities.
Our review also indicated that PA may buffer or amplify the negative impact of climate change on health. In our review, any dose of PA was shown to be beneficial to health, and greater than the potential harmful effects of climate change.35,40,46,47,60,67,68,70,71,75,77,93,95,96 However, some contradicting results were also observed when analyses were further stratified by the exposure of air pollution in addition to PA levels. Specifically, the benefits of PA were attenuated or canceled in highly polluted areas, particularly for lung health.73,74,81,84,85,88,92,100 Authors of these studies speculated that those who engage in outdoor PA in highly polluted areas may experience the detrimental effect of air pollution on health. Nonetheless, studies that detected no interactions between climate change and PA showed that higher PA is positively associated with health in less polluted areas.70,89 Combined, the health benefits of PA remain apparent, regardless of the exposure to air pollution; however, engaging in PA outdoors in highly polluted areas may negate these benefits. This is discouraging given that outdoor PA, particularly in the form of play and active travel holds the potential to bring numerous health benefits to humans in the face of challenges in today’s world, including climate change.15,105,106 Future work investigating the health benefits, as well as potentially harmful effects of PA by time, type, intensity, location, and context in relation to climate change, particularly air pollution, is warranted to generate evidence for informed policy making.107
The potential adverse impact of climate change, particularly air pollution, on PA was highlighted in previous reviews.5–7 Though evidence is marginal, one study included in our review examined a potential mediating role of air pollution on the relationship between indoor/outdoor PA and lung function among 151 children (6–14 y) living in New York City.31 It concluded that though outdoor PA was associated with lower lung function during warmer months, this association was not mediated by higher exposure to outdoor pollution during outdoor PA. More research is warranted to confirm these findings; however, lower PA and subsequent adverse health impacts may not be solely attributable to air pollution. These findings were also confirmed by the synthesis of moderation in our review. Specifically, when the moderating role of climate change, predominantly air pollution, was examined on the relationship between PA and health, the positive association between PA and health did not change at different levels of exposure to air pollutants.31,47,49,50,53,55,57–59,63,64
Relevant evidence on sleep and SB in relation to climate change and health was limited and no informed conclusions can be made through our review. Sleep was examined as a mediator27,44 in 2 studies, and as a moderator in 2 other studies,52,103 and the role it played was marginal for mediation and fair for moderation, but moderation was only based on one study.103 These findings indicate that sleep unlikely explains or modifies the relationship between climate change and health. Climate change could lead to detrimental sleep health, such as sleep disturbance, short sleep duration, and poor sleep quality.5,11 Because the behavior–health outcome relationship is more immediate and direct with sleep, while the health impact of PA and SB could be more long-term and less direct, sleep health may be more relevant to the everyday experience of climate change, such as extreme heat and related psychological distress, also known as eco-anxiety.108,109 Only 2 studies examined SB as a mediator and found no interaction between air pollution and SB on depression52 or blood pressure.45 In both studies, SB was added in addition to PA and other health-related behaviors. Indeed, the theorized link between climate change, SB, and health is not established. However, given that SB can be replaced with PA,110 an indirect role that SB may play through PA cannot be dismissed.
The potential bidirectional relationship between climate change and behavioral outcomes and the impact of climate change on health outcomes were investigated, respectively, in previous reviews.5–7,11 Building on these reviews and based on the evidence from both high-income countries and low- and middle-income countries, this systematic review provides a comprehensive synthesis of evidence regarding the potential mechanisms linking the indicators of climate change, 24-hour movement behaviors (particularly PA), and health measures. To enhance the robustness of the evidence presented in this review, future work should differentiate between indoor and outdoor locations as well as the domain of PA (eg, work, transport, leisure, play) in their assessment of PA. A notable gap in the evidence identified within this review was the absence or lack of a clear distinction between indoor and outdoor PA in relation to exposure to air pollution, leading to studies presenting mixed or contradictory findings regarding the role of PA. Furthermore, the health impact and exposure to air pollutants may differ greatly between the domain of activity (eg, work vs play),107 as well as the time of day (eg, day vs evening). Recent work also suggested that outdoor PA, particularly in the form of play, may foster environmental stewardship that could strengthen the synergistic interplay between PA, health, and the ecosystem.105,106 The capacity to engage in PA and the ensuing health effects of such activities may diverge significantly based on the location and type within and across diverse communities, areas, countries, and regions and also based on environmental conditions (eg, extreme heat, presence of natural disasters). As such, it is important to measure and report the specific location of PA engagement. All movement behavior research would benefit from more robust measures to better explore and understand relationships, with indicators of climate change and human health.
