Skip to main content
Download PDF

Spatial association between green space and COPD mortality: a township-level ecological study in Chongqing, China

BMC Pulmonary Medicine volume 23, Article number: 89 (2023) Cite this article

Abstract

Background

There are regional differences in the effect of green space on mortality of Chronic obstructive pulmonary disease (COPD). We conduct an ecological study, using the administrative divisions of Chongqing townships in China as the basic unit, to investigate the association between COPD mortality and green space based on data of 313,013 COPD deaths in Chongqing from 2012 to 2020. Green space is defined by Fractional vegetation cover (FVC), which is further calculated based on the normalised vegetation index (NDVI) from satellite remote sensing imagery maps.

Methods

After processing the data, the non-linear relationship between green space and COPD mortality is revealed by generalised additive models; the spatial differences between green space and COPD mortality is described by geographically weighted regression models; and finally, the interpretive power and interaction of each factor on the spatial distribution of COPD mortality is examined by a geographic probe.

Results

The results show that the FVC local regression coefficients ranged from − 0.0397 to 0.0478, 63.0% of the regions in Chongqing have a positive correlation between green space and COPD mortality while 37.0% of the regions mainly in the northeast and west have a negative correlation. The interpretive power of the FVC factor on the spatial distribution of COPD mortality is 0.08.

Conclusions

Green space may be a potential risk factor for increased COPD mortality in some regions of Chongqing. This study is the first to reveal the relationship between COPD mortality and green space in Chongqing at the township scale, providing a basis for public health policy formulation in Chongqing.

Peer Review reports

Introduction

COPD is a common chronic disease with airflow obstruction and incomplete reversibility [1], and acute exacerbations of COPD increase the socio-economic burden [2]. It is a major cause of increased morbidity, mortality and health care costs for chronic diseases worldwide [3]. The prevalence of COPD in most regions of China is higher than that estimated by the World Health Organization Model [4], and the number of COPD-related deaths in China in 2013 was 910,809, accounting for 31.1% of all COPD deaths in the world [5]. In order to make effective public health policy on COPD, it is important to explore the risk factors for COPD deaths.

Green space is an integral part of the habitat, and studies have shown that people in urban areas with high green space coverage are at lower risk of chronic diseases, including cardiovascular disease [6], asthma [7] and diabetes [8]. Greater exposure to the natural environment can promote overall human health and well-being [9].

However, there are inconsistencies in the effects of green space on respiratory health. A study in the Netherlands found that increased green space would reduce the prevalence of COPD [10], and another study in the UK found similar findings [11]. However, in contrast to the two studies, a national cross-sectional study in China found that green space in community may be a risk factor for increased COPD prevalence, especially true in the northern and north-eastern China [12]. In addition, a Korean cross-sectional study and a Hong Kong cohort study found no significant association between green space and mortality from respiratory diseases [13, 14]. The inconsistency of these studies may imply that there is inter-regional variability in the effect of green space on COPD.

There is no concrete evidence for this inconsistency, but it is generally accepted that possible explanations include some subjective bias in the quantification of green space in some studies, or related to inter-regional differences in vegetation types and methods of COPD ascertainment [12]. On one hand, green space can reduce air pollution and thus reduce adverse respiratory effects [15], on the other hand, some plants may release volatile organic compounds (VOCs) [16]. A study in the USA found a positive correlation between asthma in children and levels of green space around their homes, attributed to VOCs [17].

It is an effective tool for future planning, health management and evaluation to clarify the factors contributing to the spatial pattern of disease. There are few studies on the relationship between the green space and the prevalence of COPD, and the influence of the green space on the COPD mortality may be different in different regions. In order to improve health management and control COPD mortality in Chongqing, we conducted an ecological study in Chongqing based on township administrative divisions as the basic unit, with the following objectives: (1) To investigate the spatial distribution of COPD mortality in Chongqing; (2) To reveal the influence of green space on COPD mortality in Chongqing and its regional differences.

Methods

Study area

This research is based on the administrative territorial entity of Chinese villages and towns, which is the fourth level of administrative division in China. The study area, Chongqing, is a municipality directly under the Central Government of China in southwest China, with a hilly and mountainous terrain and a population mainly concentrated in the nine districts of the main city of Chongqing. The total resident population is 32,124,000 (2022), covering an area of 82,400 square kilometers, including 38 districts and counties, with 1,031 township-level divisions [18].

