Spatial coupling and individual-level evidence: linking rice cadmium exposure to liver cancer in a high-risk area of China.

Introduction
Primary liver cancer develops through a multistep process of sustained hepatic injury, ultimately leading to carcinogenesis. Established risk factors include hepatitis B virus (HBV) infection (Hardell et al., 1984), hepatitis C virus (HCV) infection (Su et al., 2008), metabolic diseases (Sasaki et al., 2010; Chen et al., 2024), alcohol intake (Feng et al., 2017), aflatoxin intake (Deng et al., 2007; Ren et al., 2008), and exposure to environmental carcinogens (Hardell et al., 1984). Among environmental carcinogens, cadmium (Cd) is classified as a Group I human carcinogen (Vukićević, 2012). It primarily enters the human body via inhalation and dietary intake (Cui et al., 2004; Mamun et al., 2018; Wu et al., 2020), is absorbed into cells in ionic form (Genchi et al., 2012), transported in the blood via erythrocytes and albumin, and accumulates in the liver (Tinkov et al., 2018), subsequently inducing chronic liver injury (Wu et al., 2012). Studies have indicated an association between human Cd exposure and cancers in multiple organs, including the liver (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024). Diet, particularly through staple foods like rice, constitutes the major pathway for human Cd exposure (Lin et al., 2024). These findings provide a theoretical foundation for understanding the hepatocarcinogenic mechanisms of environmental Cd exposure. However, current knowledge regarding Cd’s hepatotoxic effects is largely derived from laboratory-based cytotoxicity studies and pathological investigations focusing on acute, high-dose exposure in humans. A significant research gap exists in the form of robust, large-scale epidemiological data validating the relationship between chronic, low-level environmental Cd exposure and primary liver cancer at a regional scale.
Guangxi Zhuang Autonomous Region is a high-incidence area for primary liver cancer, both nationally and globally, with a crude incidence rate of 41.65 per 100,000 (An et al., 2019; Zheng et al., 2018)—far exceeding the world average (15 per 100,000). Concurrently, the median Cd concentration in rice grains from Guangxi is 0.70 mg/kg, which is 12 times higher than the national average for China (0.06 mg/kg) (Shi et al., 2022). Our team’s preliminary research has revealed that the high-incidence clusters of primary liver cancer in Guangxi are predominantly distributed in central and southwestern Guangxi, showing a notable spatial coupling with areas of high rice Cd concentration. Nevertheless, the underlying mechanisms remain unclear, necessitating systematic investigation into the relationship between rice Cd and liver cancer in Guangxi.
To achieve these objectives, this study focused on liver cancer incidence, rice grain cadmium (Cd), and human blood Cd levels in Guangxi. The key tasks undertaken were as follows: (1) systematic collection of liver cancer incidence rates and rice grain Cd data from 44 counties (cities/districts) to analyze the spatial coupling and correlation between liver cancer distribution and external Cd exposure from rice; (2) collection of blood samples from 105 liver cancer patients and 105 healthy controls for Cd analysis; (3) collection of 216 rice grain samples and blood samples from 316 residents in five typical areas; (4) comprehensive analysis of the relationships between liver cancer incidence, external Cd exposure from rice, and internal Cd exposure reflected by blood Cd levels, aiming to clarify the primary exposure pathway and elucidate the impact of rice-derived Cd on liver cancer in Guangxi. This work will elucidate the relationship between dietary Cd exposure from rice and liver cancer in Guangxi, providing scientific evidence and practical implications for understanding carcinogenic mechanisms and developing prevention strategies in high-risk regions.

Materials and methods

Study subjects and data sources
The study encompassed the following components: data from 44 counties (cities/districts), five typical study areas, a liver cancer patient group, and a control group. Regional Data (44 counties/cities/districts): Liver cancer incidence data were obtained from the 2022 Guangxi Tumor Registry Annual Report (Tang et al., 2024). Fundamental data on rice grain Cd concentrations were sourced from the publication “Geochemical Survey and Research on Land Quality in Guangxi” (Zhang et al., 2021). Typical Study Areas: Five typical areas were selected to achieve broad spatial representation across Guangxi from north to south. These included Celing Township (Northern Guangxi), Gula Town (Central Guangxi), Houlu Township and Qingfeng Town (Central-Southern Guangxi), and Changle Town (Southern Guangxi). From these areas, 316 human blood samples and 175 rice grain samples were specifically collected for this study. Human Subjects and Sample Collection: Liver Cancer Patient Group: A total of 105 blood samples were collected from patients diagnosed with liver cancer. These participants were recruited from Binyang County (6 samples), Hengzhou City (43 samples), urban Nanning (26 samples), and Qinbei District (30 samples). Control Group: For comparison, 105 blood samples were collected from healthy controls primarily residing in Nanning City.