It is important to acknowledge that our search may not have captured all eligible articles, possibly due to the selection of databases we utilized. Nevertheless, we updated our search in April 2024 to ensure the inclusion of the most up-to-date evidence. Considering the growing interest in this research topic, it is crucial that the review be periodically updated to provide timely insights for informing policy-making decisions. While this review offers a comprehensive overview of the connections between indicators associated with climate change, movement behaviors, and health, it is important to note that these relationships may vary depending on the specific indicators employed. Consequently, policies aimed at addressing particular health outcomes, such as mental health, should also take into account the specific climate change and behavioral indicators that are considered most pertinent to the desired outcomes. In a similar vein, this review provided relationships by different indicators of climate change, movement behaviors, and health when sufficient evidence was available. However, the potential mechanisms were explained by aggregating results from different indicators of climate change, which may neglect the potential complexity of these relationships. As more studies examine behavioral mediators and moderators of the relationship between different indicators of climate change and health outcomes, future reviews on this topic should generate evidence specific to each climate change indicator and health outcome and the potential role a specific movement behavior may play in that relationship. Furthermore, regarding the large number of reported observations based on studies examining green/blue spaces, we considered these as climate change indicators due to their prominence in the literature, which demonstrates their mediating role in the relationship between environmental factors and health measures. This also became apparent during our screening process. Nonetheless, we recognize that this inclusion may appear selective and that the exclusion of other adaptation strategies, such as urban or personal cooling strategies, could limit the comprehensiveness of our review. Future reviews should aim to include a broader range of adaptation strategies to provide a more comprehensive understanding of how different environmental modifications as adaptation strategies to climate change impact movement behavior and health outcomes related to climate change. Finally, subgroup analyses were not conducted due to the lack of stratification based on subgroups in the included studies (12%). Developing strategies for climate change adaptation, preparedness, and resilience with an equity lens is important given that disadvantaged countries, communities, and individuals are disproportionally impacted by climate change.2,111,112
Conclusions
A comprehensive, solution-driven systems approach is essential for planetary health. Our review suggested that PA may play both mediating and moderating roles in the link between climate change indicators and health outcomes. However, it is important to note that the evidence was marginal with fair to moderate strength, with high uncertainty. While PA may mitigate the adverse health impact of climate change, further evidence is needed to fully integrate PA into future climate change mitigation, adaptation, and resilience strategies.
Acknowledgments
The authors thank Ajaypal Bains, Shawn Hakimi, Lucy Li, and Charlotte Lipin for their assistance during the review process. Funding source: This study was funded by Queen’s Research Opportunities Fund—Catalyst Grant.
References
- 1.↑
Hottest July ever signals “era of global boiling has arrived” says UN chief. 2023. Accessed August 16, 2023. https://news.un.org/en/story/2023/07/1139162
- 2.↑
Intergovernmental Panel on Climate Change. Synthesis Report of the IPCC Sixth Assessment Report—Longer Report. Intergovernmental Panel on Climate Change; 2023.