Data

Health Data COPD mortality data was collected from Chongqing Center for Disease Prevention and Control from all districts and counties, with a total of 313,013 COPD deaths from 2012 to 2020. The COPD mortality rate for each township was then obtained by field calculation in QGIS.

Green Space In this study, FVC was used to quantify the green space of each area. FVC quantifies the denseness of vegetation and reflects its growth status, which is an important fundamental data for describing ecosystems and has been widely used in various fields [19,20,21]. In this study, the Sentinel-2A satellite was selected by the Google Earth Engine platform to obtain remote sensing images with less than 20% cloud cover in Chongqing in 2020, and the NDVI data were obtained by de-clouding, stitching and calculating. The vegetation coverage was then retrieved from the NDVI data using a pixel-wise dichotomous model [22], which is expressed as:

$${f}_{vc}=\frac{NDVI-{NDVI}_{soil}}{{NDVI}_{veg}-{NDVI}_{soil}}$$
(1)

The \({NDVI}_{soil}\) is the minimum value of the pure soil image, and \({NDVI}_{veg}\) is the maximum value of the pure vegetation image. The values of the cumulative frequency of the NDVI image were selected as the values of \({NDVI}_{soil}\) and \({NDVI}_{veg}\) respectively based on the 5% and 95% confidence intervals in the study. The NDVI frequency histograms were calculated after the removal of the water areas in order to avoid the influence of the large water areas. The final FVC raster data was obtained at 10 m resolution. Based on the raster data, the mean FVC values for each township in Chongqing were calculated in QGIS (Fig. 1).

Fig. 1
figure 1

Mean values of FVC by quintile for each township

Full size image

Other data: Population data for each township are based on the sixth census (2010), from the Chongqing Municipal Bureau of Statistics (2022), where the proportion of elderly population is calculated from the number of people over 65 years in each region and the proportion of gender is calculated by females. Air pollution is widely considered to be significantly associated with COPD mortality [23,24,25]. PM2.5 and PM10 data were obtained from a 1 km resolution daily raster data set produced by the State Key Laboratory of Remote Sensing Science, Beijing Normal University [26, 27], and the annual average raster data of PM2.5 and PM10 were calculated separately in this paper by the raster package of R.

The spatial distribution of medical resources also has an important impact on the spatial distribution of COPD mortality. In this paper, Point of Interest (POI) spatial distribution data of medical resources within Chongqing city of Gaode Map was obtained through python, and a total of 1777 POI points were obtained. In order to conduct further statistical analysis, the kernel density of these POI points was estimated by QGIS, where the influence radius was set to 10 km, and the corresponding weights were assigned according to the different levels of medical institutions of the POI points, and finally the raster of the density distribution of medical institutions within Chongqing was calculated.

Statistical analysis

The article summarizes the characteristics of the data in descriptive statistics. At the same time, in order to explore the spatial relationship of COPD mortality, a spatial autocorrelation analysis of COPD mortality was performed, and its spatial correlation was expressed by the Moran index [28, 29]. Then, to explore the binary relationship between the variables, a spearman correlation analysis [30] was performed on each variable to obtain the binary correlation coefficient between the variables. Based on the results of the spearman correlation coefficients, variables with significant collinearity and non-significant correlation with COPD mortality were removed and finally three variables of FVC, proportion of elderly population, and density of health facilities, were selected for regression analysis.

Based on the results of the spearman correlation analysis, the study first attempted to describe the relationship between COPD mortality and FVC using GAMs Model, which is able to examine the relationship between the dependent variable and multiple independent variables as well as to fit the model through a non-linear smoothing term [31]. The general expression for GAMs is:

$$g(Y)=\alpha +\sum_{j=1}^{n} {f}_{i}\left({x}_{j}\right)+\varepsilon$$
(2)

where Y is the dependent variable, \(g(Y)\) is the link function, \(\alpha\) is the intercept term, \({x}_{j}\) is the independent variable, \({f}_{i}\left({x}_{j}\right)\) is the smoothing function, and \(\varepsilon\) is the random error. Where the smoothing parameters will be calculated by the restricted maximum likelihood method of smoothing (REML) to ensure stable and reliable results [32]. The process of constructing the GAMs in this paper consists of three models with increasing adjustment levels, as follows:

$$\mathrm{Model} 1: g\left(Deaths\right)=\alpha +s\left(FVC\right)+\varepsilon$$
(3)
$$\mathrm{Model }2: g\left(Deaths\right)=\alpha +s\left(FVC\right)+s(Hospital)+\varepsilon$$
(4)
$$\mathrm{Model }3: g\left(Deaths\right)=\alpha +s\left(FVC\right)+s(Hospital)+s(Older)+\varepsilon$$
(5)

where \(Deaths\) is the COPD mortality rate in each township, \(FVC\) is the vegetation cover, \(Hospital\) is the spatial distribution density of medical institutions, and \(s()\) is the natural spline smoothing function.