Sample collection

Human blood collection
Patients and Urban Nanning controls: In the patient–control comparisons, despite the established roles of gender and age as risk factors for liver cancer, these variables were not adjusted for as covariates due to study design limitations. The selection of patients and controls was primarily based on their residence in the same geographic area, ensuring comparable exposure levels to rice-derived cadmium. Patients and controls were long-term local residents with comparable gender ratios and ages predominantly clustered between 60 and 85 years, which may have partially mitigated heterogeneity. Nevertheless, residual confounding cannot be ruled out, and practical constraints prevented age-stratified sampling. Future studies with larger cohorts and long-term follow-up are needed to validate these findings.
Biological sample collection, including blood from liver cancer patients, control subjects, and residents of typical areas, was conducted by a team from Guangxi Medical University. The procedure was as follows:Pre-collection: Patient identity was verified. A disposable blood collection needle, vacuum blood collection tubes, a tourniquet, disinfectant swabs, and sterile gloves were utilized. The subject remained seated or recumbent, with the phlebotomy site (typically the median cubital vein) exposed.

Venipuncture: A tourniquet was applied to the upper arm. The skin was disinfected in a circular motion (diameter ≥ 5 cm) with 75% alcohol, centered on the intended puncture site. After the site air-dried, the needle was inserted at a 30° angle. Upon observing blood flashback, the needle was stabilized, and the vacuum tube was connected to collect the required blood volume.

Post-collection Handling: After collecting 20 mL of blood, the tubes were gently inverted 5–8 times (anticoagulant tubes required mixing), avoiding vigorous shaking.

Storage and Transport: Following collection and on-site documentation, blood samples were refrigerated and transported to the laboratory within 8 h for analysis, processed according to standard protocols. Samples were stored at − 2 °C ± 5 °C for no more than 5 days.

Rice sample collection
In five representative areas, rice samples were collected at the administrative village level following a principle of spatially uniform distribution. The collection and preservation protocols strictly adhered to the Geochemical Evaluation Specification of Land Quality (DZ/T 0295–2016). The specific procedure involved selecting three representative sub-locations within each paddy field. At each sub-location, an 80 cm × 80 cm quadrat was established. From each quadrat, 80 to 120 rice plants were selected, and their panicles were combined to create a composite sample, which was placed in a mesh bag (total sample weight: 3–4 kg). The entire rice plant, including the roots, was uprooted, and soil adhering to the roots was shaken onto a disposable plastic sheet. The root soil from all sub-locations was then thoroughly mixed to form a single composite root soil sample.