- 3.↑
Pongsiri MJ, Gatzweiler FW, Bassi AM, Haines A, Demassieux F. The need for a systems approach to planetary health. Lancet Planet Health. 2017;1(7):e257–e259. doi:
- 4.↑
Reis R, Hunter RF, Garcia L, Salvo D. What the physical activity community can do for climate action and planetary health? J Phys Act Health. 2022;19(1):2–3. doi:
- 5.↑
Zisis E, Hakimi S, Lee EY. Climate change, 24-hour movement behaviors, and health: a mini umbrella review. Glob Health Res Policy. 2021;6(1):15. doi:
- 6.↑
Bernard P, Chevance G, Kingsbury C, et al. Climate change, physical activity and sport: a systematic review. Sports Med. 2021;51(5):1041–1059. doi:
- 7.↑
Kim YB, McCurdy AP, Lamboglia CG, et al. Ambient air pollution and movement behaviours: a scoping review. Health Place. 2021;72:102676. doi:
- 8.↑
Beggs PJ, Zhang Y, McGushin A, et al. The 2021 report of the MJA-Lancet Countdown on health and climate change: Australia increasingly out on a limb. Med J Aust. 2021;215(9):390–392.e322. doi:
- 9.↑
Charlson F, Ali S, Augustinavicius J, et al. Global priorities for climate change and mental health research. Environ Int. 2022;158:106984. doi:
- 10.↑
Watts N, Adger WN, Ayeb-Karlsson S, et al. The Lancet Countdown: tracking progress on health and climate change. Lancet. 2017;389(10074):1151–1164. doi:
- 11.↑
Rifkin DI, Long MW, Perry MJ. Climate change and sleep: A systematic review of the literature and conceptual framework. Sleep Med Rev. 2018;42:3–9. doi:
- 12.↑
Gaston SA, Singh R, Jackson CL. The need to study the role of sleep in climate change adaptation, mitigation, and resiliency strategies across the life course. Sleep. 2023;46(7):zsad070. doi:
- 13.↑
Ross R, Tremblay M. Introduction to the Canadian 24-Hour Movement Guidelines for Adults aged 18–64 years and Adults aged 65 years or older: an integration of physical activity, sedentary behaviour, and sleep. Appl Physiol Nutr Metab. 2020;45(10):v–xi. doi:
- 14.↑
Tremblay MS, Carson V, Chaput JP, et al. Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep. Appl Physiol Nutr Metab. 2016;41(6 ):S311–S327. doi:
- 15.↑
Lee EY, Tremblay MS. Unmasking the political power of physical activity research: harnessing the “apolitical-ness” as a catalyst for addressing the challenges of our time. J Phys Act Health. 2023;20(10):897–899. doi:
- 16.↑
Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg. 2021;88:105906. doi:
- 17.↑
Elmore SA, Weston EH. Predatory journals: what they are and how to avoid them. Toxicol Pathol. 2020;48(4):607–610. doi:
- 18.↑
Colquhoun D. An investigation of the false discovery rate and the misinterpretation of p-values. R Soc Open Sci. 2014;1(3):140216. doi:
- 19.↑
Young NS, Ioannidis JP, Al-Ubaydli O. Why current publication practices may distort science. PLoS Med. 2008;5(10):e201. doi:
- 20.↑
United Nations. What is climate change? https://www.un.org/en/climatechange/what-is-climate-change
- 21.↑
Hayes AF. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. Guilford publications; 2017.
- 22.↑
MacKinnon DP, Luecken LJ. How and for whom? Mediation and moderation in health psychology. Health Psychol. 2008;27(2S):S99–S100. doi:
- 23.↑
Polanin JR, Pigott TD, Espelage DL, Grotpeter JK. Best practice guidelines for abstract screening large-evidence systematic reviews and meta-analyses. Res Synth Methods. 2019;10(3):330–342.
