Air and surface water are essential for human survival, but rapid urbanization and industrialization have led to significant environmental degradation and increased health hazards. Air pollution, particularly PM2.5, is strongly linked to lung cancer, while other pollutants like nitrogen dioxide (NO2) are associated with various cancers. Surface water pollution also poses significant health risks, but the combined impact of air and surface water pollution on cancer is not well-understood. Existing research has often examined these pollutants in isolation, lacking a comprehensive evaluation system to assess their holistic impact. This study hypothesized that air and surface water environments are spatially connected and that cancers cluster in areas with poor environmental conditions. To test these hypotheses, a spatial evaluation system harmonizing nationwide data on air and surface water quality, and cancer incidence in China was created. This system was designed to overcome the challenges posed by the heterogeneity and complexity of these datasets, allowing for a comprehensive evaluation of the environment-cancer relationship. The goal is to provide a quantifiable assessment of the relationships between multiple pollutants, multiple cancer types, and their combined effects on cancer occurrence. This large-scale study provides a unique opportunity to advance knowledge in environmental health and inform the development of effective policies for cooperative environmental governance and disease prevention.

Numerous studies have demonstrated the individual effects of air and surface water pollution on various health outcomes, including cancer. The association between PM2.5 and lung cancer is well-established. Studies have also linked nitrogen dioxide (NO2) to breast cancer and other cancers. Previous studies have examined specific types of water pollution and their local health effects. However, research on the combined effects of air and surface water pollution on cancer incidence is limited, especially at a national scale. While some studies have observed overlapping distribution patterns of various cancer types, indicating potential common environmental causes, a comprehensive evaluation system to understand the holistic relationship between the real-world environment and cancer has been lacking. This lack of comprehensive studies on multi-pollutant exposures and their impact on cancer incidence highlights the need for this research, which aims to address the fragmented nature of existing environmental carcinogenicity evidence. The study focuses on creating a method to quantify this combined effect and demonstrate a spatial connection between environmental factors and cancer incidence.

The study used a novel Spatial Evaluation System for Environment and Cancer (SESEC) to integrate nationwide data on air pollutants, surface water contaminants, and cancer incidence. Data were obtained from the China National Environmental Monitoring Centre (CNEMC), the Ministry of Ecology and Environment, and the China Cancer Registry Annual Report (2019). The prefecture-level area was defined as the basic unit. Analysis units were defined as units containing all three data components (air monitoring site, water monitoring section, and cancer registry institute). The study area included 219 analysis units covering 377 million people. Pollutant levels were averaged across multiple monitoring points within an analysis unit. The SESEC incorporated six WHO-recommended air pollutants and 13 surface water organic pollutant indicators. Threshold concentrations were set using the annual limit values from the National Ambient Air Quality Standard for air pollutants and the 75th percentiles of national levels for water pollutant indicators. Thirteen high-incidence cancer types were included. To quantify co-pollution, a three-step process was used: 1) A modified local Moran's index was used to identify spatial aggregation of pollutants. 2) The principle of combining items with similar features was applied, considering the high correlation among pollutants. 3) The space was divided into four graded levels of co-pollution based on the number of high-level pollutants in both air and water. A modified local Moran's I index was used to identify spatial patterns, considering six categories: high-high cluster (HH), low-low cluster (LL), high-low outlier (HL), low-high outlier (LH), high-not clustered (HN), and low-not clustered (LN). High-level pollutants (H) were identified based on their spatial distribution patterns. The spatial distribution and composition of the co-pollution grade were analyzed. A mixed modelling strategy combining SHAP analysis and negative binomial regression was employed to identify key pollutants affecting cancer incidence. The study also considered social factors such as per capita GDP, proportion of the population aged 65+, and urbanization rate.

The study found high cross-country heterogeneity in pollutant concentrations and cancer incidence and significant spatial autocorrelation among pollutants. The Spearman correlation coefficients were high within and across air and water pollutants, and between pollutants and cancers. The proposed co-pollution grading system effectively translated the complex environmental network into a one-dimensional quantifiable scale. 78 (35.6%) analysis units were classified as Grade IV (high co-pollution), mainly in the Beijing-Tianjin-Hebei region, Huaihe River basin, and Fen-Wei Plain. Grade I (low co-pollution) units were mainly in southern China. There was a strong spatial consistency between co-pollution grade and cancer incidence. Grade IV areas had the highest incidence rates for seven cancer types (oesophageal, gallbladder, pancreatic, kidney, stomach, breast, and lung), showing a dose-response relationship. SHAP analysis identified 19 pollutants with effects on at least one cancer type. Eight pollutants (four air pollutants: PM10, PM2.5, NO2, and O3; four water pollutants: COD_Mn, petroleum, DO, and cyanide) showed significant positive effects. Social factors were also identified as significant contributors. After adjusting for social factors, the effects of the pollutants remained stable, except for liver cancer, suggesting that social factors are more influential for liver cancer. NO2 showed a positive association with several cancer types, supporting its inclusion in the WHO ambient air quality database. PM2.5 showed an association with lung cancer and leukaemia. COD_Mn showed association with pancreatic, breast, and kidney cancers. An estimated 62,847 (7.4%) new cancer cases in 2016 were attributable to air and surface water pollution, with 69.7% occurring in Grade IV areas.

The findings demonstrate a strong spatial association between co-exposure to air and surface water pollution and cancer incidence in China. The dose-response relationship between co-pollution grade and cancer incidence provides compelling evidence for the combined impact of environmental factors on cancer risk. The identification of specific pollutants associated with particular cancer types is crucial for targeted prevention efforts. The study's large scale and comprehensive approach offer significant insights, particularly highlighting the interconnectedness of air and water pollution and their combined effects. The use of a spatially explicit approach addresses the challenges of individual exposure assessment and allows for an examination of the synergistic effects of multiple pollutants. However, the study primarily focused on spatial patterns, acknowledging the limitations of inferring temporal causality. Future research should examine the temporal dynamics and explore potential interactions between multiple pollutants and the underlying biological mechanisms contributing to cancer development. Further research examining individual exposure levels, genetic factors and potential time lags would enhance the understanding of the environment-cancer relationship.

This large-scale study provides strong evidence for the spatial consistency of co-exposure to air and surface water pollution and cancer risk in China. The findings demonstrate a dose-response relationship between the degree of co-pollution and cancer incidence, highlighting the need for a comprehensive approach to environmental governance and disease prevention. The identification of specific pollutants associated with various cancer types informs targeted interventions. Future research should focus on refining exposure assessment methods, investigating temporal relationships, exploring interaction effects, and elucidating the underlying biological mechanisms. The study's findings underscore the urgent need for collaborative efforts to mitigate environmental pollution and reduce the burden of cancer.

The study's reliance on ecological data limits the ability to draw definitive causal inferences at the individual level. The use of prefecture-level data may mask finer-scale variations in pollution levels and cancer incidence. While the study adjusted for some social factors, it may not capture all relevant confounders. The temporal correlations between pollution and cancer occurrence were not thoroughly investigated, limiting the conclusions regarding causality. The study focused primarily on spatial analysis, leaving the detailed investigation of temporal aspects for future research. Lastly, the study used existing nationwide data, potentially influenced by various reporting and monitoring methodologies that may introduce uncertainty into some of the estimations.