Sample processing and analysis

Rice grain cadmium analysis
The analyses were conducted by the Guangxi Institute of Geological and Mineral Testing in compliance with the National Food Safety Standard GB 2762–2017(National Health and Family Planning Commission of the People’s Republic of China, 2017) and the geological industry standard DZ/T 0295–2016 (National Technical Committee for Standardization of Land and Resources, 2016), respectively.
Sample Preparation: Rice grains were thoroughly washed with tap water followed by deionized water, and air-dried at room temperature. A representative subsample of 200–500 g was selected, hulled using a rice polishing machine to produce polished rice, and placed in a clean tray. The sample was dried in an oven at 55 °C, ground in a blender for 1–2 min, and returned to the oven for further drying at 55 °C. To prevent cross-contamination, the grinder (a traditional Chinese medicine grinder) was pre-cleaned by processing and discarding a small amount of sample. Subsequently, no more than 300 g of the sample was ground. The grinding duration was typically set to 2 min, repeated as necessary until a homogeneous powder was obtained (generally passing through a 20-mesh sieve, approx. 40–60 mesh). The final powder was sieved, transferred to a plastic bottle, sealed, and stored in a dry, ventilated environment pending analysis.
Digestion and ICP-MS Analysis: Precisely 0.1000–1.0000 g of the processed dry powder was weighed into a clean microwave digestion vessel. Then, 5 mL of concentrated nitric acid was added. Digestion was performed according to the microwave digestion system’s operating protocol. After completion and cooling to room temperature, the vessel was placed on a heating plate at approximately 100 °C to drive off acidic fumes until the evolution of NO₂ fumes ceased. After cooling, the digestate was transferred to a 25 mL boron-free test tube, brought to volume with deionized water, and shaken homogenously for analysis. Cadmium concentration was determined using Inductively Coupled Plasma Mass Spectrometry (ICP-MS).
Quality Control: Quality assurance was implemented according to the standard DZ/T 0295–2016 (National Technical Committee for Standardization of Land and Resources, 2016). For standard reference materials, each was analyzed 12 times to calculate the accuracy (Relative Error, RE%) and precision (Relative Standard Deviation, RSD%) of the results.

Human blood cadmium analysis
The analysis of Cd in human blood samples was performed by the Guangxi Workers’ Hospital, adhering to the group standard TGXAS 727–2024 (Guangxi Association for Standardization, 2024), entitled “Occupational Health Monitoring—Determination of 22 Elements in Blood—Inductively Coupled Plasma Mass Spectrometry.”

Data processing and graphics generation
Statistical analysis of the raw data was performed using SPSS Statistics 19.0 software. Box plots and scatter plots were generated using OriginPro 2025. For the linear equations presented in Figs. 4 and 5, the parameters were derived via ordinary least squares (OLS) regression, a method that identifies the line of best fit by minimizing the sum of squared residuals. The correlation analysis between rice grain cadmium (Cd) content and liver cancer incidence was conducted using Pearson correlation analysis. All data were tested with the One-sample Kolmogorov–Smirnov test and passed the normality test. Spatial distribution maps of liver cancer incidence, rice grain Cd content, and their overlay were created using ArcGIS software.

Results

Characteristics of liver cancer incidence in 44 counties of Guangxi
According to the 2022 Guangxi Tumor Registry Annual Report (Tang et al., 2024), the number of new liver cancer cases in the registered areas of Guangxi reached 13,170, with a crude incidence rate of 40.54 per 100,000. This rate is 1.46 times higher than the national average (27.8 per 100,000) (An et al., 2019; Zheng et al., 2018). Among the 44 counties (cities/districts), seven—Guanyang County, Qinnan District, Ziyuan County, Beiliu City, Yangshuo County, Lingchuan County, and Qingxiu District—exhibited crude incidence rates below the national average, ranging from 22.43 to 27.52 per 100,000. In contrast, the remaining 37 counties (84.09% of the total) had incidence rates exceeding the national level. Notably, five counties—Fusui County, Long’an County, Hengzhou City, Wuming District, and Daxin County—displayed particularly high rates of 74.71, 74.68, 65.48, 65.02, and 59.66 per 100,000, respectively, which are 2.15 to 2.69 times the national average (Table 1).
As shown in Fig. 1, the high-incidence areas for liver cancer within these 44 counties exhibited distinct regional clustering, primarily located between 22°20′ and 23°40′N latitude, encompassing southwestern, central, and southeastern Guangxi. The high-incidence cluster in southwestern Guangxi included Fusui County, Long’an County, Tiandong County, Daxin County, Tiandeng County, Jiangzhou District, and Longzhou County. The central Guangxi cluster comprised Xingbin District, Mashan County, Wuming District, Hengzhou City, Binyang County, Liangqing District, and Jiangnan District. The southeastern Guangxi cluster included Gangnan District, Gangbei District, and Qintang District.