- 25.↑
Chastin SF, Egerton T, Leask C, Stamatakis E. Meta‐analysis of the relationship between breaks in sedentary behavior and cardiometabolic health. Obesity. 2015;23(9):1800–1810. doi:
- 26.↑
Dzhambov A, Hartig T, Markevych I, Tilov B, Dimitrova D. Urban residential greenspace and mental health in youth: different approaches to testing multiple pathways yield different conclusions. Environ Res. 2018;160:47–59. doi:
- 27.↑
Dzhambov AM, Markevych I, Tilov B, et al. Pathways linking residential noise and air pollution to mental ill-health in young adults. Environ Res. 2018;166:458–465. doi:
- 28.↑
Dzhambov AM, Markevych I, Hartig T, et al. Multiple pathways link urban green- and bluespace to mental health in young adults. Environ Res. 2018;166:223–233. doi:
- 29.↑
Huang B, Xiao T, Grekousis G, et al. Greenness-air pollution-physical activity-hypertension association among middle-aged and older adults: evidence from urban and rural China. Environ Res. 2021;195:110836. doi:
- 30.↑
Guan Q, Zhu C, Zhang G, et al. Association of land urbanization and type 2 diabetes mellitus prevalence and mediation of greenness and physical activity in Chinese adults. Environ Pollut. 2023;337:122579. doi:
- 31.↑
Lovinsky-Desir S, Jung KH, Montilla M, et al. Locations of Adolescent Physical Activity in an Urban Environment and Their Associations with Air Pollution and Lung Function. Ann Am Thorac Soc. 2021;18(1):84–92. doi:
- 32.↑
Dzhambov AM. Residential green and blue space associated with better mental health: a pilot follow-up study in university students. Arh Hig Rada Toksikol. 2018;69(4):340–349. doi:
- 33.↑
Gascon M, Sanchez-Benavides G, Dadvand P, et al. Long-term exposure to residential green and blue spaces and anxiety and depression in adults: a cross-sectional study. Environ Res. 2018;162:231–239. doi:
- 34.↑
Huang WZ, Yang BY, Yu HY, et al. Association between community greenness and obesity in urban-dwelling Chinese adults. Sci Total Environ. 2020;702:135040. doi:
- 35.↑
James P, Hart JE, Banay RF, Laden F. Exposure to greenness and mortality in a nationwide Prospective Cohort Study of Women. Environ Health Perspect. 2016;124(9):1344–1352. doi:
- 36.↑
Mila C, Ranzani O, Sanchez M, et al. Land-use change and cardiometabolic risk factors in an urbanizing area of South India: a population-based cohort study. Environ Health Perspect. 2020;128(4):47003.
- 37.↑
Triebner K, Markevych I, Bertelsen RJ, et al. Lifelong exposure to residential greenspace and the premenstrual syndrome: a population-based study of Northern European women. Environ Int. 2022;158:106975. doi:
- 38.↑
Wang R, Helbich M, Yao Y, et al. Urban greenery and mental wellbeing in adults: cross-sectional mediation analyses on multiple pathways across different greenery measures. Environ Res. 2019;176:108535. doi:
- 39.↑
Wang R, Liu Y, Xue D, Yao Y, Liu P, Helbich M. Cross-sectional associations between long-term exposure to particulate matter and depression in China: the mediating effects of sunlight, physical activity, and neighborly reciprocity. J Affect Disord. 2019;249:8–14. doi:
- 40.↑
Wang R, Yang B, Liu P, et al. The longitudinal relationship between exposure to air pollution and depression in older adults. Int J Geriatr Psychiat. 2020;35(6):610–616. doi:
- 41.↑
Yang BY, Markevych I, Bloom MS, et al. Community greenness, blood pressure, and hypertension in urban dwellers: the 33 Communities Chinese Health Study. Environ Int. 2019;126:727–734. doi:
- 42.↑
Yang BY, Markevych I, Heinrich J, et al. Residential greenness and blood lipids in urban-dwelling adults: the 33 Communities Chinese Health Study. Environ Pollut. 2019;250:14–22. doi:
- 43.↑
Zhang N, Wang L, Zhang M, Nazroo J. Air quality and obesity at older ages in China: the role of duration, severity and pollutants. PLoS One. 2019;14(12).