Based on the results of the spatial autocorrelation analysis, GWR was performed on the independent and dependent variables in order to expose the spatial association between green space and COPD mortality. GWR is essentially an improved global regression model, where GWR fits a local regression equation at each spatial location, resulting in local regression coefficients that reflect the relationship between the independent and dependent variables for each township unit, as well as the spatial heterogeneity of each region [33, 34]. The general formula for the GWR model is:

$${y}_{i}={\beta }_{0}\left({u}_{i},{v}_{i}\right)+\sum_{k=1}^{t} {\beta }_{k}\left({u}_{i},{v}_{i}\right){x}_{ki}+{\varepsilon }_{i}$$
(6)

where \({u}_{i}\) is the latitude of the i-th location, \({v}_{i}\) is the longitude of the i-th location, \({\beta }_{0}\) is the regression constant for the i-th location, \({\beta }_{k}\) is the k-th regression parameter to be estimated for the i-th location, \({x}_{ki}\) is the observed value of the k-th variable for the i-th location, t is the number of independent variables, and \({\varepsilon }_{i}\) is the random error for i-th location. The spatial weight function of GWR affects the parameters such as local regression coefficients, and the Gauss function is used to determine the weights in this study. The value of bandwidth affects the analysis results of the model, and the method selected in this study is corrected Akaike Information Criterion (AICc) [35], which finally result in an optimal bandwidth of 18.

Finally, to further investigate the interaction between the spatial correlation intensity and variables of green space and COPD mortality, the Geo-detector was used for factor detection and interaction detection [36]. It is important to note that as the data used were continuous, the discrete transformation of spatial data into class variables was carried out before the detection of factors and interactions.

The construction of the GAMs model for this study was performed using the “mgcv” package in R (4.2.1), with the selection of spline curves relying on the “splines” package (R Core Team 2022); the GWR was constructed using the “GWmodel” package [37, 38]; the “GD” package was used for the Geo-detector [39]; and the spatial autocorrelation analysis was performed using GeoDa (1.10); the calculation and processing of raster data relied on the “raster” package [40], the “sf” package [41] and QGIS (3.26).

Results

Descriptive statistics and spearman correlation

The descriptive statistics for all variables are shown in Table 1. The multi-year mortality rates for each township administrative unit of COPD in Chongqing from 2012 to 2020 range from 0 to 0.1, with a mean of 0.0191 and a standard deviation standard deviation (SD) of 0.009. The Spearman correlation coefficient is able to describe the correlation between the variables. It takes values in the range − 1 to 1. A positive value means that the two variables are positively correlated, while a negative value means the opposite [30]. Figure 2 shows the Spearman correlation between all variables, with a correlation of 0.1 between COPD mortality and FVC (p < 0.01), and a significant correlation between FVC and PM2.5, PM10 (|correlation coefficient|> 0.85, p < 0.01), these variables were removed to avoid Multicollinearity effects. Figure 3 shows the COPD mortality rates by township administrative unit in Chongqing, with values ranging from 0 to 0.1007, with high values mainly in the central, northern and western regions of Chongqing around the municipality. The values of FVC range from 0.0223, to 0.928, with high values mainly in the southeast and northwest regions.

Table 1 Descriptive statistics
Full size table
Fig. 2
figure 2

Bivariate correlation matrix graph. Deaths is the mortality rate between 2012 and 2020 for each communal administrative unit; FVC is the fractional vegetation cover; Hospital is the nuclear density value of medical institutions in each region; Age is the percentage of population over 65 years old in each region; Female is the percentage of female population in each region. Red cells indicate a positive correlation, Blue cells a negative correlation

Full size image
Fig. 3
figure 3

COPD mortality by quintile

Full size image

Deaths is the mortality rate between 2012 and 2020 for each communal administrative unit; FVC is the fractional vegetation cover; Hospital is the nuclear density value of medical institutions in each region; Age is the percentage of population over 65 years old in each region; Female is the percentage of female population in each region.