Geochemical characteristics of Cd in rice grains in Guangxi
As shown in Table 2, the median and arithmetic mean cadmium concentrations in 3754 rice grain samples from 44 counties (cities/districts) were 0.07 mg/kg and 0.12 mg/kg, respectively. These values were comparable to the average Cd level in rice from Guangxi (0.11 mg/kg; Zhang et al., 2021), 1.2 times the national median (0.06 mg/kg; Shi et al., 2022), only 0.6 times the Chinese maximum limit for Cd in rice (0.2 mg/kg; National Health Commission & State Administration for Market Regulation, 2025), and significantly lower than the Codex Alimentarius standard (0.4 mg/kg; CXS 193–1995). This indicates that Cd concentrations in rice across the 44 counties (cities, districts) are generally low.
Spatially, rice Cd concentrations were generally lower in southwestern, western, and northeastern Guangxi, while relatively higher levels were observed in central and south-central regions (Table 3, Fig. 2). Specifically, the mean Cd concentrations in 16 counties—Mashan County, Gangnan District, Hepu County, Binyang County, Guiping City, Jiangnan District, Qintang District, Wuming District, Xingbin District, Yongning District, Gongcheng County, Xingning District, Lipu City, Long’an County, Pingnan County, and Pubei County—exceeded the Guangxi average, ranging from 0.13 to 0.24 mg/kg. Notably, five of these counties (Mashan County, Gangnan District, Hepu County, Binyang County, and Guiping City) exhibited mean Cd concentrations more than twice the Guangxi average. In contrast, rice Cd concentrations were generally low in eight counties in western Guangxi (Fusui County, Tiandeng County, Longzhou County, Daxin County, Tiandong County, Tianyang District, Jiangzhou District, and Youjiang District), with mean values ranging from 0.03 to 0.07 mg/kg—only 0.27 to 0.63 times the regional average.
As shown in Table 3, the Cd concentrations in rice grains varied significantly across the five typical areas. Celing Township exhibited the highest arithmetic mean Cd concentration (0.39 mg/kg), which was 3.5 times the average level for Guangxi. This was followed by Changle Town, with a mean concentration of 0.18 mg/kg, representing 1.64 times the regional average. The mean Cd concentration in Houlu Township was 0.10 mg/kg, comparable to the Guangxi average. In contrast, significantly lower concentrations were observed in Gula Town and Qingfeng Town, with mean values of 0.03 mg/kg and 0.05 mg/kg, respectively.

Characteristics of blood cadmium concentrations
As presented in Table 4, the blood Cd concentrations in the 105 Nanning Controls samples ranged from 0.05 to 6.32 μg/L, with an arithmetic mean of 1.62 μg/L (geometric mean: 0.72 μg/L). This level is comparable to those reported for healthy European populations (0.56–5.6 μg/L) (Dana et al., 2011) but is higher than those observed in a Korean population (arithmetic mean: 1.07 μg/L) (Seo et al., 2023), U.S. adults (arithmetic mean: 0.376 μg/L) (Orr et al., 2017), and the Nanning Controls from eight provinces in China (geometric mean: 0.49 μg/L) (Ding et al., 2014).
Notably, liver cancer patients exhibited markedly elevated blood Cd levels. The concentrations ranged from 0.37–9.10 μg/L in urban Nanning, 0.46–7.23 μg/L in Hengzhou City, 0.45–4.47 μg/L in Binyang County, and 0.60–6.23 μg/L in Qinbei District. These ranges correspond to 1.83–2.37 times, 5.22–6.75 times, and 3.11–4.44 times the levels reported for the Korean population, U.S. adults, and the general population from eight Chinese provinces, respectively.
Blood Cd concentrations in the five typical areas were elevated (Table 5). The concentrations ranged from 0.86–11.76 μg/L in Changle Town, 0.68–13.38 μg/L in Celing Township, 0.45–6.82 μg/L in Gula Town, 0.49–8.43 μg/L in Houlu Township, and 0.39–9.11 μg/L in Qingfeng Town. These ranges correspond to 1.96–3.44 times, 5.59–9.80 times, and 3.41–6.11 times the levels reported for a Korean population (arithmetic mean: 1.07 μg/L; Seo et al., 2023), U.S. adults (arithmetic mean: 0.376 μg/L; Orr et al., 2017), and the general population from eight Chinese provinces (geometric mean: 0.49 μg/L; Ding et al., 2014), respectively.