- 44.↑
Lo C-C, Liu W-T, Lu Y-H, et al. Air pollution associated with cognitive decline by the mediating effects of sleep cycle disruption and changes in brain structure in adults. Environ Sci Pollution Res Int. 2022;29(35):52355–52366. doi:
- 45.↑
Avila-Palencia I, Laeremans M, Hoffmann B, et al. Effects of physical activity and air pollution on blood pressure. Environ Res. 2019;173:387–396. doi:
- 46.↑
Coleman CJ, Yeager RA, Pond ZA, Riggs DW, Bhatnagar A, Arden Pope C 3rd. Mortality risk associated with greenness, air pollution, and physical activity in a representative U.S. cohort. Sci Total Environ. 2022;824:153848. doi:
- 47.↑
Jiang H, Zhang S, Yao X, et al. Does physical activity attenuate the association between ambient PM2.5 and physical function? Sci Total Environ. 2023;874:162501. doi:
- 48.↑
Jiang D, Wang L, Han X, et al. Short-term effects of ambient oxidation, and its interaction with fine particles on first-ever stroke: a national case-crossover study in China. Sci Total Environ. 2024;907:168017. doi:
- 49.↑
Kubesch NJ, Therming Jorgensen J, Hoffmann B, et al. Effects of leisure-time and transport-related physical activities on the risk of incident and recurrent myocardial infarction and interaction with traffic-related air pollution: a cohort study. J Am Heart Assoc. 2018;7(15):18. doi:
- 50.↑
Laeremans M, Dons E, Avila-Palencia I, et al. Short-term effects of physical activity, air pollution and their interaction on the cardiovascular and respiratory system. Environ Int. 2018;117:82–90. doi:
- 51.↑
Li Z-H, Zhong W-F, Zhang X-R, et al. Association of physical activity and air pollution exposure with the risk of type 2 diabetes: a large population-based prospective cohort study. Environ Health. 2022;21(1):106. doi:
- 52.↑
Li D, Xie J, Wang L, Sun Y, Hu Y, Tian Y. Genetic susceptibility and lifestyle modify the association of long-term air pollution exposure on major depressive disorder: a prospective study in UK Biobank. BMC Med. 2023;21(1):67. doi:
- 53.↑
Zhang Z, Hoek G, Chang LY, et al. Particulate matter air pollution, physical activity and systemic inflammation in Taiwanese adults. Int J Hyg Environ Health. 2018;221(1):41–47. doi:
- 54.↑
Gray CL, Messer LC, Rappazzo KM, Jagai JS, Grabich SC, Lobdell DT. The association between physical inactivity and obesity is modified by five domains of environmental quality in U.S. adults: a cross-sectional study. PLoS One. 2018;13(8):e0203301. doi:
- 55.↑
Andersen ZJ, Nazelle AD, Mendez MA, et al. A study of the combine effects of physical activity and air pollution on mortality in elderly urban residents: the Danish diet, cancer, and health cohort. Environ Health Perspect. 2015;123(6):557–563. doi:
- 56.↑
Elavsky S, Burda M, Cipryan L, et al. Physical activity and menopausal symptoms: evaluating the contribution of obesity, fitness, and ambient air pollution status. Menopause. 2024;31(4):310–319. doi:
- 57.↑
Elliott EG, Laden F, James P, Rimm EB, Rexrode KM, Hart JE. Interaction between long-term exposure to fine particulate matter and physical activity, and risk of cardiovascular disease and overall mortality in U.S. women. Environ Health Perspect. 2020;128(12):127012. doi:
- 58.↑
Fisher JE, Loft S, Ulrik CS, et al. Physical activity, air pollution, and the risk of asthma and chronic obstructive pulmonary disease. Am J Respir. 2016;194(7):855–865.