Spatial autocorrelation of COPD mortality

To test whether COPD mortality is spatially autocorrelated, the study used Moran's I spatial autocorrelation analysis. The Moran's I index for COPD mortality was 0.257 (p < 0.01), implying a potential dependence on the spatial distribution of COPD mortality in Chongqing. The LISA Cluster Map (Fig. 4) also shows the spatial distribution of the clusters and the corresponding confidence of each region; the high value clusters are located in the central and western part of the main urban area; while the low value clusters are distributed in the north-eastern, south-eastern and western areas of the main city.

Fig. 4
figure 4

A is the Significant level; B is Spatial autocorrelation

Full size image

Generalised additive models results

According to the variance inflation factor (VIF) test, the VIF of each variable is less than 5, and there is no multicollinearity between the variables [42]. Table 2 shows the results of the GAMs regression, where degrees of freedom greater than 1 means a non-linearly relationship between the explanatory variable and the COPD mortality; the F statistic reflects the degree of influence of the explanatory variables, the larger the F value the greater the influence; the P value is a parameter to judge the hypothesis test; and the R2 reflects the effect of the model fit. Model 1, Model 2 and Model 3 are three models with increasing adjustment levels respectively. Model 1 is a univariate regression model describing the relationship between FVC and COPD mortality; In Model 2, R2 was 0.183 after adding the medical resource covariates, and the rate of deviation explanation was 21.5%; In Model 3, the proportion of elderly population was also added and the coefficients of all variables are statistically significant, and the final model deviation interpretation rate is 32.4%.

Table 2 Parameters of GAMs results
Full size table

Geographically weighted regression models results

The relationship between COPD mortality and FVC was spatially unstable as shown in Fig. 5. From the results of regional parameter (Table 3), the FVC local regression coefficients range from − 0.0397 to 0.0478, 63.0% of the regions are positive and 37.0% are negative. The negative correlation areas are mainly located in the northeast and west of Chongqing. The local regression coefficients for Hospital range from − 0.0001 to 0.0004, a smaller range than that of the other two variables, with 19.4% of the area being positive and 81.6% being negative, the positive impact area is distributed in the southeast region. The Age local regression coefficients range from − 0.0161 to 0.2428, 91.7% of the areas are positive and 8.3% are negative. The high-value regions are mainly in the central and northeastern parts of Chongqing.

Fig. 5
figure 5

A is the R2 of the geographically weighted regression model; B is the coefficient of the variable FVC; C is the coefficient of the variable Age; D is the coefficient of the variable Hospital

Full size image
Table 3 Parameters of GWR results
Full size table

Geo-detector results

According to the Q-value (P value < 0.01) of Geo-detector factor detection, the three variables selected to explain the change in COPD mortality in Chongqing are Age, Hospital and FVC in descending order of explanatory power, with 0.257, 0.142 and 0.08 respectively. The difference in interpretive power when two factors work together on a single factor can be obtained by interactive probing. Among them, the interpretive power of FVC interaction with Age was 0.2885; the interpretive power of the interaction between FVC and Hospital was 0.2039; and the interpretive power of the interaction between Age and Hospital was 0.2846. All the results showed a two-factor enhancement, indicating that the interaction of factors had different degree of enhancement in explaining the mortality of COPD compared with single factor.

Discussion

Interpretation of results

This study found that the spatial distribution of COPD mortality in Chongqing has certain aggregation characteristics. The non-linear relationship between FVC and COPD mortality in Chongqing was described by GAMS model, and the model was made more stable by adding covariates. It was found that COPD mortality changed in segments as the FVC value increased. Based on the results of the GAMs, we hypothesized that there were regional differences in the association between COPD mortality and FVC, and the results of the GWR verified this hypothesis (Fig. 5). COPD mortality was positively correlated with FVC in 63% of the township areas in Chongqing and negatively correlated in 37.0% of the areas. It means that COPD mortality increases with increasing FVC in 63% of the township areas in Chongqing, higher vegetation cover brings higher risk of COPD mortality, COPD mortality decreases with increasing FVC in 37% of the township areas, and the negatively correlated areas are mainly located in the northeast of Chongqing. And according to the results of R2 distribution, the GWR model effect was better in this region. Overall, there was a spatially non-stationary relationship between COPD mortality and FVC in Chongqing.

Considering that regional differences in air pollution may affect the judgment of the relationship between FVC and COPD mortality, the study revealed the relationship between FVC and PM2.5, PM10 through Spearman correlation analysis, with significant negative correlation coefficients of − 0.88 and − 0.87, respectively. Implying that in Chongqing, the areas with high FVC have lower PM2.5 and PM10 concentrations. Therefore this study was able to rule out the possibility that high PM2.5 and PM10 concentrations would mask the beneficial effect of FVC on COPD mortality, making the findings more convincing. Also according to the results of the GAMs, GWR and Geo-detector models, the age factor makes the largest contribution to the distribution of COPD mortality, and the correlation coefficients for FVC are all relatively small. Such results are also in line with the reality.

Regional differences in the relationship between green space and COPD mortality is acceptable. A Belgian study including mortality data from five urban areas between 2001 and 2011 found a negative association between residential green space and respiratory disease mortality [43]; a Korean cross-sectional study and a Chinese cohort study found no significant association between green space and respiratory disease mortality [13, 14]; a national study of China by Fan [12] concluded that green space was a risk factor for increased COPD prevalence, and according to similar studies, findings on the relationship between green space and respiratory disease vary widely across regions.

In most townships in Chongqing where COPD mortality is positively associated with FVC, one possible explanation for the predominance is that areas with high FVC tend to be rural and mountainous, with limited access to medical resources, resulting in high mortality and high FVC occurring in the same areas. From Fig. 3, it can be seen that the mortality rate of COPD in the west of Chongqing is lower and the medical resource there is the best, while the townships surrounding the main urban area have correspondingly higher FVC values, but also lower availability of medical resources and higher COPD mortality rates. Another possible explanation is that some high-value FVC regions have lower urbanization with industrial structure dominated by primary and secondary industries, with high poverty and unemployment rates, low health insurance coverage, and a high proportion of the population currently or previously engaged in manual work. Studies have shown that people with low incomes are more susceptible to air pollution [13, 44], and people with low levels of education may also lack knowledge of air pollution disease prevention [45], resulting in higher COPD mortality. The regions where COPD mortality is negatively correlated with FVC in Chongqing are mainly located in the northeast. The possible reasons are that these regions have excellent natural environment and some districts and towns are all-area tourism demonstration areas, dominated by tertiary industries. The green space forms a complete natural barrier to optimize air quality and block air pollution particles, thus providing a good environment and reducing the risk of respiratory diseases [15].

There were significant regional differences in the spatial correlation between COPD mortality and green space in Chongqing. The reason for this may be related to the degree of influence of the FVC factor on the distribution of COPD mortality. According to the results of the geographic probe, the interpretive power of FVC on the spatial subdivision of COPD mortality is only 0.08, which is inferior to age and hospital availability; and according to the factor interaction detection, the interpretive power of multi-factor interaction is greater than that of FVC single factor, implying that the spatial subdivision of COPD mortality should be explained by multiple factors.

Limitations

Although the results of this study are statically significant, there are some limitations. The results of this study may be confounded by some difficult-to-measure personal factors due to the use of the township as the basic study unit, where many data were missing and only a few covariates could be taken into account. Firstly, smoking is considered to be a major cause of COPD, but the results of this study cannot eliminate the possible impact of smoking habits in the population due to the lack of data. Secondly, increased mortality of COPD may also be related to the presence of underlying disease in the cases themselves, and the presence of underlying disease in the study subjects themselves may have an impact on the findings. Meanwhile, due to data limitations, there were differences between the dependent variable COPD mortality (2012–2020) and the independent variable FVC (2020), so the study could not exclude uncertainties stemming from temporal inconsistencies [46].

Conclusion

The control of COPD mortality is of great importance to alleviate social stress and improve people's health and well-being. Due to the correlation between green space and respiratory diseases, we attempted to uncover the spatial relationship between green space and COPD mortality in Chongqing. It is expected to achieve the control of COPD mortality through planning policy. However, the results did not confirm this expectation. Green space may be a potential risk factor for increased COPD mortality in some regions of Chongqing. Therefore, this spatial differentiation needs to be taken into account in future green space planning and public health policy development in Chongqing.

Availability of data and materials

The data of PM2.5 and PM10 can get from [ChinaHighPM2.5: Big Data Seamless 1 km Ground-level PM2.5 Dataset for China] at [https://doi.org/10.5281/zenodo.6398971] and [ChinaHighPM10: Big Data Seamless 1 km Ground-level PM10 Dataset for China] at [https://doi.org/10.5281/zenodo.6449937]. The population data can get from [Chongqing Bureau of Statistics] at [http://tjj.cq.gov.cn/wap.html]. Green space data can be calculated from the Google Earth Engine platform. But the COPD mortality data will be made available from the corresponding author on reasonable request due to privacy and ethical restrictions.

References

  1. Rabe KF, Hurd S, Anzueto A, Barnes PJ, Buist SA, Calverley P, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. American Thoracic Society - AJRCCM; 2007;176:532–55.

  2. Chapman KR, Mannino DM, Soriano JB, Vermeire PA, Buist AS, Thun MJ, et al. Epidemiology and costs of chronic obstructive pulmonary disease. Eur Respir J Eur Respir Soc. 2006;27:188–207.

    Article  CAS  Google Scholar 

  3. Mannino DM, Buist AS. Global burden of COPD: risk factors, prevalence, and future trends. Lancet. 2007;370:765–73.

    Article  PubMed  Google Scholar 

  4. Fang X, Wang X, Bai C. COPD in China: the burden and importance of proper management. Chest. 2011;139:920–9.

    Article  PubMed  Google Scholar 

  5. Yin P, Wang H, Vos T, Li Y, Liu S, Liu Y, et al. A subnational analysis of mortality and prevalence of COPD in China from 1990 to 2013: findings from the global burden of disease study 2013. Chest. 2016;150:1269–80.

    Article  PubMed  Google Scholar 

  6. Rahnama MR, Shaddel L. Urban green space is spatially associated with cardiovascular disease occurrence in women of mashhad: a spatial analysis of influential factors on their presence in urban green spaces. J Urban Health. 2019;96:653–68.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Alcock I, White M, Cherrie M, Wheeler B, Taylor J, McInnes R, et al. Land cover and air pollution are associated with asthma hospitalisations: a cross-sectional study. Environ Int. 2017;109:29–41.

    Article  CAS  PubMed  Google Scholar 

  8. Astell-Burt T, Feng X, Kolt GS. Is neighborhood green space associated with a lower risk of type 2 diabetes? Evidence from 267,072 Australians. Diabetes Care. 2013;37:197–201.

    Article  PubMed  Google Scholar 

  9. Hartig T, Kahn PH. Living in cities, naturally. Sci Am Assoc Adv Sci. 2016;352:938–40.

    CAS  Google Scholar 

  10. Maas J, et al. Morbidity is related to a green living environment. J Epidemiol Commun Health. 2009;63:967–73.

    Article  CAS  Google Scholar 

  11. Sarkar C, Zhang B, Ni M, Kumari S, Bauermeister S, Gallacher J, et al. Environmental correlates of chronic obstructive pulmonary disease in 96 779 participants from the UK Biobank: a cross-sectional, observational study. Lancet Planet Health. 2019;3:e478–90.

    Article  PubMed  Google Scholar 

  12. Fan J, Guo Y, Cao Z, Cong S, Wang N, Lin H, et al. Neighborhood greenness associated with chronic obstructive pulmonary disease: a nationwide cross-sectional study in China. Environ Int. 2020;144:106042.

    Article  PubMed  Google Scholar 

  13. Wang D, Lau KK-L, Yu R, Wong SYS, Kwok TTY, Woo J. Neighbouring green space and mortality in community-dwelling elderly Hong Kong Chinese: a cohort study. BMJ Open. 2017;7:e015794.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Kim S, Kim H, Lee J-T. Interactions between ambient air particles and greenness on cause-specific mortality in seven korean metropolitan cities, 2008–2016. Int J Environ Res Public Health. 2019;16:1866.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Li L, Hart JE, Coull BA, Cao S, Spengler JD, Adamkiewicz G. Effect of residential greenness and nearby parks on respiratory and allergic diseases among middle school adolescents in a Chinese City. Int J Environ Res Public Health. 2019;16:991.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Holopainen JK, Gershenzon J. Multiple stress factors and the emission of plant VOCs. Trends Plant Sci. 2010;15:176–84.

    Article  CAS  PubMed  Google Scholar 

  17. Lovasi GS, Quinn JW, Neckerman KM, Perzanowski MS, Rundle A. Children living in areas with more street trees have lower prevalence of asthma. J Epidemiol Commun Health. 2008;62:647–9.

    Article  CAS  Google Scholar 

  18. National Bureau of Statistics [Internet]. 2020 [cited 2022 Oct 30]. Available from: http://www.stats.gov.cn/

  19. Laosuwan T, Gomasathit T, Rotjanakusol T. Application of remote sensing for temperature monitoring: the technique for land surface temperature analysis. J Ecol Eng. 2017;18:53–60.

    Article  Google Scholar 

  20. Zhang Y, Odeh IOA, Ramadan E. Assessment of land surface temperature in relation to landscape metrics and fractional vegetation cover in an urban/peri-urban region using Landsat data. Int J Remote Sensing. 2013;34:168–89.

    Article  Google Scholar 

  21. Chu D. Fractional vegetation cover. In: Chu D, editor. Remote sensing of land use and land cover in mountain region: a comprehensive study at the central Tibetan Plateau. Singapore: Springer; 2020. p. 195–207.

    Chapter  Google Scholar 

  22. Carlson TN, Ripley DA. On the relation between NDVI, fractional vegetation cover, and leaf area index. Remote Sens Environ. 1997;62:241–52.

    Article  Google Scholar 

  23. DeVries R, Kriebel D, Sama S. Outdoor air pollution and COPD-related emergency department visits, hospital admissions, and mortality: a meta-analysis. COPD J Chron Obstruct Pulmon Dis. 2017;14:113–21.

    Article  Google Scholar 

  24. Li M-H, Fan L-C, Mao B, Yang J-W, Choi AMK, Cao W-J, et al. Short-term exposure to ambient fine particulate matter increases hospitalizations and mortality in COPD: a systematic review and meta-analysis. Chest. 2016;149:447–58.

    Article  PubMed  Google Scholar 

  25. Liu Y, Sun J, Gou Y, Sun X, Li X, Yuan Z, et al. A multicity analysis of the short-term effects of air pollution on the chronic obstructive pulmonary disease hospital admissions in Shandong, China. Int J Environ Res Public Health. 2018;15:774.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Wei J, Li Z. ChinaHighPM10: big data seamless 1 km ground-level PM10 dataset for China [Internet]. Zenodo; 2020 [cited 2022 Oct 29]. Available from: https://zenodo.org/record/6449937

  27. Wei J, Li Z. ChinaHighPM2.5: big data seamless 1 km ground-level PM2.5 dataset for China [Internet]. Zenodo; 2019 [cited 2022 Oct 29]. Available from: https://doi.org/10.5281/zenodo.6398971

  28. Kelejian H, Prucha IR. On the asymptotic distribution of the Moran I test statistic with applications. J Econ. 2001;104:219–57.

    Article  Google Scholar 

  29. Getis A. Reflections on spatial autocorrelation. Reg Sci Urban Econ. 2007;37:491–6.

    Article  Google Scholar 

  30. Zar JH. Significance testing of the spearman rank correlation coefficient. J Am Stat Assoc. 1972;67:578–80.

    Article  Google Scholar 

  31. Wood SN. Generalized additive models: an introduction with R. New York: Chapman and Hall/CRC; 2006. https://doi.org/10.1201/9781420010404.

    Book  Google Scholar 

  32. Wood S. mgcv: mixed GAM computation vehicle with GCV/AIC/REML smoothness estimation. 2012 [cited 2022 Oct 29]; Available from: http://cran.r-project.org/web/packages/mgcv/index.html

  33. Brunsdon C, Fotheringham S, Charlton M. Geographically weighted regression. J R Stat Soc Ser D. 1998;47:431–43.

    Google Scholar 

  34. Marotta P. Assessing spatial relationships between race, inequality, crime, and Gonorrhea and Chlamydia in the United States. J Urban Health. 2017;94:683–98.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Hurvich CM, Tsai C-L. A corrected akaike information criterion for vector autoregressive model selection. J Time Ser Anal. 1993;14:271–9.

    Article  Google Scholar 

  36. Wang J, Li X, Christakos G, Liao Y, Zhang T, Gu X, et al. Geographical detectors-based health risk assessment and its application in the neural tube defects study of the Heshun Region, China. Int J Geogr Inf Sci. 2010;24:107–27.

    Article  CAS  Google Scholar 

  37. Lu B, Harris P, Charlton M, Brunsdon C. The GWmodel R package: further topics for exploring spatial heterogeneity using geographically weighted models. Geo-spat Inf Sci. 2014;17:85–101.

    Article  Google Scholar 

  38. Gollini I, Lu B, Charlton M, Brunsdon C, Harris P. GWmodel: an R package for exploring spatial heterogeneity using geographically weighted models. J Stat Soft. 2015;63:1–50.

    Article  Google Scholar 

  39. Song Y, Wang J, Ge Y, Xu C. An optimal parameters-based geographical detector model enhances geographic characteristics of explanatory variables for spatial heterogeneity analysis: cases with different types of spatial data. GISci Remote Sens. 2020;57:593–610.

    Article  Google Scholar 

  40. Hijmans RJ, Etten J van, Sumner M, Cheng J, Baston D, Bevan A, et al. Raster: geographic data analysis and modeling [Internet]. 2022 [cited 2022 Oct 29]. Available from: https://CRAN.R-project.org/package=raster

  41. Pebesma E. Simple features for R: standardized support for spatial vector data. R J. 2018;10:439.

    Article  Google Scholar 

  42. Johnston R, Jones K, Manley D. Confounding and collinearity in regression analysis: a cautionary tale and an alternative procedure, illustrated by studies of British voting behaviour. Qual Quant. 2018;52:1957–76.

    Article  PubMed  Google Scholar 

  43. Bauwelinck M, Casas L, Nawrot TS, Nemery B, Trabelsi S, Thomas I, et al. Residing in urban areas with higher green space is associated with lower mortality risk: A census-based cohort study with ten years of follow-up. Environ Int. 2021;148:106365.

    Article  CAS  PubMed  Google Scholar 

  44. Sarkodie SA, Strezov V, Jiang Y, Evans T. Proximate determinants of particulate matter (PM2.5) emission, mortality and life expectancy in Europe, Central Asia, Australia, Canada and the US. Sci Total Environ. 2019;683:489–97.

    Article  CAS  PubMed  Google Scholar 

  45. Deguen S, Petit C, Delbarre A, Kihal W, Padilla C, Benmarhnia T, et al. Neighbourhood characteristics and long-term air pollution levels modify the association between the short-term nitrogen dioxide concentrations and all-cause mortality in Paris. PLoS ONE. 2015;10:e0131463.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Becker DA, Browning MHEM, McAnirlin O, Yuan S, Helbich M. Is green space associated with opioid-related mortality? An ecological study at the U.S. county level. Urban For Urban Green. 2022;70:127529.

    Article  Google Scholar 

Download references

Acknowledgements

Thanks to the National Natural Science Foundation of China for supporting this article.

Funding

This study was supported by the National Natural Science Foundation of China [Grant number 51778364, 51978329].

Author information

Author notes

  1. Aiping Gou and Guanzheng Tan are Co-first authors

Authors and Affiliations

  1. College of Ecological Technology and Engineering, Shanghai Institute of Technology, Shanghai, 201418, China

    Aiping Gou & Guanzheng Tan

  2. Institute of Chronic and Non-communicable Disease Control and Prevention, Chongqing Center for Disease Control and Prevention, Chongqing, 400042, China

    Xianbin Ding, Yan Jiao & Qiang Tan

  3. College of Architecture, Nanjing Tech University, Nanjing, 211816, China

    Jiangbo Wang

  4. Department of Acupuncture, Chongqing Traditional Chinese Medicine Hospital, Chongqing, 400021, China

    Chunyan Gou

Authors
  1. Aiping Gou
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Guanzheng Tan
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xianbin Ding
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jiangbo Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Yan Jiao
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Chunyan Gou
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Qiang Tan
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization, AG; Methodology, GT; Software, GT; Data curation. YJ, CG and QT; Writing—original draft, AG and GT; Writing review and editing, XD and JW; Supervision, XD and JW. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Xianbin Ding or Jiangbo Wang.

Ethics declarations

Ethic approval and consent to participate

This study was ethically approved by National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, Ethical Review Committee (Project No:202219). This study did not involve human experiments or the use of human tissue samples. The need for informed consent was waived by National Center for Chronic and Noncommunicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, Ethical Review Committee (Project No:202219). The study was conducted in accordance to relevant guidelines and regulations.

Consent for publication

Not applicate.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gou, A., Tan, G., Ding, X. et al. Spatial association between green space and COPD mortality: a township-level ecological study in Chongqing, China. BMC Pulm Med 23, 89 (2023). https://doi.org/10.1186/s12890-023-02359-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12890-023-02359-x

Keywords

BMC Pulmonary Medicine

ISSN: 1471-2466

Contact us