Relationship between liver cancer and rice grain Cd in Guangxi
Previous studies have suggested an association between cadmium (Cd) exposure and liver cancer (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024). Interestingly, our study observed a positive correlation between liver cancer incidence and rice grain Cd levels in the high-incidence cluster of central Guangxi. In contrast, a negative correlation was noted in the southwestern high-incidence cluster—comprising Fusui County, Long’an County, Hengzhou City, Daxin County, Tiandong County, Tiandeng County, Jiangzhou District, Liangqing District, Longzhou County, Tianyang District, Xixiangtang District, and Youjiang District—where rice grain Cd concentrations were relatively low (Fig. 3).The pathogenesis of liver cancer is multifactorial, involving hepatitis virus infection, drinking water contamination, lifestyle factors, and genetic polymorphisms (Hardell et al., 1984; Su et al., 2008; Sasaki et al., 2010; Kew, 2014; Chen et al., 2016; Chen et al., 2024; Feng et al., 2017; Deng et al., 2007; Ren et al., 2008). Specifically in Fusui County, high aflatoxin exposure has been identified as a potential key risk factor (Huang et al., 2014). Mei et al. (2020) found an association between microcystin levels in water bodies and liver diseases in high-risk populations within this region. Furthermore, Jia et al. (2019) reported a strong spatial agreement between the distribution of Microcystis aeruginosa toxins and liver cancer clusters in southwestern Guangxi, while Deng et al. (2020) observed significantly higher HBsAg prevalence in this area compared to others. Therefore, the apparent negative correlation between liver cancer incidence and rice Cd in southwestern Guangxi may result from the complex interplay of multiple factors, including aflatoxin, microcystin, HBV infection, and rice Cd.
To further elucidate the specific association between liver cancer and rice grain Cd in Guangxi, the 44 counties were categorized into the Southwestern Guangxi and the regions other than Southwestern Guangxi. As shown in Figs. 4 and 5, a significant positive correlation was identified between the crude liver cancer incidence rate and rice grain Cd concentration across Guangxi. Notably, the coefficient of determination (R2) reached 0.61 (p < 0.01) in the southwestern cluster after categorization, compared to R2 = 0.38 (p < 0.01) in the general region. This indicates that, after potentially accounting for the influence of confounding factors such as aflatoxin, microcystin, and HBV, the crude incidence rate of liver cancer in Guangxi is significantly correlated with Cd levels in rice grains. This finding aligns with previous reports linking Cd exposure to liver cancer (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024) and suggests that rice Cd exposure may be one of the key contributors to the distinct regional distribution pattern of liver cancer observed in Guangxi.

Relationship between liver cancer and internal Cd exposure in Guangxi
Elevated internal cadmium exposure can cause significant liver damage. According to Seo et al. (2023), for every 1 μg/L increase in blood cadmium, ALT and AST levels increase by 3.696 U/L and 2.677 U/L, respectively. A meta-analysis has demonstrated that serum cadmium levels are significantly higher in liver cancer patients compared to the general population (SMD = 2.00, 95% CI = 1.20–2.81) (Zhang, 2018). Toxicological studies further confirm that internal Cd exposure is closely associated with cancers in multiple organs, including the liver (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024).To elucidate the observed spatial coupling and correlation between the regional distribution of liver cancer incidence and external Cd exposure from rice in Guangxi, we compared the blood Cd levels between 105 liver cancer patients and 105 controls. The results are shown in Fig. 6.
The mean blood Cd concentrations in liver cancer patients were 2.15 μg/L (urban Nanning), 2.13 μg/L (Hengzhou City), 1.96 μg/L (Binyang County), and 2.54 μg/L (Qinbei District), which are 1.96 to 2.54 times higher than the mean level (1.00 μg/L) in the general population samples collected in this study (Fig. 6). This indicates that liver cancer patients exhibit significantly elevated blood Cd levels compared to the general population, consistent with previous findings (Zhang, 2018; Zhang et al., 2023). Notably, over 70% of the patients in these four regions had blood Cd concentrations exceeding the general population’s mean level (1.00 μg/L). The proportions were particularly high in Binyang County (83.33%) and Qinbei District (86.67%), followed by urban Nanning (73.10%) and Hengzhou City (69.77%). This strong association between liver cancer and elevated internal Cd exposure corroborates previous research linking Cd exposure to liver and other cancers (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024).

Relationship between blood Cd and rice Cd in Guangxi
The preceding analyses suggest that the regional distribution of liver cancer in Guangxi may be influenced by Cd in rice grains, and that a strong association exists between liver cancer and elevated blood Cd levels. If rice grain Cd is indeed the primary exposure pathway for blood Cd in the local population, it would indicate that Cd exposure from rice is a significant factor affecting the occurrence and distribution of liver cancer in Guangxi. Among dietary sources, staple foods represent the major pathway for human Cd exposure (Lin et al., 2024). As rice is the most important staple food in Guangxi, it is likely the primary source of internal Cd exposure. To further investigate the main exposure pathway for blood Cd in Guangxi, we conducted a correlation analysis between blood and rice grain Cd concentrations across the five typical areas (Fig. 7). The rice grain Cd levels ranked as follows: Celing Township > Changle Town > Houlu Township > Qingfeng Town > Gula Town. Correspondingly, distinct differences were observed in blood Cd levels, which ranked: Changle Town > Celing Township > Houlu Township > Qingfeng Town > Gula Town. Notably, Changle Town and Celing Township, which had the highest rice grain Cd concentrations, also exhibited the highest blood Cd levels, with arithmetic means of 3.69 μg/L and 3.49 μg/L, respectively—over 5.7 times higher than those reported for the general population from eight Chinese provinces. This was followed by Houlu Township, with a mean blood Cd concentration of 2.60 μg/L (4.5 times higher). Gula Town and Qingfeng Town showed comparable blood Cd levels, approximately 3.4 times higher than the reference population. Furthermore, using the upper limit for blood Cd in healthy European populations (5.6 μg/L) as a reference (Dana et al., 2011), the proportion of outliers with elevated blood Cd increased markedly with rising rice grain Cd levels. The percentages of outliers were: Celing Township (16%) > Changle Town (12.24%) > Qingfeng Town (4%) > Houlu Township (3.12%) > Gula Town (2.56%). This trend demonstrates that blood Cd concentrations in the five typical areas increase with rising rice grain Cd levels.
At the statistical unit of the five typical areas, a significant positive correlation was observed between blood and rice grain Cd, with a correlation coefficient as high as 0.8. Although this study is limited to five typical areas with concurrent blood and rice sample collection, and the sample size is constrained, these areas span the entire north–south transect of Guangxi and possess strong spatial representativeness. Therefore, the findings can, to a considerable extent, reflect the impact of rice grain Cd on blood Cd levels in Guangxi. Collectively, these analyses strongly indicate a high consistency between blood and rice grain Cd concentrations, identifying rice grain Cd as the primary exposure pathway for blood Cd in the Guangxi population.

Discussion

Discussion on the relationship between rice Cd exposure and liver cancer in Guangxi
After categorizing the 44 counties into the southwestern cluster region and regions other than Southwestern Guangxi, a certain degree of spatial coupling was observed between the regional distribution of liver cancer incidence and external Cd exposure from rice grains. Additionally, a correlation was found between the crude liver cancer incidence rate and rice grain Cd concentration, with coefficients of determination (R2) reaching 0.61 and 0.38, respectively. These results indicating that dietary cadmium exposure from rice may be associated with the zonal distribution of liver cancer in Guangxi. Previous studies have reported higher blood Cd levels in liver cancer patients compared to the general population (Zhang et al., 2018). Internal Cd exposure has been shown to cause significant liver damage (Seo et al., 2023) and is closely associated with liver cancer and other organ malignancies (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024). Consistent with these findings, our study demonstrated that liver cancer patients exhibited elevated blood Cd levels—more than twice those of the general population samples collected in this study—aligning with the results reported by Zhang et al. (2018). Moreover, over 70% of blood samples from liver cancer patients exceeded the mean blood Cd level (1.00 μg/L) of the general population in this study. These results suggest that the occurrence of liver cancer in Guangxi may be associated with internal cadmium exposure levels in humans, providing macroscopic evidence that supports previous research on the role of high cadmium exposure in promoting liver cancer and other organ-related cancers (Genchi et al., 2020; Zhang et al., 2023; Yu et al., 2024).Furthermore, our study revealed a relatively consistent relationship between blood Cd and rice grain Cd levels in the five typical areas, with blood Cd concentrations increasing as rice grain Cd levels rose. This suggests that rice grain Cd is the primary exposure pathway for blood Cd in the Guangxi population. Therefore, based on the above analysis, it is concluded that the strong association between rice grain Cd levels and liver cancer in Guangxi represents a factor warranting attention in the study of the zonal distribution of liver cancer in the region.

Conclusions limitations and outlook
This study investigated liver cancer incidence, rice grain cadmium (Cd), and human blood Cd levels in Guangxi. Based on spatial overlay and correlation analysis of crude liver cancer incidence rates and rice grain Cd concentrations across 44 counties (cities/districts), we found that external exposure to rice grain Cd was associated with the regional distribution of liver cancer in Guangxi. In comparison to the general population, liver cancer patients in Guangxi were distinguished by higher internal exposure levels of blood cadmium. Furthermore, analysis of the relationship between blood Cd and rice grain Cd in typical areas demonstrated a significant positive correlation, indicating that rice grain Cd is the primary exposure pathway for blood Cd in the Guangxi population. These findings collectively suggest that future investigations into the zonal distribution of liver cancer in Guangxi should account for the notable association with rice grain Cd as a relevant environmental factor. This research provides an explanation for the high incidence of liver cancer in Guangxi from the perspective of environmental Cd exposure. By examining the relationship between liver cancer incidence and rice grain Cd at a provincial scale, this study provides fundamental data supporting research on the hepatocarcinogenic effects of chronic Cd exposure via rice. Our findings macroscopically align with the conclusion that chronic environmental Cd exposure can lead to liver damage (Genchi et al., 2020; Zhang et al., 2018; Seo et al., 2023; Zhang et al., 2023; Yu et al., 2024), thereby enriching the understanding of the pathogenic mechanisms of liver cancer in Guangxi and nationwide. However, this study has certain limitations: (1) Although liver cancer incidence and rice grain Cd concentration data were collected from 44 counties/cities/districts, it was not feasible to obtain blood samples from patients across all these areas for internal Cd exposure analysis. This constraint primarily resulted from the substantial number of patients involved, logistical challenges in biological sample collection, and strong concerns regarding personal privacy that created significant barriers to additional sampling. (2) The research primarily focused on spatial overlay and correlation analysis between internal/external Cd exposure and liver cancer incidence, lacking subsequent studies using human hepatocyte models and animal Cd loading models. (3) Gender and age are known risk factors for liver cancer; however, due to constraints in the study design, they were not included as covariates in the statistical analysis. Although the gender and age structure of the population in the study area was relatively balanced, and the age of the study subjects was predominantly concentrated between 60 and 85 years—which may have reduced heterogeneity to some extent—residual confounding cannot be ruled out. Additionally, obtaining stratified samples by age group posed practical challenges. Further validation through expanded sample sizes and long-term follow-up is needed in future research. (4) This study has limitations in analyzing the association between blood cadmium levels and liver cancer incidence. Firstly, blood cadmium data from liver cancer patients in four counties could not be used for regression modeling due to their fixed incidence rate of 100%. Furthermore, although blood cadmium levels and corresponding incidence rates were obtained from populations in five typical areas, the limited number of sample regions (only five data points) resulted in insufficient statistical power for regression analysis, making it difficult to draw reliable conclusions. These constraints prevented an in-depth exploration of the dose–response relationship between the two variables using regression methods. (5) The insufficient spatial matching between patient blood samples and local rice samples. Due to privacy concerns, many patients did not provide detailed addresses, and those who did were geographically dispersed across the county. This made it statistically challenging to collect a sufficient number of representative rice samples from corresponding villages or townships. (6) Key risk factors for liver cancer (e.g. HBV infection, aflatoxin exposure, microcystins) are discussed but not quantitatively controlled for. (7) As an ecological study, the findings are subject to the ecological fallacy; observed area-level associations do not necessarily imply causality at the individual level. Addressing these limitations in future research could yield more substantial practical applications for understanding liver cancer pathogenesis and developing regional prevention strategies.

Supplementary Information
Below is the link to the electronic supplementary material.