- 59.↑
Hou J, Liu X, Tu R, et al. Long-term exposure to ambient air pollution attenuated the association of physical activity with metabolic syndrome in rural Chinese adults: a cross-sectional study. Environ Int. 2020;136:105459. doi:
- 60.↑
Kim SR, Choi D, Choi S, et al. Association of combined effects of physical activity and air pollution with diabetes in older adults. Environ Int. 2020;145:106161. doi:
- 61.↑
Kim SR, Choi S, Kim K, et al. Association of the combined effects of air pollution and changes in physical activity with cardiovascular disease in young adults. Euro Heart J. 2021;42(25):2487–2497. doi:
- 62.↑
Luo H, Huang Y, Zhang Q, et al. Impacts of physical activity and particulate air pollution on the onset, progression and mortality for the comorbidity of type 2 diabetes and mood disorders. Sci Total Environ. 2023;890:164315. doi:
- 63.↑
Oudin A, Stromberg U, Jakobsson K, et al. Hospital admissions for ischemic stroke: does long-term exposure to air pollution interact with major risk factors? Cerebrovasc Dis. 2011;31(3):284–293. doi:
- 64.↑
Sun S, Cao W, Qiu H, et al. Benefits of physical activity not affected by air pollution: a prospective cohort study. Int J Epidemiol. 2020;49(1):142–152. doi:
- 65.↑
Chen L, Gao D, Ma T, et al. Could greenness modify the effects of physical activity and air pollutants on overweight and obesity among children and adolescents? Science Total Environ. 2022;832:155117. doi:
- 66.↑
Tang T, Zhou X, Zhang Y, et al. Investigation into the thermal comfort and physiological adaptability of outdoor physical training in college students. Sci Total Environ. 2022;839:155979. doi:
- 67.↑
Ao L, Zhou J, Han M, et al. The joint effects of physical activity and air pollution on type 2 diabetes in older adults. BMC Geriatr. 2022;22(1):472. doi:
- 68.↑
Chen C-H, Huang L-Y, Lee K-Y, et al. Effects of PM2.5 on skeletal muscle mass and body fat mass of the elderly in Taipei, Taiwan. Sci Rep. 2019;9(1):11176. doi:
- 69.↑
Coogan PF, White LF, Yu J, et al. Long term exposure to NO2 and diabetes incidence in the Black Women’s Health Study. Environ Res. 2016;148:360–366. doi:
- 70.↑
Endes S, Schaffner E, Caviezel S, et al. Is physical activity a modifier of the association between air pollution and arterial stiffness in older adults: the SAPALDIA cohort study. Int J Hyg Environ Health. 2017;220(6):1030–1038. doi:
- 71.↑
Eze IC, Schaffner E, Foraster M, et al. Long-term exposure to ambient air pollution and metabolic syndrome in adults. PLoS One. 2015;10(6). doi:
- 72.↑
Gandini M, Scarinzi C, Bande S, et al. Long term effect of air pollution on incident hospital admissions: results from the Italian Longitudinal Study within LIFE MED HISS project. Environ Int. 2018;121(pt 2):1087–1097. doi:
- 73.↑
Gao Y, Chan EYY, Zhu Y, Wong TW. Adverse effect of outdoor air pollution on cardiorespiratory fitness in Chinese children. Atmos Environ. 2013;64:10–17. doi:
- 74.↑
Guo Q, Zhao Y, Zhao J, et al. Physical activity attenuated the associations between ambient air pollutants and metabolic syndrome (MetS): a nationwide study across 28 provinces. Environ Pollut. 2022;315:120348. doi:
- 75.↑
Hou J, Duan Y, Liu X, et al. Associations of long-term exposure to air pollutants, physical activity and platelet traits of cardiovascular risk in a rural Chinese population. Sci Total Environ. 2020;738:140182. doi:
- 76.↑
Ju K, Lu L, Wang W, et al. Causal effects of air pollution on mental health among Adults—An exploration of susceptible populations and the role of physical activity based on a longitudinal nationwide cohort in China. Environ Res. 2023;217:114761. doi:
- 77.↑
Kim KN, Lee H, Kim JH, Jung K, Lim YH, Hong YC. Physical activity- and alcohol-dependent association between air pollution exposure and elevated liver enzyme levels: an elderly panel study. J Prev Med Public Health. 2015;48(3):151–169. doi:
- 78.↑
Kim KJ, Shin J, Choi J. Cancer risk from exposure to particulate matter and ozone according to obesity and health-related behaviors: a nationwide population-based cross-sectional study. Cancer Epidemiol Biomarkers Prev. 2019;28(2):357–362. doi:
- 79.↑
Laffan K. Every breath you take, every move you make: visits to the outdoors and physical activity help to explain the relationship between air pollution and subjective wellbeing. Ecol Econ. 2018;147:96–113. doi:
- 80.↑
Lamichhane DK, Leem JH, Kim HC. Associations between ambient particulate matter and nitrogen dioxide and chronic obstructive pulmonary diseases in adults and effect modification by demographic and lifestyle factors. Int J Environ Res Public Health. 2018;15(2):363. doi:
- 81.↑
Lao XQ, Guo C, Chang LY, et al. Long-term exposure to ambient fine particulate matter (PM2.5) and incident type 2 diabetes: a longitudinal cohort study. Diabetologia. 2019;62(5):759–769. doi:
- 82.↑
Li D, Wang JB, Yu ZB, Lin HB, Chen K. Air pollutants concentration and variation of blood glucose level among pregnant women in China: a cross-sectional study. Atmos Environ. 2020;223:117191. doi:
- 83.↑
Li N, Chen G, Liu F, et al. Associations between long-term exposure to air pollution and blood pressure and effect modifications by behavioral factors. Environ Res. 2020;182:109109. doi:
- 84.↑
Lin H, Guo Y, Di Q, et al. Ambient PM2.5 and stroke: effect modifiers and population attributable risk in six low- and middle-income countries. Stroke. 2017;48(5):1191–1197. doi:
- 85.↑
McConnell R, Berhane K, Gilliland F, et al. Asthma in exercising children exposed to ozone: a cohort study. Lancet. 2002;359(9304):386–391. doi:
- 86.↑
Park W, Jang H, Ko J, et al. Physical activity-induced modification of the association of long-term air pollution exposure with the risk of depression in older adults. Yonsei Med J. 2024;65(4):227. doi:
- 87.↑
Puett RC, Schwartz J, Hart JE, et al. Chronic particulate exposure, mortality, and coronary heart disease in the nurses’ health study. Am J Epidemiol. 2008;168(10):1161–1168. doi:
- 88.↑
Raichlen DA, Furlong M, Klimentidis YC, et al. Association of physical activity with incidence of dementia is attenuated by air pollution. Med Sci Sports Exerc. 2022;54(7):1131–1138. doi:
- 89.↑
Raza W, Krachler B, Forsberg B, Sommar JN. Does physical activity modify the association between air pollution and recurrence of cardiovascular disease? Int J Environ Res Public Health. 2021;18(5):2631. doi:
- 90.↑
Roswall N, Poulsen AH, Thacher JD, et al. Nighttime road traffic noise exposure at the least and most exposed facades and sleep medication prescription redemption-a Danish cohort study. Sleep. 2020;43(8):zsaa029. doi:
- 91.↑
Tallon LA, Manjourides J, Pun VC, Salhi C, Suh H. Cognitive impacts of ambient air pollution in the National Social Health and Aging Project (NSHAP) cohort. Environ Int. 2017;104:102–109. doi:
- 92.↑
Thiering E, Markevych I, Bruske I, et al. Associations of residential long-term air pollution exposures and satellite-derived greenness with insulin resistance in German adolescents. Environ Health Perspect. 2016;124(8):1291–1298. doi:
- 93.↑
Tu R, Hou J, Liu X, et al. Physical activity attenuated association of air pollution with estimated 10-year atherosclerotic cardiovascular disease risk in a large rural Chinese adult population: a cross-sectional study. Environ Int. 2020;140:105819. doi:
- 94.↑
Vencloviene J, Tamosiunas A, Radisauskas R, et al. The influence of the North Atlantic Oscillation index on arterial blood pressure. J Hypertens. 2019;37(3):513–521. doi:
- 95.↑
Wu H, Zhang Y, Wei J, et al. Association between short-term exposure to ambient PM1 and PM2.5 and forced vital capacity in Chinese children and adolescents. Environ Sci Pollution Res Int. 2022;29(47):71665–71675. doi:
- 96.↑
Wu M, Xie J, Zhou Z, et al. Fine particulate matter, vitamin D, physical activity, and major depressive disorder in elderly adults: results from UK Biobank. J Affect Disord. 2022;299:233–238. doi:
- 97.↑
Xu J, Zhou J, Luo P, et al. Associations of long-term exposure to ambient air pollution and physical activity with insomnia in Chinese adults. Sci Total Environ. 2021;792:148197. doi:
- 98.↑
Yang BY, Qian ZM, Li S, et al. Long-term exposure to ambient air pollution (including PM1 and metabolic syndrome: the 33 Communities Chinese Health Study (33CCHS). Environ Res. 2018;164:204–211. doi:
- 99.↑
Yu W, Sulistyoningrum DC, Gasevic D, et al. Long-term exposure to PM2.5 and fasting plasma glucose in non-diabetic adolescents in Yogyakarta, Indonesia. Environ Pollut. 2020;257:113423. doi:
- 100.↑
Yu Y, Jerrett M, Paul KC, et al. Ozone exposure, outdoor physical activity, and incident type 2 diabetes in the SALSA cohort of Older Mexican Americans. Environ Health Perspect. 2021;129(9):97004. doi:
- 101.↑
Zhang S, Wolf K, Breitner S, et al. Long-term effects of air pollution on ankle-brachial index. Environ Int. 2018;118:17–25. doi:
- 102.↑
Zhang Z, Zhao D, Hong YS, et al. Long-term particulate matter exposure and onset of depression in middle-aged men and women. Environ Health Perspect. 2019;127(7):077001. doi:
- 103.↑
Shi W, Chen C, Cui Q, et al. Sleep disturbance exacerbates the cardiac conduction abnormalities induced by persistent heavy ambient fine particulate matter pollution: a multi-center cross-sectional study. Sci Total Environ. 2022;838(pt 4):156472. doi:
- 104.↑
Li Z-H, Song W-Q, Qiu C-S, et al. Long-term air pollution exposure, habitual physical activity, and incident chronic kidney disease. Ecotoxicol Environ Safety. 2023;265:115492. doi:
- 105.↑
Lee EY, de Lannoy L, Li L, et al. Correction: Play, Learn, and Teach Outdoors—Network (PLaTO-Net): terminology, taxonomy, and ontology. Int J Behav Nutr Phys Act. 2023;20(1):2. doi:
- 106.↑
Lee EY, de Lannoy L, Li L, et al. Play, Learn, and Teach Outdoors-Network (PLaTO-Net): terminology, taxonomy, and ontology. Int J Behav Nutr Phys Act. 2022;19(1):66. doi:
- 107.↑
Hahad O, Daiber A, Munzel T. Physical activity in polluted air: an urgent call to study the health risks. Lancet Planet Health. 2023;7(4):e266–e267. doi:
- 108.↑
Leger-Goodes T, Malboeuf-Hurtubise C, Mastine T, Genereux M, Paradis PO, Camden C. Eco-anxiety in children: a scoping review of the mental health impacts of the awareness of climate change. Front Psychol. 2022;13:872544.
- 109.↑
Howard C, Huston P. The health effects of climate change: know the risks and become part of the solutions. Can Commun Dis Rep. 2019;45(5):114–118. doi:
- 110.↑
Tremblay MS, Ross R. How should we move for health? The case for the 24-hour movement paradigm. CMAJ. 2020;192(49):E1728–E1729. doi:
- 111.↑
Canadian Public Health Association. Position Statement: Climate Change and Human Health. October 2019.
- 112.↑
Lee EY, Masuda J. The “freedom” to pollute? An ecological analysis of neoliberal capitalist ideology, climate culpability, lifestyle factors, and population health risk in 124 countries. Can J Public Health. 2021;112(5):877–887. doi:
