Colorectal cancer ranks as the third most common cancer in males and the second most common in females worldwide, representing a major cause of cancer-related deaths globally [1], [2]. It is estimated that by 2030, 1 in 10 colon cancers and 1 in 4 rectal cancers will occur in individuals under 50, underscoring the urgency of this growing public health concern [3], [4], [5]. Older age, comorbidities, physical activities, unhealthy lifestyle, socioeconomic disadvantages, racial and ethnic disparities, and geographical differences all contribute to poorer outcomes [6], [7], [8], [9]. The most common comorbidities that affect survival rates in patients with colorectal cancer worldwide are cardiovascular disease, diabetes mellitus, chronic obstructive pulmonary disease (COPD), dementia and depression [10], [11], [12]. Chronic exposure to particulate matter (PM2.5), nitrogen dioxide (NO2), and sulfur dioxide (SO2) increases the risk of cardiovascular disease, diabetes mellitus, COPD, and mental health disorders, all of which are common and prognostically significant comorbidities in colorectal cancer patients [13], [14], [15]. Specifically, cancer patients undergoing chemotherapy or radiation therapy exhibit increased PM2.5-related risks for cardiopulmonary mortality [16].

Recent research utilizing data from the large UK Biobank cohort (n = 428,632) has demonstrated that prolonged exposure to ambient PM2.5, NO2, and NOx is linked to an increased risk of all-cause mortality among patients with colorectal cancer [17]. However, a previous study using the National Health Interview Survey database (n = 635,539, aged 18–84) did not find a significant association between PM2.5 exposure and all-cause mortality in colorectal cancer patients who were never-smokers [18]. Similarly, a large-scale study conducted in Hong Kong (n = 66,820) also did not identify a significant relationship between PM2.5 exposure and mortality among colorectal cancer patients [19]. In another extensive study encompassing the United States and Puerto Rico (n = 623,048), NO2 was found to be positively correlated with mortality from colorectal cancer, whereas PM2.5 showed no such association [20]. Moreover, the impact of air pollutants such as SO2, NO2, NOx, and Ozone (O3) on mortality risk among colorectal cancer patients remains uncertain.

Machine learning methods, such as those used to construct high-precision land use regression (LUR) models, substantially improve the estimation of long-term air pollution exposure levels by integrating diverse spatial, temporal, and environmental data sources, enabling fine-scale, individualized exposure assessment with greater accuracy and reduced bias compared to traditional approaches [21], [22]. Advanced algorithms (e.g., random forest, XGBoost, deep neural networks) capture complex nonlinear relationships and interactions among land use, meteorology, and pollution sources, resulting in higher explanatory power and robust validation performance for ambient air pollution exposure estimates [22].

Earlier studies were limited by their reliance on database diagnostic codes, which excluded key clinical information such as cancer stage and histological grade. Since these factors are major determinants of survival, their omission may introduce confounding and weaken or mask true associations. Therefore, our study aimed to explore the association between air pollution and all-cause mortality, identifying high-risk groups among colorectal cancer patients.